diff options
Diffstat (limited to 'arch/x86/kernel/cpu')
68 files changed, 35410 insertions, 0 deletions
diff --git a/arch/x86/kernel/cpu/.gitignore b/arch/x86/kernel/cpu/.gitignore new file mode 100644 index 000000000..0bca7ef74 --- /dev/null +++ b/arch/x86/kernel/cpu/.gitignore @@ -0,0 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only +capflags.c diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile new file mode 100644 index 000000000..93792b457 --- /dev/null +++ b/arch/x86/kernel/cpu/Makefile @@ -0,0 +1,67 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Makefile for x86-compatible CPU details, features and quirks +# + +# Don't trace early stages of a secondary CPU boot +ifdef CONFIG_FUNCTION_TRACER +CFLAGS_REMOVE_common.o = -pg +CFLAGS_REMOVE_perf_event.o = -pg +endif + +# If these files are instrumented, boot hangs during the first second. +KCOV_INSTRUMENT_common.o := n +KCOV_INSTRUMENT_perf_event.o := n + +# As above, instrumenting secondary CPU boot code causes boot hangs. +KCSAN_SANITIZE_common.o := n + +# Make sure load_percpu_segment has no stackprotector +CFLAGS_common.o := -fno-stack-protector + +obj-y := cacheinfo.o scattered.o topology.o +obj-y += common.o +obj-y += rdrand.o +obj-y += match.o +obj-y += bugs.o +obj-y += aperfmperf.o +obj-y += cpuid-deps.o +obj-y += umwait.o + +obj-$(CONFIG_PROC_FS) += proc.o +obj-$(CONFIG_X86_FEATURE_NAMES) += capflags.o powerflags.o + +obj-$(CONFIG_IA32_FEAT_CTL) += feat_ctl.o +ifdef CONFIG_CPU_SUP_INTEL +obj-y += intel.o intel_pconfig.o tsx.o +obj-$(CONFIG_PM) += intel_epb.o +endif +obj-$(CONFIG_CPU_SUP_AMD) += amd.o +obj-$(CONFIG_CPU_SUP_HYGON) += hygon.o +obj-$(CONFIG_CPU_SUP_CYRIX_32) += cyrix.o +obj-$(CONFIG_CPU_SUP_CENTAUR) += centaur.o +obj-$(CONFIG_CPU_SUP_TRANSMETA_32) += transmeta.o +obj-$(CONFIG_CPU_SUP_UMC_32) += umc.o +obj-$(CONFIG_CPU_SUP_ZHAOXIN) += zhaoxin.o + +obj-$(CONFIG_X86_MCE) += mce/ +obj-$(CONFIG_MTRR) += mtrr/ +obj-$(CONFIG_MICROCODE) += microcode/ +obj-$(CONFIG_X86_CPU_RESCTRL) += resctrl/ + +obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o + +obj-$(CONFIG_HYPERVISOR_GUEST) += vmware.o hypervisor.o mshyperv.o +obj-$(CONFIG_ACRN_GUEST) += acrn.o + +ifdef CONFIG_X86_FEATURE_NAMES +quiet_cmd_mkcapflags = MKCAP $@ + cmd_mkcapflags = $(CONFIG_SHELL) $(srctree)/$(src)/mkcapflags.sh $@ $^ + +cpufeature = $(src)/../../include/asm/cpufeatures.h +vmxfeature = $(src)/../../include/asm/vmxfeatures.h + +$(obj)/capflags.c: $(cpufeature) $(vmxfeature) $(src)/mkcapflags.sh FORCE + $(call if_changed,mkcapflags) +endif +targets += capflags.c diff --git a/arch/x86/kernel/cpu/acrn.c b/arch/x86/kernel/cpu/acrn.c new file mode 100644 index 000000000..0b2c03943 --- /dev/null +++ b/arch/x86/kernel/cpu/acrn.c @@ -0,0 +1,64 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * ACRN detection support + * + * Copyright (C) 2019 Intel Corporation. All rights reserved. + * + * Jason Chen CJ <jason.cj.chen@intel.com> + * Zhao Yakui <yakui.zhao@intel.com> + * + */ + +#include <linux/interrupt.h> +#include <asm/apic.h> +#include <asm/cpufeatures.h> +#include <asm/desc.h> +#include <asm/hypervisor.h> +#include <asm/idtentry.h> +#include <asm/irq_regs.h> + +static u32 __init acrn_detect(void) +{ + return hypervisor_cpuid_base("ACRNACRNACRN", 0); +} + +static void __init acrn_init_platform(void) +{ + /* Setup the IDT for ACRN hypervisor callback */ + alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, asm_sysvec_acrn_hv_callback); +} + +static bool acrn_x2apic_available(void) +{ + return boot_cpu_has(X86_FEATURE_X2APIC); +} + +static void (*acrn_intr_handler)(void); + +DEFINE_IDTENTRY_SYSVEC(sysvec_acrn_hv_callback) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + /* + * The hypervisor requires that the APIC EOI should be acked. + * If the APIC EOI is not acked, the APIC ISR bit for the + * HYPERVISOR_CALLBACK_VECTOR will not be cleared and then it + * will block the interrupt whose vector is lower than + * HYPERVISOR_CALLBACK_VECTOR. + */ + ack_APIC_irq(); + inc_irq_stat(irq_hv_callback_count); + + if (acrn_intr_handler) + acrn_intr_handler(); + + set_irq_regs(old_regs); +} + +const __initconst struct hypervisor_x86 x86_hyper_acrn = { + .name = "ACRN", + .detect = acrn_detect, + .type = X86_HYPER_ACRN, + .init.init_platform = acrn_init_platform, + .init.x2apic_available = acrn_x2apic_available, +}; diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c new file mode 100644 index 000000000..f29c6bed9 --- /dev/null +++ b/arch/x86/kernel/cpu/amd.c @@ -0,0 +1,1346 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include <linux/export.h> +#include <linux/bitops.h> +#include <linux/elf.h> +#include <linux/mm.h> + +#include <linux/io.h> +#include <linux/sched.h> +#include <linux/sched/clock.h> +#include <linux/random.h> +#include <linux/topology.h> +#include <asm/processor.h> +#include <asm/apic.h> +#include <asm/cacheinfo.h> +#include <asm/cpu.h> +#include <asm/spec-ctrl.h> +#include <asm/smp.h> +#include <asm/numa.h> +#include <asm/pci-direct.h> +#include <asm/delay.h> +#include <asm/debugreg.h> +#include <asm/resctrl.h> + +#ifdef CONFIG_X86_64 +# include <asm/mmconfig.h> +# include <asm/set_memory.h> +#endif + +#include "cpu.h" + +/* + * nodes_per_socket: Stores the number of nodes per socket. + * Refer to Fam15h Models 00-0fh BKDG - CPUID Fn8000_001E_ECX + * Node Identifiers[10:8] + */ +static u32 nodes_per_socket = 1; + +/* + * AMD errata checking + * + * Errata are defined as arrays of ints using the AMD_LEGACY_ERRATUM() or + * AMD_OSVW_ERRATUM() macros. The latter is intended for newer errata that + * have an OSVW id assigned, which it takes as first argument. Both take a + * variable number of family-specific model-stepping ranges created by + * AMD_MODEL_RANGE(). + * + * Example: + * + * const int amd_erratum_319[] = + * AMD_LEGACY_ERRATUM(AMD_MODEL_RANGE(0x10, 0x2, 0x1, 0x4, 0x2), + * AMD_MODEL_RANGE(0x10, 0x8, 0x0, 0x8, 0x0), + * AMD_MODEL_RANGE(0x10, 0x9, 0x0, 0x9, 0x0)); + */ + +#define AMD_LEGACY_ERRATUM(...) { -1, __VA_ARGS__, 0 } +#define AMD_OSVW_ERRATUM(osvw_id, ...) { osvw_id, __VA_ARGS__, 0 } +#define AMD_MODEL_RANGE(f, m_start, s_start, m_end, s_end) \ + ((f << 24) | (m_start << 16) | (s_start << 12) | (m_end << 4) | (s_end)) +#define AMD_MODEL_RANGE_FAMILY(range) (((range) >> 24) & 0xff) +#define AMD_MODEL_RANGE_START(range) (((range) >> 12) & 0xfff) +#define AMD_MODEL_RANGE_END(range) ((range) & 0xfff) + +static const int amd_erratum_400[] = + AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0xf, 0x41, 0x2, 0xff, 0xf), + AMD_MODEL_RANGE(0x10, 0x2, 0x1, 0xff, 0xf)); + +static const int amd_erratum_383[] = + AMD_OSVW_ERRATUM(3, AMD_MODEL_RANGE(0x10, 0, 0, 0xff, 0xf)); + +/* #1054: Instructions Retired Performance Counter May Be Inaccurate */ +static const int amd_erratum_1054[] = + AMD_LEGACY_ERRATUM(AMD_MODEL_RANGE(0x17, 0, 0, 0x2f, 0xf)); + +static const int amd_zenbleed[] = + AMD_LEGACY_ERRATUM(AMD_MODEL_RANGE(0x17, 0x30, 0x0, 0x4f, 0xf), + AMD_MODEL_RANGE(0x17, 0x60, 0x0, 0x7f, 0xf), + AMD_MODEL_RANGE(0x17, 0x90, 0x0, 0x91, 0xf), + AMD_MODEL_RANGE(0x17, 0xa0, 0x0, 0xaf, 0xf)); + +static const int amd_div0[] = + AMD_LEGACY_ERRATUM(AMD_MODEL_RANGE(0x17, 0x00, 0x0, 0x2f, 0xf), + AMD_MODEL_RANGE(0x17, 0x50, 0x0, 0x5f, 0xf)); + +static const int amd_erratum_1485[] = + AMD_LEGACY_ERRATUM(AMD_MODEL_RANGE(0x19, 0x10, 0x0, 0x1f, 0xf), + AMD_MODEL_RANGE(0x19, 0x60, 0x0, 0xaf, 0xf)); + +static bool cpu_has_amd_erratum(struct cpuinfo_x86 *cpu, const int *erratum) +{ + int osvw_id = *erratum++; + u32 range; + u32 ms; + + if (osvw_id >= 0 && osvw_id < 65536 && + cpu_has(cpu, X86_FEATURE_OSVW)) { + u64 osvw_len; + + rdmsrl(MSR_AMD64_OSVW_ID_LENGTH, osvw_len); + if (osvw_id < osvw_len) { + u64 osvw_bits; + + rdmsrl(MSR_AMD64_OSVW_STATUS + (osvw_id >> 6), + osvw_bits); + return osvw_bits & (1ULL << (osvw_id & 0x3f)); + } + } + + /* OSVW unavailable or ID unknown, match family-model-stepping range */ + ms = (cpu->x86_model << 4) | cpu->x86_stepping; + while ((range = *erratum++)) + if ((cpu->x86 == AMD_MODEL_RANGE_FAMILY(range)) && + (ms >= AMD_MODEL_RANGE_START(range)) && + (ms <= AMD_MODEL_RANGE_END(range))) + return true; + + return false; +} + +static inline int rdmsrl_amd_safe(unsigned msr, unsigned long long *p) +{ + u32 gprs[8] = { 0 }; + int err; + + WARN_ONCE((boot_cpu_data.x86 != 0xf), + "%s should only be used on K8!\n", __func__); + + gprs[1] = msr; + gprs[7] = 0x9c5a203a; + + err = rdmsr_safe_regs(gprs); + + *p = gprs[0] | ((u64)gprs[2] << 32); + + return err; +} + +static inline int wrmsrl_amd_safe(unsigned msr, unsigned long long val) +{ + u32 gprs[8] = { 0 }; + + WARN_ONCE((boot_cpu_data.x86 != 0xf), + "%s should only be used on K8!\n", __func__); + + gprs[0] = (u32)val; + gprs[1] = msr; + gprs[2] = val >> 32; + gprs[7] = 0x9c5a203a; + + return wrmsr_safe_regs(gprs); +} + +/* + * B step AMD K6 before B 9730xxxx have hardware bugs that can cause + * misexecution of code under Linux. Owners of such processors should + * contact AMD for precise details and a CPU swap. + * + * See http://www.multimania.com/poulot/k6bug.html + * and section 2.6.2 of "AMD-K6 Processor Revision Guide - Model 6" + * (Publication # 21266 Issue Date: August 1998) + * + * The following test is erm.. interesting. AMD neglected to up + * the chip setting when fixing the bug but they also tweaked some + * performance at the same time.. + */ + +#ifdef CONFIG_X86_32 +extern __visible void vide(void); +__asm__(".text\n" + ".globl vide\n" + ".type vide, @function\n" + ".align 4\n" + "vide: ret\n"); +#endif + +static void init_amd_k5(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 +/* + * General Systems BIOSen alias the cpu frequency registers + * of the Elan at 0x000df000. Unfortunately, one of the Linux + * drivers subsequently pokes it, and changes the CPU speed. + * Workaround : Remove the unneeded alias. + */ +#define CBAR (0xfffc) /* Configuration Base Address (32-bit) */ +#define CBAR_ENB (0x80000000) +#define CBAR_KEY (0X000000CB) + if (c->x86_model == 9 || c->x86_model == 10) { + if (inl(CBAR) & CBAR_ENB) + outl(0 | CBAR_KEY, CBAR); + } +#endif +} + +static void init_amd_k6(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + u32 l, h; + int mbytes = get_num_physpages() >> (20-PAGE_SHIFT); + + if (c->x86_model < 6) { + /* Based on AMD doc 20734R - June 2000 */ + if (c->x86_model == 0) { + clear_cpu_cap(c, X86_FEATURE_APIC); + set_cpu_cap(c, X86_FEATURE_PGE); + } + return; + } + + if (c->x86_model == 6 && c->x86_stepping == 1) { + const int K6_BUG_LOOP = 1000000; + int n; + void (*f_vide)(void); + u64 d, d2; + + pr_info("AMD K6 stepping B detected - "); + + /* + * It looks like AMD fixed the 2.6.2 bug and improved indirect + * calls at the same time. + */ + + n = K6_BUG_LOOP; + f_vide = vide; + OPTIMIZER_HIDE_VAR(f_vide); + d = rdtsc(); + while (n--) + f_vide(); + d2 = rdtsc(); + d = d2-d; + + if (d > 20*K6_BUG_LOOP) + pr_cont("system stability may be impaired when more than 32 MB are used.\n"); + else + pr_cont("probably OK (after B9730xxxx).\n"); + } + + /* K6 with old style WHCR */ + if (c->x86_model < 8 || + (c->x86_model == 8 && c->x86_stepping < 8)) { + /* We can only write allocate on the low 508Mb */ + if (mbytes > 508) + mbytes = 508; + + rdmsr(MSR_K6_WHCR, l, h); + if ((l&0x0000FFFF) == 0) { + unsigned long flags; + l = (1<<0)|((mbytes/4)<<1); + local_irq_save(flags); + wbinvd(); + wrmsr(MSR_K6_WHCR, l, h); + local_irq_restore(flags); + pr_info("Enabling old style K6 write allocation for %d Mb\n", + mbytes); + } + return; + } + + if ((c->x86_model == 8 && c->x86_stepping > 7) || + c->x86_model == 9 || c->x86_model == 13) { + /* The more serious chips .. */ + + if (mbytes > 4092) + mbytes = 4092; + + rdmsr(MSR_K6_WHCR, l, h); + if ((l&0xFFFF0000) == 0) { + unsigned long flags; + l = ((mbytes>>2)<<22)|(1<<16); + local_irq_save(flags); + wbinvd(); + wrmsr(MSR_K6_WHCR, l, h); + local_irq_restore(flags); + pr_info("Enabling new style K6 write allocation for %d Mb\n", + mbytes); + } + + return; + } + + if (c->x86_model == 10) { + /* AMD Geode LX is model 10 */ + /* placeholder for any needed mods */ + return; + } +#endif +} + +static void init_amd_k7(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + u32 l, h; + + /* + * Bit 15 of Athlon specific MSR 15, needs to be 0 + * to enable SSE on Palomino/Morgan/Barton CPU's. + * If the BIOS didn't enable it already, enable it here. + */ + if (c->x86_model >= 6 && c->x86_model <= 10) { + if (!cpu_has(c, X86_FEATURE_XMM)) { + pr_info("Enabling disabled K7/SSE Support.\n"); + msr_clear_bit(MSR_K7_HWCR, 15); + set_cpu_cap(c, X86_FEATURE_XMM); + } + } + + /* + * It's been determined by AMD that Athlons since model 8 stepping 1 + * are more robust with CLK_CTL set to 200xxxxx instead of 600xxxxx + * As per AMD technical note 27212 0.2 + */ + if ((c->x86_model == 8 && c->x86_stepping >= 1) || (c->x86_model > 8)) { + rdmsr(MSR_K7_CLK_CTL, l, h); + if ((l & 0xfff00000) != 0x20000000) { + pr_info("CPU: CLK_CTL MSR was %x. Reprogramming to %x\n", + l, ((l & 0x000fffff)|0x20000000)); + wrmsr(MSR_K7_CLK_CTL, (l & 0x000fffff)|0x20000000, h); + } + } + + /* calling is from identify_secondary_cpu() ? */ + if (!c->cpu_index) + return; + + /* + * Certain Athlons might work (for various values of 'work') in SMP + * but they are not certified as MP capable. + */ + /* Athlon 660/661 is valid. */ + if ((c->x86_model == 6) && ((c->x86_stepping == 0) || + (c->x86_stepping == 1))) + return; + + /* Duron 670 is valid */ + if ((c->x86_model == 7) && (c->x86_stepping == 0)) + return; + + /* + * Athlon 662, Duron 671, and Athlon >model 7 have capability + * bit. It's worth noting that the A5 stepping (662) of some + * Athlon XP's have the MP bit set. + * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for + * more. + */ + if (((c->x86_model == 6) && (c->x86_stepping >= 2)) || + ((c->x86_model == 7) && (c->x86_stepping >= 1)) || + (c->x86_model > 7)) + if (cpu_has(c, X86_FEATURE_MP)) + return; + + /* If we get here, not a certified SMP capable AMD system. */ + + /* + * Don't taint if we are running SMP kernel on a single non-MP + * approved Athlon + */ + WARN_ONCE(1, "WARNING: This combination of AMD" + " processors is not suitable for SMP.\n"); + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); +#endif +} + +#ifdef CONFIG_NUMA +/* + * To workaround broken NUMA config. Read the comment in + * srat_detect_node(). + */ +static int nearby_node(int apicid) +{ + int i, node; + + for (i = apicid - 1; i >= 0; i--) { + node = __apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { + node = __apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + return first_node(node_online_map); /* Shouldn't happen */ +} +#endif + +/* + * Fix up cpu_core_id for pre-F17h systems to be in the + * [0 .. cores_per_node - 1] range. Not really needed but + * kept so as not to break existing setups. + */ +static void legacy_fixup_core_id(struct cpuinfo_x86 *c) +{ + u32 cus_per_node; + + if (c->x86 >= 0x17) + return; + + cus_per_node = c->x86_max_cores / nodes_per_socket; + c->cpu_core_id %= cus_per_node; +} + +/* + * Fixup core topology information for + * (1) AMD multi-node processors + * Assumption: Number of cores in each internal node is the same. + * (2) AMD processors supporting compute units + */ +static void amd_get_topology(struct cpuinfo_x86 *c) +{ + int cpu = smp_processor_id(); + + /* get information required for multi-node processors */ + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + int err; + u32 eax, ebx, ecx, edx; + + cpuid(0x8000001e, &eax, &ebx, &ecx, &edx); + + c->cpu_die_id = ecx & 0xff; + + if (c->x86 == 0x15) + c->cu_id = ebx & 0xff; + + if (c->x86 >= 0x17) { + c->cpu_core_id = ebx & 0xff; + + if (smp_num_siblings > 1) + c->x86_max_cores /= smp_num_siblings; + } + + /* + * In case leaf B is available, use it to derive + * topology information. + */ + err = detect_extended_topology(c); + if (!err) + c->x86_coreid_bits = get_count_order(c->x86_max_cores); + + cacheinfo_amd_init_llc_id(c, cpu); + + } else if (cpu_has(c, X86_FEATURE_NODEID_MSR)) { + u64 value; + + rdmsrl(MSR_FAM10H_NODE_ID, value); + c->cpu_die_id = value & 7; + + per_cpu(cpu_llc_id, cpu) = c->cpu_die_id; + } else + return; + + if (nodes_per_socket > 1) { + set_cpu_cap(c, X86_FEATURE_AMD_DCM); + legacy_fixup_core_id(c); + } +} + +/* + * On a AMD dual core setup the lower bits of the APIC id distinguish the cores. + * Assumes number of cores is a power of two. + */ +static void amd_detect_cmp(struct cpuinfo_x86 *c) +{ + unsigned bits; + int cpu = smp_processor_id(); + + bits = c->x86_coreid_bits; + /* Low order bits define the core id (index of core in socket) */ + c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); + /* Convert the initial APIC ID into the socket ID */ + c->phys_proc_id = c->initial_apicid >> bits; + /* use socket ID also for last level cache */ + per_cpu(cpu_llc_id, cpu) = c->cpu_die_id = c->phys_proc_id; +} + +static void amd_detect_ppin(struct cpuinfo_x86 *c) +{ + unsigned long long val; + + if (!cpu_has(c, X86_FEATURE_AMD_PPIN)) + return; + + /* When PPIN is defined in CPUID, still need to check PPIN_CTL MSR */ + if (rdmsrl_safe(MSR_AMD_PPIN_CTL, &val)) + goto clear_ppin; + + /* PPIN is locked in disabled mode, clear feature bit */ + if ((val & 3UL) == 1UL) + goto clear_ppin; + + /* If PPIN is disabled, try to enable it */ + if (!(val & 2UL)) { + wrmsrl_safe(MSR_AMD_PPIN_CTL, val | 2UL); + rdmsrl_safe(MSR_AMD_PPIN_CTL, &val); + } + + /* If PPIN_EN bit is 1, return from here; otherwise fall through */ + if (val & 2UL) + return; + +clear_ppin: + clear_cpu_cap(c, X86_FEATURE_AMD_PPIN); +} + +u16 amd_get_nb_id(int cpu) +{ + return per_cpu(cpu_llc_id, cpu); +} +EXPORT_SYMBOL_GPL(amd_get_nb_id); + +u32 amd_get_nodes_per_socket(void) +{ + return nodes_per_socket; +} +EXPORT_SYMBOL_GPL(amd_get_nodes_per_socket); + +static void srat_detect_node(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_NUMA + int cpu = smp_processor_id(); + int node; + unsigned apicid = c->apicid; + + node = numa_cpu_node(cpu); + if (node == NUMA_NO_NODE) + node = per_cpu(cpu_llc_id, cpu); + + /* + * On multi-fabric platform (e.g. Numascale NumaChip) a + * platform-specific handler needs to be called to fixup some + * IDs of the CPU. + */ + if (x86_cpuinit.fixup_cpu_id) + x86_cpuinit.fixup_cpu_id(c, node); + + if (!node_online(node)) { + /* + * Two possibilities here: + * + * - The CPU is missing memory and no node was created. In + * that case try picking one from a nearby CPU. + * + * - The APIC IDs differ from the HyperTransport node IDs + * which the K8 northbridge parsing fills in. Assume + * they are all increased by a constant offset, but in + * the same order as the HT nodeids. If that doesn't + * result in a usable node fall back to the path for the + * previous case. + * + * This workaround operates directly on the mapping between + * APIC ID and NUMA node, assuming certain relationship + * between APIC ID, HT node ID and NUMA topology. As going + * through CPU mapping may alter the outcome, directly + * access __apicid_to_node[]. + */ + int ht_nodeid = c->initial_apicid; + + if (__apicid_to_node[ht_nodeid] != NUMA_NO_NODE) + node = __apicid_to_node[ht_nodeid]; + /* Pick a nearby node */ + if (!node_online(node)) + node = nearby_node(apicid); + } + numa_set_node(cpu, node); +#endif +} + +static void early_init_amd_mc(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned bits, ecx; + + /* Multi core CPU? */ + if (c->extended_cpuid_level < 0x80000008) + return; + + ecx = cpuid_ecx(0x80000008); + + c->x86_max_cores = (ecx & 0xff) + 1; + + /* CPU telling us the core id bits shift? */ + bits = (ecx >> 12) & 0xF; + + /* Otherwise recompute */ + if (bits == 0) { + while ((1 << bits) < c->x86_max_cores) + bits++; + } + + c->x86_coreid_bits = bits; +#endif +} + +static void bsp_init_amd(struct cpuinfo_x86 *c) +{ + +#ifdef CONFIG_X86_64 + if (c->x86 >= 0xf) { + unsigned long long tseg; + + /* + * Split up direct mapping around the TSEG SMM area. + * Don't do it for gbpages because there seems very little + * benefit in doing so. + */ + if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg)) { + unsigned long pfn = tseg >> PAGE_SHIFT; + + pr_debug("tseg: %010llx\n", tseg); + if (pfn_range_is_mapped(pfn, pfn + 1)) + set_memory_4k((unsigned long)__va(tseg), 1); + } + } +#endif + + if (cpu_has(c, X86_FEATURE_CONSTANT_TSC)) { + + if (c->x86 > 0x10 || + (c->x86 == 0x10 && c->x86_model >= 0x2)) { + u64 val; + + rdmsrl(MSR_K7_HWCR, val); + if (!(val & BIT(24))) + pr_warn(FW_BUG "TSC doesn't count with P0 frequency!\n"); + } + } + + if (c->x86 == 0x15) { + unsigned long upperbit; + u32 cpuid, assoc; + + cpuid = cpuid_edx(0x80000005); + assoc = cpuid >> 16 & 0xff; + upperbit = ((cpuid >> 24) << 10) / assoc; + + va_align.mask = (upperbit - 1) & PAGE_MASK; + va_align.flags = ALIGN_VA_32 | ALIGN_VA_64; + + /* A random value per boot for bit slice [12:upper_bit) */ + va_align.bits = get_random_int() & va_align.mask; + } + + if (cpu_has(c, X86_FEATURE_MWAITX)) + use_mwaitx_delay(); + + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + u32 ecx; + + ecx = cpuid_ecx(0x8000001e); + __max_die_per_package = nodes_per_socket = ((ecx >> 8) & 7) + 1; + } else if (boot_cpu_has(X86_FEATURE_NODEID_MSR)) { + u64 value; + + rdmsrl(MSR_FAM10H_NODE_ID, value); + __max_die_per_package = nodes_per_socket = ((value >> 3) & 7) + 1; + } + + if (!boot_cpu_has(X86_FEATURE_AMD_SSBD) && + !boot_cpu_has(X86_FEATURE_VIRT_SSBD) && + c->x86 >= 0x15 && c->x86 <= 0x17) { + unsigned int bit; + + switch (c->x86) { + case 0x15: bit = 54; break; + case 0x16: bit = 33; break; + case 0x17: bit = 10; break; + default: return; + } + /* + * Try to cache the base value so further operations can + * avoid RMW. If that faults, do not enable SSBD. + */ + if (!rdmsrl_safe(MSR_AMD64_LS_CFG, &x86_amd_ls_cfg_base)) { + setup_force_cpu_cap(X86_FEATURE_LS_CFG_SSBD); + setup_force_cpu_cap(X86_FEATURE_SSBD); + x86_amd_ls_cfg_ssbd_mask = 1ULL << bit; + } + } + + resctrl_cpu_detect(c); +} + +static void early_detect_mem_encrypt(struct cpuinfo_x86 *c) +{ + u64 msr; + + /* + * BIOS support is required for SME and SEV. + * For SME: If BIOS has enabled SME then adjust x86_phys_bits by + * the SME physical address space reduction value. + * If BIOS has not enabled SME then don't advertise the + * SME feature (set in scattered.c). + * For SEV: If BIOS has not enabled SEV then don't advertise the + * SEV and SEV_ES feature (set in scattered.c). + * + * In all cases, since support for SME and SEV requires long mode, + * don't advertise the feature under CONFIG_X86_32. + */ + if (cpu_has(c, X86_FEATURE_SME) || cpu_has(c, X86_FEATURE_SEV)) { + /* Check if memory encryption is enabled */ + rdmsrl(MSR_K8_SYSCFG, msr); + if (!(msr & MSR_K8_SYSCFG_MEM_ENCRYPT)) + goto clear_all; + + /* + * Always adjust physical address bits. Even though this + * will be a value above 32-bits this is still done for + * CONFIG_X86_32 so that accurate values are reported. + */ + c->x86_phys_bits -= (cpuid_ebx(0x8000001f) >> 6) & 0x3f; + + if (IS_ENABLED(CONFIG_X86_32)) + goto clear_all; + + rdmsrl(MSR_K7_HWCR, msr); + if (!(msr & MSR_K7_HWCR_SMMLOCK)) + goto clear_sev; + + return; + +clear_all: + setup_clear_cpu_cap(X86_FEATURE_SME); +clear_sev: + setup_clear_cpu_cap(X86_FEATURE_SEV); + setup_clear_cpu_cap(X86_FEATURE_SEV_ES); + } +} + +static void early_init_amd(struct cpuinfo_x86 *c) +{ + u64 value; + u32 dummy; + + early_init_amd_mc(c); + +#ifdef CONFIG_X86_32 + if (c->x86 == 6) + set_cpu_cap(c, X86_FEATURE_K7); +#endif + + if (c->x86 >= 0xf) + set_cpu_cap(c, X86_FEATURE_K8); + + rdmsr_safe(MSR_AMD64_PATCH_LEVEL, &c->microcode, &dummy); + + /* + * c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate + * with P/T states and does not stop in deep C-states + */ + if (c->x86_power & (1 << 8)) { + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + } + + /* Bit 12 of 8000_0007 edx is accumulated power mechanism. */ + if (c->x86_power & BIT(12)) + set_cpu_cap(c, X86_FEATURE_ACC_POWER); + +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSCALL32); +#else + /* Set MTRR capability flag if appropriate */ + if (c->x86 == 5) + if (c->x86_model == 13 || c->x86_model == 9 || + (c->x86_model == 8 && c->x86_stepping >= 8)) + set_cpu_cap(c, X86_FEATURE_K6_MTRR); +#endif +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_PCI) + /* + * ApicID can always be treated as an 8-bit value for AMD APIC versions + * >= 0x10, but even old K8s came out of reset with version 0x10. So, we + * can safely set X86_FEATURE_EXTD_APICID unconditionally for families + * after 16h. + */ + if (boot_cpu_has(X86_FEATURE_APIC)) { + if (c->x86 > 0x16) + set_cpu_cap(c, X86_FEATURE_EXTD_APICID); + else if (c->x86 >= 0xf) { + /* check CPU config space for extended APIC ID */ + unsigned int val; + + val = read_pci_config(0, 24, 0, 0x68); + if ((val >> 17 & 0x3) == 0x3) + set_cpu_cap(c, X86_FEATURE_EXTD_APICID); + } + } +#endif + + /* + * This is only needed to tell the kernel whether to use VMCALL + * and VMMCALL. VMMCALL is never executed except under virt, so + * we can set it unconditionally. + */ + set_cpu_cap(c, X86_FEATURE_VMMCALL); + + /* F16h erratum 793, CVE-2013-6885 */ + if (c->x86 == 0x16 && c->x86_model <= 0xf) + msr_set_bit(MSR_AMD64_LS_CFG, 15); + + /* + * Check whether the machine is affected by erratum 400. This is + * used to select the proper idle routine and to enable the check + * whether the machine is affected in arch_post_acpi_init(), which + * sets the X86_BUG_AMD_APIC_C1E bug depending on the MSR check. + */ + if (cpu_has_amd_erratum(c, amd_erratum_400)) + set_cpu_bug(c, X86_BUG_AMD_E400); + + early_detect_mem_encrypt(c); + + /* Re-enable TopologyExtensions if switched off by BIOS */ + if (c->x86 == 0x15 && + (c->x86_model >= 0x10 && c->x86_model <= 0x6f) && + !cpu_has(c, X86_FEATURE_TOPOEXT)) { + + if (msr_set_bit(0xc0011005, 54) > 0) { + rdmsrl(0xc0011005, value); + if (value & BIT_64(54)) { + set_cpu_cap(c, X86_FEATURE_TOPOEXT); + pr_info_once(FW_INFO "CPU: Re-enabling disabled Topology Extensions Support.\n"); + } + } + } + + if (cpu_has(c, X86_FEATURE_TOPOEXT)) + smp_num_siblings = ((cpuid_ebx(0x8000001e) >> 8) & 0xff) + 1; +} + +static void init_amd_k8(struct cpuinfo_x86 *c) +{ + u32 level; + u64 value; + + /* On C+ stepping K8 rep microcode works well for copy/memset */ + level = cpuid_eax(1); + if ((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + + /* + * Some BIOSes incorrectly force this feature, but only K8 revision D + * (model = 0x14) and later actually support it. + * (AMD Erratum #110, docId: 25759). + */ + if (c->x86_model < 0x14 && cpu_has(c, X86_FEATURE_LAHF_LM)) { + clear_cpu_cap(c, X86_FEATURE_LAHF_LM); + if (!rdmsrl_amd_safe(0xc001100d, &value)) { + value &= ~BIT_64(32); + wrmsrl_amd_safe(0xc001100d, value); + } + } + + if (!c->x86_model_id[0]) + strcpy(c->x86_model_id, "Hammer"); + +#ifdef CONFIG_SMP + /* + * Disable TLB flush filter by setting HWCR.FFDIS on K8 + * bit 6 of msr C001_0015 + * + * Errata 63 for SH-B3 steppings + * Errata 122 for all steppings (F+ have it disabled by default) + */ + msr_set_bit(MSR_K7_HWCR, 6); +#endif + set_cpu_bug(c, X86_BUG_SWAPGS_FENCE); +} + +static void init_amd_gh(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_MMCONF_FAM10H + /* do this for boot cpu */ + if (c == &boot_cpu_data) + check_enable_amd_mmconf_dmi(); + + fam10h_check_enable_mmcfg(); +#endif + + /* + * Disable GART TLB Walk Errors on Fam10h. We do this here because this + * is always needed when GART is enabled, even in a kernel which has no + * MCE support built in. BIOS should disable GartTlbWlk Errors already. + * If it doesn't, we do it here as suggested by the BKDG. + * + * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012 + */ + msr_set_bit(MSR_AMD64_MCx_MASK(4), 10); + + /* + * On family 10h BIOS may not have properly enabled WC+ support, causing + * it to be converted to CD memtype. This may result in performance + * degradation for certain nested-paging guests. Prevent this conversion + * by clearing bit 24 in MSR_AMD64_BU_CFG2. + * + * NOTE: we want to use the _safe accessors so as not to #GP kvm + * guests on older kvm hosts. + */ + msr_clear_bit(MSR_AMD64_BU_CFG2, 24); + + if (cpu_has_amd_erratum(c, amd_erratum_383)) + set_cpu_bug(c, X86_BUG_AMD_TLB_MMATCH); +} + +static void init_amd_ln(struct cpuinfo_x86 *c) +{ + /* + * Apply erratum 665 fix unconditionally so machines without a BIOS + * fix work. + */ + msr_set_bit(MSR_AMD64_DE_CFG, 31); +} + +static bool rdrand_force; + +static int __init rdrand_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + if (!strcmp(str, "force")) + rdrand_force = true; + else + return -EINVAL; + + return 0; +} +early_param("rdrand", rdrand_cmdline); + +static void clear_rdrand_cpuid_bit(struct cpuinfo_x86 *c) +{ + /* + * Saving of the MSR used to hide the RDRAND support during + * suspend/resume is done by arch/x86/power/cpu.c, which is + * dependent on CONFIG_PM_SLEEP. + */ + if (!IS_ENABLED(CONFIG_PM_SLEEP)) + return; + + /* + * The nordrand option can clear X86_FEATURE_RDRAND, so check for + * RDRAND support using the CPUID function directly. + */ + if (!(cpuid_ecx(1) & BIT(30)) || rdrand_force) + return; + + msr_clear_bit(MSR_AMD64_CPUID_FN_1, 62); + + /* + * Verify that the CPUID change has occurred in case the kernel is + * running virtualized and the hypervisor doesn't support the MSR. + */ + if (cpuid_ecx(1) & BIT(30)) { + pr_info_once("BIOS may not properly restore RDRAND after suspend, but hypervisor does not support hiding RDRAND via CPUID.\n"); + return; + } + + clear_cpu_cap(c, X86_FEATURE_RDRAND); + pr_info_once("BIOS may not properly restore RDRAND after suspend, hiding RDRAND via CPUID. Use rdrand=force to reenable.\n"); +} + +static void init_amd_jg(struct cpuinfo_x86 *c) +{ + /* + * Some BIOS implementations do not restore proper RDRAND support + * across suspend and resume. Check on whether to hide the RDRAND + * instruction support via CPUID. + */ + clear_rdrand_cpuid_bit(c); +} + +static void init_amd_bd(struct cpuinfo_x86 *c) +{ + u64 value; + + /* + * The way access filter has a performance penalty on some workloads. + * Disable it on the affected CPUs. + */ + if ((c->x86_model >= 0x02) && (c->x86_model < 0x20)) { + if (!rdmsrl_safe(MSR_F15H_IC_CFG, &value) && !(value & 0x1E)) { + value |= 0x1E; + wrmsrl_safe(MSR_F15H_IC_CFG, value); + } + } + + /* + * Some BIOS implementations do not restore proper RDRAND support + * across suspend and resume. Check on whether to hide the RDRAND + * instruction support via CPUID. + */ + clear_rdrand_cpuid_bit(c); +} + +void init_spectral_chicken(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_CPU_UNRET_ENTRY + u64 value; + + /* + * On Zen2 we offer this chicken (bit) on the altar of Speculation. + * + * This suppresses speculation from the middle of a basic block, i.e. it + * suppresses non-branch predictions. + * + * We use STIBP as a heuristic to filter out Zen2 from the rest of F17H + */ + if (!cpu_has(c, X86_FEATURE_HYPERVISOR) && cpu_has(c, X86_FEATURE_AMD_STIBP)) { + if (!rdmsrl_safe(MSR_ZEN2_SPECTRAL_CHICKEN, &value)) { + value |= MSR_ZEN2_SPECTRAL_CHICKEN_BIT; + wrmsrl_safe(MSR_ZEN2_SPECTRAL_CHICKEN, value); + } + } +#endif + /* + * Work around Erratum 1386. The XSAVES instruction malfunctions in + * certain circumstances on Zen1/2 uarch, and not all parts have had + * updated microcode at the time of writing (March 2023). + * + * Affected parts all have no supervisor XSAVE states, meaning that + * the XSAVEC instruction (which works fine) is equivalent. + */ + clear_cpu_cap(c, X86_FEATURE_XSAVES); +} + +static void init_amd_zn(struct cpuinfo_x86 *c) +{ + set_cpu_cap(c, X86_FEATURE_ZEN); + +#ifdef CONFIG_NUMA + node_reclaim_distance = 32; +#endif + + /* Fix up CPUID bits, but only if not virtualised. */ + if (!cpu_has(c, X86_FEATURE_HYPERVISOR)) { + + /* Erratum 1076: CPB feature bit not being set in CPUID. */ + if (!cpu_has(c, X86_FEATURE_CPB)) + set_cpu_cap(c, X86_FEATURE_CPB); + + /* + * Zen3 (Fam19 model < 0x10) parts are not susceptible to + * Branch Type Confusion, but predate the allocation of the + * BTC_NO bit. + */ + if (c->x86 == 0x19 && !cpu_has(c, X86_FEATURE_BTC_NO)) + set_cpu_cap(c, X86_FEATURE_BTC_NO); + } +} + +static bool cpu_has_zenbleed_microcode(void) +{ + u32 good_rev = 0; + + switch (boot_cpu_data.x86_model) { + case 0x30 ... 0x3f: good_rev = 0x0830107a; break; + case 0x60 ... 0x67: good_rev = 0x0860010b; break; + case 0x68 ... 0x6f: good_rev = 0x08608105; break; + case 0x70 ... 0x7f: good_rev = 0x08701032; break; + case 0xa0 ... 0xaf: good_rev = 0x08a00008; break; + + default: + return false; + break; + } + + if (boot_cpu_data.microcode < good_rev) + return false; + + return true; +} + +static void zenbleed_check(struct cpuinfo_x86 *c) +{ + if (!cpu_has_amd_erratum(c, amd_zenbleed)) + return; + + if (cpu_has(c, X86_FEATURE_HYPERVISOR)) + return; + + if (!cpu_has(c, X86_FEATURE_AVX)) + return; + + if (!cpu_has_zenbleed_microcode()) { + pr_notice_once("Zenbleed: please update your microcode for the most optimal fix\n"); + msr_set_bit(MSR_AMD64_DE_CFG, MSR_AMD64_DE_CFG_ZEN2_FP_BACKUP_FIX_BIT); + } else { + msr_clear_bit(MSR_AMD64_DE_CFG, MSR_AMD64_DE_CFG_ZEN2_FP_BACKUP_FIX_BIT); + } +} + +static void init_amd(struct cpuinfo_x86 *c) +{ + early_init_amd(c); + + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); + + if (c->x86 >= 0x10) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + + /* get apicid instead of initial apic id from cpuid */ + c->apicid = hard_smp_processor_id(); + + /* K6s reports MCEs but don't actually have all the MSRs */ + if (c->x86 < 6) + clear_cpu_cap(c, X86_FEATURE_MCE); + + switch (c->x86) { + case 4: init_amd_k5(c); break; + case 5: init_amd_k6(c); break; + case 6: init_amd_k7(c); break; + case 0xf: init_amd_k8(c); break; + case 0x10: init_amd_gh(c); break; + case 0x12: init_amd_ln(c); break; + case 0x15: init_amd_bd(c); break; + case 0x16: init_amd_jg(c); break; + case 0x17: init_spectral_chicken(c); + fallthrough; + case 0x19: init_amd_zn(c); break; + } + + /* + * Enable workaround for FXSAVE leak on CPUs + * without a XSaveErPtr feature + */ + if ((c->x86 >= 6) && (!cpu_has(c, X86_FEATURE_XSAVEERPTR))) + set_cpu_bug(c, X86_BUG_FXSAVE_LEAK); + + cpu_detect_cache_sizes(c); + + amd_detect_cmp(c); + amd_get_topology(c); + srat_detect_node(c); + amd_detect_ppin(c); + + init_amd_cacheinfo(c); + + if (cpu_has(c, X86_FEATURE_XMM2)) { + /* + * Use LFENCE for execution serialization. On families which + * don't have that MSR, LFENCE is already serializing. + * msr_set_bit() uses the safe accessors, too, even if the MSR + * is not present. + */ + msr_set_bit(MSR_AMD64_DE_CFG, + MSR_AMD64_DE_CFG_LFENCE_SERIALIZE_BIT); + + /* A serializing LFENCE stops RDTSC speculation */ + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); + } + + /* + * Family 0x12 and above processors have APIC timer + * running in deep C states. + */ + if (c->x86 > 0x11) + set_cpu_cap(c, X86_FEATURE_ARAT); + + /* 3DNow or LM implies PREFETCHW */ + if (!cpu_has(c, X86_FEATURE_3DNOWPREFETCH)) + if (cpu_has(c, X86_FEATURE_3DNOW) || cpu_has(c, X86_FEATURE_LM)) + set_cpu_cap(c, X86_FEATURE_3DNOWPREFETCH); + + /* AMD CPUs don't reset SS attributes on SYSRET, Xen does. */ + if (!cpu_has(c, X86_FEATURE_XENPV)) + set_cpu_bug(c, X86_BUG_SYSRET_SS_ATTRS); + + /* + * Turn on the Instructions Retired free counter on machines not + * susceptible to erratum #1054 "Instructions Retired Performance + * Counter May Be Inaccurate". + */ + if (cpu_has(c, X86_FEATURE_IRPERF) && + !cpu_has_amd_erratum(c, amd_erratum_1054)) + msr_set_bit(MSR_K7_HWCR, MSR_K7_HWCR_IRPERF_EN_BIT); + + check_null_seg_clears_base(c); + + zenbleed_check(c); + + if (cpu_has_amd_erratum(c, amd_div0)) { + pr_notice_once("AMD Zen1 DIV0 bug detected. Disable SMT for full protection.\n"); + setup_force_cpu_bug(X86_BUG_DIV0); + } + + if (!cpu_has(c, X86_FEATURE_HYPERVISOR) && + cpu_has_amd_erratum(c, amd_erratum_1485)) + msr_set_bit(MSR_ZEN4_BP_CFG, MSR_ZEN4_BP_CFG_SHARED_BTB_FIX_BIT); +} + +#ifdef CONFIG_X86_32 +static unsigned int amd_size_cache(struct cpuinfo_x86 *c, unsigned int size) +{ + /* AMD errata T13 (order #21922) */ + if (c->x86 == 6) { + /* Duron Rev A0 */ + if (c->x86_model == 3 && c->x86_stepping == 0) + size = 64; + /* Tbird rev A1/A2 */ + if (c->x86_model == 4 && + (c->x86_stepping == 0 || c->x86_stepping == 1)) + size = 256; + } + return size; +} +#endif + +static void cpu_detect_tlb_amd(struct cpuinfo_x86 *c) +{ + u32 ebx, eax, ecx, edx; + u16 mask = 0xfff; + + if (c->x86 < 0xf) + return; + + if (c->extended_cpuid_level < 0x80000006) + return; + + cpuid(0x80000006, &eax, &ebx, &ecx, &edx); + + tlb_lld_4k[ENTRIES] = (ebx >> 16) & mask; + tlb_lli_4k[ENTRIES] = ebx & mask; + + /* + * K8 doesn't have 2M/4M entries in the L2 TLB so read out the L1 TLB + * characteristics from the CPUID function 0x80000005 instead. + */ + if (c->x86 == 0xf) { + cpuid(0x80000005, &eax, &ebx, &ecx, &edx); + mask = 0xff; + } + + /* Handle DTLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!((eax >> 16) & mask)) + tlb_lld_2m[ENTRIES] = (cpuid_eax(0x80000005) >> 16) & 0xff; + else + tlb_lld_2m[ENTRIES] = (eax >> 16) & mask; + + /* a 4M entry uses two 2M entries */ + tlb_lld_4m[ENTRIES] = tlb_lld_2m[ENTRIES] >> 1; + + /* Handle ITLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!(eax & mask)) { + /* Erratum 658 */ + if (c->x86 == 0x15 && c->x86_model <= 0x1f) { + tlb_lli_2m[ENTRIES] = 1024; + } else { + cpuid(0x80000005, &eax, &ebx, &ecx, &edx); + tlb_lli_2m[ENTRIES] = eax & 0xff; + } + } else + tlb_lli_2m[ENTRIES] = eax & mask; + + tlb_lli_4m[ENTRIES] = tlb_lli_2m[ENTRIES] >> 1; +} + +static const struct cpu_dev amd_cpu_dev = { + .c_vendor = "AMD", + .c_ident = { "AuthenticAMD" }, +#ifdef CONFIG_X86_32 + .legacy_models = { + { .family = 4, .model_names = + { + [3] = "486 DX/2", + [7] = "486 DX/2-WB", + [8] = "486 DX/4", + [9] = "486 DX/4-WB", + [14] = "Am5x86-WT", + [15] = "Am5x86-WB" + } + }, + }, + .legacy_cache_size = amd_size_cache, +#endif + .c_early_init = early_init_amd, + .c_detect_tlb = cpu_detect_tlb_amd, + .c_bsp_init = bsp_init_amd, + .c_init = init_amd, + .c_x86_vendor = X86_VENDOR_AMD, +}; + +cpu_dev_register(amd_cpu_dev); + +void set_dr_addr_mask(unsigned long mask, int dr) +{ + if (!boot_cpu_has(X86_FEATURE_BPEXT)) + return; + + switch (dr) { + case 0: + wrmsr(MSR_F16H_DR0_ADDR_MASK, mask, 0); + break; + case 1: + case 2: + case 3: + wrmsr(MSR_F16H_DR1_ADDR_MASK - 1 + dr, mask, 0); + break; + default: + break; + } +} + +bool cpu_has_ibpb_brtype_microcode(void) +{ + switch (boot_cpu_data.x86) { + /* Zen1/2 IBPB flushes branch type predictions too. */ + case 0x17: + return boot_cpu_has(X86_FEATURE_AMD_IBPB); + case 0x19: + /* Poke the MSR bit on Zen3/4 to check its presence. */ + if (!wrmsrl_safe(MSR_IA32_PRED_CMD, PRED_CMD_SBPB)) { + setup_force_cpu_cap(X86_FEATURE_SBPB); + return true; + } else { + return false; + } + default: + return false; + } +} + +static void zenbleed_check_cpu(void *unused) +{ + struct cpuinfo_x86 *c = &cpu_data(smp_processor_id()); + + zenbleed_check(c); +} + +void amd_check_microcode(void) +{ + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) + return; + + on_each_cpu(zenbleed_check_cpu, NULL, 1); +} + +/* + * Issue a DIV 0/1 insn to clear any division data from previous DIV + * operations. + */ +void noinstr amd_clear_divider(void) +{ + asm volatile(ALTERNATIVE("", "div %2\n\t", X86_BUG_DIV0) + :: "a" (0), "d" (0), "r" (1)); +} +EXPORT_SYMBOL_GPL(amd_clear_divider); diff --git a/arch/x86/kernel/cpu/aperfmperf.c b/arch/x86/kernel/cpu/aperfmperf.c new file mode 100644 index 000000000..e2f319dc9 --- /dev/null +++ b/arch/x86/kernel/cpu/aperfmperf.c @@ -0,0 +1,137 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * x86 APERF/MPERF KHz calculation for + * /sys/.../cpufreq/scaling_cur_freq + * + * Copyright (C) 2017 Intel Corp. + * Author: Len Brown <len.brown@intel.com> + */ + +#include <linux/delay.h> +#include <linux/ktime.h> +#include <linux/math64.h> +#include <linux/percpu.h> +#include <linux/cpufreq.h> +#include <linux/smp.h> +#include <linux/sched/isolation.h> + +#include "cpu.h" + +struct aperfmperf_sample { + unsigned int khz; + ktime_t time; + u64 aperf; + u64 mperf; +}; + +static DEFINE_PER_CPU(struct aperfmperf_sample, samples); + +#define APERFMPERF_CACHE_THRESHOLD_MS 10 +#define APERFMPERF_REFRESH_DELAY_MS 10 +#define APERFMPERF_STALE_THRESHOLD_MS 1000 + +/* + * aperfmperf_snapshot_khz() + * On the current CPU, snapshot APERF, MPERF, and jiffies + * unless we already did it within 10ms + * calculate kHz, save snapshot + */ +static void aperfmperf_snapshot_khz(void *dummy) +{ + u64 aperf, aperf_delta; + u64 mperf, mperf_delta; + struct aperfmperf_sample *s = this_cpu_ptr(&samples); + unsigned long flags; + + local_irq_save(flags); + rdmsrl(MSR_IA32_APERF, aperf); + rdmsrl(MSR_IA32_MPERF, mperf); + local_irq_restore(flags); + + aperf_delta = aperf - s->aperf; + mperf_delta = mperf - s->mperf; + + /* + * There is no architectural guarantee that MPERF + * increments faster than we can read it. + */ + if (mperf_delta == 0) + return; + + s->time = ktime_get(); + s->aperf = aperf; + s->mperf = mperf; + s->khz = div64_u64((cpu_khz * aperf_delta), mperf_delta); +} + +static bool aperfmperf_snapshot_cpu(int cpu, ktime_t now, bool wait) +{ + s64 time_delta = ktime_ms_delta(now, per_cpu(samples.time, cpu)); + + /* Don't bother re-computing within the cache threshold time. */ + if (time_delta < APERFMPERF_CACHE_THRESHOLD_MS) + return true; + + smp_call_function_single(cpu, aperfmperf_snapshot_khz, NULL, wait); + + /* Return false if the previous iteration was too long ago. */ + return time_delta <= APERFMPERF_STALE_THRESHOLD_MS; +} + +unsigned int aperfmperf_get_khz(int cpu) +{ + if (!cpu_khz) + return 0; + + if (!boot_cpu_has(X86_FEATURE_APERFMPERF)) + return 0; + + if (!housekeeping_cpu(cpu, HK_FLAG_MISC)) + return 0; + + aperfmperf_snapshot_cpu(cpu, ktime_get(), true); + return per_cpu(samples.khz, cpu); +} + +void arch_freq_prepare_all(void) +{ + ktime_t now = ktime_get(); + bool wait = false; + int cpu; + + if (!cpu_khz) + return; + + if (!boot_cpu_has(X86_FEATURE_APERFMPERF)) + return; + + for_each_online_cpu(cpu) { + if (!housekeeping_cpu(cpu, HK_FLAG_MISC)) + continue; + if (!aperfmperf_snapshot_cpu(cpu, now, false)) + wait = true; + } + + if (wait) + msleep(APERFMPERF_REFRESH_DELAY_MS); +} + +unsigned int arch_freq_get_on_cpu(int cpu) +{ + if (!cpu_khz) + return 0; + + if (!boot_cpu_has(X86_FEATURE_APERFMPERF)) + return 0; + + if (!housekeeping_cpu(cpu, HK_FLAG_MISC)) + return 0; + + if (aperfmperf_snapshot_cpu(cpu, ktime_get(), true)) + return per_cpu(samples.khz, cpu); + + msleep(APERFMPERF_REFRESH_DELAY_MS); + smp_call_function_single(cpu, aperfmperf_snapshot_khz, NULL, 1); + + return per_cpu(samples.khz, cpu); +} diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c new file mode 100644 index 000000000..d9fda0b6e --- /dev/null +++ b/arch/x86/kernel/cpu/bugs.c @@ -0,0 +1,2757 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) 1994 Linus Torvalds + * + * Cyrix stuff, June 1998 by: + * - Rafael R. Reilova (moved everything from head.S), + * <rreilova@ececs.uc.edu> + * - Channing Corn (tests & fixes), + * - Andrew D. Balsa (code cleanup). + */ +#include <linux/init.h> +#include <linux/cpu.h> +#include <linux/module.h> +#include <linux/nospec.h> +#include <linux/prctl.h> +#include <linux/sched/smt.h> +#include <linux/pgtable.h> +#include <linux/bpf.h> + +#include <asm/spec-ctrl.h> +#include <asm/cmdline.h> +#include <asm/bugs.h> +#include <asm/processor.h> +#include <asm/processor-flags.h> +#include <asm/fpu/internal.h> +#include <asm/msr.h> +#include <asm/vmx.h> +#include <asm/paravirt.h> +#include <asm/intel-family.h> +#include <asm/e820/api.h> +#include <asm/hypervisor.h> +#include <asm/tlbflush.h> + +#include "cpu.h" + +static void __init spectre_v1_select_mitigation(void); +static void __init spectre_v2_select_mitigation(void); +static void __init retbleed_select_mitigation(void); +static void __init spectre_v2_user_select_mitigation(void); +static void __init ssb_select_mitigation(void); +static void __init l1tf_select_mitigation(void); +static void __init mds_select_mitigation(void); +static void __init md_clear_update_mitigation(void); +static void __init md_clear_select_mitigation(void); +static void __init taa_select_mitigation(void); +static void __init mmio_select_mitigation(void); +static void __init srbds_select_mitigation(void); +static void __init gds_select_mitigation(void); +static void __init srso_select_mitigation(void); + +/* The base value of the SPEC_CTRL MSR without task-specific bits set */ +u64 x86_spec_ctrl_base; +EXPORT_SYMBOL_GPL(x86_spec_ctrl_base); + +/* The current value of the SPEC_CTRL MSR with task-specific bits set */ +DEFINE_PER_CPU(u64, x86_spec_ctrl_current); +EXPORT_SYMBOL_GPL(x86_spec_ctrl_current); + +u64 x86_pred_cmd __ro_after_init = PRED_CMD_IBPB; +EXPORT_SYMBOL_GPL(x86_pred_cmd); + +static DEFINE_MUTEX(spec_ctrl_mutex); + +void (*x86_return_thunk)(void) __ro_after_init = &__x86_return_thunk; + +/* Update SPEC_CTRL MSR and its cached copy unconditionally */ +static void update_spec_ctrl(u64 val) +{ + this_cpu_write(x86_spec_ctrl_current, val); + wrmsrl(MSR_IA32_SPEC_CTRL, val); +} + +/* + * Keep track of the SPEC_CTRL MSR value for the current task, which may differ + * from x86_spec_ctrl_base due to STIBP/SSB in __speculation_ctrl_update(). + */ +void update_spec_ctrl_cond(u64 val) +{ + if (this_cpu_read(x86_spec_ctrl_current) == val) + return; + + this_cpu_write(x86_spec_ctrl_current, val); + + /* + * When KERNEL_IBRS this MSR is written on return-to-user, unless + * forced the update can be delayed until that time. + */ + if (!cpu_feature_enabled(X86_FEATURE_KERNEL_IBRS)) + wrmsrl(MSR_IA32_SPEC_CTRL, val); +} + +u64 spec_ctrl_current(void) +{ + return this_cpu_read(x86_spec_ctrl_current); +} +EXPORT_SYMBOL_GPL(spec_ctrl_current); + +/* + * AMD specific MSR info for Speculative Store Bypass control. + * x86_amd_ls_cfg_ssbd_mask is initialized in identify_boot_cpu(). + */ +u64 __ro_after_init x86_amd_ls_cfg_base; +u64 __ro_after_init x86_amd_ls_cfg_ssbd_mask; + +/* Control conditional STIBP in switch_to() */ +DEFINE_STATIC_KEY_FALSE(switch_to_cond_stibp); +/* Control conditional IBPB in switch_mm() */ +DEFINE_STATIC_KEY_FALSE(switch_mm_cond_ibpb); +/* Control unconditional IBPB in switch_mm() */ +DEFINE_STATIC_KEY_FALSE(switch_mm_always_ibpb); + +/* Control MDS CPU buffer clear before returning to user space */ +DEFINE_STATIC_KEY_FALSE(mds_user_clear); +EXPORT_SYMBOL_GPL(mds_user_clear); +/* Control MDS CPU buffer clear before idling (halt, mwait) */ +DEFINE_STATIC_KEY_FALSE(mds_idle_clear); +EXPORT_SYMBOL_GPL(mds_idle_clear); + +/* Controls CPU Fill buffer clear before KVM guest MMIO accesses */ +DEFINE_STATIC_KEY_FALSE(mmio_stale_data_clear); +EXPORT_SYMBOL_GPL(mmio_stale_data_clear); + +void __init cpu_select_mitigations(void) +{ + /* + * Read the SPEC_CTRL MSR to account for reserved bits which may + * have unknown values. AMD64_LS_CFG MSR is cached in the early AMD + * init code as it is not enumerated and depends on the family. + */ + if (cpu_feature_enabled(X86_FEATURE_MSR_SPEC_CTRL)) { + rdmsrl(MSR_IA32_SPEC_CTRL, x86_spec_ctrl_base); + + /* + * Previously running kernel (kexec), may have some controls + * turned ON. Clear them and let the mitigations setup below + * rediscover them based on configuration. + */ + x86_spec_ctrl_base &= ~SPEC_CTRL_MITIGATIONS_MASK; + } + + /* Select the proper CPU mitigations before patching alternatives: */ + spectre_v1_select_mitigation(); + spectre_v2_select_mitigation(); + /* + * retbleed_select_mitigation() relies on the state set by + * spectre_v2_select_mitigation(); specifically it wants to know about + * spectre_v2=ibrs. + */ + retbleed_select_mitigation(); + /* + * spectre_v2_user_select_mitigation() relies on the state set by + * retbleed_select_mitigation(); specifically the STIBP selection is + * forced for UNRET or IBPB. + */ + spectre_v2_user_select_mitigation(); + ssb_select_mitigation(); + l1tf_select_mitigation(); + md_clear_select_mitigation(); + srbds_select_mitigation(); + + /* + * srso_select_mitigation() depends and must run after + * retbleed_select_mitigation(). + */ + srso_select_mitigation(); + gds_select_mitigation(); +} + +/* + * NOTE: For VMX, this function is not called in the vmexit path. + * It uses vmx_spec_ctrl_restore_host() instead. + */ +void +x86_virt_spec_ctrl(u64 guest_spec_ctrl, u64 guest_virt_spec_ctrl, bool setguest) +{ + u64 msrval, guestval = guest_spec_ctrl, hostval = spec_ctrl_current(); + struct thread_info *ti = current_thread_info(); + + if (static_cpu_has(X86_FEATURE_MSR_SPEC_CTRL)) { + if (hostval != guestval) { + msrval = setguest ? guestval : hostval; + wrmsrl(MSR_IA32_SPEC_CTRL, msrval); + } + } + + /* + * If SSBD is not handled in MSR_SPEC_CTRL on AMD, update + * MSR_AMD64_L2_CFG or MSR_VIRT_SPEC_CTRL if supported. + */ + if (!static_cpu_has(X86_FEATURE_LS_CFG_SSBD) && + !static_cpu_has(X86_FEATURE_VIRT_SSBD)) + return; + + /* + * If the host has SSBD mitigation enabled, force it in the host's + * virtual MSR value. If its not permanently enabled, evaluate + * current's TIF_SSBD thread flag. + */ + if (static_cpu_has(X86_FEATURE_SPEC_STORE_BYPASS_DISABLE)) + hostval = SPEC_CTRL_SSBD; + else + hostval = ssbd_tif_to_spec_ctrl(ti->flags); + + /* Sanitize the guest value */ + guestval = guest_virt_spec_ctrl & SPEC_CTRL_SSBD; + + if (hostval != guestval) { + unsigned long tif; + + tif = setguest ? ssbd_spec_ctrl_to_tif(guestval) : + ssbd_spec_ctrl_to_tif(hostval); + + speculation_ctrl_update(tif); + } +} +EXPORT_SYMBOL_GPL(x86_virt_spec_ctrl); + +static void x86_amd_ssb_disable(void) +{ + u64 msrval = x86_amd_ls_cfg_base | x86_amd_ls_cfg_ssbd_mask; + + if (boot_cpu_has(X86_FEATURE_VIRT_SSBD)) + wrmsrl(MSR_AMD64_VIRT_SPEC_CTRL, SPEC_CTRL_SSBD); + else if (boot_cpu_has(X86_FEATURE_LS_CFG_SSBD)) + wrmsrl(MSR_AMD64_LS_CFG, msrval); +} + +#undef pr_fmt +#define pr_fmt(fmt) "MDS: " fmt + +/* Default mitigation for MDS-affected CPUs */ +static enum mds_mitigations mds_mitigation __ro_after_init = MDS_MITIGATION_FULL; +static bool mds_nosmt __ro_after_init = false; + +static const char * const mds_strings[] = { + [MDS_MITIGATION_OFF] = "Vulnerable", + [MDS_MITIGATION_FULL] = "Mitigation: Clear CPU buffers", + [MDS_MITIGATION_VMWERV] = "Vulnerable: Clear CPU buffers attempted, no microcode", +}; + +static void __init mds_select_mitigation(void) +{ + if (!boot_cpu_has_bug(X86_BUG_MDS) || cpu_mitigations_off()) { + mds_mitigation = MDS_MITIGATION_OFF; + return; + } + + if (mds_mitigation == MDS_MITIGATION_FULL) { + if (!boot_cpu_has(X86_FEATURE_MD_CLEAR)) + mds_mitigation = MDS_MITIGATION_VMWERV; + + static_branch_enable(&mds_user_clear); + + if (!boot_cpu_has(X86_BUG_MSBDS_ONLY) && + (mds_nosmt || cpu_mitigations_auto_nosmt())) + cpu_smt_disable(false); + } +} + +static int __init mds_cmdline(char *str) +{ + if (!boot_cpu_has_bug(X86_BUG_MDS)) + return 0; + + if (!str) + return -EINVAL; + + if (!strcmp(str, "off")) + mds_mitigation = MDS_MITIGATION_OFF; + else if (!strcmp(str, "full")) + mds_mitigation = MDS_MITIGATION_FULL; + else if (!strcmp(str, "full,nosmt")) { + mds_mitigation = MDS_MITIGATION_FULL; + mds_nosmt = true; + } + + return 0; +} +early_param("mds", mds_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "TAA: " fmt + +enum taa_mitigations { + TAA_MITIGATION_OFF, + TAA_MITIGATION_UCODE_NEEDED, + TAA_MITIGATION_VERW, + TAA_MITIGATION_TSX_DISABLED, +}; + +/* Default mitigation for TAA-affected CPUs */ +static enum taa_mitigations taa_mitigation __ro_after_init = TAA_MITIGATION_VERW; +static bool taa_nosmt __ro_after_init; + +static const char * const taa_strings[] = { + [TAA_MITIGATION_OFF] = "Vulnerable", + [TAA_MITIGATION_UCODE_NEEDED] = "Vulnerable: Clear CPU buffers attempted, no microcode", + [TAA_MITIGATION_VERW] = "Mitigation: Clear CPU buffers", + [TAA_MITIGATION_TSX_DISABLED] = "Mitigation: TSX disabled", +}; + +static void __init taa_select_mitigation(void) +{ + u64 ia32_cap; + + if (!boot_cpu_has_bug(X86_BUG_TAA)) { + taa_mitigation = TAA_MITIGATION_OFF; + return; + } + + /* TSX previously disabled by tsx=off */ + if (!boot_cpu_has(X86_FEATURE_RTM)) { + taa_mitigation = TAA_MITIGATION_TSX_DISABLED; + return; + } + + if (cpu_mitigations_off()) { + taa_mitigation = TAA_MITIGATION_OFF; + return; + } + + /* + * TAA mitigation via VERW is turned off if both + * tsx_async_abort=off and mds=off are specified. + */ + if (taa_mitigation == TAA_MITIGATION_OFF && + mds_mitigation == MDS_MITIGATION_OFF) + return; + + if (boot_cpu_has(X86_FEATURE_MD_CLEAR)) + taa_mitigation = TAA_MITIGATION_VERW; + else + taa_mitigation = TAA_MITIGATION_UCODE_NEEDED; + + /* + * VERW doesn't clear the CPU buffers when MD_CLEAR=1 and MDS_NO=1. + * A microcode update fixes this behavior to clear CPU buffers. It also + * adds support for MSR_IA32_TSX_CTRL which is enumerated by the + * ARCH_CAP_TSX_CTRL_MSR bit. + * + * On MDS_NO=1 CPUs if ARCH_CAP_TSX_CTRL_MSR is not set, microcode + * update is required. + */ + ia32_cap = x86_read_arch_cap_msr(); + if ( (ia32_cap & ARCH_CAP_MDS_NO) && + !(ia32_cap & ARCH_CAP_TSX_CTRL_MSR)) + taa_mitigation = TAA_MITIGATION_UCODE_NEEDED; + + /* + * TSX is enabled, select alternate mitigation for TAA which is + * the same as MDS. Enable MDS static branch to clear CPU buffers. + * + * For guests that can't determine whether the correct microcode is + * present on host, enable the mitigation for UCODE_NEEDED as well. + */ + static_branch_enable(&mds_user_clear); + + if (taa_nosmt || cpu_mitigations_auto_nosmt()) + cpu_smt_disable(false); +} + +static int __init tsx_async_abort_parse_cmdline(char *str) +{ + if (!boot_cpu_has_bug(X86_BUG_TAA)) + return 0; + + if (!str) + return -EINVAL; + + if (!strcmp(str, "off")) { + taa_mitigation = TAA_MITIGATION_OFF; + } else if (!strcmp(str, "full")) { + taa_mitigation = TAA_MITIGATION_VERW; + } else if (!strcmp(str, "full,nosmt")) { + taa_mitigation = TAA_MITIGATION_VERW; + taa_nosmt = true; + } + + return 0; +} +early_param("tsx_async_abort", tsx_async_abort_parse_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "MMIO Stale Data: " fmt + +enum mmio_mitigations { + MMIO_MITIGATION_OFF, + MMIO_MITIGATION_UCODE_NEEDED, + MMIO_MITIGATION_VERW, +}; + +/* Default mitigation for Processor MMIO Stale Data vulnerabilities */ +static enum mmio_mitigations mmio_mitigation __ro_after_init = MMIO_MITIGATION_VERW; +static bool mmio_nosmt __ro_after_init = false; + +static const char * const mmio_strings[] = { + [MMIO_MITIGATION_OFF] = "Vulnerable", + [MMIO_MITIGATION_UCODE_NEEDED] = "Vulnerable: Clear CPU buffers attempted, no microcode", + [MMIO_MITIGATION_VERW] = "Mitigation: Clear CPU buffers", +}; + +static void __init mmio_select_mitigation(void) +{ + u64 ia32_cap; + + if (!boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA) || + boot_cpu_has_bug(X86_BUG_MMIO_UNKNOWN) || + cpu_mitigations_off()) { + mmio_mitigation = MMIO_MITIGATION_OFF; + return; + } + + if (mmio_mitigation == MMIO_MITIGATION_OFF) + return; + + ia32_cap = x86_read_arch_cap_msr(); + + /* + * Enable CPU buffer clear mitigation for host and VMM, if also affected + * by MDS or TAA. Otherwise, enable mitigation for VMM only. + */ + if (boot_cpu_has_bug(X86_BUG_MDS) || (boot_cpu_has_bug(X86_BUG_TAA) && + boot_cpu_has(X86_FEATURE_RTM))) + static_branch_enable(&mds_user_clear); + else + static_branch_enable(&mmio_stale_data_clear); + + /* + * If Processor-MMIO-Stale-Data bug is present and Fill Buffer data can + * be propagated to uncore buffers, clearing the Fill buffers on idle + * is required irrespective of SMT state. + */ + if (!(ia32_cap & ARCH_CAP_FBSDP_NO)) + static_branch_enable(&mds_idle_clear); + + /* + * Check if the system has the right microcode. + * + * CPU Fill buffer clear mitigation is enumerated by either an explicit + * FB_CLEAR or by the presence of both MD_CLEAR and L1D_FLUSH on MDS + * affected systems. + */ + if ((ia32_cap & ARCH_CAP_FB_CLEAR) || + (boot_cpu_has(X86_FEATURE_MD_CLEAR) && + boot_cpu_has(X86_FEATURE_FLUSH_L1D) && + !(ia32_cap & ARCH_CAP_MDS_NO))) + mmio_mitigation = MMIO_MITIGATION_VERW; + else + mmio_mitigation = MMIO_MITIGATION_UCODE_NEEDED; + + if (mmio_nosmt || cpu_mitigations_auto_nosmt()) + cpu_smt_disable(false); +} + +static int __init mmio_stale_data_parse_cmdline(char *str) +{ + if (!boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) + return 0; + + if (!str) + return -EINVAL; + + if (!strcmp(str, "off")) { + mmio_mitigation = MMIO_MITIGATION_OFF; + } else if (!strcmp(str, "full")) { + mmio_mitigation = MMIO_MITIGATION_VERW; + } else if (!strcmp(str, "full,nosmt")) { + mmio_mitigation = MMIO_MITIGATION_VERW; + mmio_nosmt = true; + } + + return 0; +} +early_param("mmio_stale_data", mmio_stale_data_parse_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "" fmt + +static void __init md_clear_update_mitigation(void) +{ + if (cpu_mitigations_off()) + return; + + if (!static_key_enabled(&mds_user_clear)) + goto out; + + /* + * mds_user_clear is now enabled. Update MDS, TAA and MMIO Stale Data + * mitigation, if necessary. + */ + if (mds_mitigation == MDS_MITIGATION_OFF && + boot_cpu_has_bug(X86_BUG_MDS)) { + mds_mitigation = MDS_MITIGATION_FULL; + mds_select_mitigation(); + } + if (taa_mitigation == TAA_MITIGATION_OFF && + boot_cpu_has_bug(X86_BUG_TAA)) { + taa_mitigation = TAA_MITIGATION_VERW; + taa_select_mitigation(); + } + if (mmio_mitigation == MMIO_MITIGATION_OFF && + boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) { + mmio_mitigation = MMIO_MITIGATION_VERW; + mmio_select_mitigation(); + } +out: + if (boot_cpu_has_bug(X86_BUG_MDS)) + pr_info("MDS: %s\n", mds_strings[mds_mitigation]); + if (boot_cpu_has_bug(X86_BUG_TAA)) + pr_info("TAA: %s\n", taa_strings[taa_mitigation]); + if (boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) + pr_info("MMIO Stale Data: %s\n", mmio_strings[mmio_mitigation]); + else if (boot_cpu_has_bug(X86_BUG_MMIO_UNKNOWN)) + pr_info("MMIO Stale Data: Unknown: No mitigations\n"); +} + +static void __init md_clear_select_mitigation(void) +{ + mds_select_mitigation(); + taa_select_mitigation(); + mmio_select_mitigation(); + + /* + * As MDS, TAA and MMIO Stale Data mitigations are inter-related, update + * and print their mitigation after MDS, TAA and MMIO Stale Data + * mitigation selection is done. + */ + md_clear_update_mitigation(); +} + +#undef pr_fmt +#define pr_fmt(fmt) "SRBDS: " fmt + +enum srbds_mitigations { + SRBDS_MITIGATION_OFF, + SRBDS_MITIGATION_UCODE_NEEDED, + SRBDS_MITIGATION_FULL, + SRBDS_MITIGATION_TSX_OFF, + SRBDS_MITIGATION_HYPERVISOR, +}; + +static enum srbds_mitigations srbds_mitigation __ro_after_init = SRBDS_MITIGATION_FULL; + +static const char * const srbds_strings[] = { + [SRBDS_MITIGATION_OFF] = "Vulnerable", + [SRBDS_MITIGATION_UCODE_NEEDED] = "Vulnerable: No microcode", + [SRBDS_MITIGATION_FULL] = "Mitigation: Microcode", + [SRBDS_MITIGATION_TSX_OFF] = "Mitigation: TSX disabled", + [SRBDS_MITIGATION_HYPERVISOR] = "Unknown: Dependent on hypervisor status", +}; + +static bool srbds_off; + +void update_srbds_msr(void) +{ + u64 mcu_ctrl; + + if (!boot_cpu_has_bug(X86_BUG_SRBDS)) + return; + + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) + return; + + if (srbds_mitigation == SRBDS_MITIGATION_UCODE_NEEDED) + return; + + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); + + switch (srbds_mitigation) { + case SRBDS_MITIGATION_OFF: + case SRBDS_MITIGATION_TSX_OFF: + mcu_ctrl |= RNGDS_MITG_DIS; + break; + case SRBDS_MITIGATION_FULL: + mcu_ctrl &= ~RNGDS_MITG_DIS; + break; + default: + break; + } + + wrmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); +} + +static void __init srbds_select_mitigation(void) +{ + u64 ia32_cap; + + if (!boot_cpu_has_bug(X86_BUG_SRBDS)) + return; + + /* + * Check to see if this is one of the MDS_NO systems supporting TSX that + * are only exposed to SRBDS when TSX is enabled or when CPU is affected + * by Processor MMIO Stale Data vulnerability. + */ + ia32_cap = x86_read_arch_cap_msr(); + if ((ia32_cap & ARCH_CAP_MDS_NO) && !boot_cpu_has(X86_FEATURE_RTM) && + !boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) + srbds_mitigation = SRBDS_MITIGATION_TSX_OFF; + else if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) + srbds_mitigation = SRBDS_MITIGATION_HYPERVISOR; + else if (!boot_cpu_has(X86_FEATURE_SRBDS_CTRL)) + srbds_mitigation = SRBDS_MITIGATION_UCODE_NEEDED; + else if (cpu_mitigations_off() || srbds_off) + srbds_mitigation = SRBDS_MITIGATION_OFF; + + update_srbds_msr(); + pr_info("%s\n", srbds_strings[srbds_mitigation]); +} + +static int __init srbds_parse_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + if (!boot_cpu_has_bug(X86_BUG_SRBDS)) + return 0; + + srbds_off = !strcmp(str, "off"); + return 0; +} +early_param("srbds", srbds_parse_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "GDS: " fmt + +enum gds_mitigations { + GDS_MITIGATION_OFF, + GDS_MITIGATION_UCODE_NEEDED, + GDS_MITIGATION_FORCE, + GDS_MITIGATION_FULL, + GDS_MITIGATION_FULL_LOCKED, + GDS_MITIGATION_HYPERVISOR, +}; + +#if IS_ENABLED(CONFIG_GDS_FORCE_MITIGATION) +static enum gds_mitigations gds_mitigation __ro_after_init = GDS_MITIGATION_FORCE; +#else +static enum gds_mitigations gds_mitigation __ro_after_init = GDS_MITIGATION_FULL; +#endif + +static const char * const gds_strings[] = { + [GDS_MITIGATION_OFF] = "Vulnerable", + [GDS_MITIGATION_UCODE_NEEDED] = "Vulnerable: No microcode", + [GDS_MITIGATION_FORCE] = "Mitigation: AVX disabled, no microcode", + [GDS_MITIGATION_FULL] = "Mitigation: Microcode", + [GDS_MITIGATION_FULL_LOCKED] = "Mitigation: Microcode (locked)", + [GDS_MITIGATION_HYPERVISOR] = "Unknown: Dependent on hypervisor status", +}; + +bool gds_ucode_mitigated(void) +{ + return (gds_mitigation == GDS_MITIGATION_FULL || + gds_mitigation == GDS_MITIGATION_FULL_LOCKED); +} +EXPORT_SYMBOL_GPL(gds_ucode_mitigated); + +void update_gds_msr(void) +{ + u64 mcu_ctrl_after; + u64 mcu_ctrl; + + switch (gds_mitigation) { + case GDS_MITIGATION_OFF: + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); + mcu_ctrl |= GDS_MITG_DIS; + break; + case GDS_MITIGATION_FULL_LOCKED: + /* + * The LOCKED state comes from the boot CPU. APs might not have + * the same state. Make sure the mitigation is enabled on all + * CPUs. + */ + case GDS_MITIGATION_FULL: + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); + mcu_ctrl &= ~GDS_MITG_DIS; + break; + case GDS_MITIGATION_FORCE: + case GDS_MITIGATION_UCODE_NEEDED: + case GDS_MITIGATION_HYPERVISOR: + return; + }; + + wrmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); + + /* + * Check to make sure that the WRMSR value was not ignored. Writes to + * GDS_MITG_DIS will be ignored if this processor is locked but the boot + * processor was not. + */ + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl_after); + WARN_ON_ONCE(mcu_ctrl != mcu_ctrl_after); +} + +static void __init gds_select_mitigation(void) +{ + u64 mcu_ctrl; + + if (!boot_cpu_has_bug(X86_BUG_GDS)) + return; + + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) { + gds_mitigation = GDS_MITIGATION_HYPERVISOR; + goto out; + } + + if (cpu_mitigations_off()) + gds_mitigation = GDS_MITIGATION_OFF; + /* Will verify below that mitigation _can_ be disabled */ + + /* No microcode */ + if (!(x86_read_arch_cap_msr() & ARCH_CAP_GDS_CTRL)) { + if (gds_mitigation == GDS_MITIGATION_FORCE) { + /* + * This only needs to be done on the boot CPU so do it + * here rather than in update_gds_msr() + */ + setup_clear_cpu_cap(X86_FEATURE_AVX); + pr_warn("Microcode update needed! Disabling AVX as mitigation.\n"); + } else { + gds_mitigation = GDS_MITIGATION_UCODE_NEEDED; + } + goto out; + } + + /* Microcode has mitigation, use it */ + if (gds_mitigation == GDS_MITIGATION_FORCE) + gds_mitigation = GDS_MITIGATION_FULL; + + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_ctrl); + if (mcu_ctrl & GDS_MITG_LOCKED) { + if (gds_mitigation == GDS_MITIGATION_OFF) + pr_warn("Mitigation locked. Disable failed.\n"); + + /* + * The mitigation is selected from the boot CPU. All other CPUs + * _should_ have the same state. If the boot CPU isn't locked + * but others are then update_gds_msr() will WARN() of the state + * mismatch. If the boot CPU is locked update_gds_msr() will + * ensure the other CPUs have the mitigation enabled. + */ + gds_mitigation = GDS_MITIGATION_FULL_LOCKED; + } + + update_gds_msr(); +out: + pr_info("%s\n", gds_strings[gds_mitigation]); +} + +static int __init gds_parse_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + if (!boot_cpu_has_bug(X86_BUG_GDS)) + return 0; + + if (!strcmp(str, "off")) + gds_mitigation = GDS_MITIGATION_OFF; + else if (!strcmp(str, "force")) + gds_mitigation = GDS_MITIGATION_FORCE; + + return 0; +} +early_param("gather_data_sampling", gds_parse_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "Spectre V1 : " fmt + +enum spectre_v1_mitigation { + SPECTRE_V1_MITIGATION_NONE, + SPECTRE_V1_MITIGATION_AUTO, +}; + +static enum spectre_v1_mitigation spectre_v1_mitigation __ro_after_init = + SPECTRE_V1_MITIGATION_AUTO; + +static const char * const spectre_v1_strings[] = { + [SPECTRE_V1_MITIGATION_NONE] = "Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers", + [SPECTRE_V1_MITIGATION_AUTO] = "Mitigation: usercopy/swapgs barriers and __user pointer sanitization", +}; + +/* + * Does SMAP provide full mitigation against speculative kernel access to + * userspace? + */ +static bool smap_works_speculatively(void) +{ + if (!boot_cpu_has(X86_FEATURE_SMAP)) + return false; + + /* + * On CPUs which are vulnerable to Meltdown, SMAP does not + * prevent speculative access to user data in the L1 cache. + * Consider SMAP to be non-functional as a mitigation on these + * CPUs. + */ + if (boot_cpu_has(X86_BUG_CPU_MELTDOWN)) + return false; + + return true; +} + +static void __init spectre_v1_select_mitigation(void) +{ + if (!boot_cpu_has_bug(X86_BUG_SPECTRE_V1) || cpu_mitigations_off()) { + spectre_v1_mitigation = SPECTRE_V1_MITIGATION_NONE; + return; + } + + if (spectre_v1_mitigation == SPECTRE_V1_MITIGATION_AUTO) { + /* + * With Spectre v1, a user can speculatively control either + * path of a conditional swapgs with a user-controlled GS + * value. The mitigation is to add lfences to both code paths. + * + * If FSGSBASE is enabled, the user can put a kernel address in + * GS, in which case SMAP provides no protection. + * + * If FSGSBASE is disabled, the user can only put a user space + * address in GS. That makes an attack harder, but still + * possible if there's no SMAP protection. + */ + if (boot_cpu_has(X86_FEATURE_FSGSBASE) || + !smap_works_speculatively()) { + /* + * Mitigation can be provided from SWAPGS itself or + * PTI as the CR3 write in the Meltdown mitigation + * is serializing. + * + * If neither is there, mitigate with an LFENCE to + * stop speculation through swapgs. + */ + if (boot_cpu_has_bug(X86_BUG_SWAPGS) && + !boot_cpu_has(X86_FEATURE_PTI)) + setup_force_cpu_cap(X86_FEATURE_FENCE_SWAPGS_USER); + + /* + * Enable lfences in the kernel entry (non-swapgs) + * paths, to prevent user entry from speculatively + * skipping swapgs. + */ + setup_force_cpu_cap(X86_FEATURE_FENCE_SWAPGS_KERNEL); + } + } + + pr_info("%s\n", spectre_v1_strings[spectre_v1_mitigation]); +} + +static int __init nospectre_v1_cmdline(char *str) +{ + spectre_v1_mitigation = SPECTRE_V1_MITIGATION_NONE; + return 0; +} +early_param("nospectre_v1", nospectre_v1_cmdline); + +static enum spectre_v2_mitigation spectre_v2_enabled __ro_after_init = + SPECTRE_V2_NONE; + +#undef pr_fmt +#define pr_fmt(fmt) "RETBleed: " fmt + +enum retbleed_mitigation { + RETBLEED_MITIGATION_NONE, + RETBLEED_MITIGATION_UNRET, + RETBLEED_MITIGATION_IBPB, + RETBLEED_MITIGATION_IBRS, + RETBLEED_MITIGATION_EIBRS, +}; + +enum retbleed_mitigation_cmd { + RETBLEED_CMD_OFF, + RETBLEED_CMD_AUTO, + RETBLEED_CMD_UNRET, + RETBLEED_CMD_IBPB, +}; + +const char * const retbleed_strings[] = { + [RETBLEED_MITIGATION_NONE] = "Vulnerable", + [RETBLEED_MITIGATION_UNRET] = "Mitigation: untrained return thunk", + [RETBLEED_MITIGATION_IBPB] = "Mitigation: IBPB", + [RETBLEED_MITIGATION_IBRS] = "Mitigation: IBRS", + [RETBLEED_MITIGATION_EIBRS] = "Mitigation: Enhanced IBRS", +}; + +static enum retbleed_mitigation retbleed_mitigation __ro_after_init = + RETBLEED_MITIGATION_NONE; +static enum retbleed_mitigation_cmd retbleed_cmd __ro_after_init = + RETBLEED_CMD_AUTO; + +static int __ro_after_init retbleed_nosmt = false; + +static int __init retbleed_parse_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + while (str) { + char *next = strchr(str, ','); + if (next) { + *next = 0; + next++; + } + + if (!strcmp(str, "off")) { + retbleed_cmd = RETBLEED_CMD_OFF; + } else if (!strcmp(str, "auto")) { + retbleed_cmd = RETBLEED_CMD_AUTO; + } else if (!strcmp(str, "unret")) { + retbleed_cmd = RETBLEED_CMD_UNRET; + } else if (!strcmp(str, "ibpb")) { + retbleed_cmd = RETBLEED_CMD_IBPB; + } else if (!strcmp(str, "nosmt")) { + retbleed_nosmt = true; + } else { + pr_err("Ignoring unknown retbleed option (%s).", str); + } + + str = next; + } + + return 0; +} +early_param("retbleed", retbleed_parse_cmdline); + +#define RETBLEED_UNTRAIN_MSG "WARNING: BTB untrained return thunk mitigation is only effective on AMD/Hygon!\n" +#define RETBLEED_INTEL_MSG "WARNING: Spectre v2 mitigation leaves CPU vulnerable to RETBleed attacks, data leaks possible!\n" + +static void __init retbleed_select_mitigation(void) +{ + bool mitigate_smt = false; + + if (!boot_cpu_has_bug(X86_BUG_RETBLEED) || cpu_mitigations_off()) + return; + + switch (retbleed_cmd) { + case RETBLEED_CMD_OFF: + return; + + case RETBLEED_CMD_UNRET: + if (IS_ENABLED(CONFIG_CPU_UNRET_ENTRY)) { + retbleed_mitigation = RETBLEED_MITIGATION_UNRET; + } else { + pr_err("WARNING: kernel not compiled with CPU_UNRET_ENTRY.\n"); + goto do_cmd_auto; + } + break; + + case RETBLEED_CMD_IBPB: + if (!boot_cpu_has(X86_FEATURE_IBPB)) { + pr_err("WARNING: CPU does not support IBPB.\n"); + goto do_cmd_auto; + } else if (IS_ENABLED(CONFIG_CPU_IBPB_ENTRY)) { + retbleed_mitigation = RETBLEED_MITIGATION_IBPB; + } else { + pr_err("WARNING: kernel not compiled with CPU_IBPB_ENTRY.\n"); + goto do_cmd_auto; + } + break; + +do_cmd_auto: + case RETBLEED_CMD_AUTO: + default: + if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD || + boot_cpu_data.x86_vendor == X86_VENDOR_HYGON) { + if (IS_ENABLED(CONFIG_CPU_UNRET_ENTRY)) + retbleed_mitigation = RETBLEED_MITIGATION_UNRET; + else if (IS_ENABLED(CONFIG_CPU_IBPB_ENTRY) && boot_cpu_has(X86_FEATURE_IBPB)) + retbleed_mitigation = RETBLEED_MITIGATION_IBPB; + } + + /* + * The Intel mitigation (IBRS or eIBRS) was already selected in + * spectre_v2_select_mitigation(). 'retbleed_mitigation' will + * be set accordingly below. + */ + + break; + } + + switch (retbleed_mitigation) { + case RETBLEED_MITIGATION_UNRET: + setup_force_cpu_cap(X86_FEATURE_RETHUNK); + setup_force_cpu_cap(X86_FEATURE_UNRET); + + if (IS_ENABLED(CONFIG_RETHUNK)) + x86_return_thunk = retbleed_return_thunk; + + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD && + boot_cpu_data.x86_vendor != X86_VENDOR_HYGON) + pr_err(RETBLEED_UNTRAIN_MSG); + + mitigate_smt = true; + break; + + case RETBLEED_MITIGATION_IBPB: + setup_force_cpu_cap(X86_FEATURE_ENTRY_IBPB); + mitigate_smt = true; + break; + + default: + break; + } + + if (mitigate_smt && !boot_cpu_has(X86_FEATURE_STIBP) && + (retbleed_nosmt || cpu_mitigations_auto_nosmt())) + cpu_smt_disable(false); + + /* + * Let IBRS trump all on Intel without affecting the effects of the + * retbleed= cmdline option. + */ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) { + switch (spectre_v2_enabled) { + case SPECTRE_V2_IBRS: + retbleed_mitigation = RETBLEED_MITIGATION_IBRS; + break; + case SPECTRE_V2_EIBRS: + case SPECTRE_V2_EIBRS_RETPOLINE: + case SPECTRE_V2_EIBRS_LFENCE: + retbleed_mitigation = RETBLEED_MITIGATION_EIBRS; + break; + default: + pr_err(RETBLEED_INTEL_MSG); + } + } + + pr_info("%s\n", retbleed_strings[retbleed_mitigation]); +} + +#undef pr_fmt +#define pr_fmt(fmt) "Spectre V2 : " fmt + +static enum spectre_v2_user_mitigation spectre_v2_user_stibp __ro_after_init = + SPECTRE_V2_USER_NONE; +static enum spectre_v2_user_mitigation spectre_v2_user_ibpb __ro_after_init = + SPECTRE_V2_USER_NONE; + +#ifdef CONFIG_RETPOLINE +static bool spectre_v2_bad_module; + +bool retpoline_module_ok(bool has_retpoline) +{ + if (spectre_v2_enabled == SPECTRE_V2_NONE || has_retpoline) + return true; + + pr_err("System may be vulnerable to spectre v2\n"); + spectre_v2_bad_module = true; + return false; +} + +static inline const char *spectre_v2_module_string(void) +{ + return spectre_v2_bad_module ? " - vulnerable module loaded" : ""; +} +#else +static inline const char *spectre_v2_module_string(void) { return ""; } +#endif + +#define SPECTRE_V2_LFENCE_MSG "WARNING: LFENCE mitigation is not recommended for this CPU, data leaks possible!\n" +#define SPECTRE_V2_EIBRS_EBPF_MSG "WARNING: Unprivileged eBPF is enabled with eIBRS on, data leaks possible via Spectre v2 BHB attacks!\n" +#define SPECTRE_V2_EIBRS_LFENCE_EBPF_SMT_MSG "WARNING: Unprivileged eBPF is enabled with eIBRS+LFENCE mitigation and SMT, data leaks possible via Spectre v2 BHB attacks!\n" +#define SPECTRE_V2_IBRS_PERF_MSG "WARNING: IBRS mitigation selected on Enhanced IBRS CPU, this may cause unnecessary performance loss\n" + +#ifdef CONFIG_BPF_SYSCALL +void unpriv_ebpf_notify(int new_state) +{ + if (new_state) + return; + + /* Unprivileged eBPF is enabled */ + + switch (spectre_v2_enabled) { + case SPECTRE_V2_EIBRS: + pr_err(SPECTRE_V2_EIBRS_EBPF_MSG); + break; + case SPECTRE_V2_EIBRS_LFENCE: + if (sched_smt_active()) + pr_err(SPECTRE_V2_EIBRS_LFENCE_EBPF_SMT_MSG); + break; + default: + break; + } +} +#endif + +static inline bool match_option(const char *arg, int arglen, const char *opt) +{ + int len = strlen(opt); + + return len == arglen && !strncmp(arg, opt, len); +} + +/* The kernel command line selection for spectre v2 */ +enum spectre_v2_mitigation_cmd { + SPECTRE_V2_CMD_NONE, + SPECTRE_V2_CMD_AUTO, + SPECTRE_V2_CMD_FORCE, + SPECTRE_V2_CMD_RETPOLINE, + SPECTRE_V2_CMD_RETPOLINE_GENERIC, + SPECTRE_V2_CMD_RETPOLINE_LFENCE, + SPECTRE_V2_CMD_EIBRS, + SPECTRE_V2_CMD_EIBRS_RETPOLINE, + SPECTRE_V2_CMD_EIBRS_LFENCE, + SPECTRE_V2_CMD_IBRS, +}; + +enum spectre_v2_user_cmd { + SPECTRE_V2_USER_CMD_NONE, + SPECTRE_V2_USER_CMD_AUTO, + SPECTRE_V2_USER_CMD_FORCE, + SPECTRE_V2_USER_CMD_PRCTL, + SPECTRE_V2_USER_CMD_PRCTL_IBPB, + SPECTRE_V2_USER_CMD_SECCOMP, + SPECTRE_V2_USER_CMD_SECCOMP_IBPB, +}; + +static const char * const spectre_v2_user_strings[] = { + [SPECTRE_V2_USER_NONE] = "User space: Vulnerable", + [SPECTRE_V2_USER_STRICT] = "User space: Mitigation: STIBP protection", + [SPECTRE_V2_USER_STRICT_PREFERRED] = "User space: Mitigation: STIBP always-on protection", + [SPECTRE_V2_USER_PRCTL] = "User space: Mitigation: STIBP via prctl", + [SPECTRE_V2_USER_SECCOMP] = "User space: Mitigation: STIBP via seccomp and prctl", +}; + +static const struct { + const char *option; + enum spectre_v2_user_cmd cmd; + bool secure; +} v2_user_options[] __initconst = { + { "auto", SPECTRE_V2_USER_CMD_AUTO, false }, + { "off", SPECTRE_V2_USER_CMD_NONE, false }, + { "on", SPECTRE_V2_USER_CMD_FORCE, true }, + { "prctl", SPECTRE_V2_USER_CMD_PRCTL, false }, + { "prctl,ibpb", SPECTRE_V2_USER_CMD_PRCTL_IBPB, false }, + { "seccomp", SPECTRE_V2_USER_CMD_SECCOMP, false }, + { "seccomp,ibpb", SPECTRE_V2_USER_CMD_SECCOMP_IBPB, false }, +}; + +static void __init spec_v2_user_print_cond(const char *reason, bool secure) +{ + if (boot_cpu_has_bug(X86_BUG_SPECTRE_V2) != secure) + pr_info("spectre_v2_user=%s forced on command line.\n", reason); +} + +static __ro_after_init enum spectre_v2_mitigation_cmd spectre_v2_cmd; + +static enum spectre_v2_user_cmd __init +spectre_v2_parse_user_cmdline(void) +{ + char arg[20]; + int ret, i; + + switch (spectre_v2_cmd) { + case SPECTRE_V2_CMD_NONE: + return SPECTRE_V2_USER_CMD_NONE; + case SPECTRE_V2_CMD_FORCE: + return SPECTRE_V2_USER_CMD_FORCE; + default: + break; + } + + ret = cmdline_find_option(boot_command_line, "spectre_v2_user", + arg, sizeof(arg)); + if (ret < 0) + return SPECTRE_V2_USER_CMD_AUTO; + + for (i = 0; i < ARRAY_SIZE(v2_user_options); i++) { + if (match_option(arg, ret, v2_user_options[i].option)) { + spec_v2_user_print_cond(v2_user_options[i].option, + v2_user_options[i].secure); + return v2_user_options[i].cmd; + } + } + + pr_err("Unknown user space protection option (%s). Switching to AUTO select\n", arg); + return SPECTRE_V2_USER_CMD_AUTO; +} + +static inline bool spectre_v2_in_eibrs_mode(enum spectre_v2_mitigation mode) +{ + return mode == SPECTRE_V2_EIBRS || + mode == SPECTRE_V2_EIBRS_RETPOLINE || + mode == SPECTRE_V2_EIBRS_LFENCE; +} + +static inline bool spectre_v2_in_ibrs_mode(enum spectre_v2_mitigation mode) +{ + return spectre_v2_in_eibrs_mode(mode) || mode == SPECTRE_V2_IBRS; +} + +static void __init +spectre_v2_user_select_mitigation(void) +{ + enum spectre_v2_user_mitigation mode = SPECTRE_V2_USER_NONE; + bool smt_possible = IS_ENABLED(CONFIG_SMP); + enum spectre_v2_user_cmd cmd; + + if (!boot_cpu_has(X86_FEATURE_IBPB) && !boot_cpu_has(X86_FEATURE_STIBP)) + return; + + if (cpu_smt_control == CPU_SMT_FORCE_DISABLED || + cpu_smt_control == CPU_SMT_NOT_SUPPORTED) + smt_possible = false; + + cmd = spectre_v2_parse_user_cmdline(); + switch (cmd) { + case SPECTRE_V2_USER_CMD_NONE: + goto set_mode; + case SPECTRE_V2_USER_CMD_FORCE: + mode = SPECTRE_V2_USER_STRICT; + break; + case SPECTRE_V2_USER_CMD_PRCTL: + case SPECTRE_V2_USER_CMD_PRCTL_IBPB: + mode = SPECTRE_V2_USER_PRCTL; + break; + case SPECTRE_V2_USER_CMD_AUTO: + case SPECTRE_V2_USER_CMD_SECCOMP: + case SPECTRE_V2_USER_CMD_SECCOMP_IBPB: + if (IS_ENABLED(CONFIG_SECCOMP)) + mode = SPECTRE_V2_USER_SECCOMP; + else + mode = SPECTRE_V2_USER_PRCTL; + break; + } + + /* Initialize Indirect Branch Prediction Barrier */ + if (boot_cpu_has(X86_FEATURE_IBPB)) { + setup_force_cpu_cap(X86_FEATURE_USE_IBPB); + + spectre_v2_user_ibpb = mode; + switch (cmd) { + case SPECTRE_V2_USER_CMD_FORCE: + case SPECTRE_V2_USER_CMD_PRCTL_IBPB: + case SPECTRE_V2_USER_CMD_SECCOMP_IBPB: + static_branch_enable(&switch_mm_always_ibpb); + spectre_v2_user_ibpb = SPECTRE_V2_USER_STRICT; + break; + case SPECTRE_V2_USER_CMD_PRCTL: + case SPECTRE_V2_USER_CMD_AUTO: + case SPECTRE_V2_USER_CMD_SECCOMP: + static_branch_enable(&switch_mm_cond_ibpb); + break; + default: + break; + } + + pr_info("mitigation: Enabling %s Indirect Branch Prediction Barrier\n", + static_key_enabled(&switch_mm_always_ibpb) ? + "always-on" : "conditional"); + } + + /* + * If no STIBP, enhanced IBRS is enabled, or SMT impossible, STIBP + * is not required. + * + * Enhanced IBRS also protects against cross-thread branch target + * injection in user-mode as the IBRS bit remains always set which + * implicitly enables cross-thread protections. However, in legacy IBRS + * mode, the IBRS bit is set only on kernel entry and cleared on return + * to userspace. This disables the implicit cross-thread protection, + * so allow for STIBP to be selected in that case. + */ + if (!boot_cpu_has(X86_FEATURE_STIBP) || + !smt_possible || + spectre_v2_in_eibrs_mode(spectre_v2_enabled)) + return; + + /* + * At this point, an STIBP mode other than "off" has been set. + * If STIBP support is not being forced, check if STIBP always-on + * is preferred. + */ + if (mode != SPECTRE_V2_USER_STRICT && + boot_cpu_has(X86_FEATURE_AMD_STIBP_ALWAYS_ON)) + mode = SPECTRE_V2_USER_STRICT_PREFERRED; + + if (retbleed_mitigation == RETBLEED_MITIGATION_UNRET || + retbleed_mitigation == RETBLEED_MITIGATION_IBPB) { + if (mode != SPECTRE_V2_USER_STRICT && + mode != SPECTRE_V2_USER_STRICT_PREFERRED) + pr_info("Selecting STIBP always-on mode to complement retbleed mitigation\n"); + mode = SPECTRE_V2_USER_STRICT_PREFERRED; + } + + spectre_v2_user_stibp = mode; + +set_mode: + pr_info("%s\n", spectre_v2_user_strings[mode]); +} + +static const char * const spectre_v2_strings[] = { + [SPECTRE_V2_NONE] = "Vulnerable", + [SPECTRE_V2_RETPOLINE] = "Mitigation: Retpolines", + [SPECTRE_V2_LFENCE] = "Mitigation: LFENCE", + [SPECTRE_V2_EIBRS] = "Mitigation: Enhanced IBRS", + [SPECTRE_V2_EIBRS_LFENCE] = "Mitigation: Enhanced IBRS + LFENCE", + [SPECTRE_V2_EIBRS_RETPOLINE] = "Mitigation: Enhanced IBRS + Retpolines", + [SPECTRE_V2_IBRS] = "Mitigation: IBRS", +}; + +static const struct { + const char *option; + enum spectre_v2_mitigation_cmd cmd; + bool secure; +} mitigation_options[] __initconst = { + { "off", SPECTRE_V2_CMD_NONE, false }, + { "on", SPECTRE_V2_CMD_FORCE, true }, + { "retpoline", SPECTRE_V2_CMD_RETPOLINE, false }, + { "retpoline,amd", SPECTRE_V2_CMD_RETPOLINE_LFENCE, false }, + { "retpoline,lfence", SPECTRE_V2_CMD_RETPOLINE_LFENCE, false }, + { "retpoline,generic", SPECTRE_V2_CMD_RETPOLINE_GENERIC, false }, + { "eibrs", SPECTRE_V2_CMD_EIBRS, false }, + { "eibrs,lfence", SPECTRE_V2_CMD_EIBRS_LFENCE, false }, + { "eibrs,retpoline", SPECTRE_V2_CMD_EIBRS_RETPOLINE, false }, + { "auto", SPECTRE_V2_CMD_AUTO, false }, + { "ibrs", SPECTRE_V2_CMD_IBRS, false }, +}; + +static void __init spec_v2_print_cond(const char *reason, bool secure) +{ + if (boot_cpu_has_bug(X86_BUG_SPECTRE_V2) != secure) + pr_info("%s selected on command line.\n", reason); +} + +static enum spectre_v2_mitigation_cmd __init spectre_v2_parse_cmdline(void) +{ + enum spectre_v2_mitigation_cmd cmd = SPECTRE_V2_CMD_AUTO; + char arg[20]; + int ret, i; + + if (cmdline_find_option_bool(boot_command_line, "nospectre_v2") || + cpu_mitigations_off()) + return SPECTRE_V2_CMD_NONE; + + ret = cmdline_find_option(boot_command_line, "spectre_v2", arg, sizeof(arg)); + if (ret < 0) + return SPECTRE_V2_CMD_AUTO; + + for (i = 0; i < ARRAY_SIZE(mitigation_options); i++) { + if (!match_option(arg, ret, mitigation_options[i].option)) + continue; + cmd = mitigation_options[i].cmd; + break; + } + + if (i >= ARRAY_SIZE(mitigation_options)) { + pr_err("unknown option (%s). Switching to AUTO select\n", arg); + return SPECTRE_V2_CMD_AUTO; + } + + if ((cmd == SPECTRE_V2_CMD_RETPOLINE || + cmd == SPECTRE_V2_CMD_RETPOLINE_LFENCE || + cmd == SPECTRE_V2_CMD_RETPOLINE_GENERIC || + cmd == SPECTRE_V2_CMD_EIBRS_LFENCE || + cmd == SPECTRE_V2_CMD_EIBRS_RETPOLINE) && + !IS_ENABLED(CONFIG_RETPOLINE)) { + pr_err("%s selected but not compiled in. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if ((cmd == SPECTRE_V2_CMD_EIBRS || + cmd == SPECTRE_V2_CMD_EIBRS_LFENCE || + cmd == SPECTRE_V2_CMD_EIBRS_RETPOLINE) && + !boot_cpu_has(X86_FEATURE_IBRS_ENHANCED)) { + pr_err("%s selected but CPU doesn't have eIBRS. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if ((cmd == SPECTRE_V2_CMD_RETPOLINE_LFENCE || + cmd == SPECTRE_V2_CMD_EIBRS_LFENCE) && + !boot_cpu_has(X86_FEATURE_LFENCE_RDTSC)) { + pr_err("%s selected, but CPU doesn't have a serializing LFENCE. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if (cmd == SPECTRE_V2_CMD_IBRS && !IS_ENABLED(CONFIG_CPU_IBRS_ENTRY)) { + pr_err("%s selected but not compiled in. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if (cmd == SPECTRE_V2_CMD_IBRS && boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) { + pr_err("%s selected but not Intel CPU. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if (cmd == SPECTRE_V2_CMD_IBRS && !boot_cpu_has(X86_FEATURE_IBRS)) { + pr_err("%s selected but CPU doesn't have IBRS. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + if (cmd == SPECTRE_V2_CMD_IBRS && boot_cpu_has(X86_FEATURE_XENPV)) { + pr_err("%s selected but running as XenPV guest. Switching to AUTO select\n", + mitigation_options[i].option); + return SPECTRE_V2_CMD_AUTO; + } + + spec_v2_print_cond(mitigation_options[i].option, + mitigation_options[i].secure); + return cmd; +} + +static enum spectre_v2_mitigation __init spectre_v2_select_retpoline(void) +{ + if (!IS_ENABLED(CONFIG_RETPOLINE)) { + pr_err("Kernel not compiled with retpoline; no mitigation available!"); + return SPECTRE_V2_NONE; + } + + return SPECTRE_V2_RETPOLINE; +} + +/* Disable in-kernel use of non-RSB RET predictors */ +static void __init spec_ctrl_disable_kernel_rrsba(void) +{ + u64 ia32_cap; + + if (!boot_cpu_has(X86_FEATURE_RRSBA_CTRL)) + return; + + ia32_cap = x86_read_arch_cap_msr(); + + if (ia32_cap & ARCH_CAP_RRSBA) { + x86_spec_ctrl_base |= SPEC_CTRL_RRSBA_DIS_S; + update_spec_ctrl(x86_spec_ctrl_base); + } +} + +static void __init spectre_v2_determine_rsb_fill_type_at_vmexit(enum spectre_v2_mitigation mode) +{ + /* + * Similar to context switches, there are two types of RSB attacks + * after VM exit: + * + * 1) RSB underflow + * + * 2) Poisoned RSB entry + * + * When retpoline is enabled, both are mitigated by filling/clearing + * the RSB. + * + * When IBRS is enabled, while #1 would be mitigated by the IBRS branch + * prediction isolation protections, RSB still needs to be cleared + * because of #2. Note that SMEP provides no protection here, unlike + * user-space-poisoned RSB entries. + * + * eIBRS should protect against RSB poisoning, but if the EIBRS_PBRSB + * bug is present then a LITE version of RSB protection is required, + * just a single call needs to retire before a RET is executed. + */ + switch (mode) { + case SPECTRE_V2_NONE: + return; + + case SPECTRE_V2_EIBRS_LFENCE: + case SPECTRE_V2_EIBRS: + if (boot_cpu_has_bug(X86_BUG_EIBRS_PBRSB)) { + setup_force_cpu_cap(X86_FEATURE_RSB_VMEXIT_LITE); + pr_info("Spectre v2 / PBRSB-eIBRS: Retire a single CALL on VMEXIT\n"); + } + return; + + case SPECTRE_V2_EIBRS_RETPOLINE: + case SPECTRE_V2_RETPOLINE: + case SPECTRE_V2_LFENCE: + case SPECTRE_V2_IBRS: + setup_force_cpu_cap(X86_FEATURE_RSB_VMEXIT); + pr_info("Spectre v2 / SpectreRSB : Filling RSB on VMEXIT\n"); + return; + } + + pr_warn_once("Unknown Spectre v2 mode, disabling RSB mitigation at VM exit"); + dump_stack(); +} + +static void __init spectre_v2_select_mitigation(void) +{ + enum spectre_v2_mitigation_cmd cmd = spectre_v2_parse_cmdline(); + enum spectre_v2_mitigation mode = SPECTRE_V2_NONE; + + /* + * If the CPU is not affected and the command line mode is NONE or AUTO + * then nothing to do. + */ + if (!boot_cpu_has_bug(X86_BUG_SPECTRE_V2) && + (cmd == SPECTRE_V2_CMD_NONE || cmd == SPECTRE_V2_CMD_AUTO)) + return; + + switch (cmd) { + case SPECTRE_V2_CMD_NONE: + return; + + case SPECTRE_V2_CMD_FORCE: + case SPECTRE_V2_CMD_AUTO: + if (boot_cpu_has(X86_FEATURE_IBRS_ENHANCED)) { + mode = SPECTRE_V2_EIBRS; + break; + } + + if (IS_ENABLED(CONFIG_CPU_IBRS_ENTRY) && + boot_cpu_has_bug(X86_BUG_RETBLEED) && + retbleed_cmd != RETBLEED_CMD_OFF && + boot_cpu_has(X86_FEATURE_IBRS) && + boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) { + mode = SPECTRE_V2_IBRS; + break; + } + + mode = spectre_v2_select_retpoline(); + break; + + case SPECTRE_V2_CMD_RETPOLINE_LFENCE: + pr_err(SPECTRE_V2_LFENCE_MSG); + mode = SPECTRE_V2_LFENCE; + break; + + case SPECTRE_V2_CMD_RETPOLINE_GENERIC: + mode = SPECTRE_V2_RETPOLINE; + break; + + case SPECTRE_V2_CMD_RETPOLINE: + mode = spectre_v2_select_retpoline(); + break; + + case SPECTRE_V2_CMD_IBRS: + mode = SPECTRE_V2_IBRS; + break; + + case SPECTRE_V2_CMD_EIBRS: + mode = SPECTRE_V2_EIBRS; + break; + + case SPECTRE_V2_CMD_EIBRS_LFENCE: + mode = SPECTRE_V2_EIBRS_LFENCE; + break; + + case SPECTRE_V2_CMD_EIBRS_RETPOLINE: + mode = SPECTRE_V2_EIBRS_RETPOLINE; + break; + } + + if (mode == SPECTRE_V2_EIBRS && unprivileged_ebpf_enabled()) + pr_err(SPECTRE_V2_EIBRS_EBPF_MSG); + + if (spectre_v2_in_ibrs_mode(mode)) { + x86_spec_ctrl_base |= SPEC_CTRL_IBRS; + update_spec_ctrl(x86_spec_ctrl_base); + } + + switch (mode) { + case SPECTRE_V2_NONE: + case SPECTRE_V2_EIBRS: + break; + + case SPECTRE_V2_IBRS: + setup_force_cpu_cap(X86_FEATURE_KERNEL_IBRS); + if (boot_cpu_has(X86_FEATURE_IBRS_ENHANCED)) + pr_warn(SPECTRE_V2_IBRS_PERF_MSG); + break; + + case SPECTRE_V2_LFENCE: + case SPECTRE_V2_EIBRS_LFENCE: + setup_force_cpu_cap(X86_FEATURE_RETPOLINE_LFENCE); + fallthrough; + + case SPECTRE_V2_RETPOLINE: + case SPECTRE_V2_EIBRS_RETPOLINE: + setup_force_cpu_cap(X86_FEATURE_RETPOLINE); + break; + } + + /* + * Disable alternate RSB predictions in kernel when indirect CALLs and + * JMPs gets protection against BHI and Intramode-BTI, but RET + * prediction from a non-RSB predictor is still a risk. + */ + if (mode == SPECTRE_V2_EIBRS_LFENCE || + mode == SPECTRE_V2_EIBRS_RETPOLINE || + mode == SPECTRE_V2_RETPOLINE) + spec_ctrl_disable_kernel_rrsba(); + + spectre_v2_enabled = mode; + pr_info("%s\n", spectre_v2_strings[mode]); + + /* + * If Spectre v2 protection has been enabled, fill the RSB during a + * context switch. In general there are two types of RSB attacks + * across context switches, for which the CALLs/RETs may be unbalanced. + * + * 1) RSB underflow + * + * Some Intel parts have "bottomless RSB". When the RSB is empty, + * speculated return targets may come from the branch predictor, + * which could have a user-poisoned BTB or BHB entry. + * + * AMD has it even worse: *all* returns are speculated from the BTB, + * regardless of the state of the RSB. + * + * When IBRS or eIBRS is enabled, the "user -> kernel" attack + * scenario is mitigated by the IBRS branch prediction isolation + * properties, so the RSB buffer filling wouldn't be necessary to + * protect against this type of attack. + * + * The "user -> user" attack scenario is mitigated by RSB filling. + * + * 2) Poisoned RSB entry + * + * If the 'next' in-kernel return stack is shorter than 'prev', + * 'next' could be tricked into speculating with a user-poisoned RSB + * entry. + * + * The "user -> kernel" attack scenario is mitigated by SMEP and + * eIBRS. + * + * The "user -> user" scenario, also known as SpectreBHB, requires + * RSB clearing. + * + * So to mitigate all cases, unconditionally fill RSB on context + * switches. + * + * FIXME: Is this pointless for retbleed-affected AMD? + */ + setup_force_cpu_cap(X86_FEATURE_RSB_CTXSW); + pr_info("Spectre v2 / SpectreRSB mitigation: Filling RSB on context switch\n"); + + spectre_v2_determine_rsb_fill_type_at_vmexit(mode); + + /* + * Retpoline protects the kernel, but doesn't protect firmware. IBRS + * and Enhanced IBRS protect firmware too, so enable IBRS around + * firmware calls only when IBRS / Enhanced IBRS aren't otherwise + * enabled. + * + * Use "mode" to check Enhanced IBRS instead of boot_cpu_has(), because + * the user might select retpoline on the kernel command line and if + * the CPU supports Enhanced IBRS, kernel might un-intentionally not + * enable IBRS around firmware calls. + */ + if (boot_cpu_has_bug(X86_BUG_RETBLEED) && + boot_cpu_has(X86_FEATURE_IBPB) && + (boot_cpu_data.x86_vendor == X86_VENDOR_AMD || + boot_cpu_data.x86_vendor == X86_VENDOR_HYGON)) { + + if (retbleed_cmd != RETBLEED_CMD_IBPB) { + setup_force_cpu_cap(X86_FEATURE_USE_IBPB_FW); + pr_info("Enabling Speculation Barrier for firmware calls\n"); + } + + } else if (boot_cpu_has(X86_FEATURE_IBRS) && !spectre_v2_in_ibrs_mode(mode)) { + setup_force_cpu_cap(X86_FEATURE_USE_IBRS_FW); + pr_info("Enabling Restricted Speculation for firmware calls\n"); + } + + /* Set up IBPB and STIBP depending on the general spectre V2 command */ + spectre_v2_cmd = cmd; +} + +static void update_stibp_msr(void * __unused) +{ + u64 val = spec_ctrl_current() | (x86_spec_ctrl_base & SPEC_CTRL_STIBP); + update_spec_ctrl(val); +} + +/* Update x86_spec_ctrl_base in case SMT state changed. */ +static void update_stibp_strict(void) +{ + u64 mask = x86_spec_ctrl_base & ~SPEC_CTRL_STIBP; + + if (sched_smt_active()) + mask |= SPEC_CTRL_STIBP; + + if (mask == x86_spec_ctrl_base) + return; + + pr_info("Update user space SMT mitigation: STIBP %s\n", + mask & SPEC_CTRL_STIBP ? "always-on" : "off"); + x86_spec_ctrl_base = mask; + on_each_cpu(update_stibp_msr, NULL, 1); +} + +/* Update the static key controlling the evaluation of TIF_SPEC_IB */ +static void update_indir_branch_cond(void) +{ + if (sched_smt_active()) + static_branch_enable(&switch_to_cond_stibp); + else + static_branch_disable(&switch_to_cond_stibp); +} + +#undef pr_fmt +#define pr_fmt(fmt) fmt + +/* Update the static key controlling the MDS CPU buffer clear in idle */ +static void update_mds_branch_idle(void) +{ + u64 ia32_cap = x86_read_arch_cap_msr(); + + /* + * Enable the idle clearing if SMT is active on CPUs which are + * affected only by MSBDS and not any other MDS variant. + * + * The other variants cannot be mitigated when SMT is enabled, so + * clearing the buffers on idle just to prevent the Store Buffer + * repartitioning leak would be a window dressing exercise. + */ + if (!boot_cpu_has_bug(X86_BUG_MSBDS_ONLY)) + return; + + if (sched_smt_active()) { + static_branch_enable(&mds_idle_clear); + } else if (mmio_mitigation == MMIO_MITIGATION_OFF || + (ia32_cap & ARCH_CAP_FBSDP_NO)) { + static_branch_disable(&mds_idle_clear); + } +} + +#define MDS_MSG_SMT "MDS CPU bug present and SMT on, data leak possible. See https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/mds.html for more details.\n" +#define TAA_MSG_SMT "TAA CPU bug present and SMT on, data leak possible. See https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/tsx_async_abort.html for more details.\n" +#define MMIO_MSG_SMT "MMIO Stale Data CPU bug present and SMT on, data leak possible. See https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/processor_mmio_stale_data.html for more details.\n" + +void cpu_bugs_smt_update(void) +{ + mutex_lock(&spec_ctrl_mutex); + + if (sched_smt_active() && unprivileged_ebpf_enabled() && + spectre_v2_enabled == SPECTRE_V2_EIBRS_LFENCE) + pr_warn_once(SPECTRE_V2_EIBRS_LFENCE_EBPF_SMT_MSG); + + switch (spectre_v2_user_stibp) { + case SPECTRE_V2_USER_NONE: + break; + case SPECTRE_V2_USER_STRICT: + case SPECTRE_V2_USER_STRICT_PREFERRED: + update_stibp_strict(); + break; + case SPECTRE_V2_USER_PRCTL: + case SPECTRE_V2_USER_SECCOMP: + update_indir_branch_cond(); + break; + } + + switch (mds_mitigation) { + case MDS_MITIGATION_FULL: + case MDS_MITIGATION_VMWERV: + if (sched_smt_active() && !boot_cpu_has(X86_BUG_MSBDS_ONLY)) + pr_warn_once(MDS_MSG_SMT); + update_mds_branch_idle(); + break; + case MDS_MITIGATION_OFF: + break; + } + + switch (taa_mitigation) { + case TAA_MITIGATION_VERW: + case TAA_MITIGATION_UCODE_NEEDED: + if (sched_smt_active()) + pr_warn_once(TAA_MSG_SMT); + break; + case TAA_MITIGATION_TSX_DISABLED: + case TAA_MITIGATION_OFF: + break; + } + + switch (mmio_mitigation) { + case MMIO_MITIGATION_VERW: + case MMIO_MITIGATION_UCODE_NEEDED: + if (sched_smt_active()) + pr_warn_once(MMIO_MSG_SMT); + break; + case MMIO_MITIGATION_OFF: + break; + } + + mutex_unlock(&spec_ctrl_mutex); +} + +#undef pr_fmt +#define pr_fmt(fmt) "Speculative Store Bypass: " fmt + +static enum ssb_mitigation ssb_mode __ro_after_init = SPEC_STORE_BYPASS_NONE; + +/* The kernel command line selection */ +enum ssb_mitigation_cmd { + SPEC_STORE_BYPASS_CMD_NONE, + SPEC_STORE_BYPASS_CMD_AUTO, + SPEC_STORE_BYPASS_CMD_ON, + SPEC_STORE_BYPASS_CMD_PRCTL, + SPEC_STORE_BYPASS_CMD_SECCOMP, +}; + +static const char * const ssb_strings[] = { + [SPEC_STORE_BYPASS_NONE] = "Vulnerable", + [SPEC_STORE_BYPASS_DISABLE] = "Mitigation: Speculative Store Bypass disabled", + [SPEC_STORE_BYPASS_PRCTL] = "Mitigation: Speculative Store Bypass disabled via prctl", + [SPEC_STORE_BYPASS_SECCOMP] = "Mitigation: Speculative Store Bypass disabled via prctl and seccomp", +}; + +static const struct { + const char *option; + enum ssb_mitigation_cmd cmd; +} ssb_mitigation_options[] __initconst = { + { "auto", SPEC_STORE_BYPASS_CMD_AUTO }, /* Platform decides */ + { "on", SPEC_STORE_BYPASS_CMD_ON }, /* Disable Speculative Store Bypass */ + { "off", SPEC_STORE_BYPASS_CMD_NONE }, /* Don't touch Speculative Store Bypass */ + { "prctl", SPEC_STORE_BYPASS_CMD_PRCTL }, /* Disable Speculative Store Bypass via prctl */ + { "seccomp", SPEC_STORE_BYPASS_CMD_SECCOMP }, /* Disable Speculative Store Bypass via prctl and seccomp */ +}; + +static enum ssb_mitigation_cmd __init ssb_parse_cmdline(void) +{ + enum ssb_mitigation_cmd cmd = SPEC_STORE_BYPASS_CMD_AUTO; + char arg[20]; + int ret, i; + + if (cmdline_find_option_bool(boot_command_line, "nospec_store_bypass_disable") || + cpu_mitigations_off()) { + return SPEC_STORE_BYPASS_CMD_NONE; + } else { + ret = cmdline_find_option(boot_command_line, "spec_store_bypass_disable", + arg, sizeof(arg)); + if (ret < 0) + return SPEC_STORE_BYPASS_CMD_AUTO; + + for (i = 0; i < ARRAY_SIZE(ssb_mitigation_options); i++) { + if (!match_option(arg, ret, ssb_mitigation_options[i].option)) + continue; + + cmd = ssb_mitigation_options[i].cmd; + break; + } + + if (i >= ARRAY_SIZE(ssb_mitigation_options)) { + pr_err("unknown option (%s). Switching to AUTO select\n", arg); + return SPEC_STORE_BYPASS_CMD_AUTO; + } + } + + return cmd; +} + +static enum ssb_mitigation __init __ssb_select_mitigation(void) +{ + enum ssb_mitigation mode = SPEC_STORE_BYPASS_NONE; + enum ssb_mitigation_cmd cmd; + + if (!boot_cpu_has(X86_FEATURE_SSBD)) + return mode; + + cmd = ssb_parse_cmdline(); + if (!boot_cpu_has_bug(X86_BUG_SPEC_STORE_BYPASS) && + (cmd == SPEC_STORE_BYPASS_CMD_NONE || + cmd == SPEC_STORE_BYPASS_CMD_AUTO)) + return mode; + + switch (cmd) { + case SPEC_STORE_BYPASS_CMD_AUTO: + case SPEC_STORE_BYPASS_CMD_SECCOMP: + /* + * Choose prctl+seccomp as the default mode if seccomp is + * enabled. + */ + if (IS_ENABLED(CONFIG_SECCOMP)) + mode = SPEC_STORE_BYPASS_SECCOMP; + else + mode = SPEC_STORE_BYPASS_PRCTL; + break; + case SPEC_STORE_BYPASS_CMD_ON: + mode = SPEC_STORE_BYPASS_DISABLE; + break; + case SPEC_STORE_BYPASS_CMD_PRCTL: + mode = SPEC_STORE_BYPASS_PRCTL; + break; + case SPEC_STORE_BYPASS_CMD_NONE: + break; + } + + /* + * We have three CPU feature flags that are in play here: + * - X86_BUG_SPEC_STORE_BYPASS - CPU is susceptible. + * - X86_FEATURE_SSBD - CPU is able to turn off speculative store bypass + * - X86_FEATURE_SPEC_STORE_BYPASS_DISABLE - engage the mitigation + */ + if (mode == SPEC_STORE_BYPASS_DISABLE) { + setup_force_cpu_cap(X86_FEATURE_SPEC_STORE_BYPASS_DISABLE); + /* + * Intel uses the SPEC CTRL MSR Bit(2) for this, while AMD may + * use a completely different MSR and bit dependent on family. + */ + if (!static_cpu_has(X86_FEATURE_SPEC_CTRL_SSBD) && + !static_cpu_has(X86_FEATURE_AMD_SSBD)) { + x86_amd_ssb_disable(); + } else { + x86_spec_ctrl_base |= SPEC_CTRL_SSBD; + update_spec_ctrl(x86_spec_ctrl_base); + } + } + + return mode; +} + +static void ssb_select_mitigation(void) +{ + ssb_mode = __ssb_select_mitigation(); + + if (boot_cpu_has_bug(X86_BUG_SPEC_STORE_BYPASS)) + pr_info("%s\n", ssb_strings[ssb_mode]); +} + +#undef pr_fmt +#define pr_fmt(fmt) "Speculation prctl: " fmt + +static void task_update_spec_tif(struct task_struct *tsk) +{ + /* Force the update of the real TIF bits */ + set_tsk_thread_flag(tsk, TIF_SPEC_FORCE_UPDATE); + + /* + * Immediately update the speculation control MSRs for the current + * task, but for a non-current task delay setting the CPU + * mitigation until it is scheduled next. + * + * This can only happen for SECCOMP mitigation. For PRCTL it's + * always the current task. + */ + if (tsk == current) + speculation_ctrl_update_current(); +} + +static int ssb_prctl_set(struct task_struct *task, unsigned long ctrl) +{ + if (ssb_mode != SPEC_STORE_BYPASS_PRCTL && + ssb_mode != SPEC_STORE_BYPASS_SECCOMP) + return -ENXIO; + + switch (ctrl) { + case PR_SPEC_ENABLE: + /* If speculation is force disabled, enable is not allowed */ + if (task_spec_ssb_force_disable(task)) + return -EPERM; + task_clear_spec_ssb_disable(task); + task_clear_spec_ssb_noexec(task); + task_update_spec_tif(task); + break; + case PR_SPEC_DISABLE: + task_set_spec_ssb_disable(task); + task_clear_spec_ssb_noexec(task); + task_update_spec_tif(task); + break; + case PR_SPEC_FORCE_DISABLE: + task_set_spec_ssb_disable(task); + task_set_spec_ssb_force_disable(task); + task_clear_spec_ssb_noexec(task); + task_update_spec_tif(task); + break; + case PR_SPEC_DISABLE_NOEXEC: + if (task_spec_ssb_force_disable(task)) + return -EPERM; + task_set_spec_ssb_disable(task); + task_set_spec_ssb_noexec(task); + task_update_spec_tif(task); + break; + default: + return -ERANGE; + } + return 0; +} + +static bool is_spec_ib_user_controlled(void) +{ + return spectre_v2_user_ibpb == SPECTRE_V2_USER_PRCTL || + spectre_v2_user_ibpb == SPECTRE_V2_USER_SECCOMP || + spectre_v2_user_stibp == SPECTRE_V2_USER_PRCTL || + spectre_v2_user_stibp == SPECTRE_V2_USER_SECCOMP; +} + +static int ib_prctl_set(struct task_struct *task, unsigned long ctrl) +{ + switch (ctrl) { + case PR_SPEC_ENABLE: + if (spectre_v2_user_ibpb == SPECTRE_V2_USER_NONE && + spectre_v2_user_stibp == SPECTRE_V2_USER_NONE) + return 0; + + /* + * With strict mode for both IBPB and STIBP, the instruction + * code paths avoid checking this task flag and instead, + * unconditionally run the instruction. However, STIBP and IBPB + * are independent and either can be set to conditionally + * enabled regardless of the mode of the other. + * + * If either is set to conditional, allow the task flag to be + * updated, unless it was force-disabled by a previous prctl + * call. Currently, this is possible on an AMD CPU which has the + * feature X86_FEATURE_AMD_STIBP_ALWAYS_ON. In this case, if the + * kernel is booted with 'spectre_v2_user=seccomp', then + * spectre_v2_user_ibpb == SPECTRE_V2_USER_SECCOMP and + * spectre_v2_user_stibp == SPECTRE_V2_USER_STRICT_PREFERRED. + */ + if (!is_spec_ib_user_controlled() || + task_spec_ib_force_disable(task)) + return -EPERM; + + task_clear_spec_ib_disable(task); + task_update_spec_tif(task); + break; + case PR_SPEC_DISABLE: + case PR_SPEC_FORCE_DISABLE: + /* + * Indirect branch speculation is always allowed when + * mitigation is force disabled. + */ + if (spectre_v2_user_ibpb == SPECTRE_V2_USER_NONE && + spectre_v2_user_stibp == SPECTRE_V2_USER_NONE) + return -EPERM; + + if (!is_spec_ib_user_controlled()) + return 0; + + task_set_spec_ib_disable(task); + if (ctrl == PR_SPEC_FORCE_DISABLE) + task_set_spec_ib_force_disable(task); + task_update_spec_tif(task); + if (task == current) + indirect_branch_prediction_barrier(); + break; + default: + return -ERANGE; + } + return 0; +} + +int arch_prctl_spec_ctrl_set(struct task_struct *task, unsigned long which, + unsigned long ctrl) +{ + switch (which) { + case PR_SPEC_STORE_BYPASS: + return ssb_prctl_set(task, ctrl); + case PR_SPEC_INDIRECT_BRANCH: + return ib_prctl_set(task, ctrl); + default: + return -ENODEV; + } +} + +#ifdef CONFIG_SECCOMP +void arch_seccomp_spec_mitigate(struct task_struct *task) +{ + if (ssb_mode == SPEC_STORE_BYPASS_SECCOMP) + ssb_prctl_set(task, PR_SPEC_FORCE_DISABLE); + if (spectre_v2_user_ibpb == SPECTRE_V2_USER_SECCOMP || + spectre_v2_user_stibp == SPECTRE_V2_USER_SECCOMP) + ib_prctl_set(task, PR_SPEC_FORCE_DISABLE); +} +#endif + +static int ssb_prctl_get(struct task_struct *task) +{ + switch (ssb_mode) { + case SPEC_STORE_BYPASS_DISABLE: + return PR_SPEC_DISABLE; + case SPEC_STORE_BYPASS_SECCOMP: + case SPEC_STORE_BYPASS_PRCTL: + if (task_spec_ssb_force_disable(task)) + return PR_SPEC_PRCTL | PR_SPEC_FORCE_DISABLE; + if (task_spec_ssb_noexec(task)) + return PR_SPEC_PRCTL | PR_SPEC_DISABLE_NOEXEC; + if (task_spec_ssb_disable(task)) + return PR_SPEC_PRCTL | PR_SPEC_DISABLE; + return PR_SPEC_PRCTL | PR_SPEC_ENABLE; + default: + if (boot_cpu_has_bug(X86_BUG_SPEC_STORE_BYPASS)) + return PR_SPEC_ENABLE; + return PR_SPEC_NOT_AFFECTED; + } +} + +static int ib_prctl_get(struct task_struct *task) +{ + if (!boot_cpu_has_bug(X86_BUG_SPECTRE_V2)) + return PR_SPEC_NOT_AFFECTED; + + if (spectre_v2_user_ibpb == SPECTRE_V2_USER_NONE && + spectre_v2_user_stibp == SPECTRE_V2_USER_NONE) + return PR_SPEC_ENABLE; + else if (is_spec_ib_user_controlled()) { + if (task_spec_ib_force_disable(task)) + return PR_SPEC_PRCTL | PR_SPEC_FORCE_DISABLE; + if (task_spec_ib_disable(task)) + return PR_SPEC_PRCTL | PR_SPEC_DISABLE; + return PR_SPEC_PRCTL | PR_SPEC_ENABLE; + } else if (spectre_v2_user_ibpb == SPECTRE_V2_USER_STRICT || + spectre_v2_user_stibp == SPECTRE_V2_USER_STRICT || + spectre_v2_user_stibp == SPECTRE_V2_USER_STRICT_PREFERRED) + return PR_SPEC_DISABLE; + else + return PR_SPEC_NOT_AFFECTED; +} + +int arch_prctl_spec_ctrl_get(struct task_struct *task, unsigned long which) +{ + switch (which) { + case PR_SPEC_STORE_BYPASS: + return ssb_prctl_get(task); + case PR_SPEC_INDIRECT_BRANCH: + return ib_prctl_get(task); + default: + return -ENODEV; + } +} + +void x86_spec_ctrl_setup_ap(void) +{ + if (boot_cpu_has(X86_FEATURE_MSR_SPEC_CTRL)) + update_spec_ctrl(x86_spec_ctrl_base); + + if (ssb_mode == SPEC_STORE_BYPASS_DISABLE) + x86_amd_ssb_disable(); +} + +bool itlb_multihit_kvm_mitigation; +EXPORT_SYMBOL_GPL(itlb_multihit_kvm_mitigation); + +#undef pr_fmt +#define pr_fmt(fmt) "L1TF: " fmt + +/* Default mitigation for L1TF-affected CPUs */ +enum l1tf_mitigations l1tf_mitigation __ro_after_init = L1TF_MITIGATION_FLUSH; +#if IS_ENABLED(CONFIG_KVM_INTEL) +EXPORT_SYMBOL_GPL(l1tf_mitigation); +#endif +enum vmx_l1d_flush_state l1tf_vmx_mitigation = VMENTER_L1D_FLUSH_AUTO; +EXPORT_SYMBOL_GPL(l1tf_vmx_mitigation); + +/* + * These CPUs all support 44bits physical address space internally in the + * cache but CPUID can report a smaller number of physical address bits. + * + * The L1TF mitigation uses the top most address bit for the inversion of + * non present PTEs. When the installed memory reaches into the top most + * address bit due to memory holes, which has been observed on machines + * which report 36bits physical address bits and have 32G RAM installed, + * then the mitigation range check in l1tf_select_mitigation() triggers. + * This is a false positive because the mitigation is still possible due to + * the fact that the cache uses 44bit internally. Use the cache bits + * instead of the reported physical bits and adjust them on the affected + * machines to 44bit if the reported bits are less than 44. + */ +static void override_cache_bits(struct cpuinfo_x86 *c) +{ + if (c->x86 != 6) + return; + + switch (c->x86_model) { + case INTEL_FAM6_NEHALEM: + case INTEL_FAM6_WESTMERE: + case INTEL_FAM6_SANDYBRIDGE: + case INTEL_FAM6_IVYBRIDGE: + case INTEL_FAM6_HASWELL: + case INTEL_FAM6_HASWELL_L: + case INTEL_FAM6_HASWELL_G: + case INTEL_FAM6_BROADWELL: + case INTEL_FAM6_BROADWELL_G: + case INTEL_FAM6_SKYLAKE_L: + case INTEL_FAM6_SKYLAKE: + case INTEL_FAM6_KABYLAKE_L: + case INTEL_FAM6_KABYLAKE: + if (c->x86_cache_bits < 44) + c->x86_cache_bits = 44; + break; + } +} + +static void __init l1tf_select_mitigation(void) +{ + u64 half_pa; + + if (!boot_cpu_has_bug(X86_BUG_L1TF)) + return; + + if (cpu_mitigations_off()) + l1tf_mitigation = L1TF_MITIGATION_OFF; + else if (cpu_mitigations_auto_nosmt()) + l1tf_mitigation = L1TF_MITIGATION_FLUSH_NOSMT; + + override_cache_bits(&boot_cpu_data); + + switch (l1tf_mitigation) { + case L1TF_MITIGATION_OFF: + case L1TF_MITIGATION_FLUSH_NOWARN: + case L1TF_MITIGATION_FLUSH: + break; + case L1TF_MITIGATION_FLUSH_NOSMT: + case L1TF_MITIGATION_FULL: + cpu_smt_disable(false); + break; + case L1TF_MITIGATION_FULL_FORCE: + cpu_smt_disable(true); + break; + } + +#if CONFIG_PGTABLE_LEVELS == 2 + pr_warn("Kernel not compiled for PAE. No mitigation for L1TF\n"); + return; +#endif + + half_pa = (u64)l1tf_pfn_limit() << PAGE_SHIFT; + if (l1tf_mitigation != L1TF_MITIGATION_OFF && + e820__mapped_any(half_pa, ULLONG_MAX - half_pa, E820_TYPE_RAM)) { + pr_warn("System has more than MAX_PA/2 memory. L1TF mitigation not effective.\n"); + pr_info("You may make it effective by booting the kernel with mem=%llu parameter.\n", + half_pa); + pr_info("However, doing so will make a part of your RAM unusable.\n"); + pr_info("Reading https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/l1tf.html might help you decide.\n"); + return; + } + + setup_force_cpu_cap(X86_FEATURE_L1TF_PTEINV); +} + +static int __init l1tf_cmdline(char *str) +{ + if (!boot_cpu_has_bug(X86_BUG_L1TF)) + return 0; + + if (!str) + return -EINVAL; + + if (!strcmp(str, "off")) + l1tf_mitigation = L1TF_MITIGATION_OFF; + else if (!strcmp(str, "flush,nowarn")) + l1tf_mitigation = L1TF_MITIGATION_FLUSH_NOWARN; + else if (!strcmp(str, "flush")) + l1tf_mitigation = L1TF_MITIGATION_FLUSH; + else if (!strcmp(str, "flush,nosmt")) + l1tf_mitigation = L1TF_MITIGATION_FLUSH_NOSMT; + else if (!strcmp(str, "full")) + l1tf_mitigation = L1TF_MITIGATION_FULL; + else if (!strcmp(str, "full,force")) + l1tf_mitigation = L1TF_MITIGATION_FULL_FORCE; + + return 0; +} +early_param("l1tf", l1tf_cmdline); + +#undef pr_fmt +#define pr_fmt(fmt) "Speculative Return Stack Overflow: " fmt + +enum srso_mitigation { + SRSO_MITIGATION_NONE, + SRSO_MITIGATION_MICROCODE, + SRSO_MITIGATION_SAFE_RET, + SRSO_MITIGATION_IBPB, + SRSO_MITIGATION_IBPB_ON_VMEXIT, +}; + +enum srso_mitigation_cmd { + SRSO_CMD_OFF, + SRSO_CMD_MICROCODE, + SRSO_CMD_SAFE_RET, + SRSO_CMD_IBPB, + SRSO_CMD_IBPB_ON_VMEXIT, +}; + +static const char * const srso_strings[] = { + [SRSO_MITIGATION_NONE] = "Vulnerable", + [SRSO_MITIGATION_MICROCODE] = "Mitigation: microcode", + [SRSO_MITIGATION_SAFE_RET] = "Mitigation: safe RET", + [SRSO_MITIGATION_IBPB] = "Mitigation: IBPB", + [SRSO_MITIGATION_IBPB_ON_VMEXIT] = "Mitigation: IBPB on VMEXIT only" +}; + +static enum srso_mitigation srso_mitigation __ro_after_init = SRSO_MITIGATION_NONE; +static enum srso_mitigation_cmd srso_cmd __ro_after_init = SRSO_CMD_SAFE_RET; + +static int __init srso_parse_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + if (!strcmp(str, "off")) + srso_cmd = SRSO_CMD_OFF; + else if (!strcmp(str, "microcode")) + srso_cmd = SRSO_CMD_MICROCODE; + else if (!strcmp(str, "safe-ret")) + srso_cmd = SRSO_CMD_SAFE_RET; + else if (!strcmp(str, "ibpb")) + srso_cmd = SRSO_CMD_IBPB; + else if (!strcmp(str, "ibpb-vmexit")) + srso_cmd = SRSO_CMD_IBPB_ON_VMEXIT; + else + pr_err("Ignoring unknown SRSO option (%s).", str); + + return 0; +} +early_param("spec_rstack_overflow", srso_parse_cmdline); + +#define SRSO_NOTICE "WARNING: See https://kernel.org/doc/html/latest/admin-guide/hw-vuln/srso.html for mitigation options." + +static void __init srso_select_mitigation(void) +{ + bool has_microcode; + + if (!boot_cpu_has_bug(X86_BUG_SRSO) || cpu_mitigations_off()) + goto pred_cmd; + + /* + * The first check is for the kernel running as a guest in order + * for guests to verify whether IBPB is a viable mitigation. + */ + has_microcode = boot_cpu_has(X86_FEATURE_IBPB_BRTYPE) || cpu_has_ibpb_brtype_microcode(); + if (!has_microcode) { + pr_warn("IBPB-extending microcode not applied!\n"); + pr_warn(SRSO_NOTICE); + } else { + /* + * Enable the synthetic (even if in a real CPUID leaf) + * flags for guests. + */ + setup_force_cpu_cap(X86_FEATURE_IBPB_BRTYPE); + + /* + * Zen1/2 with SMT off aren't vulnerable after the right + * IBPB microcode has been applied. + */ + if (boot_cpu_data.x86 < 0x19 && !cpu_smt_possible()) { + setup_force_cpu_cap(X86_FEATURE_SRSO_NO); + return; + } + } + + if (retbleed_mitigation == RETBLEED_MITIGATION_IBPB) { + if (has_microcode) { + pr_err("Retbleed IBPB mitigation enabled, using same for SRSO\n"); + srso_mitigation = SRSO_MITIGATION_IBPB; + goto pred_cmd; + } + } + + switch (srso_cmd) { + case SRSO_CMD_OFF: + goto pred_cmd; + + case SRSO_CMD_MICROCODE: + if (has_microcode) { + srso_mitigation = SRSO_MITIGATION_MICROCODE; + pr_warn(SRSO_NOTICE); + } + break; + + case SRSO_CMD_SAFE_RET: + if (IS_ENABLED(CONFIG_CPU_SRSO)) { + /* + * Enable the return thunk for generated code + * like ftrace, static_call, etc. + */ + setup_force_cpu_cap(X86_FEATURE_RETHUNK); + setup_force_cpu_cap(X86_FEATURE_UNRET); + + if (boot_cpu_data.x86 == 0x19) { + setup_force_cpu_cap(X86_FEATURE_SRSO_ALIAS); + x86_return_thunk = srso_alias_return_thunk; + } else { + setup_force_cpu_cap(X86_FEATURE_SRSO); + x86_return_thunk = srso_return_thunk; + } + srso_mitigation = SRSO_MITIGATION_SAFE_RET; + } else { + pr_err("WARNING: kernel not compiled with CPU_SRSO.\n"); + goto pred_cmd; + } + break; + + case SRSO_CMD_IBPB: + if (IS_ENABLED(CONFIG_CPU_IBPB_ENTRY)) { + if (has_microcode) { + setup_force_cpu_cap(X86_FEATURE_ENTRY_IBPB); + srso_mitigation = SRSO_MITIGATION_IBPB; + } + } else { + pr_err("WARNING: kernel not compiled with CPU_IBPB_ENTRY.\n"); + goto pred_cmd; + } + break; + + case SRSO_CMD_IBPB_ON_VMEXIT: + if (IS_ENABLED(CONFIG_CPU_SRSO)) { + if (!boot_cpu_has(X86_FEATURE_ENTRY_IBPB) && has_microcode) { + setup_force_cpu_cap(X86_FEATURE_IBPB_ON_VMEXIT); + srso_mitigation = SRSO_MITIGATION_IBPB_ON_VMEXIT; + } + } else { + pr_err("WARNING: kernel not compiled with CPU_SRSO.\n"); + goto pred_cmd; + } + break; + + default: + break; + } + + pr_info("%s%s\n", srso_strings[srso_mitigation], (has_microcode ? "" : ", no microcode")); + +pred_cmd: + if ((!boot_cpu_has_bug(X86_BUG_SRSO) || srso_cmd == SRSO_CMD_OFF) && + boot_cpu_has(X86_FEATURE_SBPB)) + x86_pred_cmd = PRED_CMD_SBPB; +} + +#undef pr_fmt +#define pr_fmt(fmt) fmt + +#ifdef CONFIG_SYSFS + +#define L1TF_DEFAULT_MSG "Mitigation: PTE Inversion" + +#if IS_ENABLED(CONFIG_KVM_INTEL) +static const char * const l1tf_vmx_states[] = { + [VMENTER_L1D_FLUSH_AUTO] = "auto", + [VMENTER_L1D_FLUSH_NEVER] = "vulnerable", + [VMENTER_L1D_FLUSH_COND] = "conditional cache flushes", + [VMENTER_L1D_FLUSH_ALWAYS] = "cache flushes", + [VMENTER_L1D_FLUSH_EPT_DISABLED] = "EPT disabled", + [VMENTER_L1D_FLUSH_NOT_REQUIRED] = "flush not necessary" +}; + +static ssize_t l1tf_show_state(char *buf) +{ + if (l1tf_vmx_mitigation == VMENTER_L1D_FLUSH_AUTO) + return sprintf(buf, "%s\n", L1TF_DEFAULT_MSG); + + if (l1tf_vmx_mitigation == VMENTER_L1D_FLUSH_EPT_DISABLED || + (l1tf_vmx_mitigation == VMENTER_L1D_FLUSH_NEVER && + sched_smt_active())) { + return sprintf(buf, "%s; VMX: %s\n", L1TF_DEFAULT_MSG, + l1tf_vmx_states[l1tf_vmx_mitigation]); + } + + return sprintf(buf, "%s; VMX: %s, SMT %s\n", L1TF_DEFAULT_MSG, + l1tf_vmx_states[l1tf_vmx_mitigation], + sched_smt_active() ? "vulnerable" : "disabled"); +} + +static ssize_t itlb_multihit_show_state(char *buf) +{ + if (!boot_cpu_has(X86_FEATURE_MSR_IA32_FEAT_CTL) || + !boot_cpu_has(X86_FEATURE_VMX)) + return sprintf(buf, "KVM: Mitigation: VMX unsupported\n"); + else if (!(cr4_read_shadow() & X86_CR4_VMXE)) + return sprintf(buf, "KVM: Mitigation: VMX disabled\n"); + else if (itlb_multihit_kvm_mitigation) + return sprintf(buf, "KVM: Mitigation: Split huge pages\n"); + else + return sprintf(buf, "KVM: Vulnerable\n"); +} +#else +static ssize_t l1tf_show_state(char *buf) +{ + return sprintf(buf, "%s\n", L1TF_DEFAULT_MSG); +} + +static ssize_t itlb_multihit_show_state(char *buf) +{ + return sprintf(buf, "Processor vulnerable\n"); +} +#endif + +static ssize_t mds_show_state(char *buf) +{ + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) { + return sprintf(buf, "%s; SMT Host state unknown\n", + mds_strings[mds_mitigation]); + } + + if (boot_cpu_has(X86_BUG_MSBDS_ONLY)) { + return sprintf(buf, "%s; SMT %s\n", mds_strings[mds_mitigation], + (mds_mitigation == MDS_MITIGATION_OFF ? "vulnerable" : + sched_smt_active() ? "mitigated" : "disabled")); + } + + return sprintf(buf, "%s; SMT %s\n", mds_strings[mds_mitigation], + sched_smt_active() ? "vulnerable" : "disabled"); +} + +static ssize_t tsx_async_abort_show_state(char *buf) +{ + if ((taa_mitigation == TAA_MITIGATION_TSX_DISABLED) || + (taa_mitigation == TAA_MITIGATION_OFF)) + return sprintf(buf, "%s\n", taa_strings[taa_mitigation]); + + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) { + return sprintf(buf, "%s; SMT Host state unknown\n", + taa_strings[taa_mitigation]); + } + + return sprintf(buf, "%s; SMT %s\n", taa_strings[taa_mitigation], + sched_smt_active() ? "vulnerable" : "disabled"); +} + +static ssize_t mmio_stale_data_show_state(char *buf) +{ + if (boot_cpu_has_bug(X86_BUG_MMIO_UNKNOWN)) + return sysfs_emit(buf, "Unknown: No mitigations\n"); + + if (mmio_mitigation == MMIO_MITIGATION_OFF) + return sysfs_emit(buf, "%s\n", mmio_strings[mmio_mitigation]); + + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) { + return sysfs_emit(buf, "%s; SMT Host state unknown\n", + mmio_strings[mmio_mitigation]); + } + + return sysfs_emit(buf, "%s; SMT %s\n", mmio_strings[mmio_mitigation], + sched_smt_active() ? "vulnerable" : "disabled"); +} + +static char *stibp_state(void) +{ + if (spectre_v2_in_eibrs_mode(spectre_v2_enabled)) + return ""; + + switch (spectre_v2_user_stibp) { + case SPECTRE_V2_USER_NONE: + return ", STIBP: disabled"; + case SPECTRE_V2_USER_STRICT: + return ", STIBP: forced"; + case SPECTRE_V2_USER_STRICT_PREFERRED: + return ", STIBP: always-on"; + case SPECTRE_V2_USER_PRCTL: + case SPECTRE_V2_USER_SECCOMP: + if (static_key_enabled(&switch_to_cond_stibp)) + return ", STIBP: conditional"; + } + return ""; +} + +static char *ibpb_state(void) +{ + if (boot_cpu_has(X86_FEATURE_IBPB)) { + if (static_key_enabled(&switch_mm_always_ibpb)) + return ", IBPB: always-on"; + if (static_key_enabled(&switch_mm_cond_ibpb)) + return ", IBPB: conditional"; + return ", IBPB: disabled"; + } + return ""; +} + +static char *pbrsb_eibrs_state(void) +{ + if (boot_cpu_has_bug(X86_BUG_EIBRS_PBRSB)) { + if (boot_cpu_has(X86_FEATURE_RSB_VMEXIT_LITE) || + boot_cpu_has(X86_FEATURE_RSB_VMEXIT)) + return ", PBRSB-eIBRS: SW sequence"; + else + return ", PBRSB-eIBRS: Vulnerable"; + } else { + return ", PBRSB-eIBRS: Not affected"; + } +} + +static ssize_t spectre_v2_show_state(char *buf) +{ + if (spectre_v2_enabled == SPECTRE_V2_LFENCE) + return sprintf(buf, "Vulnerable: LFENCE\n"); + + if (spectre_v2_enabled == SPECTRE_V2_EIBRS && unprivileged_ebpf_enabled()) + return sprintf(buf, "Vulnerable: eIBRS with unprivileged eBPF\n"); + + if (sched_smt_active() && unprivileged_ebpf_enabled() && + spectre_v2_enabled == SPECTRE_V2_EIBRS_LFENCE) + return sprintf(buf, "Vulnerable: eIBRS+LFENCE with unprivileged eBPF and SMT\n"); + + return sprintf(buf, "%s%s%s%s%s%s%s\n", + spectre_v2_strings[spectre_v2_enabled], + ibpb_state(), + boot_cpu_has(X86_FEATURE_USE_IBRS_FW) ? ", IBRS_FW" : "", + stibp_state(), + boot_cpu_has(X86_FEATURE_RSB_CTXSW) ? ", RSB filling" : "", + pbrsb_eibrs_state(), + spectre_v2_module_string()); +} + +static ssize_t srbds_show_state(char *buf) +{ + return sprintf(buf, "%s\n", srbds_strings[srbds_mitigation]); +} + +static ssize_t retbleed_show_state(char *buf) +{ + if (retbleed_mitigation == RETBLEED_MITIGATION_UNRET || + retbleed_mitigation == RETBLEED_MITIGATION_IBPB) { + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD && + boot_cpu_data.x86_vendor != X86_VENDOR_HYGON) + return sprintf(buf, "Vulnerable: untrained return thunk / IBPB on non-AMD based uarch\n"); + + return sprintf(buf, "%s; SMT %s\n", + retbleed_strings[retbleed_mitigation], + !sched_smt_active() ? "disabled" : + spectre_v2_user_stibp == SPECTRE_V2_USER_STRICT || + spectre_v2_user_stibp == SPECTRE_V2_USER_STRICT_PREFERRED ? + "enabled with STIBP protection" : "vulnerable"); + } + + return sprintf(buf, "%s\n", retbleed_strings[retbleed_mitigation]); +} + +static ssize_t gds_show_state(char *buf) +{ + return sysfs_emit(buf, "%s\n", gds_strings[gds_mitigation]); +} + +static ssize_t srso_show_state(char *buf) +{ + if (boot_cpu_has(X86_FEATURE_SRSO_NO)) + return sysfs_emit(buf, "Mitigation: SMT disabled\n"); + + return sysfs_emit(buf, "%s%s\n", + srso_strings[srso_mitigation], + boot_cpu_has(X86_FEATURE_IBPB_BRTYPE) ? "" : ", no microcode"); +} + +static ssize_t cpu_show_common(struct device *dev, struct device_attribute *attr, + char *buf, unsigned int bug) +{ + if (!boot_cpu_has_bug(bug)) + return sprintf(buf, "Not affected\n"); + + switch (bug) { + case X86_BUG_CPU_MELTDOWN: + if (boot_cpu_has(X86_FEATURE_PTI)) + return sprintf(buf, "Mitigation: PTI\n"); + + if (hypervisor_is_type(X86_HYPER_XEN_PV)) + return sprintf(buf, "Unknown (XEN PV detected, hypervisor mitigation required)\n"); + + break; + + case X86_BUG_SPECTRE_V1: + return sprintf(buf, "%s\n", spectre_v1_strings[spectre_v1_mitigation]); + + case X86_BUG_SPECTRE_V2: + return spectre_v2_show_state(buf); + + case X86_BUG_SPEC_STORE_BYPASS: + return sprintf(buf, "%s\n", ssb_strings[ssb_mode]); + + case X86_BUG_L1TF: + if (boot_cpu_has(X86_FEATURE_L1TF_PTEINV)) + return l1tf_show_state(buf); + break; + + case X86_BUG_MDS: + return mds_show_state(buf); + + case X86_BUG_TAA: + return tsx_async_abort_show_state(buf); + + case X86_BUG_ITLB_MULTIHIT: + return itlb_multihit_show_state(buf); + + case X86_BUG_SRBDS: + return srbds_show_state(buf); + + case X86_BUG_MMIO_STALE_DATA: + case X86_BUG_MMIO_UNKNOWN: + return mmio_stale_data_show_state(buf); + + case X86_BUG_RETBLEED: + return retbleed_show_state(buf); + + case X86_BUG_GDS: + return gds_show_state(buf); + + case X86_BUG_SRSO: + return srso_show_state(buf); + + default: + break; + } + + return sprintf(buf, "Vulnerable\n"); +} + +ssize_t cpu_show_meltdown(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_CPU_MELTDOWN); +} + +ssize_t cpu_show_spectre_v1(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_SPECTRE_V1); +} + +ssize_t cpu_show_spectre_v2(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_SPECTRE_V2); +} + +ssize_t cpu_show_spec_store_bypass(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_SPEC_STORE_BYPASS); +} + +ssize_t cpu_show_l1tf(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_L1TF); +} + +ssize_t cpu_show_mds(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_MDS); +} + +ssize_t cpu_show_tsx_async_abort(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_TAA); +} + +ssize_t cpu_show_itlb_multihit(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_ITLB_MULTIHIT); +} + +ssize_t cpu_show_srbds(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_SRBDS); +} + +ssize_t cpu_show_mmio_stale_data(struct device *dev, struct device_attribute *attr, char *buf) +{ + if (boot_cpu_has_bug(X86_BUG_MMIO_UNKNOWN)) + return cpu_show_common(dev, attr, buf, X86_BUG_MMIO_UNKNOWN); + else + return cpu_show_common(dev, attr, buf, X86_BUG_MMIO_STALE_DATA); +} + +ssize_t cpu_show_retbleed(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_RETBLEED); +} + +ssize_t cpu_show_gds(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_GDS); +} + +ssize_t cpu_show_spec_rstack_overflow(struct device *dev, struct device_attribute *attr, char *buf) +{ + return cpu_show_common(dev, attr, buf, X86_BUG_SRSO); +} +#endif diff --git a/arch/x86/kernel/cpu/cacheinfo.c b/arch/x86/kernel/cpu/cacheinfo.c new file mode 100644 index 000000000..b458b0fd9 --- /dev/null +++ b/arch/x86/kernel/cpu/cacheinfo.c @@ -0,0 +1,1035 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Routines to identify caches on Intel CPU. + * + * Changes: + * Venkatesh Pallipadi : Adding cache identification through cpuid(4) + * Ashok Raj <ashok.raj@intel.com>: Work with CPU hotplug infrastructure. + * Andi Kleen / Andreas Herrmann : CPUID4 emulation on AMD. + */ + +#include <linux/slab.h> +#include <linux/cacheinfo.h> +#include <linux/cpu.h> +#include <linux/sched.h> +#include <linux/capability.h> +#include <linux/sysfs.h> +#include <linux/pci.h> + +#include <asm/cpufeature.h> +#include <asm/cacheinfo.h> +#include <asm/amd_nb.h> +#include <asm/smp.h> + +#include "cpu.h" + +#define LVL_1_INST 1 +#define LVL_1_DATA 2 +#define LVL_2 3 +#define LVL_3 4 +#define LVL_TRACE 5 + +struct _cache_table { + unsigned char descriptor; + char cache_type; + short size; +}; + +#define MB(x) ((x) * 1024) + +/* All the cache descriptor types we care about (no TLB or + trace cache entries) */ + +static const struct _cache_table cache_table[] = +{ + { 0x06, LVL_1_INST, 8 }, /* 4-way set assoc, 32 byte line size */ + { 0x08, LVL_1_INST, 16 }, /* 4-way set assoc, 32 byte line size */ + { 0x09, LVL_1_INST, 32 }, /* 4-way set assoc, 64 byte line size */ + { 0x0a, LVL_1_DATA, 8 }, /* 2 way set assoc, 32 byte line size */ + { 0x0c, LVL_1_DATA, 16 }, /* 4-way set assoc, 32 byte line size */ + { 0x0d, LVL_1_DATA, 16 }, /* 4-way set assoc, 64 byte line size */ + { 0x0e, LVL_1_DATA, 24 }, /* 6-way set assoc, 64 byte line size */ + { 0x21, LVL_2, 256 }, /* 8-way set assoc, 64 byte line size */ + { 0x22, LVL_3, 512 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x23, LVL_3, MB(1) }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x25, LVL_3, MB(2) }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x29, LVL_3, MB(4) }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x2c, LVL_1_DATA, 32 }, /* 8-way set assoc, 64 byte line size */ + { 0x30, LVL_1_INST, 32 }, /* 8-way set assoc, 64 byte line size */ + { 0x39, LVL_2, 128 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x3a, LVL_2, 192 }, /* 6-way set assoc, sectored cache, 64 byte line size */ + { 0x3b, LVL_2, 128 }, /* 2-way set assoc, sectored cache, 64 byte line size */ + { 0x3c, LVL_2, 256 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x3d, LVL_2, 384 }, /* 6-way set assoc, sectored cache, 64 byte line size */ + { 0x3e, LVL_2, 512 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x3f, LVL_2, 256 }, /* 2-way set assoc, 64 byte line size */ + { 0x41, LVL_2, 128 }, /* 4-way set assoc, 32 byte line size */ + { 0x42, LVL_2, 256 }, /* 4-way set assoc, 32 byte line size */ + { 0x43, LVL_2, 512 }, /* 4-way set assoc, 32 byte line size */ + { 0x44, LVL_2, MB(1) }, /* 4-way set assoc, 32 byte line size */ + { 0x45, LVL_2, MB(2) }, /* 4-way set assoc, 32 byte line size */ + { 0x46, LVL_3, MB(4) }, /* 4-way set assoc, 64 byte line size */ + { 0x47, LVL_3, MB(8) }, /* 8-way set assoc, 64 byte line size */ + { 0x48, LVL_2, MB(3) }, /* 12-way set assoc, 64 byte line size */ + { 0x49, LVL_3, MB(4) }, /* 16-way set assoc, 64 byte line size */ + { 0x4a, LVL_3, MB(6) }, /* 12-way set assoc, 64 byte line size */ + { 0x4b, LVL_3, MB(8) }, /* 16-way set assoc, 64 byte line size */ + { 0x4c, LVL_3, MB(12) }, /* 12-way set assoc, 64 byte line size */ + { 0x4d, LVL_3, MB(16) }, /* 16-way set assoc, 64 byte line size */ + { 0x4e, LVL_2, MB(6) }, /* 24-way set assoc, 64 byte line size */ + { 0x60, LVL_1_DATA, 16 }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x66, LVL_1_DATA, 8 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x67, LVL_1_DATA, 16 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x68, LVL_1_DATA, 32 }, /* 4-way set assoc, sectored cache, 64 byte line size */ + { 0x70, LVL_TRACE, 12 }, /* 8-way set assoc */ + { 0x71, LVL_TRACE, 16 }, /* 8-way set assoc */ + { 0x72, LVL_TRACE, 32 }, /* 8-way set assoc */ + { 0x73, LVL_TRACE, 64 }, /* 8-way set assoc */ + { 0x78, LVL_2, MB(1) }, /* 4-way set assoc, 64 byte line size */ + { 0x79, LVL_2, 128 }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x7a, LVL_2, 256 }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x7b, LVL_2, 512 }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x7c, LVL_2, MB(1) }, /* 8-way set assoc, sectored cache, 64 byte line size */ + { 0x7d, LVL_2, MB(2) }, /* 8-way set assoc, 64 byte line size */ + { 0x7f, LVL_2, 512 }, /* 2-way set assoc, 64 byte line size */ + { 0x80, LVL_2, 512 }, /* 8-way set assoc, 64 byte line size */ + { 0x82, LVL_2, 256 }, /* 8-way set assoc, 32 byte line size */ + { 0x83, LVL_2, 512 }, /* 8-way set assoc, 32 byte line size */ + { 0x84, LVL_2, MB(1) }, /* 8-way set assoc, 32 byte line size */ + { 0x85, LVL_2, MB(2) }, /* 8-way set assoc, 32 byte line size */ + { 0x86, LVL_2, 512 }, /* 4-way set assoc, 64 byte line size */ + { 0x87, LVL_2, MB(1) }, /* 8-way set assoc, 64 byte line size */ + { 0xd0, LVL_3, 512 }, /* 4-way set assoc, 64 byte line size */ + { 0xd1, LVL_3, MB(1) }, /* 4-way set assoc, 64 byte line size */ + { 0xd2, LVL_3, MB(2) }, /* 4-way set assoc, 64 byte line size */ + { 0xd6, LVL_3, MB(1) }, /* 8-way set assoc, 64 byte line size */ + { 0xd7, LVL_3, MB(2) }, /* 8-way set assoc, 64 byte line size */ + { 0xd8, LVL_3, MB(4) }, /* 12-way set assoc, 64 byte line size */ + { 0xdc, LVL_3, MB(2) }, /* 12-way set assoc, 64 byte line size */ + { 0xdd, LVL_3, MB(4) }, /* 12-way set assoc, 64 byte line size */ + { 0xde, LVL_3, MB(8) }, /* 12-way set assoc, 64 byte line size */ + { 0xe2, LVL_3, MB(2) }, /* 16-way set assoc, 64 byte line size */ + { 0xe3, LVL_3, MB(4) }, /* 16-way set assoc, 64 byte line size */ + { 0xe4, LVL_3, MB(8) }, /* 16-way set assoc, 64 byte line size */ + { 0xea, LVL_3, MB(12) }, /* 24-way set assoc, 64 byte line size */ + { 0xeb, LVL_3, MB(18) }, /* 24-way set assoc, 64 byte line size */ + { 0xec, LVL_3, MB(24) }, /* 24-way set assoc, 64 byte line size */ + { 0x00, 0, 0} +}; + + +enum _cache_type { + CTYPE_NULL = 0, + CTYPE_DATA = 1, + CTYPE_INST = 2, + CTYPE_UNIFIED = 3 +}; + +union _cpuid4_leaf_eax { + struct { + enum _cache_type type:5; + unsigned int level:3; + unsigned int is_self_initializing:1; + unsigned int is_fully_associative:1; + unsigned int reserved:4; + unsigned int num_threads_sharing:12; + unsigned int num_cores_on_die:6; + } split; + u32 full; +}; + +union _cpuid4_leaf_ebx { + struct { + unsigned int coherency_line_size:12; + unsigned int physical_line_partition:10; + unsigned int ways_of_associativity:10; + } split; + u32 full; +}; + +union _cpuid4_leaf_ecx { + struct { + unsigned int number_of_sets:32; + } split; + u32 full; +}; + +struct _cpuid4_info_regs { + union _cpuid4_leaf_eax eax; + union _cpuid4_leaf_ebx ebx; + union _cpuid4_leaf_ecx ecx; + unsigned int id; + unsigned long size; + struct amd_northbridge *nb; +}; + +static unsigned short num_cache_leaves; + +/* AMD doesn't have CPUID4. Emulate it here to report the same + information to the user. This makes some assumptions about the machine: + L2 not shared, no SMT etc. that is currently true on AMD CPUs. + + In theory the TLBs could be reported as fake type (they are in "dummy"). + Maybe later */ +union l1_cache { + struct { + unsigned line_size:8; + unsigned lines_per_tag:8; + unsigned assoc:8; + unsigned size_in_kb:8; + }; + unsigned val; +}; + +union l2_cache { + struct { + unsigned line_size:8; + unsigned lines_per_tag:4; + unsigned assoc:4; + unsigned size_in_kb:16; + }; + unsigned val; +}; + +union l3_cache { + struct { + unsigned line_size:8; + unsigned lines_per_tag:4; + unsigned assoc:4; + unsigned res:2; + unsigned size_encoded:14; + }; + unsigned val; +}; + +static const unsigned short assocs[] = { + [1] = 1, + [2] = 2, + [4] = 4, + [6] = 8, + [8] = 16, + [0xa] = 32, + [0xb] = 48, + [0xc] = 64, + [0xd] = 96, + [0xe] = 128, + [0xf] = 0xffff /* fully associative - no way to show this currently */ +}; + +static const unsigned char levels[] = { 1, 1, 2, 3 }; +static const unsigned char types[] = { 1, 2, 3, 3 }; + +static const enum cache_type cache_type_map[] = { + [CTYPE_NULL] = CACHE_TYPE_NOCACHE, + [CTYPE_DATA] = CACHE_TYPE_DATA, + [CTYPE_INST] = CACHE_TYPE_INST, + [CTYPE_UNIFIED] = CACHE_TYPE_UNIFIED, +}; + +static void +amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax, + union _cpuid4_leaf_ebx *ebx, + union _cpuid4_leaf_ecx *ecx) +{ + unsigned dummy; + unsigned line_size, lines_per_tag, assoc, size_in_kb; + union l1_cache l1i, l1d; + union l2_cache l2; + union l3_cache l3; + union l1_cache *l1 = &l1d; + + eax->full = 0; + ebx->full = 0; + ecx->full = 0; + + cpuid(0x80000005, &dummy, &dummy, &l1d.val, &l1i.val); + cpuid(0x80000006, &dummy, &dummy, &l2.val, &l3.val); + + switch (leaf) { + case 1: + l1 = &l1i; + fallthrough; + case 0: + if (!l1->val) + return; + assoc = assocs[l1->assoc]; + line_size = l1->line_size; + lines_per_tag = l1->lines_per_tag; + size_in_kb = l1->size_in_kb; + break; + case 2: + if (!l2.val) + return; + assoc = assocs[l2.assoc]; + line_size = l2.line_size; + lines_per_tag = l2.lines_per_tag; + /* cpu_data has errata corrections for K7 applied */ + size_in_kb = __this_cpu_read(cpu_info.x86_cache_size); + break; + case 3: + if (!l3.val) + return; + assoc = assocs[l3.assoc]; + line_size = l3.line_size; + lines_per_tag = l3.lines_per_tag; + size_in_kb = l3.size_encoded * 512; + if (boot_cpu_has(X86_FEATURE_AMD_DCM)) { + size_in_kb = size_in_kb >> 1; + assoc = assoc >> 1; + } + break; + default: + return; + } + + eax->split.is_self_initializing = 1; + eax->split.type = types[leaf]; + eax->split.level = levels[leaf]; + eax->split.num_threads_sharing = 0; + eax->split.num_cores_on_die = __this_cpu_read(cpu_info.x86_max_cores) - 1; + + + if (assoc == 0xffff) + eax->split.is_fully_associative = 1; + ebx->split.coherency_line_size = line_size - 1; + ebx->split.ways_of_associativity = assoc - 1; + ebx->split.physical_line_partition = lines_per_tag - 1; + ecx->split.number_of_sets = (size_in_kb * 1024) / line_size / + (ebx->split.ways_of_associativity + 1) - 1; +} + +#if defined(CONFIG_AMD_NB) && defined(CONFIG_SYSFS) + +/* + * L3 cache descriptors + */ +static void amd_calc_l3_indices(struct amd_northbridge *nb) +{ + struct amd_l3_cache *l3 = &nb->l3_cache; + unsigned int sc0, sc1, sc2, sc3; + u32 val = 0; + + pci_read_config_dword(nb->misc, 0x1C4, &val); + + /* calculate subcache sizes */ + l3->subcaches[0] = sc0 = !(val & BIT(0)); + l3->subcaches[1] = sc1 = !(val & BIT(4)); + + if (boot_cpu_data.x86 == 0x15) { + l3->subcaches[0] = sc0 += !(val & BIT(1)); + l3->subcaches[1] = sc1 += !(val & BIT(5)); + } + + l3->subcaches[2] = sc2 = !(val & BIT(8)) + !(val & BIT(9)); + l3->subcaches[3] = sc3 = !(val & BIT(12)) + !(val & BIT(13)); + + l3->indices = (max(max3(sc0, sc1, sc2), sc3) << 10) - 1; +} + +/* + * check whether a slot used for disabling an L3 index is occupied. + * @l3: L3 cache descriptor + * @slot: slot number (0..1) + * + * @returns: the disabled index if used or negative value if slot free. + */ +static int amd_get_l3_disable_slot(struct amd_northbridge *nb, unsigned slot) +{ + unsigned int reg = 0; + + pci_read_config_dword(nb->misc, 0x1BC + slot * 4, ®); + + /* check whether this slot is activated already */ + if (reg & (3UL << 30)) + return reg & 0xfff; + + return -1; +} + +static ssize_t show_cache_disable(struct cacheinfo *this_leaf, char *buf, + unsigned int slot) +{ + int index; + struct amd_northbridge *nb = this_leaf->priv; + + index = amd_get_l3_disable_slot(nb, slot); + if (index >= 0) + return sprintf(buf, "%d\n", index); + + return sprintf(buf, "FREE\n"); +} + +#define SHOW_CACHE_DISABLE(slot) \ +static ssize_t \ +cache_disable_##slot##_show(struct device *dev, \ + struct device_attribute *attr, char *buf) \ +{ \ + struct cacheinfo *this_leaf = dev_get_drvdata(dev); \ + return show_cache_disable(this_leaf, buf, slot); \ +} +SHOW_CACHE_DISABLE(0) +SHOW_CACHE_DISABLE(1) + +static void amd_l3_disable_index(struct amd_northbridge *nb, int cpu, + unsigned slot, unsigned long idx) +{ + int i; + + idx |= BIT(30); + + /* + * disable index in all 4 subcaches + */ + for (i = 0; i < 4; i++) { + u32 reg = idx | (i << 20); + + if (!nb->l3_cache.subcaches[i]) + continue; + + pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg); + + /* + * We need to WBINVD on a core on the node containing the L3 + * cache which indices we disable therefore a simple wbinvd() + * is not sufficient. + */ + wbinvd_on_cpu(cpu); + + reg |= BIT(31); + pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg); + } +} + +/* + * disable a L3 cache index by using a disable-slot + * + * @l3: L3 cache descriptor + * @cpu: A CPU on the node containing the L3 cache + * @slot: slot number (0..1) + * @index: index to disable + * + * @return: 0 on success, error status on failure + */ +static int amd_set_l3_disable_slot(struct amd_northbridge *nb, int cpu, + unsigned slot, unsigned long index) +{ + int ret = 0; + + /* check if @slot is already used or the index is already disabled */ + ret = amd_get_l3_disable_slot(nb, slot); + if (ret >= 0) + return -EEXIST; + + if (index > nb->l3_cache.indices) + return -EINVAL; + + /* check whether the other slot has disabled the same index already */ + if (index == amd_get_l3_disable_slot(nb, !slot)) + return -EEXIST; + + amd_l3_disable_index(nb, cpu, slot, index); + + return 0; +} + +static ssize_t store_cache_disable(struct cacheinfo *this_leaf, + const char *buf, size_t count, + unsigned int slot) +{ + unsigned long val = 0; + int cpu, err = 0; + struct amd_northbridge *nb = this_leaf->priv; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + cpu = cpumask_first(&this_leaf->shared_cpu_map); + + if (kstrtoul(buf, 10, &val) < 0) + return -EINVAL; + + err = amd_set_l3_disable_slot(nb, cpu, slot, val); + if (err) { + if (err == -EEXIST) + pr_warn("L3 slot %d in use/index already disabled!\n", + slot); + return err; + } + return count; +} + +#define STORE_CACHE_DISABLE(slot) \ +static ssize_t \ +cache_disable_##slot##_store(struct device *dev, \ + struct device_attribute *attr, \ + const char *buf, size_t count) \ +{ \ + struct cacheinfo *this_leaf = dev_get_drvdata(dev); \ + return store_cache_disable(this_leaf, buf, count, slot); \ +} +STORE_CACHE_DISABLE(0) +STORE_CACHE_DISABLE(1) + +static ssize_t subcaches_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct cacheinfo *this_leaf = dev_get_drvdata(dev); + int cpu = cpumask_first(&this_leaf->shared_cpu_map); + + return sprintf(buf, "%x\n", amd_get_subcaches(cpu)); +} + +static ssize_t subcaches_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + struct cacheinfo *this_leaf = dev_get_drvdata(dev); + int cpu = cpumask_first(&this_leaf->shared_cpu_map); + unsigned long val; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (kstrtoul(buf, 16, &val) < 0) + return -EINVAL; + + if (amd_set_subcaches(cpu, val)) + return -EINVAL; + + return count; +} + +static DEVICE_ATTR_RW(cache_disable_0); +static DEVICE_ATTR_RW(cache_disable_1); +static DEVICE_ATTR_RW(subcaches); + +static umode_t +cache_private_attrs_is_visible(struct kobject *kobj, + struct attribute *attr, int unused) +{ + struct device *dev = kobj_to_dev(kobj); + struct cacheinfo *this_leaf = dev_get_drvdata(dev); + umode_t mode = attr->mode; + + if (!this_leaf->priv) + return 0; + + if ((attr == &dev_attr_subcaches.attr) && + amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + return mode; + + if ((attr == &dev_attr_cache_disable_0.attr || + attr == &dev_attr_cache_disable_1.attr) && + amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) + return mode; + + return 0; +} + +static struct attribute_group cache_private_group = { + .is_visible = cache_private_attrs_is_visible, +}; + +static void init_amd_l3_attrs(void) +{ + int n = 1; + static struct attribute **amd_l3_attrs; + + if (amd_l3_attrs) /* already initialized */ + return; + + if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) + n += 2; + if (amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + n += 1; + + amd_l3_attrs = kcalloc(n, sizeof(*amd_l3_attrs), GFP_KERNEL); + if (!amd_l3_attrs) + return; + + n = 0; + if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) { + amd_l3_attrs[n++] = &dev_attr_cache_disable_0.attr; + amd_l3_attrs[n++] = &dev_attr_cache_disable_1.attr; + } + if (amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + amd_l3_attrs[n++] = &dev_attr_subcaches.attr; + + cache_private_group.attrs = amd_l3_attrs; +} + +const struct attribute_group * +cache_get_priv_group(struct cacheinfo *this_leaf) +{ + struct amd_northbridge *nb = this_leaf->priv; + + if (this_leaf->level < 3 || !nb) + return NULL; + + if (nb && nb->l3_cache.indices) + init_amd_l3_attrs(); + + return &cache_private_group; +} + +static void amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf, int index) +{ + int node; + + /* only for L3, and not in virtualized environments */ + if (index < 3) + return; + + node = amd_get_nb_id(smp_processor_id()); + this_leaf->nb = node_to_amd_nb(node); + if (this_leaf->nb && !this_leaf->nb->l3_cache.indices) + amd_calc_l3_indices(this_leaf->nb); +} +#else +#define amd_init_l3_cache(x, y) +#endif /* CONFIG_AMD_NB && CONFIG_SYSFS */ + +static int +cpuid4_cache_lookup_regs(int index, struct _cpuid4_info_regs *this_leaf) +{ + union _cpuid4_leaf_eax eax; + union _cpuid4_leaf_ebx ebx; + union _cpuid4_leaf_ecx ecx; + unsigned edx; + + if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) { + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) + cpuid_count(0x8000001d, index, &eax.full, + &ebx.full, &ecx.full, &edx); + else + amd_cpuid4(index, &eax, &ebx, &ecx); + amd_init_l3_cache(this_leaf, index); + } else if (boot_cpu_data.x86_vendor == X86_VENDOR_HYGON) { + cpuid_count(0x8000001d, index, &eax.full, + &ebx.full, &ecx.full, &edx); + amd_init_l3_cache(this_leaf, index); + } else { + cpuid_count(4, index, &eax.full, &ebx.full, &ecx.full, &edx); + } + + if (eax.split.type == CTYPE_NULL) + return -EIO; /* better error ? */ + + this_leaf->eax = eax; + this_leaf->ebx = ebx; + this_leaf->ecx = ecx; + this_leaf->size = (ecx.split.number_of_sets + 1) * + (ebx.split.coherency_line_size + 1) * + (ebx.split.physical_line_partition + 1) * + (ebx.split.ways_of_associativity + 1); + return 0; +} + +static int find_num_cache_leaves(struct cpuinfo_x86 *c) +{ + unsigned int eax, ebx, ecx, edx, op; + union _cpuid4_leaf_eax cache_eax; + int i = -1; + + if (c->x86_vendor == X86_VENDOR_AMD || + c->x86_vendor == X86_VENDOR_HYGON) + op = 0x8000001d; + else + op = 4; + + do { + ++i; + /* Do cpuid(op) loop to find out num_cache_leaves */ + cpuid_count(op, i, &eax, &ebx, &ecx, &edx); + cache_eax.full = eax; + } while (cache_eax.split.type != CTYPE_NULL); + return i; +} + +void cacheinfo_amd_init_llc_id(struct cpuinfo_x86 *c, int cpu) +{ + /* + * We may have multiple LLCs if L3 caches exist, so check if we + * have an L3 cache by looking at the L3 cache CPUID leaf. + */ + if (!cpuid_edx(0x80000006)) + return; + + if (c->x86 < 0x17) { + /* LLC is at the node level. */ + per_cpu(cpu_llc_id, cpu) = c->cpu_die_id; + } else if (c->x86 == 0x17 && c->x86_model <= 0x1F) { + /* + * LLC is at the core complex level. + * Core complex ID is ApicId[3] for these processors. + */ + per_cpu(cpu_llc_id, cpu) = c->apicid >> 3; + } else { + /* + * LLC ID is calculated from the number of threads sharing the + * cache. + * */ + u32 eax, ebx, ecx, edx, num_sharing_cache = 0; + u32 llc_index = find_num_cache_leaves(c) - 1; + + cpuid_count(0x8000001d, llc_index, &eax, &ebx, &ecx, &edx); + if (eax) + num_sharing_cache = ((eax >> 14) & 0xfff) + 1; + + if (num_sharing_cache) { + int bits = get_count_order(num_sharing_cache); + + per_cpu(cpu_llc_id, cpu) = c->apicid >> bits; + } + } +} + +void cacheinfo_hygon_init_llc_id(struct cpuinfo_x86 *c, int cpu) +{ + /* + * We may have multiple LLCs if L3 caches exist, so check if we + * have an L3 cache by looking at the L3 cache CPUID leaf. + */ + if (!cpuid_edx(0x80000006)) + return; + + /* + * LLC is at the core complex level. + * Core complex ID is ApicId[3] for these processors. + */ + per_cpu(cpu_llc_id, cpu) = c->apicid >> 3; +} + +void init_amd_cacheinfo(struct cpuinfo_x86 *c) +{ + + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + num_cache_leaves = find_num_cache_leaves(c); + } else if (c->extended_cpuid_level >= 0x80000006) { + if (cpuid_edx(0x80000006) & 0xf000) + num_cache_leaves = 4; + else + num_cache_leaves = 3; + } +} + +void init_hygon_cacheinfo(struct cpuinfo_x86 *c) +{ + num_cache_leaves = find_num_cache_leaves(c); +} + +void init_intel_cacheinfo(struct cpuinfo_x86 *c) +{ + /* Cache sizes */ + unsigned int trace = 0, l1i = 0, l1d = 0, l2 = 0, l3 = 0; + unsigned int new_l1d = 0, new_l1i = 0; /* Cache sizes from cpuid(4) */ + unsigned int new_l2 = 0, new_l3 = 0, i; /* Cache sizes from cpuid(4) */ + unsigned int l2_id = 0, l3_id = 0, num_threads_sharing, index_msb; +#ifdef CONFIG_SMP + unsigned int cpu = c->cpu_index; +#endif + + if (c->cpuid_level > 3) { + static int is_initialized; + + if (is_initialized == 0) { + /* Init num_cache_leaves from boot CPU */ + num_cache_leaves = find_num_cache_leaves(c); + is_initialized++; + } + + /* + * Whenever possible use cpuid(4), deterministic cache + * parameters cpuid leaf to find the cache details + */ + for (i = 0; i < num_cache_leaves; i++) { + struct _cpuid4_info_regs this_leaf = {}; + int retval; + + retval = cpuid4_cache_lookup_regs(i, &this_leaf); + if (retval < 0) + continue; + + switch (this_leaf.eax.split.level) { + case 1: + if (this_leaf.eax.split.type == CTYPE_DATA) + new_l1d = this_leaf.size/1024; + else if (this_leaf.eax.split.type == CTYPE_INST) + new_l1i = this_leaf.size/1024; + break; + case 2: + new_l2 = this_leaf.size/1024; + num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; + index_msb = get_count_order(num_threads_sharing); + l2_id = c->apicid & ~((1 << index_msb) - 1); + break; + case 3: + new_l3 = this_leaf.size/1024; + num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; + index_msb = get_count_order(num_threads_sharing); + l3_id = c->apicid & ~((1 << index_msb) - 1); + break; + default: + break; + } + } + } + /* + * Don't use cpuid2 if cpuid4 is supported. For P4, we use cpuid2 for + * trace cache + */ + if ((num_cache_leaves == 0 || c->x86 == 15) && c->cpuid_level > 1) { + /* supports eax=2 call */ + int j, n; + unsigned int regs[4]; + unsigned char *dp = (unsigned char *)regs; + int only_trace = 0; + + if (num_cache_leaves != 0 && c->x86 == 15) + only_trace = 1; + + /* Number of times to iterate */ + n = cpuid_eax(2) & 0xFF; + + for (i = 0 ; i < n ; i++) { + cpuid(2, ®s[0], ®s[1], ®s[2], ®s[3]); + + /* If bit 31 is set, this is an unknown format */ + for (j = 0 ; j < 3 ; j++) + if (regs[j] & (1 << 31)) + regs[j] = 0; + + /* Byte 0 is level count, not a descriptor */ + for (j = 1 ; j < 16 ; j++) { + unsigned char des = dp[j]; + unsigned char k = 0; + + /* look up this descriptor in the table */ + while (cache_table[k].descriptor != 0) { + if (cache_table[k].descriptor == des) { + if (only_trace && cache_table[k].cache_type != LVL_TRACE) + break; + switch (cache_table[k].cache_type) { + case LVL_1_INST: + l1i += cache_table[k].size; + break; + case LVL_1_DATA: + l1d += cache_table[k].size; + break; + case LVL_2: + l2 += cache_table[k].size; + break; + case LVL_3: + l3 += cache_table[k].size; + break; + case LVL_TRACE: + trace += cache_table[k].size; + break; + } + + break; + } + + k++; + } + } + } + } + + if (new_l1d) + l1d = new_l1d; + + if (new_l1i) + l1i = new_l1i; + + if (new_l2) { + l2 = new_l2; +#ifdef CONFIG_SMP + per_cpu(cpu_llc_id, cpu) = l2_id; +#endif + } + + if (new_l3) { + l3 = new_l3; +#ifdef CONFIG_SMP + per_cpu(cpu_llc_id, cpu) = l3_id; +#endif + } + +#ifdef CONFIG_SMP + /* + * If cpu_llc_id is not yet set, this means cpuid_level < 4 which in + * turns means that the only possibility is SMT (as indicated in + * cpuid1). Since cpuid2 doesn't specify shared caches, and we know + * that SMT shares all caches, we can unconditionally set cpu_llc_id to + * c->phys_proc_id. + */ + if (per_cpu(cpu_llc_id, cpu) == BAD_APICID) + per_cpu(cpu_llc_id, cpu) = c->phys_proc_id; +#endif + + c->x86_cache_size = l3 ? l3 : (l2 ? l2 : (l1i+l1d)); + + if (!l2) + cpu_detect_cache_sizes(c); +} + +static int __cache_amd_cpumap_setup(unsigned int cpu, int index, + struct _cpuid4_info_regs *base) +{ + struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); + struct cacheinfo *this_leaf; + int i, sibling; + + /* + * For L3, always use the pre-calculated cpu_llc_shared_mask + * to derive shared_cpu_map. + */ + if (index == 3) { + for_each_cpu(i, cpu_llc_shared_mask(cpu)) { + this_cpu_ci = get_cpu_cacheinfo(i); + if (!this_cpu_ci->info_list) + continue; + this_leaf = this_cpu_ci->info_list + index; + for_each_cpu(sibling, cpu_llc_shared_mask(cpu)) { + if (!cpu_online(sibling)) + continue; + cpumask_set_cpu(sibling, + &this_leaf->shared_cpu_map); + } + } + } else if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + unsigned int apicid, nshared, first, last; + + nshared = base->eax.split.num_threads_sharing + 1; + apicid = cpu_data(cpu).apicid; + first = apicid - (apicid % nshared); + last = first + nshared - 1; + + for_each_online_cpu(i) { + this_cpu_ci = get_cpu_cacheinfo(i); + if (!this_cpu_ci->info_list) + continue; + + apicid = cpu_data(i).apicid; + if ((apicid < first) || (apicid > last)) + continue; + + this_leaf = this_cpu_ci->info_list + index; + + for_each_online_cpu(sibling) { + apicid = cpu_data(sibling).apicid; + if ((apicid < first) || (apicid > last)) + continue; + cpumask_set_cpu(sibling, + &this_leaf->shared_cpu_map); + } + } + } else + return 0; + + return 1; +} + +static void __cache_cpumap_setup(unsigned int cpu, int index, + struct _cpuid4_info_regs *base) +{ + struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); + struct cacheinfo *this_leaf, *sibling_leaf; + unsigned long num_threads_sharing; + int index_msb, i; + struct cpuinfo_x86 *c = &cpu_data(cpu); + + if (c->x86_vendor == X86_VENDOR_AMD || + c->x86_vendor == X86_VENDOR_HYGON) { + if (__cache_amd_cpumap_setup(cpu, index, base)) + return; + } + + this_leaf = this_cpu_ci->info_list + index; + num_threads_sharing = 1 + base->eax.split.num_threads_sharing; + + cpumask_set_cpu(cpu, &this_leaf->shared_cpu_map); + if (num_threads_sharing == 1) + return; + + index_msb = get_count_order(num_threads_sharing); + + for_each_online_cpu(i) + if (cpu_data(i).apicid >> index_msb == c->apicid >> index_msb) { + struct cpu_cacheinfo *sib_cpu_ci = get_cpu_cacheinfo(i); + + if (i == cpu || !sib_cpu_ci->info_list) + continue;/* skip if itself or no cacheinfo */ + sibling_leaf = sib_cpu_ci->info_list + index; + cpumask_set_cpu(i, &this_leaf->shared_cpu_map); + cpumask_set_cpu(cpu, &sibling_leaf->shared_cpu_map); + } +} + +static void ci_leaf_init(struct cacheinfo *this_leaf, + struct _cpuid4_info_regs *base) +{ + this_leaf->id = base->id; + this_leaf->attributes = CACHE_ID; + this_leaf->level = base->eax.split.level; + this_leaf->type = cache_type_map[base->eax.split.type]; + this_leaf->coherency_line_size = + base->ebx.split.coherency_line_size + 1; + this_leaf->ways_of_associativity = + base->ebx.split.ways_of_associativity + 1; + this_leaf->size = base->size; + this_leaf->number_of_sets = base->ecx.split.number_of_sets + 1; + this_leaf->physical_line_partition = + base->ebx.split.physical_line_partition + 1; + this_leaf->priv = base->nb; +} + +int init_cache_level(unsigned int cpu) +{ + struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); + + if (!num_cache_leaves) + return -ENOENT; + if (!this_cpu_ci) + return -EINVAL; + this_cpu_ci->num_levels = 3; + this_cpu_ci->num_leaves = num_cache_leaves; + return 0; +} + +/* + * The max shared threads number comes from CPUID.4:EAX[25-14] with input + * ECX as cache index. Then right shift apicid by the number's order to get + * cache id for this cache node. + */ +static void get_cache_id(int cpu, struct _cpuid4_info_regs *id4_regs) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + unsigned long num_threads_sharing; + int index_msb; + + num_threads_sharing = 1 + id4_regs->eax.split.num_threads_sharing; + index_msb = get_count_order(num_threads_sharing); + id4_regs->id = c->apicid >> index_msb; +} + +int populate_cache_leaves(unsigned int cpu) +{ + unsigned int idx, ret; + struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); + struct cacheinfo *this_leaf = this_cpu_ci->info_list; + struct _cpuid4_info_regs id4_regs = {}; + + for (idx = 0; idx < this_cpu_ci->num_leaves; idx++) { + ret = cpuid4_cache_lookup_regs(idx, &id4_regs); + if (ret) + return ret; + get_cache_id(cpu, &id4_regs); + ci_leaf_init(this_leaf++, &id4_regs); + __cache_cpumap_setup(cpu, idx, &id4_regs); + } + this_cpu_ci->cpu_map_populated = true; + + return 0; +} diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c new file mode 100644 index 000000000..345f7d905 --- /dev/null +++ b/arch/x86/kernel/cpu/centaur.c @@ -0,0 +1,251 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <linux/sched.h> +#include <linux/sched/clock.h> + +#include <asm/cpu.h> +#include <asm/cpufeature.h> +#include <asm/e820/api.h> +#include <asm/mtrr.h> +#include <asm/msr.h> + +#include "cpu.h" + +#define ACE_PRESENT (1 << 6) +#define ACE_ENABLED (1 << 7) +#define ACE_FCR (1 << 28) /* MSR_VIA_FCR */ + +#define RNG_PRESENT (1 << 2) +#define RNG_ENABLED (1 << 3) +#define RNG_ENABLE (1 << 6) /* MSR_VIA_RNG */ + +static void init_c3(struct cpuinfo_x86 *c) +{ + u32 lo, hi; + + /* Test for Centaur Extended Feature Flags presence */ + if (cpuid_eax(0xC0000000) >= 0xC0000001) { + u32 tmp = cpuid_edx(0xC0000001); + + /* enable ACE unit, if present and disabled */ + if ((tmp & (ACE_PRESENT | ACE_ENABLED)) == ACE_PRESENT) { + rdmsr(MSR_VIA_FCR, lo, hi); + lo |= ACE_FCR; /* enable ACE unit */ + wrmsr(MSR_VIA_FCR, lo, hi); + pr_info("CPU: Enabled ACE h/w crypto\n"); + } + + /* enable RNG unit, if present and disabled */ + if ((tmp & (RNG_PRESENT | RNG_ENABLED)) == RNG_PRESENT) { + rdmsr(MSR_VIA_RNG, lo, hi); + lo |= RNG_ENABLE; /* enable RNG unit */ + wrmsr(MSR_VIA_RNG, lo, hi); + pr_info("CPU: Enabled h/w RNG\n"); + } + + /* store Centaur Extended Feature Flags as + * word 5 of the CPU capability bit array + */ + c->x86_capability[CPUID_C000_0001_EDX] = cpuid_edx(0xC0000001); + } +#ifdef CONFIG_X86_32 + /* Cyrix III family needs CX8 & PGE explicitly enabled. */ + if (c->x86_model >= 6 && c->x86_model <= 13) { + rdmsr(MSR_VIA_FCR, lo, hi); + lo |= (1<<1 | 1<<7); + wrmsr(MSR_VIA_FCR, lo, hi); + set_cpu_cap(c, X86_FEATURE_CX8); + } + + /* Before Nehemiah, the C3's had 3dNOW! */ + if (c->x86_model >= 6 && c->x86_model < 9) + set_cpu_cap(c, X86_FEATURE_3DNOW); +#endif + if (c->x86 == 0x6 && c->x86_model >= 0xf) { + c->x86_cache_alignment = c->x86_clflush_size * 2; + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + } + + if (c->x86 >= 7) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); +} + +enum { + ECX8 = 1<<1, + EIERRINT = 1<<2, + DPM = 1<<3, + DMCE = 1<<4, + DSTPCLK = 1<<5, + ELINEAR = 1<<6, + DSMC = 1<<7, + DTLOCK = 1<<8, + EDCTLB = 1<<8, + EMMX = 1<<9, + DPDC = 1<<11, + EBRPRED = 1<<12, + DIC = 1<<13, + DDC = 1<<14, + DNA = 1<<15, + ERETSTK = 1<<16, + E2MMX = 1<<19, + EAMD3D = 1<<20, +}; + +static void early_init_centaur(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + /* Emulate MTRRs using Centaur's MCR. */ + if (c->x86 == 5) + set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR); +#endif + if ((c->x86 == 6 && c->x86_model >= 0xf) || + (c->x86 >= 7)) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +#endif + if (c->x86_power & (1 << 8)) { + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + } +} + +static void init_centaur(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + char *name; + u32 fcr_set = 0; + u32 fcr_clr = 0; + u32 lo, hi, newlo; + u32 aa, bb, cc, dd; + + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); +#endif + early_init_centaur(c); + init_intel_cacheinfo(c); + detect_num_cpu_cores(c); +#ifdef CONFIG_X86_32 + detect_ht(c); +#endif + + if (c->cpuid_level > 9) { + unsigned int eax = cpuid_eax(10); + + /* + * Check for version and the number of counters + * Version(eax[7:0]) can't be 0; + * Counters(eax[15:8]) should be greater than 1; + */ + if ((eax & 0xff) && (((eax >> 8) & 0xff) > 1)) + set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); + } + +#ifdef CONFIG_X86_32 + if (c->x86 == 5) { + switch (c->x86_model) { + case 4: + name = "C6"; + fcr_set = ECX8|DSMC|EDCTLB|EMMX|ERETSTK; + fcr_clr = DPDC; + pr_notice("Disabling bugged TSC.\n"); + clear_cpu_cap(c, X86_FEATURE_TSC); + break; + case 8: + switch (c->x86_stepping) { + default: + name = "2"; + break; + case 7 ... 9: + name = "2A"; + break; + case 10 ... 15: + name = "2B"; + break; + } + fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| + E2MMX|EAMD3D; + fcr_clr = DPDC; + break; + case 9: + name = "3"; + fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| + E2MMX|EAMD3D; + fcr_clr = DPDC; + break; + default: + name = "??"; + } + + rdmsr(MSR_IDT_FCR1, lo, hi); + newlo = (lo|fcr_set) & (~fcr_clr); + + if (newlo != lo) { + pr_info("Centaur FCR was 0x%X now 0x%X\n", + lo, newlo); + wrmsr(MSR_IDT_FCR1, newlo, hi); + } else { + pr_info("Centaur FCR is 0x%X\n", lo); + } + /* Emulate MTRRs using Centaur's MCR. */ + set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR); + /* Report CX8 */ + set_cpu_cap(c, X86_FEATURE_CX8); + /* Set 3DNow! on Winchip 2 and above. */ + if (c->x86_model >= 8) + set_cpu_cap(c, X86_FEATURE_3DNOW); + /* See if we can find out some more. */ + if (cpuid_eax(0x80000000) >= 0x80000005) { + /* Yes, we can. */ + cpuid(0x80000005, &aa, &bb, &cc, &dd); + /* Add L1 data and code cache sizes. */ + c->x86_cache_size = (cc>>24)+(dd>>24); + } + sprintf(c->x86_model_id, "WinChip %s", name); + } +#endif + if (c->x86 == 6 || c->x86 >= 7) + init_c3(c); +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); +#endif + + init_ia32_feat_ctl(c); +} + +#ifdef CONFIG_X86_32 +static unsigned int +centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) +{ + /* VIA C3 CPUs (670-68F) need further shifting. */ + if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8))) + size >>= 8; + + /* + * There's also an erratum in Nehemiah stepping 1, which + * returns '65KB' instead of '64KB' + * - Note, it seems this may only be in engineering samples. + */ + if ((c->x86 == 6) && (c->x86_model == 9) && + (c->x86_stepping == 1) && (size == 65)) + size -= 1; + return size; +} +#endif + +static const struct cpu_dev centaur_cpu_dev = { + .c_vendor = "Centaur", + .c_ident = { "CentaurHauls" }, + .c_early_init = early_init_centaur, + .c_init = init_centaur, +#ifdef CONFIG_X86_32 + .legacy_cache_size = centaur_size_cache, +#endif + .c_x86_vendor = X86_VENDOR_CENTAUR, +}; + +cpu_dev_register(centaur_cpu_dev); diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c new file mode 100644 index 000000000..4ecc6072e --- /dev/null +++ b/arch/x86/kernel/cpu/common.c @@ -0,0 +1,2285 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* cpu_feature_enabled() cannot be used this early */ +#define USE_EARLY_PGTABLE_L5 + +#include <linux/memblock.h> +#include <linux/linkage.h> +#include <linux/bitops.h> +#include <linux/kernel.h> +#include <linux/export.h> +#include <linux/percpu.h> +#include <linux/string.h> +#include <linux/ctype.h> +#include <linux/delay.h> +#include <linux/sched/mm.h> +#include <linux/sched/clock.h> +#include <linux/sched/task.h> +#include <linux/sched/smt.h> +#include <linux/init.h> +#include <linux/kprobes.h> +#include <linux/kgdb.h> +#include <linux/mem_encrypt.h> +#include <linux/smp.h> +#include <linux/cpu.h> +#include <linux/io.h> +#include <linux/syscore_ops.h> +#include <linux/pgtable.h> +#include <linux/utsname.h> + +#include <asm/alternative.h> +#include <asm/cmdline.h> +#include <asm/stackprotector.h> +#include <asm/perf_event.h> +#include <asm/mmu_context.h> +#include <asm/doublefault.h> +#include <asm/archrandom.h> +#include <asm/hypervisor.h> +#include <asm/processor.h> +#include <asm/tlbflush.h> +#include <asm/debugreg.h> +#include <asm/sections.h> +#include <asm/vsyscall.h> +#include <linux/topology.h> +#include <linux/cpumask.h> +#include <linux/atomic.h> +#include <asm/proto.h> +#include <asm/setup.h> +#include <asm/apic.h> +#include <asm/desc.h> +#include <asm/fpu/internal.h> +#include <asm/mtrr.h> +#include <asm/hwcap2.h> +#include <linux/numa.h> +#include <asm/numa.h> +#include <asm/asm.h> +#include <asm/bugs.h> +#include <asm/cpu.h> +#include <asm/mce.h> +#include <asm/msr.h> +#include <asm/memtype.h> +#include <asm/microcode.h> +#include <asm/microcode_intel.h> +#include <asm/intel-family.h> +#include <asm/cpu_device_id.h> +#include <asm/uv/uv.h> +#include <asm/set_memory.h> + +#include "cpu.h" + +u32 elf_hwcap2 __read_mostly; + +/* all of these masks are initialized in setup_cpu_local_masks() */ +cpumask_var_t cpu_initialized_mask; +cpumask_var_t cpu_callout_mask; +cpumask_var_t cpu_callin_mask; + +/* representing cpus for which sibling maps can be computed */ +cpumask_var_t cpu_sibling_setup_mask; + +/* Number of siblings per CPU package */ +int smp_num_siblings = 1; +EXPORT_SYMBOL(smp_num_siblings); + +/* Last level cache ID of each logical CPU */ +DEFINE_PER_CPU_READ_MOSTLY(u16, cpu_llc_id) = BAD_APICID; + +/* correctly size the local cpu masks */ +void __init setup_cpu_local_masks(void) +{ + alloc_bootmem_cpumask_var(&cpu_initialized_mask); + alloc_bootmem_cpumask_var(&cpu_callin_mask); + alloc_bootmem_cpumask_var(&cpu_callout_mask); + alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask); +} + +static void default_init(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_64 + cpu_detect_cache_sizes(c); +#else + /* Not much we can do here... */ + /* Check if at least it has cpuid */ + if (c->cpuid_level == -1) { + /* No cpuid. It must be an ancient CPU */ + if (c->x86 == 4) + strcpy(c->x86_model_id, "486"); + else if (c->x86 == 3) + strcpy(c->x86_model_id, "386"); + } +#endif +} + +static const struct cpu_dev default_cpu = { + .c_init = default_init, + .c_vendor = "Unknown", + .c_x86_vendor = X86_VENDOR_UNKNOWN, +}; + +static const struct cpu_dev *this_cpu = &default_cpu; + +DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { +#ifdef CONFIG_X86_64 + /* + * We need valid kernel segments for data and code in long mode too + * IRET will check the segment types kkeil 2000/10/28 + * Also sysret mandates a special GDT layout + * + * TLS descriptors are currently at a different place compared to i386. + * Hopefully nobody expects them at a fixed place (Wine?) + */ + [GDT_ENTRY_KERNEL32_CS] = GDT_ENTRY_INIT(0xc09b, 0, 0xfffff), + [GDT_ENTRY_KERNEL_CS] = GDT_ENTRY_INIT(0xa09b, 0, 0xfffff), + [GDT_ENTRY_KERNEL_DS] = GDT_ENTRY_INIT(0xc093, 0, 0xfffff), + [GDT_ENTRY_DEFAULT_USER32_CS] = GDT_ENTRY_INIT(0xc0fb, 0, 0xfffff), + [GDT_ENTRY_DEFAULT_USER_DS] = GDT_ENTRY_INIT(0xc0f3, 0, 0xfffff), + [GDT_ENTRY_DEFAULT_USER_CS] = GDT_ENTRY_INIT(0xa0fb, 0, 0xfffff), +#else + [GDT_ENTRY_KERNEL_CS] = GDT_ENTRY_INIT(0xc09a, 0, 0xfffff), + [GDT_ENTRY_KERNEL_DS] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), + [GDT_ENTRY_DEFAULT_USER_CS] = GDT_ENTRY_INIT(0xc0fa, 0, 0xfffff), + [GDT_ENTRY_DEFAULT_USER_DS] = GDT_ENTRY_INIT(0xc0f2, 0, 0xfffff), + /* + * Segments used for calling PnP BIOS have byte granularity. + * They code segments and data segments have fixed 64k limits, + * the transfer segment sizes are set at run time. + */ + /* 32-bit code */ + [GDT_ENTRY_PNPBIOS_CS32] = GDT_ENTRY_INIT(0x409a, 0, 0xffff), + /* 16-bit code */ + [GDT_ENTRY_PNPBIOS_CS16] = GDT_ENTRY_INIT(0x009a, 0, 0xffff), + /* 16-bit data */ + [GDT_ENTRY_PNPBIOS_DS] = GDT_ENTRY_INIT(0x0092, 0, 0xffff), + /* 16-bit data */ + [GDT_ENTRY_PNPBIOS_TS1] = GDT_ENTRY_INIT(0x0092, 0, 0), + /* 16-bit data */ + [GDT_ENTRY_PNPBIOS_TS2] = GDT_ENTRY_INIT(0x0092, 0, 0), + /* + * The APM segments have byte granularity and their bases + * are set at run time. All have 64k limits. + */ + /* 32-bit code */ + [GDT_ENTRY_APMBIOS_BASE] = GDT_ENTRY_INIT(0x409a, 0, 0xffff), + /* 16-bit code */ + [GDT_ENTRY_APMBIOS_BASE+1] = GDT_ENTRY_INIT(0x009a, 0, 0xffff), + /* data */ + [GDT_ENTRY_APMBIOS_BASE+2] = GDT_ENTRY_INIT(0x4092, 0, 0xffff), + + [GDT_ENTRY_ESPFIX_SS] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), + [GDT_ENTRY_PERCPU] = GDT_ENTRY_INIT(0xc092, 0, 0xfffff), + GDT_STACK_CANARY_INIT +#endif +} }; +EXPORT_PER_CPU_SYMBOL_GPL(gdt_page); + +#ifdef CONFIG_X86_64 +static int __init x86_nopcid_setup(char *s) +{ + /* nopcid doesn't accept parameters */ + if (s) + return -EINVAL; + + /* do not emit a message if the feature is not present */ + if (!boot_cpu_has(X86_FEATURE_PCID)) + return 0; + + setup_clear_cpu_cap(X86_FEATURE_PCID); + pr_info("nopcid: PCID feature disabled\n"); + return 0; +} +early_param("nopcid", x86_nopcid_setup); +#endif + +static int __init x86_noinvpcid_setup(char *s) +{ + /* noinvpcid doesn't accept parameters */ + if (s) + return -EINVAL; + + /* do not emit a message if the feature is not present */ + if (!boot_cpu_has(X86_FEATURE_INVPCID)) + return 0; + + setup_clear_cpu_cap(X86_FEATURE_INVPCID); + pr_info("noinvpcid: INVPCID feature disabled\n"); + return 0; +} +early_param("noinvpcid", x86_noinvpcid_setup); + +#ifdef CONFIG_X86_32 +static int cachesize_override = -1; +static int disable_x86_serial_nr = 1; + +static int __init cachesize_setup(char *str) +{ + get_option(&str, &cachesize_override); + return 1; +} +__setup("cachesize=", cachesize_setup); + +static int __init x86_sep_setup(char *s) +{ + setup_clear_cpu_cap(X86_FEATURE_SEP); + return 1; +} +__setup("nosep", x86_sep_setup); + +/* Standard macro to see if a specific flag is changeable */ +static inline int flag_is_changeable_p(u32 flag) +{ + u32 f1, f2; + + /* + * Cyrix and IDT cpus allow disabling of CPUID + * so the code below may return different results + * when it is executed before and after enabling + * the CPUID. Add "volatile" to not allow gcc to + * optimize the subsequent calls to this function. + */ + asm volatile ("pushfl \n\t" + "pushfl \n\t" + "popl %0 \n\t" + "movl %0, %1 \n\t" + "xorl %2, %0 \n\t" + "pushl %0 \n\t" + "popfl \n\t" + "pushfl \n\t" + "popl %0 \n\t" + "popfl \n\t" + + : "=&r" (f1), "=&r" (f2) + : "ir" (flag)); + + return ((f1^f2) & flag) != 0; +} + +/* Probe for the CPUID instruction */ +int have_cpuid_p(void) +{ + return flag_is_changeable_p(X86_EFLAGS_ID); +} + +static void squash_the_stupid_serial_number(struct cpuinfo_x86 *c) +{ + unsigned long lo, hi; + + if (!cpu_has(c, X86_FEATURE_PN) || !disable_x86_serial_nr) + return; + + /* Disable processor serial number: */ + + rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi); + lo |= 0x200000; + wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi); + + pr_notice("CPU serial number disabled.\n"); + clear_cpu_cap(c, X86_FEATURE_PN); + + /* Disabling the serial number may affect the cpuid level */ + c->cpuid_level = cpuid_eax(0); +} + +static int __init x86_serial_nr_setup(char *s) +{ + disable_x86_serial_nr = 0; + return 1; +} +__setup("serialnumber", x86_serial_nr_setup); +#else +static inline int flag_is_changeable_p(u32 flag) +{ + return 1; +} +static inline void squash_the_stupid_serial_number(struct cpuinfo_x86 *c) +{ +} +#endif + +static __init int setup_disable_smep(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_SMEP); + return 1; +} +__setup("nosmep", setup_disable_smep); + +static __always_inline void setup_smep(struct cpuinfo_x86 *c) +{ + if (cpu_has(c, X86_FEATURE_SMEP)) + cr4_set_bits(X86_CR4_SMEP); +} + +static __init int setup_disable_smap(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_SMAP); + return 1; +} +__setup("nosmap", setup_disable_smap); + +static __always_inline void setup_smap(struct cpuinfo_x86 *c) +{ + unsigned long eflags = native_save_fl(); + + /* This should have been cleared long ago */ + BUG_ON(eflags & X86_EFLAGS_AC); + + if (cpu_has(c, X86_FEATURE_SMAP)) { +#ifdef CONFIG_X86_SMAP + cr4_set_bits(X86_CR4_SMAP); +#else + clear_cpu_cap(c, X86_FEATURE_SMAP); + cr4_clear_bits(X86_CR4_SMAP); +#endif + } +} + +static __always_inline void setup_umip(struct cpuinfo_x86 *c) +{ + /* Check the boot processor, plus build option for UMIP. */ + if (!cpu_feature_enabled(X86_FEATURE_UMIP)) + goto out; + + /* Check the current processor's cpuid bits. */ + if (!cpu_has(c, X86_FEATURE_UMIP)) + goto out; + + cr4_set_bits(X86_CR4_UMIP); + + pr_info_once("x86/cpu: User Mode Instruction Prevention (UMIP) activated\n"); + + return; + +out: + /* + * Make sure UMIP is disabled in case it was enabled in a + * previous boot (e.g., via kexec). + */ + cr4_clear_bits(X86_CR4_UMIP); +} + +/* These bits should not change their value after CPU init is finished. */ +static const unsigned long cr4_pinned_mask = + X86_CR4_SMEP | X86_CR4_SMAP | X86_CR4_UMIP | X86_CR4_FSGSBASE; +static DEFINE_STATIC_KEY_FALSE_RO(cr_pinning); +static unsigned long cr4_pinned_bits __ro_after_init; + +void native_write_cr0(unsigned long val) +{ + unsigned long bits_missing = 0; + +set_register: + asm volatile("mov %0,%%cr0": "+r" (val) : : "memory"); + + if (static_branch_likely(&cr_pinning)) { + if (unlikely((val & X86_CR0_WP) != X86_CR0_WP)) { + bits_missing = X86_CR0_WP; + val |= bits_missing; + goto set_register; + } + /* Warn after we've set the missing bits. */ + WARN_ONCE(bits_missing, "CR0 WP bit went missing!?\n"); + } +} +EXPORT_SYMBOL(native_write_cr0); + +void native_write_cr4(unsigned long val) +{ + unsigned long bits_changed = 0; + +set_register: + asm volatile("mov %0,%%cr4": "+r" (val) : : "memory"); + + if (static_branch_likely(&cr_pinning)) { + if (unlikely((val & cr4_pinned_mask) != cr4_pinned_bits)) { + bits_changed = (val & cr4_pinned_mask) ^ cr4_pinned_bits; + val = (val & ~cr4_pinned_mask) | cr4_pinned_bits; + goto set_register; + } + /* Warn after we've corrected the changed bits. */ + WARN_ONCE(bits_changed, "pinned CR4 bits changed: 0x%lx!?\n", + bits_changed); + } +} +#if IS_MODULE(CONFIG_LKDTM) +EXPORT_SYMBOL_GPL(native_write_cr4); +#endif + +void cr4_update_irqsoff(unsigned long set, unsigned long clear) +{ + unsigned long newval, cr4 = this_cpu_read(cpu_tlbstate.cr4); + + lockdep_assert_irqs_disabled(); + + newval = (cr4 & ~clear) | set; + if (newval != cr4) { + this_cpu_write(cpu_tlbstate.cr4, newval); + __write_cr4(newval); + } +} +EXPORT_SYMBOL(cr4_update_irqsoff); + +/* Read the CR4 shadow. */ +unsigned long cr4_read_shadow(void) +{ + return this_cpu_read(cpu_tlbstate.cr4); +} +EXPORT_SYMBOL_GPL(cr4_read_shadow); + +void cr4_init(void) +{ + unsigned long cr4 = __read_cr4(); + + if (boot_cpu_has(X86_FEATURE_PCID)) + cr4 |= X86_CR4_PCIDE; + if (static_branch_likely(&cr_pinning)) + cr4 = (cr4 & ~cr4_pinned_mask) | cr4_pinned_bits; + + __write_cr4(cr4); + + /* Initialize cr4 shadow for this CPU. */ + this_cpu_write(cpu_tlbstate.cr4, cr4); +} + +/* + * Once CPU feature detection is finished (and boot params have been + * parsed), record any of the sensitive CR bits that are set, and + * enable CR pinning. + */ +static void __init setup_cr_pinning(void) +{ + cr4_pinned_bits = this_cpu_read(cpu_tlbstate.cr4) & cr4_pinned_mask; + static_key_enable(&cr_pinning.key); +} + +static __init int x86_nofsgsbase_setup(char *arg) +{ + /* Require an exact match without trailing characters. */ + if (strlen(arg)) + return 0; + + /* Do not emit a message if the feature is not present. */ + if (!boot_cpu_has(X86_FEATURE_FSGSBASE)) + return 1; + + setup_clear_cpu_cap(X86_FEATURE_FSGSBASE); + pr_info("FSGSBASE disabled via kernel command line\n"); + return 1; +} +__setup("nofsgsbase", x86_nofsgsbase_setup); + +/* + * Protection Keys are not available in 32-bit mode. + */ +static bool pku_disabled; + +static __always_inline void setup_pku(struct cpuinfo_x86 *c) +{ + /* check the boot processor, plus compile options for PKU: */ + if (!cpu_feature_enabled(X86_FEATURE_PKU)) + return; + /* checks the actual processor's cpuid bits: */ + if (!cpu_has(c, X86_FEATURE_PKU)) + return; + if (pku_disabled) + return; + + cr4_set_bits(X86_CR4_PKE); + /* + * Seting X86_CR4_PKE will cause the X86_FEATURE_OSPKE + * cpuid bit to be set. We need to ensure that we + * update that bit in this CPU's "cpu_info". + */ + set_cpu_cap(c, X86_FEATURE_OSPKE); +} + +#ifdef CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS +static __init int setup_disable_pku(char *arg) +{ + /* + * Do not clear the X86_FEATURE_PKU bit. All of the + * runtime checks are against OSPKE so clearing the + * bit does nothing. + * + * This way, we will see "pku" in cpuinfo, but not + * "ospke", which is exactly what we want. It shows + * that the CPU has PKU, but the OS has not enabled it. + * This happens to be exactly how a system would look + * if we disabled the config option. + */ + pr_info("x86: 'nopku' specified, disabling Memory Protection Keys\n"); + pku_disabled = true; + return 1; +} +__setup("nopku", setup_disable_pku); +#endif /* CONFIG_X86_64 */ + +/* + * Some CPU features depend on higher CPUID levels, which may not always + * be available due to CPUID level capping or broken virtualization + * software. Add those features to this table to auto-disable them. + */ +struct cpuid_dependent_feature { + u32 feature; + u32 level; +}; + +static const struct cpuid_dependent_feature +cpuid_dependent_features[] = { + { X86_FEATURE_MWAIT, 0x00000005 }, + { X86_FEATURE_DCA, 0x00000009 }, + { X86_FEATURE_XSAVE, 0x0000000d }, + { 0, 0 } +}; + +static void filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) +{ + const struct cpuid_dependent_feature *df; + + for (df = cpuid_dependent_features; df->feature; df++) { + + if (!cpu_has(c, df->feature)) + continue; + /* + * Note: cpuid_level is set to -1 if unavailable, but + * extended_extended_level is set to 0 if unavailable + * and the legitimate extended levels are all negative + * when signed; hence the weird messing around with + * signs here... + */ + if (!((s32)df->level < 0 ? + (u32)df->level > (u32)c->extended_cpuid_level : + (s32)df->level > (s32)c->cpuid_level)) + continue; + + clear_cpu_cap(c, df->feature); + if (!warn) + continue; + + pr_warn("CPU: CPU feature " X86_CAP_FMT " disabled, no CPUID level 0x%x\n", + x86_cap_flag(df->feature), df->level); + } +} + +/* + * Naming convention should be: <Name> [(<Codename>)] + * This table only is used unless init_<vendor>() below doesn't set it; + * in particular, if CPUID levels 0x80000002..4 are supported, this + * isn't used + */ + +/* Look up CPU names by table lookup. */ +static const char *table_lookup_model(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + const struct legacy_cpu_model_info *info; + + if (c->x86_model >= 16) + return NULL; /* Range check */ + + if (!this_cpu) + return NULL; + + info = this_cpu->legacy_models; + + while (info->family) { + if (info->family == c->x86) + return info->model_names[c->x86_model]; + info++; + } +#endif + return NULL; /* Not found */ +} + +/* Aligned to unsigned long to avoid split lock in atomic bitmap ops */ +__u32 cpu_caps_cleared[NCAPINTS + NBUGINTS] __aligned(sizeof(unsigned long)); +__u32 cpu_caps_set[NCAPINTS + NBUGINTS] __aligned(sizeof(unsigned long)); + +void load_percpu_segment(int cpu) +{ +#ifdef CONFIG_X86_32 + loadsegment(fs, __KERNEL_PERCPU); +#else + __loadsegment_simple(gs, 0); + wrmsrl(MSR_GS_BASE, cpu_kernelmode_gs_base(cpu)); +#endif + load_stack_canary_segment(); +} + +#ifdef CONFIG_X86_32 +/* The 32-bit entry code needs to find cpu_entry_area. */ +DEFINE_PER_CPU(struct cpu_entry_area *, cpu_entry_area); +#endif + +/* Load the original GDT from the per-cpu structure */ +void load_direct_gdt(int cpu) +{ + struct desc_ptr gdt_descr; + + gdt_descr.address = (long)get_cpu_gdt_rw(cpu); + gdt_descr.size = GDT_SIZE - 1; + load_gdt(&gdt_descr); +} +EXPORT_SYMBOL_GPL(load_direct_gdt); + +/* Load a fixmap remapping of the per-cpu GDT */ +void load_fixmap_gdt(int cpu) +{ + struct desc_ptr gdt_descr; + + gdt_descr.address = (long)get_cpu_gdt_ro(cpu); + gdt_descr.size = GDT_SIZE - 1; + load_gdt(&gdt_descr); +} +EXPORT_SYMBOL_GPL(load_fixmap_gdt); + +/* + * Current gdt points %fs at the "master" per-cpu area: after this, + * it's on the real one. + */ +void switch_to_new_gdt(int cpu) +{ + /* Load the original GDT */ + load_direct_gdt(cpu); + /* Reload the per-cpu base */ + load_percpu_segment(cpu); +} + +static const struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {}; + +static void get_model_name(struct cpuinfo_x86 *c) +{ + unsigned int *v; + char *p, *q, *s; + + if (c->extended_cpuid_level < 0x80000004) + return; + + v = (unsigned int *)c->x86_model_id; + cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]); + cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]); + cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]); + c->x86_model_id[48] = 0; + + /* Trim whitespace */ + p = q = s = &c->x86_model_id[0]; + + while (*p == ' ') + p++; + + while (*p) { + /* Note the last non-whitespace index */ + if (!isspace(*p)) + s = q; + + *q++ = *p++; + } + + *(s + 1) = '\0'; +} + +void detect_num_cpu_cores(struct cpuinfo_x86 *c) +{ + unsigned int eax, ebx, ecx, edx; + + c->x86_max_cores = 1; + if (!IS_ENABLED(CONFIG_SMP) || c->cpuid_level < 4) + return; + + cpuid_count(4, 0, &eax, &ebx, &ecx, &edx); + if (eax & 0x1f) + c->x86_max_cores = (eax >> 26) + 1; +} + +void cpu_detect_cache_sizes(struct cpuinfo_x86 *c) +{ + unsigned int n, dummy, ebx, ecx, edx, l2size; + + n = c->extended_cpuid_level; + + if (n >= 0x80000005) { + cpuid(0x80000005, &dummy, &ebx, &ecx, &edx); + c->x86_cache_size = (ecx>>24) + (edx>>24); +#ifdef CONFIG_X86_64 + /* On K8 L1 TLB is inclusive, so don't count it */ + c->x86_tlbsize = 0; +#endif + } + + if (n < 0x80000006) /* Some chips just has a large L1. */ + return; + + cpuid(0x80000006, &dummy, &ebx, &ecx, &edx); + l2size = ecx >> 16; + +#ifdef CONFIG_X86_64 + c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff); +#else + /* do processor-specific cache resizing */ + if (this_cpu->legacy_cache_size) + l2size = this_cpu->legacy_cache_size(c, l2size); + + /* Allow user to override all this if necessary. */ + if (cachesize_override != -1) + l2size = cachesize_override; + + if (l2size == 0) + return; /* Again, no L2 cache is possible */ +#endif + + c->x86_cache_size = l2size; +} + +u16 __read_mostly tlb_lli_4k[NR_INFO]; +u16 __read_mostly tlb_lli_2m[NR_INFO]; +u16 __read_mostly tlb_lli_4m[NR_INFO]; +u16 __read_mostly tlb_lld_4k[NR_INFO]; +u16 __read_mostly tlb_lld_2m[NR_INFO]; +u16 __read_mostly tlb_lld_4m[NR_INFO]; +u16 __read_mostly tlb_lld_1g[NR_INFO]; + +static void cpu_detect_tlb(struct cpuinfo_x86 *c) +{ + if (this_cpu->c_detect_tlb) + this_cpu->c_detect_tlb(c); + + pr_info("Last level iTLB entries: 4KB %d, 2MB %d, 4MB %d\n", + tlb_lli_4k[ENTRIES], tlb_lli_2m[ENTRIES], + tlb_lli_4m[ENTRIES]); + + pr_info("Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d, 1GB %d\n", + tlb_lld_4k[ENTRIES], tlb_lld_2m[ENTRIES], + tlb_lld_4m[ENTRIES], tlb_lld_1g[ENTRIES]); +} + +int detect_ht_early(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + u32 eax, ebx, ecx, edx; + + if (!cpu_has(c, X86_FEATURE_HT)) + return -1; + + if (cpu_has(c, X86_FEATURE_CMP_LEGACY)) + return -1; + + if (cpu_has(c, X86_FEATURE_XTOPOLOGY)) + return -1; + + cpuid(1, &eax, &ebx, &ecx, &edx); + + smp_num_siblings = (ebx & 0xff0000) >> 16; + if (smp_num_siblings == 1) + pr_info_once("CPU0: Hyper-Threading is disabled\n"); +#endif + return 0; +} + +void detect_ht(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + int index_msb, core_bits; + + if (detect_ht_early(c) < 0) + return; + + index_msb = get_count_order(smp_num_siblings); + c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb); + + smp_num_siblings = smp_num_siblings / c->x86_max_cores; + + index_msb = get_count_order(smp_num_siblings); + + core_bits = get_count_order(c->x86_max_cores); + + c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) & + ((1 << core_bits) - 1); +#endif +} + +static void get_cpu_vendor(struct cpuinfo_x86 *c) +{ + char *v = c->x86_vendor_id; + int i; + + for (i = 0; i < X86_VENDOR_NUM; i++) { + if (!cpu_devs[i]) + break; + + if (!strcmp(v, cpu_devs[i]->c_ident[0]) || + (cpu_devs[i]->c_ident[1] && + !strcmp(v, cpu_devs[i]->c_ident[1]))) { + + this_cpu = cpu_devs[i]; + c->x86_vendor = this_cpu->c_x86_vendor; + return; + } + } + + pr_err_once("CPU: vendor_id '%s' unknown, using generic init.\n" \ + "CPU: Your system may be unstable.\n", v); + + c->x86_vendor = X86_VENDOR_UNKNOWN; + this_cpu = &default_cpu; +} + +void cpu_detect(struct cpuinfo_x86 *c) +{ + /* Get vendor name */ + cpuid(0x00000000, (unsigned int *)&c->cpuid_level, + (unsigned int *)&c->x86_vendor_id[0], + (unsigned int *)&c->x86_vendor_id[8], + (unsigned int *)&c->x86_vendor_id[4]); + + c->x86 = 4; + /* Intel-defined flags: level 0x00000001 */ + if (c->cpuid_level >= 0x00000001) { + u32 junk, tfms, cap0, misc; + + cpuid(0x00000001, &tfms, &misc, &junk, &cap0); + c->x86 = x86_family(tfms); + c->x86_model = x86_model(tfms); + c->x86_stepping = x86_stepping(tfms); + + if (cap0 & (1<<19)) { + c->x86_clflush_size = ((misc >> 8) & 0xff) * 8; + c->x86_cache_alignment = c->x86_clflush_size; + } + } +} + +static void apply_forced_caps(struct cpuinfo_x86 *c) +{ + int i; + + for (i = 0; i < NCAPINTS + NBUGINTS; i++) { + c->x86_capability[i] &= ~cpu_caps_cleared[i]; + c->x86_capability[i] |= cpu_caps_set[i]; + } +} + +static void init_speculation_control(struct cpuinfo_x86 *c) +{ + /* + * The Intel SPEC_CTRL CPUID bit implies IBRS and IBPB support, + * and they also have a different bit for STIBP support. Also, + * a hypervisor might have set the individual AMD bits even on + * Intel CPUs, for finer-grained selection of what's available. + */ + if (cpu_has(c, X86_FEATURE_SPEC_CTRL)) { + set_cpu_cap(c, X86_FEATURE_IBRS); + set_cpu_cap(c, X86_FEATURE_IBPB); + set_cpu_cap(c, X86_FEATURE_MSR_SPEC_CTRL); + } + + if (cpu_has(c, X86_FEATURE_INTEL_STIBP)) + set_cpu_cap(c, X86_FEATURE_STIBP); + + if (cpu_has(c, X86_FEATURE_SPEC_CTRL_SSBD) || + cpu_has(c, X86_FEATURE_VIRT_SSBD)) + set_cpu_cap(c, X86_FEATURE_SSBD); + + if (cpu_has(c, X86_FEATURE_AMD_IBRS)) { + set_cpu_cap(c, X86_FEATURE_IBRS); + set_cpu_cap(c, X86_FEATURE_MSR_SPEC_CTRL); + } + + if (cpu_has(c, X86_FEATURE_AMD_IBPB)) + set_cpu_cap(c, X86_FEATURE_IBPB); + + if (cpu_has(c, X86_FEATURE_AMD_STIBP)) { + set_cpu_cap(c, X86_FEATURE_STIBP); + set_cpu_cap(c, X86_FEATURE_MSR_SPEC_CTRL); + } + + if (cpu_has(c, X86_FEATURE_AMD_SSBD)) { + set_cpu_cap(c, X86_FEATURE_SSBD); + set_cpu_cap(c, X86_FEATURE_MSR_SPEC_CTRL); + clear_cpu_cap(c, X86_FEATURE_VIRT_SSBD); + } +} + +void get_cpu_cap(struct cpuinfo_x86 *c) +{ + u32 eax, ebx, ecx, edx; + + /* Intel-defined flags: level 0x00000001 */ + if (c->cpuid_level >= 0x00000001) { + cpuid(0x00000001, &eax, &ebx, &ecx, &edx); + + c->x86_capability[CPUID_1_ECX] = ecx; + c->x86_capability[CPUID_1_EDX] = edx; + } + + /* Thermal and Power Management Leaf: level 0x00000006 (eax) */ + if (c->cpuid_level >= 0x00000006) + c->x86_capability[CPUID_6_EAX] = cpuid_eax(0x00000006); + + /* Additional Intel-defined flags: level 0x00000007 */ + if (c->cpuid_level >= 0x00000007) { + cpuid_count(0x00000007, 0, &eax, &ebx, &ecx, &edx); + c->x86_capability[CPUID_7_0_EBX] = ebx; + c->x86_capability[CPUID_7_ECX] = ecx; + c->x86_capability[CPUID_7_EDX] = edx; + + /* Check valid sub-leaf index before accessing it */ + if (eax >= 1) { + cpuid_count(0x00000007, 1, &eax, &ebx, &ecx, &edx); + c->x86_capability[CPUID_7_1_EAX] = eax; + } + } + + /* Extended state features: level 0x0000000d */ + if (c->cpuid_level >= 0x0000000d) { + cpuid_count(0x0000000d, 1, &eax, &ebx, &ecx, &edx); + + c->x86_capability[CPUID_D_1_EAX] = eax; + } + + /* AMD-defined flags: level 0x80000001 */ + eax = cpuid_eax(0x80000000); + c->extended_cpuid_level = eax; + + if ((eax & 0xffff0000) == 0x80000000) { + if (eax >= 0x80000001) { + cpuid(0x80000001, &eax, &ebx, &ecx, &edx); + + c->x86_capability[CPUID_8000_0001_ECX] = ecx; + c->x86_capability[CPUID_8000_0001_EDX] = edx; + } + } + + if (c->extended_cpuid_level >= 0x80000007) { + cpuid(0x80000007, &eax, &ebx, &ecx, &edx); + + c->x86_capability[CPUID_8000_0007_EBX] = ebx; + c->x86_power = edx; + } + + if (c->extended_cpuid_level >= 0x80000008) { + cpuid(0x80000008, &eax, &ebx, &ecx, &edx); + c->x86_capability[CPUID_8000_0008_EBX] = ebx; + } + + if (c->extended_cpuid_level >= 0x8000000a) + c->x86_capability[CPUID_8000_000A_EDX] = cpuid_edx(0x8000000a); + + if (c->extended_cpuid_level >= 0x8000001f) + c->x86_capability[CPUID_8000_001F_EAX] = cpuid_eax(0x8000001f); + + if (c->extended_cpuid_level >= 0x80000021) + c->x86_capability[CPUID_8000_0021_EAX] = cpuid_eax(0x80000021); + + init_scattered_cpuid_features(c); + init_speculation_control(c); + + /* + * Clear/Set all flags overridden by options, after probe. + * This needs to happen each time we re-probe, which may happen + * several times during CPU initialization. + */ + apply_forced_caps(c); +} + +void get_cpu_address_sizes(struct cpuinfo_x86 *c) +{ + u32 eax, ebx, ecx, edx; + + if (c->extended_cpuid_level >= 0x80000008) { + cpuid(0x80000008, &eax, &ebx, &ecx, &edx); + + c->x86_virt_bits = (eax >> 8) & 0xff; + c->x86_phys_bits = eax & 0xff; + } +#ifdef CONFIG_X86_32 + else if (cpu_has(c, X86_FEATURE_PAE) || cpu_has(c, X86_FEATURE_PSE36)) + c->x86_phys_bits = 36; +#endif + c->x86_cache_bits = c->x86_phys_bits; +} + +static void identify_cpu_without_cpuid(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + int i; + + /* + * First of all, decide if this is a 486 or higher + * It's a 486 if we can modify the AC flag + */ + if (flag_is_changeable_p(X86_EFLAGS_AC)) + c->x86 = 4; + else + c->x86 = 3; + + for (i = 0; i < X86_VENDOR_NUM; i++) + if (cpu_devs[i] && cpu_devs[i]->c_identify) { + c->x86_vendor_id[0] = 0; + cpu_devs[i]->c_identify(c); + if (c->x86_vendor_id[0]) { + get_cpu_vendor(c); + break; + } + } +#endif +} + +#define NO_SPECULATION BIT(0) +#define NO_MELTDOWN BIT(1) +#define NO_SSB BIT(2) +#define NO_L1TF BIT(3) +#define NO_MDS BIT(4) +#define MSBDS_ONLY BIT(5) +#define NO_SWAPGS BIT(6) +#define NO_ITLB_MULTIHIT BIT(7) +#define NO_SPECTRE_V2 BIT(8) +#define NO_MMIO BIT(9) +#define NO_EIBRS_PBRSB BIT(10) + +#define VULNWL(vendor, family, model, whitelist) \ + X86_MATCH_VENDOR_FAM_MODEL(vendor, family, model, whitelist) + +#define VULNWL_INTEL(model, whitelist) \ + VULNWL(INTEL, 6, INTEL_FAM6_##model, whitelist) + +#define VULNWL_AMD(family, whitelist) \ + VULNWL(AMD, family, X86_MODEL_ANY, whitelist) + +#define VULNWL_HYGON(family, whitelist) \ + VULNWL(HYGON, family, X86_MODEL_ANY, whitelist) + +static const __initconst struct x86_cpu_id cpu_vuln_whitelist[] = { + VULNWL(ANY, 4, X86_MODEL_ANY, NO_SPECULATION), + VULNWL(CENTAUR, 5, X86_MODEL_ANY, NO_SPECULATION), + VULNWL(INTEL, 5, X86_MODEL_ANY, NO_SPECULATION), + VULNWL(NSC, 5, X86_MODEL_ANY, NO_SPECULATION), + + /* Intel Family 6 */ + VULNWL_INTEL(TIGERLAKE, NO_MMIO), + VULNWL_INTEL(TIGERLAKE_L, NO_MMIO), + VULNWL_INTEL(ALDERLAKE, NO_MMIO), + VULNWL_INTEL(ALDERLAKE_L, NO_MMIO), + + VULNWL_INTEL(ATOM_SALTWELL, NO_SPECULATION | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_SALTWELL_TABLET, NO_SPECULATION | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_SALTWELL_MID, NO_SPECULATION | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_BONNELL, NO_SPECULATION | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_BONNELL_MID, NO_SPECULATION | NO_ITLB_MULTIHIT), + + VULNWL_INTEL(ATOM_SILVERMONT, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_SILVERMONT_D, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_SILVERMONT_MID, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_AIRMONT, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(XEON_PHI_KNL, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(XEON_PHI_KNM, NO_SSB | NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + + VULNWL_INTEL(CORE_YONAH, NO_SSB), + + VULNWL_INTEL(ATOM_AIRMONT_MID, NO_L1TF | MSBDS_ONLY | NO_SWAPGS | NO_ITLB_MULTIHIT), + VULNWL_INTEL(ATOM_AIRMONT_NP, NO_L1TF | NO_SWAPGS | NO_ITLB_MULTIHIT), + + VULNWL_INTEL(ATOM_GOLDMONT, NO_MDS | NO_L1TF | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_INTEL(ATOM_GOLDMONT_D, NO_MDS | NO_L1TF | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_INTEL(ATOM_GOLDMONT_PLUS, NO_MDS | NO_L1TF | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO | NO_EIBRS_PBRSB), + + /* + * Technically, swapgs isn't serializing on AMD (despite it previously + * being documented as such in the APM). But according to AMD, %gs is + * updated non-speculatively, and the issuing of %gs-relative memory + * operands will be blocked until the %gs update completes, which is + * good enough for our purposes. + */ + + VULNWL_INTEL(ATOM_TREMONT, NO_EIBRS_PBRSB), + VULNWL_INTEL(ATOM_TREMONT_L, NO_EIBRS_PBRSB), + VULNWL_INTEL(ATOM_TREMONT_D, NO_ITLB_MULTIHIT | NO_EIBRS_PBRSB), + + /* AMD Family 0xf - 0x12 */ + VULNWL_AMD(0x0f, NO_MELTDOWN | NO_SSB | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_AMD(0x10, NO_MELTDOWN | NO_SSB | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_AMD(0x11, NO_MELTDOWN | NO_SSB | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_AMD(0x12, NO_MELTDOWN | NO_SSB | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + + /* FAMILY_ANY must be last, otherwise 0x0f - 0x12 matches won't work */ + VULNWL_AMD(X86_FAMILY_ANY, NO_MELTDOWN | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + VULNWL_HYGON(X86_FAMILY_ANY, NO_MELTDOWN | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT | NO_MMIO), + + /* Zhaoxin Family 7 */ + VULNWL(CENTAUR, 7, X86_MODEL_ANY, NO_SPECTRE_V2 | NO_SWAPGS | NO_MMIO), + VULNWL(ZHAOXIN, 7, X86_MODEL_ANY, NO_SPECTRE_V2 | NO_SWAPGS | NO_MMIO), + {} +}; + +#define VULNBL(vendor, family, model, blacklist) \ + X86_MATCH_VENDOR_FAM_MODEL(vendor, family, model, blacklist) + +#define VULNBL_INTEL_STEPPINGS(model, steppings, issues) \ + X86_MATCH_VENDOR_FAM_MODEL_STEPPINGS_FEATURE(INTEL, 6, \ + INTEL_FAM6_##model, steppings, \ + X86_FEATURE_ANY, issues) + +#define VULNBL_AMD(family, blacklist) \ + VULNBL(AMD, family, X86_MODEL_ANY, blacklist) + +#define VULNBL_HYGON(family, blacklist) \ + VULNBL(HYGON, family, X86_MODEL_ANY, blacklist) + +#define SRBDS BIT(0) +/* CPU is affected by X86_BUG_MMIO_STALE_DATA */ +#define MMIO BIT(1) +/* CPU is affected by Shared Buffers Data Sampling (SBDS), a variant of X86_BUG_MMIO_STALE_DATA */ +#define MMIO_SBDS BIT(2) +/* CPU is affected by RETbleed, speculating where you would not expect it */ +#define RETBLEED BIT(3) +/* CPU is affected by SMT (cross-thread) return predictions */ +#define SMT_RSB BIT(4) +/* CPU is affected by SRSO */ +#define SRSO BIT(5) +/* CPU is affected by GDS */ +#define GDS BIT(6) + +static const struct x86_cpu_id cpu_vuln_blacklist[] __initconst = { + VULNBL_INTEL_STEPPINGS(IVYBRIDGE, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(HASWELL, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(HASWELL_L, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(HASWELL_G, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(HASWELL_X, X86_STEPPING_ANY, MMIO), + VULNBL_INTEL_STEPPINGS(BROADWELL_D, X86_STEPPING_ANY, MMIO), + VULNBL_INTEL_STEPPINGS(BROADWELL_G, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(BROADWELL_X, X86_STEPPING_ANY, MMIO), + VULNBL_INTEL_STEPPINGS(BROADWELL, X86_STEPPING_ANY, SRBDS), + VULNBL_INTEL_STEPPINGS(SKYLAKE_X, X86_STEPPING_ANY, MMIO | RETBLEED | GDS), + VULNBL_INTEL_STEPPINGS(SKYLAKE_L, X86_STEPPING_ANY, MMIO | RETBLEED | GDS | SRBDS), + VULNBL_INTEL_STEPPINGS(SKYLAKE, X86_STEPPING_ANY, MMIO | RETBLEED | GDS | SRBDS), + VULNBL_INTEL_STEPPINGS(KABYLAKE_L, X86_STEPPING_ANY, MMIO | RETBLEED | GDS | SRBDS), + VULNBL_INTEL_STEPPINGS(KABYLAKE, X86_STEPPING_ANY, MMIO | RETBLEED | GDS | SRBDS), + VULNBL_INTEL_STEPPINGS(CANNONLAKE_L, X86_STEPPING_ANY, RETBLEED), + VULNBL_INTEL_STEPPINGS(ICELAKE_L, X86_STEPPING_ANY, MMIO | MMIO_SBDS | RETBLEED | GDS), + VULNBL_INTEL_STEPPINGS(ICELAKE_D, X86_STEPPING_ANY, MMIO | GDS), + VULNBL_INTEL_STEPPINGS(ICELAKE_X, X86_STEPPING_ANY, MMIO | GDS), + VULNBL_INTEL_STEPPINGS(COMETLAKE, X86_STEPPING_ANY, MMIO | MMIO_SBDS | RETBLEED | GDS), + VULNBL_INTEL_STEPPINGS(COMETLAKE_L, X86_STEPPINGS(0x0, 0x0), MMIO | RETBLEED), + VULNBL_INTEL_STEPPINGS(COMETLAKE_L, X86_STEPPING_ANY, MMIO | MMIO_SBDS | RETBLEED | GDS), + VULNBL_INTEL_STEPPINGS(TIGERLAKE_L, X86_STEPPING_ANY, GDS), + VULNBL_INTEL_STEPPINGS(TIGERLAKE, X86_STEPPING_ANY, GDS), + VULNBL_INTEL_STEPPINGS(LAKEFIELD, X86_STEPPING_ANY, MMIO | MMIO_SBDS | RETBLEED), + VULNBL_INTEL_STEPPINGS(ROCKETLAKE, X86_STEPPING_ANY, MMIO | RETBLEED | GDS), + VULNBL_INTEL_STEPPINGS(ATOM_TREMONT, X86_STEPPING_ANY, MMIO | MMIO_SBDS), + VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_D, X86_STEPPING_ANY, MMIO), + VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_L, X86_STEPPING_ANY, MMIO | MMIO_SBDS), + + VULNBL_AMD(0x15, RETBLEED), + VULNBL_AMD(0x16, RETBLEED), + VULNBL_AMD(0x17, RETBLEED | SRSO), + VULNBL_HYGON(0x18, RETBLEED), + VULNBL_AMD(0x19, SRSO), + {} +}; + +static bool __init cpu_matches(const struct x86_cpu_id *table, unsigned long which) +{ + const struct x86_cpu_id *m = x86_match_cpu(table); + + return m && !!(m->driver_data & which); +} + +u64 x86_read_arch_cap_msr(void) +{ + u64 ia32_cap = 0; + + if (boot_cpu_has(X86_FEATURE_ARCH_CAPABILITIES)) + rdmsrl(MSR_IA32_ARCH_CAPABILITIES, ia32_cap); + + return ia32_cap; +} + +static bool arch_cap_mmio_immune(u64 ia32_cap) +{ + return (ia32_cap & ARCH_CAP_FBSDP_NO && + ia32_cap & ARCH_CAP_PSDP_NO && + ia32_cap & ARCH_CAP_SBDR_SSDP_NO); +} + +static void __init cpu_set_bug_bits(struct cpuinfo_x86 *c) +{ + u64 ia32_cap = x86_read_arch_cap_msr(); + + /* Set ITLB_MULTIHIT bug if cpu is not in the whitelist and not mitigated */ + if (!cpu_matches(cpu_vuln_whitelist, NO_ITLB_MULTIHIT) && + !(ia32_cap & ARCH_CAP_PSCHANGE_MC_NO)) + setup_force_cpu_bug(X86_BUG_ITLB_MULTIHIT); + + if (cpu_matches(cpu_vuln_whitelist, NO_SPECULATION)) + return; + + setup_force_cpu_bug(X86_BUG_SPECTRE_V1); + + if (!cpu_matches(cpu_vuln_whitelist, NO_SPECTRE_V2)) + setup_force_cpu_bug(X86_BUG_SPECTRE_V2); + + if (!cpu_matches(cpu_vuln_whitelist, NO_SSB) && + !(ia32_cap & ARCH_CAP_SSB_NO) && + !cpu_has(c, X86_FEATURE_AMD_SSB_NO)) + setup_force_cpu_bug(X86_BUG_SPEC_STORE_BYPASS); + + if (ia32_cap & ARCH_CAP_IBRS_ALL) + setup_force_cpu_cap(X86_FEATURE_IBRS_ENHANCED); + + if (!cpu_matches(cpu_vuln_whitelist, NO_MDS) && + !(ia32_cap & ARCH_CAP_MDS_NO)) { + setup_force_cpu_bug(X86_BUG_MDS); + if (cpu_matches(cpu_vuln_whitelist, MSBDS_ONLY)) + setup_force_cpu_bug(X86_BUG_MSBDS_ONLY); + } + + if (!cpu_matches(cpu_vuln_whitelist, NO_SWAPGS)) + setup_force_cpu_bug(X86_BUG_SWAPGS); + + /* + * When the CPU is not mitigated for TAA (TAA_NO=0) set TAA bug when: + * - TSX is supported or + * - TSX_CTRL is present + * + * TSX_CTRL check is needed for cases when TSX could be disabled before + * the kernel boot e.g. kexec. + * TSX_CTRL check alone is not sufficient for cases when the microcode + * update is not present or running as guest that don't get TSX_CTRL. + */ + if (!(ia32_cap & ARCH_CAP_TAA_NO) && + (cpu_has(c, X86_FEATURE_RTM) || + (ia32_cap & ARCH_CAP_TSX_CTRL_MSR))) + setup_force_cpu_bug(X86_BUG_TAA); + + /* + * SRBDS affects CPUs which support RDRAND or RDSEED and are listed + * in the vulnerability blacklist. + * + * Some of the implications and mitigation of Shared Buffers Data + * Sampling (SBDS) are similar to SRBDS. Give SBDS same treatment as + * SRBDS. + */ + if ((cpu_has(c, X86_FEATURE_RDRAND) || + cpu_has(c, X86_FEATURE_RDSEED)) && + cpu_matches(cpu_vuln_blacklist, SRBDS | MMIO_SBDS)) + setup_force_cpu_bug(X86_BUG_SRBDS); + + /* + * Processor MMIO Stale Data bug enumeration + * + * Affected CPU list is generally enough to enumerate the vulnerability, + * but for virtualization case check for ARCH_CAP MSR bits also, VMM may + * not want the guest to enumerate the bug. + * + * Set X86_BUG_MMIO_UNKNOWN for CPUs that are neither in the blacklist, + * nor in the whitelist and also don't enumerate MSR ARCH_CAP MMIO bits. + */ + if (!arch_cap_mmio_immune(ia32_cap)) { + if (cpu_matches(cpu_vuln_blacklist, MMIO)) + setup_force_cpu_bug(X86_BUG_MMIO_STALE_DATA); + else if (!cpu_matches(cpu_vuln_whitelist, NO_MMIO)) + setup_force_cpu_bug(X86_BUG_MMIO_UNKNOWN); + } + + if (!cpu_has(c, X86_FEATURE_BTC_NO)) { + if (cpu_matches(cpu_vuln_blacklist, RETBLEED) || (ia32_cap & ARCH_CAP_RSBA)) + setup_force_cpu_bug(X86_BUG_RETBLEED); + } + + if (cpu_has(c, X86_FEATURE_IBRS_ENHANCED) && + !cpu_matches(cpu_vuln_whitelist, NO_EIBRS_PBRSB) && + !(ia32_cap & ARCH_CAP_PBRSB_NO)) + setup_force_cpu_bug(X86_BUG_EIBRS_PBRSB); + + /* + * Check if CPU is vulnerable to GDS. If running in a virtual machine on + * an affected processor, the VMM may have disabled the use of GATHER by + * disabling AVX2. The only way to do this in HW is to clear XCR0[2], + * which means that AVX will be disabled. + */ + if (cpu_matches(cpu_vuln_blacklist, GDS) && !(ia32_cap & ARCH_CAP_GDS_NO) && + boot_cpu_has(X86_FEATURE_AVX)) + setup_force_cpu_bug(X86_BUG_GDS); + + if (!cpu_has(c, X86_FEATURE_SRSO_NO)) { + if (cpu_matches(cpu_vuln_blacklist, SRSO)) + setup_force_cpu_bug(X86_BUG_SRSO); + } + + if (cpu_matches(cpu_vuln_whitelist, NO_MELTDOWN)) + return; + + /* Rogue Data Cache Load? No! */ + if (ia32_cap & ARCH_CAP_RDCL_NO) + return; + + setup_force_cpu_bug(X86_BUG_CPU_MELTDOWN); + + if (cpu_matches(cpu_vuln_whitelist, NO_L1TF)) + return; + + setup_force_cpu_bug(X86_BUG_L1TF); +} + +/* + * The NOPL instruction is supposed to exist on all CPUs of family >= 6; + * unfortunately, that's not true in practice because of early VIA + * chips and (more importantly) broken virtualizers that are not easy + * to detect. In the latter case it doesn't even *fail* reliably, so + * probing for it doesn't even work. Disable it completely on 32-bit + * unless we can find a reliable way to detect all the broken cases. + * Enable it explicitly on 64-bit for non-constant inputs of cpu_has(). + */ +static void detect_nopl(void) +{ +#ifdef CONFIG_X86_32 + setup_clear_cpu_cap(X86_FEATURE_NOPL); +#else + setup_force_cpu_cap(X86_FEATURE_NOPL); +#endif +} + +/* + * We parse cpu parameters early because fpu__init_system() is executed + * before parse_early_param(). + */ +static void __init cpu_parse_early_param(void) +{ + char arg[128]; + char *argptr = arg; + int arglen, res, bit; + +#ifdef CONFIG_X86_32 + if (cmdline_find_option_bool(boot_command_line, "no387")) +#ifdef CONFIG_MATH_EMULATION + setup_clear_cpu_cap(X86_FEATURE_FPU); +#else + pr_err("Option 'no387' required CONFIG_MATH_EMULATION enabled.\n"); +#endif + + if (cmdline_find_option_bool(boot_command_line, "nofxsr")) + setup_clear_cpu_cap(X86_FEATURE_FXSR); +#endif + + if (cmdline_find_option_bool(boot_command_line, "noxsave")) + setup_clear_cpu_cap(X86_FEATURE_XSAVE); + + if (cmdline_find_option_bool(boot_command_line, "noxsaveopt")) + setup_clear_cpu_cap(X86_FEATURE_XSAVEOPT); + + if (cmdline_find_option_bool(boot_command_line, "noxsaves")) + setup_clear_cpu_cap(X86_FEATURE_XSAVES); + + arglen = cmdline_find_option(boot_command_line, "clearcpuid", arg, sizeof(arg)); + if (arglen <= 0) + return; + + pr_info("Clearing CPUID bits:"); + do { + res = get_option(&argptr, &bit); + if (res == 0 || res == 3) + break; + + /* If the argument was too long, the last bit may be cut off */ + if (res == 1 && arglen >= sizeof(arg)) + break; + + if (bit >= 0 && bit < NCAPINTS * 32) { + pr_cont(" " X86_CAP_FMT, x86_cap_flag(bit)); + setup_clear_cpu_cap(bit); + } + } while (res == 2); + pr_cont("\n"); +} + +/* + * Do minimum CPU detection early. + * Fields really needed: vendor, cpuid_level, family, model, mask, + * cache alignment. + * The others are not touched to avoid unwanted side effects. + * + * WARNING: this function is only called on the boot CPU. Don't add code + * here that is supposed to run on all CPUs. + */ +static void __init early_identify_cpu(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_64 + c->x86_clflush_size = 64; + c->x86_phys_bits = 36; + c->x86_virt_bits = 48; +#else + c->x86_clflush_size = 32; + c->x86_phys_bits = 32; + c->x86_virt_bits = 32; +#endif + c->x86_cache_alignment = c->x86_clflush_size; + + memset(&c->x86_capability, 0, sizeof(c->x86_capability)); + c->extended_cpuid_level = 0; + + if (!have_cpuid_p()) + identify_cpu_without_cpuid(c); + + /* cyrix could have cpuid enabled via c_identify()*/ + if (have_cpuid_p()) { + cpu_detect(c); + get_cpu_vendor(c); + get_cpu_cap(c); + get_cpu_address_sizes(c); + setup_force_cpu_cap(X86_FEATURE_CPUID); + cpu_parse_early_param(); + + if (this_cpu->c_early_init) + this_cpu->c_early_init(c); + + c->cpu_index = 0; + filter_cpuid_features(c, false); + + if (this_cpu->c_bsp_init) + this_cpu->c_bsp_init(c); + } else { + setup_clear_cpu_cap(X86_FEATURE_CPUID); + } + + setup_force_cpu_cap(X86_FEATURE_ALWAYS); + + cpu_set_bug_bits(c); + + cpu_set_core_cap_bits(c); + +#ifdef CONFIG_X86_32 + /* + * Regardless of whether PCID is enumerated, the SDM says + * that it can't be enabled in 32-bit mode. + */ + setup_clear_cpu_cap(X86_FEATURE_PCID); +#endif + + /* + * Later in the boot process pgtable_l5_enabled() relies on + * cpu_feature_enabled(X86_FEATURE_LA57). If 5-level paging is not + * enabled by this point we need to clear the feature bit to avoid + * false-positives at the later stage. + * + * pgtable_l5_enabled() can be false here for several reasons: + * - 5-level paging is disabled compile-time; + * - it's 32-bit kernel; + * - machine doesn't support 5-level paging; + * - user specified 'no5lvl' in kernel command line. + */ + if (!pgtable_l5_enabled()) + setup_clear_cpu_cap(X86_FEATURE_LA57); + + detect_nopl(); +} + +void __init early_cpu_init(void) +{ + const struct cpu_dev *const *cdev; + int count = 0; + +#ifdef CONFIG_PROCESSOR_SELECT + pr_info("KERNEL supported cpus:\n"); +#endif + + for (cdev = __x86_cpu_dev_start; cdev < __x86_cpu_dev_end; cdev++) { + const struct cpu_dev *cpudev = *cdev; + + if (count >= X86_VENDOR_NUM) + break; + cpu_devs[count] = cpudev; + count++; + +#ifdef CONFIG_PROCESSOR_SELECT + { + unsigned int j; + + for (j = 0; j < 2; j++) { + if (!cpudev->c_ident[j]) + continue; + pr_info(" %s %s\n", cpudev->c_vendor, + cpudev->c_ident[j]); + } + } +#endif + } + early_identify_cpu(&boot_cpu_data); +} + +static bool detect_null_seg_behavior(void) +{ + /* + * Empirically, writing zero to a segment selector on AMD does + * not clear the base, whereas writing zero to a segment + * selector on Intel does clear the base. Intel's behavior + * allows slightly faster context switches in the common case + * where GS is unused by the prev and next threads. + * + * Since neither vendor documents this anywhere that I can see, + * detect it directly instead of hardcoding the choice by + * vendor. + * + * I've designated AMD's behavior as the "bug" because it's + * counterintuitive and less friendly. + */ + + unsigned long old_base, tmp; + rdmsrl(MSR_FS_BASE, old_base); + wrmsrl(MSR_FS_BASE, 1); + loadsegment(fs, 0); + rdmsrl(MSR_FS_BASE, tmp); + wrmsrl(MSR_FS_BASE, old_base); + return tmp == 0; +} + +void check_null_seg_clears_base(struct cpuinfo_x86 *c) +{ + /* BUG_NULL_SEG is only relevant with 64bit userspace */ + if (!IS_ENABLED(CONFIG_X86_64)) + return; + + /* Zen3 CPUs advertise Null Selector Clears Base in CPUID. */ + if (c->extended_cpuid_level >= 0x80000021 && + cpuid_eax(0x80000021) & BIT(6)) + return; + + /* + * CPUID bit above wasn't set. If this kernel is still running + * as a HV guest, then the HV has decided not to advertize + * that CPUID bit for whatever reason. For example, one + * member of the migration pool might be vulnerable. Which + * means, the bug is present: set the BUG flag and return. + */ + if (cpu_has(c, X86_FEATURE_HYPERVISOR)) { + set_cpu_bug(c, X86_BUG_NULL_SEG); + return; + } + + /* + * Zen2 CPUs also have this behaviour, but no CPUID bit. + * 0x18 is the respective family for Hygon. + */ + if ((c->x86 == 0x17 || c->x86 == 0x18) && + detect_null_seg_behavior()) + return; + + /* All the remaining ones are affected */ + set_cpu_bug(c, X86_BUG_NULL_SEG); +} + +static void generic_identify(struct cpuinfo_x86 *c) +{ + c->extended_cpuid_level = 0; + + if (!have_cpuid_p()) + identify_cpu_without_cpuid(c); + + /* cyrix could have cpuid enabled via c_identify()*/ + if (!have_cpuid_p()) + return; + + cpu_detect(c); + + get_cpu_vendor(c); + + get_cpu_cap(c); + + get_cpu_address_sizes(c); + + if (c->cpuid_level >= 0x00000001) { + c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xFF; +#ifdef CONFIG_X86_32 +# ifdef CONFIG_SMP + c->apicid = apic->phys_pkg_id(c->initial_apicid, 0); +# else + c->apicid = c->initial_apicid; +# endif +#endif + c->phys_proc_id = c->initial_apicid; + } + + get_model_name(c); /* Default name */ + + /* + * ESPFIX is a strange bug. All real CPUs have it. Paravirt + * systems that run Linux at CPL > 0 may or may not have the + * issue, but, even if they have the issue, there's absolutely + * nothing we can do about it because we can't use the real IRET + * instruction. + * + * NB: For the time being, only 32-bit kernels support + * X86_BUG_ESPFIX as such. 64-bit kernels directly choose + * whether to apply espfix using paravirt hooks. If any + * non-paravirt system ever shows up that does *not* have the + * ESPFIX issue, we can change this. + */ +#ifdef CONFIG_X86_32 + set_cpu_bug(c, X86_BUG_ESPFIX); +#endif +} + +/* + * Validate that ACPI/mptables have the same information about the + * effective APIC id and update the package map. + */ +static void validate_apic_and_package_id(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int apicid, cpu = smp_processor_id(); + + apicid = apic->cpu_present_to_apicid(cpu); + + if (apicid != c->apicid) { + pr_err(FW_BUG "CPU%u: APIC id mismatch. Firmware: %x APIC: %x\n", + cpu, apicid, c->initial_apicid); + } + BUG_ON(topology_update_package_map(c->phys_proc_id, cpu)); + BUG_ON(topology_update_die_map(c->cpu_die_id, cpu)); +#else + c->logical_proc_id = 0; +#endif +} + +/* + * This does the hard work of actually picking apart the CPU stuff... + */ +static void identify_cpu(struct cpuinfo_x86 *c) +{ + int i; + + c->loops_per_jiffy = loops_per_jiffy; + c->x86_cache_size = 0; + c->x86_vendor = X86_VENDOR_UNKNOWN; + c->x86_model = c->x86_stepping = 0; /* So far unknown... */ + c->x86_vendor_id[0] = '\0'; /* Unset */ + c->x86_model_id[0] = '\0'; /* Unset */ + c->x86_max_cores = 1; + c->x86_coreid_bits = 0; + c->cu_id = 0xff; +#ifdef CONFIG_X86_64 + c->x86_clflush_size = 64; + c->x86_phys_bits = 36; + c->x86_virt_bits = 48; +#else + c->cpuid_level = -1; /* CPUID not detected */ + c->x86_clflush_size = 32; + c->x86_phys_bits = 32; + c->x86_virt_bits = 32; +#endif + c->x86_cache_alignment = c->x86_clflush_size; + memset(&c->x86_capability, 0, sizeof(c->x86_capability)); +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + memset(&c->vmx_capability, 0, sizeof(c->vmx_capability)); +#endif + + generic_identify(c); + + if (this_cpu->c_identify) + this_cpu->c_identify(c); + + /* Clear/Set all flags overridden by options, after probe */ + apply_forced_caps(c); + +#ifdef CONFIG_X86_64 + c->apicid = apic->phys_pkg_id(c->initial_apicid, 0); +#endif + + /* + * Vendor-specific initialization. In this section we + * canonicalize the feature flags, meaning if there are + * features a certain CPU supports which CPUID doesn't + * tell us, CPUID claiming incorrect flags, or other bugs, + * we handle them here. + * + * At the end of this section, c->x86_capability better + * indicate the features this CPU genuinely supports! + */ + if (this_cpu->c_init) + this_cpu->c_init(c); + + /* Disable the PN if appropriate */ + squash_the_stupid_serial_number(c); + + /* Set up SMEP/SMAP/UMIP */ + setup_smep(c); + setup_smap(c); + setup_umip(c); + + /* Enable FSGSBASE instructions if available. */ + if (cpu_has(c, X86_FEATURE_FSGSBASE)) { + cr4_set_bits(X86_CR4_FSGSBASE); + elf_hwcap2 |= HWCAP2_FSGSBASE; + } + + /* + * The vendor-specific functions might have changed features. + * Now we do "generic changes." + */ + + /* Filter out anything that depends on CPUID levels we don't have */ + filter_cpuid_features(c, true); + + /* If the model name is still unset, do table lookup. */ + if (!c->x86_model_id[0]) { + const char *p; + p = table_lookup_model(c); + if (p) + strcpy(c->x86_model_id, p); + else + /* Last resort... */ + sprintf(c->x86_model_id, "%02x/%02x", + c->x86, c->x86_model); + } + +#ifdef CONFIG_X86_64 + detect_ht(c); +#endif + + x86_init_rdrand(c); + setup_pku(c); + + /* + * Clear/Set all flags overridden by options, need do it + * before following smp all cpus cap AND. + */ + apply_forced_caps(c); + + /* + * On SMP, boot_cpu_data holds the common feature set between + * all CPUs; so make sure that we indicate which features are + * common between the CPUs. The first time this routine gets + * executed, c == &boot_cpu_data. + */ + if (c != &boot_cpu_data) { + /* AND the already accumulated flags with these */ + for (i = 0; i < NCAPINTS; i++) + boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; + + /* OR, i.e. replicate the bug flags */ + for (i = NCAPINTS; i < NCAPINTS + NBUGINTS; i++) + c->x86_capability[i] |= boot_cpu_data.x86_capability[i]; + } + + /* Init Machine Check Exception if available. */ + mcheck_cpu_init(c); + + select_idle_routine(c); + +#ifdef CONFIG_NUMA + numa_add_cpu(smp_processor_id()); +#endif +} + +/* + * Set up the CPU state needed to execute SYSENTER/SYSEXIT instructions + * on 32-bit kernels: + */ +#ifdef CONFIG_X86_32 +void enable_sep_cpu(void) +{ + struct tss_struct *tss; + int cpu; + + if (!boot_cpu_has(X86_FEATURE_SEP)) + return; + + cpu = get_cpu(); + tss = &per_cpu(cpu_tss_rw, cpu); + + /* + * We cache MSR_IA32_SYSENTER_CS's value in the TSS's ss1 field -- + * see the big comment in struct x86_hw_tss's definition. + */ + + tss->x86_tss.ss1 = __KERNEL_CS; + wrmsr(MSR_IA32_SYSENTER_CS, tss->x86_tss.ss1, 0); + wrmsr(MSR_IA32_SYSENTER_ESP, (unsigned long)(cpu_entry_stack(cpu) + 1), 0); + wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long)entry_SYSENTER_32, 0); + + put_cpu(); +} +#endif + +void __init identify_boot_cpu(void) +{ + identify_cpu(&boot_cpu_data); +#ifdef CONFIG_X86_32 + sysenter_setup(); + enable_sep_cpu(); +#endif + cpu_detect_tlb(&boot_cpu_data); + setup_cr_pinning(); + + tsx_init(); +} + +void identify_secondary_cpu(struct cpuinfo_x86 *c) +{ + BUG_ON(c == &boot_cpu_data); + identify_cpu(c); +#ifdef CONFIG_X86_32 + enable_sep_cpu(); +#endif + mtrr_ap_init(); + validate_apic_and_package_id(c); + x86_spec_ctrl_setup_ap(); + update_srbds_msr(); + if (boot_cpu_has_bug(X86_BUG_GDS)) + update_gds_msr(); +} + +static __init int setup_noclflush(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_CLFLUSH); + setup_clear_cpu_cap(X86_FEATURE_CLFLUSHOPT); + return 1; +} +__setup("noclflush", setup_noclflush); + +void print_cpu_info(struct cpuinfo_x86 *c) +{ + const char *vendor = NULL; + + if (c->x86_vendor < X86_VENDOR_NUM) { + vendor = this_cpu->c_vendor; + } else { + if (c->cpuid_level >= 0) + vendor = c->x86_vendor_id; + } + + if (vendor && !strstr(c->x86_model_id, vendor)) + pr_cont("%s ", vendor); + + if (c->x86_model_id[0]) + pr_cont("%s", c->x86_model_id); + else + pr_cont("%d86", c->x86); + + pr_cont(" (family: 0x%x, model: 0x%x", c->x86, c->x86_model); + + if (c->x86_stepping || c->cpuid_level >= 0) + pr_cont(", stepping: 0x%x)\n", c->x86_stepping); + else + pr_cont(")\n"); +} + +/* + * clearcpuid= was already parsed in fpu__init_parse_early_param. + * But we need to keep a dummy __setup around otherwise it would + * show up as an environment variable for init. + */ +static __init int setup_clearcpuid(char *arg) +{ + return 1; +} +__setup("clearcpuid=", setup_clearcpuid); + +#ifdef CONFIG_X86_64 +DEFINE_PER_CPU_FIRST(struct fixed_percpu_data, + fixed_percpu_data) __aligned(PAGE_SIZE) __visible; +EXPORT_PER_CPU_SYMBOL_GPL(fixed_percpu_data); + +/* + * The following percpu variables are hot. Align current_task to + * cacheline size such that they fall in the same cacheline. + */ +DEFINE_PER_CPU(struct task_struct *, current_task) ____cacheline_aligned = + &init_task; +EXPORT_PER_CPU_SYMBOL(current_task); + +DEFINE_PER_CPU(struct irq_stack *, hardirq_stack_ptr); +DEFINE_PER_CPU(unsigned int, irq_count) __visible = -1; + +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); + +/* May not be marked __init: used by software suspend */ +void syscall_init(void) +{ + wrmsr(MSR_STAR, 0, (__USER32_CS << 16) | __KERNEL_CS); + wrmsrl(MSR_LSTAR, (unsigned long)entry_SYSCALL_64); + +#ifdef CONFIG_IA32_EMULATION + wrmsrl(MSR_CSTAR, (unsigned long)entry_SYSCALL_compat); + /* + * This only works on Intel CPUs. + * On AMD CPUs these MSRs are 32-bit, CPU truncates MSR_IA32_SYSENTER_EIP. + * This does not cause SYSENTER to jump to the wrong location, because + * AMD doesn't allow SYSENTER in long mode (either 32- or 64-bit). + */ + wrmsrl_safe(MSR_IA32_SYSENTER_CS, (u64)__KERNEL_CS); + wrmsrl_safe(MSR_IA32_SYSENTER_ESP, + (unsigned long)(cpu_entry_stack(smp_processor_id()) + 1)); + wrmsrl_safe(MSR_IA32_SYSENTER_EIP, (u64)entry_SYSENTER_compat); +#else + wrmsrl(MSR_CSTAR, (unsigned long)ignore_sysret); + wrmsrl_safe(MSR_IA32_SYSENTER_CS, (u64)GDT_ENTRY_INVALID_SEG); + wrmsrl_safe(MSR_IA32_SYSENTER_ESP, 0ULL); + wrmsrl_safe(MSR_IA32_SYSENTER_EIP, 0ULL); +#endif + + /* Flags to clear on syscall */ + wrmsrl(MSR_SYSCALL_MASK, + X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF| + X86_EFLAGS_IOPL|X86_EFLAGS_AC|X86_EFLAGS_NT); +} + +#else /* CONFIG_X86_64 */ + +DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; +EXPORT_PER_CPU_SYMBOL(current_task); +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); + +/* + * On x86_32, vm86 modifies tss.sp0, so sp0 isn't a reliable way to find + * the top of the kernel stack. Use an extra percpu variable to track the + * top of the kernel stack directly. + */ +DEFINE_PER_CPU(unsigned long, cpu_current_top_of_stack) = + (unsigned long)&init_thread_union + THREAD_SIZE; +EXPORT_PER_CPU_SYMBOL(cpu_current_top_of_stack); + +#ifdef CONFIG_STACKPROTECTOR +DEFINE_PER_CPU_ALIGNED(struct stack_canary, stack_canary); +#endif + +#endif /* CONFIG_X86_64 */ + +/* + * Clear all 6 debug registers: + */ +static void clear_all_debug_regs(void) +{ + int i; + + for (i = 0; i < 8; i++) { + /* Ignore db4, db5 */ + if ((i == 4) || (i == 5)) + continue; + + set_debugreg(0, i); + } +} + +#ifdef CONFIG_KGDB +/* + * Restore debug regs if using kgdbwait and you have a kernel debugger + * connection established. + */ +static void dbg_restore_debug_regs(void) +{ + if (unlikely(kgdb_connected && arch_kgdb_ops.correct_hw_break)) + arch_kgdb_ops.correct_hw_break(); +} +#else /* ! CONFIG_KGDB */ +#define dbg_restore_debug_regs() +#endif /* ! CONFIG_KGDB */ + +static void wait_for_master_cpu(int cpu) +{ +#ifdef CONFIG_SMP + /* + * wait for ACK from master CPU before continuing + * with AP initialization + */ + WARN_ON(cpumask_test_and_set_cpu(cpu, cpu_initialized_mask)); + while (!cpumask_test_cpu(cpu, cpu_callout_mask)) + cpu_relax(); +#endif +} + +#ifdef CONFIG_X86_64 +static inline void setup_getcpu(int cpu) +{ + unsigned long cpudata = vdso_encode_cpunode(cpu, early_cpu_to_node(cpu)); + struct desc_struct d = { }; + + if (boot_cpu_has(X86_FEATURE_RDTSCP) || boot_cpu_has(X86_FEATURE_RDPID)) + write_rdtscp_aux(cpudata); + + /* Store CPU and node number in limit. */ + d.limit0 = cpudata; + d.limit1 = cpudata >> 16; + + d.type = 5; /* RO data, expand down, accessed */ + d.dpl = 3; /* Visible to user code */ + d.s = 1; /* Not a system segment */ + d.p = 1; /* Present */ + d.d = 1; /* 32-bit */ + + write_gdt_entry(get_cpu_gdt_rw(cpu), GDT_ENTRY_CPUNODE, &d, DESCTYPE_S); +} + +static inline void ucode_cpu_init(int cpu) +{ + if (cpu) + load_ucode_ap(); +} + +static inline void tss_setup_ist(struct tss_struct *tss) +{ + /* Set up the per-CPU TSS IST stacks */ + tss->x86_tss.ist[IST_INDEX_DF] = __this_cpu_ist_top_va(DF); + tss->x86_tss.ist[IST_INDEX_NMI] = __this_cpu_ist_top_va(NMI); + tss->x86_tss.ist[IST_INDEX_DB] = __this_cpu_ist_top_va(DB); + tss->x86_tss.ist[IST_INDEX_MCE] = __this_cpu_ist_top_va(MCE); + /* Only mapped when SEV-ES is active */ + tss->x86_tss.ist[IST_INDEX_VC] = __this_cpu_ist_top_va(VC); +} + +#else /* CONFIG_X86_64 */ + +static inline void setup_getcpu(int cpu) { } + +static inline void ucode_cpu_init(int cpu) +{ + show_ucode_info_early(); +} + +static inline void tss_setup_ist(struct tss_struct *tss) { } + +#endif /* !CONFIG_X86_64 */ + +static inline void tss_setup_io_bitmap(struct tss_struct *tss) +{ + tss->x86_tss.io_bitmap_base = IO_BITMAP_OFFSET_INVALID; + +#ifdef CONFIG_X86_IOPL_IOPERM + tss->io_bitmap.prev_max = 0; + tss->io_bitmap.prev_sequence = 0; + memset(tss->io_bitmap.bitmap, 0xff, sizeof(tss->io_bitmap.bitmap)); + /* + * Invalidate the extra array entry past the end of the all + * permission bitmap as required by the hardware. + */ + tss->io_bitmap.mapall[IO_BITMAP_LONGS] = ~0UL; +#endif +} + +/* + * Setup everything needed to handle exceptions from the IDT, including the IST + * exceptions which use paranoid_entry(). + */ +void cpu_init_exception_handling(void) +{ + struct tss_struct *tss = this_cpu_ptr(&cpu_tss_rw); + int cpu = raw_smp_processor_id(); + + /* paranoid_entry() gets the CPU number from the GDT */ + setup_getcpu(cpu); + + /* IST vectors need TSS to be set up. */ + tss_setup_ist(tss); + tss_setup_io_bitmap(tss); + set_tss_desc(cpu, &get_cpu_entry_area(cpu)->tss.x86_tss); + + load_TR_desc(); + + /* Finally load the IDT */ + load_current_idt(); +} + +/* + * cpu_init() initializes state that is per-CPU. Some data is already + * initialized (naturally) in the bootstrap process, such as the GDT. We + * reload it nevertheless, this function acts as a 'CPU state barrier', + * nothing should get across. + */ +void cpu_init(void) +{ + struct task_struct *cur = current; + int cpu = raw_smp_processor_id(); + + wait_for_master_cpu(cpu); + + ucode_cpu_init(cpu); + +#ifdef CONFIG_NUMA + if (this_cpu_read(numa_node) == 0 && + early_cpu_to_node(cpu) != NUMA_NO_NODE) + set_numa_node(early_cpu_to_node(cpu)); +#endif + pr_debug("Initializing CPU#%d\n", cpu); + + if (IS_ENABLED(CONFIG_X86_64) || cpu_feature_enabled(X86_FEATURE_VME) || + boot_cpu_has(X86_FEATURE_TSC) || boot_cpu_has(X86_FEATURE_DE)) + cr4_clear_bits(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE); + + /* + * Initialize the per-CPU GDT with the boot GDT, + * and set up the GDT descriptor: + */ + switch_to_new_gdt(cpu); + + if (IS_ENABLED(CONFIG_X86_64)) { + loadsegment(fs, 0); + memset(cur->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8); + syscall_init(); + + wrmsrl(MSR_FS_BASE, 0); + wrmsrl(MSR_KERNEL_GS_BASE, 0); + barrier(); + + x2apic_setup(); + } + + mmgrab(&init_mm); + cur->active_mm = &init_mm; + BUG_ON(cur->mm); + initialize_tlbstate_and_flush(); + enter_lazy_tlb(&init_mm, cur); + + /* + * sp0 points to the entry trampoline stack regardless of what task + * is running. + */ + load_sp0((unsigned long)(cpu_entry_stack(cpu) + 1)); + + load_mm_ldt(&init_mm); + + clear_all_debug_regs(); + dbg_restore_debug_regs(); + + doublefault_init_cpu_tss(); + + if (is_uv_system()) + uv_cpu_init(); + + load_fixmap_gdt(cpu); +} + +#ifdef CONFIG_SMP +void cpu_init_secondary(void) +{ + /* + * Relies on the BP having set-up the IDT tables, which are loaded + * on this CPU in cpu_init_exception_handling(). + */ + cpu_init_exception_handling(); + cpu_init(); + fpu__init_cpu(); +} +#endif + +#ifdef CONFIG_MICROCODE_LATE_LOADING +/** + * store_cpu_caps() - Store a snapshot of CPU capabilities + * @curr_info: Pointer where to store it + * + * Returns: None + */ +void store_cpu_caps(struct cpuinfo_x86 *curr_info) +{ + /* Reload CPUID max function as it might've changed. */ + curr_info->cpuid_level = cpuid_eax(0); + + /* Copy all capability leafs and pick up the synthetic ones. */ + memcpy(&curr_info->x86_capability, &boot_cpu_data.x86_capability, + sizeof(curr_info->x86_capability)); + + /* Get the hardware CPUID leafs */ + get_cpu_cap(curr_info); +} + +/** + * microcode_check() - Check if any CPU capabilities changed after an update. + * @prev_info: CPU capabilities stored before an update. + * + * The microcode loader calls this upon late microcode load to recheck features, + * only when microcode has been updated. Caller holds microcode_mutex and CPU + * hotplug lock. + * + * Return: None + */ +void microcode_check(struct cpuinfo_x86 *prev_info) +{ + struct cpuinfo_x86 curr_info; + + perf_check_microcode(); + + amd_check_microcode(); + + store_cpu_caps(&curr_info); + + if (!memcmp(&prev_info->x86_capability, &curr_info.x86_capability, + sizeof(prev_info->x86_capability))) + return; + + pr_warn("x86/CPU: CPU features have changed after loading microcode, but might not take effect.\n"); + pr_warn("x86/CPU: Please consider either early loading through initrd/built-in or a potential BIOS update.\n"); +} +#endif + +/* + * Invoked from core CPU hotplug code after hotplug operations + */ +void arch_smt_update(void) +{ + /* Handle the speculative execution misfeatures */ + cpu_bugs_smt_update(); + /* Check whether IPI broadcasting can be enabled */ + apic_smt_update(); +} + +void __init arch_cpu_finalize_init(void) +{ + identify_boot_cpu(); + + /* + * identify_boot_cpu() initialized SMT support information, let the + * core code know. + */ + cpu_smt_check_topology(); + + if (!IS_ENABLED(CONFIG_SMP)) { + pr_info("CPU: "); + print_cpu_info(&boot_cpu_data); + } + + cpu_select_mitigations(); + + arch_smt_update(); + + if (IS_ENABLED(CONFIG_X86_32)) { + /* + * Check whether this is a real i386 which is not longer + * supported and fixup the utsname. + */ + if (boot_cpu_data.x86 < 4) + panic("Kernel requires i486+ for 'invlpg' and other features"); + + init_utsname()->machine[1] = + '0' + (boot_cpu_data.x86 > 6 ? 6 : boot_cpu_data.x86); + } + + /* + * Must be before alternatives because it might set or clear + * feature bits. + */ + fpu__init_system(); + fpu__init_cpu(); + + alternative_instructions(); + + if (IS_ENABLED(CONFIG_X86_64)) { + /* + * Make sure the first 2MB area is not mapped by huge pages + * There are typically fixed size MTRRs in there and overlapping + * MTRRs into large pages causes slow downs. + * + * Right now we don't do that with gbpages because there seems + * very little benefit for that case. + */ + if (!direct_gbpages) + set_memory_4k((unsigned long)__va(0), 1); + } else { + fpu__init_check_bugs(); + } + + /* + * This needs to be called before any devices perform DMA + * operations that might use the SWIOTLB bounce buffers. It will + * mark the bounce buffers as decrypted so that their usage will + * not cause "plain-text" data to be decrypted when accessed. It + * must be called after late_time_init() so that Hyper-V x86/x64 + * hypercalls work when the SWIOTLB bounce buffers are decrypted. + */ + mem_encrypt_init(); +} diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h new file mode 100644 index 000000000..66e19380a --- /dev/null +++ b/arch/x86/kernel/cpu/cpu.h @@ -0,0 +1,89 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef ARCH_X86_CPU_H +#define ARCH_X86_CPU_H + +/* attempt to consolidate cpu attributes */ +struct cpu_dev { + const char *c_vendor; + + /* some have two possibilities for cpuid string */ + const char *c_ident[2]; + + void (*c_early_init)(struct cpuinfo_x86 *); + void (*c_bsp_init)(struct cpuinfo_x86 *); + void (*c_init)(struct cpuinfo_x86 *); + void (*c_identify)(struct cpuinfo_x86 *); + void (*c_detect_tlb)(struct cpuinfo_x86 *); + int c_x86_vendor; +#ifdef CONFIG_X86_32 + /* Optional vendor specific routine to obtain the cache size. */ + unsigned int (*legacy_cache_size)(struct cpuinfo_x86 *, + unsigned int); + + /* Family/stepping-based lookup table for model names. */ + struct legacy_cpu_model_info { + int family; + const char *model_names[16]; + } legacy_models[5]; +#endif +}; + +struct _tlb_table { + unsigned char descriptor; + char tlb_type; + unsigned int entries; + /* unsigned int ways; */ + char info[128]; +}; + +#define cpu_dev_register(cpu_devX) \ + static const struct cpu_dev *const __cpu_dev_##cpu_devX __used \ + __section(".x86_cpu_dev.init") = \ + &cpu_devX; + +extern const struct cpu_dev *const __x86_cpu_dev_start[], + *const __x86_cpu_dev_end[]; + +#ifdef CONFIG_CPU_SUP_INTEL +enum tsx_ctrl_states { + TSX_CTRL_ENABLE, + TSX_CTRL_DISABLE, + TSX_CTRL_NOT_SUPPORTED, +}; + +extern __ro_after_init enum tsx_ctrl_states tsx_ctrl_state; + +extern void __init tsx_init(void); +extern void tsx_enable(void); +extern void tsx_disable(void); +#else +static inline void tsx_init(void) { } +#endif /* CONFIG_CPU_SUP_INTEL */ + +extern void init_spectral_chicken(struct cpuinfo_x86 *c); + +extern void get_cpu_cap(struct cpuinfo_x86 *c); +extern void get_cpu_address_sizes(struct cpuinfo_x86 *c); +extern void cpu_detect_cache_sizes(struct cpuinfo_x86 *c); +extern void init_scattered_cpuid_features(struct cpuinfo_x86 *c); +extern void init_intel_cacheinfo(struct cpuinfo_x86 *c); +extern void init_amd_cacheinfo(struct cpuinfo_x86 *c); +extern void init_hygon_cacheinfo(struct cpuinfo_x86 *c); + +extern void detect_num_cpu_cores(struct cpuinfo_x86 *c); +extern int detect_extended_topology_early(struct cpuinfo_x86 *c); +extern int detect_extended_topology(struct cpuinfo_x86 *c); +extern int detect_ht_early(struct cpuinfo_x86 *c); +extern void detect_ht(struct cpuinfo_x86 *c); +extern void check_null_seg_clears_base(struct cpuinfo_x86 *c); + +unsigned int aperfmperf_get_khz(int cpu); +void cpu_select_mitigations(void); + +extern void x86_spec_ctrl_setup_ap(void); +extern void update_srbds_msr(void); +extern void update_gds_msr(void); + +extern u64 x86_read_arch_cap_msr(void); + +#endif /* ARCH_X86_CPU_H */ diff --git a/arch/x86/kernel/cpu/cpuid-deps.c b/arch/x86/kernel/cpu/cpuid-deps.c new file mode 100644 index 000000000..d50224199 --- /dev/null +++ b/arch/x86/kernel/cpu/cpuid-deps.c @@ -0,0 +1,133 @@ +/* Declare dependencies between CPUIDs */ +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <asm/cpufeature.h> + +struct cpuid_dep { + unsigned int feature; + unsigned int depends; +}; + +/* + * Table of CPUID features that depend on others. + * + * This only includes dependencies that can be usefully disabled, not + * features part of the base set (like FPU). + * + * Note this all is not __init / __initdata because it can be + * called from cpu hotplug. It shouldn't do anything in this case, + * but it's difficult to tell that to the init reference checker. + */ +static const struct cpuid_dep cpuid_deps[] = { + { X86_FEATURE_FXSR, X86_FEATURE_FPU }, + { X86_FEATURE_XSAVEOPT, X86_FEATURE_XSAVE }, + { X86_FEATURE_XSAVEC, X86_FEATURE_XSAVE }, + { X86_FEATURE_XSAVES, X86_FEATURE_XSAVE }, + { X86_FEATURE_AVX, X86_FEATURE_XSAVE }, + { X86_FEATURE_PKU, X86_FEATURE_XSAVE }, + { X86_FEATURE_MPX, X86_FEATURE_XSAVE }, + { X86_FEATURE_XGETBV1, X86_FEATURE_XSAVE }, + { X86_FEATURE_CMOV, X86_FEATURE_FXSR }, + { X86_FEATURE_MMX, X86_FEATURE_FXSR }, + { X86_FEATURE_MMXEXT, X86_FEATURE_MMX }, + { X86_FEATURE_FXSR_OPT, X86_FEATURE_FXSR }, + { X86_FEATURE_XSAVE, X86_FEATURE_FXSR }, + { X86_FEATURE_XMM, X86_FEATURE_FXSR }, + { X86_FEATURE_XMM2, X86_FEATURE_XMM }, + { X86_FEATURE_XMM3, X86_FEATURE_XMM2 }, + { X86_FEATURE_XMM4_1, X86_FEATURE_XMM2 }, + { X86_FEATURE_XMM4_2, X86_FEATURE_XMM2 }, + { X86_FEATURE_XMM3, X86_FEATURE_XMM2 }, + { X86_FEATURE_PCLMULQDQ, X86_FEATURE_XMM2 }, + { X86_FEATURE_SSSE3, X86_FEATURE_XMM2, }, + { X86_FEATURE_F16C, X86_FEATURE_XMM2, }, + { X86_FEATURE_AES, X86_FEATURE_XMM2 }, + { X86_FEATURE_SHA_NI, X86_FEATURE_XMM2 }, + { X86_FEATURE_FMA, X86_FEATURE_AVX }, + { X86_FEATURE_AVX2, X86_FEATURE_AVX, }, + { X86_FEATURE_AVX512F, X86_FEATURE_AVX, }, + { X86_FEATURE_AVX512IFMA, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512PF, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512ER, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512CD, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512DQ, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512BW, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512VL, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512VBMI, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512_VBMI2, X86_FEATURE_AVX512VL }, + { X86_FEATURE_GFNI, X86_FEATURE_AVX512VL }, + { X86_FEATURE_VAES, X86_FEATURE_AVX512VL }, + { X86_FEATURE_VPCLMULQDQ, X86_FEATURE_AVX512VL }, + { X86_FEATURE_AVX512_VNNI, X86_FEATURE_AVX512VL }, + { X86_FEATURE_AVX512_BITALG, X86_FEATURE_AVX512VL }, + { X86_FEATURE_AVX512_4VNNIW, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512_4FMAPS, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512_VPOPCNTDQ, X86_FEATURE_AVX512F }, + { X86_FEATURE_AVX512_VP2INTERSECT, X86_FEATURE_AVX512VL }, + { X86_FEATURE_CQM_OCCUP_LLC, X86_FEATURE_CQM_LLC }, + { X86_FEATURE_CQM_MBM_TOTAL, X86_FEATURE_CQM_LLC }, + { X86_FEATURE_CQM_MBM_LOCAL, X86_FEATURE_CQM_LLC }, + { X86_FEATURE_AVX512_BF16, X86_FEATURE_AVX512VL }, + { X86_FEATURE_ENQCMD, X86_FEATURE_XSAVES }, + { X86_FEATURE_PER_THREAD_MBA, X86_FEATURE_MBA }, + {} +}; + +static inline void clear_feature(struct cpuinfo_x86 *c, unsigned int feature) +{ + /* + * Note: This could use the non atomic __*_bit() variants, but the + * rest of the cpufeature code uses atomics as well, so keep it for + * consistency. Cleanup all of it separately. + */ + if (!c) { + clear_cpu_cap(&boot_cpu_data, feature); + set_bit(feature, (unsigned long *)cpu_caps_cleared); + } else { + clear_bit(feature, (unsigned long *)c->x86_capability); + } +} + +/* Take the capabilities and the BUG bits into account */ +#define MAX_FEATURE_BITS ((NCAPINTS + NBUGINTS) * sizeof(u32) * 8) + +static void do_clear_cpu_cap(struct cpuinfo_x86 *c, unsigned int feature) +{ + DECLARE_BITMAP(disable, MAX_FEATURE_BITS); + const struct cpuid_dep *d; + bool changed; + + if (WARN_ON(feature >= MAX_FEATURE_BITS)) + return; + + clear_feature(c, feature); + + /* Collect all features to disable, handling dependencies */ + memset(disable, 0, sizeof(disable)); + __set_bit(feature, disable); + + /* Loop until we get a stable state. */ + do { + changed = false; + for (d = cpuid_deps; d->feature; d++) { + if (!test_bit(d->depends, disable)) + continue; + if (__test_and_set_bit(d->feature, disable)) + continue; + + changed = true; + clear_feature(c, d->feature); + } + } while (changed); +} + +void clear_cpu_cap(struct cpuinfo_x86 *c, unsigned int feature) +{ + do_clear_cpu_cap(c, feature); +} + +void setup_clear_cpu_cap(unsigned int feature) +{ + do_clear_cpu_cap(NULL, feature); +} diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c new file mode 100644 index 000000000..1d9b8aaea --- /dev/null +++ b/arch/x86/kernel/cpu/cyrix.c @@ -0,0 +1,466 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/bitops.h> +#include <linux/delay.h> +#include <linux/pci.h> +#include <asm/dma.h> +#include <linux/io.h> +#include <asm/processor-cyrix.h> +#include <asm/processor-flags.h> +#include <linux/timer.h> +#include <asm/pci-direct.h> +#include <asm/tsc.h> +#include <asm/cpufeature.h> +#include <linux/sched.h> +#include <linux/sched/clock.h> + +#include "cpu.h" + +/* + * Read NSC/Cyrix DEVID registers (DIR) to get more detailed info. about the CPU + */ +static void __do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) +{ + unsigned char ccr2, ccr3; + + /* we test for DEVID by checking whether CCR3 is writable */ + ccr3 = getCx86(CX86_CCR3); + setCx86(CX86_CCR3, ccr3 ^ 0x80); + getCx86(0xc0); /* dummy to change bus */ + + if (getCx86(CX86_CCR3) == ccr3) { /* no DEVID regs. */ + ccr2 = getCx86(CX86_CCR2); + setCx86(CX86_CCR2, ccr2 ^ 0x04); + getCx86(0xc0); /* dummy */ + + if (getCx86(CX86_CCR2) == ccr2) /* old Cx486SLC/DLC */ + *dir0 = 0xfd; + else { /* Cx486S A step */ + setCx86(CX86_CCR2, ccr2); + *dir0 = 0xfe; + } + } else { + setCx86(CX86_CCR3, ccr3); /* restore CCR3 */ + + /* read DIR0 and DIR1 CPU registers */ + *dir0 = getCx86(CX86_DIR0); + *dir1 = getCx86(CX86_DIR1); + } +} + +static void do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) +{ + unsigned long flags; + + local_irq_save(flags); + __do_cyrix_devid(dir0, dir1); + local_irq_restore(flags); +} +/* + * Cx86_dir0_msb is a HACK needed by check_cx686_cpuid/slop in bugs.h in + * order to identify the Cyrix CPU model after we're out of setup.c + * + * Actually since bugs.h doesn't even reference this perhaps someone should + * fix the documentation ??? + */ +static unsigned char Cx86_dir0_msb = 0; + +static const char Cx86_model[][9] = { + "Cx486", "Cx486", "5x86 ", "6x86", "MediaGX ", "6x86MX ", + "M II ", "Unknown" +}; +static const char Cx486_name[][5] = { + "SLC", "DLC", "SLC2", "DLC2", "SRx", "DRx", + "SRx2", "DRx2" +}; +static const char Cx486S_name[][4] = { + "S", "S2", "Se", "S2e" +}; +static const char Cx486D_name[][4] = { + "DX", "DX2", "?", "?", "?", "DX4" +}; +static char Cx86_cb[] = "?.5x Core/Bus Clock"; +static const char cyrix_model_mult1[] = "12??43"; +static const char cyrix_model_mult2[] = "12233445"; + +/* + * Reset the slow-loop (SLOP) bit on the 686(L) which is set by some old + * BIOSes for compatibility with DOS games. This makes the udelay loop + * work correctly, and improves performance. + * + * FIXME: our newer udelay uses the tsc. We don't need to frob with SLOP + */ + +static void check_cx686_slop(struct cpuinfo_x86 *c) +{ + unsigned long flags; + + if (Cx86_dir0_msb == 3) { + unsigned char ccr3, ccr5; + + local_irq_save(flags); + ccr3 = getCx86(CX86_CCR3); + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */ + ccr5 = getCx86(CX86_CCR5); + if (ccr5 & 2) + setCx86(CX86_CCR5, ccr5 & 0xfd); /* reset SLOP */ + setCx86(CX86_CCR3, ccr3); /* disable MAPEN */ + local_irq_restore(flags); + + if (ccr5 & 2) { /* possible wrong calibration done */ + pr_info("Recalibrating delay loop with SLOP bit reset\n"); + calibrate_delay(); + c->loops_per_jiffy = loops_per_jiffy; + } + } +} + + +static void set_cx86_reorder(void) +{ + u8 ccr3; + + pr_info("Enable Memory access reorder on Cyrix/NSC processor.\n"); + ccr3 = getCx86(CX86_CCR3); + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */ + + /* Load/Store Serialize to mem access disable (=reorder it) */ + setCx86(CX86_PCR0, getCx86(CX86_PCR0) & ~0x80); + /* set load/store serialize from 1GB to 4GB */ + ccr3 |= 0xe0; + setCx86(CX86_CCR3, ccr3); +} + +static void set_cx86_memwb(void) +{ + pr_info("Enable Memory-Write-back mode on Cyrix/NSC processor.\n"); + + /* CCR2 bit 2: unlock NW bit */ + setCx86(CX86_CCR2, getCx86(CX86_CCR2) & ~0x04); + /* set 'Not Write-through' */ + write_cr0(read_cr0() | X86_CR0_NW); + /* CCR2 bit 2: lock NW bit and set WT1 */ + setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x14); +} + +/* + * Configure later MediaGX and/or Geode processor. + */ + +static void geode_configure(void) +{ + unsigned long flags; + u8 ccr3; + local_irq_save(flags); + + /* Suspend on halt power saving and enable #SUSP pin */ + setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x88); + + ccr3 = getCx86(CX86_CCR3); + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */ + + + /* FPU fast, DTE cache, Mem bypass */ + setCx86(CX86_CCR4, getCx86(CX86_CCR4) | 0x38); + setCx86(CX86_CCR3, ccr3); /* disable MAPEN */ + + set_cx86_memwb(); + set_cx86_reorder(); + + local_irq_restore(flags); +} + +static void early_init_cyrix(struct cpuinfo_x86 *c) +{ + unsigned char dir0, dir0_msn, dir1 = 0; + + __do_cyrix_devid(&dir0, &dir1); + dir0_msn = dir0 >> 4; /* identifies CPU "family" */ + + switch (dir0_msn) { + case 3: /* 6x86/6x86L */ + /* Emulate MTRRs using Cyrix's ARRs. */ + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); + break; + case 5: /* 6x86MX/M II */ + /* Emulate MTRRs using Cyrix's ARRs. */ + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); + break; + } +} + +static void init_cyrix(struct cpuinfo_x86 *c) +{ + unsigned char dir0, dir0_msn, dir0_lsn, dir1 = 0; + char *buf = c->x86_model_id; + const char *p = NULL; + + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); + + /* Cyrix used bit 24 in extended (AMD) CPUID for Cyrix MMX extensions */ + if (test_cpu_cap(c, 1*32+24)) { + clear_cpu_cap(c, 1*32+24); + set_cpu_cap(c, X86_FEATURE_CXMMX); + } + + do_cyrix_devid(&dir0, &dir1); + + check_cx686_slop(c); + + Cx86_dir0_msb = dir0_msn = dir0 >> 4; /* identifies CPU "family" */ + dir0_lsn = dir0 & 0xf; /* model or clock multiplier */ + + /* common case step number/rev -- exceptions handled below */ + c->x86_model = (dir1 >> 4) + 1; + c->x86_stepping = dir1 & 0xf; + + /* Now cook; the original recipe is by Channing Corn, from Cyrix. + * We do the same thing for each generation: we work out + * the model, multiplier and stepping. Black magic included, + * to make the silicon step/rev numbers match the printed ones. + */ + + switch (dir0_msn) { + unsigned char tmp; + + case 0: /* Cx486SLC/DLC/SRx/DRx */ + p = Cx486_name[dir0_lsn & 7]; + break; + + case 1: /* Cx486S/DX/DX2/DX4 */ + p = (dir0_lsn & 8) ? Cx486D_name[dir0_lsn & 5] + : Cx486S_name[dir0_lsn & 3]; + break; + + case 2: /* 5x86 */ + Cx86_cb[2] = cyrix_model_mult1[dir0_lsn & 5]; + p = Cx86_cb+2; + break; + + case 3: /* 6x86/6x86L */ + Cx86_cb[1] = ' '; + Cx86_cb[2] = cyrix_model_mult1[dir0_lsn & 5]; + if (dir1 > 0x21) { /* 686L */ + Cx86_cb[0] = 'L'; + p = Cx86_cb; + (c->x86_model)++; + } else /* 686 */ + p = Cx86_cb+1; + /* Emulate MTRRs using Cyrix's ARRs. */ + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); + /* 6x86's contain this bug */ + set_cpu_bug(c, X86_BUG_COMA); + break; + + case 4: /* MediaGX/GXm or Geode GXM/GXLV/GX1 */ + case 11: /* GX1 with inverted Device ID */ +#ifdef CONFIG_PCI + { + u32 vendor, device; + /* + * It isn't really a PCI quirk directly, but the cure is the + * same. The MediaGX has deep magic SMM stuff that handles the + * SB emulation. It throws away the fifo on disable_dma() which + * is wrong and ruins the audio. + * + * Bug2: VSA1 has a wrap bug so that using maximum sized DMA + * causes bad things. According to NatSemi VSA2 has another + * bug to do with 'hlt'. I've not seen any boards using VSA2 + * and X doesn't seem to support it either so who cares 8). + * VSA1 we work around however. + */ + + pr_info("Working around Cyrix MediaGX virtual DMA bugs.\n"); + isa_dma_bridge_buggy = 2; + + /* We do this before the PCI layer is running. However we + are safe here as we know the bridge must be a Cyrix + companion and must be present */ + vendor = read_pci_config_16(0, 0, 0x12, PCI_VENDOR_ID); + device = read_pci_config_16(0, 0, 0x12, PCI_DEVICE_ID); + + /* + * The 5510/5520 companion chips have a funky PIT. + */ + if (vendor == PCI_VENDOR_ID_CYRIX && + (device == PCI_DEVICE_ID_CYRIX_5510 || + device == PCI_DEVICE_ID_CYRIX_5520)) + mark_tsc_unstable("cyrix 5510/5520 detected"); + } +#endif + c->x86_cache_size = 16; /* Yep 16K integrated cache thats it */ + + /* GXm supports extended cpuid levels 'ala' AMD */ + if (c->cpuid_level == 2) { + /* Enable cxMMX extensions (GX1 Datasheet 54) */ + setCx86(CX86_CCR7, getCx86(CX86_CCR7) | 1); + + /* + * GXm : 0x30 ... 0x5f GXm datasheet 51 + * GXlv: 0x6x GXlv datasheet 54 + * ? : 0x7x + * GX1 : 0x8x GX1 datasheet 56 + */ + if ((0x30 <= dir1 && dir1 <= 0x6f) || + (0x80 <= dir1 && dir1 <= 0x8f)) + geode_configure(); + return; + } else { /* MediaGX */ + Cx86_cb[2] = (dir0_lsn & 1) ? '3' : '4'; + p = Cx86_cb+2; + c->x86_model = (dir1 & 0x20) ? 1 : 2; + } + break; + + case 5: /* 6x86MX/M II */ + if (dir1 > 7) { + dir0_msn++; /* M II */ + /* Enable MMX extensions (App note 108) */ + setCx86(CX86_CCR7, getCx86(CX86_CCR7)|1); + } else { + /* A 6x86MX - it has the bug. */ + set_cpu_bug(c, X86_BUG_COMA); + } + tmp = (!(dir0_lsn & 7) || dir0_lsn & 1) ? 2 : 0; + Cx86_cb[tmp] = cyrix_model_mult2[dir0_lsn & 7]; + p = Cx86_cb+tmp; + if (((dir1 & 0x0f) > 4) || ((dir1 & 0xf0) == 0x20)) + (c->x86_model)++; + /* Emulate MTRRs using Cyrix's ARRs. */ + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); + break; + + case 0xf: /* Cyrix 486 without DEVID registers */ + switch (dir0_lsn) { + case 0xd: /* either a 486SLC or DLC w/o DEVID */ + dir0_msn = 0; + p = Cx486_name[!!boot_cpu_has(X86_FEATURE_FPU)]; + break; + + case 0xe: /* a 486S A step */ + dir0_msn = 0; + p = Cx486S_name[0]; + break; + } + break; + + default: /* unknown (shouldn't happen, we know everyone ;-) */ + dir0_msn = 7; + break; + } + strcpy(buf, Cx86_model[dir0_msn & 7]); + if (p) + strcat(buf, p); + return; +} + +/* + * Handle National Semiconductor branded processors + */ +static void init_nsc(struct cpuinfo_x86 *c) +{ + /* + * There may be GX1 processors in the wild that are branded + * NSC and not Cyrix. + * + * This function only handles the GX processor, and kicks every + * thing else to the Cyrix init function above - that should + * cover any processors that might have been branded differently + * after NSC acquired Cyrix. + * + * If this breaks your GX1 horribly, please e-mail + * info-linux@ldcmail.amd.com to tell us. + */ + + /* Handle the GX (Formally known as the GX2) */ + + if (c->x86 == 5 && c->x86_model == 5) + cpu_detect_cache_sizes(c); + else + init_cyrix(c); +} + +/* + * Cyrix CPUs without cpuid or with cpuid not yet enabled can be detected + * by the fact that they preserve the flags across the division of 5/2. + * PII and PPro exhibit this behavior too, but they have cpuid available. + */ + +/* + * Perform the Cyrix 5/2 test. A Cyrix won't change + * the flags, while other 486 chips will. + */ +static inline int test_cyrix_52div(void) +{ + unsigned int test; + + __asm__ __volatile__( + "sahf\n\t" /* clear flags (%eax = 0x0005) */ + "div %b2\n\t" /* divide 5 by 2 */ + "lahf" /* store flags into %ah */ + : "=a" (test) + : "0" (5), "q" (2) + : "cc"); + + /* AH is 0x02 on Cyrix after the divide.. */ + return (unsigned char) (test >> 8) == 0x02; +} + +static void cyrix_identify(struct cpuinfo_x86 *c) +{ + /* Detect Cyrix with disabled CPUID */ + if (c->x86 == 4 && test_cyrix_52div()) { + unsigned char dir0, dir1; + + strcpy(c->x86_vendor_id, "CyrixInstead"); + c->x86_vendor = X86_VENDOR_CYRIX; + + /* Actually enable cpuid on the older cyrix */ + + /* Retrieve CPU revisions */ + + do_cyrix_devid(&dir0, &dir1); + + dir0 >>= 4; + + /* Check it is an affected model */ + + if (dir0 == 5 || dir0 == 3) { + unsigned char ccr3; + unsigned long flags; + pr_info("Enabling CPUID on Cyrix processor.\n"); + local_irq_save(flags); + ccr3 = getCx86(CX86_CCR3); + /* enable MAPEN */ + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); + /* enable cpuid */ + setCx86(CX86_CCR4, getCx86(CX86_CCR4) | 0x80); + /* disable MAPEN */ + setCx86(CX86_CCR3, ccr3); + local_irq_restore(flags); + } + } +} + +static const struct cpu_dev cyrix_cpu_dev = { + .c_vendor = "Cyrix", + .c_ident = { "CyrixInstead" }, + .c_early_init = early_init_cyrix, + .c_init = init_cyrix, + .c_identify = cyrix_identify, + .c_x86_vendor = X86_VENDOR_CYRIX, +}; + +cpu_dev_register(cyrix_cpu_dev); + +static const struct cpu_dev nsc_cpu_dev = { + .c_vendor = "NSC", + .c_ident = { "Geode by NSC" }, + .c_init = init_nsc, + .c_x86_vendor = X86_VENDOR_NSC, +}; + +cpu_dev_register(nsc_cpu_dev); diff --git a/arch/x86/kernel/cpu/feat_ctl.c b/arch/x86/kernel/cpu/feat_ctl.c new file mode 100644 index 000000000..d75418512 --- /dev/null +++ b/arch/x86/kernel/cpu/feat_ctl.c @@ -0,0 +1,146 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/tboot.h> + +#include <asm/cpu.h> +#include <asm/cpufeature.h> +#include <asm/msr-index.h> +#include <asm/processor.h> +#include <asm/vmx.h> + +#undef pr_fmt +#define pr_fmt(fmt) "x86/cpu: " fmt + +#ifdef CONFIG_X86_VMX_FEATURE_NAMES +enum vmx_feature_leafs { + MISC_FEATURES = 0, + PRIMARY_CTLS, + SECONDARY_CTLS, + NR_VMX_FEATURE_WORDS, +}; + +#define VMX_F(x) BIT(VMX_FEATURE_##x & 0x1f) + +static void init_vmx_capabilities(struct cpuinfo_x86 *c) +{ + u32 supported, funcs, ept, vpid, ign; + + BUILD_BUG_ON(NVMXINTS != NR_VMX_FEATURE_WORDS); + + /* + * The high bits contain the allowed-1 settings, i.e. features that can + * be turned on. The low bits contain the allowed-0 settings, i.e. + * features that can be turned off. Ignore the allowed-0 settings, + * if a feature can be turned on then it's supported. + * + * Use raw rdmsr() for primary processor controls and pin controls MSRs + * as they exist on any CPU that supports VMX, i.e. we want the WARN if + * the RDMSR faults. + */ + rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, ign, supported); + c->vmx_capability[PRIMARY_CTLS] = supported; + + rdmsr_safe(MSR_IA32_VMX_PROCBASED_CTLS2, &ign, &supported); + c->vmx_capability[SECONDARY_CTLS] = supported; + + rdmsr(MSR_IA32_VMX_PINBASED_CTLS, ign, supported); + rdmsr_safe(MSR_IA32_VMX_VMFUNC, &ign, &funcs); + + /* + * Except for EPT+VPID, which enumerates support for both in a single + * MSR, low for EPT, high for VPID. + */ + rdmsr_safe(MSR_IA32_VMX_EPT_VPID_CAP, &ept, &vpid); + + /* Pin, EPT, VPID and VM-Func are merged into a single word. */ + WARN_ON_ONCE(supported >> 16); + WARN_ON_ONCE(funcs >> 4); + c->vmx_capability[MISC_FEATURES] = (supported & 0xffff) | + ((vpid & 0x1) << 16) | + ((funcs & 0xf) << 28); + + /* EPT bits are full on scattered and must be manually handled. */ + if (ept & VMX_EPT_EXECUTE_ONLY_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_EXECUTE_ONLY); + if (ept & VMX_EPT_AD_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_AD); + if (ept & VMX_EPT_1GB_PAGE_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_1GB); + + /* Synthetic APIC features that are aggregates of multiple features. */ + if ((c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(VIRT_APIC_ACCESSES))) + c->vmx_capability[MISC_FEATURES] |= VMX_F(FLEXPRIORITY); + + if ((c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(APIC_REGISTER_VIRT)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(VIRT_INTR_DELIVERY)) && + (c->vmx_capability[MISC_FEATURES] & VMX_F(POSTED_INTR))) + c->vmx_capability[MISC_FEATURES] |= VMX_F(APICV); + + /* Set the synthetic cpufeatures to preserve /proc/cpuinfo's ABI. */ + if (c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) + set_cpu_cap(c, X86_FEATURE_TPR_SHADOW); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(FLEXPRIORITY)) + set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(VIRTUAL_NMIS)) + set_cpu_cap(c, X86_FEATURE_VNMI); + if (c->vmx_capability[SECONDARY_CTLS] & VMX_F(EPT)) + set_cpu_cap(c, X86_FEATURE_EPT); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(EPT_AD)) + set_cpu_cap(c, X86_FEATURE_EPT_AD); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(VPID)) + set_cpu_cap(c, X86_FEATURE_VPID); +} +#endif /* CONFIG_X86_VMX_FEATURE_NAMES */ + +void init_ia32_feat_ctl(struct cpuinfo_x86 *c) +{ + bool tboot = tboot_enabled(); + u64 msr; + + if (rdmsrl_safe(MSR_IA32_FEAT_CTL, &msr)) { + clear_cpu_cap(c, X86_FEATURE_VMX); + return; + } + + if (msr & FEAT_CTL_LOCKED) + goto update_caps; + + /* + * Ignore whatever value BIOS left in the MSR to avoid enabling random + * features or faulting on the WRMSR. + */ + msr = FEAT_CTL_LOCKED; + + /* + * Enable VMX if and only if the kernel may do VMXON at some point, + * i.e. KVM is enabled, to avoid unnecessarily adding an attack vector + * for the kernel, e.g. using VMX to hide malicious code. + */ + if (cpu_has(c, X86_FEATURE_VMX) && IS_ENABLED(CONFIG_KVM_INTEL)) { + msr |= FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX; + + if (tboot) + msr |= FEAT_CTL_VMX_ENABLED_INSIDE_SMX; + } + + wrmsrl(MSR_IA32_FEAT_CTL, msr); + +update_caps: + set_cpu_cap(c, X86_FEATURE_MSR_IA32_FEAT_CTL); + + if (!cpu_has(c, X86_FEATURE_VMX)) + return; + + if ( (tboot && !(msr & FEAT_CTL_VMX_ENABLED_INSIDE_SMX)) || + (!tboot && !(msr & FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX))) { + if (IS_ENABLED(CONFIG_KVM_INTEL)) + pr_err_once("VMX (%s TXT) disabled by BIOS\n", + tboot ? "inside" : "outside"); + clear_cpu_cap(c, X86_FEATURE_VMX); + } else { +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + init_vmx_capabilities(c); +#endif + } +} diff --git a/arch/x86/kernel/cpu/hygon.c b/arch/x86/kernel/cpu/hygon.c new file mode 100644 index 000000000..3f5c00b15 --- /dev/null +++ b/arch/x86/kernel/cpu/hygon.c @@ -0,0 +1,410 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Hygon Processor Support for Linux + * + * Copyright (C) 2018 Chengdu Haiguang IC Design Co., Ltd. + * + * Author: Pu Wen <puwen@hygon.cn> + */ +#include <linux/io.h> + +#include <asm/cpu.h> +#include <asm/smp.h> +#include <asm/numa.h> +#include <asm/cacheinfo.h> +#include <asm/spec-ctrl.h> +#include <asm/delay.h> +#ifdef CONFIG_X86_64 +# include <asm/set_memory.h> +#endif + +#include "cpu.h" + +#define APICID_SOCKET_ID_BIT 6 + +/* + * nodes_per_socket: Stores the number of nodes per socket. + * Refer to CPUID Fn8000_001E_ECX Node Identifiers[10:8] + */ +static u32 nodes_per_socket = 1; + +#ifdef CONFIG_NUMA +/* + * To workaround broken NUMA config. Read the comment in + * srat_detect_node(). + */ +static int nearby_node(int apicid) +{ + int i, node; + + for (i = apicid - 1; i >= 0; i--) { + node = __apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { + node = __apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + return first_node(node_online_map); /* Shouldn't happen */ +} +#endif + +static void hygon_get_topology_early(struct cpuinfo_x86 *c) +{ + if (cpu_has(c, X86_FEATURE_TOPOEXT)) + smp_num_siblings = ((cpuid_ebx(0x8000001e) >> 8) & 0xff) + 1; +} + +/* + * Fixup core topology information for + * (1) Hygon multi-node processors + * Assumption: Number of cores in each internal node is the same. + * (2) Hygon processors supporting compute units + */ +static void hygon_get_topology(struct cpuinfo_x86 *c) +{ + int cpu = smp_processor_id(); + + /* get information required for multi-node processors */ + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + int err; + u32 eax, ebx, ecx, edx; + + cpuid(0x8000001e, &eax, &ebx, &ecx, &edx); + + c->cpu_die_id = ecx & 0xff; + + c->cpu_core_id = ebx & 0xff; + + if (smp_num_siblings > 1) + c->x86_max_cores /= smp_num_siblings; + + /* + * In case leaf B is available, use it to derive + * topology information. + */ + err = detect_extended_topology(c); + if (!err) + c->x86_coreid_bits = get_count_order(c->x86_max_cores); + + /* + * Socket ID is ApicId[6] for the processors with model <= 0x3 + * when running on host. + */ + if (!boot_cpu_has(X86_FEATURE_HYPERVISOR) && c->x86_model <= 0x3) + c->phys_proc_id = c->apicid >> APICID_SOCKET_ID_BIT; + + cacheinfo_hygon_init_llc_id(c, cpu); + } else if (cpu_has(c, X86_FEATURE_NODEID_MSR)) { + u64 value; + + rdmsrl(MSR_FAM10H_NODE_ID, value); + c->cpu_die_id = value & 7; + + per_cpu(cpu_llc_id, cpu) = c->cpu_die_id; + } else + return; + + if (nodes_per_socket > 1) + set_cpu_cap(c, X86_FEATURE_AMD_DCM); +} + +/* + * On Hygon setup the lower bits of the APIC id distinguish the cores. + * Assumes number of cores is a power of two. + */ +static void hygon_detect_cmp(struct cpuinfo_x86 *c) +{ + unsigned int bits; + int cpu = smp_processor_id(); + + bits = c->x86_coreid_bits; + /* Low order bits define the core id (index of core in socket) */ + c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); + /* Convert the initial APIC ID into the socket ID */ + c->phys_proc_id = c->initial_apicid >> bits; + /* use socket ID also for last level cache */ + per_cpu(cpu_llc_id, cpu) = c->cpu_die_id = c->phys_proc_id; +} + +static void srat_detect_node(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_NUMA + int cpu = smp_processor_id(); + int node; + unsigned int apicid = c->apicid; + + node = numa_cpu_node(cpu); + if (node == NUMA_NO_NODE) + node = per_cpu(cpu_llc_id, cpu); + + /* + * On multi-fabric platform (e.g. Numascale NumaChip) a + * platform-specific handler needs to be called to fixup some + * IDs of the CPU. + */ + if (x86_cpuinit.fixup_cpu_id) + x86_cpuinit.fixup_cpu_id(c, node); + + if (!node_online(node)) { + /* + * Two possibilities here: + * + * - The CPU is missing memory and no node was created. In + * that case try picking one from a nearby CPU. + * + * - The APIC IDs differ from the HyperTransport node IDs. + * Assume they are all increased by a constant offset, but + * in the same order as the HT nodeids. If that doesn't + * result in a usable node fall back to the path for the + * previous case. + * + * This workaround operates directly on the mapping between + * APIC ID and NUMA node, assuming certain relationship + * between APIC ID, HT node ID and NUMA topology. As going + * through CPU mapping may alter the outcome, directly + * access __apicid_to_node[]. + */ + int ht_nodeid = c->initial_apicid; + + if (__apicid_to_node[ht_nodeid] != NUMA_NO_NODE) + node = __apicid_to_node[ht_nodeid]; + /* Pick a nearby node */ + if (!node_online(node)) + node = nearby_node(apicid); + } + numa_set_node(cpu, node); +#endif +} + +static void early_init_hygon_mc(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int bits, ecx; + + /* Multi core CPU? */ + if (c->extended_cpuid_level < 0x80000008) + return; + + ecx = cpuid_ecx(0x80000008); + + c->x86_max_cores = (ecx & 0xff) + 1; + + /* CPU telling us the core id bits shift? */ + bits = (ecx >> 12) & 0xF; + + /* Otherwise recompute */ + if (bits == 0) { + while ((1 << bits) < c->x86_max_cores) + bits++; + } + + c->x86_coreid_bits = bits; +#endif +} + +static void bsp_init_hygon(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_64 + unsigned long long tseg; + + /* + * Split up direct mapping around the TSEG SMM area. + * Don't do it for gbpages because there seems very little + * benefit in doing so. + */ + if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg)) { + unsigned long pfn = tseg >> PAGE_SHIFT; + + pr_debug("tseg: %010llx\n", tseg); + if (pfn_range_is_mapped(pfn, pfn + 1)) + set_memory_4k((unsigned long)__va(tseg), 1); + } +#endif + + if (cpu_has(c, X86_FEATURE_CONSTANT_TSC)) { + u64 val; + + rdmsrl(MSR_K7_HWCR, val); + if (!(val & BIT(24))) + pr_warn(FW_BUG "TSC doesn't count with P0 frequency!\n"); + } + + if (cpu_has(c, X86_FEATURE_MWAITX)) + use_mwaitx_delay(); + + if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + u32 ecx; + + ecx = cpuid_ecx(0x8000001e); + nodes_per_socket = ((ecx >> 8) & 7) + 1; + } else if (boot_cpu_has(X86_FEATURE_NODEID_MSR)) { + u64 value; + + rdmsrl(MSR_FAM10H_NODE_ID, value); + nodes_per_socket = ((value >> 3) & 7) + 1; + } + + if (!boot_cpu_has(X86_FEATURE_AMD_SSBD) && + !boot_cpu_has(X86_FEATURE_VIRT_SSBD)) { + /* + * Try to cache the base value so further operations can + * avoid RMW. If that faults, do not enable SSBD. + */ + if (!rdmsrl_safe(MSR_AMD64_LS_CFG, &x86_amd_ls_cfg_base)) { + setup_force_cpu_cap(X86_FEATURE_LS_CFG_SSBD); + setup_force_cpu_cap(X86_FEATURE_SSBD); + x86_amd_ls_cfg_ssbd_mask = 1ULL << 10; + } + } +} + +static void early_init_hygon(struct cpuinfo_x86 *c) +{ + u32 dummy; + + early_init_hygon_mc(c); + + set_cpu_cap(c, X86_FEATURE_K8); + + rdmsr_safe(MSR_AMD64_PATCH_LEVEL, &c->microcode, &dummy); + + /* + * c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate + * with P/T states and does not stop in deep C-states + */ + if (c->x86_power & (1 << 8)) { + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + } + + /* Bit 12 of 8000_0007 edx is accumulated power mechanism. */ + if (c->x86_power & BIT(12)) + set_cpu_cap(c, X86_FEATURE_ACC_POWER); + +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSCALL32); +#endif + +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_PCI) + /* + * ApicID can always be treated as an 8-bit value for Hygon APIC So, we + * can safely set X86_FEATURE_EXTD_APICID unconditionally. + */ + if (boot_cpu_has(X86_FEATURE_APIC)) + set_cpu_cap(c, X86_FEATURE_EXTD_APICID); +#endif + + /* + * This is only needed to tell the kernel whether to use VMCALL + * and VMMCALL. VMMCALL is never executed except under virt, so + * we can set it unconditionally. + */ + set_cpu_cap(c, X86_FEATURE_VMMCALL); + + hygon_get_topology_early(c); +} + +static void init_hygon(struct cpuinfo_x86 *c) +{ + early_init_hygon(c); + + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); + + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + + /* get apicid instead of initial apic id from cpuid */ + c->apicid = hard_smp_processor_id(); + + /* + * XXX someone from Hygon needs to confirm this DTRT + * + init_spectral_chicken(c); + */ + + set_cpu_cap(c, X86_FEATURE_ZEN); + set_cpu_cap(c, X86_FEATURE_CPB); + + cpu_detect_cache_sizes(c); + + hygon_detect_cmp(c); + hygon_get_topology(c); + srat_detect_node(c); + + init_hygon_cacheinfo(c); + + if (cpu_has(c, X86_FEATURE_XMM2)) { + /* + * Use LFENCE for execution serialization. On families which + * don't have that MSR, LFENCE is already serializing. + * msr_set_bit() uses the safe accessors, too, even if the MSR + * is not present. + */ + msr_set_bit(MSR_AMD64_DE_CFG, + MSR_AMD64_DE_CFG_LFENCE_SERIALIZE_BIT); + + /* A serializing LFENCE stops RDTSC speculation */ + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); + } + + /* + * Hygon processors have APIC timer running in deep C states. + */ + set_cpu_cap(c, X86_FEATURE_ARAT); + + /* Hygon CPUs don't reset SS attributes on SYSRET, Xen does. */ + if (!cpu_has(c, X86_FEATURE_XENPV)) + set_cpu_bug(c, X86_BUG_SYSRET_SS_ATTRS); + + check_null_seg_clears_base(c); +} + +static void cpu_detect_tlb_hygon(struct cpuinfo_x86 *c) +{ + u32 ebx, eax, ecx, edx; + u16 mask = 0xfff; + + if (c->extended_cpuid_level < 0x80000006) + return; + + cpuid(0x80000006, &eax, &ebx, &ecx, &edx); + + tlb_lld_4k[ENTRIES] = (ebx >> 16) & mask; + tlb_lli_4k[ENTRIES] = ebx & mask; + + /* Handle DTLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!((eax >> 16) & mask)) + tlb_lld_2m[ENTRIES] = (cpuid_eax(0x80000005) >> 16) & 0xff; + else + tlb_lld_2m[ENTRIES] = (eax >> 16) & mask; + + /* a 4M entry uses two 2M entries */ + tlb_lld_4m[ENTRIES] = tlb_lld_2m[ENTRIES] >> 1; + + /* Handle ITLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!(eax & mask)) { + cpuid(0x80000005, &eax, &ebx, &ecx, &edx); + tlb_lli_2m[ENTRIES] = eax & 0xff; + } else + tlb_lli_2m[ENTRIES] = eax & mask; + + tlb_lli_4m[ENTRIES] = tlb_lli_2m[ENTRIES] >> 1; +} + +static const struct cpu_dev hygon_cpu_dev = { + .c_vendor = "Hygon", + .c_ident = { "HygonGenuine" }, + .c_early_init = early_init_hygon, + .c_detect_tlb = cpu_detect_tlb_hygon, + .c_bsp_init = bsp_init_hygon, + .c_init = init_hygon, + .c_x86_vendor = X86_VENDOR_HYGON, +}; + +cpu_dev_register(hygon_cpu_dev); diff --git a/arch/x86/kernel/cpu/hypervisor.c b/arch/x86/kernel/cpu/hypervisor.c new file mode 100644 index 000000000..553bfbfc3 --- /dev/null +++ b/arch/x86/kernel/cpu/hypervisor.c @@ -0,0 +1,109 @@ +/* + * Common hypervisor code + * + * Copyright (C) 2008, VMware, Inc. + * Author : Alok N Kataria <akataria@vmware.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ + +#include <linux/init.h> +#include <linux/export.h> +#include <asm/processor.h> +#include <asm/hypervisor.h> + +static const __initconst struct hypervisor_x86 * const hypervisors[] = +{ +#ifdef CONFIG_XEN_PV + &x86_hyper_xen_pv, +#endif +#ifdef CONFIG_XEN_PVHVM + &x86_hyper_xen_hvm, +#endif + &x86_hyper_vmware, + &x86_hyper_ms_hyperv, +#ifdef CONFIG_KVM_GUEST + &x86_hyper_kvm, +#endif +#ifdef CONFIG_JAILHOUSE_GUEST + &x86_hyper_jailhouse, +#endif +#ifdef CONFIG_ACRN_GUEST + &x86_hyper_acrn, +#endif +}; + +enum x86_hypervisor_type x86_hyper_type; +EXPORT_SYMBOL(x86_hyper_type); + +bool __initdata nopv; +static __init int parse_nopv(char *arg) +{ + nopv = true; + return 0; +} +early_param("nopv", parse_nopv); + +static inline const struct hypervisor_x86 * __init +detect_hypervisor_vendor(void) +{ + const struct hypervisor_x86 *h = NULL, * const *p; + uint32_t pri, max_pri = 0; + + for (p = hypervisors; p < hypervisors + ARRAY_SIZE(hypervisors); p++) { + if (unlikely(nopv) && !(*p)->ignore_nopv) + continue; + + pri = (*p)->detect(); + if (pri > max_pri) { + max_pri = pri; + h = *p; + } + } + + if (h) + pr_info("Hypervisor detected: %s\n", h->name); + + return h; +} + +static void __init copy_array(const void *src, void *target, unsigned int size) +{ + unsigned int i, n = size / sizeof(void *); + const void * const *from = (const void * const *)src; + const void **to = (const void **)target; + + for (i = 0; i < n; i++) + if (from[i]) + to[i] = from[i]; +} + +void __init init_hypervisor_platform(void) +{ + const struct hypervisor_x86 *h; + + h = detect_hypervisor_vendor(); + + if (!h) + return; + + copy_array(&h->init, &x86_init.hyper, sizeof(h->init)); + copy_array(&h->runtime, &x86_platform.hyper, sizeof(h->runtime)); + + x86_hyper_type = h->type; + x86_init.hyper.init_platform(); +} diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c new file mode 100644 index 000000000..c6ad53e38 --- /dev/null +++ b/arch/x86/kernel/cpu/intel.c @@ -0,0 +1,1194 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/kernel.h> +#include <linux/pgtable.h> + +#include <linux/string.h> +#include <linux/bitops.h> +#include <linux/smp.h> +#include <linux/sched.h> +#include <linux/sched/clock.h> +#include <linux/thread_info.h> +#include <linux/init.h> +#include <linux/uaccess.h> + +#include <asm/cpufeature.h> +#include <asm/msr.h> +#include <asm/bugs.h> +#include <asm/cpu.h> +#include <asm/intel-family.h> +#include <asm/microcode_intel.h> +#include <asm/hwcap2.h> +#include <asm/elf.h> +#include <asm/cpu_device_id.h> +#include <asm/cmdline.h> +#include <asm/traps.h> +#include <asm/resctrl.h> +#include <asm/numa.h> + +#ifdef CONFIG_X86_64 +#include <linux/topology.h> +#endif + +#include "cpu.h" + +#ifdef CONFIG_X86_LOCAL_APIC +#include <asm/mpspec.h> +#include <asm/apic.h> +#endif + +enum split_lock_detect_state { + sld_off = 0, + sld_warn, + sld_fatal, +}; + +/* + * Default to sld_off because most systems do not support split lock detection + * split_lock_setup() will switch this to sld_warn on systems that support + * split lock detect, unless there is a command line override. + */ +static enum split_lock_detect_state sld_state __ro_after_init = sld_off; +static u64 msr_test_ctrl_cache __ro_after_init; + +/* + * With a name like MSR_TEST_CTL it should go without saying, but don't touch + * MSR_TEST_CTL unless the CPU is one of the whitelisted models. Writing it + * on CPUs that do not support SLD can cause fireworks, even when writing '0'. + */ +static bool cpu_model_supports_sld __ro_after_init; + +/* + * Processors which have self-snooping capability can handle conflicting + * memory type across CPUs by snooping its own cache. However, there exists + * CPU models in which having conflicting memory types still leads to + * unpredictable behavior, machine check errors, or hangs. Clear this + * feature to prevent its use on machines with known erratas. + */ +static void check_memory_type_self_snoop_errata(struct cpuinfo_x86 *c) +{ + switch (c->x86_model) { + case INTEL_FAM6_CORE_YONAH: + case INTEL_FAM6_CORE2_MEROM: + case INTEL_FAM6_CORE2_MEROM_L: + case INTEL_FAM6_CORE2_PENRYN: + case INTEL_FAM6_CORE2_DUNNINGTON: + case INTEL_FAM6_NEHALEM: + case INTEL_FAM6_NEHALEM_G: + case INTEL_FAM6_NEHALEM_EP: + case INTEL_FAM6_NEHALEM_EX: + case INTEL_FAM6_WESTMERE: + case INTEL_FAM6_WESTMERE_EP: + case INTEL_FAM6_SANDYBRIDGE: + setup_clear_cpu_cap(X86_FEATURE_SELFSNOOP); + } +} + +static bool ring3mwait_disabled __read_mostly; + +static int __init ring3mwait_disable(char *__unused) +{ + ring3mwait_disabled = true; + return 1; +} +__setup("ring3mwait=disable", ring3mwait_disable); + +static void probe_xeon_phi_r3mwait(struct cpuinfo_x86 *c) +{ + /* + * Ring 3 MONITOR/MWAIT feature cannot be detected without + * cpu model and family comparison. + */ + if (c->x86 != 6) + return; + switch (c->x86_model) { + case INTEL_FAM6_XEON_PHI_KNL: + case INTEL_FAM6_XEON_PHI_KNM: + break; + default: + return; + } + + if (ring3mwait_disabled) + return; + + set_cpu_cap(c, X86_FEATURE_RING3MWAIT); + this_cpu_or(msr_misc_features_shadow, + 1UL << MSR_MISC_FEATURES_ENABLES_RING3MWAIT_BIT); + + if (c == &boot_cpu_data) + ELF_HWCAP2 |= HWCAP2_RING3MWAIT; +} + +/* + * Early microcode releases for the Spectre v2 mitigation were broken. + * Information taken from; + * - https://newsroom.intel.com/wp-content/uploads/sites/11/2018/03/microcode-update-guidance.pdf + * - https://kb.vmware.com/s/article/52345 + * - Microcode revisions observed in the wild + * - Release note from 20180108 microcode release + */ +struct sku_microcode { + u8 model; + u8 stepping; + u32 microcode; +}; +static const struct sku_microcode spectre_bad_microcodes[] = { + { INTEL_FAM6_KABYLAKE, 0x0B, 0x80 }, + { INTEL_FAM6_KABYLAKE, 0x0A, 0x80 }, + { INTEL_FAM6_KABYLAKE, 0x09, 0x80 }, + { INTEL_FAM6_KABYLAKE_L, 0x0A, 0x80 }, + { INTEL_FAM6_KABYLAKE_L, 0x09, 0x80 }, + { INTEL_FAM6_SKYLAKE_X, 0x03, 0x0100013e }, + { INTEL_FAM6_SKYLAKE_X, 0x04, 0x0200003c }, + { INTEL_FAM6_BROADWELL, 0x04, 0x28 }, + { INTEL_FAM6_BROADWELL_G, 0x01, 0x1b }, + { INTEL_FAM6_BROADWELL_D, 0x02, 0x14 }, + { INTEL_FAM6_BROADWELL_D, 0x03, 0x07000011 }, + { INTEL_FAM6_BROADWELL_X, 0x01, 0x0b000025 }, + { INTEL_FAM6_HASWELL_L, 0x01, 0x21 }, + { INTEL_FAM6_HASWELL_G, 0x01, 0x18 }, + { INTEL_FAM6_HASWELL, 0x03, 0x23 }, + { INTEL_FAM6_HASWELL_X, 0x02, 0x3b }, + { INTEL_FAM6_HASWELL_X, 0x04, 0x10 }, + { INTEL_FAM6_IVYBRIDGE_X, 0x04, 0x42a }, + /* Observed in the wild */ + { INTEL_FAM6_SANDYBRIDGE_X, 0x06, 0x61b }, + { INTEL_FAM6_SANDYBRIDGE_X, 0x07, 0x712 }, +}; + +static bool bad_spectre_microcode(struct cpuinfo_x86 *c) +{ + int i; + + /* + * We know that the hypervisor lie to us on the microcode version so + * we may as well hope that it is running the correct version. + */ + if (cpu_has(c, X86_FEATURE_HYPERVISOR)) + return false; + + if (c->x86 != 6) + return false; + + for (i = 0; i < ARRAY_SIZE(spectre_bad_microcodes); i++) { + if (c->x86_model == spectre_bad_microcodes[i].model && + c->x86_stepping == spectre_bad_microcodes[i].stepping) + return (c->microcode <= spectre_bad_microcodes[i].microcode); + } + return false; +} + +static void early_init_intel(struct cpuinfo_x86 *c) +{ + u64 misc_enable; + + /* Unmask CPUID levels if masked: */ + if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { + if (msr_clear_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_LIMIT_CPUID_BIT) > 0) { + c->cpuid_level = cpuid_eax(0); + get_cpu_cap(c); + } + } + + if ((c->x86 == 0xf && c->x86_model >= 0x03) || + (c->x86 == 0x6 && c->x86_model >= 0x0e)) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + + if (c->x86 >= 6 && !cpu_has(c, X86_FEATURE_IA64)) + c->microcode = intel_get_microcode_revision(); + + /* Now if any of them are set, check the blacklist and clear the lot */ + if ((cpu_has(c, X86_FEATURE_SPEC_CTRL) || + cpu_has(c, X86_FEATURE_INTEL_STIBP) || + cpu_has(c, X86_FEATURE_IBRS) || cpu_has(c, X86_FEATURE_IBPB) || + cpu_has(c, X86_FEATURE_STIBP)) && bad_spectre_microcode(c)) { + pr_warn("Intel Spectre v2 broken microcode detected; disabling Speculation Control\n"); + setup_clear_cpu_cap(X86_FEATURE_IBRS); + setup_clear_cpu_cap(X86_FEATURE_IBPB); + setup_clear_cpu_cap(X86_FEATURE_STIBP); + setup_clear_cpu_cap(X86_FEATURE_SPEC_CTRL); + setup_clear_cpu_cap(X86_FEATURE_MSR_SPEC_CTRL); + setup_clear_cpu_cap(X86_FEATURE_INTEL_STIBP); + setup_clear_cpu_cap(X86_FEATURE_SSBD); + setup_clear_cpu_cap(X86_FEATURE_SPEC_CTRL_SSBD); + } + + /* + * Atom erratum AAE44/AAF40/AAG38/AAH41: + * + * A race condition between speculative fetches and invalidating + * a large page. This is worked around in microcode, but we + * need the microcode to have already been loaded... so if it is + * not, recommend a BIOS update and disable large pages. + */ + if (c->x86 == 6 && c->x86_model == 0x1c && c->x86_stepping <= 2 && + c->microcode < 0x20e) { + pr_warn("Atom PSE erratum detected, BIOS microcode update recommended\n"); + clear_cpu_cap(c, X86_FEATURE_PSE); + } + +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +#else + /* Netburst reports 64 bytes clflush size, but does IO in 128 bytes */ + if (c->x86 == 15 && c->x86_cache_alignment == 64) + c->x86_cache_alignment = 128; +#endif + + /* CPUID workaround for 0F33/0F34 CPU */ + if (c->x86 == 0xF && c->x86_model == 0x3 + && (c->x86_stepping == 0x3 || c->x86_stepping == 0x4)) + c->x86_phys_bits = 36; + + /* + * c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate + * with P/T states and does not stop in deep C-states. + * + * It is also reliable across cores and sockets. (but not across + * cabinets - we turn it off in that case explicitly.) + */ + if (c->x86_power & (1 << 8)) { + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + } + + /* Penwell and Cloverview have the TSC which doesn't sleep on S3 */ + if (c->x86 == 6) { + switch (c->x86_model) { + case INTEL_FAM6_ATOM_SALTWELL_MID: + case INTEL_FAM6_ATOM_SALTWELL_TABLET: + case INTEL_FAM6_ATOM_SILVERMONT_MID: + case INTEL_FAM6_ATOM_AIRMONT_NP: + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC_S3); + break; + default: + break; + } + } + + /* + * There is a known erratum on Pentium III and Core Solo + * and Core Duo CPUs. + * " Page with PAT set to WC while associated MTRR is UC + * may consolidate to UC " + * Because of this erratum, it is better to stick with + * setting WC in MTRR rather than using PAT on these CPUs. + * + * Enable PAT WC only on P4, Core 2 or later CPUs. + */ + if (c->x86 == 6 && c->x86_model < 15) + clear_cpu_cap(c, X86_FEATURE_PAT); + + /* + * If fast string is not enabled in IA32_MISC_ENABLE for any reason, + * clear the fast string and enhanced fast string CPU capabilities. + */ + if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { + rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); + if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) { + pr_info("Disabled fast string operations\n"); + setup_clear_cpu_cap(X86_FEATURE_REP_GOOD); + setup_clear_cpu_cap(X86_FEATURE_ERMS); + } + } + + /* + * Intel Quark Core DevMan_001.pdf section 6.4.11 + * "The operating system also is required to invalidate (i.e., flush) + * the TLB when any changes are made to any of the page table entries. + * The operating system must reload CR3 to cause the TLB to be flushed" + * + * As a result, boot_cpu_has(X86_FEATURE_PGE) in arch/x86/include/asm/tlbflush.h + * should be false so that __flush_tlb_all() causes CR3 insted of CR4.PGE + * to be modified. + */ + if (c->x86 == 5 && c->x86_model == 9) { + pr_info("Disabling PGE capability bit\n"); + setup_clear_cpu_cap(X86_FEATURE_PGE); + } + + if (c->cpuid_level >= 0x00000001) { + u32 eax, ebx, ecx, edx; + + cpuid(0x00000001, &eax, &ebx, &ecx, &edx); + /* + * If HTT (EDX[28]) is set EBX[16:23] contain the number of + * apicids which are reserved per package. Store the resulting + * shift value for the package management code. + */ + if (edx & (1U << 28)) + c->x86_coreid_bits = get_count_order((ebx >> 16) & 0xff); + } + + check_memory_type_self_snoop_errata(c); + + /* + * Get the number of SMT siblings early from the extended topology + * leaf, if available. Otherwise try the legacy SMT detection. + */ + if (detect_extended_topology_early(c) < 0) + detect_ht_early(c); +} + +static void bsp_init_intel(struct cpuinfo_x86 *c) +{ + resctrl_cpu_detect(c); +} + +#ifdef CONFIG_X86_32 +/* + * Early probe support logic for ppro memory erratum #50 + * + * This is called before we do cpu ident work + */ + +int ppro_with_ram_bug(void) +{ + /* Uses data from early_cpu_detect now */ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && + boot_cpu_data.x86 == 6 && + boot_cpu_data.x86_model == 1 && + boot_cpu_data.x86_stepping < 8) { + pr_info("Pentium Pro with Errata#50 detected. Taking evasive action.\n"); + return 1; + } + return 0; +} + +static void intel_smp_check(struct cpuinfo_x86 *c) +{ + /* calling is from identify_secondary_cpu() ? */ + if (!c->cpu_index) + return; + + /* + * Mask B, Pentium, but not Pentium MMX + */ + if (c->x86 == 5 && + c->x86_stepping >= 1 && c->x86_stepping <= 4 && + c->x86_model <= 3) { + /* + * Remember we have B step Pentia with bugs + */ + WARN_ONCE(1, "WARNING: SMP operation may be unreliable" + "with B stepping processors.\n"); + } +} + +static int forcepae; +static int __init forcepae_setup(char *__unused) +{ + forcepae = 1; + return 1; +} +__setup("forcepae", forcepae_setup); + +static void intel_workarounds(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_F00F_BUG + /* + * All models of Pentium and Pentium with MMX technology CPUs + * have the F0 0F bug, which lets nonprivileged users lock up the + * system. Announce that the fault handler will be checking for it. + * The Quark is also family 5, but does not have the same bug. + */ + clear_cpu_bug(c, X86_BUG_F00F); + if (c->x86 == 5 && c->x86_model < 9) { + static int f00f_workaround_enabled; + + set_cpu_bug(c, X86_BUG_F00F); + if (!f00f_workaround_enabled) { + pr_notice("Intel Pentium with F0 0F bug - workaround enabled.\n"); + f00f_workaround_enabled = 1; + } + } +#endif + + /* + * SEP CPUID bug: Pentium Pro reports SEP but doesn't have it until + * model 3 mask 3 + */ + if ((c->x86<<8 | c->x86_model<<4 | c->x86_stepping) < 0x633) + clear_cpu_cap(c, X86_FEATURE_SEP); + + /* + * PAE CPUID issue: many Pentium M report no PAE but may have a + * functionally usable PAE implementation. + * Forcefully enable PAE if kernel parameter "forcepae" is present. + */ + if (forcepae) { + pr_warn("PAE forced!\n"); + set_cpu_cap(c, X86_FEATURE_PAE); + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); + } + + /* + * P4 Xeon erratum 037 workaround. + * Hardware prefetcher may cause stale data to be loaded into the cache. + */ + if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_stepping == 1)) { + if (msr_set_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE_BIT) > 0) { + pr_info("CPU: C0 stepping P4 Xeon detected.\n"); + pr_info("CPU: Disabling hardware prefetching (Erratum 037)\n"); + } + } + + /* + * See if we have a good local APIC by checking for buggy Pentia, + * i.e. all B steppings and the C2 stepping of P54C when using their + * integrated APIC (see 11AP erratum in "Pentium Processor + * Specification Update"). + */ + if (boot_cpu_has(X86_FEATURE_APIC) && (c->x86<<8 | c->x86_model<<4) == 0x520 && + (c->x86_stepping < 0x6 || c->x86_stepping == 0xb)) + set_cpu_bug(c, X86_BUG_11AP); + + +#ifdef CONFIG_X86_INTEL_USERCOPY + /* + * Set up the preferred alignment for movsl bulk memory moves + */ + switch (c->x86) { + case 4: /* 486: untested */ + break; + case 5: /* Old Pentia: untested */ + break; + case 6: /* PII/PIII only like movsl with 8-byte alignment */ + movsl_mask.mask = 7; + break; + case 15: /* P4 is OK down to 8-byte alignment */ + movsl_mask.mask = 7; + break; + } +#endif + + intel_smp_check(c); +} +#else +static void intel_workarounds(struct cpuinfo_x86 *c) +{ +} +#endif + +static void srat_detect_node(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_NUMA + unsigned node; + int cpu = smp_processor_id(); + + /* Don't do the funky fallback heuristics the AMD version employs + for now. */ + node = numa_cpu_node(cpu); + if (node == NUMA_NO_NODE || !node_online(node)) { + /* reuse the value from init_cpu_to_node() */ + node = cpu_to_node(cpu); + } + numa_set_node(cpu, node); +#endif +} + +#define MSR_IA32_TME_ACTIVATE 0x982 + +/* Helpers to access TME_ACTIVATE MSR */ +#define TME_ACTIVATE_LOCKED(x) (x & 0x1) +#define TME_ACTIVATE_ENABLED(x) (x & 0x2) + +#define TME_ACTIVATE_POLICY(x) ((x >> 4) & 0xf) /* Bits 7:4 */ +#define TME_ACTIVATE_POLICY_AES_XTS_128 0 + +#define TME_ACTIVATE_KEYID_BITS(x) ((x >> 32) & 0xf) /* Bits 35:32 */ + +#define TME_ACTIVATE_CRYPTO_ALGS(x) ((x >> 48) & 0xffff) /* Bits 63:48 */ +#define TME_ACTIVATE_CRYPTO_AES_XTS_128 1 + +/* Values for mktme_status (SW only construct) */ +#define MKTME_ENABLED 0 +#define MKTME_DISABLED 1 +#define MKTME_UNINITIALIZED 2 +static int mktme_status = MKTME_UNINITIALIZED; + +static void detect_tme(struct cpuinfo_x86 *c) +{ + u64 tme_activate, tme_policy, tme_crypto_algs; + int keyid_bits = 0, nr_keyids = 0; + static u64 tme_activate_cpu0 = 0; + + rdmsrl(MSR_IA32_TME_ACTIVATE, tme_activate); + + if (mktme_status != MKTME_UNINITIALIZED) { + if (tme_activate != tme_activate_cpu0) { + /* Broken BIOS? */ + pr_err_once("x86/tme: configuration is inconsistent between CPUs\n"); + pr_err_once("x86/tme: MKTME is not usable\n"); + mktme_status = MKTME_DISABLED; + + /* Proceed. We may need to exclude bits from x86_phys_bits. */ + } + } else { + tme_activate_cpu0 = tme_activate; + } + + if (!TME_ACTIVATE_LOCKED(tme_activate) || !TME_ACTIVATE_ENABLED(tme_activate)) { + pr_info_once("x86/tme: not enabled by BIOS\n"); + mktme_status = MKTME_DISABLED; + return; + } + + if (mktme_status != MKTME_UNINITIALIZED) + goto detect_keyid_bits; + + pr_info("x86/tme: enabled by BIOS\n"); + + tme_policy = TME_ACTIVATE_POLICY(tme_activate); + if (tme_policy != TME_ACTIVATE_POLICY_AES_XTS_128) + pr_warn("x86/tme: Unknown policy is active: %#llx\n", tme_policy); + + tme_crypto_algs = TME_ACTIVATE_CRYPTO_ALGS(tme_activate); + if (!(tme_crypto_algs & TME_ACTIVATE_CRYPTO_AES_XTS_128)) { + pr_err("x86/mktme: No known encryption algorithm is supported: %#llx\n", + tme_crypto_algs); + mktme_status = MKTME_DISABLED; + } +detect_keyid_bits: + keyid_bits = TME_ACTIVATE_KEYID_BITS(tme_activate); + nr_keyids = (1UL << keyid_bits) - 1; + if (nr_keyids) { + pr_info_once("x86/mktme: enabled by BIOS\n"); + pr_info_once("x86/mktme: %d KeyIDs available\n", nr_keyids); + } else { + pr_info_once("x86/mktme: disabled by BIOS\n"); + } + + if (mktme_status == MKTME_UNINITIALIZED) { + /* MKTME is usable */ + mktme_status = MKTME_ENABLED; + } + + /* + * KeyID bits effectively lower the number of physical address + * bits. Update cpuinfo_x86::x86_phys_bits accordingly. + */ + c->x86_phys_bits -= keyid_bits; +} + +static void init_cpuid_fault(struct cpuinfo_x86 *c) +{ + u64 msr; + + if (!rdmsrl_safe(MSR_PLATFORM_INFO, &msr)) { + if (msr & MSR_PLATFORM_INFO_CPUID_FAULT) + set_cpu_cap(c, X86_FEATURE_CPUID_FAULT); + } +} + +static void init_intel_misc_features(struct cpuinfo_x86 *c) +{ + u64 msr; + + if (rdmsrl_safe(MSR_MISC_FEATURES_ENABLES, &msr)) + return; + + /* Clear all MISC features */ + this_cpu_write(msr_misc_features_shadow, 0); + + /* Check features and update capabilities and shadow control bits */ + init_cpuid_fault(c); + probe_xeon_phi_r3mwait(c); + + msr = this_cpu_read(msr_misc_features_shadow); + wrmsrl(MSR_MISC_FEATURES_ENABLES, msr); +} + +static void split_lock_init(void); + +static void init_intel(struct cpuinfo_x86 *c) +{ + early_init_intel(c); + + intel_workarounds(c); + + /* + * Detect the extended topology information if available. This + * will reinitialise the initial_apicid which will be used + * in init_intel_cacheinfo() + */ + detect_extended_topology(c); + + if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { + /* + * let's use the legacy cpuid vector 0x1 and 0x4 for topology + * detection. + */ + detect_num_cpu_cores(c); +#ifdef CONFIG_X86_32 + detect_ht(c); +#endif + } + + init_intel_cacheinfo(c); + + if (c->cpuid_level > 9) { + unsigned eax = cpuid_eax(10); + /* Check for version and the number of counters */ + if ((eax & 0xff) && (((eax>>8) & 0xff) > 1)) + set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); + } + + if (cpu_has(c, X86_FEATURE_XMM2)) + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); + + if (boot_cpu_has(X86_FEATURE_DS)) { + unsigned int l1, l2; + + rdmsr(MSR_IA32_MISC_ENABLE, l1, l2); + if (!(l1 & (1<<11))) + set_cpu_cap(c, X86_FEATURE_BTS); + if (!(l1 & (1<<12))) + set_cpu_cap(c, X86_FEATURE_PEBS); + } + + if (c->x86 == 6 && boot_cpu_has(X86_FEATURE_CLFLUSH) && + (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) + set_cpu_bug(c, X86_BUG_CLFLUSH_MONITOR); + + if (c->x86 == 6 && boot_cpu_has(X86_FEATURE_MWAIT) && + ((c->x86_model == INTEL_FAM6_ATOM_GOLDMONT))) + set_cpu_bug(c, X86_BUG_MONITOR); + +#ifdef CONFIG_X86_64 + if (c->x86 == 15) + c->x86_cache_alignment = c->x86_clflush_size * 2; + if (c->x86 == 6) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); +#else + /* + * Names for the Pentium II/Celeron processors + * detectable only by also checking the cache size. + * Dixon is NOT a Celeron. + */ + if (c->x86 == 6) { + unsigned int l2 = c->x86_cache_size; + char *p = NULL; + + switch (c->x86_model) { + case 5: + if (l2 == 0) + p = "Celeron (Covington)"; + else if (l2 == 256) + p = "Mobile Pentium II (Dixon)"; + break; + + case 6: + if (l2 == 128) + p = "Celeron (Mendocino)"; + else if (c->x86_stepping == 0 || c->x86_stepping == 5) + p = "Celeron-A"; + break; + + case 8: + if (l2 == 128) + p = "Celeron (Coppermine)"; + break; + } + + if (p) + strcpy(c->x86_model_id, p); + } + + if (c->x86 == 15) + set_cpu_cap(c, X86_FEATURE_P4); + if (c->x86 == 6) + set_cpu_cap(c, X86_FEATURE_P3); +#endif + + /* Work around errata */ + srat_detect_node(c); + + init_ia32_feat_ctl(c); + + if (cpu_has(c, X86_FEATURE_TME)) + detect_tme(c); + + init_intel_misc_features(c); + + if (tsx_ctrl_state == TSX_CTRL_ENABLE) + tsx_enable(); + if (tsx_ctrl_state == TSX_CTRL_DISABLE) + tsx_disable(); + + split_lock_init(); +} + +#ifdef CONFIG_X86_32 +static unsigned int intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) +{ + /* + * Intel PIII Tualatin. This comes in two flavours. + * One has 256kb of cache, the other 512. We have no way + * to determine which, so we use a boottime override + * for the 512kb model, and assume 256 otherwise. + */ + if ((c->x86 == 6) && (c->x86_model == 11) && (size == 0)) + size = 256; + + /* + * Intel Quark SoC X1000 contains a 4-way set associative + * 16K cache with a 16 byte cache line and 256 lines per tag + */ + if ((c->x86 == 5) && (c->x86_model == 9)) + size = 16; + return size; +} +#endif + +#define TLB_INST_4K 0x01 +#define TLB_INST_4M 0x02 +#define TLB_INST_2M_4M 0x03 + +#define TLB_INST_ALL 0x05 +#define TLB_INST_1G 0x06 + +#define TLB_DATA_4K 0x11 +#define TLB_DATA_4M 0x12 +#define TLB_DATA_2M_4M 0x13 +#define TLB_DATA_4K_4M 0x14 + +#define TLB_DATA_1G 0x16 + +#define TLB_DATA0_4K 0x21 +#define TLB_DATA0_4M 0x22 +#define TLB_DATA0_2M_4M 0x23 + +#define STLB_4K 0x41 +#define STLB_4K_2M 0x42 + +static const struct _tlb_table intel_tlb_table[] = { + { 0x01, TLB_INST_4K, 32, " TLB_INST 4 KByte pages, 4-way set associative" }, + { 0x02, TLB_INST_4M, 2, " TLB_INST 4 MByte pages, full associative" }, + { 0x03, TLB_DATA_4K, 64, " TLB_DATA 4 KByte pages, 4-way set associative" }, + { 0x04, TLB_DATA_4M, 8, " TLB_DATA 4 MByte pages, 4-way set associative" }, + { 0x05, TLB_DATA_4M, 32, " TLB_DATA 4 MByte pages, 4-way set associative" }, + { 0x0b, TLB_INST_4M, 4, " TLB_INST 4 MByte pages, 4-way set associative" }, + { 0x4f, TLB_INST_4K, 32, " TLB_INST 4 KByte pages" }, + { 0x50, TLB_INST_ALL, 64, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x51, TLB_INST_ALL, 128, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x52, TLB_INST_ALL, 256, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x55, TLB_INST_2M_4M, 7, " TLB_INST 2-MByte or 4-MByte pages, fully associative" }, + { 0x56, TLB_DATA0_4M, 16, " TLB_DATA0 4 MByte pages, 4-way set associative" }, + { 0x57, TLB_DATA0_4K, 16, " TLB_DATA0 4 KByte pages, 4-way associative" }, + { 0x59, TLB_DATA0_4K, 16, " TLB_DATA0 4 KByte pages, fully associative" }, + { 0x5a, TLB_DATA0_2M_4M, 32, " TLB_DATA0 2-MByte or 4 MByte pages, 4-way set associative" }, + { 0x5b, TLB_DATA_4K_4M, 64, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x5c, TLB_DATA_4K_4M, 128, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x5d, TLB_DATA_4K_4M, 256, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x61, TLB_INST_4K, 48, " TLB_INST 4 KByte pages, full associative" }, + { 0x63, TLB_DATA_1G, 4, " TLB_DATA 1 GByte pages, 4-way set associative" }, + { 0x6b, TLB_DATA_4K, 256, " TLB_DATA 4 KByte pages, 8-way associative" }, + { 0x6c, TLB_DATA_2M_4M, 128, " TLB_DATA 2 MByte or 4 MByte pages, 8-way associative" }, + { 0x6d, TLB_DATA_1G, 16, " TLB_DATA 1 GByte pages, fully associative" }, + { 0x76, TLB_INST_2M_4M, 8, " TLB_INST 2-MByte or 4-MByte pages, fully associative" }, + { 0xb0, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 4-way set associative" }, + { 0xb1, TLB_INST_2M_4M, 4, " TLB_INST 2M pages, 4-way, 8 entries or 4M pages, 4-way entries" }, + { 0xb2, TLB_INST_4K, 64, " TLB_INST 4KByte pages, 4-way set associative" }, + { 0xb3, TLB_DATA_4K, 128, " TLB_DATA 4 KByte pages, 4-way set associative" }, + { 0xb4, TLB_DATA_4K, 256, " TLB_DATA 4 KByte pages, 4-way associative" }, + { 0xb5, TLB_INST_4K, 64, " TLB_INST 4 KByte pages, 8-way set associative" }, + { 0xb6, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 8-way set associative" }, + { 0xba, TLB_DATA_4K, 64, " TLB_DATA 4 KByte pages, 4-way associative" }, + { 0xc0, TLB_DATA_4K_4M, 8, " TLB_DATA 4 KByte and 4 MByte pages, 4-way associative" }, + { 0xc1, STLB_4K_2M, 1024, " STLB 4 KByte and 2 MByte pages, 8-way associative" }, + { 0xc2, TLB_DATA_2M_4M, 16, " TLB_DATA 2 MByte/4MByte pages, 4-way associative" }, + { 0xca, STLB_4K, 512, " STLB 4 KByte pages, 4-way associative" }, + { 0x00, 0, 0 } +}; + +static void intel_tlb_lookup(const unsigned char desc) +{ + unsigned char k; + if (desc == 0) + return; + + /* look up this descriptor in the table */ + for (k = 0; intel_tlb_table[k].descriptor != desc && + intel_tlb_table[k].descriptor != 0; k++) + ; + + if (intel_tlb_table[k].tlb_type == 0) + return; + + switch (intel_tlb_table[k].tlb_type) { + case STLB_4K: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case STLB_4K_2M: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_ALL: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_4K: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_4M: + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_2M_4M: + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4K: + case TLB_DATA0_4K: + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4M: + case TLB_DATA0_4M: + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_2M_4M: + case TLB_DATA0_2M_4M: + if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4K_4M: + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_1G: + if (tlb_lld_1g[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_1g[ENTRIES] = intel_tlb_table[k].entries; + break; + } +} + +static void intel_detect_tlb(struct cpuinfo_x86 *c) +{ + int i, j, n; + unsigned int regs[4]; + unsigned char *desc = (unsigned char *)regs; + + if (c->cpuid_level < 2) + return; + + /* Number of times to iterate */ + n = cpuid_eax(2) & 0xFF; + + for (i = 0 ; i < n ; i++) { + cpuid(2, ®s[0], ®s[1], ®s[2], ®s[3]); + + /* If bit 31 is set, this is an unknown format */ + for (j = 0 ; j < 3 ; j++) + if (regs[j] & (1 << 31)) + regs[j] = 0; + + /* Byte 0 is level count, not a descriptor */ + for (j = 1 ; j < 16 ; j++) + intel_tlb_lookup(desc[j]); + } +} + +static const struct cpu_dev intel_cpu_dev = { + .c_vendor = "Intel", + .c_ident = { "GenuineIntel" }, +#ifdef CONFIG_X86_32 + .legacy_models = { + { .family = 4, .model_names = + { + [0] = "486 DX-25/33", + [1] = "486 DX-50", + [2] = "486 SX", + [3] = "486 DX/2", + [4] = "486 SL", + [5] = "486 SX/2", + [7] = "486 DX/2-WB", + [8] = "486 DX/4", + [9] = "486 DX/4-WB" + } + }, + { .family = 5, .model_names = + { + [0] = "Pentium 60/66 A-step", + [1] = "Pentium 60/66", + [2] = "Pentium 75 - 200", + [3] = "OverDrive PODP5V83", + [4] = "Pentium MMX", + [7] = "Mobile Pentium 75 - 200", + [8] = "Mobile Pentium MMX", + [9] = "Quark SoC X1000", + } + }, + { .family = 6, .model_names = + { + [0] = "Pentium Pro A-step", + [1] = "Pentium Pro", + [3] = "Pentium II (Klamath)", + [4] = "Pentium II (Deschutes)", + [5] = "Pentium II (Deschutes)", + [6] = "Mobile Pentium II", + [7] = "Pentium III (Katmai)", + [8] = "Pentium III (Coppermine)", + [10] = "Pentium III (Cascades)", + [11] = "Pentium III (Tualatin)", + } + }, + { .family = 15, .model_names = + { + [0] = "Pentium 4 (Unknown)", + [1] = "Pentium 4 (Willamette)", + [2] = "Pentium 4 (Northwood)", + [4] = "Pentium 4 (Foster)", + [5] = "Pentium 4 (Foster)", + } + }, + }, + .legacy_cache_size = intel_size_cache, +#endif + .c_detect_tlb = intel_detect_tlb, + .c_early_init = early_init_intel, + .c_bsp_init = bsp_init_intel, + .c_init = init_intel, + .c_x86_vendor = X86_VENDOR_INTEL, +}; + +cpu_dev_register(intel_cpu_dev); + +#undef pr_fmt +#define pr_fmt(fmt) "x86/split lock detection: " fmt + +static const struct { + const char *option; + enum split_lock_detect_state state; +} sld_options[] __initconst = { + { "off", sld_off }, + { "warn", sld_warn }, + { "fatal", sld_fatal }, +}; + +static inline bool match_option(const char *arg, int arglen, const char *opt) +{ + int len = strlen(opt); + + return len == arglen && !strncmp(arg, opt, len); +} + +static bool split_lock_verify_msr(bool on) +{ + u64 ctrl, tmp; + + if (rdmsrl_safe(MSR_TEST_CTRL, &ctrl)) + return false; + if (on) + ctrl |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + else + ctrl &= ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + if (wrmsrl_safe(MSR_TEST_CTRL, ctrl)) + return false; + rdmsrl(MSR_TEST_CTRL, tmp); + return ctrl == tmp; +} + +static void __init split_lock_setup(void) +{ + enum split_lock_detect_state state = sld_warn; + char arg[20]; + int i, ret; + + if (!split_lock_verify_msr(false)) { + pr_info("MSR access failed: Disabled\n"); + return; + } + + ret = cmdline_find_option(boot_command_line, "split_lock_detect", + arg, sizeof(arg)); + if (ret >= 0) { + for (i = 0; i < ARRAY_SIZE(sld_options); i++) { + if (match_option(arg, ret, sld_options[i].option)) { + state = sld_options[i].state; + break; + } + } + } + + switch (state) { + case sld_off: + pr_info("disabled\n"); + return; + case sld_warn: + pr_info("warning about user-space split_locks\n"); + break; + case sld_fatal: + pr_info("sending SIGBUS on user-space split_locks\n"); + break; + } + + rdmsrl(MSR_TEST_CTRL, msr_test_ctrl_cache); + + if (!split_lock_verify_msr(true)) { + pr_info("MSR access failed: Disabled\n"); + return; + } + + sld_state = state; + setup_force_cpu_cap(X86_FEATURE_SPLIT_LOCK_DETECT); +} + +/* + * MSR_TEST_CTRL is per core, but we treat it like a per CPU MSR. Locking + * is not implemented as one thread could undo the setting of the other + * thread immediately after dropping the lock anyway. + */ +static void sld_update_msr(bool on) +{ + u64 test_ctrl_val = msr_test_ctrl_cache; + + if (on) + test_ctrl_val |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + + wrmsrl(MSR_TEST_CTRL, test_ctrl_val); +} + +static void split_lock_init(void) +{ + if (cpu_model_supports_sld) + split_lock_verify_msr(sld_state != sld_off); +} + +static void split_lock_warn(unsigned long ip) +{ + pr_warn_ratelimited("#AC: %s/%d took a split_lock trap at address: 0x%lx\n", + current->comm, current->pid, ip); + + /* + * Disable the split lock detection for this task so it can make + * progress and set TIF_SLD so the detection is re-enabled via + * switch_to_sld() when the task is scheduled out. + */ + sld_update_msr(false); + set_tsk_thread_flag(current, TIF_SLD); +} + +bool handle_guest_split_lock(unsigned long ip) +{ + if (sld_state == sld_warn) { + split_lock_warn(ip); + return true; + } + + pr_warn_once("#AC: %s/%d %s split_lock trap at address: 0x%lx\n", + current->comm, current->pid, + sld_state == sld_fatal ? "fatal" : "bogus", ip); + + current->thread.error_code = 0; + current->thread.trap_nr = X86_TRAP_AC; + force_sig_fault(SIGBUS, BUS_ADRALN, NULL); + return false; +} +EXPORT_SYMBOL_GPL(handle_guest_split_lock); + +bool handle_user_split_lock(struct pt_regs *regs, long error_code) +{ + if ((regs->flags & X86_EFLAGS_AC) || sld_state == sld_fatal) + return false; + split_lock_warn(regs->ip); + return true; +} + +/* + * This function is called only when switching between tasks with + * different split-lock detection modes. It sets the MSR for the + * mode of the new task. This is right most of the time, but since + * the MSR is shared by hyperthreads on a physical core there can + * be glitches when the two threads need different modes. + */ +void switch_to_sld(unsigned long tifn) +{ + sld_update_msr(!(tifn & _TIF_SLD)); +} + +/* + * Bits in the IA32_CORE_CAPABILITIES are not architectural, so they should + * only be trusted if it is confirmed that a CPU model implements a + * specific feature at a particular bit position. + * + * The possible driver data field values: + * + * - 0: CPU models that are known to have the per-core split-lock detection + * feature even though they do not enumerate IA32_CORE_CAPABILITIES. + * + * - 1: CPU models which may enumerate IA32_CORE_CAPABILITIES and if so use + * bit 5 to enumerate the per-core split-lock detection feature. + */ +static const struct x86_cpu_id split_lock_cpu_ids[] __initconst = { + X86_MATCH_INTEL_FAM6_MODEL(ICELAKE_X, 0), + X86_MATCH_INTEL_FAM6_MODEL(ICELAKE_L, 0), + X86_MATCH_INTEL_FAM6_MODEL(ICELAKE_D, 0), + X86_MATCH_INTEL_FAM6_MODEL(ATOM_TREMONT, 1), + X86_MATCH_INTEL_FAM6_MODEL(ATOM_TREMONT_D, 1), + X86_MATCH_INTEL_FAM6_MODEL(ATOM_TREMONT_L, 1), + X86_MATCH_INTEL_FAM6_MODEL(TIGERLAKE_L, 1), + X86_MATCH_INTEL_FAM6_MODEL(TIGERLAKE, 1), + X86_MATCH_INTEL_FAM6_MODEL(SAPPHIRERAPIDS_X, 1), + X86_MATCH_INTEL_FAM6_MODEL(ALDERLAKE, 1), + X86_MATCH_INTEL_FAM6_MODEL(ALDERLAKE_L, 1), + {} +}; + +void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c) +{ + const struct x86_cpu_id *m; + u64 ia32_core_caps; + + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) + return; + + m = x86_match_cpu(split_lock_cpu_ids); + if (!m) + return; + + switch (m->driver_data) { + case 0: + break; + case 1: + if (!cpu_has(c, X86_FEATURE_CORE_CAPABILITIES)) + return; + rdmsrl(MSR_IA32_CORE_CAPS, ia32_core_caps); + if (!(ia32_core_caps & MSR_IA32_CORE_CAPS_SPLIT_LOCK_DETECT)) + return; + break; + default: + return; + } + + cpu_model_supports_sld = true; + split_lock_setup(); +} diff --git a/arch/x86/kernel/cpu/intel_epb.c b/arch/x86/kernel/cpu/intel_epb.c new file mode 100644 index 000000000..f4dd73396 --- /dev/null +++ b/arch/x86/kernel/cpu/intel_epb.c @@ -0,0 +1,216 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Performance and Energy Bias Hint support. + * + * Copyright (C) 2019 Intel Corporation + * + * Author: + * Rafael J. Wysocki <rafael.j.wysocki@intel.com> + */ + +#include <linux/cpuhotplug.h> +#include <linux/cpu.h> +#include <linux/device.h> +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/syscore_ops.h> +#include <linux/pm.h> + +#include <asm/cpufeature.h> +#include <asm/msr.h> + +/** + * DOC: overview + * + * The Performance and Energy Bias Hint (EPB) allows software to specify its + * preference with respect to the power-performance tradeoffs present in the + * processor. Generally, the EPB is expected to be set by user space (directly + * via sysfs or with the help of the x86_energy_perf_policy tool), but there are + * two reasons for the kernel to update it. + * + * First, there are systems where the platform firmware resets the EPB during + * system-wide transitions from sleep states back into the working state + * effectively causing the previous EPB updates by user space to be lost. + * Thus the kernel needs to save the current EPB values for all CPUs during + * system-wide transitions to sleep states and restore them on the way back to + * the working state. That can be achieved by saving EPB for secondary CPUs + * when they are taken offline during transitions into system sleep states and + * for the boot CPU in a syscore suspend operation, so that it can be restored + * for the boot CPU in a syscore resume operation and for the other CPUs when + * they are brought back online. However, CPUs that are already offline when + * a system-wide PM transition is started are not taken offline again, but their + * EPB values may still be reset by the platform firmware during the transition, + * so in fact it is necessary to save the EPB of any CPU taken offline and to + * restore it when the given CPU goes back online at all times. + * + * Second, on many systems the initial EPB value coming from the platform + * firmware is 0 ('performance') and at least on some of them that is because + * the platform firmware does not initialize EPB at all with the assumption that + * the OS will do that anyway. That sometimes is problematic, as it may cause + * the system battery to drain too fast, for example, so it is better to adjust + * it on CPU bring-up and if the initial EPB value for a given CPU is 0, the + * kernel changes it to 6 ('normal'). + */ + +static DEFINE_PER_CPU(u8, saved_epb); + +#define EPB_MASK 0x0fULL +#define EPB_SAVED 0x10ULL +#define MAX_EPB EPB_MASK + +static int intel_epb_save(void) +{ + u64 epb; + + rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb); + /* + * Ensure that saved_epb will always be nonzero after this write even if + * the EPB value read from the MSR is 0. + */ + this_cpu_write(saved_epb, (epb & EPB_MASK) | EPB_SAVED); + + return 0; +} + +static void intel_epb_restore(void) +{ + u64 val = this_cpu_read(saved_epb); + u64 epb; + + rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb); + if (val) { + val &= EPB_MASK; + } else { + /* + * Because intel_epb_save() has not run for the current CPU yet, + * it is going online for the first time, so if its EPB value is + * 0 ('performance') at this point, assume that it has not been + * initialized by the platform firmware and set it to 6 + * ('normal'). + */ + val = epb & EPB_MASK; + if (val == ENERGY_PERF_BIAS_PERFORMANCE) { + val = ENERGY_PERF_BIAS_NORMAL; + pr_warn_once("ENERGY_PERF_BIAS: Set to 'normal', was 'performance'\n"); + } + } + wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, (epb & ~EPB_MASK) | val); +} + +static struct syscore_ops intel_epb_syscore_ops = { + .suspend = intel_epb_save, + .resume = intel_epb_restore, +}; + +static const char * const energy_perf_strings[] = { + "performance", + "balance-performance", + "normal", + "balance-power", + "power" +}; +static const u8 energ_perf_values[] = { + ENERGY_PERF_BIAS_PERFORMANCE, + ENERGY_PERF_BIAS_BALANCE_PERFORMANCE, + ENERGY_PERF_BIAS_NORMAL, + ENERGY_PERF_BIAS_BALANCE_POWERSAVE, + ENERGY_PERF_BIAS_POWERSAVE +}; + +static ssize_t energy_perf_bias_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + unsigned int cpu = dev->id; + u64 epb; + int ret; + + ret = rdmsrl_on_cpu(cpu, MSR_IA32_ENERGY_PERF_BIAS, &epb); + if (ret < 0) + return ret; + + return sprintf(buf, "%llu\n", epb); +} + +static ssize_t energy_perf_bias_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + unsigned int cpu = dev->id; + u64 epb, val; + int ret; + + ret = __sysfs_match_string(energy_perf_strings, + ARRAY_SIZE(energy_perf_strings), buf); + if (ret >= 0) + val = energ_perf_values[ret]; + else if (kstrtou64(buf, 0, &val) || val > MAX_EPB) + return -EINVAL; + + ret = rdmsrl_on_cpu(cpu, MSR_IA32_ENERGY_PERF_BIAS, &epb); + if (ret < 0) + return ret; + + ret = wrmsrl_on_cpu(cpu, MSR_IA32_ENERGY_PERF_BIAS, + (epb & ~EPB_MASK) | val); + if (ret < 0) + return ret; + + return count; +} + +static DEVICE_ATTR_RW(energy_perf_bias); + +static struct attribute *intel_epb_attrs[] = { + &dev_attr_energy_perf_bias.attr, + NULL +}; + +static const struct attribute_group intel_epb_attr_group = { + .name = power_group_name, + .attrs = intel_epb_attrs +}; + +static int intel_epb_online(unsigned int cpu) +{ + struct device *cpu_dev = get_cpu_device(cpu); + + intel_epb_restore(); + if (!cpuhp_tasks_frozen) + sysfs_merge_group(&cpu_dev->kobj, &intel_epb_attr_group); + + return 0; +} + +static int intel_epb_offline(unsigned int cpu) +{ + struct device *cpu_dev = get_cpu_device(cpu); + + if (!cpuhp_tasks_frozen) + sysfs_unmerge_group(&cpu_dev->kobj, &intel_epb_attr_group); + + intel_epb_save(); + return 0; +} + +static __init int intel_epb_init(void) +{ + int ret; + + if (!boot_cpu_has(X86_FEATURE_EPB)) + return -ENODEV; + + ret = cpuhp_setup_state(CPUHP_AP_X86_INTEL_EPB_ONLINE, + "x86/intel/epb:online", intel_epb_online, + intel_epb_offline); + if (ret < 0) + goto err_out_online; + + register_syscore_ops(&intel_epb_syscore_ops); + return 0; + +err_out_online: + cpuhp_remove_state(CPUHP_AP_X86_INTEL_EPB_ONLINE); + return ret; +} +subsys_initcall(intel_epb_init); diff --git a/arch/x86/kernel/cpu/intel_pconfig.c b/arch/x86/kernel/cpu/intel_pconfig.c new file mode 100644 index 000000000..0771a905b --- /dev/null +++ b/arch/x86/kernel/cpu/intel_pconfig.c @@ -0,0 +1,82 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel PCONFIG instruction support. + * + * Copyright (C) 2017 Intel Corporation + * + * Author: + * Kirill A. Shutemov <kirill.shutemov@linux.intel.com> + */ + +#include <asm/cpufeature.h> +#include <asm/intel_pconfig.h> + +#define PCONFIG_CPUID 0x1b + +#define PCONFIG_CPUID_SUBLEAF_MASK ((1 << 12) - 1) + +/* Subleaf type (EAX) for PCONFIG CPUID leaf (0x1B) */ +enum { + PCONFIG_CPUID_SUBLEAF_INVALID = 0, + PCONFIG_CPUID_SUBLEAF_TARGETID = 1, +}; + +/* Bitmask of supported targets */ +static u64 targets_supported __read_mostly; + +int pconfig_target_supported(enum pconfig_target target) +{ + /* + * We would need to re-think the implementation once we get > 64 + * PCONFIG targets. Spec allows up to 2^32 targets. + */ + BUILD_BUG_ON(PCONFIG_TARGET_NR >= 64); + + if (WARN_ON_ONCE(target >= 64)) + return 0; + return targets_supported & (1ULL << target); +} + +static int __init intel_pconfig_init(void) +{ + int subleaf; + + if (!boot_cpu_has(X86_FEATURE_PCONFIG)) + return 0; + + /* + * Scan subleafs of PCONFIG CPUID leaf. + * + * Subleafs of the same type need not to be consecutive. + * + * Stop on the first invalid subleaf type. All subleafs after the first + * invalid are invalid too. + */ + for (subleaf = 0; subleaf < INT_MAX; subleaf++) { + struct cpuid_regs regs; + + cpuid_count(PCONFIG_CPUID, subleaf, + ®s.eax, ®s.ebx, ®s.ecx, ®s.edx); + + switch (regs.eax & PCONFIG_CPUID_SUBLEAF_MASK) { + case PCONFIG_CPUID_SUBLEAF_INVALID: + /* Stop on the first invalid subleaf */ + goto out; + case PCONFIG_CPUID_SUBLEAF_TARGETID: + /* Mark supported PCONFIG targets */ + if (regs.ebx < 64) + targets_supported |= (1ULL << regs.ebx); + if (regs.ecx < 64) + targets_supported |= (1ULL << regs.ecx); + if (regs.edx < 64) + targets_supported |= (1ULL << regs.edx); + break; + default: + /* Unknown CPUID.PCONFIG subleaf: ignore */ + break; + } + } +out: + return 0; +} +arch_initcall(intel_pconfig_init); diff --git a/arch/x86/kernel/cpu/match.c b/arch/x86/kernel/cpu/match.c new file mode 100644 index 000000000..ad6776081 --- /dev/null +++ b/arch/x86/kernel/cpu/match.c @@ -0,0 +1,91 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <asm/cpu_device_id.h> +#include <asm/cpufeature.h> +#include <linux/cpu.h> +#include <linux/export.h> +#include <linux/slab.h> + +/** + * x86_match_cpu - match current CPU again an array of x86_cpu_ids + * @match: Pointer to array of x86_cpu_ids. Last entry terminated with + * {}. + * + * Return the entry if the current CPU matches the entries in the + * passed x86_cpu_id match table. Otherwise NULL. The match table + * contains vendor (X86_VENDOR_*), family, model and feature bits or + * respective wildcard entries. + * + * A typical table entry would be to match a specific CPU + * + * X86_MATCH_VENDOR_FAM_MODEL_FEATURE(INTEL, 6, INTEL_FAM6_BROADWELL, + * X86_FEATURE_ANY, NULL); + * + * Fields can be wildcarded with %X86_VENDOR_ANY, %X86_FAMILY_ANY, + * %X86_MODEL_ANY, %X86_FEATURE_ANY (except for vendor) + * + * asm/cpu_device_id.h contains a set of useful macros which are shortcuts + * for various common selections. The above can be shortened to: + * + * X86_MATCH_INTEL_FAM6_MODEL(BROADWELL, NULL); + * + * Arrays used to match for this should also be declared using + * MODULE_DEVICE_TABLE(x86cpu, ...) + * + * This always matches against the boot cpu, assuming models and features are + * consistent over all CPUs. + */ +const struct x86_cpu_id *x86_match_cpu(const struct x86_cpu_id *match) +{ + const struct x86_cpu_id *m; + struct cpuinfo_x86 *c = &boot_cpu_data; + + for (m = match; + m->vendor | m->family | m->model | m->steppings | m->feature; + m++) { + if (m->vendor != X86_VENDOR_ANY && c->x86_vendor != m->vendor) + continue; + if (m->family != X86_FAMILY_ANY && c->x86 != m->family) + continue; + if (m->model != X86_MODEL_ANY && c->x86_model != m->model) + continue; + if (m->steppings != X86_STEPPING_ANY && + !(BIT(c->x86_stepping) & m->steppings)) + continue; + if (m->feature != X86_FEATURE_ANY && !cpu_has(c, m->feature)) + continue; + return m; + } + return NULL; +} +EXPORT_SYMBOL(x86_match_cpu); + +static const struct x86_cpu_desc * +x86_match_cpu_with_stepping(const struct x86_cpu_desc *match) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + const struct x86_cpu_desc *m; + + for (m = match; m->x86_family | m->x86_model; m++) { + if (c->x86_vendor != m->x86_vendor) + continue; + if (c->x86 != m->x86_family) + continue; + if (c->x86_model != m->x86_model) + continue; + if (c->x86_stepping != m->x86_stepping) + continue; + return m; + } + return NULL; +} + +bool x86_cpu_has_min_microcode_rev(const struct x86_cpu_desc *table) +{ + const struct x86_cpu_desc *res = x86_match_cpu_with_stepping(table); + + if (!res || res->x86_microcode_rev > boot_cpu_data.microcode) + return false; + + return true; +} +EXPORT_SYMBOL_GPL(x86_cpu_has_min_microcode_rev); diff --git a/arch/x86/kernel/cpu/mce/Makefile b/arch/x86/kernel/cpu/mce/Makefile new file mode 100644 index 000000000..9f020c994 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/Makefile @@ -0,0 +1,16 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-y = core.o severity.o genpool.o + +obj-$(CONFIG_X86_ANCIENT_MCE) += winchip.o p5.o +obj-$(CONFIG_X86_MCE_INTEL) += intel.o +obj-$(CONFIG_X86_MCE_AMD) += amd.o +obj-$(CONFIG_X86_MCE_THRESHOLD) += threshold.o + +mce-inject-y := inject.o +obj-$(CONFIG_X86_MCE_INJECT) += mce-inject.o + +obj-$(CONFIG_X86_THERMAL_VECTOR) += therm_throt.o + +obj-$(CONFIG_ACPI_APEI) += apei.o + +obj-$(CONFIG_X86_MCELOG_LEGACY) += dev-mcelog.o diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c new file mode 100644 index 000000000..cd8db6b9c --- /dev/null +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -0,0 +1,1541 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * (c) 2005-2016 Advanced Micro Devices, Inc. + * + * Written by Jacob Shin - AMD, Inc. + * Maintained by: Borislav Petkov <bp@alien8.de> + * + * All MC4_MISCi registers are shared between cores on a node. + */ +#include <linux/interrupt.h> +#include <linux/notifier.h> +#include <linux/kobject.h> +#include <linux/percpu.h> +#include <linux/errno.h> +#include <linux/sched.h> +#include <linux/sysfs.h> +#include <linux/slab.h> +#include <linux/init.h> +#include <linux/cpu.h> +#include <linux/smp.h> +#include <linux/string.h> + +#include <asm/amd_nb.h> +#include <asm/traps.h> +#include <asm/apic.h> +#include <asm/mce.h> +#include <asm/msr.h> +#include <asm/trace/irq_vectors.h> + +#include "internal.h" + +#define NR_BLOCKS 5 +#define THRESHOLD_MAX 0xFFF +#define INT_TYPE_APIC 0x00020000 +#define MASK_VALID_HI 0x80000000 +#define MASK_CNTP_HI 0x40000000 +#define MASK_LOCKED_HI 0x20000000 +#define MASK_LVTOFF_HI 0x00F00000 +#define MASK_COUNT_EN_HI 0x00080000 +#define MASK_INT_TYPE_HI 0x00060000 +#define MASK_OVERFLOW_HI 0x00010000 +#define MASK_ERR_COUNT_HI 0x00000FFF +#define MASK_BLKPTR_LO 0xFF000000 +#define MCG_XBLK_ADDR 0xC0000400 + +/* Deferred error settings */ +#define MSR_CU_DEF_ERR 0xC0000410 +#define MASK_DEF_LVTOFF 0x000000F0 +#define MASK_DEF_INT_TYPE 0x00000006 +#define DEF_LVT_OFF 0x2 +#define DEF_INT_TYPE_APIC 0x2 + +/* Scalable MCA: */ + +/* Threshold LVT offset is at MSR0xC0000410[15:12] */ +#define SMCA_THR_LVT_OFF 0xF000 + +static bool thresholding_irq_en; + +static const char * const th_names[] = { + "load_store", + "insn_fetch", + "combined_unit", + "decode_unit", + "northbridge", + "execution_unit", +}; + +static const char * const smca_umc_block_names[] = { + "dram_ecc", + "misc_umc" +}; + +struct smca_bank_name { + const char *name; /* Short name for sysfs */ + const char *long_name; /* Long name for pretty-printing */ +}; + +static struct smca_bank_name smca_names[] = { + [SMCA_LS] = { "load_store", "Load Store Unit" }, + [SMCA_LS_V2] = { "load_store", "Load Store Unit" }, + [SMCA_IF] = { "insn_fetch", "Instruction Fetch Unit" }, + [SMCA_L2_CACHE] = { "l2_cache", "L2 Cache" }, + [SMCA_DE] = { "decode_unit", "Decode Unit" }, + [SMCA_RESERVED] = { "reserved", "Reserved" }, + [SMCA_EX] = { "execution_unit", "Execution Unit" }, + [SMCA_FP] = { "floating_point", "Floating Point Unit" }, + [SMCA_L3_CACHE] = { "l3_cache", "L3 Cache" }, + [SMCA_CS] = { "coherent_slave", "Coherent Slave" }, + [SMCA_CS_V2] = { "coherent_slave", "Coherent Slave" }, + [SMCA_PIE] = { "pie", "Power, Interrupts, etc." }, + [SMCA_UMC] = { "umc", "Unified Memory Controller" }, + [SMCA_PB] = { "param_block", "Parameter Block" }, + [SMCA_PSP] = { "psp", "Platform Security Processor" }, + [SMCA_PSP_V2] = { "psp", "Platform Security Processor" }, + [SMCA_SMU] = { "smu", "System Management Unit" }, + [SMCA_SMU_V2] = { "smu", "System Management Unit" }, + [SMCA_MP5] = { "mp5", "Microprocessor 5 Unit" }, + [SMCA_NBIO] = { "nbio", "Northbridge IO Unit" }, + [SMCA_PCIE] = { "pcie", "PCI Express Unit" }, +}; + +static const char *smca_get_name(enum smca_bank_types t) +{ + if (t >= N_SMCA_BANK_TYPES) + return NULL; + + return smca_names[t].name; +} + +const char *smca_get_long_name(enum smca_bank_types t) +{ + if (t >= N_SMCA_BANK_TYPES) + return NULL; + + return smca_names[t].long_name; +} +EXPORT_SYMBOL_GPL(smca_get_long_name); + +static enum smca_bank_types smca_get_bank_type(unsigned int bank) +{ + struct smca_bank *b; + + if (bank >= MAX_NR_BANKS) + return N_SMCA_BANK_TYPES; + + b = &smca_banks[bank]; + if (!b->hwid) + return N_SMCA_BANK_TYPES; + + return b->hwid->bank_type; +} + +static struct smca_hwid smca_hwid_mcatypes[] = { + /* { bank_type, hwid_mcatype } */ + + /* Reserved type */ + { SMCA_RESERVED, HWID_MCATYPE(0x00, 0x0) }, + + /* ZN Core (HWID=0xB0) MCA types */ + { SMCA_LS, HWID_MCATYPE(0xB0, 0x0) }, + { SMCA_LS_V2, HWID_MCATYPE(0xB0, 0x10) }, + { SMCA_IF, HWID_MCATYPE(0xB0, 0x1) }, + { SMCA_L2_CACHE, HWID_MCATYPE(0xB0, 0x2) }, + { SMCA_DE, HWID_MCATYPE(0xB0, 0x3) }, + /* HWID 0xB0 MCATYPE 0x4 is Reserved */ + { SMCA_EX, HWID_MCATYPE(0xB0, 0x5) }, + { SMCA_FP, HWID_MCATYPE(0xB0, 0x6) }, + { SMCA_L3_CACHE, HWID_MCATYPE(0xB0, 0x7) }, + + /* Data Fabric MCA types */ + { SMCA_CS, HWID_MCATYPE(0x2E, 0x0) }, + { SMCA_PIE, HWID_MCATYPE(0x2E, 0x1) }, + { SMCA_CS_V2, HWID_MCATYPE(0x2E, 0x2) }, + + /* Unified Memory Controller MCA type */ + { SMCA_UMC, HWID_MCATYPE(0x96, 0x0) }, + + /* Parameter Block MCA type */ + { SMCA_PB, HWID_MCATYPE(0x05, 0x0) }, + + /* Platform Security Processor MCA type */ + { SMCA_PSP, HWID_MCATYPE(0xFF, 0x0) }, + { SMCA_PSP_V2, HWID_MCATYPE(0xFF, 0x1) }, + + /* System Management Unit MCA type */ + { SMCA_SMU, HWID_MCATYPE(0x01, 0x0) }, + { SMCA_SMU_V2, HWID_MCATYPE(0x01, 0x1) }, + + /* Microprocessor 5 Unit MCA type */ + { SMCA_MP5, HWID_MCATYPE(0x01, 0x2) }, + + /* Northbridge IO Unit MCA type */ + { SMCA_NBIO, HWID_MCATYPE(0x18, 0x0) }, + + /* PCI Express Unit MCA type */ + { SMCA_PCIE, HWID_MCATYPE(0x46, 0x0) }, +}; + +struct smca_bank smca_banks[MAX_NR_BANKS]; +EXPORT_SYMBOL_GPL(smca_banks); + +/* + * In SMCA enabled processors, we can have multiple banks for a given IP type. + * So to define a unique name for each bank, we use a temp c-string to append + * the MCA_IPID[InstanceId] to type's name in get_name(). + * + * InstanceId is 32 bits which is 8 characters. Make sure MAX_MCATYPE_NAME_LEN + * is greater than 8 plus 1 (for underscore) plus length of longest type name. + */ +#define MAX_MCATYPE_NAME_LEN 30 +static char buf_mcatype[MAX_MCATYPE_NAME_LEN]; + +static DEFINE_PER_CPU(struct threshold_bank **, threshold_banks); + +/* + * A list of the banks enabled on each logical CPU. Controls which respective + * descriptors to initialize later in mce_threshold_create_device(). + */ +static DEFINE_PER_CPU(u64, bank_map); + +/* Map of banks that have more than MCA_MISC0 available. */ +static DEFINE_PER_CPU(u64, smca_misc_banks_map); + +static void amd_threshold_interrupt(void); +static void amd_deferred_error_interrupt(void); + +static void default_deferred_error_interrupt(void) +{ + pr_err("Unexpected deferred interrupt at vector %x\n", DEFERRED_ERROR_VECTOR); +} +void (*deferred_error_int_vector)(void) = default_deferred_error_interrupt; + +static void smca_set_misc_banks_map(unsigned int bank, unsigned int cpu) +{ + u32 low, high; + + /* + * For SMCA enabled processors, BLKPTR field of the first MISC register + * (MCx_MISC0) indicates presence of additional MISC regs set (MISC1-4). + */ + if (rdmsr_safe(MSR_AMD64_SMCA_MCx_CONFIG(bank), &low, &high)) + return; + + if (!(low & MCI_CONFIG_MCAX)) + return; + + if (rdmsr_safe(MSR_AMD64_SMCA_MCx_MISC(bank), &low, &high)) + return; + + if (low & MASK_BLKPTR_LO) + per_cpu(smca_misc_banks_map, cpu) |= BIT_ULL(bank); + +} + +static void smca_configure(unsigned int bank, unsigned int cpu) +{ + unsigned int i, hwid_mcatype; + struct smca_hwid *s_hwid; + u32 high, low; + u32 smca_config = MSR_AMD64_SMCA_MCx_CONFIG(bank); + + /* Set appropriate bits in MCA_CONFIG */ + if (!rdmsr_safe(smca_config, &low, &high)) { + /* + * OS is required to set the MCAX bit to acknowledge that it is + * now using the new MSR ranges and new registers under each + * bank. It also means that the OS will configure deferred + * errors in the new MCx_CONFIG register. If the bit is not set, + * uncorrectable errors will cause a system panic. + * + * MCA_CONFIG[MCAX] is bit 32 (0 in the high portion of the MSR.) + */ + high |= BIT(0); + + /* + * SMCA sets the Deferred Error Interrupt type per bank. + * + * MCA_CONFIG[DeferredIntTypeSupported] is bit 5, and tells us + * if the DeferredIntType bit field is available. + * + * MCA_CONFIG[DeferredIntType] is bits [38:37] ([6:5] in the + * high portion of the MSR). OS should set this to 0x1 to enable + * APIC based interrupt. First, check that no interrupt has been + * set. + */ + if ((low & BIT(5)) && !((high >> 5) & 0x3)) + high |= BIT(5); + + wrmsr(smca_config, low, high); + } + + smca_set_misc_banks_map(bank, cpu); + + /* Return early if this bank was already initialized. */ + if (smca_banks[bank].hwid && smca_banks[bank].hwid->hwid_mcatype != 0) + return; + + if (rdmsr_safe(MSR_AMD64_SMCA_MCx_IPID(bank), &low, &high)) { + pr_warn("Failed to read MCA_IPID for bank %d\n", bank); + return; + } + + hwid_mcatype = HWID_MCATYPE(high & MCI_IPID_HWID, + (high & MCI_IPID_MCATYPE) >> 16); + + for (i = 0; i < ARRAY_SIZE(smca_hwid_mcatypes); i++) { + s_hwid = &smca_hwid_mcatypes[i]; + if (hwid_mcatype == s_hwid->hwid_mcatype) { + smca_banks[bank].hwid = s_hwid; + smca_banks[bank].id = low; + smca_banks[bank].sysfs_id = s_hwid->count++; + break; + } + } +} + +struct thresh_restart { + struct threshold_block *b; + int reset; + int set_lvt_off; + int lvt_off; + u16 old_limit; +}; + +static inline bool is_shared_bank(int bank) +{ + /* + * Scalable MCA provides for only one core to have access to the MSRs of + * a shared bank. + */ + if (mce_flags.smca) + return false; + + /* Bank 4 is for northbridge reporting and is thus shared */ + return (bank == 4); +} + +static const char *bank4_names(const struct threshold_block *b) +{ + switch (b->address) { + /* MSR4_MISC0 */ + case 0x00000413: + return "dram"; + + case 0xc0000408: + return "ht_links"; + + case 0xc0000409: + return "l3_cache"; + + default: + WARN(1, "Funny MSR: 0x%08x\n", b->address); + return ""; + } +}; + + +static bool lvt_interrupt_supported(unsigned int bank, u32 msr_high_bits) +{ + /* + * bank 4 supports APIC LVT interrupts implicitly since forever. + */ + if (bank == 4) + return true; + + /* + * IntP: interrupt present; if this bit is set, the thresholding + * bank can generate APIC LVT interrupts + */ + return msr_high_bits & BIT(28); +} + +static int lvt_off_valid(struct threshold_block *b, int apic, u32 lo, u32 hi) +{ + int msr = (hi & MASK_LVTOFF_HI) >> 20; + + if (apic < 0) { + pr_err(FW_BUG "cpu %d, failed to setup threshold interrupt " + "for bank %d, block %d (MSR%08X=0x%x%08x)\n", b->cpu, + b->bank, b->block, b->address, hi, lo); + return 0; + } + + if (apic != msr) { + /* + * On SMCA CPUs, LVT offset is programmed at a different MSR, and + * the BIOS provides the value. The original field where LVT offset + * was set is reserved. Return early here: + */ + if (mce_flags.smca) + return 0; + + pr_err(FW_BUG "cpu %d, invalid threshold interrupt offset %d " + "for bank %d, block %d (MSR%08X=0x%x%08x)\n", + b->cpu, apic, b->bank, b->block, b->address, hi, lo); + return 0; + } + + return 1; +}; + +/* Reprogram MCx_MISC MSR behind this threshold bank. */ +static void threshold_restart_bank(void *_tr) +{ + struct thresh_restart *tr = _tr; + u32 hi, lo; + + /* sysfs write might race against an offline operation */ + if (!this_cpu_read(threshold_banks) && !tr->set_lvt_off) + return; + + rdmsr(tr->b->address, lo, hi); + + if (tr->b->threshold_limit < (hi & THRESHOLD_MAX)) + tr->reset = 1; /* limit cannot be lower than err count */ + + if (tr->reset) { /* reset err count and overflow bit */ + hi = + (hi & ~(MASK_ERR_COUNT_HI | MASK_OVERFLOW_HI)) | + (THRESHOLD_MAX - tr->b->threshold_limit); + } else if (tr->old_limit) { /* change limit w/o reset */ + int new_count = (hi & THRESHOLD_MAX) + + (tr->old_limit - tr->b->threshold_limit); + + hi = (hi & ~MASK_ERR_COUNT_HI) | + (new_count & THRESHOLD_MAX); + } + + /* clear IntType */ + hi &= ~MASK_INT_TYPE_HI; + + if (!tr->b->interrupt_capable) + goto done; + + if (tr->set_lvt_off) { + if (lvt_off_valid(tr->b, tr->lvt_off, lo, hi)) { + /* set new lvt offset */ + hi &= ~MASK_LVTOFF_HI; + hi |= tr->lvt_off << 20; + } + } + + if (tr->b->interrupt_enable) + hi |= INT_TYPE_APIC; + + done: + + hi |= MASK_COUNT_EN_HI; + wrmsr(tr->b->address, lo, hi); +} + +static void mce_threshold_block_init(struct threshold_block *b, int offset) +{ + struct thresh_restart tr = { + .b = b, + .set_lvt_off = 1, + .lvt_off = offset, + }; + + b->threshold_limit = THRESHOLD_MAX; + threshold_restart_bank(&tr); +}; + +static int setup_APIC_mce_threshold(int reserved, int new) +{ + if (reserved < 0 && !setup_APIC_eilvt(new, THRESHOLD_APIC_VECTOR, + APIC_EILVT_MSG_FIX, 0)) + return new; + + return reserved; +} + +static int setup_APIC_deferred_error(int reserved, int new) +{ + if (reserved < 0 && !setup_APIC_eilvt(new, DEFERRED_ERROR_VECTOR, + APIC_EILVT_MSG_FIX, 0)) + return new; + + return reserved; +} + +static void deferred_error_interrupt_enable(struct cpuinfo_x86 *c) +{ + u32 low = 0, high = 0; + int def_offset = -1, def_new; + + if (rdmsr_safe(MSR_CU_DEF_ERR, &low, &high)) + return; + + def_new = (low & MASK_DEF_LVTOFF) >> 4; + if (!(low & MASK_DEF_LVTOFF)) { + pr_err(FW_BUG "Your BIOS is not setting up LVT offset 0x2 for deferred error IRQs correctly.\n"); + def_new = DEF_LVT_OFF; + low = (low & ~MASK_DEF_LVTOFF) | (DEF_LVT_OFF << 4); + } + + def_offset = setup_APIC_deferred_error(def_offset, def_new); + if ((def_offset == def_new) && + (deferred_error_int_vector != amd_deferred_error_interrupt)) + deferred_error_int_vector = amd_deferred_error_interrupt; + + if (!mce_flags.smca) + low = (low & ~MASK_DEF_INT_TYPE) | DEF_INT_TYPE_APIC; + + wrmsr(MSR_CU_DEF_ERR, low, high); +} + +static u32 smca_get_block_address(unsigned int bank, unsigned int block, + unsigned int cpu) +{ + if (!block) + return MSR_AMD64_SMCA_MCx_MISC(bank); + + if (!(per_cpu(smca_misc_banks_map, cpu) & BIT_ULL(bank))) + return 0; + + return MSR_AMD64_SMCA_MCx_MISCy(bank, block - 1); +} + +static u32 get_block_address(u32 current_addr, u32 low, u32 high, + unsigned int bank, unsigned int block, + unsigned int cpu) +{ + u32 addr = 0, offset = 0; + + if ((bank >= per_cpu(mce_num_banks, cpu)) || (block >= NR_BLOCKS)) + return addr; + + if (mce_flags.smca) + return smca_get_block_address(bank, block, cpu); + + /* Fall back to method we used for older processors: */ + switch (block) { + case 0: + addr = mca_msr_reg(bank, MCA_MISC); + break; + case 1: + offset = ((low & MASK_BLKPTR_LO) >> 21); + if (offset) + addr = MCG_XBLK_ADDR + offset; + break; + default: + addr = ++current_addr; + } + return addr; +} + +static int +prepare_threshold_block(unsigned int bank, unsigned int block, u32 addr, + int offset, u32 misc_high) +{ + unsigned int cpu = smp_processor_id(); + u32 smca_low, smca_high; + struct threshold_block b; + int new; + + if (!block) + per_cpu(bank_map, cpu) |= BIT_ULL(bank); + + memset(&b, 0, sizeof(b)); + b.cpu = cpu; + b.bank = bank; + b.block = block; + b.address = addr; + b.interrupt_capable = lvt_interrupt_supported(bank, misc_high); + + if (!b.interrupt_capable) + goto done; + + b.interrupt_enable = 1; + + if (!mce_flags.smca) { + new = (misc_high & MASK_LVTOFF_HI) >> 20; + goto set_offset; + } + + /* Gather LVT offset for thresholding: */ + if (rdmsr_safe(MSR_CU_DEF_ERR, &smca_low, &smca_high)) + goto out; + + new = (smca_low & SMCA_THR_LVT_OFF) >> 12; + +set_offset: + offset = setup_APIC_mce_threshold(offset, new); + if (offset == new) + thresholding_irq_en = true; + +done: + mce_threshold_block_init(&b, offset); + +out: + return offset; +} + +bool amd_filter_mce(struct mce *m) +{ + enum smca_bank_types bank_type = smca_get_bank_type(m->bank); + struct cpuinfo_x86 *c = &boot_cpu_data; + + /* See Family 17h Models 10h-2Fh Erratum #1114. */ + if (c->x86 == 0x17 && + c->x86_model >= 0x10 && c->x86_model <= 0x2F && + bank_type == SMCA_IF && XEC(m->status, 0x3f) == 10) + return true; + + /* NB GART TLB error reporting is disabled by default. */ + if (c->x86 < 0x17) { + if (m->bank == 4 && XEC(m->status, 0x1f) == 0x5) + return true; + } + + return false; +} + +/* + * Turn off thresholding banks for the following conditions: + * - MC4_MISC thresholding is not supported on Family 0x15. + * - Prevent possible spurious interrupts from the IF bank on Family 0x17 + * Models 0x10-0x2F due to Erratum #1114. + */ +static void disable_err_thresholding(struct cpuinfo_x86 *c, unsigned int bank) +{ + int i, num_msrs; + u64 hwcr; + bool need_toggle; + u32 msrs[NR_BLOCKS]; + + if (c->x86 == 0x15 && bank == 4) { + msrs[0] = 0x00000413; /* MC4_MISC0 */ + msrs[1] = 0xc0000408; /* MC4_MISC1 */ + num_msrs = 2; + } else if (c->x86 == 0x17 && + (c->x86_model >= 0x10 && c->x86_model <= 0x2F)) { + + if (smca_get_bank_type(bank) != SMCA_IF) + return; + + msrs[0] = MSR_AMD64_SMCA_MCx_MISC(bank); + num_msrs = 1; + } else { + return; + } + + rdmsrl(MSR_K7_HWCR, hwcr); + + /* McStatusWrEn has to be set */ + need_toggle = !(hwcr & BIT(18)); + if (need_toggle) + wrmsrl(MSR_K7_HWCR, hwcr | BIT(18)); + + /* Clear CntP bit safely */ + for (i = 0; i < num_msrs; i++) + msr_clear_bit(msrs[i], 62); + + /* restore old settings */ + if (need_toggle) + wrmsrl(MSR_K7_HWCR, hwcr); +} + +/* cpu init entry point, called from mce.c with preempt off */ +void mce_amd_feature_init(struct cpuinfo_x86 *c) +{ + unsigned int bank, block, cpu = smp_processor_id(); + u32 low = 0, high = 0, address = 0; + int offset = -1; + + + for (bank = 0; bank < this_cpu_read(mce_num_banks); ++bank) { + if (mce_flags.smca) + smca_configure(bank, cpu); + + disable_err_thresholding(c, bank); + + for (block = 0; block < NR_BLOCKS; ++block) { + address = get_block_address(address, low, high, bank, block, cpu); + if (!address) + break; + + if (rdmsr_safe(address, &low, &high)) + break; + + if (!(high & MASK_VALID_HI)) + continue; + + if (!(high & MASK_CNTP_HI) || + (high & MASK_LOCKED_HI)) + continue; + + offset = prepare_threshold_block(bank, block, address, offset, high); + } + } + + if (mce_flags.succor) + deferred_error_interrupt_enable(c); +} + +int umc_normaddr_to_sysaddr(u64 norm_addr, u16 nid, u8 umc, u64 *sys_addr) +{ + u64 dram_base_addr, dram_limit_addr, dram_hole_base; + /* We start from the normalized address */ + u64 ret_addr = norm_addr; + + u32 tmp; + + u8 die_id_shift, die_id_mask, socket_id_shift, socket_id_mask; + u8 intlv_num_dies, intlv_num_chan, intlv_num_sockets; + u8 intlv_addr_sel, intlv_addr_bit; + u8 num_intlv_bits, hashed_bit; + u8 lgcy_mmio_hole_en, base = 0; + u8 cs_mask, cs_id = 0; + bool hash_enabled = false; + + /* Read D18F0x1B4 (DramOffset), check if base 1 is used. */ + if (amd_df_indirect_read(nid, 0, 0x1B4, umc, &tmp)) + goto out_err; + + /* Remove HiAddrOffset from normalized address, if enabled: */ + if (tmp & BIT(0)) { + u64 hi_addr_offset = (tmp & GENMASK_ULL(31, 20)) << 8; + + if (norm_addr >= hi_addr_offset) { + ret_addr -= hi_addr_offset; + base = 1; + } + } + + /* Read D18F0x110 (DramBaseAddress). */ + if (amd_df_indirect_read(nid, 0, 0x110 + (8 * base), umc, &tmp)) + goto out_err; + + /* Check if address range is valid. */ + if (!(tmp & BIT(0))) { + pr_err("%s: Invalid DramBaseAddress range: 0x%x.\n", + __func__, tmp); + goto out_err; + } + + lgcy_mmio_hole_en = tmp & BIT(1); + intlv_num_chan = (tmp >> 4) & 0xF; + intlv_addr_sel = (tmp >> 8) & 0x7; + dram_base_addr = (tmp & GENMASK_ULL(31, 12)) << 16; + + /* {0, 1, 2, 3} map to address bits {8, 9, 10, 11} respectively */ + if (intlv_addr_sel > 3) { + pr_err("%s: Invalid interleave address select %d.\n", + __func__, intlv_addr_sel); + goto out_err; + } + + /* Read D18F0x114 (DramLimitAddress). */ + if (amd_df_indirect_read(nid, 0, 0x114 + (8 * base), umc, &tmp)) + goto out_err; + + intlv_num_sockets = (tmp >> 8) & 0x1; + intlv_num_dies = (tmp >> 10) & 0x3; + dram_limit_addr = ((tmp & GENMASK_ULL(31, 12)) << 16) | GENMASK_ULL(27, 0); + + intlv_addr_bit = intlv_addr_sel + 8; + + /* Re-use intlv_num_chan by setting it equal to log2(#channels) */ + switch (intlv_num_chan) { + case 0: intlv_num_chan = 0; break; + case 1: intlv_num_chan = 1; break; + case 3: intlv_num_chan = 2; break; + case 5: intlv_num_chan = 3; break; + case 7: intlv_num_chan = 4; break; + + case 8: intlv_num_chan = 1; + hash_enabled = true; + break; + default: + pr_err("%s: Invalid number of interleaved channels %d.\n", + __func__, intlv_num_chan); + goto out_err; + } + + num_intlv_bits = intlv_num_chan; + + if (intlv_num_dies > 2) { + pr_err("%s: Invalid number of interleaved nodes/dies %d.\n", + __func__, intlv_num_dies); + goto out_err; + } + + num_intlv_bits += intlv_num_dies; + + /* Add a bit if sockets are interleaved. */ + num_intlv_bits += intlv_num_sockets; + + /* Assert num_intlv_bits <= 4 */ + if (num_intlv_bits > 4) { + pr_err("%s: Invalid interleave bits %d.\n", + __func__, num_intlv_bits); + goto out_err; + } + + if (num_intlv_bits > 0) { + u64 temp_addr_x, temp_addr_i, temp_addr_y; + u8 die_id_bit, sock_id_bit, cs_fabric_id; + + /* + * Read FabricBlockInstanceInformation3_CS[BlockFabricID]. + * This is the fabric id for this coherent slave. Use + * umc/channel# as instance id of the coherent slave + * for FICAA. + */ + if (amd_df_indirect_read(nid, 0, 0x50, umc, &tmp)) + goto out_err; + + cs_fabric_id = (tmp >> 8) & 0xFF; + die_id_bit = 0; + + /* If interleaved over more than 1 channel: */ + if (intlv_num_chan) { + die_id_bit = intlv_num_chan; + cs_mask = (1 << die_id_bit) - 1; + cs_id = cs_fabric_id & cs_mask; + } + + sock_id_bit = die_id_bit; + + /* Read D18F1x208 (SystemFabricIdMask). */ + if (intlv_num_dies || intlv_num_sockets) + if (amd_df_indirect_read(nid, 1, 0x208, umc, &tmp)) + goto out_err; + + /* If interleaved over more than 1 die. */ + if (intlv_num_dies) { + sock_id_bit = die_id_bit + intlv_num_dies; + die_id_shift = (tmp >> 24) & 0xF; + die_id_mask = (tmp >> 8) & 0xFF; + + cs_id |= ((cs_fabric_id & die_id_mask) >> die_id_shift) << die_id_bit; + } + + /* If interleaved over more than 1 socket. */ + if (intlv_num_sockets) { + socket_id_shift = (tmp >> 28) & 0xF; + socket_id_mask = (tmp >> 16) & 0xFF; + + cs_id |= ((cs_fabric_id & socket_id_mask) >> socket_id_shift) << sock_id_bit; + } + + /* + * The pre-interleaved address consists of XXXXXXIIIYYYYY + * where III is the ID for this CS, and XXXXXXYYYYY are the + * address bits from the post-interleaved address. + * "num_intlv_bits" has been calculated to tell us how many "I" + * bits there are. "intlv_addr_bit" tells us how many "Y" bits + * there are (where "I" starts). + */ + temp_addr_y = ret_addr & GENMASK_ULL(intlv_addr_bit-1, 0); + temp_addr_i = (cs_id << intlv_addr_bit); + temp_addr_x = (ret_addr & GENMASK_ULL(63, intlv_addr_bit)) << num_intlv_bits; + ret_addr = temp_addr_x | temp_addr_i | temp_addr_y; + } + + /* Add dram base address */ + ret_addr += dram_base_addr; + + /* If legacy MMIO hole enabled */ + if (lgcy_mmio_hole_en) { + if (amd_df_indirect_read(nid, 0, 0x104, umc, &tmp)) + goto out_err; + + dram_hole_base = tmp & GENMASK(31, 24); + if (ret_addr >= dram_hole_base) + ret_addr += (BIT_ULL(32) - dram_hole_base); + } + + if (hash_enabled) { + /* Save some parentheses and grab ls-bit at the end. */ + hashed_bit = (ret_addr >> 12) ^ + (ret_addr >> 18) ^ + (ret_addr >> 21) ^ + (ret_addr >> 30) ^ + cs_id; + + hashed_bit &= BIT(0); + + if (hashed_bit != ((ret_addr >> intlv_addr_bit) & BIT(0))) + ret_addr ^= BIT(intlv_addr_bit); + } + + /* Is calculated system address is above DRAM limit address? */ + if (ret_addr > dram_limit_addr) + goto out_err; + + *sys_addr = ret_addr; + return 0; + +out_err: + return -EINVAL; +} +EXPORT_SYMBOL_GPL(umc_normaddr_to_sysaddr); + +bool amd_mce_is_memory_error(struct mce *m) +{ + /* ErrCodeExt[20:16] */ + u8 xec = (m->status >> 16) & 0x1f; + + if (mce_flags.smca) + return smca_get_bank_type(m->bank) == SMCA_UMC && xec == 0x0; + + return m->bank == 4 && xec == 0x8; +} + +static void __log_error(unsigned int bank, u64 status, u64 addr, u64 misc) +{ + struct mce m; + + mce_setup(&m); + + m.status = status; + m.misc = misc; + m.bank = bank; + m.tsc = rdtsc(); + + if (m.status & MCI_STATUS_ADDRV) { + m.addr = addr; + + /* + * Extract [55:<lsb>] where lsb is the least significant + * *valid* bit of the address bits. + */ + if (mce_flags.smca) { + u8 lsb = (m.addr >> 56) & 0x3f; + + m.addr &= GENMASK_ULL(55, lsb); + } + } + + if (mce_flags.smca) { + rdmsrl(MSR_AMD64_SMCA_MCx_IPID(bank), m.ipid); + + if (m.status & MCI_STATUS_SYNDV) + rdmsrl(MSR_AMD64_SMCA_MCx_SYND(bank), m.synd); + } + + mce_log(&m); +} + +DEFINE_IDTENTRY_SYSVEC(sysvec_deferred_error) +{ + trace_deferred_error_apic_entry(DEFERRED_ERROR_VECTOR); + inc_irq_stat(irq_deferred_error_count); + deferred_error_int_vector(); + trace_deferred_error_apic_exit(DEFERRED_ERROR_VECTOR); + ack_APIC_irq(); +} + +/* + * Returns true if the logged error is deferred. False, otherwise. + */ +static inline bool +_log_error_bank(unsigned int bank, u32 msr_stat, u32 msr_addr, u64 misc) +{ + u64 status, addr = 0; + + rdmsrl(msr_stat, status); + if (!(status & MCI_STATUS_VAL)) + return false; + + if (status & MCI_STATUS_ADDRV) + rdmsrl(msr_addr, addr); + + __log_error(bank, status, addr, misc); + + wrmsrl(msr_stat, 0); + + return status & MCI_STATUS_DEFERRED; +} + +static bool _log_error_deferred(unsigned int bank, u32 misc) +{ + if (!_log_error_bank(bank, mca_msr_reg(bank, MCA_STATUS), + mca_msr_reg(bank, MCA_ADDR), misc)) + return false; + + /* + * Non-SMCA systems don't have MCA_DESTAT/MCA_DEADDR registers. + * Return true here to avoid accessing these registers. + */ + if (!mce_flags.smca) + return true; + + /* Clear MCA_DESTAT if the deferred error was logged from MCA_STATUS. */ + wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(bank), 0); + return true; +} + +/* + * We have three scenarios for checking for Deferred errors: + * + * 1) Non-SMCA systems check MCA_STATUS and log error if found. + * 2) SMCA systems check MCA_STATUS. If error is found then log it and also + * clear MCA_DESTAT. + * 3) SMCA systems check MCA_DESTAT, if error was not found in MCA_STATUS, and + * log it. + */ +static void log_error_deferred(unsigned int bank) +{ + if (_log_error_deferred(bank, 0)) + return; + + /* + * Only deferred errors are logged in MCA_DE{STAT,ADDR} so just check + * for a valid error. + */ + _log_error_bank(bank, MSR_AMD64_SMCA_MCx_DESTAT(bank), + MSR_AMD64_SMCA_MCx_DEADDR(bank), 0); +} + +/* APIC interrupt handler for deferred errors */ +static void amd_deferred_error_interrupt(void) +{ + unsigned int bank; + + for (bank = 0; bank < this_cpu_read(mce_num_banks); ++bank) + log_error_deferred(bank); +} + +static void log_error_thresholding(unsigned int bank, u64 misc) +{ + _log_error_deferred(bank, misc); +} + +static void log_and_reset_block(struct threshold_block *block) +{ + struct thresh_restart tr; + u32 low = 0, high = 0; + + if (!block) + return; + + if (rdmsr_safe(block->address, &low, &high)) + return; + + if (!(high & MASK_OVERFLOW_HI)) + return; + + /* Log the MCE which caused the threshold event. */ + log_error_thresholding(block->bank, ((u64)high << 32) | low); + + /* Reset threshold block after logging error. */ + memset(&tr, 0, sizeof(tr)); + tr.b = block; + threshold_restart_bank(&tr); +} + +/* + * Threshold interrupt handler will service THRESHOLD_APIC_VECTOR. The interrupt + * goes off when error_count reaches threshold_limit. + */ +static void amd_threshold_interrupt(void) +{ + struct threshold_block *first_block = NULL, *block = NULL, *tmp = NULL; + struct threshold_bank **bp = this_cpu_read(threshold_banks); + unsigned int bank, cpu = smp_processor_id(); + + /* + * Validate that the threshold bank has been initialized already. The + * handler is installed at boot time, but on a hotplug event the + * interrupt might fire before the data has been initialized. + */ + if (!bp) + return; + + for (bank = 0; bank < this_cpu_read(mce_num_banks); ++bank) { + if (!(per_cpu(bank_map, cpu) & BIT_ULL(bank))) + continue; + + first_block = bp[bank]->blocks; + if (!first_block) + continue; + + /* + * The first block is also the head of the list. Check it first + * before iterating over the rest. + */ + log_and_reset_block(first_block); + list_for_each_entry_safe(block, tmp, &first_block->miscj, miscj) + log_and_reset_block(block); + } +} + +/* + * Sysfs Interface + */ + +struct threshold_attr { + struct attribute attr; + ssize_t (*show) (struct threshold_block *, char *); + ssize_t (*store) (struct threshold_block *, const char *, size_t count); +}; + +#define SHOW_FIELDS(name) \ +static ssize_t show_ ## name(struct threshold_block *b, char *buf) \ +{ \ + return sprintf(buf, "%lu\n", (unsigned long) b->name); \ +} +SHOW_FIELDS(interrupt_enable) +SHOW_FIELDS(threshold_limit) + +static ssize_t +store_interrupt_enable(struct threshold_block *b, const char *buf, size_t size) +{ + struct thresh_restart tr; + unsigned long new; + + if (!b->interrupt_capable) + return -EINVAL; + + if (kstrtoul(buf, 0, &new) < 0) + return -EINVAL; + + b->interrupt_enable = !!new; + + memset(&tr, 0, sizeof(tr)); + tr.b = b; + + if (smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1)) + return -ENODEV; + + return size; +} + +static ssize_t +store_threshold_limit(struct threshold_block *b, const char *buf, size_t size) +{ + struct thresh_restart tr; + unsigned long new; + + if (kstrtoul(buf, 0, &new) < 0) + return -EINVAL; + + if (new > THRESHOLD_MAX) + new = THRESHOLD_MAX; + if (new < 1) + new = 1; + + memset(&tr, 0, sizeof(tr)); + tr.old_limit = b->threshold_limit; + b->threshold_limit = new; + tr.b = b; + + if (smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1)) + return -ENODEV; + + return size; +} + +static ssize_t show_error_count(struct threshold_block *b, char *buf) +{ + u32 lo, hi; + + /* CPU might be offline by now */ + if (rdmsr_on_cpu(b->cpu, b->address, &lo, &hi)) + return -ENODEV; + + return sprintf(buf, "%u\n", ((hi & THRESHOLD_MAX) - + (THRESHOLD_MAX - b->threshold_limit))); +} + +static struct threshold_attr error_count = { + .attr = {.name = __stringify(error_count), .mode = 0444 }, + .show = show_error_count, +}; + +#define RW_ATTR(val) \ +static struct threshold_attr val = { \ + .attr = {.name = __stringify(val), .mode = 0644 }, \ + .show = show_## val, \ + .store = store_## val, \ +}; + +RW_ATTR(interrupt_enable); +RW_ATTR(threshold_limit); + +static struct attribute *default_attrs[] = { + &threshold_limit.attr, + &error_count.attr, + NULL, /* possibly interrupt_enable if supported, see below */ + NULL, +}; + +#define to_block(k) container_of(k, struct threshold_block, kobj) +#define to_attr(a) container_of(a, struct threshold_attr, attr) + +static ssize_t show(struct kobject *kobj, struct attribute *attr, char *buf) +{ + struct threshold_block *b = to_block(kobj); + struct threshold_attr *a = to_attr(attr); + ssize_t ret; + + ret = a->show ? a->show(b, buf) : -EIO; + + return ret; +} + +static ssize_t store(struct kobject *kobj, struct attribute *attr, + const char *buf, size_t count) +{ + struct threshold_block *b = to_block(kobj); + struct threshold_attr *a = to_attr(attr); + ssize_t ret; + + ret = a->store ? a->store(b, buf, count) : -EIO; + + return ret; +} + +static const struct sysfs_ops threshold_ops = { + .show = show, + .store = store, +}; + +static void threshold_block_release(struct kobject *kobj); + +static struct kobj_type threshold_ktype = { + .sysfs_ops = &threshold_ops, + .default_attrs = default_attrs, + .release = threshold_block_release, +}; + +static const char *get_name(unsigned int bank, struct threshold_block *b) +{ + enum smca_bank_types bank_type; + + if (!mce_flags.smca) { + if (b && bank == 4) + return bank4_names(b); + + return th_names[bank]; + } + + bank_type = smca_get_bank_type(bank); + if (bank_type >= N_SMCA_BANK_TYPES) + return NULL; + + if (b && bank_type == SMCA_UMC) { + if (b->block < ARRAY_SIZE(smca_umc_block_names)) + return smca_umc_block_names[b->block]; + return NULL; + } + + if (smca_banks[bank].hwid->count == 1) + return smca_get_name(bank_type); + + snprintf(buf_mcatype, MAX_MCATYPE_NAME_LEN, + "%s_%x", smca_get_name(bank_type), + smca_banks[bank].sysfs_id); + return buf_mcatype; +} + +static int allocate_threshold_blocks(unsigned int cpu, struct threshold_bank *tb, + unsigned int bank, unsigned int block, + u32 address) +{ + struct threshold_block *b = NULL; + u32 low, high; + int err; + + if ((bank >= this_cpu_read(mce_num_banks)) || (block >= NR_BLOCKS)) + return 0; + + if (rdmsr_safe(address, &low, &high)) + return 0; + + if (!(high & MASK_VALID_HI)) { + if (block) + goto recurse; + else + return 0; + } + + if (!(high & MASK_CNTP_HI) || + (high & MASK_LOCKED_HI)) + goto recurse; + + b = kzalloc(sizeof(struct threshold_block), GFP_KERNEL); + if (!b) + return -ENOMEM; + + b->block = block; + b->bank = bank; + b->cpu = cpu; + b->address = address; + b->interrupt_enable = 0; + b->interrupt_capable = lvt_interrupt_supported(bank, high); + b->threshold_limit = THRESHOLD_MAX; + + if (b->interrupt_capable) { + threshold_ktype.default_attrs[2] = &interrupt_enable.attr; + b->interrupt_enable = 1; + } else { + threshold_ktype.default_attrs[2] = NULL; + } + + INIT_LIST_HEAD(&b->miscj); + + /* This is safe as @tb is not visible yet */ + if (tb->blocks) + list_add(&b->miscj, &tb->blocks->miscj); + else + tb->blocks = b; + + err = kobject_init_and_add(&b->kobj, &threshold_ktype, tb->kobj, get_name(bank, b)); + if (err) + goto out_free; +recurse: + address = get_block_address(address, low, high, bank, ++block, cpu); + if (!address) + return 0; + + err = allocate_threshold_blocks(cpu, tb, bank, block, address); + if (err) + goto out_free; + + if (b) + kobject_uevent(&b->kobj, KOBJ_ADD); + + return 0; + +out_free: + if (b) { + list_del(&b->miscj); + kobject_put(&b->kobj); + } + return err; +} + +static int __threshold_add_blocks(struct threshold_bank *b) +{ + struct list_head *head = &b->blocks->miscj; + struct threshold_block *pos = NULL; + struct threshold_block *tmp = NULL; + int err = 0; + + err = kobject_add(&b->blocks->kobj, b->kobj, b->blocks->kobj.name); + if (err) + return err; + + list_for_each_entry_safe(pos, tmp, head, miscj) { + + err = kobject_add(&pos->kobj, b->kobj, pos->kobj.name); + if (err) { + list_for_each_entry_safe_reverse(pos, tmp, head, miscj) + kobject_del(&pos->kobj); + + return err; + } + } + return err; +} + +static int threshold_create_bank(struct threshold_bank **bp, unsigned int cpu, + unsigned int bank) +{ + struct device *dev = this_cpu_read(mce_device); + struct amd_northbridge *nb = NULL; + struct threshold_bank *b = NULL; + const char *name = get_name(bank, NULL); + int err = 0; + + if (!dev) + return -ENODEV; + + if (is_shared_bank(bank)) { + nb = node_to_amd_nb(amd_get_nb_id(cpu)); + + /* threshold descriptor already initialized on this node? */ + if (nb && nb->bank4) { + /* yes, use it */ + b = nb->bank4; + err = kobject_add(b->kobj, &dev->kobj, name); + if (err) + goto out; + + bp[bank] = b; + refcount_inc(&b->cpus); + + err = __threshold_add_blocks(b); + + goto out; + } + } + + b = kzalloc(sizeof(struct threshold_bank), GFP_KERNEL); + if (!b) { + err = -ENOMEM; + goto out; + } + + /* Associate the bank with the per-CPU MCE device */ + b->kobj = kobject_create_and_add(name, &dev->kobj); + if (!b->kobj) { + err = -EINVAL; + goto out_free; + } + + if (is_shared_bank(bank)) { + b->shared = 1; + refcount_set(&b->cpus, 1); + + /* nb is already initialized, see above */ + if (nb) { + WARN_ON(nb->bank4); + nb->bank4 = b; + } + } + + err = allocate_threshold_blocks(cpu, b, bank, 0, mca_msr_reg(bank, MCA_MISC)); + if (err) + goto out_kobj; + + bp[bank] = b; + return 0; + +out_kobj: + kobject_put(b->kobj); +out_free: + kfree(b); +out: + return err; +} + +static void threshold_block_release(struct kobject *kobj) +{ + kfree(to_block(kobj)); +} + +static void deallocate_threshold_blocks(struct threshold_bank *bank) +{ + struct threshold_block *pos, *tmp; + + list_for_each_entry_safe(pos, tmp, &bank->blocks->miscj, miscj) { + list_del(&pos->miscj); + kobject_put(&pos->kobj); + } + + kobject_put(&bank->blocks->kobj); +} + +static void __threshold_remove_blocks(struct threshold_bank *b) +{ + struct threshold_block *pos = NULL; + struct threshold_block *tmp = NULL; + + kobject_del(b->kobj); + + list_for_each_entry_safe(pos, tmp, &b->blocks->miscj, miscj) + kobject_del(&pos->kobj); +} + +static void threshold_remove_bank(struct threshold_bank *bank) +{ + struct amd_northbridge *nb; + + if (!bank->blocks) + goto out_free; + + if (!bank->shared) + goto out_dealloc; + + if (!refcount_dec_and_test(&bank->cpus)) { + __threshold_remove_blocks(bank); + return; + } else { + /* + * The last CPU on this node using the shared bank is going + * away, remove that bank now. + */ + nb = node_to_amd_nb(amd_get_nb_id(smp_processor_id())); + nb->bank4 = NULL; + } + +out_dealloc: + deallocate_threshold_blocks(bank); + +out_free: + kobject_put(bank->kobj); + kfree(bank); +} + +static void __threshold_remove_device(struct threshold_bank **bp) +{ + unsigned int bank, numbanks = this_cpu_read(mce_num_banks); + + for (bank = 0; bank < numbanks; bank++) { + if (!bp[bank]) + continue; + + threshold_remove_bank(bp[bank]); + bp[bank] = NULL; + } + kfree(bp); +} + +int mce_threshold_remove_device(unsigned int cpu) +{ + struct threshold_bank **bp = this_cpu_read(threshold_banks); + + if (!bp) + return 0; + + /* + * Clear the pointer before cleaning up, so that the interrupt won't + * touch anything of this. + */ + this_cpu_write(threshold_banks, NULL); + + __threshold_remove_device(bp); + return 0; +} + +/** + * mce_threshold_create_device - Create the per-CPU MCE threshold device + * @cpu: The plugged in CPU + * + * Create directories and files for all valid threshold banks. + * + * This is invoked from the CPU hotplug callback which was installed in + * mcheck_init_device(). The invocation happens in context of the hotplug + * thread running on @cpu. The callback is invoked on all CPUs which are + * online when the callback is installed or during a real hotplug event. + */ +int mce_threshold_create_device(unsigned int cpu) +{ + unsigned int numbanks, bank; + struct threshold_bank **bp; + int err; + + if (!mce_flags.amd_threshold) + return 0; + + bp = this_cpu_read(threshold_banks); + if (bp) + return 0; + + numbanks = this_cpu_read(mce_num_banks); + bp = kcalloc(numbanks, sizeof(*bp), GFP_KERNEL); + if (!bp) + return -ENOMEM; + + for (bank = 0; bank < numbanks; ++bank) { + if (!(this_cpu_read(bank_map) & BIT_ULL(bank))) + continue; + err = threshold_create_bank(bp, cpu, bank); + if (err) { + __threshold_remove_device(bp); + return err; + } + } + this_cpu_write(threshold_banks, bp); + + if (thresholding_irq_en) + mce_threshold_vector = amd_threshold_interrupt; + return 0; +} diff --git a/arch/x86/kernel/cpu/mce/apei.c b/arch/x86/kernel/cpu/mce/apei.c new file mode 100644 index 000000000..af8d37962 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/apei.c @@ -0,0 +1,145 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Bridge between MCE and APEI + * + * On some machine, corrected memory errors are reported via APEI + * generic hardware error source (GHES) instead of corrected Machine + * Check. These corrected memory errors can be reported to user space + * through /dev/mcelog via faking a corrected Machine Check, so that + * the error memory page can be offlined by /sbin/mcelog if the error + * count for one page is beyond the threshold. + * + * For fatal MCE, save MCE record into persistent storage via ERST, so + * that the MCE record can be logged after reboot via ERST. + * + * Copyright 2010 Intel Corp. + * Author: Huang Ying <ying.huang@intel.com> + */ + +#include <linux/export.h> +#include <linux/kernel.h> +#include <linux/acpi.h> +#include <linux/cper.h> +#include <acpi/apei.h> +#include <acpi/ghes.h> +#include <asm/mce.h> + +#include "internal.h" + +void apei_mce_report_mem_error(int severity, struct cper_sec_mem_err *mem_err) +{ + struct mce m; + + if (!(mem_err->validation_bits & CPER_MEM_VALID_PA)) + return; + + mce_setup(&m); + m.bank = -1; + /* Fake a memory read error with unknown channel */ + m.status = MCI_STATUS_VAL | MCI_STATUS_EN | MCI_STATUS_ADDRV | 0x9f; + + if (severity >= GHES_SEV_RECOVERABLE) + m.status |= MCI_STATUS_UC; + + if (severity >= GHES_SEV_PANIC) { + m.status |= MCI_STATUS_PCC; + m.tsc = rdtsc(); + } + + m.addr = mem_err->physical_addr; + mce_log(&m); +} +EXPORT_SYMBOL_GPL(apei_mce_report_mem_error); + +#define CPER_CREATOR_MCE \ + GUID_INIT(0x75a574e3, 0x5052, 0x4b29, 0x8a, 0x8e, 0xbe, 0x2c, \ + 0x64, 0x90, 0xb8, 0x9d) +#define CPER_SECTION_TYPE_MCE \ + GUID_INIT(0xfe08ffbe, 0x95e4, 0x4be7, 0xbc, 0x73, 0x40, 0x96, \ + 0x04, 0x4a, 0x38, 0xfc) + +/* + * CPER specification (in UEFI specification 2.3 appendix N) requires + * byte-packed. + */ +struct cper_mce_record { + struct cper_record_header hdr; + struct cper_section_descriptor sec_hdr; + struct mce mce; +} __packed; + +int apei_write_mce(struct mce *m) +{ + struct cper_mce_record rcd; + + memset(&rcd, 0, sizeof(rcd)); + memcpy(rcd.hdr.signature, CPER_SIG_RECORD, CPER_SIG_SIZE); + rcd.hdr.revision = CPER_RECORD_REV; + rcd.hdr.signature_end = CPER_SIG_END; + rcd.hdr.section_count = 1; + rcd.hdr.error_severity = CPER_SEV_FATAL; + /* timestamp, platform_id, partition_id are all invalid */ + rcd.hdr.validation_bits = 0; + rcd.hdr.record_length = sizeof(rcd); + rcd.hdr.creator_id = CPER_CREATOR_MCE; + rcd.hdr.notification_type = CPER_NOTIFY_MCE; + rcd.hdr.record_id = cper_next_record_id(); + rcd.hdr.flags = CPER_HW_ERROR_FLAGS_PREVERR; + + rcd.sec_hdr.section_offset = (void *)&rcd.mce - (void *)&rcd; + rcd.sec_hdr.section_length = sizeof(rcd.mce); + rcd.sec_hdr.revision = CPER_SEC_REV; + /* fru_id and fru_text is invalid */ + rcd.sec_hdr.validation_bits = 0; + rcd.sec_hdr.flags = CPER_SEC_PRIMARY; + rcd.sec_hdr.section_type = CPER_SECTION_TYPE_MCE; + rcd.sec_hdr.section_severity = CPER_SEV_FATAL; + + memcpy(&rcd.mce, m, sizeof(*m)); + + return erst_write(&rcd.hdr); +} + +ssize_t apei_read_mce(struct mce *m, u64 *record_id) +{ + struct cper_mce_record rcd; + int rc, pos; + + rc = erst_get_record_id_begin(&pos); + if (rc) + return rc; +retry: + rc = erst_get_record_id_next(&pos, record_id); + if (rc) + goto out; + /* no more record */ + if (*record_id == APEI_ERST_INVALID_RECORD_ID) + goto out; + rc = erst_read(*record_id, &rcd.hdr, sizeof(rcd)); + /* someone else has cleared the record, try next one */ + if (rc == -ENOENT) + goto retry; + else if (rc < 0) + goto out; + /* try to skip other type records in storage */ + else if (rc != sizeof(rcd) || + !guid_equal(&rcd.hdr.creator_id, &CPER_CREATOR_MCE)) + goto retry; + memcpy(m, &rcd.mce, sizeof(*m)); + rc = sizeof(*m); +out: + erst_get_record_id_end(); + + return rc; +} + +/* Check whether there is record in ERST */ +int apei_check_mce(void) +{ + return erst_get_record_count(); +} + +int apei_clear_mce(u64 record_id) +{ + return erst_clear(record_id); +} diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c new file mode 100644 index 000000000..0b7c81389 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/core.c @@ -0,0 +1,2794 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Machine check handler. + * + * K8 parts Copyright 2002,2003 Andi Kleen, SuSE Labs. + * Rest from unknown author(s). + * 2004 Andi Kleen. Rewrote most of it. + * Copyright 2008 Intel Corporation + * Author: Andi Kleen + */ + +#include <linux/thread_info.h> +#include <linux/capability.h> +#include <linux/miscdevice.h> +#include <linux/ratelimit.h> +#include <linux/rcupdate.h> +#include <linux/kobject.h> +#include <linux/uaccess.h> +#include <linux/kdebug.h> +#include <linux/kernel.h> +#include <linux/percpu.h> +#include <linux/string.h> +#include <linux/device.h> +#include <linux/syscore_ops.h> +#include <linux/delay.h> +#include <linux/ctype.h> +#include <linux/sched.h> +#include <linux/sysfs.h> +#include <linux/types.h> +#include <linux/slab.h> +#include <linux/init.h> +#include <linux/kmod.h> +#include <linux/poll.h> +#include <linux/nmi.h> +#include <linux/cpu.h> +#include <linux/ras.h> +#include <linux/smp.h> +#include <linux/fs.h> +#include <linux/mm.h> +#include <linux/debugfs.h> +#include <linux/irq_work.h> +#include <linux/export.h> +#include <linux/set_memory.h> +#include <linux/sync_core.h> +#include <linux/task_work.h> +#include <linux/hardirq.h> + +#include <asm/intel-family.h> +#include <asm/processor.h> +#include <asm/traps.h> +#include <asm/tlbflush.h> +#include <asm/mce.h> +#include <asm/msr.h> +#include <asm/reboot.h> + +#include "internal.h" + +/* sysfs synchronization */ +static DEFINE_MUTEX(mce_sysfs_mutex); + +#define CREATE_TRACE_POINTS +#include <trace/events/mce.h> + +#define SPINUNIT 100 /* 100ns */ + +DEFINE_PER_CPU(unsigned, mce_exception_count); + +DEFINE_PER_CPU_READ_MOSTLY(unsigned int, mce_num_banks); + +struct mce_bank { + u64 ctl; /* subevents to enable */ + bool init; /* initialise bank? */ +}; +static DEFINE_PER_CPU_READ_MOSTLY(struct mce_bank[MAX_NR_BANKS], mce_banks_array); + +#define ATTR_LEN 16 +/* One object for each MCE bank, shared by all CPUs */ +struct mce_bank_dev { + struct device_attribute attr; /* device attribute */ + char attrname[ATTR_LEN]; /* attribute name */ + u8 bank; /* bank number */ +}; +static struct mce_bank_dev mce_bank_devs[MAX_NR_BANKS]; + +struct mce_vendor_flags mce_flags __read_mostly; + +struct mca_config mca_cfg __read_mostly = { + .bootlog = -1, + /* + * Tolerant levels: + * 0: always panic on uncorrected errors, log corrected errors + * 1: panic or SIGBUS on uncorrected errors, log corrected errors + * 2: SIGBUS or log uncorrected errors (if possible), log corr. errors + * 3: never panic or SIGBUS, log all errors (for testing only) + */ + .tolerant = 1, + .monarch_timeout = -1 +}; + +static DEFINE_PER_CPU(struct mce, mces_seen); +static unsigned long mce_need_notify; +static int cpu_missing; + +/* + * MCA banks polled by the period polling timer for corrected events. + * With Intel CMCI, this only has MCA banks which do not support CMCI (if any). + */ +DEFINE_PER_CPU(mce_banks_t, mce_poll_banks) = { + [0 ... BITS_TO_LONGS(MAX_NR_BANKS)-1] = ~0UL +}; + +/* + * MCA banks controlled through firmware first for corrected errors. + * This is a global list of banks for which we won't enable CMCI and we + * won't poll. Firmware controls these banks and is responsible for + * reporting corrected errors through GHES. Uncorrected/recoverable + * errors are still notified through a machine check. + */ +mce_banks_t mce_banks_ce_disabled; + +static struct work_struct mce_work; +static struct irq_work mce_irq_work; + +static void (*quirk_no_way_out)(int bank, struct mce *m, struct pt_regs *regs); + +/* + * CPU/chipset specific EDAC code can register a notifier call here to print + * MCE errors in a human-readable form. + */ +BLOCKING_NOTIFIER_HEAD(x86_mce_decoder_chain); + +/* Do initial initialization of a struct mce */ +noinstr void mce_setup(struct mce *m) +{ + memset(m, 0, sizeof(struct mce)); + m->cpu = m->extcpu = smp_processor_id(); + /* need the internal __ version to avoid deadlocks */ + m->time = __ktime_get_real_seconds(); + m->cpuvendor = boot_cpu_data.x86_vendor; + m->cpuid = cpuid_eax(1); + m->socketid = cpu_data(m->extcpu).phys_proc_id; + m->apicid = cpu_data(m->extcpu).initial_apicid; + m->mcgcap = __rdmsr(MSR_IA32_MCG_CAP); + + if (this_cpu_has(X86_FEATURE_INTEL_PPIN)) + m->ppin = __rdmsr(MSR_PPIN); + else if (this_cpu_has(X86_FEATURE_AMD_PPIN)) + m->ppin = __rdmsr(MSR_AMD_PPIN); + + m->microcode = boot_cpu_data.microcode; +} + +DEFINE_PER_CPU(struct mce, injectm); +EXPORT_PER_CPU_SYMBOL_GPL(injectm); + +void mce_log(struct mce *m) +{ + if (!mce_gen_pool_add(m)) + irq_work_queue(&mce_irq_work); +} +EXPORT_SYMBOL_GPL(mce_log); + +void mce_register_decode_chain(struct notifier_block *nb) +{ + if (WARN_ON(nb->priority < MCE_PRIO_LOWEST || + nb->priority > MCE_PRIO_HIGHEST)) + return; + + blocking_notifier_chain_register(&x86_mce_decoder_chain, nb); +} +EXPORT_SYMBOL_GPL(mce_register_decode_chain); + +void mce_unregister_decode_chain(struct notifier_block *nb) +{ + blocking_notifier_chain_unregister(&x86_mce_decoder_chain, nb); +} +EXPORT_SYMBOL_GPL(mce_unregister_decode_chain); + +u32 mca_msr_reg(int bank, enum mca_msr reg) +{ + if (mce_flags.smca) { + switch (reg) { + case MCA_CTL: return MSR_AMD64_SMCA_MCx_CTL(bank); + case MCA_ADDR: return MSR_AMD64_SMCA_MCx_ADDR(bank); + case MCA_MISC: return MSR_AMD64_SMCA_MCx_MISC(bank); + case MCA_STATUS: return MSR_AMD64_SMCA_MCx_STATUS(bank); + } + } + + switch (reg) { + case MCA_CTL: return MSR_IA32_MCx_CTL(bank); + case MCA_ADDR: return MSR_IA32_MCx_ADDR(bank); + case MCA_MISC: return MSR_IA32_MCx_MISC(bank); + case MCA_STATUS: return MSR_IA32_MCx_STATUS(bank); + } + + return 0; +} + +static void __print_mce(struct mce *m) +{ + pr_emerg(HW_ERR "CPU %d: Machine Check%s: %Lx Bank %d: %016Lx\n", + m->extcpu, + (m->mcgstatus & MCG_STATUS_MCIP ? " Exception" : ""), + m->mcgstatus, m->bank, m->status); + + if (m->ip) { + pr_emerg(HW_ERR "RIP%s %02x:<%016Lx> ", + !(m->mcgstatus & MCG_STATUS_EIPV) ? " !INEXACT!" : "", + m->cs, m->ip); + + if (m->cs == __KERNEL_CS) + pr_cont("{%pS}", (void *)(unsigned long)m->ip); + pr_cont("\n"); + } + + pr_emerg(HW_ERR "TSC %llx ", m->tsc); + if (m->addr) + pr_cont("ADDR %llx ", m->addr); + if (m->misc) + pr_cont("MISC %llx ", m->misc); + if (m->ppin) + pr_cont("PPIN %llx ", m->ppin); + + if (mce_flags.smca) { + if (m->synd) + pr_cont("SYND %llx ", m->synd); + if (m->ipid) + pr_cont("IPID %llx ", m->ipid); + } + + pr_cont("\n"); + + /* + * Note this output is parsed by external tools and old fields + * should not be changed. + */ + pr_emerg(HW_ERR "PROCESSOR %u:%x TIME %llu SOCKET %u APIC %x microcode %x\n", + m->cpuvendor, m->cpuid, m->time, m->socketid, m->apicid, + m->microcode); +} + +static void print_mce(struct mce *m) +{ + __print_mce(m); + + if (m->cpuvendor != X86_VENDOR_AMD && m->cpuvendor != X86_VENDOR_HYGON) + pr_emerg_ratelimited(HW_ERR "Run the above through 'mcelog --ascii'\n"); +} + +#define PANIC_TIMEOUT 5 /* 5 seconds */ + +static atomic_t mce_panicked; + +static int fake_panic; +static atomic_t mce_fake_panicked; + +/* Panic in progress. Enable interrupts and wait for final IPI */ +static void wait_for_panic(void) +{ + long timeout = PANIC_TIMEOUT*USEC_PER_SEC; + + preempt_disable(); + local_irq_enable(); + while (timeout-- > 0) + udelay(1); + if (panic_timeout == 0) + panic_timeout = mca_cfg.panic_timeout; + panic("Panicing machine check CPU died"); +} + +static noinstr void mce_panic(const char *msg, struct mce *final, char *exp) +{ + struct llist_node *pending; + struct mce_evt_llist *l; + int apei_err = 0; + + /* + * Allow instrumentation around external facilities usage. Not that it + * matters a whole lot since the machine is going to panic anyway. + */ + instrumentation_begin(); + + if (!fake_panic) { + /* + * Make sure only one CPU runs in machine check panic + */ + if (atomic_inc_return(&mce_panicked) > 1) + wait_for_panic(); + barrier(); + + bust_spinlocks(1); + console_verbose(); + } else { + /* Don't log too much for fake panic */ + if (atomic_inc_return(&mce_fake_panicked) > 1) + goto out; + } + pending = mce_gen_pool_prepare_records(); + /* First print corrected ones that are still unlogged */ + llist_for_each_entry(l, pending, llnode) { + struct mce *m = &l->mce; + if (!(m->status & MCI_STATUS_UC)) { + print_mce(m); + if (!apei_err) + apei_err = apei_write_mce(m); + } + } + /* Now print uncorrected but with the final one last */ + llist_for_each_entry(l, pending, llnode) { + struct mce *m = &l->mce; + if (!(m->status & MCI_STATUS_UC)) + continue; + if (!final || mce_cmp(m, final)) { + print_mce(m); + if (!apei_err) + apei_err = apei_write_mce(m); + } + } + if (final) { + print_mce(final); + if (!apei_err) + apei_err = apei_write_mce(final); + } + if (cpu_missing) + pr_emerg(HW_ERR "Some CPUs didn't answer in synchronization\n"); + if (exp) + pr_emerg(HW_ERR "Machine check: %s\n", exp); + if (!fake_panic) { + if (panic_timeout == 0) + panic_timeout = mca_cfg.panic_timeout; + panic(msg); + } else + pr_emerg(HW_ERR "Fake kernel panic: %s\n", msg); + +out: + instrumentation_end(); +} + +/* Support code for software error injection */ + +static int msr_to_offset(u32 msr) +{ + unsigned bank = __this_cpu_read(injectm.bank); + + if (msr == mca_cfg.rip_msr) + return offsetof(struct mce, ip); + if (msr == mca_msr_reg(bank, MCA_STATUS)) + return offsetof(struct mce, status); + if (msr == mca_msr_reg(bank, MCA_ADDR)) + return offsetof(struct mce, addr); + if (msr == mca_msr_reg(bank, MCA_MISC)) + return offsetof(struct mce, misc); + if (msr == MSR_IA32_MCG_STATUS) + return offsetof(struct mce, mcgstatus); + return -1; +} + +__visible bool ex_handler_rdmsr_fault(const struct exception_table_entry *fixup, + struct pt_regs *regs, int trapnr, + unsigned long error_code, + unsigned long fault_addr) +{ + pr_emerg("MSR access error: RDMSR from 0x%x at rIP: 0x%lx (%pS)\n", + (unsigned int)regs->cx, regs->ip, (void *)regs->ip); + + show_stack_regs(regs); + + panic("MCA architectural violation!\n"); + + while (true) + cpu_relax(); + + return true; +} + +/* MSR access wrappers used for error injection */ +static noinstr u64 mce_rdmsrl(u32 msr) +{ + DECLARE_ARGS(val, low, high); + + if (__this_cpu_read(injectm.finished)) { + int offset; + u64 ret; + + instrumentation_begin(); + + offset = msr_to_offset(msr); + if (offset < 0) + ret = 0; + else + ret = *(u64 *)((char *)this_cpu_ptr(&injectm) + offset); + + instrumentation_end(); + + return ret; + } + + /* + * RDMSR on MCA MSRs should not fault. If they do, this is very much an + * architectural violation and needs to be reported to hw vendor. Panic + * the box to not allow any further progress. + */ + asm volatile("1: rdmsr\n" + "2:\n" + _ASM_EXTABLE_HANDLE(1b, 2b, ex_handler_rdmsr_fault) + : EAX_EDX_RET(val, low, high) : "c" (msr)); + + + return EAX_EDX_VAL(val, low, high); +} + +__visible bool ex_handler_wrmsr_fault(const struct exception_table_entry *fixup, + struct pt_regs *regs, int trapnr, + unsigned long error_code, + unsigned long fault_addr) +{ + pr_emerg("MSR access error: WRMSR to 0x%x (tried to write 0x%08x%08x) at rIP: 0x%lx (%pS)\n", + (unsigned int)regs->cx, (unsigned int)regs->dx, (unsigned int)regs->ax, + regs->ip, (void *)regs->ip); + + show_stack_regs(regs); + + panic("MCA architectural violation!\n"); + + while (true) + cpu_relax(); + + return true; +} + +static noinstr void mce_wrmsrl(u32 msr, u64 v) +{ + u32 low, high; + + if (__this_cpu_read(injectm.finished)) { + int offset; + + instrumentation_begin(); + + offset = msr_to_offset(msr); + if (offset >= 0) + *(u64 *)((char *)this_cpu_ptr(&injectm) + offset) = v; + + instrumentation_end(); + + return; + } + + low = (u32)v; + high = (u32)(v >> 32); + + /* See comment in mce_rdmsrl() */ + asm volatile("1: wrmsr\n" + "2:\n" + _ASM_EXTABLE_HANDLE(1b, 2b, ex_handler_wrmsr_fault) + : : "c" (msr), "a"(low), "d" (high) : "memory"); +} + +/* + * Collect all global (w.r.t. this processor) status about this machine + * check into our "mce" struct so that we can use it later to assess + * the severity of the problem as we read per-bank specific details. + */ +static inline void mce_gather_info(struct mce *m, struct pt_regs *regs) +{ + mce_setup(m); + + m->mcgstatus = mce_rdmsrl(MSR_IA32_MCG_STATUS); + if (regs) { + /* + * Get the address of the instruction at the time of + * the machine check error. + */ + if (m->mcgstatus & (MCG_STATUS_RIPV|MCG_STATUS_EIPV)) { + m->ip = regs->ip; + m->cs = regs->cs; + + /* + * When in VM86 mode make the cs look like ring 3 + * always. This is a lie, but it's better than passing + * the additional vm86 bit around everywhere. + */ + if (v8086_mode(regs)) + m->cs |= 3; + } + /* Use accurate RIP reporting if available. */ + if (mca_cfg.rip_msr) + m->ip = mce_rdmsrl(mca_cfg.rip_msr); + } +} + +int mce_available(struct cpuinfo_x86 *c) +{ + if (mca_cfg.disabled) + return 0; + return cpu_has(c, X86_FEATURE_MCE) && cpu_has(c, X86_FEATURE_MCA); +} + +static void mce_schedule_work(void) +{ + if (!mce_gen_pool_empty()) + schedule_work(&mce_work); +} + +static void mce_irq_work_cb(struct irq_work *entry) +{ + mce_schedule_work(); +} + +/* + * Check if the address reported by the CPU is in a format we can parse. + * It would be possible to add code for most other cases, but all would + * be somewhat complicated (e.g. segment offset would require an instruction + * parser). So only support physical addresses up to page granuality for now. + */ +int mce_usable_address(struct mce *m) +{ + if (!(m->status & MCI_STATUS_ADDRV)) + return 0; + + /* Checks after this one are Intel/Zhaoxin-specific: */ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL && + boot_cpu_data.x86_vendor != X86_VENDOR_ZHAOXIN) + return 1; + + if (!(m->status & MCI_STATUS_MISCV)) + return 0; + + if (MCI_MISC_ADDR_LSB(m->misc) > PAGE_SHIFT) + return 0; + + if (MCI_MISC_ADDR_MODE(m->misc) != MCI_MISC_ADDR_PHYS) + return 0; + + return 1; +} +EXPORT_SYMBOL_GPL(mce_usable_address); + +bool mce_is_memory_error(struct mce *m) +{ + switch (m->cpuvendor) { + case X86_VENDOR_AMD: + case X86_VENDOR_HYGON: + return amd_mce_is_memory_error(m); + + case X86_VENDOR_INTEL: + case X86_VENDOR_ZHAOXIN: + /* + * Intel SDM Volume 3B - 15.9.2 Compound Error Codes + * + * Bit 7 of the MCACOD field of IA32_MCi_STATUS is used for + * indicating a memory error. Bit 8 is used for indicating a + * cache hierarchy error. The combination of bit 2 and bit 3 + * is used for indicating a `generic' cache hierarchy error + * But we can't just blindly check the above bits, because if + * bit 11 is set, then it is a bus/interconnect error - and + * either way the above bits just gives more detail on what + * bus/interconnect error happened. Note that bit 12 can be + * ignored, as it's the "filter" bit. + */ + return (m->status & 0xef80) == BIT(7) || + (m->status & 0xef00) == BIT(8) || + (m->status & 0xeffc) == 0xc; + + default: + return false; + } +} +EXPORT_SYMBOL_GPL(mce_is_memory_error); + +static bool whole_page(struct mce *m) +{ + if (!mca_cfg.ser || !(m->status & MCI_STATUS_MISCV)) + return true; + + return MCI_MISC_ADDR_LSB(m->misc) >= PAGE_SHIFT; +} + +bool mce_is_correctable(struct mce *m) +{ + if (m->cpuvendor == X86_VENDOR_AMD && m->status & MCI_STATUS_DEFERRED) + return false; + + if (m->cpuvendor == X86_VENDOR_HYGON && m->status & MCI_STATUS_DEFERRED) + return false; + + if (m->status & MCI_STATUS_UC) + return false; + + return true; +} +EXPORT_SYMBOL_GPL(mce_is_correctable); + +static int mce_early_notifier(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct mce *m = (struct mce *)data; + + if (!m) + return NOTIFY_DONE; + + /* Emit the trace record: */ + trace_mce_record(m); + + set_bit(0, &mce_need_notify); + + mce_notify_irq(); + + return NOTIFY_DONE; +} + +static struct notifier_block early_nb = { + .notifier_call = mce_early_notifier, + .priority = MCE_PRIO_EARLY, +}; + +static int uc_decode_notifier(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct mce *mce = (struct mce *)data; + unsigned long pfn; + + if (!mce || !mce_usable_address(mce)) + return NOTIFY_DONE; + + if (mce->severity != MCE_AO_SEVERITY && + mce->severity != MCE_DEFERRED_SEVERITY) + return NOTIFY_DONE; + + pfn = mce->addr >> PAGE_SHIFT; + if (!memory_failure(pfn, 0)) { + set_mce_nospec(pfn, whole_page(mce)); + mce->kflags |= MCE_HANDLED_UC; + } + + return NOTIFY_OK; +} + +static struct notifier_block mce_uc_nb = { + .notifier_call = uc_decode_notifier, + .priority = MCE_PRIO_UC, +}; + +static int mce_default_notifier(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct mce *m = (struct mce *)data; + + if (!m) + return NOTIFY_DONE; + + if (mca_cfg.print_all || !m->kflags) + __print_mce(m); + + return NOTIFY_DONE; +} + +static struct notifier_block mce_default_nb = { + .notifier_call = mce_default_notifier, + /* lowest prio, we want it to run last. */ + .priority = MCE_PRIO_LOWEST, +}; + +/* + * Read ADDR and MISC registers. + */ +static noinstr void mce_read_aux(struct mce *m, int i) +{ + if (m->status & MCI_STATUS_MISCV) + m->misc = mce_rdmsrl(mca_msr_reg(i, MCA_MISC)); + + if (m->status & MCI_STATUS_ADDRV) { + m->addr = mce_rdmsrl(mca_msr_reg(i, MCA_ADDR)); + + /* + * Mask the reported address by the reported granularity. + */ + if (mca_cfg.ser && (m->status & MCI_STATUS_MISCV)) { + u8 shift = MCI_MISC_ADDR_LSB(m->misc); + m->addr >>= shift; + m->addr <<= shift; + } + + /* + * Extract [55:<lsb>] where lsb is the least significant + * *valid* bit of the address bits. + */ + if (mce_flags.smca) { + u8 lsb = (m->addr >> 56) & 0x3f; + + m->addr &= GENMASK_ULL(55, lsb); + } + } + + if (mce_flags.smca) { + m->ipid = mce_rdmsrl(MSR_AMD64_SMCA_MCx_IPID(i)); + + if (m->status & MCI_STATUS_SYNDV) + m->synd = mce_rdmsrl(MSR_AMD64_SMCA_MCx_SYND(i)); + } +} + +DEFINE_PER_CPU(unsigned, mce_poll_count); + +/* + * Poll for corrected events or events that happened before reset. + * Those are just logged through /dev/mcelog. + * + * This is executed in standard interrupt context. + * + * Note: spec recommends to panic for fatal unsignalled + * errors here. However this would be quite problematic -- + * we would need to reimplement the Monarch handling and + * it would mess up the exclusion between exception handler + * and poll handler -- * so we skip this for now. + * These cases should not happen anyways, or only when the CPU + * is already totally * confused. In this case it's likely it will + * not fully execute the machine check handler either. + */ +bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + bool error_seen = false; + struct mce m; + int i; + + this_cpu_inc(mce_poll_count); + + mce_gather_info(&m, NULL); + + if (flags & MCP_TIMESTAMP) + m.tsc = rdtsc(); + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + if (!mce_banks[i].ctl || !test_bit(i, *b)) + continue; + + m.misc = 0; + m.addr = 0; + m.bank = i; + + barrier(); + m.status = mce_rdmsrl(mca_msr_reg(i, MCA_STATUS)); + + /* If this entry is not valid, ignore it */ + if (!(m.status & MCI_STATUS_VAL)) + continue; + + /* + * If we are logging everything (at CPU online) or this + * is a corrected error, then we must log it. + */ + if ((flags & MCP_UC) || !(m.status & MCI_STATUS_UC)) + goto log_it; + + /* + * Newer Intel systems that support software error + * recovery need to make additional checks. Other + * CPUs should skip over uncorrected errors, but log + * everything else. + */ + if (!mca_cfg.ser) { + if (m.status & MCI_STATUS_UC) + continue; + goto log_it; + } + + /* Log "not enabled" (speculative) errors */ + if (!(m.status & MCI_STATUS_EN)) + goto log_it; + + /* + * Log UCNA (SDM: 15.6.3 "UCR Error Classification") + * UC == 1 && PCC == 0 && S == 0 + */ + if (!(m.status & MCI_STATUS_PCC) && !(m.status & MCI_STATUS_S)) + goto log_it; + + /* + * Skip anything else. Presumption is that our read of this + * bank is racing with a machine check. Leave the log alone + * for do_machine_check() to deal with it. + */ + continue; + +log_it: + error_seen = true; + + if (flags & MCP_DONTLOG) + goto clear_it; + + mce_read_aux(&m, i); + m.severity = mce_severity(&m, NULL, mca_cfg.tolerant, NULL, false); + /* + * Don't get the IP here because it's unlikely to + * have anything to do with the actual error location. + */ + + if (mca_cfg.dont_log_ce && !mce_usable_address(&m)) + goto clear_it; + + if (flags & MCP_QUEUE_LOG) + mce_gen_pool_add(&m); + else + mce_log(&m); + +clear_it: + /* + * Clear state for this bank. + */ + mce_wrmsrl(mca_msr_reg(i, MCA_STATUS), 0); + } + + /* + * Don't clear MCG_STATUS here because it's only defined for + * exceptions. + */ + + sync_core(); + + return error_seen; +} +EXPORT_SYMBOL_GPL(machine_check_poll); + +/* + * Do a quick check if any of the events requires a panic. + * This decides if we keep the events around or clear them. + */ +static int mce_no_way_out(struct mce *m, char **msg, unsigned long *validp, + struct pt_regs *regs) +{ + char *tmp = *msg; + int i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + m->status = mce_rdmsrl(mca_msr_reg(i, MCA_STATUS)); + if (!(m->status & MCI_STATUS_VAL)) + continue; + + __set_bit(i, validp); + if (quirk_no_way_out) + quirk_no_way_out(i, m, regs); + + m->bank = i; + if (mce_severity(m, regs, mca_cfg.tolerant, &tmp, true) >= MCE_PANIC_SEVERITY) { + mce_read_aux(m, i); + *msg = tmp; + return 1; + } + } + return 0; +} + +/* + * Variable to establish order between CPUs while scanning. + * Each CPU spins initially until executing is equal its number. + */ +static atomic_t mce_executing; + +/* + * Defines order of CPUs on entry. First CPU becomes Monarch. + */ +static atomic_t mce_callin; + +/* + * Check if a timeout waiting for other CPUs happened. + */ +static int mce_timed_out(u64 *t, const char *msg) +{ + /* + * The others already did panic for some reason. + * Bail out like in a timeout. + * rmb() to tell the compiler that system_state + * might have been modified by someone else. + */ + rmb(); + if (atomic_read(&mce_panicked)) + wait_for_panic(); + if (!mca_cfg.monarch_timeout) + goto out; + if ((s64)*t < SPINUNIT) { + if (mca_cfg.tolerant <= 1) + mce_panic(msg, NULL, NULL); + cpu_missing = 1; + return 1; + } + *t -= SPINUNIT; +out: + touch_nmi_watchdog(); + return 0; +} + +/* + * The Monarch's reign. The Monarch is the CPU who entered + * the machine check handler first. It waits for the others to + * raise the exception too and then grades them. When any + * error is fatal panic. Only then let the others continue. + * + * The other CPUs entering the MCE handler will be controlled by the + * Monarch. They are called Subjects. + * + * This way we prevent any potential data corruption in a unrecoverable case + * and also makes sure always all CPU's errors are examined. + * + * Also this detects the case of a machine check event coming from outer + * space (not detected by any CPUs) In this case some external agent wants + * us to shut down, so panic too. + * + * The other CPUs might still decide to panic if the handler happens + * in a unrecoverable place, but in this case the system is in a semi-stable + * state and won't corrupt anything by itself. It's ok to let the others + * continue for a bit first. + * + * All the spin loops have timeouts; when a timeout happens a CPU + * typically elects itself to be Monarch. + */ +static void mce_reign(void) +{ + int cpu; + struct mce *m = NULL; + int global_worst = 0; + char *msg = NULL; + + /* + * This CPU is the Monarch and the other CPUs have run + * through their handlers. + * Grade the severity of the errors of all the CPUs. + */ + for_each_possible_cpu(cpu) { + struct mce *mtmp = &per_cpu(mces_seen, cpu); + + if (mtmp->severity > global_worst) { + global_worst = mtmp->severity; + m = &per_cpu(mces_seen, cpu); + } + } + + /* + * Cannot recover? Panic here then. + * This dumps all the mces in the log buffer and stops the + * other CPUs. + */ + if (m && global_worst >= MCE_PANIC_SEVERITY && mca_cfg.tolerant < 3) { + /* call mce_severity() to get "msg" for panic */ + mce_severity(m, NULL, mca_cfg.tolerant, &msg, true); + mce_panic("Fatal machine check", m, msg); + } + + /* + * For UC somewhere we let the CPU who detects it handle it. + * Also must let continue the others, otherwise the handling + * CPU could deadlock on a lock. + */ + + /* + * No machine check event found. Must be some external + * source or one CPU is hung. Panic. + */ + if (global_worst <= MCE_KEEP_SEVERITY && mca_cfg.tolerant < 3) + mce_panic("Fatal machine check from unknown source", NULL, NULL); + + /* + * Now clear all the mces_seen so that they don't reappear on + * the next mce. + */ + for_each_possible_cpu(cpu) + memset(&per_cpu(mces_seen, cpu), 0, sizeof(struct mce)); +} + +static atomic_t global_nwo; + +/* + * Start of Monarch synchronization. This waits until all CPUs have + * entered the exception handler and then determines if any of them + * saw a fatal event that requires panic. Then it executes them + * in the entry order. + * TBD double check parallel CPU hotunplug + */ +static int mce_start(int *no_way_out) +{ + int order; + int cpus = num_online_cpus(); + u64 timeout = (u64)mca_cfg.monarch_timeout * NSEC_PER_USEC; + + if (!timeout) + return -1; + + atomic_add(*no_way_out, &global_nwo); + /* + * Rely on the implied barrier below, such that global_nwo + * is updated before mce_callin. + */ + order = atomic_inc_return(&mce_callin); + + /* + * Wait for everyone. + */ + while (atomic_read(&mce_callin) != cpus) { + if (mce_timed_out(&timeout, + "Timeout: Not all CPUs entered broadcast exception handler")) { + atomic_set(&global_nwo, 0); + return -1; + } + ndelay(SPINUNIT); + } + + /* + * mce_callin should be read before global_nwo + */ + smp_rmb(); + + if (order == 1) { + /* + * Monarch: Starts executing now, the others wait. + */ + atomic_set(&mce_executing, 1); + } else { + /* + * Subject: Now start the scanning loop one by one in + * the original callin order. + * This way when there are any shared banks it will be + * only seen by one CPU before cleared, avoiding duplicates. + */ + while (atomic_read(&mce_executing) < order) { + if (mce_timed_out(&timeout, + "Timeout: Subject CPUs unable to finish machine check processing")) { + atomic_set(&global_nwo, 0); + return -1; + } + ndelay(SPINUNIT); + } + } + + /* + * Cache the global no_way_out state. + */ + *no_way_out = atomic_read(&global_nwo); + + return order; +} + +/* + * Synchronize between CPUs after main scanning loop. + * This invokes the bulk of the Monarch processing. + */ +static noinstr int mce_end(int order) +{ + u64 timeout = (u64)mca_cfg.monarch_timeout * NSEC_PER_USEC; + int ret = -1; + + /* Allow instrumentation around external facilities. */ + instrumentation_begin(); + + if (!timeout) + goto reset; + if (order < 0) + goto reset; + + /* + * Allow others to run. + */ + atomic_inc(&mce_executing); + + if (order == 1) { + /* CHECKME: Can this race with a parallel hotplug? */ + int cpus = num_online_cpus(); + + /* + * Monarch: Wait for everyone to go through their scanning + * loops. + */ + while (atomic_read(&mce_executing) <= cpus) { + if (mce_timed_out(&timeout, + "Timeout: Monarch CPU unable to finish machine check processing")) + goto reset; + ndelay(SPINUNIT); + } + + mce_reign(); + barrier(); + ret = 0; + } else { + /* + * Subject: Wait for Monarch to finish. + */ + while (atomic_read(&mce_executing) != 0) { + if (mce_timed_out(&timeout, + "Timeout: Monarch CPU did not finish machine check processing")) + goto reset; + ndelay(SPINUNIT); + } + + /* + * Don't reset anything. That's done by the Monarch. + */ + ret = 0; + goto out; + } + + /* + * Reset all global state. + */ +reset: + atomic_set(&global_nwo, 0); + atomic_set(&mce_callin, 0); + barrier(); + + /* + * Let others run again. + */ + atomic_set(&mce_executing, 0); + +out: + instrumentation_end(); + + return ret; +} + +static void mce_clear_state(unsigned long *toclear) +{ + int i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + if (test_bit(i, toclear)) + mce_wrmsrl(mca_msr_reg(i, MCA_STATUS), 0); + } +} + +/* + * Cases where we avoid rendezvous handler timeout: + * 1) If this CPU is offline. + * + * 2) If crashing_cpu was set, e.g. we're entering kdump and we need to + * skip those CPUs which remain looping in the 1st kernel - see + * crash_nmi_callback(). + * + * Note: there still is a small window between kexec-ing and the new, + * kdump kernel establishing a new #MC handler where a broadcasted MCE + * might not get handled properly. + */ +static noinstr bool mce_check_crashing_cpu(void) +{ + unsigned int cpu = smp_processor_id(); + + if (arch_cpu_is_offline(cpu) || + (crashing_cpu != -1 && crashing_cpu != cpu)) { + u64 mcgstatus; + + mcgstatus = __rdmsr(MSR_IA32_MCG_STATUS); + + if (boot_cpu_data.x86_vendor == X86_VENDOR_ZHAOXIN) { + if (mcgstatus & MCG_STATUS_LMCES) + return false; + } + + if (mcgstatus & MCG_STATUS_RIPV) { + __wrmsr(MSR_IA32_MCG_STATUS, 0, 0); + return true; + } + } + return false; +} + +static void __mc_scan_banks(struct mce *m, struct pt_regs *regs, struct mce *final, + unsigned long *toclear, unsigned long *valid_banks, + int no_way_out, int *worst) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + struct mca_config *cfg = &mca_cfg; + int severity, i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + __clear_bit(i, toclear); + if (!test_bit(i, valid_banks)) + continue; + + if (!mce_banks[i].ctl) + continue; + + m->misc = 0; + m->addr = 0; + m->bank = i; + + m->status = mce_rdmsrl(mca_msr_reg(i, MCA_STATUS)); + if (!(m->status & MCI_STATUS_VAL)) + continue; + + /* + * Corrected or non-signaled errors are handled by + * machine_check_poll(). Leave them alone, unless this panics. + */ + if (!(m->status & (cfg->ser ? MCI_STATUS_S : MCI_STATUS_UC)) && + !no_way_out) + continue; + + /* Set taint even when machine check was not enabled. */ + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); + + severity = mce_severity(m, regs, cfg->tolerant, NULL, true); + + /* + * When machine check was for corrected/deferred handler don't + * touch, unless we're panicking. + */ + if ((severity == MCE_KEEP_SEVERITY || + severity == MCE_UCNA_SEVERITY) && !no_way_out) + continue; + + __set_bit(i, toclear); + + /* Machine check event was not enabled. Clear, but ignore. */ + if (severity == MCE_NO_SEVERITY) + continue; + + mce_read_aux(m, i); + + /* assuming valid severity level != 0 */ + m->severity = severity; + + mce_log(m); + + if (severity > *worst) { + *final = *m; + *worst = severity; + } + } + + /* mce_clear_state will clear *final, save locally for use later */ + *m = *final; +} + +static void kill_me_now(struct callback_head *ch) +{ + struct task_struct *p = container_of(ch, struct task_struct, mce_kill_me); + + p->mce_count = 0; + force_sig(SIGBUS); +} + +static void kill_me_maybe(struct callback_head *cb) +{ + struct task_struct *p = container_of(cb, struct task_struct, mce_kill_me); + int flags = MF_ACTION_REQUIRED; + + p->mce_count = 0; + pr_err("Uncorrected hardware memory error in user-access at %llx", p->mce_addr); + + if (!p->mce_ripv) + flags |= MF_MUST_KILL; + + if (!memory_failure(p->mce_addr >> PAGE_SHIFT, flags) && + !(p->mce_kflags & MCE_IN_KERNEL_COPYIN)) { + set_mce_nospec(p->mce_addr >> PAGE_SHIFT, p->mce_whole_page); + sync_core(); + return; + } + + if (p->mce_vaddr != (void __user *)-1l) { + force_sig_mceerr(BUS_MCEERR_AR, p->mce_vaddr, PAGE_SHIFT); + } else { + pr_err("Memory error not recovered"); + kill_me_now(cb); + } +} + +static void queue_task_work(struct mce *m, char *msg, int kill_current_task) +{ + int count = ++current->mce_count; + + /* First call, save all the details */ + if (count == 1) { + current->mce_addr = m->addr; + current->mce_kflags = m->kflags; + current->mce_ripv = !!(m->mcgstatus & MCG_STATUS_RIPV); + current->mce_whole_page = whole_page(m); + + if (kill_current_task) + current->mce_kill_me.func = kill_me_now; + else + current->mce_kill_me.func = kill_me_maybe; + } + + /* Ten is likely overkill. Don't expect more than two faults before task_work() */ + if (count > 10) + mce_panic("Too many consecutive machine checks while accessing user data", m, msg); + + /* Second or later call, make sure page address matches the one from first call */ + if (count > 1 && (current->mce_addr >> PAGE_SHIFT) != (m->addr >> PAGE_SHIFT)) + mce_panic("Consecutive machine checks to different user pages", m, msg); + + /* Do not call task_work_add() more than once */ + if (count > 1) + return; + + task_work_add(current, ¤t->mce_kill_me, TWA_RESUME); +} + +/* + * The actual machine check handler. This only handles real + * exceptions when something got corrupted coming in through int 18. + * + * This is executed in NMI context not subject to normal locking rules. This + * implies that most kernel services cannot be safely used. Don't even + * think about putting a printk in there! + * + * On Intel systems this is entered on all CPUs in parallel through + * MCE broadcast. However some CPUs might be broken beyond repair, + * so be always careful when synchronizing with others. + * + * Tracing and kprobes are disabled: if we interrupted a kernel context + * with IF=1, we need to minimize stack usage. There are also recursion + * issues: if the machine check was due to a failure of the memory + * backing the user stack, tracing that reads the user stack will cause + * potentially infinite recursion. + */ +noinstr void do_machine_check(struct pt_regs *regs) +{ + DECLARE_BITMAP(valid_banks, MAX_NR_BANKS); + DECLARE_BITMAP(toclear, MAX_NR_BANKS); + struct mca_config *cfg = &mca_cfg; + struct mce m, *final; + char *msg = NULL; + int worst = 0; + + /* + * Establish sequential order between the CPUs entering the machine + * check handler. + */ + int order = -1; + + /* + * If no_way_out gets set, there is no safe way to recover from this + * MCE. If mca_cfg.tolerant is cranked up, we'll try anyway. + */ + int no_way_out = 0; + + /* + * If kill_it gets set, there might be a way to recover from this + * error. + */ + int kill_it = 0; + + /* + * MCEs are always local on AMD. Same is determined by MCG_STATUS_LMCES + * on Intel. + */ + int lmce = 1; + + this_cpu_inc(mce_exception_count); + + mce_gather_info(&m, regs); + m.tsc = rdtsc(); + + final = this_cpu_ptr(&mces_seen); + *final = m; + + memset(valid_banks, 0, sizeof(valid_banks)); + no_way_out = mce_no_way_out(&m, &msg, valid_banks, regs); + + barrier(); + + /* + * When no restart IP might need to kill or panic. + * Assume the worst for now, but if we find the + * severity is MCE_AR_SEVERITY we have other options. + */ + if (!(m.mcgstatus & MCG_STATUS_RIPV)) + kill_it = 1; + + /* + * Check if this MCE is signaled to only this logical processor, + * on Intel, Zhaoxin only. + */ + if (m.cpuvendor == X86_VENDOR_INTEL || + m.cpuvendor == X86_VENDOR_ZHAOXIN) + lmce = m.mcgstatus & MCG_STATUS_LMCES; + + /* + * Local machine check may already know that we have to panic. + * Broadcast machine check begins rendezvous in mce_start() + * Go through all banks in exclusion of the other CPUs. This way we + * don't report duplicated events on shared banks because the first one + * to see it will clear it. + */ + if (lmce) { + if (no_way_out) + mce_panic("Fatal local machine check", &m, msg); + } else { + order = mce_start(&no_way_out); + } + + __mc_scan_banks(&m, regs, final, toclear, valid_banks, no_way_out, &worst); + + if (!no_way_out) + mce_clear_state(toclear); + + /* + * Do most of the synchronization with other CPUs. + * When there's any problem use only local no_way_out state. + */ + if (!lmce) { + if (mce_end(order) < 0) { + if (!no_way_out) + no_way_out = worst >= MCE_PANIC_SEVERITY; + } + } else { + /* + * If there was a fatal machine check we should have + * already called mce_panic earlier in this function. + * Since we re-read the banks, we might have found + * something new. Check again to see if we found a + * fatal error. We call "mce_severity()" again to + * make sure we have the right "msg". + */ + if (worst >= MCE_PANIC_SEVERITY && mca_cfg.tolerant < 3) { + mce_severity(&m, regs, cfg->tolerant, &msg, true); + mce_panic("Local fatal machine check!", &m, msg); + } + } + + /* + * If tolerant is at an insane level we drop requests to kill + * processes and continue even when there is no way out. + */ + if (cfg->tolerant == 3) + kill_it = 0; + else if (no_way_out) + mce_panic("Fatal machine check on current CPU", &m, msg); + + if (worst > 0) + irq_work_queue(&mce_irq_work); + + if (worst != MCE_AR_SEVERITY && !kill_it) + goto out; + + /* + * Enable instrumentation around the external facilities like + * task_work_add() (via queue_task_work()), fixup_exception() etc. + * For now, that is. Fixing this properly would need a lot more involved + * reorganization. + */ + instrumentation_begin(); + + /* Fault was in user mode and we need to take some action */ + if ((m.cs & 3) == 3) { + /* If this triggers there is no way to recover. Die hard. */ + BUG_ON(!on_thread_stack() || !user_mode(regs)); + + queue_task_work(&m, msg, kill_it); + + } else { + /* + * Handle an MCE which has happened in kernel space but from + * which the kernel can recover: ex_has_fault_handler() has + * already verified that the rIP at which the error happened is + * a rIP from which the kernel can recover (by jumping to + * recovery code specified in _ASM_EXTABLE_FAULT()) and the + * corresponding exception handler which would do that is the + * proper one. + */ + if (m.kflags & MCE_IN_KERNEL_RECOV) { + if (!fixup_exception(regs, X86_TRAP_MC, 0, 0)) + mce_panic("Failed kernel mode recovery", &m, msg); + } + + if (m.kflags & MCE_IN_KERNEL_COPYIN) + queue_task_work(&m, msg, kill_it); + } + + instrumentation_end(); + +out: + mce_wrmsrl(MSR_IA32_MCG_STATUS, 0); +} +EXPORT_SYMBOL_GPL(do_machine_check); + +#ifndef CONFIG_MEMORY_FAILURE +int memory_failure(unsigned long pfn, int flags) +{ + /* mce_severity() should not hand us an ACTION_REQUIRED error */ + BUG_ON(flags & MF_ACTION_REQUIRED); + pr_err("Uncorrected memory error in page 0x%lx ignored\n" + "Rebuild kernel with CONFIG_MEMORY_FAILURE=y for smarter handling\n", + pfn); + + return 0; +} +#endif + +/* + * Periodic polling timer for "silent" machine check errors. If the + * poller finds an MCE, poll 2x faster. When the poller finds no more + * errors, poll 2x slower (up to check_interval seconds). + */ +static unsigned long check_interval = INITIAL_CHECK_INTERVAL; + +static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ +static DEFINE_PER_CPU(struct timer_list, mce_timer); + +static unsigned long mce_adjust_timer_default(unsigned long interval) +{ + return interval; +} + +static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; + +static void __start_timer(struct timer_list *t, unsigned long interval) +{ + unsigned long when = jiffies + interval; + unsigned long flags; + + local_irq_save(flags); + + if (!timer_pending(t) || time_before(when, t->expires)) + mod_timer(t, round_jiffies(when)); + + local_irq_restore(flags); +} + +static void mce_timer_fn(struct timer_list *t) +{ + struct timer_list *cpu_t = this_cpu_ptr(&mce_timer); + unsigned long iv; + + WARN_ON(cpu_t != t); + + iv = __this_cpu_read(mce_next_interval); + + if (mce_available(this_cpu_ptr(&cpu_info))) { + machine_check_poll(0, this_cpu_ptr(&mce_poll_banks)); + + if (mce_intel_cmci_poll()) { + iv = mce_adjust_timer(iv); + goto done; + } + } + + /* + * Alert userspace if needed. If we logged an MCE, reduce the polling + * interval, otherwise increase the polling interval. + */ + if (mce_notify_irq()) + iv = max(iv / 2, (unsigned long) HZ/100); + else + iv = min(iv * 2, round_jiffies_relative(check_interval * HZ)); + +done: + __this_cpu_write(mce_next_interval, iv); + __start_timer(t, iv); +} + +/* + * Ensure that the timer is firing in @interval from now. + */ +void mce_timer_kick(unsigned long interval) +{ + struct timer_list *t = this_cpu_ptr(&mce_timer); + unsigned long iv = __this_cpu_read(mce_next_interval); + + __start_timer(t, interval); + + if (interval < iv) + __this_cpu_write(mce_next_interval, interval); +} + +/* Must not be called in IRQ context where del_timer_sync() can deadlock */ +static void mce_timer_delete_all(void) +{ + int cpu; + + for_each_online_cpu(cpu) + del_timer_sync(&per_cpu(mce_timer, cpu)); +} + +/* + * Notify the user(s) about new machine check events. + * Can be called from interrupt context, but not from machine check/NMI + * context. + */ +int mce_notify_irq(void) +{ + /* Not more than two messages every minute */ + static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); + + if (test_and_clear_bit(0, &mce_need_notify)) { + mce_work_trigger(); + + if (__ratelimit(&ratelimit)) + pr_info(HW_ERR "Machine check events logged\n"); + + return 1; + } + return 0; +} +EXPORT_SYMBOL_GPL(mce_notify_irq); + +static void __mcheck_cpu_mce_banks_init(void) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + u8 n_banks = this_cpu_read(mce_num_banks); + int i; + + for (i = 0; i < n_banks; i++) { + struct mce_bank *b = &mce_banks[i]; + + /* + * Init them all, __mcheck_cpu_apply_quirks() is going to apply + * the required vendor quirks before + * __mcheck_cpu_init_clear_banks() does the final bank setup. + */ + b->ctl = -1ULL; + b->init = 1; + } +} + +/* + * Initialize Machine Checks for a CPU. + */ +static void __mcheck_cpu_cap_init(void) +{ + u64 cap; + u8 b; + + rdmsrl(MSR_IA32_MCG_CAP, cap); + + b = cap & MCG_BANKCNT_MASK; + + if (b > MAX_NR_BANKS) { + pr_warn("CPU%d: Using only %u machine check banks out of %u\n", + smp_processor_id(), MAX_NR_BANKS, b); + b = MAX_NR_BANKS; + } + + this_cpu_write(mce_num_banks, b); + + __mcheck_cpu_mce_banks_init(); + + /* Use accurate RIP reporting if available. */ + if ((cap & MCG_EXT_P) && MCG_EXT_CNT(cap) >= 9) + mca_cfg.rip_msr = MSR_IA32_MCG_EIP; + + if (cap & MCG_SER_P) + mca_cfg.ser = 1; +} + +static void __mcheck_cpu_init_generic(void) +{ + enum mcp_flags m_fl = 0; + mce_banks_t all_banks; + u64 cap; + + if (!mca_cfg.bootlog) + m_fl = MCP_DONTLOG; + + /* + * Log the machine checks left over from the previous reset. Log them + * only, do not start processing them. That will happen in mcheck_late_init() + * when all consumers have been registered on the notifier chain. + */ + bitmap_fill(all_banks, MAX_NR_BANKS); + machine_check_poll(MCP_UC | MCP_QUEUE_LOG | m_fl, &all_banks); + + cr4_set_bits(X86_CR4_MCE); + + rdmsrl(MSR_IA32_MCG_CAP, cap); + if (cap & MCG_CTL_P) + wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); +} + +static void __mcheck_cpu_init_clear_banks(void) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + int i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + struct mce_bank *b = &mce_banks[i]; + + if (!b->init) + continue; + wrmsrl(mca_msr_reg(i, MCA_CTL), b->ctl); + wrmsrl(mca_msr_reg(i, MCA_STATUS), 0); + } +} + +/* + * Do a final check to see if there are any unused/RAZ banks. + * + * This must be done after the banks have been initialized and any quirks have + * been applied. + * + * Do not call this from any user-initiated flows, e.g. CPU hotplug or sysfs. + * Otherwise, a user who disables a bank will not be able to re-enable it + * without a system reboot. + */ +static void __mcheck_cpu_check_banks(void) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + u64 msrval; + int i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + struct mce_bank *b = &mce_banks[i]; + + if (!b->init) + continue; + + rdmsrl(mca_msr_reg(i, MCA_CTL), msrval); + b->init = !!msrval; + } +} + +/* + * During IFU recovery Sandy Bridge -EP4S processors set the RIPV and + * EIPV bits in MCG_STATUS to zero on the affected logical processor (SDM + * Vol 3B Table 15-20). But this confuses both the code that determines + * whether the machine check occurred in kernel or user mode, and also + * the severity assessment code. Pretend that EIPV was set, and take the + * ip/cs values from the pt_regs that mce_gather_info() ignored earlier. + */ +static void quirk_sandybridge_ifu(int bank, struct mce *m, struct pt_regs *regs) +{ + if (bank != 0) + return; + if ((m->mcgstatus & (MCG_STATUS_EIPV|MCG_STATUS_RIPV)) != 0) + return; + if ((m->status & (MCI_STATUS_OVER|MCI_STATUS_UC| + MCI_STATUS_EN|MCI_STATUS_MISCV|MCI_STATUS_ADDRV| + MCI_STATUS_PCC|MCI_STATUS_S|MCI_STATUS_AR| + MCACOD)) != + (MCI_STATUS_UC|MCI_STATUS_EN| + MCI_STATUS_MISCV|MCI_STATUS_ADDRV|MCI_STATUS_S| + MCI_STATUS_AR|MCACOD_INSTR)) + return; + + m->mcgstatus |= MCG_STATUS_EIPV; + m->ip = regs->ip; + m->cs = regs->cs; +} + +/* Add per CPU specific workarounds here */ +static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + struct mca_config *cfg = &mca_cfg; + + if (c->x86_vendor == X86_VENDOR_UNKNOWN) { + pr_info("unknown CPU type - not enabling MCE support\n"); + return -EOPNOTSUPP; + } + + /* This should be disabled by the BIOS, but isn't always */ + if (c->x86_vendor == X86_VENDOR_AMD) { + if (c->x86 == 15 && this_cpu_read(mce_num_banks) > 4) { + /* + * disable GART TBL walk error reporting, which + * trips off incorrectly with the IOMMU & 3ware + * & Cerberus: + */ + clear_bit(10, (unsigned long *)&mce_banks[4].ctl); + } + if (c->x86 < 0x11 && cfg->bootlog < 0) { + /* + * Lots of broken BIOS around that don't clear them + * by default and leave crap in there. Don't log: + */ + cfg->bootlog = 0; + } + /* + * Various K7s with broken bank 0 around. Always disable + * by default. + */ + if (c->x86 == 6 && this_cpu_read(mce_num_banks) > 0) + mce_banks[0].ctl = 0; + + /* + * overflow_recov is supported for F15h Models 00h-0fh + * even though we don't have a CPUID bit for it. + */ + if (c->x86 == 0x15 && c->x86_model <= 0xf) + mce_flags.overflow_recov = 1; + + } + + if (c->x86_vendor == X86_VENDOR_INTEL) { + /* + * SDM documents that on family 6 bank 0 should not be written + * because it aliases to another special BIOS controlled + * register. + * But it's not aliased anymore on model 0x1a+ + * Don't ignore bank 0 completely because there could be a + * valid event later, merely don't write CTL0. + */ + + if (c->x86 == 6 && c->x86_model < 0x1A && this_cpu_read(mce_num_banks) > 0) + mce_banks[0].init = 0; + + /* + * All newer Intel systems support MCE broadcasting. Enable + * synchronization with a one second timeout. + */ + if ((c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xe)) && + cfg->monarch_timeout < 0) + cfg->monarch_timeout = USEC_PER_SEC; + + /* + * There are also broken BIOSes on some Pentium M and + * earlier systems: + */ + if (c->x86 == 6 && c->x86_model <= 13 && cfg->bootlog < 0) + cfg->bootlog = 0; + + if (c->x86 == 6 && c->x86_model == 45) + quirk_no_way_out = quirk_sandybridge_ifu; + } + + if (c->x86_vendor == X86_VENDOR_ZHAOXIN) { + /* + * All newer Zhaoxin CPUs support MCE broadcasting. Enable + * synchronization with a one second timeout. + */ + if (c->x86 > 6 || (c->x86_model == 0x19 || c->x86_model == 0x1f)) { + if (cfg->monarch_timeout < 0) + cfg->monarch_timeout = USEC_PER_SEC; + } + } + + if (cfg->monarch_timeout < 0) + cfg->monarch_timeout = 0; + if (cfg->bootlog != 0) + cfg->panic_timeout = 30; + + return 0; +} + +static int __mcheck_cpu_ancient_init(struct cpuinfo_x86 *c) +{ + if (c->x86 != 5) + return 0; + + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + intel_p5_mcheck_init(c); + return 1; + break; + case X86_VENDOR_CENTAUR: + winchip_mcheck_init(c); + return 1; + break; + default: + return 0; + } + + return 0; +} + +/* + * Init basic CPU features needed for early decoding of MCEs. + */ +static void __mcheck_cpu_init_early(struct cpuinfo_x86 *c) +{ + if (c->x86_vendor == X86_VENDOR_AMD || c->x86_vendor == X86_VENDOR_HYGON) { + mce_flags.overflow_recov = !!cpu_has(c, X86_FEATURE_OVERFLOW_RECOV); + mce_flags.succor = !!cpu_has(c, X86_FEATURE_SUCCOR); + mce_flags.smca = !!cpu_has(c, X86_FEATURE_SMCA); + mce_flags.amd_threshold = 1; + } +} + +static void mce_centaur_feature_init(struct cpuinfo_x86 *c) +{ + struct mca_config *cfg = &mca_cfg; + + /* + * All newer Centaur CPUs support MCE broadcasting. Enable + * synchronization with a one second timeout. + */ + if ((c->x86 == 6 && c->x86_model == 0xf && c->x86_stepping >= 0xe) || + c->x86 > 6) { + if (cfg->monarch_timeout < 0) + cfg->monarch_timeout = USEC_PER_SEC; + } +} + +static void mce_zhaoxin_feature_init(struct cpuinfo_x86 *c) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + + /* + * These CPUs have MCA bank 8 which reports only one error type called + * SVAD (System View Address Decoder). The reporting of that error is + * controlled by IA32_MC8.CTL.0. + * + * If enabled, prefetching on these CPUs will cause SVAD MCE when + * virtual machines start and result in a system panic. Always disable + * bank 8 SVAD error by default. + */ + if ((c->x86 == 7 && c->x86_model == 0x1b) || + (c->x86_model == 0x19 || c->x86_model == 0x1f)) { + if (this_cpu_read(mce_num_banks) > 8) + mce_banks[8].ctl = 0; + } + + intel_init_cmci(); + intel_init_lmce(); + mce_adjust_timer = cmci_intel_adjust_timer; +} + +static void mce_zhaoxin_feature_clear(struct cpuinfo_x86 *c) +{ + intel_clear_lmce(); +} + +static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c) +{ + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + mce_intel_feature_init(c); + mce_adjust_timer = cmci_intel_adjust_timer; + break; + + case X86_VENDOR_AMD: { + mce_amd_feature_init(c); + break; + } + + case X86_VENDOR_HYGON: + mce_hygon_feature_init(c); + break; + + case X86_VENDOR_CENTAUR: + mce_centaur_feature_init(c); + break; + + case X86_VENDOR_ZHAOXIN: + mce_zhaoxin_feature_init(c); + break; + + default: + break; + } +} + +static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) +{ + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + mce_intel_feature_clear(c); + break; + + case X86_VENDOR_ZHAOXIN: + mce_zhaoxin_feature_clear(c); + break; + + default: + break; + } +} + +static void mce_start_timer(struct timer_list *t) +{ + unsigned long iv = check_interval * HZ; + + if (mca_cfg.ignore_ce || !iv) + return; + + this_cpu_write(mce_next_interval, iv); + __start_timer(t, iv); +} + +static void __mcheck_cpu_setup_timer(void) +{ + struct timer_list *t = this_cpu_ptr(&mce_timer); + + timer_setup(t, mce_timer_fn, TIMER_PINNED); +} + +static void __mcheck_cpu_init_timer(void) +{ + struct timer_list *t = this_cpu_ptr(&mce_timer); + + timer_setup(t, mce_timer_fn, TIMER_PINNED); + mce_start_timer(t); +} + +bool filter_mce(struct mce *m) +{ + if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) + return amd_filter_mce(m); + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + return intel_filter_mce(m); + + return false; +} + +/* Handle unconfigured int18 (should never happen) */ +static noinstr void unexpected_machine_check(struct pt_regs *regs) +{ + instrumentation_begin(); + pr_err("CPU#%d: Unexpected int18 (Machine Check)\n", + smp_processor_id()); + instrumentation_end(); +} + +/* Call the installed machine check handler for this CPU setup. */ +void (*machine_check_vector)(struct pt_regs *) = unexpected_machine_check; + +static __always_inline void exc_machine_check_kernel(struct pt_regs *regs) +{ + irqentry_state_t irq_state; + + WARN_ON_ONCE(user_mode(regs)); + + /* + * Only required when from kernel mode. See + * mce_check_crashing_cpu() for details. + */ + if (machine_check_vector == do_machine_check && + mce_check_crashing_cpu()) + return; + + irq_state = irqentry_nmi_enter(regs); + /* + * The call targets are marked noinstr, but objtool can't figure + * that out because it's an indirect call. Annotate it. + */ + instrumentation_begin(); + trace_hardirqs_off_finish(); + machine_check_vector(regs); + if (regs->flags & X86_EFLAGS_IF) + trace_hardirqs_on_prepare(); + instrumentation_end(); + irqentry_nmi_exit(regs, irq_state); +} + +static __always_inline void exc_machine_check_user(struct pt_regs *regs) +{ + irqentry_enter_from_user_mode(regs); + instrumentation_begin(); + machine_check_vector(regs); + instrumentation_end(); + irqentry_exit_to_user_mode(regs); +} + +#ifdef CONFIG_X86_64 +/* MCE hit kernel mode */ +DEFINE_IDTENTRY_MCE(exc_machine_check) +{ + unsigned long dr7; + + dr7 = local_db_save(); + exc_machine_check_kernel(regs); + local_db_restore(dr7); +} + +/* The user mode variant. */ +DEFINE_IDTENTRY_MCE_USER(exc_machine_check) +{ + unsigned long dr7; + + dr7 = local_db_save(); + exc_machine_check_user(regs); + local_db_restore(dr7); +} +#else +/* 32bit unified entry point */ +DEFINE_IDTENTRY_RAW(exc_machine_check) +{ + unsigned long dr7; + + dr7 = local_db_save(); + if (user_mode(regs)) + exc_machine_check_user(regs); + else + exc_machine_check_kernel(regs); + local_db_restore(dr7); +} +#endif + +/* + * Called for each booted CPU to set up machine checks. + * Must be called with preempt off: + */ +void mcheck_cpu_init(struct cpuinfo_x86 *c) +{ + if (mca_cfg.disabled) + return; + + if (__mcheck_cpu_ancient_init(c)) + return; + + if (!mce_available(c)) + return; + + __mcheck_cpu_cap_init(); + + if (__mcheck_cpu_apply_quirks(c) < 0) { + mca_cfg.disabled = 1; + return; + } + + if (mce_gen_pool_init()) { + mca_cfg.disabled = 1; + pr_emerg("Couldn't allocate MCE records pool!\n"); + return; + } + + machine_check_vector = do_machine_check; + + __mcheck_cpu_init_early(c); + __mcheck_cpu_init_generic(); + __mcheck_cpu_init_vendor(c); + __mcheck_cpu_init_clear_banks(); + __mcheck_cpu_check_banks(); + __mcheck_cpu_setup_timer(); +} + +/* + * Called for each booted CPU to clear some machine checks opt-ins + */ +void mcheck_cpu_clear(struct cpuinfo_x86 *c) +{ + if (mca_cfg.disabled) + return; + + if (!mce_available(c)) + return; + + /* + * Possibly to clear general settings generic to x86 + * __mcheck_cpu_clear_generic(c); + */ + __mcheck_cpu_clear_vendor(c); + +} + +static void __mce_disable_bank(void *arg) +{ + int bank = *((int *)arg); + __clear_bit(bank, this_cpu_ptr(mce_poll_banks)); + cmci_disable_bank(bank); +} + +void mce_disable_bank(int bank) +{ + if (bank >= this_cpu_read(mce_num_banks)) { + pr_warn(FW_BUG + "Ignoring request to disable invalid MCA bank %d.\n", + bank); + return; + } + set_bit(bank, mce_banks_ce_disabled); + on_each_cpu(__mce_disable_bank, &bank, 1); +} + +/* + * mce=off Disables machine check + * mce=no_cmci Disables CMCI + * mce=no_lmce Disables LMCE + * mce=dont_log_ce Clears corrected events silently, no log created for CEs. + * mce=print_all Print all machine check logs to console + * mce=ignore_ce Disables polling and CMCI, corrected events are not cleared. + * mce=TOLERANCELEVEL[,monarchtimeout] (number, see above) + * monarchtimeout is how long to wait for other CPUs on machine + * check, or 0 to not wait + * mce=bootlog Log MCEs from before booting. Disabled by default on AMD Fam10h + and older. + * mce=nobootlog Don't log MCEs from before booting. + * mce=bios_cmci_threshold Don't program the CMCI threshold + * mce=recovery force enable copy_mc_fragile() + */ +static int __init mcheck_enable(char *str) +{ + struct mca_config *cfg = &mca_cfg; + + if (*str == 0) { + enable_p5_mce(); + return 1; + } + if (*str == '=') + str++; + if (!strcmp(str, "off")) + cfg->disabled = 1; + else if (!strcmp(str, "no_cmci")) + cfg->cmci_disabled = true; + else if (!strcmp(str, "no_lmce")) + cfg->lmce_disabled = 1; + else if (!strcmp(str, "dont_log_ce")) + cfg->dont_log_ce = true; + else if (!strcmp(str, "print_all")) + cfg->print_all = true; + else if (!strcmp(str, "ignore_ce")) + cfg->ignore_ce = true; + else if (!strcmp(str, "bootlog") || !strcmp(str, "nobootlog")) + cfg->bootlog = (str[0] == 'b'); + else if (!strcmp(str, "bios_cmci_threshold")) + cfg->bios_cmci_threshold = 1; + else if (!strcmp(str, "recovery")) + cfg->recovery = 1; + else if (isdigit(str[0])) { + if (get_option(&str, &cfg->tolerant) == 2) + get_option(&str, &(cfg->monarch_timeout)); + } else { + pr_info("mce argument %s ignored. Please use /sys\n", str); + return 0; + } + return 1; +} +__setup("mce", mcheck_enable); + +int __init mcheck_init(void) +{ + mcheck_intel_therm_init(); + mce_register_decode_chain(&early_nb); + mce_register_decode_chain(&mce_uc_nb); + mce_register_decode_chain(&mce_default_nb); + mcheck_vendor_init_severity(); + + INIT_WORK(&mce_work, mce_gen_pool_process); + init_irq_work(&mce_irq_work, mce_irq_work_cb); + + return 0; +} + +/* + * mce_syscore: PM support + */ + +/* + * Disable machine checks on suspend and shutdown. We can't really handle + * them later. + */ +static void mce_disable_error_reporting(void) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + int i; + + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + struct mce_bank *b = &mce_banks[i]; + + if (b->init) + wrmsrl(mca_msr_reg(i, MCA_CTL), 0); + } + return; +} + +static void vendor_disable_error_reporting(void) +{ + /* + * Don't clear on Intel or AMD or Hygon or Zhaoxin CPUs. Some of these + * MSRs are socket-wide. Disabling them for just a single offlined CPU + * is bad, since it will inhibit reporting for all shared resources on + * the socket like the last level cache (LLC), the integrated memory + * controller (iMC), etc. + */ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL || + boot_cpu_data.x86_vendor == X86_VENDOR_HYGON || + boot_cpu_data.x86_vendor == X86_VENDOR_AMD || + boot_cpu_data.x86_vendor == X86_VENDOR_ZHAOXIN) + return; + + mce_disable_error_reporting(); +} + +static int mce_syscore_suspend(void) +{ + vendor_disable_error_reporting(); + return 0; +} + +static void mce_syscore_shutdown(void) +{ + vendor_disable_error_reporting(); +} + +/* + * On resume clear all MCE state. Don't want to see leftovers from the BIOS. + * Only one CPU is active at this time, the others get re-added later using + * CPU hotplug: + */ +static void mce_syscore_resume(void) +{ + __mcheck_cpu_init_generic(); + __mcheck_cpu_init_vendor(raw_cpu_ptr(&cpu_info)); + __mcheck_cpu_init_clear_banks(); +} + +static struct syscore_ops mce_syscore_ops = { + .suspend = mce_syscore_suspend, + .shutdown = mce_syscore_shutdown, + .resume = mce_syscore_resume, +}; + +/* + * mce_device: Sysfs support + */ + +static void mce_cpu_restart(void *data) +{ + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + __mcheck_cpu_init_generic(); + __mcheck_cpu_init_clear_banks(); + __mcheck_cpu_init_timer(); +} + +/* Reinit MCEs after user configuration changes */ +static void mce_restart(void) +{ + mce_timer_delete_all(); + on_each_cpu(mce_cpu_restart, NULL, 1); + mce_schedule_work(); +} + +/* Toggle features for corrected errors */ +static void mce_disable_cmci(void *data) +{ + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + cmci_clear(); +} + +static void mce_enable_ce(void *all) +{ + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + cmci_reenable(); + cmci_recheck(); + if (all) + __mcheck_cpu_init_timer(); +} + +static struct bus_type mce_subsys = { + .name = "machinecheck", + .dev_name = "machinecheck", +}; + +DEFINE_PER_CPU(struct device *, mce_device); + +static inline struct mce_bank_dev *attr_to_bank(struct device_attribute *attr) +{ + return container_of(attr, struct mce_bank_dev, attr); +} + +static ssize_t show_bank(struct device *s, struct device_attribute *attr, + char *buf) +{ + u8 bank = attr_to_bank(attr)->bank; + struct mce_bank *b; + + if (bank >= per_cpu(mce_num_banks, s->id)) + return -EINVAL; + + b = &per_cpu(mce_banks_array, s->id)[bank]; + + if (!b->init) + return -ENODEV; + + return sprintf(buf, "%llx\n", b->ctl); +} + +static ssize_t set_bank(struct device *s, struct device_attribute *attr, + const char *buf, size_t size) +{ + u8 bank = attr_to_bank(attr)->bank; + struct mce_bank *b; + u64 new; + + if (kstrtou64(buf, 0, &new) < 0) + return -EINVAL; + + if (bank >= per_cpu(mce_num_banks, s->id)) + return -EINVAL; + + b = &per_cpu(mce_banks_array, s->id)[bank]; + + if (!b->init) + return -ENODEV; + + b->ctl = new; + mce_restart(); + + return size; +} + +static ssize_t set_ignore_ce(struct device *s, + struct device_attribute *attr, + const char *buf, size_t size) +{ + u64 new; + + if (kstrtou64(buf, 0, &new) < 0) + return -EINVAL; + + mutex_lock(&mce_sysfs_mutex); + if (mca_cfg.ignore_ce ^ !!new) { + if (new) { + /* disable ce features */ + mce_timer_delete_all(); + on_each_cpu(mce_disable_cmci, NULL, 1); + mca_cfg.ignore_ce = true; + } else { + /* enable ce features */ + mca_cfg.ignore_ce = false; + on_each_cpu(mce_enable_ce, (void *)1, 1); + } + } + mutex_unlock(&mce_sysfs_mutex); + + return size; +} + +static ssize_t set_cmci_disabled(struct device *s, + struct device_attribute *attr, + const char *buf, size_t size) +{ + u64 new; + + if (kstrtou64(buf, 0, &new) < 0) + return -EINVAL; + + mutex_lock(&mce_sysfs_mutex); + if (mca_cfg.cmci_disabled ^ !!new) { + if (new) { + /* disable cmci */ + on_each_cpu(mce_disable_cmci, NULL, 1); + mca_cfg.cmci_disabled = true; + } else { + /* enable cmci */ + mca_cfg.cmci_disabled = false; + on_each_cpu(mce_enable_ce, NULL, 1); + } + } + mutex_unlock(&mce_sysfs_mutex); + + return size; +} + +static ssize_t store_int_with_restart(struct device *s, + struct device_attribute *attr, + const char *buf, size_t size) +{ + unsigned long old_check_interval = check_interval; + ssize_t ret = device_store_ulong(s, attr, buf, size); + + if (check_interval == old_check_interval) + return ret; + + mutex_lock(&mce_sysfs_mutex); + mce_restart(); + mutex_unlock(&mce_sysfs_mutex); + + return ret; +} + +static DEVICE_INT_ATTR(tolerant, 0644, mca_cfg.tolerant); +static DEVICE_INT_ATTR(monarch_timeout, 0644, mca_cfg.monarch_timeout); +static DEVICE_BOOL_ATTR(dont_log_ce, 0644, mca_cfg.dont_log_ce); +static DEVICE_BOOL_ATTR(print_all, 0644, mca_cfg.print_all); + +static struct dev_ext_attribute dev_attr_check_interval = { + __ATTR(check_interval, 0644, device_show_int, store_int_with_restart), + &check_interval +}; + +static struct dev_ext_attribute dev_attr_ignore_ce = { + __ATTR(ignore_ce, 0644, device_show_bool, set_ignore_ce), + &mca_cfg.ignore_ce +}; + +static struct dev_ext_attribute dev_attr_cmci_disabled = { + __ATTR(cmci_disabled, 0644, device_show_bool, set_cmci_disabled), + &mca_cfg.cmci_disabled +}; + +static struct device_attribute *mce_device_attrs[] = { + &dev_attr_tolerant.attr, + &dev_attr_check_interval.attr, +#ifdef CONFIG_X86_MCELOG_LEGACY + &dev_attr_trigger, +#endif + &dev_attr_monarch_timeout.attr, + &dev_attr_dont_log_ce.attr, + &dev_attr_print_all.attr, + &dev_attr_ignore_ce.attr, + &dev_attr_cmci_disabled.attr, + NULL +}; + +static cpumask_var_t mce_device_initialized; + +static void mce_device_release(struct device *dev) +{ + kfree(dev); +} + +/* Per CPU device init. All of the CPUs still share the same bank device: */ +static int mce_device_create(unsigned int cpu) +{ + struct device *dev; + int err; + int i, j; + + if (!mce_available(&boot_cpu_data)) + return -EIO; + + dev = per_cpu(mce_device, cpu); + if (dev) + return 0; + + dev = kzalloc(sizeof(*dev), GFP_KERNEL); + if (!dev) + return -ENOMEM; + dev->id = cpu; + dev->bus = &mce_subsys; + dev->release = &mce_device_release; + + err = device_register(dev); + if (err) { + put_device(dev); + return err; + } + + for (i = 0; mce_device_attrs[i]; i++) { + err = device_create_file(dev, mce_device_attrs[i]); + if (err) + goto error; + } + for (j = 0; j < per_cpu(mce_num_banks, cpu); j++) { + err = device_create_file(dev, &mce_bank_devs[j].attr); + if (err) + goto error2; + } + cpumask_set_cpu(cpu, mce_device_initialized); + per_cpu(mce_device, cpu) = dev; + + return 0; +error2: + while (--j >= 0) + device_remove_file(dev, &mce_bank_devs[j].attr); +error: + while (--i >= 0) + device_remove_file(dev, mce_device_attrs[i]); + + device_unregister(dev); + + return err; +} + +static void mce_device_remove(unsigned int cpu) +{ + struct device *dev = per_cpu(mce_device, cpu); + int i; + + if (!cpumask_test_cpu(cpu, mce_device_initialized)) + return; + + for (i = 0; mce_device_attrs[i]; i++) + device_remove_file(dev, mce_device_attrs[i]); + + for (i = 0; i < per_cpu(mce_num_banks, cpu); i++) + device_remove_file(dev, &mce_bank_devs[i].attr); + + device_unregister(dev); + cpumask_clear_cpu(cpu, mce_device_initialized); + per_cpu(mce_device, cpu) = NULL; +} + +/* Make sure there are no machine checks on offlined CPUs. */ +static void mce_disable_cpu(void) +{ + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + + if (!cpuhp_tasks_frozen) + cmci_clear(); + + vendor_disable_error_reporting(); +} + +static void mce_reenable_cpu(void) +{ + struct mce_bank *mce_banks = this_cpu_ptr(mce_banks_array); + int i; + + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + + if (!cpuhp_tasks_frozen) + cmci_reenable(); + for (i = 0; i < this_cpu_read(mce_num_banks); i++) { + struct mce_bank *b = &mce_banks[i]; + + if (b->init) + wrmsrl(mca_msr_reg(i, MCA_CTL), b->ctl); + } +} + +static int mce_cpu_dead(unsigned int cpu) +{ + mce_intel_hcpu_update(cpu); + + /* intentionally ignoring frozen here */ + if (!cpuhp_tasks_frozen) + cmci_rediscover(); + return 0; +} + +static int mce_cpu_online(unsigned int cpu) +{ + struct timer_list *t = this_cpu_ptr(&mce_timer); + int ret; + + mce_device_create(cpu); + + ret = mce_threshold_create_device(cpu); + if (ret) { + mce_device_remove(cpu); + return ret; + } + mce_reenable_cpu(); + mce_start_timer(t); + return 0; +} + +static int mce_cpu_pre_down(unsigned int cpu) +{ + struct timer_list *t = this_cpu_ptr(&mce_timer); + + mce_disable_cpu(); + del_timer_sync(t); + mce_threshold_remove_device(cpu); + mce_device_remove(cpu); + return 0; +} + +static __init void mce_init_banks(void) +{ + int i; + + for (i = 0; i < MAX_NR_BANKS; i++) { + struct mce_bank_dev *b = &mce_bank_devs[i]; + struct device_attribute *a = &b->attr; + + b->bank = i; + + sysfs_attr_init(&a->attr); + a->attr.name = b->attrname; + snprintf(b->attrname, ATTR_LEN, "bank%d", i); + + a->attr.mode = 0644; + a->show = show_bank; + a->store = set_bank; + } +} + +/* + * When running on XEN, this initcall is ordered against the XEN mcelog + * initcall: + * + * device_initcall(xen_late_init_mcelog); + * device_initcall_sync(mcheck_init_device); + */ +static __init int mcheck_init_device(void) +{ + int err; + + /* + * Check if we have a spare virtual bit. This will only become + * a problem if/when we move beyond 5-level page tables. + */ + MAYBE_BUILD_BUG_ON(__VIRTUAL_MASK_SHIFT >= 63); + + if (!mce_available(&boot_cpu_data)) { + err = -EIO; + goto err_out; + } + + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; + } + + mce_init_banks(); + + err = subsys_system_register(&mce_subsys, NULL); + if (err) + goto err_out_mem; + + err = cpuhp_setup_state(CPUHP_X86_MCE_DEAD, "x86/mce:dead", NULL, + mce_cpu_dead); + if (err) + goto err_out_mem; + + /* + * Invokes mce_cpu_online() on all CPUs which are online when + * the state is installed. + */ + err = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "x86/mce:online", + mce_cpu_online, mce_cpu_pre_down); + if (err < 0) + goto err_out_online; + + register_syscore_ops(&mce_syscore_ops); + + return 0; + +err_out_online: + cpuhp_remove_state(CPUHP_X86_MCE_DEAD); + +err_out_mem: + free_cpumask_var(mce_device_initialized); + +err_out: + pr_err("Unable to init MCE device (rc: %d)\n", err); + + return err; +} +device_initcall_sync(mcheck_init_device); + +/* + * Old style boot options parsing. Only for compatibility. + */ +static int __init mcheck_disable(char *str) +{ + mca_cfg.disabled = 1; + return 1; +} +__setup("nomce", mcheck_disable); + +#ifdef CONFIG_DEBUG_FS +struct dentry *mce_get_debugfs_dir(void) +{ + static struct dentry *dmce; + + if (!dmce) + dmce = debugfs_create_dir("mce", NULL); + + return dmce; +} + +static void mce_reset(void) +{ + cpu_missing = 0; + atomic_set(&mce_fake_panicked, 0); + atomic_set(&mce_executing, 0); + atomic_set(&mce_callin, 0); + atomic_set(&global_nwo, 0); +} + +static int fake_panic_get(void *data, u64 *val) +{ + *val = fake_panic; + return 0; +} + +static int fake_panic_set(void *data, u64 val) +{ + mce_reset(); + fake_panic = val; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(fake_panic_fops, fake_panic_get, fake_panic_set, + "%llu\n"); + +static void __init mcheck_debugfs_init(void) +{ + struct dentry *dmce; + + dmce = mce_get_debugfs_dir(); + debugfs_create_file_unsafe("fake_panic", 0444, dmce, NULL, + &fake_panic_fops); +} +#else +static void __init mcheck_debugfs_init(void) { } +#endif + +static int __init mcheck_late_init(void) +{ + if (mca_cfg.recovery) + enable_copy_mc_fragile(); + + mcheck_debugfs_init(); + + /* + * Flush out everything that has been logged during early boot, now that + * everything has been initialized (workqueues, decoders, ...). + */ + mce_schedule_work(); + + return 0; +} +late_initcall(mcheck_late_init); diff --git a/arch/x86/kernel/cpu/mce/dev-mcelog.c b/arch/x86/kernel/cpu/mce/dev-mcelog.c new file mode 100644 index 000000000..100fbeebd --- /dev/null +++ b/arch/x86/kernel/cpu/mce/dev-mcelog.c @@ -0,0 +1,374 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * /dev/mcelog driver + * + * K8 parts Copyright 2002,2003 Andi Kleen, SuSE Labs. + * Rest from unknown author(s). + * 2004 Andi Kleen. Rewrote most of it. + * Copyright 2008 Intel Corporation + * Author: Andi Kleen + */ + +#include <linux/miscdevice.h> +#include <linux/slab.h> +#include <linux/kmod.h> +#include <linux/poll.h> + +#include "internal.h" + +static BLOCKING_NOTIFIER_HEAD(mce_injector_chain); + +static DEFINE_MUTEX(mce_chrdev_read_mutex); + +static char mce_helper[128]; +static char *mce_helper_argv[2] = { mce_helper, NULL }; + +/* + * Lockless MCE logging infrastructure. + * This avoids deadlocks on printk locks without having to break locks. Also + * separate MCEs from kernel messages to avoid bogus bug reports. + */ + +static struct mce_log_buffer *mcelog; + +static DECLARE_WAIT_QUEUE_HEAD(mce_chrdev_wait); + +static int dev_mce_log(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct mce *mce = (struct mce *)data; + unsigned int entry; + + if (mce->kflags & MCE_HANDLED_CEC) + return NOTIFY_DONE; + + mutex_lock(&mce_chrdev_read_mutex); + + entry = mcelog->next; + + /* + * When the buffer fills up discard new entries. Assume that the + * earlier errors are the more interesting ones: + */ + if (entry >= mcelog->len) { + set_bit(MCE_OVERFLOW, (unsigned long *)&mcelog->flags); + goto unlock; + } + + mcelog->next = entry + 1; + + memcpy(mcelog->entry + entry, mce, sizeof(struct mce)); + mcelog->entry[entry].finished = 1; + mcelog->entry[entry].kflags = 0; + + /* wake processes polling /dev/mcelog */ + wake_up_interruptible(&mce_chrdev_wait); + +unlock: + mutex_unlock(&mce_chrdev_read_mutex); + + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) + mce->kflags |= MCE_HANDLED_MCELOG; + + return NOTIFY_OK; +} + +static struct notifier_block dev_mcelog_nb = { + .notifier_call = dev_mce_log, + .priority = MCE_PRIO_MCELOG, +}; + +static void mce_do_trigger(struct work_struct *work) +{ + call_usermodehelper(mce_helper, mce_helper_argv, NULL, UMH_NO_WAIT); +} + +static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + + +void mce_work_trigger(void) +{ + if (mce_helper[0]) + schedule_work(&mce_trigger_work); +} + +static ssize_t +show_trigger(struct device *s, struct device_attribute *attr, char *buf) +{ + strcpy(buf, mce_helper); + strcat(buf, "\n"); + return strlen(mce_helper) + 1; +} + +static ssize_t set_trigger(struct device *s, struct device_attribute *attr, + const char *buf, size_t siz) +{ + char *p; + + strncpy(mce_helper, buf, sizeof(mce_helper)); + mce_helper[sizeof(mce_helper)-1] = 0; + p = strchr(mce_helper, '\n'); + + if (p) + *p = 0; + + return strlen(mce_helper) + !!p; +} + +DEVICE_ATTR(trigger, 0644, show_trigger, set_trigger); + +/* + * mce_chrdev: Character device /dev/mcelog to read and clear the MCE log. + */ + +static DEFINE_SPINLOCK(mce_chrdev_state_lock); +static int mce_chrdev_open_count; /* #times opened */ +static int mce_chrdev_open_exclu; /* already open exclusive? */ + +static int mce_chrdev_open(struct inode *inode, struct file *file) +{ + spin_lock(&mce_chrdev_state_lock); + + if (mce_chrdev_open_exclu || + (mce_chrdev_open_count && (file->f_flags & O_EXCL))) { + spin_unlock(&mce_chrdev_state_lock); + + return -EBUSY; + } + + if (file->f_flags & O_EXCL) + mce_chrdev_open_exclu = 1; + mce_chrdev_open_count++; + + spin_unlock(&mce_chrdev_state_lock); + + return nonseekable_open(inode, file); +} + +static int mce_chrdev_release(struct inode *inode, struct file *file) +{ + spin_lock(&mce_chrdev_state_lock); + + mce_chrdev_open_count--; + mce_chrdev_open_exclu = 0; + + spin_unlock(&mce_chrdev_state_lock); + + return 0; +} + +static int mce_apei_read_done; + +/* Collect MCE record of previous boot in persistent storage via APEI ERST. */ +static int __mce_read_apei(char __user **ubuf, size_t usize) +{ + int rc; + u64 record_id; + struct mce m; + + if (usize < sizeof(struct mce)) + return -EINVAL; + + rc = apei_read_mce(&m, &record_id); + /* Error or no more MCE record */ + if (rc <= 0) { + mce_apei_read_done = 1; + /* + * When ERST is disabled, mce_chrdev_read() should return + * "no record" instead of "no device." + */ + if (rc == -ENODEV) + return 0; + return rc; + } + rc = -EFAULT; + if (copy_to_user(*ubuf, &m, sizeof(struct mce))) + return rc; + /* + * In fact, we should have cleared the record after that has + * been flushed to the disk or sent to network in + * /sbin/mcelog, but we have no interface to support that now, + * so just clear it to avoid duplication. + */ + rc = apei_clear_mce(record_id); + if (rc) { + mce_apei_read_done = 1; + return rc; + } + *ubuf += sizeof(struct mce); + + return 0; +} + +static ssize_t mce_chrdev_read(struct file *filp, char __user *ubuf, + size_t usize, loff_t *off) +{ + char __user *buf = ubuf; + unsigned next; + int i, err; + + mutex_lock(&mce_chrdev_read_mutex); + + if (!mce_apei_read_done) { + err = __mce_read_apei(&buf, usize); + if (err || buf != ubuf) + goto out; + } + + /* Only supports full reads right now */ + err = -EINVAL; + if (*off != 0 || usize < mcelog->len * sizeof(struct mce)) + goto out; + + next = mcelog->next; + err = 0; + + for (i = 0; i < next; i++) { + struct mce *m = &mcelog->entry[i]; + + err |= copy_to_user(buf, m, sizeof(*m)); + buf += sizeof(*m); + } + + memset(mcelog->entry, 0, next * sizeof(struct mce)); + mcelog->next = 0; + + if (err) + err = -EFAULT; + +out: + mutex_unlock(&mce_chrdev_read_mutex); + + return err ? err : buf - ubuf; +} + +static __poll_t mce_chrdev_poll(struct file *file, poll_table *wait) +{ + poll_wait(file, &mce_chrdev_wait, wait); + if (READ_ONCE(mcelog->next)) + return EPOLLIN | EPOLLRDNORM; + if (!mce_apei_read_done && apei_check_mce()) + return EPOLLIN | EPOLLRDNORM; + return 0; +} + +static long mce_chrdev_ioctl(struct file *f, unsigned int cmd, + unsigned long arg) +{ + int __user *p = (int __user *)arg; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + switch (cmd) { + case MCE_GET_RECORD_LEN: + return put_user(sizeof(struct mce), p); + case MCE_GET_LOG_LEN: + return put_user(mcelog->len, p); + case MCE_GETCLEAR_FLAGS: { + unsigned flags; + + do { + flags = mcelog->flags; + } while (cmpxchg(&mcelog->flags, flags, 0) != flags); + + return put_user(flags, p); + } + default: + return -ENOTTY; + } +} + +void mce_register_injector_chain(struct notifier_block *nb) +{ + blocking_notifier_chain_register(&mce_injector_chain, nb); +} +EXPORT_SYMBOL_GPL(mce_register_injector_chain); + +void mce_unregister_injector_chain(struct notifier_block *nb) +{ + blocking_notifier_chain_unregister(&mce_injector_chain, nb); +} +EXPORT_SYMBOL_GPL(mce_unregister_injector_chain); + +static ssize_t mce_chrdev_write(struct file *filp, const char __user *ubuf, + size_t usize, loff_t *off) +{ + struct mce m; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + /* + * There are some cases where real MSR reads could slip + * through. + */ + if (!boot_cpu_has(X86_FEATURE_MCE) || !boot_cpu_has(X86_FEATURE_MCA)) + return -EIO; + + if ((unsigned long)usize > sizeof(struct mce)) + usize = sizeof(struct mce); + if (copy_from_user(&m, ubuf, usize)) + return -EFAULT; + + if (m.extcpu >= num_possible_cpus() || !cpu_online(m.extcpu)) + return -EINVAL; + + /* + * Need to give user space some time to set everything up, + * so do it a jiffie or two later everywhere. + */ + schedule_timeout(2); + + blocking_notifier_call_chain(&mce_injector_chain, 0, &m); + + return usize; +} + +static const struct file_operations mce_chrdev_ops = { + .open = mce_chrdev_open, + .release = mce_chrdev_release, + .read = mce_chrdev_read, + .write = mce_chrdev_write, + .poll = mce_chrdev_poll, + .unlocked_ioctl = mce_chrdev_ioctl, + .compat_ioctl = compat_ptr_ioctl, + .llseek = no_llseek, +}; + +static struct miscdevice mce_chrdev_device = { + MISC_MCELOG_MINOR, + "mcelog", + &mce_chrdev_ops, +}; + +static __init int dev_mcelog_init_device(void) +{ + int mce_log_len; + int err; + + mce_log_len = max(MCE_LOG_MIN_LEN, num_online_cpus()); + mcelog = kzalloc(struct_size(mcelog, entry, mce_log_len), GFP_KERNEL); + if (!mcelog) + return -ENOMEM; + + memcpy(mcelog->signature, MCE_LOG_SIGNATURE, sizeof(mcelog->signature)); + mcelog->len = mce_log_len; + mcelog->recordlen = sizeof(struct mce); + + /* register character device /dev/mcelog */ + err = misc_register(&mce_chrdev_device); + if (err) { + if (err == -EBUSY) + /* Xen dom0 might have registered the device already. */ + pr_info("Unable to init device /dev/mcelog, already registered"); + else + pr_err("Unable to init device /dev/mcelog (rc: %d)\n", err); + + kfree(mcelog); + return err; + } + + mce_register_decode_chain(&dev_mcelog_nb); + return 0; +} +device_initcall_sync(dev_mcelog_init_device); diff --git a/arch/x86/kernel/cpu/mce/genpool.c b/arch/x86/kernel/cpu/mce/genpool.c new file mode 100644 index 000000000..fbe8b61c3 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/genpool.c @@ -0,0 +1,147 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * MCE event pool management in MCE context + * + * Copyright (C) 2015 Intel Corp. + * Author: Chen, Gong <gong.chen@linux.intel.com> + */ +#include <linux/smp.h> +#include <linux/mm.h> +#include <linux/genalloc.h> +#include <linux/llist.h> +#include "internal.h" + +/* + * printk() is not safe in MCE context. This is a lock-less memory allocator + * used to save error information organized in a lock-less list. + * + * This memory pool is only to be used to save MCE records in MCE context. + * MCE events are rare, so a fixed size memory pool should be enough. Use + * 2 pages to save MCE events for now (~80 MCE records at most). + */ +#define MCE_POOLSZ (2 * PAGE_SIZE) + +static struct gen_pool *mce_evt_pool; +static LLIST_HEAD(mce_event_llist); +static char gen_pool_buf[MCE_POOLSZ]; + +/* + * Compare the record "t" with each of the records on list "l" to see if + * an equivalent one is present in the list. + */ +static bool is_duplicate_mce_record(struct mce_evt_llist *t, struct mce_evt_llist *l) +{ + struct mce_evt_llist *node; + struct mce *m1, *m2; + + m1 = &t->mce; + + llist_for_each_entry(node, &l->llnode, llnode) { + m2 = &node->mce; + + if (!mce_cmp(m1, m2)) + return true; + } + return false; +} + +/* + * The system has panicked - we'd like to peruse the list of MCE records + * that have been queued, but not seen by anyone yet. The list is in + * reverse time order, so we need to reverse it. While doing that we can + * also drop duplicate records (these were logged because some banks are + * shared between cores or by all threads on a socket). + */ +struct llist_node *mce_gen_pool_prepare_records(void) +{ + struct llist_node *head; + LLIST_HEAD(new_head); + struct mce_evt_llist *node, *t; + + head = llist_del_all(&mce_event_llist); + if (!head) + return NULL; + + /* squeeze out duplicates while reversing order */ + llist_for_each_entry_safe(node, t, head, llnode) { + if (!is_duplicate_mce_record(node, t)) + llist_add(&node->llnode, &new_head); + } + + return new_head.first; +} + +void mce_gen_pool_process(struct work_struct *__unused) +{ + struct llist_node *head; + struct mce_evt_llist *node, *tmp; + struct mce *mce; + + head = llist_del_all(&mce_event_llist); + if (!head) + return; + + head = llist_reverse_order(head); + llist_for_each_entry_safe(node, tmp, head, llnode) { + mce = &node->mce; + blocking_notifier_call_chain(&x86_mce_decoder_chain, 0, mce); + gen_pool_free(mce_evt_pool, (unsigned long)node, sizeof(*node)); + } +} + +bool mce_gen_pool_empty(void) +{ + return llist_empty(&mce_event_llist); +} + +int mce_gen_pool_add(struct mce *mce) +{ + struct mce_evt_llist *node; + + if (filter_mce(mce)) + return -EINVAL; + + if (!mce_evt_pool) + return -EINVAL; + + node = (void *)gen_pool_alloc(mce_evt_pool, sizeof(*node)); + if (!node) { + pr_warn_ratelimited("MCE records pool full!\n"); + return -ENOMEM; + } + + memcpy(&node->mce, mce, sizeof(*mce)); + llist_add(&node->llnode, &mce_event_llist); + + return 0; +} + +static int mce_gen_pool_create(void) +{ + struct gen_pool *tmpp; + int ret = -ENOMEM; + + tmpp = gen_pool_create(ilog2(sizeof(struct mce_evt_llist)), -1); + if (!tmpp) + goto out; + + ret = gen_pool_add(tmpp, (unsigned long)gen_pool_buf, MCE_POOLSZ, -1); + if (ret) { + gen_pool_destroy(tmpp); + goto out; + } + + mce_evt_pool = tmpp; + +out: + return ret; +} + +int mce_gen_pool_init(void) +{ + /* Just init mce_gen_pool once. */ + if (mce_evt_pool) + return 0; + + return mce_gen_pool_create(); +} diff --git a/arch/x86/kernel/cpu/mce/inject.c b/arch/x86/kernel/cpu/mce/inject.c new file mode 100644 index 000000000..e7808309d --- /dev/null +++ b/arch/x86/kernel/cpu/mce/inject.c @@ -0,0 +1,705 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Machine check injection support. + * Copyright 2008 Intel Corporation. + * + * Authors: + * Andi Kleen + * Ying Huang + * + * The AMD part (from mce_amd_inj.c): a simple MCE injection facility + * for testing different aspects of the RAS code. This driver should be + * built as module so that it can be loaded on production kernels for + * testing purposes. + * + * Copyright (c) 2010-17: Borislav Petkov <bp@alien8.de> + * Advanced Micro Devices Inc. + */ + +#include <linux/cpu.h> +#include <linux/debugfs.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/notifier.h> +#include <linux/pci.h> +#include <linux/uaccess.h> + +#include <asm/amd_nb.h> +#include <asm/apic.h> +#include <asm/irq_vectors.h> +#include <asm/mce.h> +#include <asm/nmi.h> +#include <asm/smp.h> + +#include "internal.h" + +/* + * Collect all the MCi_XXX settings + */ +static struct mce i_mce; +static struct dentry *dfs_inj; + +#define MAX_FLAG_OPT_SIZE 4 +#define NBCFG 0x44 + +enum injection_type { + SW_INJ = 0, /* SW injection, simply decode the error */ + HW_INJ, /* Trigger a #MC */ + DFR_INT_INJ, /* Trigger Deferred error interrupt */ + THR_INT_INJ, /* Trigger threshold interrupt */ + N_INJ_TYPES, +}; + +static const char * const flags_options[] = { + [SW_INJ] = "sw", + [HW_INJ] = "hw", + [DFR_INT_INJ] = "df", + [THR_INT_INJ] = "th", + NULL +}; + +/* Set default injection to SW_INJ */ +static enum injection_type inj_type = SW_INJ; + +#define MCE_INJECT_SET(reg) \ +static int inj_##reg##_set(void *data, u64 val) \ +{ \ + struct mce *m = (struct mce *)data; \ + \ + m->reg = val; \ + return 0; \ +} + +MCE_INJECT_SET(status); +MCE_INJECT_SET(misc); +MCE_INJECT_SET(addr); +MCE_INJECT_SET(synd); + +#define MCE_INJECT_GET(reg) \ +static int inj_##reg##_get(void *data, u64 *val) \ +{ \ + struct mce *m = (struct mce *)data; \ + \ + *val = m->reg; \ + return 0; \ +} + +MCE_INJECT_GET(status); +MCE_INJECT_GET(misc); +MCE_INJECT_GET(addr); +MCE_INJECT_GET(synd); + +DEFINE_SIMPLE_ATTRIBUTE(status_fops, inj_status_get, inj_status_set, "%llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(misc_fops, inj_misc_get, inj_misc_set, "%llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(addr_fops, inj_addr_get, inj_addr_set, "%llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(synd_fops, inj_synd_get, inj_synd_set, "%llx\n"); + +static void setup_inj_struct(struct mce *m) +{ + memset(m, 0, sizeof(struct mce)); + + m->cpuvendor = boot_cpu_data.x86_vendor; + m->time = ktime_get_real_seconds(); + m->cpuid = cpuid_eax(1); + m->microcode = boot_cpu_data.microcode; +} + +/* Update fake mce registers on current CPU. */ +static void inject_mce(struct mce *m) +{ + struct mce *i = &per_cpu(injectm, m->extcpu); + + /* Make sure no one reads partially written injectm */ + i->finished = 0; + mb(); + m->finished = 0; + /* First set the fields after finished */ + i->extcpu = m->extcpu; + mb(); + /* Now write record in order, finished last (except above) */ + memcpy(i, m, sizeof(struct mce)); + /* Finally activate it */ + mb(); + i->finished = 1; +} + +static void raise_poll(struct mce *m) +{ + unsigned long flags; + mce_banks_t b; + + memset(&b, 0xff, sizeof(mce_banks_t)); + local_irq_save(flags); + machine_check_poll(0, &b); + local_irq_restore(flags); + m->finished = 0; +} + +static void raise_exception(struct mce *m, struct pt_regs *pregs) +{ + struct pt_regs regs; + unsigned long flags; + + if (!pregs) { + memset(®s, 0, sizeof(struct pt_regs)); + regs.ip = m->ip; + regs.cs = m->cs; + pregs = ®s; + } + /* do_machine_check() expects interrupts disabled -- at least */ + local_irq_save(flags); + do_machine_check(pregs); + local_irq_restore(flags); + m->finished = 0; +} + +static cpumask_var_t mce_inject_cpumask; +static DEFINE_MUTEX(mce_inject_mutex); + +static int mce_raise_notify(unsigned int cmd, struct pt_regs *regs) +{ + int cpu = smp_processor_id(); + struct mce *m = this_cpu_ptr(&injectm); + if (!cpumask_test_cpu(cpu, mce_inject_cpumask)) + return NMI_DONE; + cpumask_clear_cpu(cpu, mce_inject_cpumask); + if (m->inject_flags & MCJ_EXCEPTION) + raise_exception(m, regs); + else if (m->status) + raise_poll(m); + return NMI_HANDLED; +} + +static void mce_irq_ipi(void *info) +{ + int cpu = smp_processor_id(); + struct mce *m = this_cpu_ptr(&injectm); + + if (cpumask_test_cpu(cpu, mce_inject_cpumask) && + m->inject_flags & MCJ_EXCEPTION) { + cpumask_clear_cpu(cpu, mce_inject_cpumask); + raise_exception(m, NULL); + } +} + +/* Inject mce on current CPU */ +static int raise_local(void) +{ + struct mce *m = this_cpu_ptr(&injectm); + int context = MCJ_CTX(m->inject_flags); + int ret = 0; + int cpu = m->extcpu; + + if (m->inject_flags & MCJ_EXCEPTION) { + pr_info("Triggering MCE exception on CPU %d\n", cpu); + switch (context) { + case MCJ_CTX_IRQ: + /* + * Could do more to fake interrupts like + * calling irq_enter, but the necessary + * machinery isn't exported currently. + */ + fallthrough; + case MCJ_CTX_PROCESS: + raise_exception(m, NULL); + break; + default: + pr_info("Invalid MCE context\n"); + ret = -EINVAL; + } + pr_info("MCE exception done on CPU %d\n", cpu); + } else if (m->status) { + pr_info("Starting machine check poll CPU %d\n", cpu); + raise_poll(m); + mce_notify_irq(); + pr_info("Machine check poll done on CPU %d\n", cpu); + } else + m->finished = 0; + + return ret; +} + +static void __maybe_unused raise_mce(struct mce *m) +{ + int context = MCJ_CTX(m->inject_flags); + + inject_mce(m); + + if (context == MCJ_CTX_RANDOM) + return; + + if (m->inject_flags & (MCJ_IRQ_BROADCAST | MCJ_NMI_BROADCAST)) { + unsigned long start; + int cpu; + + get_online_cpus(); + cpumask_copy(mce_inject_cpumask, cpu_online_mask); + cpumask_clear_cpu(get_cpu(), mce_inject_cpumask); + for_each_online_cpu(cpu) { + struct mce *mcpu = &per_cpu(injectm, cpu); + if (!mcpu->finished || + MCJ_CTX(mcpu->inject_flags) != MCJ_CTX_RANDOM) + cpumask_clear_cpu(cpu, mce_inject_cpumask); + } + if (!cpumask_empty(mce_inject_cpumask)) { + if (m->inject_flags & MCJ_IRQ_BROADCAST) { + /* + * don't wait because mce_irq_ipi is necessary + * to be sync with following raise_local + */ + preempt_disable(); + smp_call_function_many(mce_inject_cpumask, + mce_irq_ipi, NULL, 0); + preempt_enable(); + } else if (m->inject_flags & MCJ_NMI_BROADCAST) + apic->send_IPI_mask(mce_inject_cpumask, + NMI_VECTOR); + } + start = jiffies; + while (!cpumask_empty(mce_inject_cpumask)) { + if (!time_before(jiffies, start + 2*HZ)) { + pr_err("Timeout waiting for mce inject %lx\n", + *cpumask_bits(mce_inject_cpumask)); + break; + } + cpu_relax(); + } + raise_local(); + put_cpu(); + put_online_cpus(); + } else { + preempt_disable(); + raise_local(); + preempt_enable(); + } +} + +static int mce_inject_raise(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct mce *m = (struct mce *)data; + + if (!m) + return NOTIFY_DONE; + + mutex_lock(&mce_inject_mutex); + raise_mce(m); + mutex_unlock(&mce_inject_mutex); + + return NOTIFY_DONE; +} + +static struct notifier_block inject_nb = { + .notifier_call = mce_inject_raise, +}; + +/* + * Caller needs to be make sure this cpu doesn't disappear + * from under us, i.e.: get_cpu/put_cpu. + */ +static int toggle_hw_mce_inject(unsigned int cpu, bool enable) +{ + u32 l, h; + int err; + + err = rdmsr_on_cpu(cpu, MSR_K7_HWCR, &l, &h); + if (err) { + pr_err("%s: error reading HWCR\n", __func__); + return err; + } + + enable ? (l |= BIT(18)) : (l &= ~BIT(18)); + + err = wrmsr_on_cpu(cpu, MSR_K7_HWCR, l, h); + if (err) + pr_err("%s: error writing HWCR\n", __func__); + + return err; +} + +static int __set_inj(const char *buf) +{ + int i; + + for (i = 0; i < N_INJ_TYPES; i++) { + if (!strncmp(flags_options[i], buf, strlen(flags_options[i]))) { + inj_type = i; + return 0; + } + } + return -EINVAL; +} + +static ssize_t flags_read(struct file *filp, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char buf[MAX_FLAG_OPT_SIZE]; + int n; + + n = sprintf(buf, "%s\n", flags_options[inj_type]); + + return simple_read_from_buffer(ubuf, cnt, ppos, buf, n); +} + +static ssize_t flags_write(struct file *filp, const char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char buf[MAX_FLAG_OPT_SIZE], *__buf; + int err; + + if (!cnt || cnt > MAX_FLAG_OPT_SIZE) + return -EINVAL; + + if (copy_from_user(&buf, ubuf, cnt)) + return -EFAULT; + + buf[cnt - 1] = 0; + + /* strip whitespace */ + __buf = strstrip(buf); + + err = __set_inj(__buf); + if (err) { + pr_err("%s: Invalid flags value: %s\n", __func__, __buf); + return err; + } + + *ppos += cnt; + + return cnt; +} + +static const struct file_operations flags_fops = { + .read = flags_read, + .write = flags_write, + .llseek = generic_file_llseek, +}; + +/* + * On which CPU to inject? + */ +MCE_INJECT_GET(extcpu); + +static int inj_extcpu_set(void *data, u64 val) +{ + struct mce *m = (struct mce *)data; + + if (val >= nr_cpu_ids || !cpu_online(val)) { + pr_err("%s: Invalid CPU: %llu\n", __func__, val); + return -EINVAL; + } + m->extcpu = val; + return 0; +} + +DEFINE_SIMPLE_ATTRIBUTE(extcpu_fops, inj_extcpu_get, inj_extcpu_set, "%llu\n"); + +static void trigger_mce(void *info) +{ + asm volatile("int $18"); +} + +static void trigger_dfr_int(void *info) +{ + asm volatile("int %0" :: "i" (DEFERRED_ERROR_VECTOR)); +} + +static void trigger_thr_int(void *info) +{ + asm volatile("int %0" :: "i" (THRESHOLD_APIC_VECTOR)); +} + +static u32 get_nbc_for_node(int node_id) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + u32 cores_per_node; + + cores_per_node = (c->x86_max_cores * smp_num_siblings) / amd_get_nodes_per_socket(); + + return cores_per_node * node_id; +} + +static void toggle_nb_mca_mst_cpu(u16 nid) +{ + struct amd_northbridge *nb; + struct pci_dev *F3; + u32 val; + int err; + + nb = node_to_amd_nb(nid); + if (!nb) + return; + + F3 = nb->misc; + if (!F3) + return; + + err = pci_read_config_dword(F3, NBCFG, &val); + if (err) { + pr_err("%s: Error reading F%dx%03x.\n", + __func__, PCI_FUNC(F3->devfn), NBCFG); + return; + } + + if (val & BIT(27)) + return; + + pr_err("%s: Set D18F3x44[NbMcaToMstCpuEn] which BIOS hasn't done.\n", + __func__); + + val |= BIT(27); + err = pci_write_config_dword(F3, NBCFG, val); + if (err) + pr_err("%s: Error writing F%dx%03x.\n", + __func__, PCI_FUNC(F3->devfn), NBCFG); +} + +static void prepare_msrs(void *info) +{ + struct mce m = *(struct mce *)info; + u8 b = m.bank; + + wrmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus); + + if (boot_cpu_has(X86_FEATURE_SMCA)) { + if (m.inject_flags == DFR_INT_INJ) { + wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(b), m.status); + wrmsrl(MSR_AMD64_SMCA_MCx_DEADDR(b), m.addr); + } else { + wrmsrl(MSR_AMD64_SMCA_MCx_STATUS(b), m.status); + wrmsrl(MSR_AMD64_SMCA_MCx_ADDR(b), m.addr); + } + + wrmsrl(MSR_AMD64_SMCA_MCx_MISC(b), m.misc); + wrmsrl(MSR_AMD64_SMCA_MCx_SYND(b), m.synd); + } else { + wrmsrl(MSR_IA32_MCx_STATUS(b), m.status); + wrmsrl(MSR_IA32_MCx_ADDR(b), m.addr); + wrmsrl(MSR_IA32_MCx_MISC(b), m.misc); + } +} + +static void do_inject(void) +{ + u64 mcg_status = 0; + unsigned int cpu = i_mce.extcpu; + u8 b = i_mce.bank; + + i_mce.tsc = rdtsc_ordered(); + + if (i_mce.misc) + i_mce.status |= MCI_STATUS_MISCV; + + if (i_mce.synd) + i_mce.status |= MCI_STATUS_SYNDV; + + if (inj_type == SW_INJ) { + mce_log(&i_mce); + return; + } + + /* prep MCE global settings for the injection */ + mcg_status = MCG_STATUS_MCIP | MCG_STATUS_EIPV; + + if (!(i_mce.status & MCI_STATUS_PCC)) + mcg_status |= MCG_STATUS_RIPV; + + /* + * Ensure necessary status bits for deferred errors: + * - MCx_STATUS[Deferred]: make sure it is a deferred error + * - MCx_STATUS[UC] cleared: deferred errors are _not_ UC + */ + if (inj_type == DFR_INT_INJ) { + i_mce.status |= MCI_STATUS_DEFERRED; + i_mce.status &= ~MCI_STATUS_UC; + } + + /* + * For multi node CPUs, logging and reporting of bank 4 errors happens + * only on the node base core. Refer to D18F3x44[NbMcaToMstCpuEn] for + * Fam10h and later BKDGs. + */ + if (boot_cpu_has(X86_FEATURE_AMD_DCM) && + b == 4 && + boot_cpu_data.x86 < 0x17) { + toggle_nb_mca_mst_cpu(amd_get_nb_id(cpu)); + cpu = get_nbc_for_node(amd_get_nb_id(cpu)); + } + + get_online_cpus(); + if (!cpu_online(cpu)) + goto err; + + toggle_hw_mce_inject(cpu, true); + + i_mce.mcgstatus = mcg_status; + i_mce.inject_flags = inj_type; + smp_call_function_single(cpu, prepare_msrs, &i_mce, 0); + + toggle_hw_mce_inject(cpu, false); + + switch (inj_type) { + case DFR_INT_INJ: + smp_call_function_single(cpu, trigger_dfr_int, NULL, 0); + break; + case THR_INT_INJ: + smp_call_function_single(cpu, trigger_thr_int, NULL, 0); + break; + default: + smp_call_function_single(cpu, trigger_mce, NULL, 0); + } + +err: + put_online_cpus(); + +} + +/* + * This denotes into which bank we're injecting and triggers + * the injection, at the same time. + */ +static int inj_bank_set(void *data, u64 val) +{ + struct mce *m = (struct mce *)data; + u8 n_banks; + u64 cap; + + /* Get bank count on target CPU so we can handle non-uniform values. */ + rdmsrl_on_cpu(m->extcpu, MSR_IA32_MCG_CAP, &cap); + n_banks = cap & MCG_BANKCNT_MASK; + + if (val >= n_banks) { + pr_err("MCA bank %llu non-existent on CPU%d\n", val, m->extcpu); + return -EINVAL; + } + + m->bank = val; + do_inject(); + + /* Reset injection struct */ + setup_inj_struct(&i_mce); + + return 0; +} + +MCE_INJECT_GET(bank); + +DEFINE_SIMPLE_ATTRIBUTE(bank_fops, inj_bank_get, inj_bank_set, "%llu\n"); + +static const char readme_msg[] = +"Description of the files and their usages:\n" +"\n" +"Note1: i refers to the bank number below.\n" +"Note2: See respective BKDGs for the exact bit definitions of the files below\n" +"as they mirror the hardware registers.\n" +"\n" +"status:\t Set MCi_STATUS: the bits in that MSR control the error type and\n" +"\t attributes of the error which caused the MCE.\n" +"\n" +"misc:\t Set MCi_MISC: provide auxiliary info about the error. It is mostly\n" +"\t used for error thresholding purposes and its validity is indicated by\n" +"\t MCi_STATUS[MiscV].\n" +"\n" +"synd:\t Set MCi_SYND: provide syndrome info about the error. Only valid on\n" +"\t Scalable MCA systems, and its validity is indicated by MCi_STATUS[SyndV].\n" +"\n" +"addr:\t Error address value to be written to MCi_ADDR. Log address information\n" +"\t associated with the error.\n" +"\n" +"cpu:\t The CPU to inject the error on.\n" +"\n" +"bank:\t Specify the bank you want to inject the error into: the number of\n" +"\t banks in a processor varies and is family/model-specific, therefore, the\n" +"\t supplied value is sanity-checked. Setting the bank value also triggers the\n" +"\t injection.\n" +"\n" +"flags:\t Injection type to be performed. Writing to this file will trigger a\n" +"\t real machine check, an APIC interrupt or invoke the error decoder routines\n" +"\t for AMD processors.\n" +"\n" +"\t Allowed error injection types:\n" +"\t - \"sw\": Software error injection. Decode error to a human-readable \n" +"\t format only. Safe to use.\n" +"\t - \"hw\": Hardware error injection. Causes the #MC exception handler to \n" +"\t handle the error. Be warned: might cause system panic if MCi_STATUS[PCC] \n" +"\t is set. Therefore, consider setting (debugfs_mountpoint)/mce/fake_panic \n" +"\t before injecting.\n" +"\t - \"df\": Trigger APIC interrupt for Deferred error. Causes deferred \n" +"\t error APIC interrupt handler to handle the error if the feature is \n" +"\t is present in hardware. \n" +"\t - \"th\": Trigger APIC interrupt for Threshold errors. Causes threshold \n" +"\t APIC interrupt handler to handle the error. \n" +"\n"; + +static ssize_t +inj_readme_read(struct file *filp, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + return simple_read_from_buffer(ubuf, cnt, ppos, + readme_msg, strlen(readme_msg)); +} + +static const struct file_operations readme_fops = { + .read = inj_readme_read, +}; + +static struct dfs_node { + char *name; + const struct file_operations *fops; + umode_t perm; +} dfs_fls[] = { + { .name = "status", .fops = &status_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "misc", .fops = &misc_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "addr", .fops = &addr_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "synd", .fops = &synd_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "bank", .fops = &bank_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "flags", .fops = &flags_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "cpu", .fops = &extcpu_fops, .perm = S_IRUSR | S_IWUSR }, + { .name = "README", .fops = &readme_fops, .perm = S_IRUSR | S_IRGRP | S_IROTH }, +}; + +static void __init debugfs_init(void) +{ + unsigned int i; + + dfs_inj = debugfs_create_dir("mce-inject", NULL); + + for (i = 0; i < ARRAY_SIZE(dfs_fls); i++) + debugfs_create_file(dfs_fls[i].name, dfs_fls[i].perm, dfs_inj, + &i_mce, dfs_fls[i].fops); +} + +static int __init inject_init(void) +{ + if (!alloc_cpumask_var(&mce_inject_cpumask, GFP_KERNEL)) + return -ENOMEM; + + debugfs_init(); + + register_nmi_handler(NMI_LOCAL, mce_raise_notify, 0, "mce_notify"); + mce_register_injector_chain(&inject_nb); + + setup_inj_struct(&i_mce); + + pr_info("Machine check injector initialized\n"); + + return 0; +} + +static void __exit inject_exit(void) +{ + + mce_unregister_injector_chain(&inject_nb); + unregister_nmi_handler(NMI_LOCAL, "mce_notify"); + + debugfs_remove_recursive(dfs_inj); + dfs_inj = NULL; + + memset(&dfs_fls, 0, sizeof(dfs_fls)); + + free_cpumask_var(mce_inject_cpumask); +} + +module_init(inject_init); +module_exit(inject_exit); +MODULE_LICENSE("GPL"); diff --git a/arch/x86/kernel/cpu/mce/intel.c b/arch/x86/kernel/cpu/mce/intel.c new file mode 100644 index 000000000..886d4648c --- /dev/null +++ b/arch/x86/kernel/cpu/mce/intel.c @@ -0,0 +1,543 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel specific MCE features. + * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca> + * Copyright (C) 2008, 2009 Intel Corporation + * Author: Andi Kleen + */ + +#include <linux/gfp.h> +#include <linux/interrupt.h> +#include <linux/percpu.h> +#include <linux/sched.h> +#include <linux/cpumask.h> +#include <asm/apic.h> +#include <asm/cpufeature.h> +#include <asm/intel-family.h> +#include <asm/processor.h> +#include <asm/msr.h> +#include <asm/mce.h> + +#include "internal.h" + +/* + * Support for Intel Correct Machine Check Interrupts. This allows + * the CPU to raise an interrupt when a corrected machine check happened. + * Normally we pick those up using a regular polling timer. + * Also supports reliable discovery of shared banks. + */ + +/* + * CMCI can be delivered to multiple cpus that share a machine check bank + * so we need to designate a single cpu to process errors logged in each bank + * in the interrupt handler (otherwise we would have many races and potential + * double reporting of the same error). + * Note that this can change when a cpu is offlined or brought online since + * some MCA banks are shared across cpus. When a cpu is offlined, cmci_clear() + * disables CMCI on all banks owned by the cpu and clears this bitfield. At + * this point, cmci_rediscover() kicks in and a different cpu may end up + * taking ownership of some of the shared MCA banks that were previously + * owned by the offlined cpu. + */ +static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); + +/* + * CMCI storm detection backoff counter + * + * During storm, we reset this counter to INITIAL_CHECK_INTERVAL in case we've + * encountered an error. If not, we decrement it by one. We signal the end of + * the CMCI storm when it reaches 0. + */ +static DEFINE_PER_CPU(int, cmci_backoff_cnt); + +/* + * cmci_discover_lock protects against parallel discovery attempts + * which could race against each other. + */ +static DEFINE_RAW_SPINLOCK(cmci_discover_lock); + +#define CMCI_THRESHOLD 1 +#define CMCI_POLL_INTERVAL (30 * HZ) +#define CMCI_STORM_INTERVAL (HZ) +#define CMCI_STORM_THRESHOLD 15 + +static DEFINE_PER_CPU(unsigned long, cmci_time_stamp); +static DEFINE_PER_CPU(unsigned int, cmci_storm_cnt); +static DEFINE_PER_CPU(unsigned int, cmci_storm_state); + +enum { + CMCI_STORM_NONE, + CMCI_STORM_ACTIVE, + CMCI_STORM_SUBSIDED, +}; + +static atomic_t cmci_storm_on_cpus; + +static int cmci_supported(int *banks) +{ + u64 cap; + + if (mca_cfg.cmci_disabled || mca_cfg.ignore_ce) + return 0; + + /* + * Vendor check is not strictly needed, but the initial + * initialization is vendor keyed and this + * makes sure none of the backdoors are entered otherwise. + */ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL && + boot_cpu_data.x86_vendor != X86_VENDOR_ZHAOXIN) + return 0; + + if (!boot_cpu_has(X86_FEATURE_APIC) || lapic_get_maxlvt() < 6) + return 0; + rdmsrl(MSR_IA32_MCG_CAP, cap); + *banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff); + return !!(cap & MCG_CMCI_P); +} + +static bool lmce_supported(void) +{ + u64 tmp; + + if (mca_cfg.lmce_disabled) + return false; + + rdmsrl(MSR_IA32_MCG_CAP, tmp); + + /* + * LMCE depends on recovery support in the processor. Hence both + * MCG_SER_P and MCG_LMCE_P should be present in MCG_CAP. + */ + if ((tmp & (MCG_SER_P | MCG_LMCE_P)) != + (MCG_SER_P | MCG_LMCE_P)) + return false; + + /* + * BIOS should indicate support for LMCE by setting bit 20 in + * IA32_FEAT_CTL without which touching MCG_EXT_CTL will generate a #GP + * fault. The MSR must also be locked for LMCE_ENABLED to take effect. + * WARN if the MSR isn't locked as init_ia32_feat_ctl() unconditionally + * locks the MSR in the event that it wasn't already locked by BIOS. + */ + rdmsrl(MSR_IA32_FEAT_CTL, tmp); + if (WARN_ON_ONCE(!(tmp & FEAT_CTL_LOCKED))) + return false; + + return tmp & FEAT_CTL_LMCE_ENABLED; +} + +bool mce_intel_cmci_poll(void) +{ + if (__this_cpu_read(cmci_storm_state) == CMCI_STORM_NONE) + return false; + + /* + * Reset the counter if we've logged an error in the last poll + * during the storm. + */ + if (machine_check_poll(0, this_cpu_ptr(&mce_banks_owned))) + this_cpu_write(cmci_backoff_cnt, INITIAL_CHECK_INTERVAL); + else + this_cpu_dec(cmci_backoff_cnt); + + return true; +} + +void mce_intel_hcpu_update(unsigned long cpu) +{ + if (per_cpu(cmci_storm_state, cpu) == CMCI_STORM_ACTIVE) + atomic_dec(&cmci_storm_on_cpus); + + per_cpu(cmci_storm_state, cpu) = CMCI_STORM_NONE; +} + +static void cmci_toggle_interrupt_mode(bool on) +{ + unsigned long flags, *owned; + int bank; + u64 val; + + raw_spin_lock_irqsave(&cmci_discover_lock, flags); + owned = this_cpu_ptr(mce_banks_owned); + for_each_set_bit(bank, owned, MAX_NR_BANKS) { + rdmsrl(MSR_IA32_MCx_CTL2(bank), val); + + if (on) + val |= MCI_CTL2_CMCI_EN; + else + val &= ~MCI_CTL2_CMCI_EN; + + wrmsrl(MSR_IA32_MCx_CTL2(bank), val); + } + raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + +unsigned long cmci_intel_adjust_timer(unsigned long interval) +{ + if ((this_cpu_read(cmci_backoff_cnt) > 0) && + (__this_cpu_read(cmci_storm_state) == CMCI_STORM_ACTIVE)) { + mce_notify_irq(); + return CMCI_STORM_INTERVAL; + } + + switch (__this_cpu_read(cmci_storm_state)) { + case CMCI_STORM_ACTIVE: + + /* + * We switch back to interrupt mode once the poll timer has + * silenced itself. That means no events recorded and the timer + * interval is back to our poll interval. + */ + __this_cpu_write(cmci_storm_state, CMCI_STORM_SUBSIDED); + if (!atomic_sub_return(1, &cmci_storm_on_cpus)) + pr_notice("CMCI storm subsided: switching to interrupt mode\n"); + + fallthrough; + + case CMCI_STORM_SUBSIDED: + /* + * We wait for all CPUs to go back to SUBSIDED state. When that + * happens we switch back to interrupt mode. + */ + if (!atomic_read(&cmci_storm_on_cpus)) { + __this_cpu_write(cmci_storm_state, CMCI_STORM_NONE); + cmci_toggle_interrupt_mode(true); + cmci_recheck(); + } + return CMCI_POLL_INTERVAL; + default: + + /* We have shiny weather. Let the poll do whatever it thinks. */ + return interval; + } +} + +static bool cmci_storm_detect(void) +{ + unsigned int cnt = __this_cpu_read(cmci_storm_cnt); + unsigned long ts = __this_cpu_read(cmci_time_stamp); + unsigned long now = jiffies; + int r; + + if (__this_cpu_read(cmci_storm_state) != CMCI_STORM_NONE) + return true; + + if (time_before_eq(now, ts + CMCI_STORM_INTERVAL)) { + cnt++; + } else { + cnt = 1; + __this_cpu_write(cmci_time_stamp, now); + } + __this_cpu_write(cmci_storm_cnt, cnt); + + if (cnt <= CMCI_STORM_THRESHOLD) + return false; + + cmci_toggle_interrupt_mode(false); + __this_cpu_write(cmci_storm_state, CMCI_STORM_ACTIVE); + r = atomic_add_return(1, &cmci_storm_on_cpus); + mce_timer_kick(CMCI_STORM_INTERVAL); + this_cpu_write(cmci_backoff_cnt, INITIAL_CHECK_INTERVAL); + + if (r == 1) + pr_notice("CMCI storm detected: switching to poll mode\n"); + return true; +} + +/* + * The interrupt handler. This is called on every event. + * Just call the poller directly to log any events. + * This could in theory increase the threshold under high load, + * but doesn't for now. + */ +static void intel_threshold_interrupt(void) +{ + if (cmci_storm_detect()) + return; + + machine_check_poll(MCP_TIMESTAMP, this_cpu_ptr(&mce_banks_owned)); +} + +/* + * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks + * on this CPU. Use the algorithm recommended in the SDM to discover shared + * banks. + */ +static void cmci_discover(int banks) +{ + unsigned long *owned = (void *)this_cpu_ptr(&mce_banks_owned); + unsigned long flags; + int i; + int bios_wrong_thresh = 0; + + raw_spin_lock_irqsave(&cmci_discover_lock, flags); + for (i = 0; i < banks; i++) { + u64 val; + int bios_zero_thresh = 0; + + if (test_bit(i, owned)) + continue; + + /* Skip banks in firmware first mode */ + if (test_bit(i, mce_banks_ce_disabled)) + continue; + + rdmsrl(MSR_IA32_MCx_CTL2(i), val); + + /* Already owned by someone else? */ + if (val & MCI_CTL2_CMCI_EN) { + clear_bit(i, owned); + __clear_bit(i, this_cpu_ptr(mce_poll_banks)); + continue; + } + + if (!mca_cfg.bios_cmci_threshold) { + val &= ~MCI_CTL2_CMCI_THRESHOLD_MASK; + val |= CMCI_THRESHOLD; + } else if (!(val & MCI_CTL2_CMCI_THRESHOLD_MASK)) { + /* + * If bios_cmci_threshold boot option was specified + * but the threshold is zero, we'll try to initialize + * it to 1. + */ + bios_zero_thresh = 1; + val |= CMCI_THRESHOLD; + } + + val |= MCI_CTL2_CMCI_EN; + wrmsrl(MSR_IA32_MCx_CTL2(i), val); + rdmsrl(MSR_IA32_MCx_CTL2(i), val); + + /* Did the enable bit stick? -- the bank supports CMCI */ + if (val & MCI_CTL2_CMCI_EN) { + set_bit(i, owned); + __clear_bit(i, this_cpu_ptr(mce_poll_banks)); + /* + * We are able to set thresholds for some banks that + * had a threshold of 0. This means the BIOS has not + * set the thresholds properly or does not work with + * this boot option. Note down now and report later. + */ + if (mca_cfg.bios_cmci_threshold && bios_zero_thresh && + (val & MCI_CTL2_CMCI_THRESHOLD_MASK)) + bios_wrong_thresh = 1; + } else { + WARN_ON(!test_bit(i, this_cpu_ptr(mce_poll_banks))); + } + } + raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); + if (mca_cfg.bios_cmci_threshold && bios_wrong_thresh) { + pr_info_once( + "bios_cmci_threshold: Some banks do not have valid thresholds set\n"); + pr_info_once( + "bios_cmci_threshold: Make sure your BIOS supports this boot option\n"); + } +} + +/* + * Just in case we missed an event during initialization check + * all the CMCI owned banks. + */ +void cmci_recheck(void) +{ + unsigned long flags; + int banks; + + if (!mce_available(raw_cpu_ptr(&cpu_info)) || !cmci_supported(&banks)) + return; + + local_irq_save(flags); + machine_check_poll(0, this_cpu_ptr(&mce_banks_owned)); + local_irq_restore(flags); +} + +/* Caller must hold the lock on cmci_discover_lock */ +static void __cmci_disable_bank(int bank) +{ + u64 val; + + if (!test_bit(bank, this_cpu_ptr(mce_banks_owned))) + return; + rdmsrl(MSR_IA32_MCx_CTL2(bank), val); + val &= ~MCI_CTL2_CMCI_EN; + wrmsrl(MSR_IA32_MCx_CTL2(bank), val); + __clear_bit(bank, this_cpu_ptr(mce_banks_owned)); +} + +/* + * Disable CMCI on this CPU for all banks it owns when it goes down. + * This allows other CPUs to claim the banks on rediscovery. + */ +void cmci_clear(void) +{ + unsigned long flags; + int i; + int banks; + + if (!cmci_supported(&banks)) + return; + raw_spin_lock_irqsave(&cmci_discover_lock, flags); + for (i = 0; i < banks; i++) + __cmci_disable_bank(i); + raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + +static void cmci_rediscover_work_func(void *arg) +{ + int banks; + + /* Recheck banks in case CPUs don't all have the same */ + if (cmci_supported(&banks)) + cmci_discover(banks); +} + +/* After a CPU went down cycle through all the others and rediscover */ +void cmci_rediscover(void) +{ + int banks; + + if (!cmci_supported(&banks)) + return; + + on_each_cpu(cmci_rediscover_work_func, NULL, 1); +} + +/* + * Reenable CMCI on this CPU in case a CPU down failed. + */ +void cmci_reenable(void) +{ + int banks; + if (cmci_supported(&banks)) + cmci_discover(banks); +} + +void cmci_disable_bank(int bank) +{ + int banks; + unsigned long flags; + + if (!cmci_supported(&banks)) + return; + + raw_spin_lock_irqsave(&cmci_discover_lock, flags); + __cmci_disable_bank(bank); + raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + +void intel_init_cmci(void) +{ + int banks; + + if (!cmci_supported(&banks)) + return; + + mce_threshold_vector = intel_threshold_interrupt; + cmci_discover(banks); + /* + * For CPU #0 this runs with still disabled APIC, but that's + * ok because only the vector is set up. We still do another + * check for the banks later for CPU #0 just to make sure + * to not miss any events. + */ + apic_write(APIC_LVTCMCI, THRESHOLD_APIC_VECTOR|APIC_DM_FIXED); + cmci_recheck(); +} + +void intel_init_lmce(void) +{ + u64 val; + + if (!lmce_supported()) + return; + + rdmsrl(MSR_IA32_MCG_EXT_CTL, val); + + if (!(val & MCG_EXT_CTL_LMCE_EN)) + wrmsrl(MSR_IA32_MCG_EXT_CTL, val | MCG_EXT_CTL_LMCE_EN); +} + +void intel_clear_lmce(void) +{ + u64 val; + + if (!lmce_supported()) + return; + + rdmsrl(MSR_IA32_MCG_EXT_CTL, val); + val &= ~MCG_EXT_CTL_LMCE_EN; + wrmsrl(MSR_IA32_MCG_EXT_CTL, val); +} + +static void intel_ppin_init(struct cpuinfo_x86 *c) +{ + unsigned long long val; + + /* + * Even if testing the presence of the MSR would be enough, we don't + * want to risk the situation where other models reuse this MSR for + * other purposes. + */ + switch (c->x86_model) { + case INTEL_FAM6_IVYBRIDGE_X: + case INTEL_FAM6_HASWELL_X: + case INTEL_FAM6_BROADWELL_D: + case INTEL_FAM6_BROADWELL_X: + case INTEL_FAM6_SKYLAKE_X: + case INTEL_FAM6_ICELAKE_X: + case INTEL_FAM6_ICELAKE_D: + case INTEL_FAM6_SAPPHIRERAPIDS_X: + case INTEL_FAM6_XEON_PHI_KNL: + case INTEL_FAM6_XEON_PHI_KNM: + + if (rdmsrl_safe(MSR_PPIN_CTL, &val)) + return; + + if ((val & 3UL) == 1UL) { + /* PPIN locked in disabled mode */ + return; + } + + /* If PPIN is disabled, try to enable */ + if (!(val & 2UL)) { + wrmsrl_safe(MSR_PPIN_CTL, val | 2UL); + rdmsrl_safe(MSR_PPIN_CTL, &val); + } + + /* Is the enable bit set? */ + if (val & 2UL) + set_cpu_cap(c, X86_FEATURE_INTEL_PPIN); + } +} + +void mce_intel_feature_init(struct cpuinfo_x86 *c) +{ + intel_init_thermal(c); + intel_init_cmci(); + intel_init_lmce(); + intel_ppin_init(c); +} + +void mce_intel_feature_clear(struct cpuinfo_x86 *c) +{ + intel_clear_lmce(); +} + +bool intel_filter_mce(struct mce *m) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + /* MCE errata HSD131, HSM142, HSW131, BDM48, HSM142 and SKX37 */ + if ((c->x86 == 6) && + ((c->x86_model == INTEL_FAM6_HASWELL) || + (c->x86_model == INTEL_FAM6_HASWELL_L) || + (c->x86_model == INTEL_FAM6_BROADWELL) || + (c->x86_model == INTEL_FAM6_HASWELL_G) || + (c->x86_model == INTEL_FAM6_SKYLAKE_X)) && + (m->bank == 0) && + ((m->status & 0xa0000000ffffffff) == 0x80000000000f0005)) + return true; + + return false; +} diff --git a/arch/x86/kernel/cpu/mce/internal.h b/arch/x86/kernel/cpu/mce/internal.h new file mode 100644 index 000000000..3a485c0d5 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/internal.h @@ -0,0 +1,199 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __X86_MCE_INTERNAL_H__ +#define __X86_MCE_INTERNAL_H__ + +#undef pr_fmt +#define pr_fmt(fmt) "mce: " fmt + +#include <linux/device.h> +#include <asm/mce.h> + +/* Pointer to the installed machine check handler for this CPU setup. */ +extern void (*machine_check_vector)(struct pt_regs *); + +enum severity_level { + MCE_NO_SEVERITY, + MCE_DEFERRED_SEVERITY, + MCE_UCNA_SEVERITY = MCE_DEFERRED_SEVERITY, + MCE_KEEP_SEVERITY, + MCE_SOME_SEVERITY, + MCE_AO_SEVERITY, + MCE_UC_SEVERITY, + MCE_AR_SEVERITY, + MCE_PANIC_SEVERITY, +}; + +extern struct blocking_notifier_head x86_mce_decoder_chain; + +#define INITIAL_CHECK_INTERVAL 5 * 60 /* 5 minutes */ + +struct mce_evt_llist { + struct llist_node llnode; + struct mce mce; +}; + +void mce_gen_pool_process(struct work_struct *__unused); +bool mce_gen_pool_empty(void); +int mce_gen_pool_add(struct mce *mce); +int mce_gen_pool_init(void); +struct llist_node *mce_gen_pool_prepare_records(void); + +extern int (*mce_severity)(struct mce *a, struct pt_regs *regs, + int tolerant, char **msg, bool is_excp); +struct dentry *mce_get_debugfs_dir(void); + +extern mce_banks_t mce_banks_ce_disabled; + +#ifdef CONFIG_X86_MCE_INTEL +unsigned long cmci_intel_adjust_timer(unsigned long interval); +bool mce_intel_cmci_poll(void); +void mce_intel_hcpu_update(unsigned long cpu); +void cmci_disable_bank(int bank); +void intel_init_cmci(void); +void intel_init_lmce(void); +void intel_clear_lmce(void); +bool intel_filter_mce(struct mce *m); +#else +# define cmci_intel_adjust_timer mce_adjust_timer_default +static inline bool mce_intel_cmci_poll(void) { return false; } +static inline void mce_intel_hcpu_update(unsigned long cpu) { } +static inline void cmci_disable_bank(int bank) { } +static inline void intel_init_cmci(void) { } +static inline void intel_init_lmce(void) { } +static inline void intel_clear_lmce(void) { } +static inline bool intel_filter_mce(struct mce *m) { return false; }; +#endif + +void mce_timer_kick(unsigned long interval); + +#ifdef CONFIG_ACPI_APEI +int apei_write_mce(struct mce *m); +ssize_t apei_read_mce(struct mce *m, u64 *record_id); +int apei_check_mce(void); +int apei_clear_mce(u64 record_id); +#else +static inline int apei_write_mce(struct mce *m) +{ + return -EINVAL; +} +static inline ssize_t apei_read_mce(struct mce *m, u64 *record_id) +{ + return 0; +} +static inline int apei_check_mce(void) +{ + return 0; +} +static inline int apei_clear_mce(u64 record_id) +{ + return -EINVAL; +} +#endif + +/* + * We consider records to be equivalent if bank+status+addr+misc all match. + * This is only used when the system is going down because of a fatal error + * to avoid cluttering the console log with essentially repeated information. + * In normal processing all errors seen are logged. + */ +static inline bool mce_cmp(struct mce *m1, struct mce *m2) +{ + return m1->bank != m2->bank || + m1->status != m2->status || + m1->addr != m2->addr || + m1->misc != m2->misc; +} + +extern struct device_attribute dev_attr_trigger; + +#ifdef CONFIG_X86_MCELOG_LEGACY +void mce_work_trigger(void); +void mce_register_injector_chain(struct notifier_block *nb); +void mce_unregister_injector_chain(struct notifier_block *nb); +#else +static inline void mce_work_trigger(void) { } +static inline void mce_register_injector_chain(struct notifier_block *nb) { } +static inline void mce_unregister_injector_chain(struct notifier_block *nb) { } +#endif + +struct mca_config { + bool dont_log_ce; + bool cmci_disabled; + bool ignore_ce; + bool print_all; + + __u64 lmce_disabled : 1, + disabled : 1, + ser : 1, + recovery : 1, + bios_cmci_threshold : 1, + __reserved : 59; + + s8 bootlog; + int tolerant; + int monarch_timeout; + int panic_timeout; + u32 rip_msr; +}; + +extern struct mca_config mca_cfg; +DECLARE_PER_CPU_READ_MOSTLY(unsigned int, mce_num_banks); + +struct mce_vendor_flags { + /* + * Indicates that overflow conditions are not fatal, when set. + */ + __u64 overflow_recov : 1, + + /* + * (AMD) SUCCOR stands for S/W UnCorrectable error COntainment and + * Recovery. It indicates support for data poisoning in HW and deferred + * error interrupts. + */ + succor : 1, + + /* + * (AMD) SMCA: This bit indicates support for Scalable MCA which expands + * the register space for each MCA bank and also increases number of + * banks. Also, to accommodate the new banks and registers, the MCA + * register space is moved to a new MSR range. + */ + smca : 1, + + /* AMD-style error thresholding banks present. */ + amd_threshold : 1, + + __reserved_0 : 60; +}; + +extern struct mce_vendor_flags mce_flags; + +enum mca_msr { + MCA_CTL, + MCA_STATUS, + MCA_ADDR, + MCA_MISC, +}; + +u32 mca_msr_reg(int bank, enum mca_msr reg); + +/* Decide whether to add MCE record to MCE event pool or filter it out. */ +extern bool filter_mce(struct mce *m); + +#ifdef CONFIG_X86_MCE_AMD +extern bool amd_filter_mce(struct mce *m); +#else +static inline bool amd_filter_mce(struct mce *m) { return false; }; +#endif + +__visible bool ex_handler_rdmsr_fault(const struct exception_table_entry *fixup, + struct pt_regs *regs, int trapnr, + unsigned long error_code, + unsigned long fault_addr); + +__visible bool ex_handler_wrmsr_fault(const struct exception_table_entry *fixup, + struct pt_regs *regs, int trapnr, + unsigned long error_code, + unsigned long fault_addr); + +#endif /* __X86_MCE_INTERNAL_H__ */ diff --git a/arch/x86/kernel/cpu/mce/p5.c b/arch/x86/kernel/cpu/mce/p5.c new file mode 100644 index 000000000..19e90cae8 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/p5.c @@ -0,0 +1,70 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * P5 specific Machine Check Exception Reporting + * (C) Copyright 2002 Alan Cox <alan@lxorguk.ukuu.org.uk> + */ +#include <linux/interrupt.h> +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/smp.h> +#include <linux/hardirq.h> + +#include <asm/processor.h> +#include <asm/traps.h> +#include <asm/tlbflush.h> +#include <asm/mce.h> +#include <asm/msr.h> + +#include "internal.h" + +/* By default disabled */ +int mce_p5_enabled __read_mostly; + +/* Machine check handler for Pentium class Intel CPUs: */ +static noinstr void pentium_machine_check(struct pt_regs *regs) +{ + u32 loaddr, hi, lotype; + + instrumentation_begin(); + rdmsr(MSR_IA32_P5_MC_ADDR, loaddr, hi); + rdmsr(MSR_IA32_P5_MC_TYPE, lotype, hi); + + pr_emerg("CPU#%d: Machine Check Exception: 0x%8X (type 0x%8X).\n", + smp_processor_id(), loaddr, lotype); + + if (lotype & (1<<5)) { + pr_emerg("CPU#%d: Possible thermal failure (CPU on fire ?).\n", + smp_processor_id()); + } + + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); + instrumentation_end(); +} + +/* Set up machine check reporting for processors with Intel style MCE: */ +void intel_p5_mcheck_init(struct cpuinfo_x86 *c) +{ + u32 l, h; + + /* Default P5 to off as its often misconnected: */ + if (!mce_p5_enabled) + return; + + /* Check for MCE support: */ + if (!cpu_has(c, X86_FEATURE_MCE)) + return; + + machine_check_vector = pentium_machine_check; + /* Make sure the vector pointer is visible before we enable MCEs: */ + wmb(); + + /* Read registers before enabling: */ + rdmsr(MSR_IA32_P5_MC_ADDR, l, h); + rdmsr(MSR_IA32_P5_MC_TYPE, l, h); + pr_info("Intel old style machine check architecture supported.\n"); + + /* Enable MCE: */ + cr4_set_bits(X86_CR4_MCE); + pr_info("Intel old style machine check reporting enabled on CPU#%d.\n", + smp_processor_id()); +} diff --git a/arch/x86/kernel/cpu/mce/severity.c b/arch/x86/kernel/cpu/mce/severity.c new file mode 100644 index 000000000..83df99131 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/severity.c @@ -0,0 +1,492 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * MCE grading rules. + * Copyright 2008, 2009 Intel Corporation. + * + * Author: Andi Kleen + */ +#include <linux/kernel.h> +#include <linux/seq_file.h> +#include <linux/init.h> +#include <linux/debugfs.h> +#include <linux/uaccess.h> + +#include <asm/mce.h> +#include <asm/intel-family.h> +#include <asm/traps.h> +#include <asm/insn.h> +#include <asm/insn-eval.h> + +#include "internal.h" + +/* + * Grade an mce by severity. In general the most severe ones are processed + * first. Since there are quite a lot of combinations test the bits in a + * table-driven way. The rules are simply processed in order, first + * match wins. + * + * Note this is only used for machine check exceptions, the corrected + * errors use much simpler rules. The exceptions still check for the corrected + * errors, but only to leave them alone for the CMCI handler (except for + * panic situations) + */ + +enum context { IN_KERNEL = 1, IN_USER = 2, IN_KERNEL_RECOV = 3 }; +enum ser { SER_REQUIRED = 1, NO_SER = 2 }; +enum exception { EXCP_CONTEXT = 1, NO_EXCP = 2 }; + +static struct severity { + u64 mask; + u64 result; + unsigned char sev; + unsigned char mcgmask; + unsigned char mcgres; + unsigned char ser; + unsigned char context; + unsigned char excp; + unsigned char covered; + unsigned char cpu_model; + unsigned char cpu_minstepping; + unsigned char bank_lo, bank_hi; + char *msg; +} severities[] = { +#define MCESEV(s, m, c...) { .sev = MCE_ ## s ## _SEVERITY, .msg = m, ## c } +#define BANK_RANGE(l, h) .bank_lo = l, .bank_hi = h +#define MODEL_STEPPING(m, s) .cpu_model = m, .cpu_minstepping = s +#define KERNEL .context = IN_KERNEL +#define USER .context = IN_USER +#define KERNEL_RECOV .context = IN_KERNEL_RECOV +#define SER .ser = SER_REQUIRED +#define NOSER .ser = NO_SER +#define EXCP .excp = EXCP_CONTEXT +#define NOEXCP .excp = NO_EXCP +#define BITCLR(x) .mask = x, .result = 0 +#define BITSET(x) .mask = x, .result = x +#define MCGMASK(x, y) .mcgmask = x, .mcgres = y +#define MASK(x, y) .mask = x, .result = y +#define MCI_UC_S (MCI_STATUS_UC|MCI_STATUS_S) +#define MCI_UC_AR (MCI_STATUS_UC|MCI_STATUS_AR) +#define MCI_UC_SAR (MCI_STATUS_UC|MCI_STATUS_S|MCI_STATUS_AR) +#define MCI_ADDR (MCI_STATUS_ADDRV|MCI_STATUS_MISCV) + + MCESEV( + NO, "Invalid", + BITCLR(MCI_STATUS_VAL) + ), + MCESEV( + NO, "Not enabled", + EXCP, BITCLR(MCI_STATUS_EN) + ), + MCESEV( + PANIC, "Processor context corrupt", + BITSET(MCI_STATUS_PCC) + ), + /* When MCIP is not set something is very confused */ + MCESEV( + PANIC, "MCIP not set in MCA handler", + EXCP, MCGMASK(MCG_STATUS_MCIP, 0) + ), + /* Neither return not error IP -- no chance to recover -> PANIC */ + MCESEV( + PANIC, "Neither restart nor error IP", + EXCP, MCGMASK(MCG_STATUS_RIPV|MCG_STATUS_EIPV, 0) + ), + MCESEV( + PANIC, "In kernel and no restart IP", + EXCP, KERNEL, MCGMASK(MCG_STATUS_RIPV, 0) + ), + MCESEV( + PANIC, "In kernel and no restart IP", + EXCP, KERNEL_RECOV, MCGMASK(MCG_STATUS_RIPV, 0) + ), + MCESEV( + KEEP, "Corrected error", + NOSER, BITCLR(MCI_STATUS_UC) + ), + /* + * known AO MCACODs reported via MCE or CMC: + * + * SRAO could be signaled either via a machine check exception or + * CMCI with the corresponding bit S 1 or 0. So we don't need to + * check bit S for SRAO. + */ + MCESEV( + AO, "Action optional: memory scrubbing error", + SER, MASK(MCI_UC_AR|MCACOD_SCRUBMSK, MCI_STATUS_UC|MCACOD_SCRUB) + ), + MCESEV( + AO, "Action optional: last level cache writeback error", + SER, MASK(MCI_UC_AR|MCACOD, MCI_STATUS_UC|MCACOD_L3WB) + ), + /* + * Quirk for Skylake/Cascade Lake. Patrol scrubber may be configured + * to report uncorrected errors using CMCI with a special signature. + * UC=0, MSCOD=0x0010, MCACOD=binary(000X 0000 1100 XXXX) reported + * in one of the memory controller banks. + * Set severity to "AO" for same action as normal patrol scrub error. + */ + MCESEV( + AO, "Uncorrected Patrol Scrub Error", + SER, MASK(MCI_STATUS_UC|MCI_ADDR|0xffffeff0, MCI_ADDR|0x001000c0), + MODEL_STEPPING(INTEL_FAM6_SKYLAKE_X, 4), BANK_RANGE(13, 18) + ), + + /* ignore OVER for UCNA */ + MCESEV( + UCNA, "Uncorrected no action required", + SER, MASK(MCI_UC_SAR, MCI_STATUS_UC) + ), + MCESEV( + PANIC, "Illegal combination (UCNA with AR=1)", + SER, + MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_STATUS_UC|MCI_STATUS_AR) + ), + MCESEV( + KEEP, "Non signalled machine check", + SER, BITCLR(MCI_STATUS_S) + ), + + MCESEV( + PANIC, "Action required with lost events", + SER, BITSET(MCI_STATUS_OVER|MCI_UC_SAR) + ), + + /* known AR MCACODs: */ +#ifdef CONFIG_MEMORY_FAILURE + MCESEV( + KEEP, "Action required but unaffected thread is continuable", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR, MCI_UC_SAR|MCI_ADDR), + MCGMASK(MCG_STATUS_RIPV|MCG_STATUS_EIPV, MCG_STATUS_RIPV) + ), + MCESEV( + AR, "Action required: data load in error recoverable area of kernel", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA), + KERNEL_RECOV + ), + MCESEV( + AR, "Action required: data load error in a user process", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA), + USER + ), + MCESEV( + AR, "Action required: instruction fetch error in a user process", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_INSTR), + USER + ), + MCESEV( + PANIC, "Data load in unrecoverable area of kernel", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA), + KERNEL + ), + MCESEV( + PANIC, "Instruction fetch error in kernel", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_INSTR), + KERNEL + ), +#endif + MCESEV( + PANIC, "Action required: unknown MCACOD", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_UC_SAR) + ), + + MCESEV( + SOME, "Action optional: unknown MCACOD", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_UC_S) + ), + MCESEV( + SOME, "Action optional with lost events", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_STATUS_OVER|MCI_UC_S) + ), + + MCESEV( + PANIC, "Overflowed uncorrected", + BITSET(MCI_STATUS_OVER|MCI_STATUS_UC) + ), + MCESEV( + UC, "Uncorrected", + BITSET(MCI_STATUS_UC) + ), + MCESEV( + SOME, "No match", + BITSET(0) + ) /* always matches. keep at end */ +}; + +#define mc_recoverable(mcg) (((mcg) & (MCG_STATUS_RIPV|MCG_STATUS_EIPV)) == \ + (MCG_STATUS_RIPV|MCG_STATUS_EIPV)) + +static bool is_copy_from_user(struct pt_regs *regs) +{ + u8 insn_buf[MAX_INSN_SIZE]; + struct insn insn; + unsigned long addr; + + if (copy_from_kernel_nofault(insn_buf, (void *)regs->ip, MAX_INSN_SIZE)) + return false; + + kernel_insn_init(&insn, insn_buf, MAX_INSN_SIZE); + insn_get_opcode(&insn); + if (!insn.opcode.got) + return false; + + switch (insn.opcode.value) { + /* MOV mem,reg */ + case 0x8A: case 0x8B: + /* MOVZ mem,reg */ + case 0xB60F: case 0xB70F: + insn_get_modrm(&insn); + insn_get_sib(&insn); + if (!insn.modrm.got || !insn.sib.got) + return false; + addr = (unsigned long)insn_get_addr_ref(&insn, regs); + break; + /* REP MOVS */ + case 0xA4: case 0xA5: + addr = regs->si; + break; + default: + return false; + } + + if (fault_in_kernel_space(addr)) + return false; + + current->mce_vaddr = (void __user *)addr; + + return true; +} + +/* + * If mcgstatus indicated that ip/cs on the stack were + * no good, then "m->cs" will be zero and we will have + * to assume the worst case (IN_KERNEL) as we actually + * have no idea what we were executing when the machine + * check hit. + * If we do have a good "m->cs" (or a faked one in the + * case we were executing in VM86 mode) we can use it to + * distinguish an exception taken in user from from one + * taken in the kernel. + */ +static int error_context(struct mce *m, struct pt_regs *regs) +{ + enum handler_type t; + + if ((m->cs & 3) == 3) + return IN_USER; + if (!mc_recoverable(m->mcgstatus)) + return IN_KERNEL; + + t = ex_get_fault_handler_type(m->ip); + if (t == EX_HANDLER_FAULT) { + m->kflags |= MCE_IN_KERNEL_RECOV; + return IN_KERNEL_RECOV; + } + if (t == EX_HANDLER_UACCESS && regs && is_copy_from_user(regs)) { + m->kflags |= MCE_IN_KERNEL_RECOV; + m->kflags |= MCE_IN_KERNEL_COPYIN; + return IN_KERNEL_RECOV; + } + + return IN_KERNEL; +} + +static int mce_severity_amd_smca(struct mce *m, enum context err_ctx) +{ + u32 addr = MSR_AMD64_SMCA_MCx_CONFIG(m->bank); + u32 low, high; + + /* + * We need to look at the following bits: + * - "succor" bit (data poisoning support), and + * - TCC bit (Task Context Corrupt) + * in MCi_STATUS to determine error severity. + */ + if (!mce_flags.succor) + return MCE_PANIC_SEVERITY; + + if (rdmsr_safe(addr, &low, &high)) + return MCE_PANIC_SEVERITY; + + /* TCC (Task context corrupt). If set and if IN_KERNEL, panic. */ + if ((low & MCI_CONFIG_MCAX) && + (m->status & MCI_STATUS_TCC) && + (err_ctx == IN_KERNEL)) + return MCE_PANIC_SEVERITY; + + /* ...otherwise invoke hwpoison handler. */ + return MCE_AR_SEVERITY; +} + +/* + * See AMD Error Scope Hierarchy table in a newer BKDG. For example + * 49125_15h_Models_30h-3Fh_BKDG.pdf, section "RAS Features" + */ +static int mce_severity_amd(struct mce *m, struct pt_regs *regs, int tolerant, + char **msg, bool is_excp) +{ + enum context ctx = error_context(m, regs); + + /* Processor Context Corrupt, no need to fumble too much, die! */ + if (m->status & MCI_STATUS_PCC) + return MCE_PANIC_SEVERITY; + + if (m->status & MCI_STATUS_UC) { + + if (ctx == IN_KERNEL) + return MCE_PANIC_SEVERITY; + + /* + * On older systems where overflow_recov flag is not present, we + * should simply panic if an error overflow occurs. If + * overflow_recov flag is present and set, then software can try + * to at least kill process to prolong system operation. + */ + if (mce_flags.overflow_recov) { + if (mce_flags.smca) + return mce_severity_amd_smca(m, ctx); + + /* kill current process */ + return MCE_AR_SEVERITY; + } else { + /* at least one error was not logged */ + if (m->status & MCI_STATUS_OVER) + return MCE_PANIC_SEVERITY; + } + + /* + * For any other case, return MCE_UC_SEVERITY so that we log the + * error and exit #MC handler. + */ + return MCE_UC_SEVERITY; + } + + /* + * deferred error: poll handler catches these and adds to mce_ring so + * memory-failure can take recovery actions. + */ + if (m->status & MCI_STATUS_DEFERRED) + return MCE_DEFERRED_SEVERITY; + + /* + * corrected error: poll handler catches these and passes responsibility + * of decoding the error to EDAC + */ + return MCE_KEEP_SEVERITY; +} + +static int mce_severity_intel(struct mce *m, struct pt_regs *regs, + int tolerant, char **msg, bool is_excp) +{ + enum exception excp = (is_excp ? EXCP_CONTEXT : NO_EXCP); + enum context ctx = error_context(m, regs); + struct severity *s; + + for (s = severities;; s++) { + if ((m->status & s->mask) != s->result) + continue; + if ((m->mcgstatus & s->mcgmask) != s->mcgres) + continue; + if (s->ser == SER_REQUIRED && !mca_cfg.ser) + continue; + if (s->ser == NO_SER && mca_cfg.ser) + continue; + if (s->context && ctx != s->context) + continue; + if (s->excp && excp != s->excp) + continue; + if (s->cpu_model && boot_cpu_data.x86_model != s->cpu_model) + continue; + if (s->cpu_minstepping && boot_cpu_data.x86_stepping < s->cpu_minstepping) + continue; + if (s->bank_lo && (m->bank < s->bank_lo || m->bank > s->bank_hi)) + continue; + if (msg) + *msg = s->msg; + s->covered = 1; + if (s->sev >= MCE_UC_SEVERITY && ctx == IN_KERNEL) { + if (tolerant < 1) + return MCE_PANIC_SEVERITY; + } + return s->sev; + } +} + +/* Default to mce_severity_intel */ +int (*mce_severity)(struct mce *m, struct pt_regs *regs, int tolerant, char **msg, bool is_excp) = + mce_severity_intel; + +void __init mcheck_vendor_init_severity(void) +{ + if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD || + boot_cpu_data.x86_vendor == X86_VENDOR_HYGON) + mce_severity = mce_severity_amd; +} + +#ifdef CONFIG_DEBUG_FS +static void *s_start(struct seq_file *f, loff_t *pos) +{ + if (*pos >= ARRAY_SIZE(severities)) + return NULL; + return &severities[*pos]; +} + +static void *s_next(struct seq_file *f, void *data, loff_t *pos) +{ + if (++(*pos) >= ARRAY_SIZE(severities)) + return NULL; + return &severities[*pos]; +} + +static void s_stop(struct seq_file *f, void *data) +{ +} + +static int s_show(struct seq_file *f, void *data) +{ + struct severity *ser = data; + seq_printf(f, "%d\t%s\n", ser->covered, ser->msg); + return 0; +} + +static const struct seq_operations severities_seq_ops = { + .start = s_start, + .next = s_next, + .stop = s_stop, + .show = s_show, +}; + +static int severities_coverage_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &severities_seq_ops); +} + +static ssize_t severities_coverage_write(struct file *file, + const char __user *ubuf, + size_t count, loff_t *ppos) +{ + int i; + for (i = 0; i < ARRAY_SIZE(severities); i++) + severities[i].covered = 0; + return count; +} + +static const struct file_operations severities_coverage_fops = { + .open = severities_coverage_open, + .release = seq_release, + .read = seq_read, + .write = severities_coverage_write, + .llseek = seq_lseek, +}; + +static int __init severities_debugfs_init(void) +{ + struct dentry *dmce; + + dmce = mce_get_debugfs_dir(); + + debugfs_create_file("severities-coverage", 0444, dmce, NULL, + &severities_coverage_fops); + return 0; +} +late_initcall(severities_debugfs_init); +#endif /* CONFIG_DEBUG_FS */ diff --git a/arch/x86/kernel/cpu/mce/therm_throt.c b/arch/x86/kernel/cpu/mce/therm_throt.c new file mode 100644 index 000000000..a7cd2d203 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/therm_throt.c @@ -0,0 +1,739 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Thermal throttle event support code (such as syslog messaging and rate + * limiting) that was factored out from x86_64 (mce_intel.c) and i386 (p4.c). + * + * This allows consistent reporting of CPU thermal throttle events. + * + * Maintains a counter in /sys that keeps track of the number of thermal + * events, such that the user knows how bad the thermal problem might be + * (since the logging to syslog is rate limited). + * + * Author: Dmitriy Zavin (dmitriyz@google.com) + * + * Credits: Adapted from Zwane Mwaikambo's original code in mce_intel.c. + * Inspired by Ross Biro's and Al Borchers' counter code. + */ +#include <linux/interrupt.h> +#include <linux/notifier.h> +#include <linux/jiffies.h> +#include <linux/kernel.h> +#include <linux/percpu.h> +#include <linux/export.h> +#include <linux/types.h> +#include <linux/init.h> +#include <linux/smp.h> +#include <linux/cpu.h> + +#include <asm/processor.h> +#include <asm/traps.h> +#include <asm/apic.h> +#include <asm/mce.h> +#include <asm/msr.h> +#include <asm/trace/irq_vectors.h> + +#include "internal.h" + +/* How long to wait between reporting thermal events */ +#define CHECK_INTERVAL (300 * HZ) + +#define THERMAL_THROTTLING_EVENT 0 +#define POWER_LIMIT_EVENT 1 + +/** + * struct _thermal_state - Represent the current thermal event state + * @next_check: Stores the next timestamp, when it is allowed + * to log the next warning message. + * @last_interrupt_time: Stores the timestamp for the last threshold + * high event. + * @therm_work: Delayed workqueue structure + * @count: Stores the current running count for thermal + * or power threshold interrupts. + * @last_count: Stores the previous running count for thermal + * or power threshold interrupts. + * @max_time_ms: This shows the maximum amount of time CPU was + * in throttled state for a single thermal + * threshold high to low state. + * @total_time_ms: This is a cumulative time during which CPU was + * in the throttled state. + * @rate_control_active: Set when a throttling message is logged. + * This is used for the purpose of rate-control. + * @new_event: Stores the last high/low status of the + * THERM_STATUS_PROCHOT or + * THERM_STATUS_POWER_LIMIT. + * @level: Stores whether this _thermal_state instance is + * for a CORE level or for PACKAGE level. + * @sample_index: Index for storing the next sample in the buffer + * temp_samples[]. + * @sample_count: Total number of samples collected in the buffer + * temp_samples[]. + * @average: The last moving average of temperature samples + * @baseline_temp: Temperature at which thermal threshold high + * interrupt was generated. + * @temp_samples: Storage for temperature samples to calculate + * moving average. + * + * This structure is used to represent data related to thermal state for a CPU. + * There is a separate storage for core and package level for each CPU. + */ +struct _thermal_state { + u64 next_check; + u64 last_interrupt_time; + struct delayed_work therm_work; + unsigned long count; + unsigned long last_count; + unsigned long max_time_ms; + unsigned long total_time_ms; + bool rate_control_active; + bool new_event; + u8 level; + u8 sample_index; + u8 sample_count; + u8 average; + u8 baseline_temp; + u8 temp_samples[3]; +}; + +struct thermal_state { + struct _thermal_state core_throttle; + struct _thermal_state core_power_limit; + struct _thermal_state package_throttle; + struct _thermal_state package_power_limit; + struct _thermal_state core_thresh0; + struct _thermal_state core_thresh1; + struct _thermal_state pkg_thresh0; + struct _thermal_state pkg_thresh1; +}; + +/* Callback to handle core threshold interrupts */ +int (*platform_thermal_notify)(__u64 msr_val); +EXPORT_SYMBOL(platform_thermal_notify); + +/* Callback to handle core package threshold_interrupts */ +int (*platform_thermal_package_notify)(__u64 msr_val); +EXPORT_SYMBOL_GPL(platform_thermal_package_notify); + +/* Callback support of rate control, return true, if + * callback has rate control */ +bool (*platform_thermal_package_rate_control)(void); +EXPORT_SYMBOL_GPL(platform_thermal_package_rate_control); + + +static DEFINE_PER_CPU(struct thermal_state, thermal_state); + +static atomic_t therm_throt_en = ATOMIC_INIT(0); + +static u32 lvtthmr_init __read_mostly; + +#ifdef CONFIG_SYSFS +#define define_therm_throt_device_one_ro(_name) \ + static DEVICE_ATTR(_name, 0444, \ + therm_throt_device_show_##_name, \ + NULL) \ + +#define define_therm_throt_device_show_func(event, name) \ + \ +static ssize_t therm_throt_device_show_##event##_##name( \ + struct device *dev, \ + struct device_attribute *attr, \ + char *buf) \ +{ \ + unsigned int cpu = dev->id; \ + ssize_t ret; \ + \ + preempt_disable(); /* CPU hotplug */ \ + if (cpu_online(cpu)) { \ + ret = sprintf(buf, "%lu\n", \ + per_cpu(thermal_state, cpu).event.name); \ + } else \ + ret = 0; \ + preempt_enable(); \ + \ + return ret; \ +} + +define_therm_throt_device_show_func(core_throttle, count); +define_therm_throt_device_one_ro(core_throttle_count); + +define_therm_throt_device_show_func(core_power_limit, count); +define_therm_throt_device_one_ro(core_power_limit_count); + +define_therm_throt_device_show_func(package_throttle, count); +define_therm_throt_device_one_ro(package_throttle_count); + +define_therm_throt_device_show_func(package_power_limit, count); +define_therm_throt_device_one_ro(package_power_limit_count); + +define_therm_throt_device_show_func(core_throttle, max_time_ms); +define_therm_throt_device_one_ro(core_throttle_max_time_ms); + +define_therm_throt_device_show_func(package_throttle, max_time_ms); +define_therm_throt_device_one_ro(package_throttle_max_time_ms); + +define_therm_throt_device_show_func(core_throttle, total_time_ms); +define_therm_throt_device_one_ro(core_throttle_total_time_ms); + +define_therm_throt_device_show_func(package_throttle, total_time_ms); +define_therm_throt_device_one_ro(package_throttle_total_time_ms); + +static struct attribute *thermal_throttle_attrs[] = { + &dev_attr_core_throttle_count.attr, + &dev_attr_core_throttle_max_time_ms.attr, + &dev_attr_core_throttle_total_time_ms.attr, + NULL +}; + +static const struct attribute_group thermal_attr_group = { + .attrs = thermal_throttle_attrs, + .name = "thermal_throttle" +}; +#endif /* CONFIG_SYSFS */ + +#define CORE_LEVEL 0 +#define PACKAGE_LEVEL 1 + +#define THERM_THROT_POLL_INTERVAL HZ +#define THERM_STATUS_PROCHOT_LOG BIT(1) + +#define THERM_STATUS_CLEAR_CORE_MASK (BIT(1) | BIT(3) | BIT(5) | BIT(7) | BIT(9) | BIT(11) | BIT(13) | BIT(15)) +#define THERM_STATUS_CLEAR_PKG_MASK (BIT(1) | BIT(3) | BIT(5) | BIT(7) | BIT(9) | BIT(11)) + +static void clear_therm_status_log(int level) +{ + int msr; + u64 mask, msr_val; + + if (level == CORE_LEVEL) { + msr = MSR_IA32_THERM_STATUS; + mask = THERM_STATUS_CLEAR_CORE_MASK; + } else { + msr = MSR_IA32_PACKAGE_THERM_STATUS; + mask = THERM_STATUS_CLEAR_PKG_MASK; + } + + rdmsrl(msr, msr_val); + msr_val &= mask; + wrmsrl(msr, msr_val & ~THERM_STATUS_PROCHOT_LOG); +} + +static void get_therm_status(int level, bool *proc_hot, u8 *temp) +{ + int msr; + u64 msr_val; + + if (level == CORE_LEVEL) + msr = MSR_IA32_THERM_STATUS; + else + msr = MSR_IA32_PACKAGE_THERM_STATUS; + + rdmsrl(msr, msr_val); + if (msr_val & THERM_STATUS_PROCHOT_LOG) + *proc_hot = true; + else + *proc_hot = false; + + *temp = (msr_val >> 16) & 0x7F; +} + +static void __maybe_unused throttle_active_work(struct work_struct *work) +{ + struct _thermal_state *state = container_of(to_delayed_work(work), + struct _thermal_state, therm_work); + unsigned int i, avg, this_cpu = smp_processor_id(); + u64 now = get_jiffies_64(); + bool hot; + u8 temp; + + get_therm_status(state->level, &hot, &temp); + /* temperature value is offset from the max so lesser means hotter */ + if (!hot && temp > state->baseline_temp) { + if (state->rate_control_active) + pr_info("CPU%d: %s temperature/speed normal (total events = %lu)\n", + this_cpu, + state->level == CORE_LEVEL ? "Core" : "Package", + state->count); + + state->rate_control_active = false; + return; + } + + if (time_before64(now, state->next_check) && + state->rate_control_active) + goto re_arm; + + state->next_check = now + CHECK_INTERVAL; + + if (state->count != state->last_count) { + /* There was one new thermal interrupt */ + state->last_count = state->count; + state->average = 0; + state->sample_count = 0; + state->sample_index = 0; + } + + state->temp_samples[state->sample_index] = temp; + state->sample_count++; + state->sample_index = (state->sample_index + 1) % ARRAY_SIZE(state->temp_samples); + if (state->sample_count < ARRAY_SIZE(state->temp_samples)) + goto re_arm; + + avg = 0; + for (i = 0; i < ARRAY_SIZE(state->temp_samples); ++i) + avg += state->temp_samples[i]; + + avg /= ARRAY_SIZE(state->temp_samples); + + if (state->average > avg) { + pr_warn("CPU%d: %s temperature is above threshold, cpu clock is throttled (total events = %lu)\n", + this_cpu, + state->level == CORE_LEVEL ? "Core" : "Package", + state->count); + state->rate_control_active = true; + } + + state->average = avg; + +re_arm: + clear_therm_status_log(state->level); + schedule_delayed_work_on(this_cpu, &state->therm_work, THERM_THROT_POLL_INTERVAL); +} + +/*** + * therm_throt_process - Process thermal throttling event from interrupt + * @curr: Whether the condition is current or not (boolean), since the + * thermal interrupt normally gets called both when the thermal + * event begins and once the event has ended. + * + * This function is called by the thermal interrupt after the + * IRQ has been acknowledged. + * + * It will take care of rate limiting and printing messages to the syslog. + */ +static void therm_throt_process(bool new_event, int event, int level) +{ + struct _thermal_state *state; + unsigned int this_cpu = smp_processor_id(); + bool old_event; + u64 now; + struct thermal_state *pstate = &per_cpu(thermal_state, this_cpu); + + now = get_jiffies_64(); + if (level == CORE_LEVEL) { + if (event == THERMAL_THROTTLING_EVENT) + state = &pstate->core_throttle; + else if (event == POWER_LIMIT_EVENT) + state = &pstate->core_power_limit; + else + return; + } else if (level == PACKAGE_LEVEL) { + if (event == THERMAL_THROTTLING_EVENT) + state = &pstate->package_throttle; + else if (event == POWER_LIMIT_EVENT) + state = &pstate->package_power_limit; + else + return; + } else + return; + + old_event = state->new_event; + state->new_event = new_event; + + if (new_event) + state->count++; + + if (event != THERMAL_THROTTLING_EVENT) + return; + + if (new_event && !state->last_interrupt_time) { + bool hot; + u8 temp; + + get_therm_status(state->level, &hot, &temp); + /* + * Ignore short temperature spike as the system is not close + * to PROCHOT. 10C offset is large enough to ignore. It is + * already dropped from the high threshold temperature. + */ + if (temp > 10) + return; + + state->baseline_temp = temp; + state->last_interrupt_time = now; + schedule_delayed_work_on(this_cpu, &state->therm_work, THERM_THROT_POLL_INTERVAL); + } else if (old_event && state->last_interrupt_time) { + unsigned long throttle_time; + + throttle_time = jiffies_delta_to_msecs(now - state->last_interrupt_time); + if (throttle_time > state->max_time_ms) + state->max_time_ms = throttle_time; + state->total_time_ms += throttle_time; + state->last_interrupt_time = 0; + } +} + +static int thresh_event_valid(int level, int event) +{ + struct _thermal_state *state; + unsigned int this_cpu = smp_processor_id(); + struct thermal_state *pstate = &per_cpu(thermal_state, this_cpu); + u64 now = get_jiffies_64(); + + if (level == PACKAGE_LEVEL) + state = (event == 0) ? &pstate->pkg_thresh0 : + &pstate->pkg_thresh1; + else + state = (event == 0) ? &pstate->core_thresh0 : + &pstate->core_thresh1; + + if (time_before64(now, state->next_check)) + return 0; + + state->next_check = now + CHECK_INTERVAL; + + return 1; +} + +static bool int_pln_enable; +static int __init int_pln_enable_setup(char *s) +{ + int_pln_enable = true; + + return 1; +} +__setup("int_pln_enable", int_pln_enable_setup); + +#ifdef CONFIG_SYSFS +/* Add/Remove thermal_throttle interface for CPU device: */ +static int thermal_throttle_add_dev(struct device *dev, unsigned int cpu) +{ + int err; + struct cpuinfo_x86 *c = &cpu_data(cpu); + + err = sysfs_create_group(&dev->kobj, &thermal_attr_group); + if (err) + return err; + + if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) { + err = sysfs_add_file_to_group(&dev->kobj, + &dev_attr_core_power_limit_count.attr, + thermal_attr_group.name); + if (err) + goto del_group; + } + + if (cpu_has(c, X86_FEATURE_PTS)) { + err = sysfs_add_file_to_group(&dev->kobj, + &dev_attr_package_throttle_count.attr, + thermal_attr_group.name); + if (err) + goto del_group; + + err = sysfs_add_file_to_group(&dev->kobj, + &dev_attr_package_throttle_max_time_ms.attr, + thermal_attr_group.name); + if (err) + goto del_group; + + err = sysfs_add_file_to_group(&dev->kobj, + &dev_attr_package_throttle_total_time_ms.attr, + thermal_attr_group.name); + if (err) + goto del_group; + + if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) { + err = sysfs_add_file_to_group(&dev->kobj, + &dev_attr_package_power_limit_count.attr, + thermal_attr_group.name); + if (err) + goto del_group; + } + } + + return 0; + +del_group: + sysfs_remove_group(&dev->kobj, &thermal_attr_group); + + return err; +} + +static void thermal_throttle_remove_dev(struct device *dev) +{ + sysfs_remove_group(&dev->kobj, &thermal_attr_group); +} + +/* Get notified when a cpu comes on/off. Be hotplug friendly. */ +static int thermal_throttle_online(unsigned int cpu) +{ + struct thermal_state *state = &per_cpu(thermal_state, cpu); + struct device *dev = get_cpu_device(cpu); + u32 l; + + state->package_throttle.level = PACKAGE_LEVEL; + state->core_throttle.level = CORE_LEVEL; + + INIT_DELAYED_WORK(&state->package_throttle.therm_work, throttle_active_work); + INIT_DELAYED_WORK(&state->core_throttle.therm_work, throttle_active_work); + + /* Unmask the thermal vector after the above workqueues are initialized. */ + l = apic_read(APIC_LVTTHMR); + apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED); + + return thermal_throttle_add_dev(dev, cpu); +} + +static int thermal_throttle_offline(unsigned int cpu) +{ + struct thermal_state *state = &per_cpu(thermal_state, cpu); + struct device *dev = get_cpu_device(cpu); + u32 l; + + /* Mask the thermal vector before draining evtl. pending work */ + l = apic_read(APIC_LVTTHMR); + apic_write(APIC_LVTTHMR, l | APIC_LVT_MASKED); + + cancel_delayed_work_sync(&state->package_throttle.therm_work); + cancel_delayed_work_sync(&state->core_throttle.therm_work); + + state->package_throttle.rate_control_active = false; + state->core_throttle.rate_control_active = false; + + thermal_throttle_remove_dev(dev); + return 0; +} + +static __init int thermal_throttle_init_device(void) +{ + int ret; + + if (!atomic_read(&therm_throt_en)) + return 0; + + ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "x86/therm:online", + thermal_throttle_online, + thermal_throttle_offline); + return ret < 0 ? ret : 0; +} +device_initcall(thermal_throttle_init_device); + +#endif /* CONFIG_SYSFS */ + +static void notify_package_thresholds(__u64 msr_val) +{ + bool notify_thres_0 = false; + bool notify_thres_1 = false; + + if (!platform_thermal_package_notify) + return; + + /* lower threshold check */ + if (msr_val & THERM_LOG_THRESHOLD0) + notify_thres_0 = true; + /* higher threshold check */ + if (msr_val & THERM_LOG_THRESHOLD1) + notify_thres_1 = true; + + if (!notify_thres_0 && !notify_thres_1) + return; + + if (platform_thermal_package_rate_control && + platform_thermal_package_rate_control()) { + /* Rate control is implemented in callback */ + platform_thermal_package_notify(msr_val); + return; + } + + /* lower threshold reached */ + if (notify_thres_0 && thresh_event_valid(PACKAGE_LEVEL, 0)) + platform_thermal_package_notify(msr_val); + /* higher threshold reached */ + if (notify_thres_1 && thresh_event_valid(PACKAGE_LEVEL, 1)) + platform_thermal_package_notify(msr_val); +} + +static void notify_thresholds(__u64 msr_val) +{ + /* check whether the interrupt handler is defined; + * otherwise simply return + */ + if (!platform_thermal_notify) + return; + + /* lower threshold reached */ + if ((msr_val & THERM_LOG_THRESHOLD0) && + thresh_event_valid(CORE_LEVEL, 0)) + platform_thermal_notify(msr_val); + /* higher threshold reached */ + if ((msr_val & THERM_LOG_THRESHOLD1) && + thresh_event_valid(CORE_LEVEL, 1)) + platform_thermal_notify(msr_val); +} + +/* Thermal transition interrupt handler */ +static void intel_thermal_interrupt(void) +{ + __u64 msr_val; + + if (static_cpu_has(X86_FEATURE_HWP)) + wrmsrl_safe(MSR_HWP_STATUS, 0); + + rdmsrl(MSR_IA32_THERM_STATUS, msr_val); + + /* Check for violation of core thermal thresholds*/ + notify_thresholds(msr_val); + + therm_throt_process(msr_val & THERM_STATUS_PROCHOT, + THERMAL_THROTTLING_EVENT, + CORE_LEVEL); + + if (this_cpu_has(X86_FEATURE_PLN) && int_pln_enable) + therm_throt_process(msr_val & THERM_STATUS_POWER_LIMIT, + POWER_LIMIT_EVENT, + CORE_LEVEL); + + if (this_cpu_has(X86_FEATURE_PTS)) { + rdmsrl(MSR_IA32_PACKAGE_THERM_STATUS, msr_val); + /* check violations of package thermal thresholds */ + notify_package_thresholds(msr_val); + therm_throt_process(msr_val & PACKAGE_THERM_STATUS_PROCHOT, + THERMAL_THROTTLING_EVENT, + PACKAGE_LEVEL); + if (this_cpu_has(X86_FEATURE_PLN) && int_pln_enable) + therm_throt_process(msr_val & + PACKAGE_THERM_STATUS_POWER_LIMIT, + POWER_LIMIT_EVENT, + PACKAGE_LEVEL); + } +} + +static void unexpected_thermal_interrupt(void) +{ + pr_err("CPU%d: Unexpected LVT thermal interrupt!\n", + smp_processor_id()); +} + +static void (*smp_thermal_vector)(void) = unexpected_thermal_interrupt; + +DEFINE_IDTENTRY_SYSVEC(sysvec_thermal) +{ + trace_thermal_apic_entry(THERMAL_APIC_VECTOR); + inc_irq_stat(irq_thermal_count); + smp_thermal_vector(); + trace_thermal_apic_exit(THERMAL_APIC_VECTOR); + ack_APIC_irq(); +} + +/* Thermal monitoring depends on APIC, ACPI and clock modulation */ +static int intel_thermal_supported(struct cpuinfo_x86 *c) +{ + if (!boot_cpu_has(X86_FEATURE_APIC)) + return 0; + if (!cpu_has(c, X86_FEATURE_ACPI) || !cpu_has(c, X86_FEATURE_ACC)) + return 0; + return 1; +} + +void __init mcheck_intel_therm_init(void) +{ + /* + * This function is only called on boot CPU. Save the init thermal + * LVT value on BSP and use that value to restore APs' thermal LVT + * entry BIOS programmed later + */ + if (intel_thermal_supported(&boot_cpu_data)) + lvtthmr_init = apic_read(APIC_LVTTHMR); +} + +void intel_init_thermal(struct cpuinfo_x86 *c) +{ + unsigned int cpu = smp_processor_id(); + int tm2 = 0; + u32 l, h; + + if (!intel_thermal_supported(c)) + return; + + /* + * First check if its enabled already, in which case there might + * be some SMM goo which handles it, so we can't even put a handler + * since it might be delivered via SMI already: + */ + rdmsr(MSR_IA32_MISC_ENABLE, l, h); + + h = lvtthmr_init; + /* + * The initial value of thermal LVT entries on all APs always reads + * 0x10000 because APs are woken up by BSP issuing INIT-SIPI-SIPI + * sequence to them and LVT registers are reset to 0s except for + * the mask bits which are set to 1s when APs receive INIT IPI. + * If BIOS takes over the thermal interrupt and sets its interrupt + * delivery mode to SMI (not fixed), it restores the value that the + * BIOS has programmed on AP based on BSP's info we saved since BIOS + * is always setting the same value for all threads/cores. + */ + if ((h & APIC_DM_FIXED_MASK) != APIC_DM_FIXED) + apic_write(APIC_LVTTHMR, lvtthmr_init); + + + if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) { + if (system_state == SYSTEM_BOOTING) + pr_debug("CPU%d: Thermal monitoring handled by SMI\n", cpu); + return; + } + + /* early Pentium M models use different method for enabling TM2 */ + if (cpu_has(c, X86_FEATURE_TM2)) { + if (c->x86 == 6 && (c->x86_model == 9 || c->x86_model == 13)) { + rdmsr(MSR_THERM2_CTL, l, h); + if (l & MSR_THERM2_CTL_TM_SELECT) + tm2 = 1; + } else if (l & MSR_IA32_MISC_ENABLE_TM2) + tm2 = 1; + } + + /* We'll mask the thermal vector in the lapic till we're ready: */ + h = THERMAL_APIC_VECTOR | APIC_DM_FIXED | APIC_LVT_MASKED; + apic_write(APIC_LVTTHMR, h); + + rdmsr(MSR_IA32_THERM_INTERRUPT, l, h); + if (cpu_has(c, X86_FEATURE_PLN) && !int_pln_enable) + wrmsr(MSR_IA32_THERM_INTERRUPT, + (l | (THERM_INT_LOW_ENABLE + | THERM_INT_HIGH_ENABLE)) & ~THERM_INT_PLN_ENABLE, h); + else if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) + wrmsr(MSR_IA32_THERM_INTERRUPT, + l | (THERM_INT_LOW_ENABLE + | THERM_INT_HIGH_ENABLE | THERM_INT_PLN_ENABLE), h); + else + wrmsr(MSR_IA32_THERM_INTERRUPT, + l | (THERM_INT_LOW_ENABLE | THERM_INT_HIGH_ENABLE), h); + + if (cpu_has(c, X86_FEATURE_PTS)) { + rdmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, l, h); + if (cpu_has(c, X86_FEATURE_PLN) && !int_pln_enable) + wrmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, + (l | (PACKAGE_THERM_INT_LOW_ENABLE + | PACKAGE_THERM_INT_HIGH_ENABLE)) + & ~PACKAGE_THERM_INT_PLN_ENABLE, h); + else if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) + wrmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, + l | (PACKAGE_THERM_INT_LOW_ENABLE + | PACKAGE_THERM_INT_HIGH_ENABLE + | PACKAGE_THERM_INT_PLN_ENABLE), h); + else + wrmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, + l | (PACKAGE_THERM_INT_LOW_ENABLE + | PACKAGE_THERM_INT_HIGH_ENABLE), h); + } + + smp_thermal_vector = intel_thermal_interrupt; + + rdmsr(MSR_IA32_MISC_ENABLE, l, h); + wrmsr(MSR_IA32_MISC_ENABLE, l | MSR_IA32_MISC_ENABLE_TM1, h); + + pr_info_once("CPU0: Thermal monitoring enabled (%s)\n", + tm2 ? "TM2" : "TM1"); + + /* enable thermal throttle processing */ + atomic_set(&therm_throt_en, 1); +} diff --git a/arch/x86/kernel/cpu/mce/threshold.c b/arch/x86/kernel/cpu/mce/threshold.c new file mode 100644 index 000000000..6a059a035 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/threshold.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Common corrected MCE threshold handler code: + */ +#include <linux/interrupt.h> +#include <linux/kernel.h> + +#include <asm/irq_vectors.h> +#include <asm/traps.h> +#include <asm/apic.h> +#include <asm/mce.h> +#include <asm/trace/irq_vectors.h> + +#include "internal.h" + +static void default_threshold_interrupt(void) +{ + pr_err("Unexpected threshold interrupt at vector %x\n", + THRESHOLD_APIC_VECTOR); +} + +void (*mce_threshold_vector)(void) = default_threshold_interrupt; + +DEFINE_IDTENTRY_SYSVEC(sysvec_threshold) +{ + trace_threshold_apic_entry(THRESHOLD_APIC_VECTOR); + inc_irq_stat(irq_threshold_count); + mce_threshold_vector(); + trace_threshold_apic_exit(THRESHOLD_APIC_VECTOR); + ack_APIC_irq(); +} diff --git a/arch/x86/kernel/cpu/mce/winchip.c b/arch/x86/kernel/cpu/mce/winchip.c new file mode 100644 index 000000000..9c9f0abd2 --- /dev/null +++ b/arch/x86/kernel/cpu/mce/winchip.c @@ -0,0 +1,45 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * IDT Winchip specific Machine Check Exception Reporting + * (C) Copyright 2002 Alan Cox <alan@lxorguk.ukuu.org.uk> + */ +#include <linux/interrupt.h> +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/hardirq.h> + +#include <asm/processor.h> +#include <asm/traps.h> +#include <asm/tlbflush.h> +#include <asm/mce.h> +#include <asm/msr.h> + +#include "internal.h" + +/* Machine check handler for WinChip C6: */ +static noinstr void winchip_machine_check(struct pt_regs *regs) +{ + instrumentation_begin(); + pr_emerg("CPU0: Machine Check Exception.\n"); + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); + instrumentation_end(); +} + +/* Set up machine check reporting on the Winchip C6 series */ +void winchip_mcheck_init(struct cpuinfo_x86 *c) +{ + u32 lo, hi; + + machine_check_vector = winchip_machine_check; + /* Make sure the vector pointer is visible before we enable MCEs: */ + wmb(); + + rdmsr(MSR_IDT_FCR1, lo, hi); + lo |= (1<<2); /* Enable EIERRINT (int 18 MCE) */ + lo &= ~(1<<4); /* Enable MCE */ + wrmsr(MSR_IDT_FCR1, lo, hi); + + cr4_set_bits(X86_CR4_MCE); + + pr_info("Winchip machine check reporting enabled on CPU#0.\n"); +} diff --git a/arch/x86/kernel/cpu/microcode/Makefile b/arch/x86/kernel/cpu/microcode/Makefile new file mode 100644 index 000000000..34098d48c --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0-only +microcode-y := core.o +obj-$(CONFIG_MICROCODE) += microcode.o +microcode-$(CONFIG_MICROCODE_INTEL) += intel.o +microcode-$(CONFIG_MICROCODE_AMD) += amd.o diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c new file mode 100644 index 000000000..936085d81 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/amd.c @@ -0,0 +1,974 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * AMD CPU Microcode Update Driver for Linux + * + * This driver allows to upgrade microcode on F10h AMD + * CPUs and later. + * + * Copyright (C) 2008-2011 Advanced Micro Devices Inc. + * 2013-2018 Borislav Petkov <bp@alien8.de> + * + * Author: Peter Oruba <peter.oruba@amd.com> + * + * Based on work by: + * Tigran Aivazian <aivazian.tigran@gmail.com> + * + * early loader: + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Jacob Shin <jacob.shin@amd.com> + * Fixes: Borislav Petkov <bp@suse.de> + */ +#define pr_fmt(fmt) "microcode: " fmt + +#include <linux/earlycpio.h> +#include <linux/firmware.h> +#include <linux/uaccess.h> +#include <linux/vmalloc.h> +#include <linux/initrd.h> +#include <linux/kernel.h> +#include <linux/pci.h> + +#include <asm/microcode_amd.h> +#include <asm/microcode.h> +#include <asm/processor.h> +#include <asm/setup.h> +#include <asm/cpu.h> +#include <asm/msr.h> + +static struct equiv_cpu_table { + unsigned int num_entries; + struct equiv_cpu_entry *entry; +} equiv_table; + +/* + * This points to the current valid container of microcode patches which we will + * save from the initrd/builtin before jettisoning its contents. @mc is the + * microcode patch we found to match. + */ +struct cont_desc { + struct microcode_amd *mc; + u32 cpuid_1_eax; + u32 psize; + u8 *data; + size_t size; +}; + +static u32 ucode_new_rev; + +/* One blob per node. */ +static u8 amd_ucode_patch[MAX_NUMNODES][PATCH_MAX_SIZE]; + +/* + * Microcode patch container file is prepended to the initrd in cpio + * format. See Documentation/x86/microcode.rst + */ +static const char +ucode_path[] __maybe_unused = "kernel/x86/microcode/AuthenticAMD.bin"; + +static u16 find_equiv_id(struct equiv_cpu_table *et, u32 sig) +{ + unsigned int i; + + if (!et || !et->num_entries) + return 0; + + for (i = 0; i < et->num_entries; i++) { + struct equiv_cpu_entry *e = &et->entry[i]; + + if (sig == e->installed_cpu) + return e->equiv_cpu; + + e++; + } + return 0; +} + +/* + * Check whether there is a valid microcode container file at the beginning + * of @buf of size @buf_size. Set @early to use this function in the early path. + */ +static bool verify_container(const u8 *buf, size_t buf_size, bool early) +{ + u32 cont_magic; + + if (buf_size <= CONTAINER_HDR_SZ) { + if (!early) + pr_debug("Truncated microcode container header.\n"); + + return false; + } + + cont_magic = *(const u32 *)buf; + if (cont_magic != UCODE_MAGIC) { + if (!early) + pr_debug("Invalid magic value (0x%08x).\n", cont_magic); + + return false; + } + + return true; +} + +/* + * Check whether there is a valid, non-truncated CPU equivalence table at the + * beginning of @buf of size @buf_size. Set @early to use this function in the + * early path. + */ +static bool verify_equivalence_table(const u8 *buf, size_t buf_size, bool early) +{ + const u32 *hdr = (const u32 *)buf; + u32 cont_type, equiv_tbl_len; + + if (!verify_container(buf, buf_size, early)) + return false; + + cont_type = hdr[1]; + if (cont_type != UCODE_EQUIV_CPU_TABLE_TYPE) { + if (!early) + pr_debug("Wrong microcode container equivalence table type: %u.\n", + cont_type); + + return false; + } + + buf_size -= CONTAINER_HDR_SZ; + + equiv_tbl_len = hdr[2]; + if (equiv_tbl_len < sizeof(struct equiv_cpu_entry) || + buf_size < equiv_tbl_len) { + if (!early) + pr_debug("Truncated equivalence table.\n"); + + return false; + } + + return true; +} + +/* + * Check whether there is a valid, non-truncated microcode patch section at the + * beginning of @buf of size @buf_size. Set @early to use this function in the + * early path. + * + * On success, @sh_psize returns the patch size according to the section header, + * to the caller. + */ +static bool +__verify_patch_section(const u8 *buf, size_t buf_size, u32 *sh_psize, bool early) +{ + u32 p_type, p_size; + const u32 *hdr; + + if (buf_size < SECTION_HDR_SIZE) { + if (!early) + pr_debug("Truncated patch section.\n"); + + return false; + } + + hdr = (const u32 *)buf; + p_type = hdr[0]; + p_size = hdr[1]; + + if (p_type != UCODE_UCODE_TYPE) { + if (!early) + pr_debug("Invalid type field (0x%x) in container file section header.\n", + p_type); + + return false; + } + + if (p_size < sizeof(struct microcode_header_amd)) { + if (!early) + pr_debug("Patch of size %u too short.\n", p_size); + + return false; + } + + *sh_psize = p_size; + + return true; +} + +/* + * Check whether the passed remaining file @buf_size is large enough to contain + * a patch of the indicated @sh_psize (and also whether this size does not + * exceed the per-family maximum). @sh_psize is the size read from the section + * header. + */ +static unsigned int __verify_patch_size(u8 family, u32 sh_psize, size_t buf_size) +{ + u32 max_size; + + if (family >= 0x15) + return min_t(u32, sh_psize, buf_size); + +#define F1XH_MPB_MAX_SIZE 2048 +#define F14H_MPB_MAX_SIZE 1824 + + switch (family) { + case 0x10 ... 0x12: + max_size = F1XH_MPB_MAX_SIZE; + break; + case 0x14: + max_size = F14H_MPB_MAX_SIZE; + break; + default: + WARN(1, "%s: WTF family: 0x%x\n", __func__, family); + return 0; + break; + } + + if (sh_psize > min_t(u32, buf_size, max_size)) + return 0; + + return sh_psize; +} + +/* + * Verify the patch in @buf. + * + * Returns: + * negative: on error + * positive: patch is not for this family, skip it + * 0: success + */ +static int +verify_patch(u8 family, const u8 *buf, size_t buf_size, u32 *patch_size, bool early) +{ + struct microcode_header_amd *mc_hdr; + unsigned int ret; + u32 sh_psize; + u16 proc_id; + u8 patch_fam; + + if (!__verify_patch_section(buf, buf_size, &sh_psize, early)) + return -1; + + /* + * The section header length is not included in this indicated size + * but is present in the leftover file length so we need to subtract + * it before passing this value to the function below. + */ + buf_size -= SECTION_HDR_SIZE; + + /* + * Check if the remaining buffer is big enough to contain a patch of + * size sh_psize, as the section claims. + */ + if (buf_size < sh_psize) { + if (!early) + pr_debug("Patch of size %u truncated.\n", sh_psize); + + return -1; + } + + ret = __verify_patch_size(family, sh_psize, buf_size); + if (!ret) { + if (!early) + pr_debug("Per-family patch size mismatch.\n"); + return -1; + } + + *patch_size = sh_psize; + + mc_hdr = (struct microcode_header_amd *)(buf + SECTION_HDR_SIZE); + if (mc_hdr->nb_dev_id || mc_hdr->sb_dev_id) { + if (!early) + pr_err("Patch-ID 0x%08x: chipset-specific code unsupported.\n", mc_hdr->patch_id); + return -1; + } + + proc_id = mc_hdr->processor_rev_id; + patch_fam = 0xf + (proc_id >> 12); + if (patch_fam != family) + return 1; + + return 0; +} + +/* + * This scans the ucode blob for the proper container as we can have multiple + * containers glued together. Returns the equivalence ID from the equivalence + * table or 0 if none found. + * Returns the amount of bytes consumed while scanning. @desc contains all the + * data we're going to use in later stages of the application. + */ +static size_t parse_container(u8 *ucode, size_t size, struct cont_desc *desc) +{ + struct equiv_cpu_table table; + size_t orig_size = size; + u32 *hdr = (u32 *)ucode; + u16 eq_id; + u8 *buf; + + if (!verify_equivalence_table(ucode, size, true)) + return 0; + + buf = ucode; + + table.entry = (struct equiv_cpu_entry *)(buf + CONTAINER_HDR_SZ); + table.num_entries = hdr[2] / sizeof(struct equiv_cpu_entry); + + /* + * Find the equivalence ID of our CPU in this table. Even if this table + * doesn't contain a patch for the CPU, scan through the whole container + * so that it can be skipped in case there are other containers appended. + */ + eq_id = find_equiv_id(&table, desc->cpuid_1_eax); + + buf += hdr[2] + CONTAINER_HDR_SZ; + size -= hdr[2] + CONTAINER_HDR_SZ; + + /* + * Scan through the rest of the container to find where it ends. We do + * some basic sanity-checking too. + */ + while (size > 0) { + struct microcode_amd *mc; + u32 patch_size; + int ret; + + ret = verify_patch(x86_family(desc->cpuid_1_eax), buf, size, &patch_size, true); + if (ret < 0) { + /* + * Patch verification failed, skip to the next + * container, if there's one: + */ + goto out; + } else if (ret > 0) { + goto skip; + } + + mc = (struct microcode_amd *)(buf + SECTION_HDR_SIZE); + if (eq_id == mc->hdr.processor_rev_id) { + desc->psize = patch_size; + desc->mc = mc; + } + +skip: + /* Skip patch section header too: */ + buf += patch_size + SECTION_HDR_SIZE; + size -= patch_size + SECTION_HDR_SIZE; + } + + /* + * If we have found a patch (desc->mc), it means we're looking at the + * container which has a patch for this CPU so return 0 to mean, @ucode + * already points to the proper container. Otherwise, we return the size + * we scanned so that we can advance to the next container in the + * buffer. + */ + if (desc->mc) { + desc->data = ucode; + desc->size = orig_size - size; + + return 0; + } + +out: + return orig_size - size; +} + +/* + * Scan the ucode blob for the proper container as we can have multiple + * containers glued together. + */ +static void scan_containers(u8 *ucode, size_t size, struct cont_desc *desc) +{ + while (size) { + size_t s = parse_container(ucode, size, desc); + if (!s) + return; + + /* catch wraparound */ + if (size >= s) { + ucode += s; + size -= s; + } else { + return; + } + } +} + +static int __apply_microcode_amd(struct microcode_amd *mc) +{ + u32 rev, dummy; + + native_wrmsrl(MSR_AMD64_PATCH_LOADER, (u64)(long)&mc->hdr.data_code); + + /* verify patch application was successful */ + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + if (rev != mc->hdr.patch_id) + return -1; + + return 0; +} + +/* + * Early load occurs before we can vmalloc(). So we look for the microcode + * patch container file in initrd, traverse equivalent cpu table, look for a + * matching microcode patch, and update, all in initrd memory in place. + * When vmalloc() is available for use later -- on 64-bit during first AP load, + * and on 32-bit during save_microcode_in_initrd_amd() -- we can call + * load_microcode_amd() to save equivalent cpu table and microcode patches in + * kernel heap memory. + * + * Returns true if container found (sets @desc), false otherwise. + */ +static bool +apply_microcode_early_amd(u32 cpuid_1_eax, void *ucode, size_t size, bool save_patch) +{ + struct cont_desc desc = { 0 }; + u8 (*patch)[PATCH_MAX_SIZE]; + struct microcode_amd *mc; + u32 rev, dummy, *new_rev; + bool ret = false; + +#ifdef CONFIG_X86_32 + new_rev = (u32 *)__pa_nodebug(&ucode_new_rev); + patch = (u8 (*)[PATCH_MAX_SIZE])__pa_nodebug(&amd_ucode_patch); +#else + new_rev = &ucode_new_rev; + patch = &amd_ucode_patch[0]; +#endif + + desc.cpuid_1_eax = cpuid_1_eax; + + scan_containers(ucode, size, &desc); + + mc = desc.mc; + if (!mc) + return ret; + + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + /* + * Allow application of the same revision to pick up SMT-specific + * changes even if the revision of the other SMT thread is already + * up-to-date. + */ + if (rev > mc->hdr.patch_id) + return ret; + + if (!__apply_microcode_amd(mc)) { + *new_rev = mc->hdr.patch_id; + ret = true; + + if (save_patch) + memcpy(patch, mc, min_t(u32, desc.psize, PATCH_MAX_SIZE)); + } + + return ret; +} + +static bool get_builtin_microcode(struct cpio_data *cp, unsigned int family) +{ +#ifdef CONFIG_X86_64 + char fw_name[36] = "amd-ucode/microcode_amd.bin"; + + if (family >= 0x15) + snprintf(fw_name, sizeof(fw_name), + "amd-ucode/microcode_amd_fam%.2xh.bin", family); + + return get_builtin_firmware(cp, fw_name); +#else + return false; +#endif +} + +static void __load_ucode_amd(unsigned int cpuid_1_eax, struct cpio_data *ret) +{ + struct ucode_cpu_info *uci; + struct cpio_data cp; + const char *path; + bool use_pa; + + if (IS_ENABLED(CONFIG_X86_32)) { + uci = (struct ucode_cpu_info *)__pa_nodebug(ucode_cpu_info); + path = (const char *)__pa_nodebug(ucode_path); + use_pa = true; + } else { + uci = ucode_cpu_info; + path = ucode_path; + use_pa = false; + } + + if (!get_builtin_microcode(&cp, x86_family(cpuid_1_eax))) + cp = find_microcode_in_initrd(path, use_pa); + + /* Needed in load_microcode_amd() */ + uci->cpu_sig.sig = cpuid_1_eax; + + *ret = cp; +} + +void __init load_ucode_amd_bsp(unsigned int cpuid_1_eax) +{ + struct cpio_data cp = { }; + + __load_ucode_amd(cpuid_1_eax, &cp); + if (!(cp.data && cp.size)) + return; + + apply_microcode_early_amd(cpuid_1_eax, cp.data, cp.size, true); +} + +void load_ucode_amd_ap(unsigned int cpuid_1_eax) +{ + struct microcode_amd *mc; + struct cpio_data cp; + u32 *new_rev, rev, dummy; + + if (IS_ENABLED(CONFIG_X86_32)) { + mc = (struct microcode_amd *)__pa_nodebug(amd_ucode_patch); + new_rev = (u32 *)__pa_nodebug(&ucode_new_rev); + } else { + mc = (struct microcode_amd *)amd_ucode_patch; + new_rev = &ucode_new_rev; + } + + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + /* + * Check whether a new patch has been saved already. Also, allow application of + * the same revision in order to pick up SMT-thread-specific configuration even + * if the sibling SMT thread already has an up-to-date revision. + */ + if (*new_rev && rev <= mc->hdr.patch_id) { + if (!__apply_microcode_amd(mc)) { + *new_rev = mc->hdr.patch_id; + return; + } + } + + __load_ucode_amd(cpuid_1_eax, &cp); + if (!(cp.data && cp.size)) + return; + + apply_microcode_early_amd(cpuid_1_eax, cp.data, cp.size, false); +} + +static enum ucode_state load_microcode_amd(u8 family, const u8 *data, size_t size); + +int __init save_microcode_in_initrd_amd(unsigned int cpuid_1_eax) +{ + struct cont_desc desc = { 0 }; + enum ucode_state ret; + struct cpio_data cp; + + cp = find_microcode_in_initrd(ucode_path, false); + if (!(cp.data && cp.size)) + return -EINVAL; + + desc.cpuid_1_eax = cpuid_1_eax; + + scan_containers(cp.data, cp.size, &desc); + if (!desc.mc) + return -EINVAL; + + ret = load_microcode_amd(x86_family(cpuid_1_eax), desc.data, desc.size); + if (ret > UCODE_UPDATED) + return -EINVAL; + + return 0; +} + +void reload_ucode_amd(unsigned int cpu) +{ + u32 rev, dummy __always_unused; + struct microcode_amd *mc; + + mc = (struct microcode_amd *)amd_ucode_patch[cpu_to_node(cpu)]; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + if (rev < mc->hdr.patch_id) { + if (!__apply_microcode_amd(mc)) { + ucode_new_rev = mc->hdr.patch_id; + pr_info("reload patch_level=0x%08x\n", ucode_new_rev); + } + } +} +static u16 __find_equiv_id(unsigned int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + return find_equiv_id(&equiv_table, uci->cpu_sig.sig); +} + +/* + * a small, trivial cache of per-family ucode patches + */ +static struct ucode_patch *cache_find_patch(u16 equiv_cpu) +{ + struct ucode_patch *p; + + list_for_each_entry(p, µcode_cache, plist) + if (p->equiv_cpu == equiv_cpu) + return p; + return NULL; +} + +static void update_cache(struct ucode_patch *new_patch) +{ + struct ucode_patch *p; + + list_for_each_entry(p, µcode_cache, plist) { + if (p->equiv_cpu == new_patch->equiv_cpu) { + if (p->patch_id >= new_patch->patch_id) { + /* we already have the latest patch */ + kfree(new_patch->data); + kfree(new_patch); + return; + } + + list_replace(&p->plist, &new_patch->plist); + kfree(p->data); + kfree(p); + return; + } + } + /* no patch found, add it */ + list_add_tail(&new_patch->plist, µcode_cache); +} + +static void free_cache(void) +{ + struct ucode_patch *p, *tmp; + + list_for_each_entry_safe(p, tmp, µcode_cache, plist) { + __list_del(p->plist.prev, p->plist.next); + kfree(p->data); + kfree(p); + } +} + +static struct ucode_patch *find_patch(unsigned int cpu) +{ + u16 equiv_id; + + equiv_id = __find_equiv_id(cpu); + if (!equiv_id) + return NULL; + + return cache_find_patch(equiv_id); +} + +static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct ucode_patch *p; + + csig->sig = cpuid_eax(0x00000001); + csig->rev = c->microcode; + + /* + * a patch could have been loaded early, set uci->mc so that + * mc_bp_resume() can call apply_microcode() + */ + p = find_patch(cpu); + if (p && (p->patch_id == csig->rev)) + uci->mc = p->data; + + pr_info("CPU%d: patch_level=0x%08x\n", cpu, csig->rev); + + return 0; +} + +static enum ucode_state apply_microcode_amd(int cpu) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct microcode_amd *mc_amd; + struct ucode_cpu_info *uci; + struct ucode_patch *p; + enum ucode_state ret; + u32 rev, dummy __always_unused; + + BUG_ON(raw_smp_processor_id() != cpu); + + uci = ucode_cpu_info + cpu; + + p = find_patch(cpu); + if (!p) + return UCODE_NFOUND; + + mc_amd = p->data; + uci->mc = p->data; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + /* need to apply patch? */ + if (rev > mc_amd->hdr.patch_id) { + ret = UCODE_OK; + goto out; + } + + if (__apply_microcode_amd(mc_amd)) { + pr_err("CPU%d: update failed for patch_level=0x%08x\n", + cpu, mc_amd->hdr.patch_id); + return UCODE_ERROR; + } + + rev = mc_amd->hdr.patch_id; + ret = UCODE_UPDATED; + + pr_info("CPU%d: new patch_level=0x%08x\n", cpu, rev); + +out: + uci->cpu_sig.rev = rev; + c->microcode = rev; + + /* Update boot_cpu_data's revision too, if we're on the BSP: */ + if (c->cpu_index == boot_cpu_data.cpu_index) + boot_cpu_data.microcode = rev; + + return ret; +} + +static size_t install_equiv_cpu_table(const u8 *buf, size_t buf_size) +{ + u32 equiv_tbl_len; + const u32 *hdr; + + if (!verify_equivalence_table(buf, buf_size, false)) + return 0; + + hdr = (const u32 *)buf; + equiv_tbl_len = hdr[2]; + + equiv_table.entry = vmalloc(equiv_tbl_len); + if (!equiv_table.entry) { + pr_err("failed to allocate equivalent CPU table\n"); + return 0; + } + + memcpy(equiv_table.entry, buf + CONTAINER_HDR_SZ, equiv_tbl_len); + equiv_table.num_entries = equiv_tbl_len / sizeof(struct equiv_cpu_entry); + + /* add header length */ + return equiv_tbl_len + CONTAINER_HDR_SZ; +} + +static void free_equiv_cpu_table(void) +{ + vfree(equiv_table.entry); + memset(&equiv_table, 0, sizeof(equiv_table)); +} + +static void cleanup(void) +{ + free_equiv_cpu_table(); + free_cache(); +} + +/* + * Return a non-negative value even if some of the checks failed so that + * we can skip over the next patch. If we return a negative value, we + * signal a grave error like a memory allocation has failed and the + * driver cannot continue functioning normally. In such cases, we tear + * down everything we've used up so far and exit. + */ +static int verify_and_add_patch(u8 family, u8 *fw, unsigned int leftover, + unsigned int *patch_size) +{ + struct microcode_header_amd *mc_hdr; + struct ucode_patch *patch; + u16 proc_id; + int ret; + + ret = verify_patch(family, fw, leftover, patch_size, false); + if (ret) + return ret; + + patch = kzalloc(sizeof(*patch), GFP_KERNEL); + if (!patch) { + pr_err("Patch allocation failure.\n"); + return -EINVAL; + } + + patch->data = kmemdup(fw + SECTION_HDR_SIZE, *patch_size, GFP_KERNEL); + if (!patch->data) { + pr_err("Patch data allocation failure.\n"); + kfree(patch); + return -EINVAL; + } + patch->size = *patch_size; + + mc_hdr = (struct microcode_header_amd *)(fw + SECTION_HDR_SIZE); + proc_id = mc_hdr->processor_rev_id; + + INIT_LIST_HEAD(&patch->plist); + patch->patch_id = mc_hdr->patch_id; + patch->equiv_cpu = proc_id; + + pr_debug("%s: Added patch_id: 0x%08x, proc_id: 0x%04x\n", + __func__, patch->patch_id, proc_id); + + /* ... and add to cache. */ + update_cache(patch); + + return 0; +} + +static enum ucode_state __load_microcode_amd(u8 family, const u8 *data, + size_t size) +{ + u8 *fw = (u8 *)data; + size_t offset; + + offset = install_equiv_cpu_table(data, size); + if (!offset) + return UCODE_ERROR; + + fw += offset; + size -= offset; + + if (*(u32 *)fw != UCODE_UCODE_TYPE) { + pr_err("invalid type field in container file section header\n"); + free_equiv_cpu_table(); + return UCODE_ERROR; + } + + while (size > 0) { + unsigned int crnt_size = 0; + int ret; + + ret = verify_and_add_patch(family, fw, size, &crnt_size); + if (ret < 0) + return UCODE_ERROR; + + fw += crnt_size + SECTION_HDR_SIZE; + size -= (crnt_size + SECTION_HDR_SIZE); + } + + return UCODE_OK; +} + +static enum ucode_state load_microcode_amd(u8 family, const u8 *data, size_t size) +{ + struct cpuinfo_x86 *c; + unsigned int nid, cpu; + struct ucode_patch *p; + enum ucode_state ret; + + /* free old equiv table */ + free_equiv_cpu_table(); + + ret = __load_microcode_amd(family, data, size); + if (ret != UCODE_OK) { + cleanup(); + return ret; + } + + for_each_node(nid) { + cpu = cpumask_first(cpumask_of_node(nid)); + c = &cpu_data(cpu); + + p = find_patch(cpu); + if (!p) + continue; + + if (c->microcode >= p->patch_id) + continue; + + ret = UCODE_NEW; + + memset(&amd_ucode_patch[nid], 0, PATCH_MAX_SIZE); + memcpy(&amd_ucode_patch[nid], p->data, min_t(u32, p->size, PATCH_MAX_SIZE)); + } + + return ret; +} + +/* + * AMD microcode firmware naming convention, up to family 15h they are in + * the legacy file: + * + * amd-ucode/microcode_amd.bin + * + * This legacy file is always smaller than 2K in size. + * + * Beginning with family 15h, they are in family-specific firmware files: + * + * amd-ucode/microcode_amd_fam15h.bin + * amd-ucode/microcode_amd_fam16h.bin + * ... + * + * These might be larger than 2K. + */ +static enum ucode_state request_microcode_amd(int cpu, struct device *device, + bool refresh_fw) +{ + char fw_name[36] = "amd-ucode/microcode_amd.bin"; + struct cpuinfo_x86 *c = &cpu_data(cpu); + enum ucode_state ret = UCODE_NFOUND; + const struct firmware *fw; + + /* reload ucode container only on the boot cpu */ + if (!refresh_fw) + return UCODE_OK; + + if (c->x86 >= 0x15) + snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); + + if (request_firmware_direct(&fw, (const char *)fw_name, device)) { + pr_debug("failed to load file %s\n", fw_name); + goto out; + } + + ret = UCODE_ERROR; + if (!verify_container(fw->data, fw->size, false)) + goto fw_release; + + ret = load_microcode_amd(c->x86, fw->data, fw->size); + + fw_release: + release_firmware(fw); + + out: + return ret; +} + +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) +{ + return UCODE_ERROR; +} + +static void microcode_fini_cpu_amd(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + uci->mc = NULL; +} + +static struct microcode_ops microcode_amd_ops = { + .request_microcode_user = request_microcode_user, + .request_microcode_fw = request_microcode_amd, + .collect_cpu_info = collect_cpu_info_amd, + .apply_microcode = apply_microcode_amd, + .microcode_fini_cpu = microcode_fini_cpu_amd, +}; + +struct microcode_ops * __init init_amd_microcode(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + if (c->x86_vendor != X86_VENDOR_AMD || c->x86 < 0x10) { + pr_warn("AMD CPU family 0x%x not supported\n", c->x86); + return NULL; + } + + if (ucode_new_rev) + pr_info_once("microcode updated early to new patch_level=0x%08x\n", + ucode_new_rev); + + return µcode_amd_ops; +} + +void __exit exit_amd_microcode(void) +{ + cleanup(); +} diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c new file mode 100644 index 000000000..24254d141 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -0,0 +1,825 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * CPU Microcode Update Driver for Linux + * + * Copyright (C) 2000-2006 Tigran Aivazian <aivazian.tigran@gmail.com> + * 2006 Shaohua Li <shaohua.li@intel.com> + * 2013-2016 Borislav Petkov <bp@alien8.de> + * + * X86 CPU microcode early update for Linux: + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * (C) 2015 Borislav Petkov <bp@alien8.de> + * + * This driver allows to upgrade microcode on x86 processors. + */ + +#define pr_fmt(fmt) "microcode: " fmt + +#include <linux/platform_device.h> +#include <linux/stop_machine.h> +#include <linux/syscore_ops.h> +#include <linux/miscdevice.h> +#include <linux/capability.h> +#include <linux/firmware.h> +#include <linux/kernel.h> +#include <linux/delay.h> +#include <linux/mutex.h> +#include <linux/cpu.h> +#include <linux/nmi.h> +#include <linux/fs.h> +#include <linux/mm.h> + +#include <asm/microcode_intel.h> +#include <asm/cpu_device_id.h> +#include <asm/microcode_amd.h> +#include <asm/perf_event.h> +#include <asm/microcode.h> +#include <asm/processor.h> +#include <asm/cmdline.h> +#include <asm/setup.h> + +#define DRIVER_VERSION "2.2" + +static struct microcode_ops *microcode_ops; +static bool dis_ucode_ldr = true; + +bool initrd_gone; + +LIST_HEAD(microcode_cache); + +/* + * Synchronization. + * + * All non cpu-hotplug-callback call sites use: + * + * - microcode_mutex to synchronize with each other; + * - cpus_read_lock/unlock() to synchronize with + * the cpu-hotplug-callback call sites. + * + * We guarantee that only a single cpu is being + * updated at any particular moment of time. + */ +static DEFINE_MUTEX(microcode_mutex); + +struct ucode_cpu_info ucode_cpu_info[NR_CPUS]; + +struct cpu_info_ctx { + struct cpu_signature *cpu_sig; + int err; +}; + +/* + * Those patch levels cannot be updated to newer ones and thus should be final. + */ +static u32 final_levels[] = { + 0x01000098, + 0x0100009f, + 0x010000af, + 0, /* T-101 terminator */ +}; + +/* + * Check the current patch level on this CPU. + * + * Returns: + * - true: if update should stop + * - false: otherwise + */ +static bool amd_check_current_patch_level(void) +{ + u32 lvl, dummy, i; + u32 *levels; + + native_rdmsr(MSR_AMD64_PATCH_LEVEL, lvl, dummy); + + if (IS_ENABLED(CONFIG_X86_32)) + levels = (u32 *)__pa_nodebug(&final_levels); + else + levels = final_levels; + + for (i = 0; levels[i]; i++) { + if (lvl == levels[i]) + return true; + } + return false; +} + +static bool __init check_loader_disabled_bsp(void) +{ + static const char *__dis_opt_str = "dis_ucode_ldr"; + +#ifdef CONFIG_X86_32 + const char *cmdline = (const char *)__pa_nodebug(boot_command_line); + const char *option = (const char *)__pa_nodebug(__dis_opt_str); + bool *res = (bool *)__pa_nodebug(&dis_ucode_ldr); + +#else /* CONFIG_X86_64 */ + const char *cmdline = boot_command_line; + const char *option = __dis_opt_str; + bool *res = &dis_ucode_ldr; +#endif + + /* + * CPUID(1).ECX[31]: reserved for hypervisor use. This is still not + * completely accurate as xen pv guests don't see that CPUID bit set but + * that's good enough as they don't land on the BSP path anyway. + */ + if (native_cpuid_ecx(1) & BIT(31)) + return *res; + + if (x86_cpuid_vendor() == X86_VENDOR_AMD) { + if (amd_check_current_patch_level()) + return *res; + } + + if (cmdline_find_option_bool(cmdline, option) <= 0) + *res = false; + + return *res; +} + +extern struct builtin_fw __start_builtin_fw[]; +extern struct builtin_fw __end_builtin_fw[]; + +bool get_builtin_firmware(struct cpio_data *cd, const char *name) +{ + struct builtin_fw *b_fw; + + for (b_fw = __start_builtin_fw; b_fw != __end_builtin_fw; b_fw++) { + if (!strcmp(name, b_fw->name)) { + cd->size = b_fw->size; + cd->data = b_fw->data; + return true; + } + } + return false; +} + +void __init load_ucode_bsp(void) +{ + unsigned int cpuid_1_eax; + bool intel = true; + + if (!have_cpuid_p()) + return; + + cpuid_1_eax = native_cpuid_eax(1); + + switch (x86_cpuid_vendor()) { + case X86_VENDOR_INTEL: + if (x86_family(cpuid_1_eax) < 6) + return; + break; + + case X86_VENDOR_AMD: + if (x86_family(cpuid_1_eax) < 0x10) + return; + intel = false; + break; + + default: + return; + } + + if (check_loader_disabled_bsp()) + return; + + if (intel) + load_ucode_intel_bsp(); + else + load_ucode_amd_bsp(cpuid_1_eax); +} + +static bool check_loader_disabled_ap(void) +{ +#ifdef CONFIG_X86_32 + return *((bool *)__pa_nodebug(&dis_ucode_ldr)); +#else + return dis_ucode_ldr; +#endif +} + +void load_ucode_ap(void) +{ + unsigned int cpuid_1_eax; + + if (check_loader_disabled_ap()) + return; + + cpuid_1_eax = native_cpuid_eax(1); + + switch (x86_cpuid_vendor()) { + case X86_VENDOR_INTEL: + if (x86_family(cpuid_1_eax) >= 6) + load_ucode_intel_ap(); + break; + case X86_VENDOR_AMD: + if (x86_family(cpuid_1_eax) >= 0x10) + load_ucode_amd_ap(cpuid_1_eax); + break; + default: + break; + } +} + +static int __init save_microcode_in_initrd(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + int ret = -EINVAL; + + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + if (c->x86 >= 6) + ret = save_microcode_in_initrd_intel(); + break; + case X86_VENDOR_AMD: + if (c->x86 >= 0x10) + ret = save_microcode_in_initrd_amd(cpuid_eax(1)); + break; + default: + break; + } + + initrd_gone = true; + + return ret; +} + +struct cpio_data find_microcode_in_initrd(const char *path, bool use_pa) +{ +#ifdef CONFIG_BLK_DEV_INITRD + unsigned long start = 0; + size_t size; + +#ifdef CONFIG_X86_32 + struct boot_params *params; + + if (use_pa) + params = (struct boot_params *)__pa_nodebug(&boot_params); + else + params = &boot_params; + + size = params->hdr.ramdisk_size; + + /* + * Set start only if we have an initrd image. We cannot use initrd_start + * because it is not set that early yet. + */ + if (size) + start = params->hdr.ramdisk_image; + +# else /* CONFIG_X86_64 */ + size = (unsigned long)boot_params.ext_ramdisk_size << 32; + size |= boot_params.hdr.ramdisk_size; + + if (size) { + start = (unsigned long)boot_params.ext_ramdisk_image << 32; + start |= boot_params.hdr.ramdisk_image; + + start += PAGE_OFFSET; + } +# endif + + /* + * Fixup the start address: after reserve_initrd() runs, initrd_start + * has the virtual address of the beginning of the initrd. It also + * possibly relocates the ramdisk. In either case, initrd_start contains + * the updated address so use that instead. + * + * initrd_gone is for the hotplug case where we've thrown out initrd + * already. + */ + if (!use_pa) { + if (initrd_gone) + return (struct cpio_data){ NULL, 0, "" }; + if (initrd_start) + start = initrd_start; + } else { + /* + * The picture with physical addresses is a bit different: we + * need to get the *physical* address to which the ramdisk was + * relocated, i.e., relocated_ramdisk (not initrd_start) and + * since we're running from physical addresses, we need to access + * relocated_ramdisk through its *physical* address too. + */ + u64 *rr = (u64 *)__pa_nodebug(&relocated_ramdisk); + if (*rr) + start = *rr; + } + + return find_cpio_data(path, (void *)start, size, NULL); +#else /* !CONFIG_BLK_DEV_INITRD */ + return (struct cpio_data){ NULL, 0, "" }; +#endif +} + +void reload_early_microcode(unsigned int cpu) +{ + int vendor, family; + + vendor = x86_cpuid_vendor(); + family = x86_cpuid_family(); + + switch (vendor) { + case X86_VENDOR_INTEL: + if (family >= 6) + reload_ucode_intel(); + break; + case X86_VENDOR_AMD: + if (family >= 0x10) + reload_ucode_amd(cpu); + break; + default: + break; + } +} + +static void collect_cpu_info_local(void *arg) +{ + struct cpu_info_ctx *ctx = arg; + + ctx->err = microcode_ops->collect_cpu_info(smp_processor_id(), + ctx->cpu_sig); +} + +static int collect_cpu_info_on_target(int cpu, struct cpu_signature *cpu_sig) +{ + struct cpu_info_ctx ctx = { .cpu_sig = cpu_sig, .err = 0 }; + int ret; + + ret = smp_call_function_single(cpu, collect_cpu_info_local, &ctx, 1); + if (!ret) + ret = ctx.err; + + return ret; +} + +static int collect_cpu_info(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + int ret; + + memset(uci, 0, sizeof(*uci)); + + ret = collect_cpu_info_on_target(cpu, &uci->cpu_sig); + if (!ret) + uci->valid = 1; + + return ret; +} + +static void apply_microcode_local(void *arg) +{ + enum ucode_state *err = arg; + + *err = microcode_ops->apply_microcode(smp_processor_id()); +} + +static int apply_microcode_on_target(int cpu) +{ + enum ucode_state err; + int ret; + + ret = smp_call_function_single(cpu, apply_microcode_local, &err, 1); + if (!ret) { + if (err == UCODE_ERROR) + ret = 1; + } + return ret; +} + +/* fake device for request_firmware */ +static struct platform_device *microcode_pdev; + +#ifdef CONFIG_MICROCODE_LATE_LOADING +/* + * Late loading dance. Why the heavy-handed stomp_machine effort? + * + * - HT siblings must be idle and not execute other code while the other sibling + * is loading microcode in order to avoid any negative interactions caused by + * the loading. + * + * - In addition, microcode update on the cores must be serialized until this + * requirement can be relaxed in the future. Right now, this is conservative + * and good. + */ +#define SPINUNIT 100 /* 100 nsec */ + +static int check_online_cpus(void) +{ + unsigned int cpu; + + /* + * Make sure all CPUs are online. It's fine for SMT to be disabled if + * all the primary threads are still online. + */ + for_each_present_cpu(cpu) { + if (topology_is_primary_thread(cpu) && !cpu_online(cpu)) { + pr_err("Not all CPUs online, aborting microcode update.\n"); + return -EINVAL; + } + } + + return 0; +} + +static atomic_t late_cpus_in; +static atomic_t late_cpus_out; + +static int __wait_for_cpus(atomic_t *t, long long timeout) +{ + int all_cpus = num_online_cpus(); + + atomic_inc(t); + + while (atomic_read(t) < all_cpus) { + if (timeout < SPINUNIT) { + pr_err("Timeout while waiting for CPUs rendezvous, remaining: %d\n", + all_cpus - atomic_read(t)); + return 1; + } + + ndelay(SPINUNIT); + timeout -= SPINUNIT; + + touch_nmi_watchdog(); + } + return 0; +} + +/* + * Returns: + * < 0 - on error + * 0 - success (no update done or microcode was updated) + */ +static int __reload_late(void *info) +{ + int cpu = smp_processor_id(); + enum ucode_state err; + int ret = 0; + + /* + * Wait for all CPUs to arrive. A load will not be attempted unless all + * CPUs show up. + * */ + if (__wait_for_cpus(&late_cpus_in, NSEC_PER_SEC)) + return -1; + + /* + * On an SMT system, it suffices to load the microcode on one sibling of + * the core because the microcode engine is shared between the threads. + * Synchronization still needs to take place so that no concurrent + * loading attempts happen on multiple threads of an SMT core. See + * below. + */ + if (cpumask_first(topology_sibling_cpumask(cpu)) == cpu) + apply_microcode_local(&err); + else + goto wait_for_siblings; + + if (err >= UCODE_NFOUND) { + if (err == UCODE_ERROR) + pr_warn("Error reloading microcode on CPU %d\n", cpu); + + ret = -1; + } + +wait_for_siblings: + if (__wait_for_cpus(&late_cpus_out, NSEC_PER_SEC)) + panic("Timeout during microcode update!\n"); + + /* + * At least one thread has completed update on each core. + * For others, simply call the update to make sure the + * per-cpu cpuinfo can be updated with right microcode + * revision. + */ + if (cpumask_first(topology_sibling_cpumask(cpu)) != cpu) + apply_microcode_local(&err); + + return ret; +} + +/* + * Reload microcode late on all CPUs. Wait for a sec until they + * all gather together. + */ +static int microcode_reload_late(void) +{ + int old = boot_cpu_data.microcode, ret; + struct cpuinfo_x86 prev_info; + + atomic_set(&late_cpus_in, 0); + atomic_set(&late_cpus_out, 0); + + /* + * Take a snapshot before the microcode update in order to compare and + * check whether any bits changed after an update. + */ + store_cpu_caps(&prev_info); + + ret = stop_machine_cpuslocked(__reload_late, NULL, cpu_online_mask); + if (!ret) { + pr_info("Reload succeeded, microcode revision: 0x%x -> 0x%x\n", + old, boot_cpu_data.microcode); + microcode_check(&prev_info); + } else { + pr_info("Reload failed, current microcode revision: 0x%x\n", + boot_cpu_data.microcode); + } + + return ret; +} + +static ssize_t reload_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t size) +{ + enum ucode_state tmp_ret = UCODE_OK; + int bsp = boot_cpu_data.cpu_index; + unsigned long val; + ssize_t ret = 0; + + ret = kstrtoul(buf, 0, &val); + if (ret) + return ret; + + if (val != 1) + return size; + + cpus_read_lock(); + + ret = check_online_cpus(); + if (ret) + goto put; + + tmp_ret = microcode_ops->request_microcode_fw(bsp, µcode_pdev->dev, true); + if (tmp_ret != UCODE_NEW) + goto put; + + mutex_lock(µcode_mutex); + ret = microcode_reload_late(); + mutex_unlock(µcode_mutex); + +put: + cpus_read_unlock(); + + if (ret == 0) + ret = size; + + return ret; +} + +static DEVICE_ATTR_WO(reload); +#endif + +static ssize_t version_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + dev->id; + + return sprintf(buf, "0x%x\n", uci->cpu_sig.rev); +} + +static ssize_t pf_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + dev->id; + + return sprintf(buf, "0x%x\n", uci->cpu_sig.pf); +} + +static DEVICE_ATTR(version, 0444, version_show, NULL); +static DEVICE_ATTR(processor_flags, 0444, pf_show, NULL); + +static struct attribute *mc_default_attrs[] = { + &dev_attr_version.attr, + &dev_attr_processor_flags.attr, + NULL +}; + +static const struct attribute_group mc_attr_group = { + .attrs = mc_default_attrs, + .name = "microcode", +}; + +static void microcode_fini_cpu(int cpu) +{ + if (microcode_ops->microcode_fini_cpu) + microcode_ops->microcode_fini_cpu(cpu); +} + +static enum ucode_state microcode_resume_cpu(int cpu) +{ + if (apply_microcode_on_target(cpu)) + return UCODE_ERROR; + + pr_debug("CPU%d updated upon resume\n", cpu); + + return UCODE_OK; +} + +static enum ucode_state microcode_init_cpu(int cpu, bool refresh_fw) +{ + enum ucode_state ustate; + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci->valid) + return UCODE_OK; + + if (collect_cpu_info(cpu)) + return UCODE_ERROR; + + /* --dimm. Trigger a delayed update? */ + if (system_state != SYSTEM_RUNNING) + return UCODE_NFOUND; + + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev, refresh_fw); + if (ustate == UCODE_NEW) { + pr_debug("CPU%d updated upon init\n", cpu); + apply_microcode_on_target(cpu); + } + + return ustate; +} + +static enum ucode_state microcode_update_cpu(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + /* Refresh CPU microcode revision after resume. */ + collect_cpu_info(cpu); + + if (uci->valid) + return microcode_resume_cpu(cpu); + + return microcode_init_cpu(cpu, false); +} + +static int mc_device_add(struct device *dev, struct subsys_interface *sif) +{ + int err, cpu = dev->id; + + if (!cpu_online(cpu)) + return 0; + + pr_debug("CPU%d added\n", cpu); + + err = sysfs_create_group(&dev->kobj, &mc_attr_group); + if (err) + return err; + + if (microcode_init_cpu(cpu, true) == UCODE_ERROR) + return -EINVAL; + + return err; +} + +static void mc_device_remove(struct device *dev, struct subsys_interface *sif) +{ + int cpu = dev->id; + + if (!cpu_online(cpu)) + return; + + pr_debug("CPU%d removed\n", cpu); + microcode_fini_cpu(cpu); + sysfs_remove_group(&dev->kobj, &mc_attr_group); +} + +static struct subsys_interface mc_cpu_interface = { + .name = "microcode", + .subsys = &cpu_subsys, + .add_dev = mc_device_add, + .remove_dev = mc_device_remove, +}; + +/** + * microcode_bsp_resume - Update boot CPU microcode during resume. + */ +void microcode_bsp_resume(void) +{ + int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci->valid && uci->mc) + microcode_ops->apply_microcode(cpu); + else if (!uci->mc) + reload_early_microcode(cpu); +} + +static struct syscore_ops mc_syscore_ops = { + .resume = microcode_bsp_resume, +}; + +static int mc_cpu_starting(unsigned int cpu) +{ + microcode_update_cpu(cpu); + pr_debug("CPU%d added\n", cpu); + return 0; +} + +static int mc_cpu_online(unsigned int cpu) +{ + struct device *dev = get_cpu_device(cpu); + + if (sysfs_create_group(&dev->kobj, &mc_attr_group)) + pr_err("Failed to create group for CPU%d\n", cpu); + return 0; +} + +static int mc_cpu_down_prep(unsigned int cpu) +{ + struct device *dev; + + dev = get_cpu_device(cpu); + /* Suspend is in progress, only remove the interface */ + sysfs_remove_group(&dev->kobj, &mc_attr_group); + pr_debug("CPU%d removed\n", cpu); + + return 0; +} + +static struct attribute *cpu_root_microcode_attrs[] = { +#ifdef CONFIG_MICROCODE_LATE_LOADING + &dev_attr_reload.attr, +#endif + NULL +}; + +static const struct attribute_group cpu_root_microcode_group = { + .name = "microcode", + .attrs = cpu_root_microcode_attrs, +}; + +int __init microcode_init(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + int error; + + if (dis_ucode_ldr) + return -EINVAL; + + if (c->x86_vendor == X86_VENDOR_INTEL) + microcode_ops = init_intel_microcode(); + else if (c->x86_vendor == X86_VENDOR_AMD) + microcode_ops = init_amd_microcode(); + else + pr_err("no support for this CPU vendor\n"); + + if (!microcode_ops) + return -ENODEV; + + microcode_pdev = platform_device_register_simple("microcode", -1, + NULL, 0); + if (IS_ERR(microcode_pdev)) + return PTR_ERR(microcode_pdev); + + cpus_read_lock(); + mutex_lock(µcode_mutex); + + error = subsys_interface_register(&mc_cpu_interface); + if (!error) + perf_check_microcode(); + mutex_unlock(µcode_mutex); + cpus_read_unlock(); + + if (error) + goto out_pdev; + + error = sysfs_create_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + if (error) { + pr_err("Error creating microcode group!\n"); + goto out_driver; + } + + register_syscore_ops(&mc_syscore_ops); + cpuhp_setup_state_nocalls(CPUHP_AP_MICROCODE_LOADER, "x86/microcode:starting", + mc_cpu_starting, NULL); + cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN, "x86/microcode:online", + mc_cpu_online, mc_cpu_down_prep); + + pr_info("Microcode Update Driver: v%s.", DRIVER_VERSION); + + return 0; + + out_driver: + cpus_read_lock(); + mutex_lock(µcode_mutex); + + subsys_interface_unregister(&mc_cpu_interface); + + mutex_unlock(µcode_mutex); + cpus_read_unlock(); + + out_pdev: + platform_device_unregister(microcode_pdev); + return error; + +} +fs_initcall(save_microcode_in_initrd); +late_initcall(microcode_init); diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c new file mode 100644 index 000000000..1ba590e6e --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -0,0 +1,996 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Intel CPU Microcode Update Driver for Linux + * + * Copyright (C) 2000-2006 Tigran Aivazian <aivazian.tigran@gmail.com> + * 2006 Shaohua Li <shaohua.li@intel.com> + * + * Intel CPU microcode early update for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + */ + +/* + * This needs to be before all headers so that pr_debug in printk.h doesn't turn + * printk calls into no_printk(). + * + *#define DEBUG + */ +#define pr_fmt(fmt) "microcode: " fmt + +#include <linux/earlycpio.h> +#include <linux/firmware.h> +#include <linux/uaccess.h> +#include <linux/vmalloc.h> +#include <linux/initrd.h> +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/cpu.h> +#include <linux/uio.h> +#include <linux/mm.h> + +#include <asm/microcode_intel.h> +#include <asm/intel-family.h> +#include <asm/processor.h> +#include <asm/tlbflush.h> +#include <asm/setup.h> +#include <asm/msr.h> + +static const char ucode_path[] = "kernel/x86/microcode/GenuineIntel.bin"; + +/* Current microcode patch used in early patching on the APs. */ +static struct microcode_intel *intel_ucode_patch; + +/* last level cache size per core */ +static int llc_size_per_core; + +static inline bool cpu_signatures_match(unsigned int s1, unsigned int p1, + unsigned int s2, unsigned int p2) +{ + if (s1 != s2) + return false; + + /* Processor flags are either both 0 ... */ + if (!p1 && !p2) + return true; + + /* ... or they intersect. */ + return p1 & p2; +} + +/* + * Returns 1 if update has been found, 0 otherwise. + */ +static int find_matching_signature(void *mc, unsigned int csig, int cpf) +{ + struct microcode_header_intel *mc_hdr = mc; + struct extended_sigtable *ext_hdr; + struct extended_signature *ext_sig; + int i; + + if (cpu_signatures_match(csig, cpf, mc_hdr->sig, mc_hdr->pf)) + return 1; + + /* Look for ext. headers: */ + if (get_totalsize(mc_hdr) <= get_datasize(mc_hdr) + MC_HEADER_SIZE) + return 0; + + ext_hdr = mc + get_datasize(mc_hdr) + MC_HEADER_SIZE; + ext_sig = (void *)ext_hdr + EXT_HEADER_SIZE; + + for (i = 0; i < ext_hdr->count; i++) { + if (cpu_signatures_match(csig, cpf, ext_sig->sig, ext_sig->pf)) + return 1; + ext_sig++; + } + return 0; +} + +/* + * Returns 1 if update has been found, 0 otherwise. + */ +static int has_newer_microcode(void *mc, unsigned int csig, int cpf, int new_rev) +{ + struct microcode_header_intel *mc_hdr = mc; + + if (mc_hdr->rev <= new_rev) + return 0; + + return find_matching_signature(mc, csig, cpf); +} + +static struct ucode_patch *memdup_patch(void *data, unsigned int size) +{ + struct ucode_patch *p; + + p = kzalloc(sizeof(struct ucode_patch), GFP_KERNEL); + if (!p) + return NULL; + + p->data = kmemdup(data, size, GFP_KERNEL); + if (!p->data) { + kfree(p); + return NULL; + } + + return p; +} + +static void save_microcode_patch(struct ucode_cpu_info *uci, void *data, unsigned int size) +{ + struct microcode_header_intel *mc_hdr, *mc_saved_hdr; + struct ucode_patch *iter, *tmp, *p = NULL; + bool prev_found = false; + unsigned int sig, pf; + + mc_hdr = (struct microcode_header_intel *)data; + + list_for_each_entry_safe(iter, tmp, µcode_cache, plist) { + mc_saved_hdr = (struct microcode_header_intel *)iter->data; + sig = mc_saved_hdr->sig; + pf = mc_saved_hdr->pf; + + if (find_matching_signature(data, sig, pf)) { + prev_found = true; + + if (mc_hdr->rev <= mc_saved_hdr->rev) + continue; + + p = memdup_patch(data, size); + if (!p) + pr_err("Error allocating buffer %p\n", data); + else { + list_replace(&iter->plist, &p->plist); + kfree(iter->data); + kfree(iter); + } + } + } + + /* + * There weren't any previous patches found in the list cache; save the + * newly found. + */ + if (!prev_found) { + p = memdup_patch(data, size); + if (!p) + pr_err("Error allocating buffer for %p\n", data); + else + list_add_tail(&p->plist, µcode_cache); + } + + if (!p) + return; + + if (!find_matching_signature(p->data, uci->cpu_sig.sig, uci->cpu_sig.pf)) + return; + + /* + * Save for early loading. On 32-bit, that needs to be a physical + * address as the APs are running from physical addresses, before + * paging has been enabled. + */ + if (IS_ENABLED(CONFIG_X86_32)) + intel_ucode_patch = (struct microcode_intel *)__pa_nodebug(p->data); + else + intel_ucode_patch = p->data; +} + +static int microcode_sanity_check(void *mc, int print_err) +{ + unsigned long total_size, data_size, ext_table_size; + struct microcode_header_intel *mc_header = mc; + struct extended_sigtable *ext_header = NULL; + u32 sum, orig_sum, ext_sigcount = 0, i; + struct extended_signature *ext_sig; + + total_size = get_totalsize(mc_header); + data_size = get_datasize(mc_header); + + if (data_size + MC_HEADER_SIZE > total_size) { + if (print_err) + pr_err("Error: bad microcode data file size.\n"); + return -EINVAL; + } + + if (mc_header->ldrver != 1 || mc_header->hdrver != 1) { + if (print_err) + pr_err("Error: invalid/unknown microcode update format.\n"); + return -EINVAL; + } + + ext_table_size = total_size - (MC_HEADER_SIZE + data_size); + if (ext_table_size) { + u32 ext_table_sum = 0; + u32 *ext_tablep; + + if ((ext_table_size < EXT_HEADER_SIZE) + || ((ext_table_size - EXT_HEADER_SIZE) % EXT_SIGNATURE_SIZE)) { + if (print_err) + pr_err("Error: truncated extended signature table.\n"); + return -EINVAL; + } + + ext_header = mc + MC_HEADER_SIZE + data_size; + if (ext_table_size != exttable_size(ext_header)) { + if (print_err) + pr_err("Error: extended signature table size mismatch.\n"); + return -EFAULT; + } + + ext_sigcount = ext_header->count; + + /* + * Check extended table checksum: the sum of all dwords that + * comprise a valid table must be 0. + */ + ext_tablep = (u32 *)ext_header; + + i = ext_table_size / sizeof(u32); + while (i--) + ext_table_sum += ext_tablep[i]; + + if (ext_table_sum) { + if (print_err) + pr_warn("Bad extended signature table checksum, aborting.\n"); + return -EINVAL; + } + } + + /* + * Calculate the checksum of update data and header. The checksum of + * valid update data and header including the extended signature table + * must be 0. + */ + orig_sum = 0; + i = (MC_HEADER_SIZE + data_size) / sizeof(u32); + while (i--) + orig_sum += ((u32 *)mc)[i]; + + if (orig_sum) { + if (print_err) + pr_err("Bad microcode data checksum, aborting.\n"); + return -EINVAL; + } + + if (!ext_table_size) + return 0; + + /* + * Check extended signature checksum: 0 => valid. + */ + for (i = 0; i < ext_sigcount; i++) { + ext_sig = (void *)ext_header + EXT_HEADER_SIZE + + EXT_SIGNATURE_SIZE * i; + + sum = (mc_header->sig + mc_header->pf + mc_header->cksum) - + (ext_sig->sig + ext_sig->pf + ext_sig->cksum); + if (sum) { + if (print_err) + pr_err("Bad extended signature checksum, aborting.\n"); + return -EINVAL; + } + } + return 0; +} + +/* + * Get microcode matching with BSP's model. Only CPUs with the same model as + * BSP can stay in the platform. + */ +static struct microcode_intel * +scan_microcode(void *data, size_t size, struct ucode_cpu_info *uci, bool save) +{ + struct microcode_header_intel *mc_header; + struct microcode_intel *patch = NULL; + unsigned int mc_size; + + while (size) { + if (size < sizeof(struct microcode_header_intel)) + break; + + mc_header = (struct microcode_header_intel *)data; + + mc_size = get_totalsize(mc_header); + if (!mc_size || + mc_size > size || + microcode_sanity_check(data, 0) < 0) + break; + + size -= mc_size; + + if (!find_matching_signature(data, uci->cpu_sig.sig, + uci->cpu_sig.pf)) { + data += mc_size; + continue; + } + + if (save) { + save_microcode_patch(uci, data, mc_size); + goto next; + } + + + if (!patch) { + if (!has_newer_microcode(data, + uci->cpu_sig.sig, + uci->cpu_sig.pf, + uci->cpu_sig.rev)) + goto next; + + } else { + struct microcode_header_intel *phdr = &patch->hdr; + + if (!has_newer_microcode(data, + phdr->sig, + phdr->pf, + phdr->rev)) + goto next; + } + + /* We have a newer patch, save it. */ + patch = data; + +next: + data += mc_size; + } + + if (size) + return NULL; + + return patch; +} + +static int collect_cpu_info_early(struct ucode_cpu_info *uci) +{ + unsigned int val[2]; + unsigned int family, model; + struct cpu_signature csig = { 0 }; + unsigned int eax, ebx, ecx, edx; + + memset(uci, 0, sizeof(*uci)); + + eax = 0x00000001; + ecx = 0; + native_cpuid(&eax, &ebx, &ecx, &edx); + csig.sig = eax; + + family = x86_family(eax); + model = x86_model(eax); + + if ((model >= 5) || (family > 6)) { + /* get processor flags from MSR 0x17 */ + native_rdmsr(MSR_IA32_PLATFORM_ID, val[0], val[1]); + csig.pf = 1 << ((val[1] >> 18) & 7); + } + + csig.rev = intel_get_microcode_revision(); + + uci->cpu_sig = csig; + uci->valid = 1; + + return 0; +} + +static void show_saved_mc(void) +{ +#ifdef DEBUG + int i = 0, j; + unsigned int sig, pf, rev, total_size, data_size, date; + struct ucode_cpu_info uci; + struct ucode_patch *p; + + if (list_empty(µcode_cache)) { + pr_debug("no microcode data saved.\n"); + return; + } + + collect_cpu_info_early(&uci); + + sig = uci.cpu_sig.sig; + pf = uci.cpu_sig.pf; + rev = uci.cpu_sig.rev; + pr_debug("CPU: sig=0x%x, pf=0x%x, rev=0x%x\n", sig, pf, rev); + + list_for_each_entry(p, µcode_cache, plist) { + struct microcode_header_intel *mc_saved_header; + struct extended_sigtable *ext_header; + struct extended_signature *ext_sig; + int ext_sigcount; + + mc_saved_header = (struct microcode_header_intel *)p->data; + + sig = mc_saved_header->sig; + pf = mc_saved_header->pf; + rev = mc_saved_header->rev; + date = mc_saved_header->date; + + total_size = get_totalsize(mc_saved_header); + data_size = get_datasize(mc_saved_header); + + pr_debug("mc_saved[%d]: sig=0x%x, pf=0x%x, rev=0x%x, total size=0x%x, date = %04x-%02x-%02x\n", + i++, sig, pf, rev, total_size, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); + + /* Look for ext. headers: */ + if (total_size <= data_size + MC_HEADER_SIZE) + continue; + + ext_header = (void *)mc_saved_header + data_size + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (j = 0; j < ext_sigcount; j++) { + sig = ext_sig->sig; + pf = ext_sig->pf; + + pr_debug("\tExtended[%d]: sig=0x%x, pf=0x%x\n", + j, sig, pf); + + ext_sig++; + } + } +#endif +} + +/* + * Save this microcode patch. It will be loaded early when a CPU is + * hot-added or resumes. + */ +static void save_mc_for_early(struct ucode_cpu_info *uci, u8 *mc, unsigned int size) +{ + /* Synchronization during CPU hotplug. */ + static DEFINE_MUTEX(x86_cpu_microcode_mutex); + + mutex_lock(&x86_cpu_microcode_mutex); + + save_microcode_patch(uci, mc, size); + show_saved_mc(); + + mutex_unlock(&x86_cpu_microcode_mutex); +} + +static bool load_builtin_intel_microcode(struct cpio_data *cp) +{ + unsigned int eax = 1, ebx, ecx = 0, edx; + char name[30]; + + if (IS_ENABLED(CONFIG_X86_32)) + return false; + + native_cpuid(&eax, &ebx, &ecx, &edx); + + sprintf(name, "intel-ucode/%02x-%02x-%02x", + x86_family(eax), x86_model(eax), x86_stepping(eax)); + + return get_builtin_firmware(cp, name); +} + +/* + * Print ucode update info. + */ +static void +print_ucode_info(struct ucode_cpu_info *uci, unsigned int date) +{ + pr_info_once("microcode updated early to revision 0x%x, date = %04x-%02x-%02x\n", + uci->cpu_sig.rev, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); +} + +#ifdef CONFIG_X86_32 + +static int delay_ucode_info; +static int current_mc_date; + +/* + * Print early updated ucode info after printk works. This is delayed info dump. + */ +void show_ucode_info_early(void) +{ + struct ucode_cpu_info uci; + + if (delay_ucode_info) { + collect_cpu_info_early(&uci); + print_ucode_info(&uci, current_mc_date); + delay_ucode_info = 0; + } +} + +/* + * At this point, we can not call printk() yet. Delay printing microcode info in + * show_ucode_info_early() until printk() works. + */ +static void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc; + int *delay_ucode_info_p; + int *current_mc_date_p; + + mc = uci->mc; + if (!mc) + return; + + delay_ucode_info_p = (int *)__pa_nodebug(&delay_ucode_info); + current_mc_date_p = (int *)__pa_nodebug(¤t_mc_date); + + *delay_ucode_info_p = 1; + *current_mc_date_p = mc->hdr.date; +} +#else + +static inline void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc; + + mc = uci->mc; + if (!mc) + return; + + print_ucode_info(uci, mc->hdr.date); +} +#endif + +static int apply_microcode_early(struct ucode_cpu_info *uci, bool early) +{ + struct microcode_intel *mc; + u32 rev; + + mc = uci->mc; + if (!mc) + return 0; + + /* + * Save us the MSR write below - which is a particular expensive + * operation - when the other hyperthread has updated the microcode + * already. + */ + rev = intel_get_microcode_revision(); + if (rev >= mc->hdr.rev) { + uci->cpu_sig.rev = rev; + return UCODE_OK; + } + + /* + * Writeback and invalidate caches before updating microcode to avoid + * internal issues depending on what the microcode is updating. + */ + native_wbinvd(); + + /* write microcode via MSR 0x79 */ + native_wrmsrl(MSR_IA32_UCODE_WRITE, (unsigned long)mc->bits); + + rev = intel_get_microcode_revision(); + if (rev != mc->hdr.rev) + return -1; + + uci->cpu_sig.rev = rev; + + if (early) + print_ucode(uci); + else + print_ucode_info(uci, mc->hdr.date); + + return 0; +} + +int __init save_microcode_in_initrd_intel(void) +{ + struct ucode_cpu_info uci; + struct cpio_data cp; + + /* + * initrd is going away, clear patch ptr. We will scan the microcode one + * last time before jettisoning and save a patch, if found. Then we will + * update that pointer too, with a stable patch address to use when + * resuming the cores. + */ + intel_ucode_patch = NULL; + + if (!load_builtin_intel_microcode(&cp)) + cp = find_microcode_in_initrd(ucode_path, false); + + if (!(cp.data && cp.size)) + return 0; + + collect_cpu_info_early(&uci); + + scan_microcode(cp.data, cp.size, &uci, true); + + show_saved_mc(); + + return 0; +} + +/* + * @res_patch, output: a pointer to the patch we found. + */ +static struct microcode_intel *__load_ucode_intel(struct ucode_cpu_info *uci) +{ + static const char *path; + struct cpio_data cp; + bool use_pa; + + if (IS_ENABLED(CONFIG_X86_32)) { + path = (const char *)__pa_nodebug(ucode_path); + use_pa = true; + } else { + path = ucode_path; + use_pa = false; + } + + /* try built-in microcode first */ + if (!load_builtin_intel_microcode(&cp)) + cp = find_microcode_in_initrd(path, use_pa); + + if (!(cp.data && cp.size)) + return NULL; + + collect_cpu_info_early(uci); + + return scan_microcode(cp.data, cp.size, uci, false); +} + +void __init load_ucode_intel_bsp(void) +{ + struct microcode_intel *patch; + struct ucode_cpu_info uci; + + patch = __load_ucode_intel(&uci); + if (!patch) + return; + + uci.mc = patch; + + apply_microcode_early(&uci, true); +} + +void load_ucode_intel_ap(void) +{ + struct microcode_intel *patch, **iup; + struct ucode_cpu_info uci; + + if (IS_ENABLED(CONFIG_X86_32)) + iup = (struct microcode_intel **) __pa_nodebug(&intel_ucode_patch); + else + iup = &intel_ucode_patch; + + if (!*iup) { + patch = __load_ucode_intel(&uci); + if (!patch) + return; + + *iup = patch; + } + + uci.mc = *iup; + + apply_microcode_early(&uci, true); +} + +static struct microcode_intel *find_patch(struct ucode_cpu_info *uci) +{ + struct microcode_header_intel *phdr; + struct ucode_patch *iter, *tmp; + + list_for_each_entry_safe(iter, tmp, µcode_cache, plist) { + + phdr = (struct microcode_header_intel *)iter->data; + + if (phdr->rev <= uci->cpu_sig.rev) + continue; + + if (!find_matching_signature(phdr, + uci->cpu_sig.sig, + uci->cpu_sig.pf)) + continue; + + return iter->data; + } + return NULL; +} + +void reload_ucode_intel(void) +{ + struct microcode_intel *p; + struct ucode_cpu_info uci; + + collect_cpu_info_early(&uci); + + p = find_patch(&uci); + if (!p) + return; + + uci.mc = p; + + apply_microcode_early(&uci, false); +} + +static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) +{ + static struct cpu_signature prev; + struct cpuinfo_x86 *c = &cpu_data(cpu_num); + unsigned int val[2]; + + memset(csig, 0, sizeof(*csig)); + + csig->sig = cpuid_eax(0x00000001); + + if ((c->x86_model >= 5) || (c->x86 > 6)) { + /* get processor flags from MSR 0x17 */ + rdmsr(MSR_IA32_PLATFORM_ID, val[0], val[1]); + csig->pf = 1 << ((val[1] >> 18) & 7); + } + + csig->rev = c->microcode; + + /* No extra locking on prev, races are harmless. */ + if (csig->sig != prev.sig || csig->pf != prev.pf || csig->rev != prev.rev) { + pr_info("sig=0x%x, pf=0x%x, revision=0x%x\n", + csig->sig, csig->pf, csig->rev); + prev = *csig; + } + + return 0; +} + +static enum ucode_state apply_microcode_intel(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct cpuinfo_x86 *c = &cpu_data(cpu); + bool bsp = c->cpu_index == boot_cpu_data.cpu_index; + struct microcode_intel *mc; + enum ucode_state ret; + static int prev_rev; + u32 rev; + + /* We should bind the task to the CPU */ + if (WARN_ON(raw_smp_processor_id() != cpu)) + return UCODE_ERROR; + + /* Look for a newer patch in our cache: */ + mc = find_patch(uci); + if (!mc) { + mc = uci->mc; + if (!mc) + return UCODE_NFOUND; + } + + /* + * Save us the MSR write below - which is a particular expensive + * operation - when the other hyperthread has updated the microcode + * already. + */ + rev = intel_get_microcode_revision(); + if (rev >= mc->hdr.rev) { + ret = UCODE_OK; + goto out; + } + + /* + * Writeback and invalidate caches before updating microcode to avoid + * internal issues depending on what the microcode is updating. + */ + native_wbinvd(); + + /* write microcode via MSR 0x79 */ + wrmsrl(MSR_IA32_UCODE_WRITE, (unsigned long)mc->bits); + + rev = intel_get_microcode_revision(); + + if (rev != mc->hdr.rev) { + pr_err("CPU%d update to revision 0x%x failed\n", + cpu, mc->hdr.rev); + return UCODE_ERROR; + } + + if (bsp && rev != prev_rev) { + pr_info("updated to revision 0x%x, date = %04x-%02x-%02x\n", + rev, + mc->hdr.date & 0xffff, + mc->hdr.date >> 24, + (mc->hdr.date >> 16) & 0xff); + prev_rev = rev; + } + + ret = UCODE_UPDATED; + +out: + uci->cpu_sig.rev = rev; + c->microcode = rev; + + /* Update boot_cpu_data's revision too, if we're on the BSP: */ + if (bsp) + boot_cpu_data.microcode = rev; + + return ret; +} + +static enum ucode_state generic_load_microcode(int cpu, struct iov_iter *iter) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + unsigned int curr_mc_size = 0, new_mc_size = 0; + enum ucode_state ret = UCODE_OK; + int new_rev = uci->cpu_sig.rev; + u8 *new_mc = NULL, *mc = NULL; + unsigned int csig, cpf; + + while (iov_iter_count(iter)) { + struct microcode_header_intel mc_header; + unsigned int mc_size, data_size; + u8 *data; + + if (!copy_from_iter_full(&mc_header, sizeof(mc_header), iter)) { + pr_err("error! Truncated or inaccessible header in microcode data file\n"); + break; + } + + mc_size = get_totalsize(&mc_header); + if (mc_size < sizeof(mc_header)) { + pr_err("error! Bad data in microcode data file (totalsize too small)\n"); + break; + } + data_size = mc_size - sizeof(mc_header); + if (data_size > iov_iter_count(iter)) { + pr_err("error! Bad data in microcode data file (truncated file?)\n"); + break; + } + + /* For performance reasons, reuse mc area when possible */ + if (!mc || mc_size > curr_mc_size) { + vfree(mc); + mc = vmalloc(mc_size); + if (!mc) + break; + curr_mc_size = mc_size; + } + + memcpy(mc, &mc_header, sizeof(mc_header)); + data = mc + sizeof(mc_header); + if (!copy_from_iter_full(data, data_size, iter) || + microcode_sanity_check(mc, 1) < 0) { + break; + } + + csig = uci->cpu_sig.sig; + cpf = uci->cpu_sig.pf; + if (has_newer_microcode(mc, csig, cpf, new_rev)) { + vfree(new_mc); + new_rev = mc_header.rev; + new_mc = mc; + new_mc_size = mc_size; + mc = NULL; /* trigger new vmalloc */ + ret = UCODE_NEW; + } + } + + vfree(mc); + + if (iov_iter_count(iter)) { + vfree(new_mc); + return UCODE_ERROR; + } + + if (!new_mc) + return UCODE_NFOUND; + + vfree(uci->mc); + uci->mc = (struct microcode_intel *)new_mc; + + /* + * If early loading microcode is supported, save this mc into + * permanent memory. So it will be loaded early when a CPU is hot added + * or resumes. + */ + save_mc_for_early(uci, new_mc, new_mc_size); + + pr_debug("CPU%d found a matching microcode update with version 0x%x (current=0x%x)\n", + cpu, new_rev, uci->cpu_sig.rev); + + return ret; +} + +static bool is_blacklisted(unsigned int cpu) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + + /* + * Late loading on model 79 with microcode revision less than 0x0b000021 + * and LLC size per core bigger than 2.5MB may result in a system hang. + * This behavior is documented in item BDF90, #334165 (Intel Xeon + * Processor E7-8800/4800 v4 Product Family). + */ + if (c->x86 == 6 && + c->x86_model == INTEL_FAM6_BROADWELL_X && + c->x86_stepping == 0x01 && + llc_size_per_core > 2621440 && + c->microcode < 0x0b000021) { + pr_err_once("Erratum BDF90: late loading with revision < 0x0b000021 (0x%x) disabled.\n", c->microcode); + pr_err_once("Please consider either early loading through initrd/built-in or a potential BIOS update.\n"); + return true; + } + + return false; +} + +static enum ucode_state request_microcode_fw(int cpu, struct device *device, + bool refresh_fw) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + const struct firmware *firmware; + struct iov_iter iter; + enum ucode_state ret; + struct kvec kvec; + char name[30]; + + if (is_blacklisted(cpu)) + return UCODE_NFOUND; + + sprintf(name, "intel-ucode/%02x-%02x-%02x", + c->x86, c->x86_model, c->x86_stepping); + + if (request_firmware_direct(&firmware, name, device)) { + pr_debug("data file %s load failed\n", name); + return UCODE_NFOUND; + } + + kvec.iov_base = (void *)firmware->data; + kvec.iov_len = firmware->size; + iov_iter_kvec(&iter, WRITE, &kvec, 1, firmware->size); + ret = generic_load_microcode(cpu, &iter); + + release_firmware(firmware); + + return ret; +} + +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) +{ + struct iov_iter iter; + struct iovec iov; + + if (is_blacklisted(cpu)) + return UCODE_NFOUND; + + iov.iov_base = (void __user *)buf; + iov.iov_len = size; + iov_iter_init(&iter, WRITE, &iov, 1, size); + + return generic_load_microcode(cpu, &iter); +} + +static struct microcode_ops microcode_intel_ops = { + .request_microcode_user = request_microcode_user, + .request_microcode_fw = request_microcode_fw, + .collect_cpu_info = collect_cpu_info, + .apply_microcode = apply_microcode_intel, +}; + +static int __init calc_llc_size_per_core(struct cpuinfo_x86 *c) +{ + u64 llc_size = c->x86_cache_size * 1024ULL; + + do_div(llc_size, c->x86_max_cores); + + return (int)llc_size; +} + +struct microcode_ops * __init init_intel_microcode(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + if (c->x86_vendor != X86_VENDOR_INTEL || c->x86 < 6 || + cpu_has(c, X86_FEATURE_IA64)) { + pr_err("Intel CPU family 0x%x not supported\n", c->x86); + return NULL; + } + + llc_size_per_core = calc_llc_size_per_core(c); + + return µcode_intel_ops; +} diff --git a/arch/x86/kernel/cpu/mkcapflags.sh b/arch/x86/kernel/cpu/mkcapflags.sh new file mode 100644 index 000000000..1db560ed2 --- /dev/null +++ b/arch/x86/kernel/cpu/mkcapflags.sh @@ -0,0 +1,74 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# Generate the x86_cap/bug_flags[] arrays from include/asm/cpufeatures.h +# + +set -e + +OUT=$1 + +dump_array() +{ + ARRAY=$1 + SIZE=$2 + PFX=$3 + POSTFIX=$4 + IN=$5 + + PFX_SZ=$(echo $PFX | wc -c) + TABS="$(printf '\t\t\t\t\t')" + + echo "const char * const $ARRAY[$SIZE] = {" + + # Iterate through any input lines starting with #define $PFX + sed -n -e 's/\t/ /g' -e "s/^ *# *define *$PFX//p" $IN | + while read i + do + # Name is everything up to the first whitespace + NAME="$(echo "$i" | sed 's/ .*//')" + + # If the /* comment */ starts with a quote string, grab that. + VALUE="$(echo "$i" | sed -n 's@.*/\* *\("[^"]*"\).*\*/@\1@p')" + [ -z "$VALUE" ] && VALUE="\"$NAME\"" + [ "$VALUE" = '""' ] && continue + + # Name is uppercase, VALUE is all lowercase + VALUE="$(echo "$VALUE" | tr A-Z a-z)" + + if [ -n "$POSTFIX" ]; then + T=$(( $PFX_SZ + $(echo $POSTFIX | wc -c) + 2 )) + TABS="$(printf '\t\t\t\t\t\t')" + TABCOUNT=$(( ( 6*8 - ($T + 1) - $(echo "$NAME" | wc -c) ) / 8 )) + printf "\t[%s - %s]%.*s = %s,\n" "$PFX$NAME" "$POSTFIX" "$TABCOUNT" "$TABS" "$VALUE" + else + TABCOUNT=$(( ( 5*8 - ($PFX_SZ + 1) - $(echo "$NAME" | wc -c) ) / 8 )) + printf "\t[%s]%.*s = %s,\n" "$PFX$NAME" "$TABCOUNT" "$TABS" "$VALUE" + fi + done + echo "};" +} + +trap 'rm "$OUT"' EXIT + +( + echo "#ifndef _ASM_X86_CPUFEATURES_H" + echo "#include <asm/cpufeatures.h>" + echo "#endif" + echo "" + + dump_array "x86_cap_flags" "NCAPINTS*32" "X86_FEATURE_" "" $2 + echo "" + + dump_array "x86_bug_flags" "NBUGINTS*32" "X86_BUG_" "NCAPINTS*32" $2 + echo "" + + echo "#ifdef CONFIG_X86_VMX_FEATURE_NAMES" + echo "#ifndef _ASM_X86_VMXFEATURES_H" + echo "#include <asm/vmxfeatures.h>" + echo "#endif" + dump_array "x86_vmx_flags" "NVMXINTS*32" "VMX_FEATURE_" "" $3 + echo "#endif /* CONFIG_X86_VMX_FEATURE_NAMES */" +) > $OUT + +trap - EXIT diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c new file mode 100644 index 000000000..021cd0677 --- /dev/null +++ b/arch/x86/kernel/cpu/mshyperv.c @@ -0,0 +1,397 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * HyperV Detection code. + * + * Copyright (C) 2010, Novell, Inc. + * Author : K. Y. Srinivasan <ksrinivasan@novell.com> + */ + +#include <linux/types.h> +#include <linux/time.h> +#include <linux/clocksource.h> +#include <linux/init.h> +#include <linux/export.h> +#include <linux/hardirq.h> +#include <linux/efi.h> +#include <linux/interrupt.h> +#include <linux/irq.h> +#include <linux/kexec.h> +#include <linux/i8253.h> +#include <linux/random.h> +#include <asm/processor.h> +#include <asm/hypervisor.h> +#include <asm/hyperv-tlfs.h> +#include <asm/mshyperv.h> +#include <asm/desc.h> +#include <asm/idtentry.h> +#include <asm/irq_regs.h> +#include <asm/i8259.h> +#include <asm/apic.h> +#include <asm/timer.h> +#include <asm/reboot.h> +#include <asm/nmi.h> +#include <clocksource/hyperv_timer.h> + +struct ms_hyperv_info ms_hyperv; +EXPORT_SYMBOL_GPL(ms_hyperv); + +#if IS_ENABLED(CONFIG_HYPERV) +static void (*vmbus_handler)(void); +static void (*hv_stimer0_handler)(void); +static void (*hv_kexec_handler)(void); +static void (*hv_crash_handler)(struct pt_regs *regs); + +DEFINE_IDTENTRY_SYSVEC(sysvec_hyperv_callback) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + inc_irq_stat(irq_hv_callback_count); + if (vmbus_handler) + vmbus_handler(); + + if (ms_hyperv.hints & HV_DEPRECATING_AEOI_RECOMMENDED) + ack_APIC_irq(); + + set_irq_regs(old_regs); +} + +int hv_setup_vmbus_irq(int irq, void (*handler)(void)) +{ + /* + * The 'irq' argument is ignored on x86/x64 because a hard-coded + * interrupt vector is used for Hyper-V interrupts. + */ + vmbus_handler = handler; + return 0; +} + +void hv_remove_vmbus_irq(void) +{ + /* We have no way to deallocate the interrupt gate */ + vmbus_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_setup_vmbus_irq); +EXPORT_SYMBOL_GPL(hv_remove_vmbus_irq); + +/* + * Routines to do per-architecture handling of stimer0 + * interrupts when in Direct Mode + */ +DEFINE_IDTENTRY_SYSVEC(sysvec_hyperv_stimer0) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + inc_irq_stat(hyperv_stimer0_count); + if (hv_stimer0_handler) + hv_stimer0_handler(); + add_interrupt_randomness(HYPERV_STIMER0_VECTOR); + ack_APIC_irq(); + + set_irq_regs(old_regs); +} + +int hv_setup_stimer0_irq(int *irq, int *vector, void (*handler)(void)) +{ + *vector = HYPERV_STIMER0_VECTOR; + *irq = -1; /* Unused on x86/x64 */ + hv_stimer0_handler = handler; + return 0; +} +EXPORT_SYMBOL_GPL(hv_setup_stimer0_irq); + +void hv_remove_stimer0_irq(int irq) +{ + /* We have no way to deallocate the interrupt gate */ + hv_stimer0_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_remove_stimer0_irq); + +void hv_setup_kexec_handler(void (*handler)(void)) +{ + hv_kexec_handler = handler; +} +EXPORT_SYMBOL_GPL(hv_setup_kexec_handler); + +void hv_remove_kexec_handler(void) +{ + hv_kexec_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_remove_kexec_handler); + +void hv_setup_crash_handler(void (*handler)(struct pt_regs *regs)) +{ + hv_crash_handler = handler; +} +EXPORT_SYMBOL_GPL(hv_setup_crash_handler); + +void hv_remove_crash_handler(void) +{ + hv_crash_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_remove_crash_handler); + +#ifdef CONFIG_KEXEC_CORE +static void hv_machine_shutdown(void) +{ + if (kexec_in_progress && hv_kexec_handler) + hv_kexec_handler(); + + /* + * Call hv_cpu_die() on all the CPUs, otherwise later the hypervisor + * corrupts the old VP Assist Pages and can crash the kexec kernel. + */ + if (kexec_in_progress && hyperv_init_cpuhp > 0) + cpuhp_remove_state(hyperv_init_cpuhp); + + /* The function calls stop_other_cpus(). */ + native_machine_shutdown(); + + /* Disable the hypercall page when there is only 1 active CPU. */ + if (kexec_in_progress) + hyperv_cleanup(); +} + +static void hv_machine_crash_shutdown(struct pt_regs *regs) +{ + if (hv_crash_handler) + hv_crash_handler(regs); + + /* The function calls crash_smp_send_stop(). */ + native_machine_crash_shutdown(regs); + + /* Disable the hypercall page when there is only 1 active CPU. */ + hyperv_cleanup(); +} +#endif /* CONFIG_KEXEC_CORE */ +#endif /* CONFIG_HYPERV */ + +static uint32_t __init ms_hyperv_platform(void) +{ + u32 eax; + u32 hyp_signature[3]; + + if (!boot_cpu_has(X86_FEATURE_HYPERVISOR)) + return 0; + + cpuid(HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS, + &eax, &hyp_signature[0], &hyp_signature[1], &hyp_signature[2]); + + if (eax >= HYPERV_CPUID_MIN && + eax <= HYPERV_CPUID_MAX && + !memcmp("Microsoft Hv", hyp_signature, 12)) + return HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS; + + return 0; +} + +static unsigned char hv_get_nmi_reason(void) +{ + return 0; +} + +#ifdef CONFIG_X86_LOCAL_APIC +/* + * Prior to WS2016 Debug-VM sends NMIs to all CPUs which makes + * it dificult to process CHANNELMSG_UNLOAD in case of crash. Handle + * unknown NMI on the first CPU which gets it. + */ +static int hv_nmi_unknown(unsigned int val, struct pt_regs *regs) +{ + static atomic_t nmi_cpu = ATOMIC_INIT(-1); + + if (!unknown_nmi_panic) + return NMI_DONE; + + if (atomic_cmpxchg(&nmi_cpu, -1, raw_smp_processor_id()) != -1) + return NMI_HANDLED; + + return NMI_DONE; +} +#endif + +static unsigned long hv_get_tsc_khz(void) +{ + unsigned long freq; + + rdmsrl(HV_X64_MSR_TSC_FREQUENCY, freq); + + return freq / 1000; +} + +#if defined(CONFIG_SMP) && IS_ENABLED(CONFIG_HYPERV) +static void __init hv_smp_prepare_boot_cpu(void) +{ + native_smp_prepare_boot_cpu(); +#if defined(CONFIG_X86_64) && defined(CONFIG_PARAVIRT_SPINLOCKS) + hv_init_spinlocks(); +#endif +} +#endif + +static void __init ms_hyperv_init_platform(void) +{ + int hv_host_info_eax; + int hv_host_info_ebx; + int hv_host_info_ecx; + int hv_host_info_edx; + +#ifdef CONFIG_PARAVIRT + pv_info.name = "Hyper-V"; +#endif + + /* + * Extract the features and hints + */ + ms_hyperv.features = cpuid_eax(HYPERV_CPUID_FEATURES); + ms_hyperv.misc_features = cpuid_edx(HYPERV_CPUID_FEATURES); + ms_hyperv.hints = cpuid_eax(HYPERV_CPUID_ENLIGHTMENT_INFO); + + pr_info("Hyper-V: features 0x%x, hints 0x%x, misc 0x%x\n", + ms_hyperv.features, ms_hyperv.hints, ms_hyperv.misc_features); + + ms_hyperv.max_vp_index = cpuid_eax(HYPERV_CPUID_IMPLEMENT_LIMITS); + ms_hyperv.max_lp_index = cpuid_ebx(HYPERV_CPUID_IMPLEMENT_LIMITS); + + pr_debug("Hyper-V: max %u virtual processors, %u logical processors\n", + ms_hyperv.max_vp_index, ms_hyperv.max_lp_index); + + /* + * Extract host information. + */ + if (cpuid_eax(HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS) >= + HYPERV_CPUID_VERSION) { + hv_host_info_eax = cpuid_eax(HYPERV_CPUID_VERSION); + hv_host_info_ebx = cpuid_ebx(HYPERV_CPUID_VERSION); + hv_host_info_ecx = cpuid_ecx(HYPERV_CPUID_VERSION); + hv_host_info_edx = cpuid_edx(HYPERV_CPUID_VERSION); + + pr_info("Hyper-V Host Build:%d-%d.%d-%d-%d.%d\n", + hv_host_info_eax, hv_host_info_ebx >> 16, + hv_host_info_ebx & 0xFFFF, hv_host_info_ecx, + hv_host_info_edx >> 24, hv_host_info_edx & 0xFFFFFF); + } + + if (ms_hyperv.features & HV_ACCESS_FREQUENCY_MSRS && + ms_hyperv.misc_features & HV_FEATURE_FREQUENCY_MSRS_AVAILABLE) { + x86_platform.calibrate_tsc = hv_get_tsc_khz; + x86_platform.calibrate_cpu = hv_get_tsc_khz; + } + + if (ms_hyperv.hints & HV_X64_ENLIGHTENED_VMCS_RECOMMENDED) { + ms_hyperv.nested_features = + cpuid_eax(HYPERV_CPUID_NESTED_FEATURES); + } + + /* + * Hyper-V expects to get crash register data or kmsg when + * crash enlightment is available and system crashes. Set + * crash_kexec_post_notifiers to be true to make sure that + * calling crash enlightment interface before running kdump + * kernel. + */ + if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) + crash_kexec_post_notifiers = true; + +#ifdef CONFIG_X86_LOCAL_APIC + if (ms_hyperv.features & HV_ACCESS_FREQUENCY_MSRS && + ms_hyperv.misc_features & HV_FEATURE_FREQUENCY_MSRS_AVAILABLE) { + /* + * Get the APIC frequency. + */ + u64 hv_lapic_frequency; + + rdmsrl(HV_X64_MSR_APIC_FREQUENCY, hv_lapic_frequency); + hv_lapic_frequency = div_u64(hv_lapic_frequency, HZ); + lapic_timer_period = hv_lapic_frequency; + pr_info("Hyper-V: LAPIC Timer Frequency: %#x\n", + lapic_timer_period); + } + + register_nmi_handler(NMI_UNKNOWN, hv_nmi_unknown, NMI_FLAG_FIRST, + "hv_nmi_unknown"); +#endif + +#ifdef CONFIG_X86_IO_APIC + no_timer_check = 1; +#endif + +#if IS_ENABLED(CONFIG_HYPERV) && defined(CONFIG_KEXEC_CORE) + machine_ops.shutdown = hv_machine_shutdown; + machine_ops.crash_shutdown = hv_machine_crash_shutdown; +#endif + if (ms_hyperv.features & HV_ACCESS_TSC_INVARIANT) { + wrmsrl(HV_X64_MSR_TSC_INVARIANT_CONTROL, 0x1); + setup_force_cpu_cap(X86_FEATURE_TSC_RELIABLE); + } + + /* + * Generation 2 instances don't support reading the NMI status from + * 0x61 port. + */ + if (efi_enabled(EFI_BOOT)) + x86_platform.get_nmi_reason = hv_get_nmi_reason; + + /* + * Hyper-V VMs have a PIT emulation quirk such that zeroing the + * counter register during PIT shutdown restarts the PIT. So it + * continues to interrupt @18.2 HZ. Setting i8253_clear_counter + * to false tells pit_shutdown() not to zero the counter so that + * the PIT really is shutdown. Generation 2 VMs don't have a PIT, + * and setting this value has no effect. + */ + i8253_clear_counter_on_shutdown = false; + +#if IS_ENABLED(CONFIG_HYPERV) + /* + * Setup the hook to get control post apic initialization. + */ + x86_platform.apic_post_init = hyperv_init; + hyperv_setup_mmu_ops(); + /* Setup the IDT for hypervisor callback */ + alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, asm_sysvec_hyperv_callback); + + /* Setup the IDT for reenlightenment notifications */ + if (ms_hyperv.features & HV_ACCESS_REENLIGHTENMENT) { + alloc_intr_gate(HYPERV_REENLIGHTENMENT_VECTOR, + asm_sysvec_hyperv_reenlightenment); + } + + /* Setup the IDT for stimer0 */ + if (ms_hyperv.misc_features & HV_STIMER_DIRECT_MODE_AVAILABLE) { + alloc_intr_gate(HYPERV_STIMER0_VECTOR, + asm_sysvec_hyperv_stimer0); + } + +# ifdef CONFIG_SMP + smp_ops.smp_prepare_boot_cpu = hv_smp_prepare_boot_cpu; +# endif + + /* + * Hyper-V doesn't provide irq remapping for IO-APIC. To enable x2apic, + * set x2apic destination mode to physcial mode when x2apic is available + * and Hyper-V IOMMU driver makes sure cpus assigned with IO-APIC irqs + * have 8-bit APIC id. + */ +# ifdef CONFIG_X86_X2APIC + if (x2apic_supported()) + x2apic_phys = 1; +# endif + + /* Register Hyper-V specific clocksource */ + hv_init_clocksource(); +#endif + /* + * TSC should be marked as unstable only after Hyper-V + * clocksource has been initialized. This ensures that the + * stability of the sched_clock is not altered. + */ + if (!(ms_hyperv.features & HV_ACCESS_TSC_INVARIANT)) + mark_tsc_unstable("running on Hyper-V"); +} + +const __initconst struct hypervisor_x86 x86_hyper_ms_hyperv = { + .name = "Microsoft Hyper-V", + .detect = ms_hyperv_platform, + .type = X86_HYPER_MS_HYPERV, + .init.init_platform = ms_hyperv_init_platform, +}; diff --git a/arch/x86/kernel/cpu/mtrr/Makefile b/arch/x86/kernel/cpu/mtrr/Makefile new file mode 100644 index 000000000..cc4f9f1cb --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/Makefile @@ -0,0 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0-only +obj-y := mtrr.o if.o generic.o cleanup.o +obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o + diff --git a/arch/x86/kernel/cpu/mtrr/amd.c b/arch/x86/kernel/cpu/mtrr/amd.c new file mode 100644 index 000000000..a65a02720 --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/amd.c @@ -0,0 +1,125 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/init.h> +#include <linux/mm.h> +#include <asm/mtrr.h> +#include <asm/msr.h> + +#include "mtrr.h" + +static void +amd_get_mtrr(unsigned int reg, unsigned long *base, + unsigned long *size, mtrr_type *type) +{ + unsigned long low, high; + + rdmsr(MSR_K6_UWCCR, low, high); + /* Upper dword is region 1, lower is region 0 */ + if (reg == 1) + low = high; + /* The base masks off on the right alignment */ + *base = (low & 0xFFFE0000) >> PAGE_SHIFT; + *type = 0; + if (low & 1) + *type = MTRR_TYPE_UNCACHABLE; + if (low & 2) + *type = MTRR_TYPE_WRCOMB; + if (!(low & 3)) { + *size = 0; + return; + } + /* + * This needs a little explaining. The size is stored as an + * inverted mask of bits of 128K granularity 15 bits long offset + * 2 bits. + * + * So to get a size we do invert the mask and add 1 to the lowest + * mask bit (4 as its 2 bits in). This gives us a size we then shift + * to turn into 128K blocks. + * + * eg 111 1111 1111 1100 is 512K + * + * invert 000 0000 0000 0011 + * +1 000 0000 0000 0100 + * *128K ... + */ + low = (~low) & 0x1FFFC; + *size = (low + 4) << (15 - PAGE_SHIFT); +} + +/** + * amd_set_mtrr - Set variable MTRR register on the local CPU. + * + * @reg The register to set. + * @base The base address of the region. + * @size The size of the region. If this is 0 the region is disabled. + * @type The type of the region. + * + * Returns nothing. + */ +static void +amd_set_mtrr(unsigned int reg, unsigned long base, unsigned long size, mtrr_type type) +{ + u32 regs[2]; + + /* + * Low is MTRR0, High MTRR 1 + */ + rdmsr(MSR_K6_UWCCR, regs[0], regs[1]); + /* + * Blank to disable + */ + if (size == 0) { + regs[reg] = 0; + } else { + /* + * Set the register to the base, the type (off by one) and an + * inverted bitmask of the size The size is the only odd + * bit. We are fed say 512K We invert this and we get 111 1111 + * 1111 1011 but if you subtract one and invert you get the + * desired 111 1111 1111 1100 mask + * + * But ~(x - 1) == ~x + 1 == -x. Two's complement rocks! + */ + regs[reg] = (-size >> (15 - PAGE_SHIFT) & 0x0001FFFC) + | (base << PAGE_SHIFT) | (type + 1); + } + + /* + * The writeback rule is quite specific. See the manual. Its + * disable local interrupts, write back the cache, set the mtrr + */ + wbinvd(); + wrmsr(MSR_K6_UWCCR, regs[0], regs[1]); +} + +static int +amd_validate_add_page(unsigned long base, unsigned long size, unsigned int type) +{ + /* + * Apply the K6 block alignment and size rules + * In order + * o Uncached or gathering only + * o 128K or bigger block + * o Power of 2 block + * o base suitably aligned to the power + */ + if (type > MTRR_TYPE_WRCOMB || size < (1 << (17 - PAGE_SHIFT)) + || (size & ~(size - 1)) - size || (base & (size - 1))) + return -EINVAL; + return 0; +} + +static const struct mtrr_ops amd_mtrr_ops = { + .vendor = X86_VENDOR_AMD, + .set = amd_set_mtrr, + .get = amd_get_mtrr, + .get_free_region = generic_get_free_region, + .validate_add_page = amd_validate_add_page, + .have_wrcomb = positive_have_wrcomb, +}; + +int __init amd_init_mtrr(void) +{ + set_mtrr_ops(&amd_mtrr_ops); + return 0; +} diff --git a/arch/x86/kernel/cpu/mtrr/centaur.c b/arch/x86/kernel/cpu/mtrr/centaur.c new file mode 100644 index 000000000..f27177816 --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/centaur.c @@ -0,0 +1,127 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/init.h> +#include <linux/mm.h> + +#include <asm/mtrr.h> +#include <asm/msr.h> + +#include "mtrr.h" + +static struct { + unsigned long high; + unsigned long low; +} centaur_mcr[8]; + +static u8 centaur_mcr_reserved; +static u8 centaur_mcr_type; /* 0 for winchip, 1 for winchip2 */ + +/** + * centaur_get_free_region - Get a free MTRR. + * + * @base: The starting (base) address of the region. + * @size: The size (in bytes) of the region. + * + * Returns: the index of the region on success, else -1 on error. + */ +static int +centaur_get_free_region(unsigned long base, unsigned long size, int replace_reg) +{ + unsigned long lbase, lsize; + mtrr_type ltype; + int i, max; + + max = num_var_ranges; + if (replace_reg >= 0 && replace_reg < max) + return replace_reg; + + for (i = 0; i < max; ++i) { + if (centaur_mcr_reserved & (1 << i)) + continue; + mtrr_if->get(i, &lbase, &lsize, <ype); + if (lsize == 0) + return i; + } + + return -ENOSPC; +} + +/* + * Report boot time MCR setups + */ +void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) +{ + centaur_mcr[mcr].low = lo; + centaur_mcr[mcr].high = hi; +} + +static void +centaur_get_mcr(unsigned int reg, unsigned long *base, + unsigned long *size, mtrr_type * type) +{ + *base = centaur_mcr[reg].high >> PAGE_SHIFT; + *size = -(centaur_mcr[reg].low & 0xfffff000) >> PAGE_SHIFT; + *type = MTRR_TYPE_WRCOMB; /* write-combining */ + + if (centaur_mcr_type == 1 && ((centaur_mcr[reg].low & 31) & 2)) + *type = MTRR_TYPE_UNCACHABLE; + if (centaur_mcr_type == 1 && (centaur_mcr[reg].low & 31) == 25) + *type = MTRR_TYPE_WRBACK; + if (centaur_mcr_type == 0 && (centaur_mcr[reg].low & 31) == 31) + *type = MTRR_TYPE_WRBACK; +} + +static void +centaur_set_mcr(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type) +{ + unsigned long low, high; + + if (size == 0) { + /* Disable */ + high = low = 0; + } else { + high = base << PAGE_SHIFT; + if (centaur_mcr_type == 0) { + /* Only support write-combining... */ + low = -size << PAGE_SHIFT | 0x1f; + } else { + if (type == MTRR_TYPE_UNCACHABLE) + low = -size << PAGE_SHIFT | 0x02; /* NC */ + else + low = -size << PAGE_SHIFT | 0x09; /* WWO, WC */ + } + } + centaur_mcr[reg].high = high; + centaur_mcr[reg].low = low; + wrmsr(MSR_IDT_MCR0 + reg, low, high); +} + +static int +centaur_validate_add_page(unsigned long base, unsigned long size, unsigned int type) +{ + /* + * FIXME: Winchip2 supports uncached + */ + if (type != MTRR_TYPE_WRCOMB && + (centaur_mcr_type == 0 || type != MTRR_TYPE_UNCACHABLE)) { + pr_warn("mtrr: only write-combining%s supported\n", + centaur_mcr_type ? " and uncacheable are" : " is"); + return -EINVAL; + } + return 0; +} + +static const struct mtrr_ops centaur_mtrr_ops = { + .vendor = X86_VENDOR_CENTAUR, + .set = centaur_set_mcr, + .get = centaur_get_mcr, + .get_free_region = centaur_get_free_region, + .validate_add_page = centaur_validate_add_page, + .have_wrcomb = positive_have_wrcomb, +}; + +int __init centaur_init_mtrr(void) +{ + set_mtrr_ops(¢aur_mtrr_ops); + return 0; +} diff --git a/arch/x86/kernel/cpu/mtrr/cleanup.c b/arch/x86/kernel/cpu/mtrr/cleanup.c new file mode 100644 index 000000000..5bd011737 --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/cleanup.c @@ -0,0 +1,987 @@ +/* + * MTRR (Memory Type Range Register) cleanup + * + * Copyright (C) 2009 Yinghai Lu + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Library General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Library General Public License for more details. + * + * You should have received a copy of the GNU Library General Public + * License along with this library; if not, write to the Free + * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/smp.h> +#include <linux/cpu.h> +#include <linux/mutex.h> +#include <linux/uaccess.h> +#include <linux/kvm_para.h> +#include <linux/range.h> + +#include <asm/processor.h> +#include <asm/e820/api.h> +#include <asm/mtrr.h> +#include <asm/msr.h> + +#include "mtrr.h" + +struct var_mtrr_range_state { + unsigned long base_pfn; + unsigned long size_pfn; + mtrr_type type; +}; + +struct var_mtrr_state { + unsigned long range_startk; + unsigned long range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + unsigned int reg; +}; + +/* Should be related to MTRR_VAR_RANGES nums */ +#define RANGE_NUM 256 + +static struct range __initdata range[RANGE_NUM]; +static int __initdata nr_range; + +static struct var_mtrr_range_state __initdata range_state[RANGE_NUM]; + +static int __initdata debug_print; +#define Dprintk(x...) do { if (debug_print) pr_debug(x); } while (0) + +#define BIOS_BUG_MSG \ + "WARNING: BIOS bug: VAR MTRR %d contains strange UC entry under 1M, check with your system vendor!\n" + +static int __init +x86_get_mtrr_mem_range(struct range *range, int nr_range, + unsigned long extra_remove_base, + unsigned long extra_remove_size) +{ + unsigned long base, size; + mtrr_type type; + int i; + + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_WRBACK) + continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; + nr_range = add_range_with_merge(range, RANGE_NUM, nr_range, + base, base + size); + } + if (debug_print) { + pr_debug("After WB checking\n"); + for (i = 0; i < nr_range; i++) + pr_debug("MTRR MAP PFN: %016llx - %016llx\n", + range[i].start, range[i].end); + } + + /* Take out UC ranges: */ + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_UNCACHABLE && + type != MTRR_TYPE_WRPROT) + continue; + size = range_state[i].size_pfn; + if (!size) + continue; + base = range_state[i].base_pfn; + if (base < (1<<(20-PAGE_SHIFT)) && mtrr_state.have_fixed && + (mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED) && + (mtrr_state.enabled & MTRR_STATE_MTRR_FIXED_ENABLED)) { + /* Var MTRR contains UC entry below 1M? Skip it: */ + pr_warn(BIOS_BUG_MSG, i); + if (base + size <= (1<<(20-PAGE_SHIFT))) + continue; + size -= (1<<(20-PAGE_SHIFT)) - base; + base = 1<<(20-PAGE_SHIFT); + } + subtract_range(range, RANGE_NUM, base, base + size); + } + if (extra_remove_size) + subtract_range(range, RANGE_NUM, extra_remove_base, + extra_remove_base + extra_remove_size); + + if (debug_print) { + pr_debug("After UC checking\n"); + for (i = 0; i < RANGE_NUM; i++) { + if (!range[i].end) + continue; + pr_debug("MTRR MAP PFN: %016llx - %016llx\n", + range[i].start, range[i].end); + } + } + + /* sort the ranges */ + nr_range = clean_sort_range(range, RANGE_NUM); + if (debug_print) { + pr_debug("After sorting\n"); + for (i = 0; i < nr_range; i++) + pr_debug("MTRR MAP PFN: %016llx - %016llx\n", + range[i].start, range[i].end); + } + + return nr_range; +} + +#ifdef CONFIG_MTRR_SANITIZER + +static unsigned long __init sum_ranges(struct range *range, int nr_range) +{ + unsigned long sum = 0; + int i; + + for (i = 0; i < nr_range; i++) + sum += range[i].end - range[i].start; + + return sum; +} + +static int enable_mtrr_cleanup __initdata = + CONFIG_MTRR_SANITIZER_ENABLE_DEFAULT; + +static int __init disable_mtrr_cleanup_setup(char *str) +{ + enable_mtrr_cleanup = 0; + return 0; +} +early_param("disable_mtrr_cleanup", disable_mtrr_cleanup_setup); + +static int __init enable_mtrr_cleanup_setup(char *str) +{ + enable_mtrr_cleanup = 1; + return 0; +} +early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup); + +static int __init mtrr_cleanup_debug_setup(char *str) +{ + debug_print = 1; + return 0; +} +early_param("mtrr_cleanup_debug", mtrr_cleanup_debug_setup); + +static void __init +set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type, unsigned int address_bits) +{ + u32 base_lo, base_hi, mask_lo, mask_hi; + u64 base, mask; + + if (!sizek) { + fill_mtrr_var_range(reg, 0, 0, 0, 0); + return; + } + + mask = (1ULL << address_bits) - 1; + mask &= ~((((u64)sizek) << 10) - 1); + + base = ((u64)basek) << 10; + + base |= type; + mask |= 0x800; + + base_lo = base & ((1ULL<<32) - 1); + base_hi = base >> 32; + + mask_lo = mask & ((1ULL<<32) - 1); + mask_hi = mask >> 32; + + fill_mtrr_var_range(reg, base_lo, base_hi, mask_lo, mask_hi); +} + +static void __init +save_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type) +{ + range_state[reg].base_pfn = basek >> (PAGE_SHIFT - 10); + range_state[reg].size_pfn = sizek >> (PAGE_SHIFT - 10); + range_state[reg].type = type; +} + +static void __init set_var_mtrr_all(unsigned int address_bits) +{ + unsigned long basek, sizek; + unsigned char type; + unsigned int reg; + + for (reg = 0; reg < num_var_ranges; reg++) { + basek = range_state[reg].base_pfn << (PAGE_SHIFT - 10); + sizek = range_state[reg].size_pfn << (PAGE_SHIFT - 10); + type = range_state[reg].type; + + set_var_mtrr(reg, basek, sizek, type, address_bits); + } +} + +static unsigned long to_size_factor(unsigned long sizek, char *factorp) +{ + unsigned long base = sizek; + char factor; + + if (base & ((1<<10) - 1)) { + /* Not MB-aligned: */ + factor = 'K'; + } else if (base & ((1<<20) - 1)) { + factor = 'M'; + base >>= 10; + } else { + factor = 'G'; + base >>= 20; + } + + *factorp = factor; + + return base; +} + +static unsigned int __init +range_to_mtrr(unsigned int reg, unsigned long range_startk, + unsigned long range_sizek, unsigned char type) +{ + if (!range_sizek || (reg >= num_var_ranges)) + return reg; + + while (range_sizek) { + unsigned long max_align, align; + unsigned long sizek; + + /* Compute the maximum size with which we can make a range: */ + if (range_startk) + max_align = __ffs(range_startk); + else + max_align = BITS_PER_LONG - 1; + + align = __fls(range_sizek); + if (align > max_align) + align = max_align; + + sizek = 1UL << align; + if (debug_print) { + char start_factor = 'K', size_factor = 'K'; + unsigned long start_base, size_base; + + start_base = to_size_factor(range_startk, &start_factor); + size_base = to_size_factor(sizek, &size_factor); + + Dprintk("Setting variable MTRR %d, " + "base: %ld%cB, range: %ld%cB, type %s\n", + reg, start_base, start_factor, + size_base, size_factor, + (type == MTRR_TYPE_UNCACHABLE) ? "UC" : + ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other") + ); + } + save_var_mtrr(reg++, range_startk, sizek, type); + range_startk += sizek; + range_sizek -= sizek; + if (reg >= num_var_ranges) + break; + } + return reg; +} + +static unsigned __init +range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek, + unsigned long sizek) +{ + unsigned long hole_basek, hole_sizek; + unsigned long second_sizek; + unsigned long range0_basek, range0_sizek; + unsigned long range_basek, range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + + hole_basek = 0; + hole_sizek = 0; + second_sizek = 0; + chunk_sizek = state->chunk_sizek; + gran_sizek = state->gran_sizek; + + /* Align with gran size, prevent small block used up MTRRs: */ + range_basek = ALIGN(state->range_startk, gran_sizek); + if ((range_basek > basek) && basek) + return second_sizek; + + state->range_sizek -= (range_basek - state->range_startk); + range_sizek = ALIGN(state->range_sizek, gran_sizek); + + while (range_sizek > state->range_sizek) { + range_sizek -= gran_sizek; + if (!range_sizek) + return 0; + } + state->range_sizek = range_sizek; + + /* Try to append some small hole: */ + range0_basek = state->range_startk; + range0_sizek = ALIGN(state->range_sizek, chunk_sizek); + + /* No increase: */ + if (range0_sizek == state->range_sizek) { + Dprintk("rangeX: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + state->range_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + state->range_sizek, MTRR_TYPE_WRBACK); + return 0; + } + + /* Only cut back when it is not the last: */ + if (sizek) { + while (range0_basek + range0_sizek > (basek + sizek)) { + if (range0_sizek >= chunk_sizek) + range0_sizek -= chunk_sizek; + else + range0_sizek = 0; + + if (!range0_sizek) + break; + } + } + +second_try: + range_basek = range0_basek + range0_sizek; + + /* One hole in the middle: */ + if (range_basek > basek && range_basek <= (basek + sizek)) + second_sizek = range_basek - basek; + + if (range0_sizek > state->range_sizek) { + + /* One hole in middle or at the end: */ + hole_sizek = range0_sizek - state->range_sizek - second_sizek; + + /* Hole size should be less than half of range0 size: */ + if (hole_sizek >= (range0_sizek >> 1) && + range0_sizek >= chunk_sizek) { + range0_sizek -= chunk_sizek; + second_sizek = 0; + hole_sizek = 0; + + goto second_try; + } + } + + if (range0_sizek) { + Dprintk("range0: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + range0_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + range0_sizek, MTRR_TYPE_WRBACK); + } + + if (range0_sizek < state->range_sizek) { + /* Need to handle left over range: */ + range_sizek = state->range_sizek - range0_sizek; + + Dprintk("range: %016lx - %016lx\n", + range_basek<<10, + (range_basek + range_sizek)<<10); + + state->reg = range_to_mtrr(state->reg, range_basek, + range_sizek, MTRR_TYPE_WRBACK); + } + + if (hole_sizek) { + hole_basek = range_basek - hole_sizek - second_sizek; + Dprintk("hole: %016lx - %016lx\n", + hole_basek<<10, + (hole_basek + hole_sizek)<<10); + state->reg = range_to_mtrr(state->reg, hole_basek, + hole_sizek, MTRR_TYPE_UNCACHABLE); + } + + return second_sizek; +} + +static void __init +set_var_mtrr_range(struct var_mtrr_state *state, unsigned long base_pfn, + unsigned long size_pfn) +{ + unsigned long basek, sizek; + unsigned long second_sizek = 0; + + if (state->reg >= num_var_ranges) + return; + + basek = base_pfn << (PAGE_SHIFT - 10); + sizek = size_pfn << (PAGE_SHIFT - 10); + + /* See if I can merge with the last range: */ + if ((basek <= 1024) || + (state->range_startk + state->range_sizek == basek)) { + unsigned long endk = basek + sizek; + state->range_sizek = endk - state->range_startk; + return; + } + /* Write the range mtrrs: */ + if (state->range_sizek != 0) + second_sizek = range_to_mtrr_with_hole(state, basek, sizek); + + /* Allocate an msr: */ + state->range_startk = basek + second_sizek; + state->range_sizek = sizek - second_sizek; +} + +/* Mininum size of mtrr block that can take hole: */ +static u64 mtrr_chunk_size __initdata = (256ULL<<20); + +static int __init parse_mtrr_chunk_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_chunk_size = memparse(p, &p); + return 0; +} +early_param("mtrr_chunk_size", parse_mtrr_chunk_size_opt); + +/* Granularity of mtrr of block: */ +static u64 mtrr_gran_size __initdata; + +static int __init parse_mtrr_gran_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_gran_size = memparse(p, &p); + return 0; +} +early_param("mtrr_gran_size", parse_mtrr_gran_size_opt); + +static unsigned long nr_mtrr_spare_reg __initdata = + CONFIG_MTRR_SANITIZER_SPARE_REG_NR_DEFAULT; + +static int __init parse_mtrr_spare_reg(char *arg) +{ + if (arg) + nr_mtrr_spare_reg = simple_strtoul(arg, NULL, 0); + return 0; +} +early_param("mtrr_spare_reg_nr", parse_mtrr_spare_reg); + +static int __init +x86_setup_var_mtrrs(struct range *range, int nr_range, + u64 chunk_size, u64 gran_size) +{ + struct var_mtrr_state var_state; + int num_reg; + int i; + + var_state.range_startk = 0; + var_state.range_sizek = 0; + var_state.reg = 0; + var_state.chunk_sizek = chunk_size >> 10; + var_state.gran_sizek = gran_size >> 10; + + memset(range_state, 0, sizeof(range_state)); + + /* Write the range: */ + for (i = 0; i < nr_range; i++) { + set_var_mtrr_range(&var_state, range[i].start, + range[i].end - range[i].start); + } + + /* Write the last range: */ + if (var_state.range_sizek != 0) + range_to_mtrr_with_hole(&var_state, 0, 0); + + num_reg = var_state.reg; + /* Clear out the extra MTRR's: */ + while (var_state.reg < num_var_ranges) { + save_var_mtrr(var_state.reg, 0, 0, 0); + var_state.reg++; + } + + return num_reg; +} + +struct mtrr_cleanup_result { + unsigned long gran_sizek; + unsigned long chunk_sizek; + unsigned long lose_cover_sizek; + unsigned int num_reg; + int bad; +}; + +/* + * gran_size: 64K, 128K, 256K, 512K, 1M, 2M, ..., 2G + * chunk size: gran_size, ..., 2G + * so we need (1+16)*8 + */ +#define NUM_RESULT 136 +#define PSHIFT (PAGE_SHIFT - 10) + +static struct mtrr_cleanup_result __initdata result[NUM_RESULT]; +static unsigned long __initdata min_loss_pfn[RANGE_NUM]; + +static void __init print_out_mtrr_range_state(void) +{ + char start_factor = 'K', size_factor = 'K'; + unsigned long start_base, size_base; + mtrr_type type; + int i; + + for (i = 0; i < num_var_ranges; i++) { + + size_base = range_state[i].size_pfn << (PAGE_SHIFT - 10); + if (!size_base) + continue; + + size_base = to_size_factor(size_base, &size_factor), + start_base = range_state[i].base_pfn << (PAGE_SHIFT - 10); + start_base = to_size_factor(start_base, &start_factor), + type = range_state[i].type; + + pr_debug("reg %d, base: %ld%cB, range: %ld%cB, type %s\n", + i, start_base, start_factor, + size_base, size_factor, + (type == MTRR_TYPE_UNCACHABLE) ? "UC" : + ((type == MTRR_TYPE_WRPROT) ? "WP" : + ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other")) + ); + } +} + +static int __init mtrr_need_cleanup(void) +{ + int i; + mtrr_type type; + unsigned long size; + /* Extra one for all 0: */ + int num[MTRR_NUM_TYPES + 1]; + + /* Check entries number: */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + size = range_state[i].size_pfn; + if (type >= MTRR_NUM_TYPES) + continue; + if (!size) + type = MTRR_NUM_TYPES; + num[type]++; + } + + /* Check if we got UC entries: */ + if (!num[MTRR_TYPE_UNCACHABLE]) + return 0; + + /* Check if we only had WB and UC */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + return 1; +} + +static unsigned long __initdata range_sums; + +static void __init +mtrr_calc_range_state(u64 chunk_size, u64 gran_size, + unsigned long x_remove_base, + unsigned long x_remove_size, int i) +{ + /* + * range_new should really be an automatic variable, but + * putting 4096 bytes on the stack is frowned upon, to put it + * mildly. It is safe to make it a static __initdata variable, + * since mtrr_calc_range_state is only called during init and + * there's no way it will call itself recursively. + */ + static struct range range_new[RANGE_NUM] __initdata; + unsigned long range_sums_new; + int nr_range_new; + int num_reg; + + /* Convert ranges to var ranges state: */ + num_reg = x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size); + + /* We got new setting in range_state, check it: */ + memset(range_new, 0, sizeof(range_new)); + nr_range_new = x86_get_mtrr_mem_range(range_new, 0, + x_remove_base, x_remove_size); + range_sums_new = sum_ranges(range_new, nr_range_new); + + result[i].chunk_sizek = chunk_size >> 10; + result[i].gran_sizek = gran_size >> 10; + result[i].num_reg = num_reg; + + if (range_sums < range_sums_new) { + result[i].lose_cover_sizek = (range_sums_new - range_sums) << PSHIFT; + result[i].bad = 1; + } else { + result[i].lose_cover_sizek = (range_sums - range_sums_new) << PSHIFT; + } + + /* Double check it: */ + if (!result[i].bad && !result[i].lose_cover_sizek) { + if (nr_range_new != nr_range || memcmp(range, range_new, sizeof(range))) + result[i].bad = 1; + } + + if (!result[i].bad && (range_sums - range_sums_new < min_loss_pfn[num_reg])) + min_loss_pfn[num_reg] = range_sums - range_sums_new; +} + +static void __init mtrr_print_out_one_result(int i) +{ + unsigned long gran_base, chunk_base, lose_base; + char gran_factor, chunk_factor, lose_factor; + + gran_base = to_size_factor(result[i].gran_sizek, &gran_factor); + chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor); + lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor); + + pr_info("%sgran_size: %ld%c \tchunk_size: %ld%c \t", + result[i].bad ? "*BAD*" : " ", + gran_base, gran_factor, chunk_base, chunk_factor); + pr_cont("num_reg: %d \tlose cover RAM: %s%ld%c\n", + result[i].num_reg, result[i].bad ? "-" : "", + lose_base, lose_factor); +} + +static int __init mtrr_search_optimal_index(void) +{ + int num_reg_good; + int index_good; + int i; + + if (nr_mtrr_spare_reg >= num_var_ranges) + nr_mtrr_spare_reg = num_var_ranges - 1; + + num_reg_good = -1; + for (i = num_var_ranges - nr_mtrr_spare_reg; i > 0; i--) { + if (!min_loss_pfn[i]) + num_reg_good = i; + } + + index_good = -1; + if (num_reg_good != -1) { + for (i = 0; i < NUM_RESULT; i++) { + if (!result[i].bad && + result[i].num_reg == num_reg_good && + !result[i].lose_cover_sizek) { + index_good = i; + break; + } + } + } + + return index_good; +} + +int __init mtrr_cleanup(unsigned address_bits) +{ + unsigned long x_remove_base, x_remove_size; + unsigned long base, size, def, dummy; + u64 chunk_size, gran_size; + mtrr_type type; + int index_good; + int i; + + if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1) + return 0; + + rdmsr(MSR_MTRRdefType, def, dummy); + def &= 0xff; + if (def != MTRR_TYPE_UNCACHABLE) + return 0; + + /* Get it and store it aside: */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } + + /* Check if we need handle it and can handle it: */ + if (!mtrr_need_cleanup()) + return 0; + + /* Print original var MTRRs at first, for debugging: */ + pr_debug("original variable MTRRs\n"); + print_out_mtrr_range_state(); + + memset(range, 0, sizeof(range)); + x_remove_size = 0; + x_remove_base = 1 << (32 - PAGE_SHIFT); + if (mtrr_tom2) + x_remove_size = (mtrr_tom2 >> PAGE_SHIFT) - x_remove_base; + + /* + * [0, 1M) should always be covered by var mtrr with WB + * and fixed mtrrs should take effect before var mtrr for it: + */ + nr_range = add_range_with_merge(range, RANGE_NUM, 0, 0, + 1ULL<<(20 - PAGE_SHIFT)); + /* add from var mtrr at last */ + nr_range = x86_get_mtrr_mem_range(range, nr_range, + x_remove_base, x_remove_size); + + range_sums = sum_ranges(range, nr_range); + pr_info("total RAM covered: %ldM\n", + range_sums >> (20 - PAGE_SHIFT)); + + if (mtrr_chunk_size && mtrr_gran_size) { + i = 0; + mtrr_calc_range_state(mtrr_chunk_size, mtrr_gran_size, + x_remove_base, x_remove_size, i); + + mtrr_print_out_one_result(i); + + if (!result[i].bad) { + set_var_mtrr_all(address_bits); + pr_debug("New variable MTRRs\n"); + print_out_mtrr_range_state(); + return 1; + } + pr_info("invalid mtrr_gran_size or mtrr_chunk_size, will find optimal one\n"); + } + + i = 0; + memset(min_loss_pfn, 0xff, sizeof(min_loss_pfn)); + memset(result, 0, sizeof(result)); + for (gran_size = (1ULL<<16); gran_size < (1ULL<<32); gran_size <<= 1) { + + for (chunk_size = gran_size; chunk_size < (1ULL<<32); + chunk_size <<= 1) { + + if (i >= NUM_RESULT) + continue; + + mtrr_calc_range_state(chunk_size, gran_size, + x_remove_base, x_remove_size, i); + if (debug_print) { + mtrr_print_out_one_result(i); + pr_info("\n"); + } + + i++; + } + } + + /* Try to find the optimal index: */ + index_good = mtrr_search_optimal_index(); + + if (index_good != -1) { + pr_info("Found optimal setting for mtrr clean up\n"); + i = index_good; + mtrr_print_out_one_result(i); + + /* Convert ranges to var ranges state: */ + chunk_size = result[i].chunk_sizek; + chunk_size <<= 10; + gran_size = result[i].gran_sizek; + gran_size <<= 10; + x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size); + set_var_mtrr_all(address_bits); + pr_debug("New variable MTRRs\n"); + print_out_mtrr_range_state(); + return 1; + } else { + /* print out all */ + for (i = 0; i < NUM_RESULT; i++) + mtrr_print_out_one_result(i); + } + + pr_info("mtrr_cleanup: can not find optimal value\n"); + pr_info("please specify mtrr_gran_size/mtrr_chunk_size\n"); + + return 0; +} +#else +int __init mtrr_cleanup(unsigned address_bits) +{ + return 0; +} +#endif + +static int disable_mtrr_trim; + +static int __init disable_mtrr_trim_setup(char *str) +{ + disable_mtrr_trim = 1; + return 0; +} +early_param("disable_mtrr_trim", disable_mtrr_trim_setup); + +/* + * Newer AMD K8s and later CPUs have a special magic MSR way to force WB + * for memory >4GB. Check for that here. + * Note this won't check if the MTRRs < 4GB where the magic bit doesn't + * apply to are wrong, but so far we don't know of any such case in the wild. + */ +#define Tom2Enabled (1U << 21) +#define Tom2ForceMemTypeWB (1U << 22) + +int __init amd_special_default_mtrr(void) +{ + u32 l, h; + + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD && + boot_cpu_data.x86_vendor != X86_VENDOR_HYGON) + return 0; + if (boot_cpu_data.x86 < 0xf) + return 0; + /* In case some hypervisor doesn't pass SYSCFG through: */ + if (rdmsr_safe(MSR_K8_SYSCFG, &l, &h) < 0) + return 0; + /* + * Memory between 4GB and top of mem is forced WB by this magic bit. + * Reserved before K8RevF, but should be zero there. + */ + if ((l & (Tom2Enabled | Tom2ForceMemTypeWB)) == + (Tom2Enabled | Tom2ForceMemTypeWB)) + return 1; + return 0; +} + +static u64 __init +real_trim_memory(unsigned long start_pfn, unsigned long limit_pfn) +{ + u64 trim_start, trim_size; + + trim_start = start_pfn; + trim_start <<= PAGE_SHIFT; + + trim_size = limit_pfn; + trim_size <<= PAGE_SHIFT; + trim_size -= trim_start; + + return e820__range_update(trim_start, trim_size, E820_TYPE_RAM, E820_TYPE_RESERVED); +} + +/** + * mtrr_trim_uncached_memory - trim RAM not covered by MTRRs + * @end_pfn: ending page frame number + * + * Some buggy BIOSes don't setup the MTRRs properly for systems with certain + * memory configurations. This routine checks that the highest MTRR matches + * the end of memory, to make sure the MTRRs having a write back type cover + * all of the memory the kernel is intending to use. If not, it'll trim any + * memory off the end by adjusting end_pfn, removing it from the kernel's + * allocation pools, warning the user with an obnoxious message. + */ +int __init mtrr_trim_uncached_memory(unsigned long end_pfn) +{ + unsigned long i, base, size, highest_pfn = 0, def, dummy; + mtrr_type type; + u64 total_trim_size; + /* extra one for all 0 */ + int num[MTRR_NUM_TYPES + 1]; + + /* + * Make sure we only trim uncachable memory on machines that + * support the Intel MTRR architecture: + */ + if (!is_cpu(INTEL) || disable_mtrr_trim) + return 0; + + rdmsr(MSR_MTRRdefType, def, dummy); + def &= 0xff; + if (def != MTRR_TYPE_UNCACHABLE) + return 0; + + /* Get it and store it aside: */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } + + /* Find highest cached pfn: */ + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_WRBACK) + continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; + if (highest_pfn < base + size) + highest_pfn = base + size; + } + + /* kvm/qemu doesn't have mtrr set right, don't trim them all: */ + if (!highest_pfn) { + pr_info("CPU MTRRs all blank - virtualized system.\n"); + return 0; + } + + /* Check entries number: */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type >= MTRR_NUM_TYPES) + continue; + size = range_state[i].size_pfn; + if (!size) + type = MTRR_NUM_TYPES; + num[type]++; + } + + /* No entry for WB? */ + if (!num[MTRR_TYPE_WRBACK]) + return 0; + + /* Check if we only had WB and UC: */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + memset(range, 0, sizeof(range)); + nr_range = 0; + if (mtrr_tom2) { + range[nr_range].start = (1ULL<<(32 - PAGE_SHIFT)); + range[nr_range].end = mtrr_tom2 >> PAGE_SHIFT; + if (highest_pfn < range[nr_range].end) + highest_pfn = range[nr_range].end; + nr_range++; + } + nr_range = x86_get_mtrr_mem_range(range, nr_range, 0, 0); + + /* Check the head: */ + total_trim_size = 0; + if (range[0].start) + total_trim_size += real_trim_memory(0, range[0].start); + + /* Check the holes: */ + for (i = 0; i < nr_range - 1; i++) { + if (range[i].end < range[i+1].start) + total_trim_size += real_trim_memory(range[i].end, + range[i+1].start); + } + + /* Check the top: */ + i = nr_range - 1; + if (range[i].end < end_pfn) + total_trim_size += real_trim_memory(range[i].end, + end_pfn); + + if (total_trim_size) { + pr_warn("WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing %lluMB of RAM.\n", + total_trim_size >> 20); + + if (!changed_by_mtrr_cleanup) + WARN_ON(1); + + pr_info("update e820 for mtrr\n"); + e820__update_table_print(); + + return 1; + } + + return 0; +} diff --git a/arch/x86/kernel/cpu/mtrr/cyrix.c b/arch/x86/kernel/cpu/mtrr/cyrix.c new file mode 100644 index 000000000..ca670919b --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/cyrix.c @@ -0,0 +1,284 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/init.h> +#include <linux/io.h> +#include <linux/mm.h> + +#include <asm/processor-cyrix.h> +#include <asm/processor-flags.h> +#include <asm/mtrr.h> +#include <asm/msr.h> + +#include "mtrr.h" + +static void +cyrix_get_arr(unsigned int reg, unsigned long *base, + unsigned long *size, mtrr_type * type) +{ + unsigned char arr, ccr3, rcr, shift; + unsigned long flags; + + arr = CX86_ARR_BASE + (reg << 1) + reg; /* avoid multiplication by 3 */ + + local_irq_save(flags); + + ccr3 = getCx86(CX86_CCR3); + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */ + ((unsigned char *)base)[3] = getCx86(arr); + ((unsigned char *)base)[2] = getCx86(arr + 1); + ((unsigned char *)base)[1] = getCx86(arr + 2); + rcr = getCx86(CX86_RCR_BASE + reg); + setCx86(CX86_CCR3, ccr3); /* disable MAPEN */ + + local_irq_restore(flags); + + shift = ((unsigned char *) base)[1] & 0x0f; + *base >>= PAGE_SHIFT; + + /* + * Power of two, at least 4K on ARR0-ARR6, 256K on ARR7 + * Note: shift==0xf means 4G, this is unsupported. + */ + if (shift) + *size = (reg < 7 ? 0x1UL : 0x40UL) << (shift - 1); + else + *size = 0; + + /* Bit 0 is Cache Enable on ARR7, Cache Disable on ARR0-ARR6 */ + if (reg < 7) { + switch (rcr) { + case 1: + *type = MTRR_TYPE_UNCACHABLE; + break; + case 8: + *type = MTRR_TYPE_WRBACK; + break; + case 9: + *type = MTRR_TYPE_WRCOMB; + break; + case 24: + default: + *type = MTRR_TYPE_WRTHROUGH; + break; + } + } else { + switch (rcr) { + case 0: + *type = MTRR_TYPE_UNCACHABLE; + break; + case 8: + *type = MTRR_TYPE_WRCOMB; + break; + case 9: + *type = MTRR_TYPE_WRBACK; + break; + case 25: + default: + *type = MTRR_TYPE_WRTHROUGH; + break; + } + } +} + +/* + * cyrix_get_free_region - get a free ARR. + * + * @base: the starting (base) address of the region. + * @size: the size (in bytes) of the region. + * + * Returns: the index of the region on success, else -1 on error. +*/ +static int +cyrix_get_free_region(unsigned long base, unsigned long size, int replace_reg) +{ + unsigned long lbase, lsize; + mtrr_type ltype; + int i; + + switch (replace_reg) { + case 7: + if (size < 0x40) + break; + fallthrough; + case 6: + case 5: + case 4: + return replace_reg; + case 3: + case 2: + case 1: + case 0: + return replace_reg; + } + /* If we are to set up a region >32M then look at ARR7 immediately */ + if (size > 0x2000) { + cyrix_get_arr(7, &lbase, &lsize, <ype); + if (lsize == 0) + return 7; + /* Else try ARR0-ARR6 first */ + } else { + for (i = 0; i < 7; i++) { + cyrix_get_arr(i, &lbase, &lsize, <ype); + if (lsize == 0) + return i; + } + /* + * ARR0-ARR6 isn't free + * try ARR7 but its size must be at least 256K + */ + cyrix_get_arr(i, &lbase, &lsize, <ype); + if ((lsize == 0) && (size >= 0x40)) + return i; + } + return -ENOSPC; +} + +static u32 cr4, ccr3; + +static void prepare_set(void) +{ + u32 cr0; + + /* Save value of CR4 and clear Page Global Enable (bit 7) */ + if (boot_cpu_has(X86_FEATURE_PGE)) { + cr4 = __read_cr4(); + __write_cr4(cr4 & ~X86_CR4_PGE); + } + + /* + * Disable and flush caches. + * Note that wbinvd flushes the TLBs as a side-effect + */ + cr0 = read_cr0() | X86_CR0_CD; + wbinvd(); + write_cr0(cr0); + wbinvd(); + + /* Cyrix ARRs - everything else was excluded at the top */ + ccr3 = getCx86(CX86_CCR3); + + /* Cyrix ARRs - everything else was excluded at the top */ + setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); +} + +static void post_set(void) +{ + /* Flush caches and TLBs */ + wbinvd(); + + /* Cyrix ARRs - everything else was excluded at the top */ + setCx86(CX86_CCR3, ccr3); + + /* Enable caches */ + write_cr0(read_cr0() & ~X86_CR0_CD); + + /* Restore value of CR4 */ + if (boot_cpu_has(X86_FEATURE_PGE)) + __write_cr4(cr4); +} + +static void cyrix_set_arr(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type) +{ + unsigned char arr, arr_type, arr_size; + + arr = CX86_ARR_BASE + (reg << 1) + reg; /* avoid multiplication by 3 */ + + /* count down from 32M (ARR0-ARR6) or from 2G (ARR7) */ + if (reg >= 7) + size >>= 6; + + size &= 0x7fff; /* make sure arr_size <= 14 */ + for (arr_size = 0; size; arr_size++, size >>= 1) + ; + + if (reg < 7) { + switch (type) { + case MTRR_TYPE_UNCACHABLE: + arr_type = 1; + break; + case MTRR_TYPE_WRCOMB: + arr_type = 9; + break; + case MTRR_TYPE_WRTHROUGH: + arr_type = 24; + break; + default: + arr_type = 8; + break; + } + } else { + switch (type) { + case MTRR_TYPE_UNCACHABLE: + arr_type = 0; + break; + case MTRR_TYPE_WRCOMB: + arr_type = 8; + break; + case MTRR_TYPE_WRTHROUGH: + arr_type = 25; + break; + default: + arr_type = 9; + break; + } + } + + prepare_set(); + + base <<= PAGE_SHIFT; + setCx86(arr + 0, ((unsigned char *)&base)[3]); + setCx86(arr + 1, ((unsigned char *)&base)[2]); + setCx86(arr + 2, (((unsigned char *)&base)[1]) | arr_size); + setCx86(CX86_RCR_BASE + reg, arr_type); + + post_set(); +} + +typedef struct { + unsigned long base; + unsigned long size; + mtrr_type type; +} arr_state_t; + +static arr_state_t arr_state[8] = { + {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL}, + {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL}, {0UL, 0UL, 0UL} +}; + +static unsigned char ccr_state[7] = { 0, 0, 0, 0, 0, 0, 0 }; + +static void cyrix_set_all(void) +{ + int i; + + prepare_set(); + + /* the CCRs are not contiguous */ + for (i = 0; i < 4; i++) + setCx86(CX86_CCR0 + i, ccr_state[i]); + for (; i < 7; i++) + setCx86(CX86_CCR4 + i, ccr_state[i]); + + for (i = 0; i < 8; i++) { + cyrix_set_arr(i, arr_state[i].base, + arr_state[i].size, arr_state[i].type); + } + + post_set(); +} + +static const struct mtrr_ops cyrix_mtrr_ops = { + .vendor = X86_VENDOR_CYRIX, + .set_all = cyrix_set_all, + .set = cyrix_set_arr, + .get = cyrix_get_arr, + .get_free_region = cyrix_get_free_region, + .validate_add_page = generic_validate_add_page, + .have_wrcomb = positive_have_wrcomb, +}; + +int __init cyrix_init_mtrr(void) +{ + set_mtrr_ops(&cyrix_mtrr_ops); + return 0; +} diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c new file mode 100644 index 000000000..a29997e6c --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -0,0 +1,925 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * This only handles 32bit MTRR on 32bit hosts. This is strictly wrong + * because MTRRs can span up to 40 bits (36bits on most modern x86) + */ +#define DEBUG + +#include <linux/export.h> +#include <linux/init.h> +#include <linux/io.h> +#include <linux/mm.h> + +#include <asm/processor-flags.h> +#include <asm/cpufeature.h> +#include <asm/tlbflush.h> +#include <asm/mtrr.h> +#include <asm/msr.h> +#include <asm/memtype.h> + +#include "mtrr.h" + +struct fixed_range_block { + int base_msr; /* start address of an MTRR block */ + int ranges; /* number of MTRRs in this block */ +}; + +static struct fixed_range_block fixed_range_blocks[] = { + { MSR_MTRRfix64K_00000, 1 }, /* one 64k MTRR */ + { MSR_MTRRfix16K_80000, 2 }, /* two 16k MTRRs */ + { MSR_MTRRfix4K_C0000, 8 }, /* eight 4k MTRRs */ + {} +}; + +static unsigned long smp_changes_mask; +static int mtrr_state_set; +u64 mtrr_tom2; + +struct mtrr_state_type mtrr_state; +EXPORT_SYMBOL_GPL(mtrr_state); + +/* + * BIOS is expected to clear MtrrFixDramModEn bit, see for example + * "BIOS and Kernel Developer's Guide for the AMD Athlon 64 and AMD + * Opteron Processors" (26094 Rev. 3.30 February 2006), section + * "13.2.1.2 SYSCFG Register": "The MtrrFixDramModEn bit should be set + * to 1 during BIOS initialization of the fixed MTRRs, then cleared to + * 0 for operation." + */ +static inline void k8_check_syscfg_dram_mod_en(void) +{ + u32 lo, hi; + + if (!((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && + (boot_cpu_data.x86 >= 0x0f))) + return; + + rdmsr(MSR_K8_SYSCFG, lo, hi); + if (lo & K8_MTRRFIXRANGE_DRAM_MODIFY) { + pr_err(FW_WARN "MTRR: CPU %u: SYSCFG[MtrrFixDramModEn]" + " not cleared by BIOS, clearing this bit\n", + smp_processor_id()); + lo &= ~K8_MTRRFIXRANGE_DRAM_MODIFY; + mtrr_wrmsr(MSR_K8_SYSCFG, lo, hi); + } +} + +/* Get the size of contiguous MTRR range */ +static u64 get_mtrr_size(u64 mask) +{ + u64 size; + + mask >>= PAGE_SHIFT; + mask |= size_or_mask; + size = -mask; + size <<= PAGE_SHIFT; + return size; +} + +/* + * Check and return the effective type for MTRR-MTRR type overlap. + * Returns 1 if the effective type is UNCACHEABLE, else returns 0 + */ +static int check_type_overlap(u8 *prev, u8 *curr) +{ + if (*prev == MTRR_TYPE_UNCACHABLE || *curr == MTRR_TYPE_UNCACHABLE) { + *prev = MTRR_TYPE_UNCACHABLE; + *curr = MTRR_TYPE_UNCACHABLE; + return 1; + } + + if ((*prev == MTRR_TYPE_WRBACK && *curr == MTRR_TYPE_WRTHROUGH) || + (*prev == MTRR_TYPE_WRTHROUGH && *curr == MTRR_TYPE_WRBACK)) { + *prev = MTRR_TYPE_WRTHROUGH; + *curr = MTRR_TYPE_WRTHROUGH; + } + + if (*prev != *curr) { + *prev = MTRR_TYPE_UNCACHABLE; + *curr = MTRR_TYPE_UNCACHABLE; + return 1; + } + + return 0; +} + +/** + * mtrr_type_lookup_fixed - look up memory type in MTRR fixed entries + * + * Return the MTRR fixed memory type of 'start'. + * + * MTRR fixed entries are divided into the following ways: + * 0x00000 - 0x7FFFF : This range is divided into eight 64KB sub-ranges + * 0x80000 - 0xBFFFF : This range is divided into sixteen 16KB sub-ranges + * 0xC0000 - 0xFFFFF : This range is divided into sixty-four 4KB sub-ranges + * + * Return Values: + * MTRR_TYPE_(type) - Matched memory type + * MTRR_TYPE_INVALID - Unmatched + */ +static u8 mtrr_type_lookup_fixed(u64 start, u64 end) +{ + int idx; + + if (start >= 0x100000) + return MTRR_TYPE_INVALID; + + /* 0x0 - 0x7FFFF */ + if (start < 0x80000) { + idx = 0; + idx += (start >> 16); + return mtrr_state.fixed_ranges[idx]; + /* 0x80000 - 0xBFFFF */ + } else if (start < 0xC0000) { + idx = 1 * 8; + idx += ((start - 0x80000) >> 14); + return mtrr_state.fixed_ranges[idx]; + } + + /* 0xC0000 - 0xFFFFF */ + idx = 3 * 8; + idx += ((start - 0xC0000) >> 12); + return mtrr_state.fixed_ranges[idx]; +} + +/** + * mtrr_type_lookup_variable - look up memory type in MTRR variable entries + * + * Return Value: + * MTRR_TYPE_(type) - Matched memory type or default memory type (unmatched) + * + * Output Arguments: + * repeat - Set to 1 when [start:end] spanned across MTRR range and type + * returned corresponds only to [start:*partial_end]. Caller has + * to lookup again for [*partial_end:end]. + * + * uniform - Set to 1 when an MTRR covers the region uniformly, i.e. the + * region is fully covered by a single MTRR entry or the default + * type. + */ +static u8 mtrr_type_lookup_variable(u64 start, u64 end, u64 *partial_end, + int *repeat, u8 *uniform) +{ + int i; + u64 base, mask; + u8 prev_match, curr_match; + + *repeat = 0; + *uniform = 1; + + prev_match = MTRR_TYPE_INVALID; + for (i = 0; i < num_var_ranges; ++i) { + unsigned short start_state, end_state, inclusive; + + if (!(mtrr_state.var_ranges[i].mask_lo & (1 << 11))) + continue; + + base = (((u64)mtrr_state.var_ranges[i].base_hi) << 32) + + (mtrr_state.var_ranges[i].base_lo & PAGE_MASK); + mask = (((u64)mtrr_state.var_ranges[i].mask_hi) << 32) + + (mtrr_state.var_ranges[i].mask_lo & PAGE_MASK); + + start_state = ((start & mask) == (base & mask)); + end_state = ((end & mask) == (base & mask)); + inclusive = ((start < base) && (end > base)); + + if ((start_state != end_state) || inclusive) { + /* + * We have start:end spanning across an MTRR. + * We split the region into either + * + * - start_state:1 + * (start:mtrr_end)(mtrr_end:end) + * - end_state:1 + * (start:mtrr_start)(mtrr_start:end) + * - inclusive:1 + * (start:mtrr_start)(mtrr_start:mtrr_end)(mtrr_end:end) + * + * depending on kind of overlap. + * + * Return the type of the first region and a pointer + * to the start of next region so that caller will be + * advised to lookup again after having adjusted start + * and end. + * + * Note: This way we handle overlaps with multiple + * entries and the default type properly. + */ + if (start_state) + *partial_end = base + get_mtrr_size(mask); + else + *partial_end = base; + + if (unlikely(*partial_end <= start)) { + WARN_ON(1); + *partial_end = start + PAGE_SIZE; + } + + end = *partial_end - 1; /* end is inclusive */ + *repeat = 1; + *uniform = 0; + } + + if ((start & mask) != (base & mask)) + continue; + + curr_match = mtrr_state.var_ranges[i].base_lo & 0xff; + if (prev_match == MTRR_TYPE_INVALID) { + prev_match = curr_match; + continue; + } + + *uniform = 0; + if (check_type_overlap(&prev_match, &curr_match)) + return curr_match; + } + + if (prev_match != MTRR_TYPE_INVALID) + return prev_match; + + return mtrr_state.def_type; +} + +/** + * mtrr_type_lookup - look up memory type in MTRR + * + * Return Values: + * MTRR_TYPE_(type) - The effective MTRR type for the region + * MTRR_TYPE_INVALID - MTRR is disabled + * + * Output Argument: + * uniform - Set to 1 when an MTRR covers the region uniformly, i.e. the + * region is fully covered by a single MTRR entry or the default + * type. + */ +u8 mtrr_type_lookup(u64 start, u64 end, u8 *uniform) +{ + u8 type, prev_type, is_uniform = 1, dummy; + int repeat; + u64 partial_end; + + /* Make end inclusive instead of exclusive */ + end--; + + if (!mtrr_state_set) + return MTRR_TYPE_INVALID; + + if (!(mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED)) + return MTRR_TYPE_INVALID; + + /* + * Look up the fixed ranges first, which take priority over + * the variable ranges. + */ + if ((start < 0x100000) && + (mtrr_state.have_fixed) && + (mtrr_state.enabled & MTRR_STATE_MTRR_FIXED_ENABLED)) { + is_uniform = 0; + type = mtrr_type_lookup_fixed(start, end); + goto out; + } + + /* + * Look up the variable ranges. Look of multiple ranges matching + * this address and pick type as per MTRR precedence. + */ + type = mtrr_type_lookup_variable(start, end, &partial_end, + &repeat, &is_uniform); + + /* + * Common path is with repeat = 0. + * However, we can have cases where [start:end] spans across some + * MTRR ranges and/or the default type. Do repeated lookups for + * that case here. + */ + while (repeat) { + prev_type = type; + start = partial_end; + is_uniform = 0; + type = mtrr_type_lookup_variable(start, end, &partial_end, + &repeat, &dummy); + + if (check_type_overlap(&prev_type, &type)) + goto out; + } + + if (mtrr_tom2 && (start >= (1ULL<<32)) && (end < mtrr_tom2)) + type = MTRR_TYPE_WRBACK; + +out: + *uniform = is_uniform; + return type; +} + +/* Get the MSR pair relating to a var range */ +static void +get_mtrr_var_range(unsigned int index, struct mtrr_var_range *vr) +{ + rdmsr(MTRRphysBase_MSR(index), vr->base_lo, vr->base_hi); + rdmsr(MTRRphysMask_MSR(index), vr->mask_lo, vr->mask_hi); +} + +/* Fill the MSR pair relating to a var range */ +void fill_mtrr_var_range(unsigned int index, + u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi) +{ + struct mtrr_var_range *vr; + + vr = mtrr_state.var_ranges; + + vr[index].base_lo = base_lo; + vr[index].base_hi = base_hi; + vr[index].mask_lo = mask_lo; + vr[index].mask_hi = mask_hi; +} + +static void get_fixed_ranges(mtrr_type *frs) +{ + unsigned int *p = (unsigned int *)frs; + int i; + + k8_check_syscfg_dram_mod_en(); + + rdmsr(MSR_MTRRfix64K_00000, p[0], p[1]); + + for (i = 0; i < 2; i++) + rdmsr(MSR_MTRRfix16K_80000 + i, p[2 + i * 2], p[3 + i * 2]); + for (i = 0; i < 8; i++) + rdmsr(MSR_MTRRfix4K_C0000 + i, p[6 + i * 2], p[7 + i * 2]); +} + +void mtrr_save_fixed_ranges(void *info) +{ + if (boot_cpu_has(X86_FEATURE_MTRR)) + get_fixed_ranges(mtrr_state.fixed_ranges); +} + +static unsigned __initdata last_fixed_start; +static unsigned __initdata last_fixed_end; +static mtrr_type __initdata last_fixed_type; + +static void __init print_fixed_last(void) +{ + if (!last_fixed_end) + return; + + pr_debug(" %05X-%05X %s\n", last_fixed_start, + last_fixed_end - 1, mtrr_attrib_to_str(last_fixed_type)); + + last_fixed_end = 0; +} + +static void __init update_fixed_last(unsigned base, unsigned end, + mtrr_type type) +{ + last_fixed_start = base; + last_fixed_end = end; + last_fixed_type = type; +} + +static void __init +print_fixed(unsigned base, unsigned step, const mtrr_type *types) +{ + unsigned i; + + for (i = 0; i < 8; ++i, ++types, base += step) { + if (last_fixed_end == 0) { + update_fixed_last(base, base + step, *types); + continue; + } + if (last_fixed_end == base && last_fixed_type == *types) { + last_fixed_end = base + step; + continue; + } + /* new segments: gap or different type */ + print_fixed_last(); + update_fixed_last(base, base + step, *types); + } +} + +static void prepare_set(void); +static void post_set(void); + +static void __init print_mtrr_state(void) +{ + unsigned int i; + int high_width; + + pr_debug("MTRR default type: %s\n", + mtrr_attrib_to_str(mtrr_state.def_type)); + if (mtrr_state.have_fixed) { + pr_debug("MTRR fixed ranges %sabled:\n", + ((mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED) && + (mtrr_state.enabled & MTRR_STATE_MTRR_FIXED_ENABLED)) ? + "en" : "dis"); + print_fixed(0x00000, 0x10000, mtrr_state.fixed_ranges + 0); + for (i = 0; i < 2; ++i) + print_fixed(0x80000 + i * 0x20000, 0x04000, + mtrr_state.fixed_ranges + (i + 1) * 8); + for (i = 0; i < 8; ++i) + print_fixed(0xC0000 + i * 0x08000, 0x01000, + mtrr_state.fixed_ranges + (i + 3) * 8); + + /* tail */ + print_fixed_last(); + } + pr_debug("MTRR variable ranges %sabled:\n", + mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED ? "en" : "dis"); + high_width = (__ffs64(size_or_mask) - (32 - PAGE_SHIFT) + 3) / 4; + + for (i = 0; i < num_var_ranges; ++i) { + if (mtrr_state.var_ranges[i].mask_lo & (1 << 11)) + pr_debug(" %u base %0*X%05X000 mask %0*X%05X000 %s\n", + i, + high_width, + mtrr_state.var_ranges[i].base_hi, + mtrr_state.var_ranges[i].base_lo >> 12, + high_width, + mtrr_state.var_ranges[i].mask_hi, + mtrr_state.var_ranges[i].mask_lo >> 12, + mtrr_attrib_to_str(mtrr_state.var_ranges[i].base_lo & 0xff)); + else + pr_debug(" %u disabled\n", i); + } + if (mtrr_tom2) + pr_debug("TOM2: %016llx aka %lldM\n", mtrr_tom2, mtrr_tom2>>20); +} + +/* PAT setup for BP. We need to go through sync steps here */ +void __init mtrr_bp_pat_init(void) +{ + unsigned long flags; + + local_irq_save(flags); + prepare_set(); + + pat_init(); + + post_set(); + local_irq_restore(flags); +} + +/* Grab all of the MTRR state for this CPU into *state */ +bool __init get_mtrr_state(void) +{ + struct mtrr_var_range *vrs; + unsigned lo, dummy; + unsigned int i; + + vrs = mtrr_state.var_ranges; + + rdmsr(MSR_MTRRcap, lo, dummy); + mtrr_state.have_fixed = (lo >> 8) & 1; + + for (i = 0; i < num_var_ranges; i++) + get_mtrr_var_range(i, &vrs[i]); + if (mtrr_state.have_fixed) + get_fixed_ranges(mtrr_state.fixed_ranges); + + rdmsr(MSR_MTRRdefType, lo, dummy); + mtrr_state.def_type = (lo & 0xff); + mtrr_state.enabled = (lo & 0xc00) >> 10; + + if (amd_special_default_mtrr()) { + unsigned low, high; + + /* TOP_MEM2 */ + rdmsr(MSR_K8_TOP_MEM2, low, high); + mtrr_tom2 = high; + mtrr_tom2 <<= 32; + mtrr_tom2 |= low; + mtrr_tom2 &= 0xffffff800000ULL; + } + + print_mtrr_state(); + + mtrr_state_set = 1; + + return !!(mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED); +} + +/* Some BIOS's are messed up and don't set all MTRRs the same! */ +void __init mtrr_state_warn(void) +{ + unsigned long mask = smp_changes_mask; + + if (!mask) + return; + if (mask & MTRR_CHANGE_MASK_FIXED) + pr_warn("mtrr: your CPUs had inconsistent fixed MTRR settings\n"); + if (mask & MTRR_CHANGE_MASK_VARIABLE) + pr_warn("mtrr: your CPUs had inconsistent variable MTRR settings\n"); + if (mask & MTRR_CHANGE_MASK_DEFTYPE) + pr_warn("mtrr: your CPUs had inconsistent MTRRdefType settings\n"); + + pr_info("mtrr: probably your BIOS does not setup all CPUs.\n"); + pr_info("mtrr: corrected configuration.\n"); +} + +/* + * Doesn't attempt to pass an error out to MTRR users + * because it's quite complicated in some cases and probably not + * worth it because the best error handling is to ignore it. + */ +void mtrr_wrmsr(unsigned msr, unsigned a, unsigned b) +{ + if (wrmsr_safe(msr, a, b) < 0) { + pr_err("MTRR: CPU %u: Writing MSR %x to %x:%x failed\n", + smp_processor_id(), msr, a, b); + } +} + +/** + * set_fixed_range - checks & updates a fixed-range MTRR if it + * differs from the value it should have + * @msr: MSR address of the MTTR which should be checked and updated + * @changed: pointer which indicates whether the MTRR needed to be changed + * @msrwords: pointer to the MSR values which the MSR should have + */ +static void set_fixed_range(int msr, bool *changed, unsigned int *msrwords) +{ + unsigned lo, hi; + + rdmsr(msr, lo, hi); + + if (lo != msrwords[0] || hi != msrwords[1]) { + mtrr_wrmsr(msr, msrwords[0], msrwords[1]); + *changed = true; + } +} + +/** + * generic_get_free_region - Get a free MTRR. + * @base: The starting (base) address of the region. + * @size: The size (in bytes) of the region. + * @replace_reg: mtrr index to be replaced; set to invalid value if none. + * + * Returns: The index of the region on success, else negative on error. + */ +int +generic_get_free_region(unsigned long base, unsigned long size, int replace_reg) +{ + unsigned long lbase, lsize; + mtrr_type ltype; + int i, max; + + max = num_var_ranges; + if (replace_reg >= 0 && replace_reg < max) + return replace_reg; + + for (i = 0; i < max; ++i) { + mtrr_if->get(i, &lbase, &lsize, <ype); + if (lsize == 0) + return i; + } + + return -ENOSPC; +} + +static void generic_get_mtrr(unsigned int reg, unsigned long *base, + unsigned long *size, mtrr_type *type) +{ + u32 mask_lo, mask_hi, base_lo, base_hi; + unsigned int hi; + u64 tmp, mask; + + /* + * get_mtrr doesn't need to update mtrr_state, also it could be called + * from any cpu, so try to print it out directly. + */ + get_cpu(); + + rdmsr(MTRRphysMask_MSR(reg), mask_lo, mask_hi); + + if ((mask_lo & 0x800) == 0) { + /* Invalid (i.e. free) range */ + *base = 0; + *size = 0; + *type = 0; + goto out_put_cpu; + } + + rdmsr(MTRRphysBase_MSR(reg), base_lo, base_hi); + + /* Work out the shifted address mask: */ + tmp = (u64)mask_hi << (32 - PAGE_SHIFT) | mask_lo >> PAGE_SHIFT; + mask = size_or_mask | tmp; + + /* Expand tmp with high bits to all 1s: */ + hi = fls64(tmp); + if (hi > 0) { + tmp |= ~((1ULL<<(hi - 1)) - 1); + + if (tmp != mask) { + pr_warn("mtrr: your BIOS has configured an incorrect mask, fixing it.\n"); + add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK); + mask = tmp; + } + } + + /* + * This works correctly if size is a power of two, i.e. a + * contiguous range: + */ + *size = -mask; + *base = (u64)base_hi << (32 - PAGE_SHIFT) | base_lo >> PAGE_SHIFT; + *type = base_lo & 0xff; + +out_put_cpu: + put_cpu(); +} + +/** + * set_fixed_ranges - checks & updates the fixed-range MTRRs if they + * differ from the saved set + * @frs: pointer to fixed-range MTRR values, saved by get_fixed_ranges() + */ +static int set_fixed_ranges(mtrr_type *frs) +{ + unsigned long long *saved = (unsigned long long *)frs; + bool changed = false; + int block = -1, range; + + k8_check_syscfg_dram_mod_en(); + + while (fixed_range_blocks[++block].ranges) { + for (range = 0; range < fixed_range_blocks[block].ranges; range++) + set_fixed_range(fixed_range_blocks[block].base_msr + range, + &changed, (unsigned int *)saved++); + } + + return changed; +} + +/* + * Set the MSR pair relating to a var range. + * Returns true if changes are made. + */ +static bool set_mtrr_var_ranges(unsigned int index, struct mtrr_var_range *vr) +{ + unsigned int lo, hi; + bool changed = false; + + rdmsr(MTRRphysBase_MSR(index), lo, hi); + if ((vr->base_lo & 0xfffff0ffUL) != (lo & 0xfffff0ffUL) + || (vr->base_hi & (size_and_mask >> (32 - PAGE_SHIFT))) != + (hi & (size_and_mask >> (32 - PAGE_SHIFT)))) { + + mtrr_wrmsr(MTRRphysBase_MSR(index), vr->base_lo, vr->base_hi); + changed = true; + } + + rdmsr(MTRRphysMask_MSR(index), lo, hi); + + if ((vr->mask_lo & 0xfffff800UL) != (lo & 0xfffff800UL) + || (vr->mask_hi & (size_and_mask >> (32 - PAGE_SHIFT))) != + (hi & (size_and_mask >> (32 - PAGE_SHIFT)))) { + mtrr_wrmsr(MTRRphysMask_MSR(index), vr->mask_lo, vr->mask_hi); + changed = true; + } + return changed; +} + +static u32 deftype_lo, deftype_hi; + +/** + * set_mtrr_state - Set the MTRR state for this CPU. + * + * NOTE: The CPU must already be in a safe state for MTRR changes. + * RETURNS: 0 if no changes made, else a mask indicating what was changed. + */ +static unsigned long set_mtrr_state(void) +{ + unsigned long change_mask = 0; + unsigned int i; + + for (i = 0; i < num_var_ranges; i++) { + if (set_mtrr_var_ranges(i, &mtrr_state.var_ranges[i])) + change_mask |= MTRR_CHANGE_MASK_VARIABLE; + } + + if (mtrr_state.have_fixed && set_fixed_ranges(mtrr_state.fixed_ranges)) + change_mask |= MTRR_CHANGE_MASK_FIXED; + + /* + * Set_mtrr_restore restores the old value of MTRRdefType, + * so to set it we fiddle with the saved value: + */ + if ((deftype_lo & 0xff) != mtrr_state.def_type + || ((deftype_lo & 0xc00) >> 10) != mtrr_state.enabled) { + + deftype_lo = (deftype_lo & ~0xcff) | mtrr_state.def_type | + (mtrr_state.enabled << 10); + change_mask |= MTRR_CHANGE_MASK_DEFTYPE; + } + + return change_mask; +} + + +static unsigned long cr4; +static DEFINE_RAW_SPINLOCK(set_atomicity_lock); + +/* + * Since we are disabling the cache don't allow any interrupts, + * they would run extremely slow and would only increase the pain. + * + * The caller must ensure that local interrupts are disabled and + * are reenabled after post_set() has been called. + */ +static void prepare_set(void) __acquires(set_atomicity_lock) +{ + unsigned long cr0; + + /* + * Note that this is not ideal + * since the cache is only flushed/disabled for this CPU while the + * MTRRs are changed, but changing this requires more invasive + * changes to the way the kernel boots + */ + + raw_spin_lock(&set_atomicity_lock); + + /* Enter the no-fill (CD=1, NW=0) cache mode and flush caches. */ + cr0 = read_cr0() | X86_CR0_CD; + write_cr0(cr0); + + /* + * Cache flushing is the most time-consuming step when programming + * the MTRRs. Fortunately, as per the Intel Software Development + * Manual, we can skip it if the processor supports cache self- + * snooping. + */ + if (!static_cpu_has(X86_FEATURE_SELFSNOOP)) + wbinvd(); + + /* Save value of CR4 and clear Page Global Enable (bit 7) */ + if (boot_cpu_has(X86_FEATURE_PGE)) { + cr4 = __read_cr4(); + __write_cr4(cr4 & ~X86_CR4_PGE); + } + + /* Flush all TLBs via a mov %cr3, %reg; mov %reg, %cr3 */ + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); + flush_tlb_local(); + + /* Save MTRR state */ + rdmsr(MSR_MTRRdefType, deftype_lo, deftype_hi); + + /* Disable MTRRs, and set the default type to uncached */ + mtrr_wrmsr(MSR_MTRRdefType, deftype_lo & ~0xcff, deftype_hi); + + /* Again, only flush caches if we have to. */ + if (!static_cpu_has(X86_FEATURE_SELFSNOOP)) + wbinvd(); +} + +static void post_set(void) __releases(set_atomicity_lock) +{ + /* Flush TLBs (no need to flush caches - they are disabled) */ + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); + flush_tlb_local(); + + /* Intel (P6) standard MTRRs */ + mtrr_wrmsr(MSR_MTRRdefType, deftype_lo, deftype_hi); + + /* Enable caches */ + write_cr0(read_cr0() & ~X86_CR0_CD); + + /* Restore value of CR4 */ + if (boot_cpu_has(X86_FEATURE_PGE)) + __write_cr4(cr4); + raw_spin_unlock(&set_atomicity_lock); +} + +static void generic_set_all(void) +{ + unsigned long mask, count; + unsigned long flags; + + local_irq_save(flags); + prepare_set(); + + /* Actually set the state */ + mask = set_mtrr_state(); + + /* also set PAT */ + pat_init(); + + post_set(); + local_irq_restore(flags); + + /* Use the atomic bitops to update the global mask */ + for (count = 0; count < sizeof(mask) * 8; ++count) { + if (mask & 0x01) + set_bit(count, &smp_changes_mask); + mask >>= 1; + } + +} + +/** + * generic_set_mtrr - set variable MTRR register on the local CPU. + * + * @reg: The register to set. + * @base: The base address of the region. + * @size: The size of the region. If this is 0 the region is disabled. + * @type: The type of the region. + * + * Returns nothing. + */ +static void generic_set_mtrr(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type) +{ + unsigned long flags; + struct mtrr_var_range *vr; + + vr = &mtrr_state.var_ranges[reg]; + + local_irq_save(flags); + prepare_set(); + + if (size == 0) { + /* + * The invalid bit is kept in the mask, so we simply + * clear the relevant mask register to disable a range. + */ + mtrr_wrmsr(MTRRphysMask_MSR(reg), 0, 0); + memset(vr, 0, sizeof(struct mtrr_var_range)); + } else { + vr->base_lo = base << PAGE_SHIFT | type; + vr->base_hi = (base & size_and_mask) >> (32 - PAGE_SHIFT); + vr->mask_lo = -size << PAGE_SHIFT | 0x800; + vr->mask_hi = (-size & size_and_mask) >> (32 - PAGE_SHIFT); + + mtrr_wrmsr(MTRRphysBase_MSR(reg), vr->base_lo, vr->base_hi); + mtrr_wrmsr(MTRRphysMask_MSR(reg), vr->mask_lo, vr->mask_hi); + } + + post_set(); + local_irq_restore(flags); +} + +int generic_validate_add_page(unsigned long base, unsigned long size, + unsigned int type) +{ + unsigned long lbase, last; + + /* + * For Intel PPro stepping <= 7 + * must be 4 MiB aligned and not touch 0x70000000 -> 0x7003FFFF + */ + if (is_cpu(INTEL) && boot_cpu_data.x86 == 6 && + boot_cpu_data.x86_model == 1 && + boot_cpu_data.x86_stepping <= 7) { + if (base & ((1 << (22 - PAGE_SHIFT)) - 1)) { + pr_warn("mtrr: base(0x%lx000) is not 4 MiB aligned\n", base); + return -EINVAL; + } + if (!(base + size < 0x70000 || base > 0x7003F) && + (type == MTRR_TYPE_WRCOMB + || type == MTRR_TYPE_WRBACK)) { + pr_warn("mtrr: writable mtrr between 0x70000000 and 0x7003FFFF may hang the CPU.\n"); + return -EINVAL; + } + } + + /* + * Check upper bits of base and last are equal and lower bits are 0 + * for base and 1 for last + */ + last = base + size - 1; + for (lbase = base; !(lbase & 1) && (last & 1); + lbase = lbase >> 1, last = last >> 1) + ; + if (lbase != last) { + pr_warn("mtrr: base(0x%lx000) is not aligned on a size(0x%lx000) boundary\n", base, size); + return -EINVAL; + } + return 0; +} + +static int generic_have_wrcomb(void) +{ + unsigned long config, dummy; + rdmsr(MSR_MTRRcap, config, dummy); + return config & (1 << 10); +} + +int positive_have_wrcomb(void) +{ + return 1; +} + +/* + * Generic structure... + */ +const struct mtrr_ops generic_mtrr_ops = { + .use_intel_if = 1, + .set_all = generic_set_all, + .get = generic_get_mtrr, + .get_free_region = generic_get_free_region, + .set = generic_set_mtrr, + .validate_add_page = generic_validate_add_page, + .have_wrcomb = generic_have_wrcomb, +}; diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c new file mode 100644 index 000000000..a5c506f6d --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/if.c @@ -0,0 +1,425 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/capability.h> +#include <linux/seq_file.h> +#include <linux/uaccess.h> +#include <linux/proc_fs.h> +#include <linux/ctype.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/init.h> + +#define LINE_SIZE 80 + +#include <asm/mtrr.h> + +#include "mtrr.h" + +#define FILE_FCOUNT(f) (((struct seq_file *)((f)->private_data))->private) + +static const char *const mtrr_strings[MTRR_NUM_TYPES] = +{ + "uncachable", /* 0 */ + "write-combining", /* 1 */ + "?", /* 2 */ + "?", /* 3 */ + "write-through", /* 4 */ + "write-protect", /* 5 */ + "write-back", /* 6 */ +}; + +const char *mtrr_attrib_to_str(int x) +{ + return (x <= 6) ? mtrr_strings[x] : "?"; +} + +#ifdef CONFIG_PROC_FS + +static int +mtrr_file_add(unsigned long base, unsigned long size, + unsigned int type, bool increment, struct file *file, int page) +{ + unsigned int *fcount = FILE_FCOUNT(file); + int reg, max; + + max = num_var_ranges; + if (fcount == NULL) { + fcount = kcalloc(max, sizeof(*fcount), GFP_KERNEL); + if (!fcount) + return -ENOMEM; + FILE_FCOUNT(file) = fcount; + } + if (!page) { + if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) + return -EINVAL; + base >>= PAGE_SHIFT; + size >>= PAGE_SHIFT; + } + reg = mtrr_add_page(base, size, type, true); + if (reg >= 0) + ++fcount[reg]; + return reg; +} + +static int +mtrr_file_del(unsigned long base, unsigned long size, + struct file *file, int page) +{ + unsigned int *fcount = FILE_FCOUNT(file); + int reg; + + if (!page) { + if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) + return -EINVAL; + base >>= PAGE_SHIFT; + size >>= PAGE_SHIFT; + } + reg = mtrr_del_page(-1, base, size); + if (reg < 0) + return reg; + if (fcount == NULL) + return reg; + if (fcount[reg] < 1) + return -EINVAL; + --fcount[reg]; + return reg; +} + +/* + * seq_file can seek but we ignore it. + * + * Format of control line: + * "base=%Lx size=%Lx type=%s" or "disable=%d" + */ +static ssize_t +mtrr_write(struct file *file, const char __user *buf, size_t len, loff_t * ppos) +{ + int i, err; + unsigned long reg; + unsigned long long base, size; + char *ptr; + char line[LINE_SIZE]; + int length; + size_t linelen; + + memset(line, 0, LINE_SIZE); + + len = min_t(size_t, len, LINE_SIZE - 1); + length = strncpy_from_user(line, buf, len); + if (length < 0) + return length; + + linelen = strlen(line); + ptr = line + linelen - 1; + if (linelen && *ptr == '\n') + *ptr = '\0'; + + if (!strncmp(line, "disable=", 8)) { + reg = simple_strtoul(line + 8, &ptr, 0); + err = mtrr_del_page(reg, 0, 0); + if (err < 0) + return err; + return len; + } + + if (strncmp(line, "base=", 5)) + return -EINVAL; + + base = simple_strtoull(line + 5, &ptr, 0); + ptr = skip_spaces(ptr); + + if (strncmp(ptr, "size=", 5)) + return -EINVAL; + + size = simple_strtoull(ptr + 5, &ptr, 0); + if ((base & 0xfff) || (size & 0xfff)) + return -EINVAL; + ptr = skip_spaces(ptr); + + if (strncmp(ptr, "type=", 5)) + return -EINVAL; + ptr = skip_spaces(ptr + 5); + + i = match_string(mtrr_strings, MTRR_NUM_TYPES, ptr); + if (i < 0) + return i; + + base >>= PAGE_SHIFT; + size >>= PAGE_SHIFT; + err = mtrr_add_page((unsigned long)base, (unsigned long)size, i, true); + if (err < 0) + return err; + return len; +} + +static long +mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) +{ + int err = 0; + mtrr_type type; + unsigned long base; + unsigned long size; + struct mtrr_sentry sentry; + struct mtrr_gentry gentry; + void __user *arg = (void __user *) __arg; + + memset(&gentry, 0, sizeof(gentry)); + + switch (cmd) { + case MTRRIOC_ADD_ENTRY: + case MTRRIOC_SET_ENTRY: + case MTRRIOC_DEL_ENTRY: + case MTRRIOC_KILL_ENTRY: + case MTRRIOC_ADD_PAGE_ENTRY: + case MTRRIOC_SET_PAGE_ENTRY: + case MTRRIOC_DEL_PAGE_ENTRY: + case MTRRIOC_KILL_PAGE_ENTRY: + if (copy_from_user(&sentry, arg, sizeof(sentry))) + return -EFAULT; + break; + case MTRRIOC_GET_ENTRY: + case MTRRIOC_GET_PAGE_ENTRY: + if (copy_from_user(&gentry, arg, sizeof(gentry))) + return -EFAULT; + break; +#ifdef CONFIG_COMPAT + case MTRRIOC32_ADD_ENTRY: + case MTRRIOC32_SET_ENTRY: + case MTRRIOC32_DEL_ENTRY: + case MTRRIOC32_KILL_ENTRY: + case MTRRIOC32_ADD_PAGE_ENTRY: + case MTRRIOC32_SET_PAGE_ENTRY: + case MTRRIOC32_DEL_PAGE_ENTRY: + case MTRRIOC32_KILL_PAGE_ENTRY: { + struct mtrr_sentry32 __user *s32; + + s32 = (struct mtrr_sentry32 __user *)__arg; + err = get_user(sentry.base, &s32->base); + err |= get_user(sentry.size, &s32->size); + err |= get_user(sentry.type, &s32->type); + if (err) + return err; + break; + } + case MTRRIOC32_GET_ENTRY: + case MTRRIOC32_GET_PAGE_ENTRY: { + struct mtrr_gentry32 __user *g32; + + g32 = (struct mtrr_gentry32 __user *)__arg; + err = get_user(gentry.regnum, &g32->regnum); + err |= get_user(gentry.base, &g32->base); + err |= get_user(gentry.size, &g32->size); + err |= get_user(gentry.type, &g32->type); + if (err) + return err; + break; + } +#endif + } + + switch (cmd) { + default: + return -ENOTTY; + case MTRRIOC_ADD_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_ADD_ENTRY: +#endif + err = + mtrr_file_add(sentry.base, sentry.size, sentry.type, true, + file, 0); + break; + case MTRRIOC_SET_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_SET_ENTRY: +#endif + err = mtrr_add(sentry.base, sentry.size, sentry.type, false); + break; + case MTRRIOC_DEL_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_DEL_ENTRY: +#endif + err = mtrr_file_del(sentry.base, sentry.size, file, 0); + break; + case MTRRIOC_KILL_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_KILL_ENTRY: +#endif + err = mtrr_del(-1, sentry.base, sentry.size); + break; + case MTRRIOC_GET_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_GET_ENTRY: +#endif + if (gentry.regnum >= num_var_ranges) + return -EINVAL; + mtrr_if->get(gentry.regnum, &base, &size, &type); + + /* Hide entries that go above 4GB */ + if (base + size - 1 >= (1UL << (8 * sizeof(gentry.size) - PAGE_SHIFT)) + || size >= (1UL << (8 * sizeof(gentry.size) - PAGE_SHIFT))) + gentry.base = gentry.size = gentry.type = 0; + else { + gentry.base = base << PAGE_SHIFT; + gentry.size = size << PAGE_SHIFT; + gentry.type = type; + } + + break; + case MTRRIOC_ADD_PAGE_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_ADD_PAGE_ENTRY: +#endif + err = + mtrr_file_add(sentry.base, sentry.size, sentry.type, true, + file, 1); + break; + case MTRRIOC_SET_PAGE_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_SET_PAGE_ENTRY: +#endif + err = + mtrr_add_page(sentry.base, sentry.size, sentry.type, false); + break; + case MTRRIOC_DEL_PAGE_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_DEL_PAGE_ENTRY: +#endif + err = mtrr_file_del(sentry.base, sentry.size, file, 1); + break; + case MTRRIOC_KILL_PAGE_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_KILL_PAGE_ENTRY: +#endif + err = mtrr_del_page(-1, sentry.base, sentry.size); + break; + case MTRRIOC_GET_PAGE_ENTRY: +#ifdef CONFIG_COMPAT + case MTRRIOC32_GET_PAGE_ENTRY: +#endif + if (gentry.regnum >= num_var_ranges) + return -EINVAL; + mtrr_if->get(gentry.regnum, &base, &size, &type); + /* Hide entries that would overflow */ + if (size != (__typeof__(gentry.size))size) + gentry.base = gentry.size = gentry.type = 0; + else { + gentry.base = base; + gentry.size = size; + gentry.type = type; + } + break; + } + + if (err) + return err; + + switch (cmd) { + case MTRRIOC_GET_ENTRY: + case MTRRIOC_GET_PAGE_ENTRY: + if (copy_to_user(arg, &gentry, sizeof(gentry))) + err = -EFAULT; + break; +#ifdef CONFIG_COMPAT + case MTRRIOC32_GET_ENTRY: + case MTRRIOC32_GET_PAGE_ENTRY: { + struct mtrr_gentry32 __user *g32; + + g32 = (struct mtrr_gentry32 __user *)__arg; + err = put_user(gentry.base, &g32->base); + err |= put_user(gentry.size, &g32->size); + err |= put_user(gentry.regnum, &g32->regnum); + err |= put_user(gentry.type, &g32->type); + break; + } +#endif + } + return err; +} + +static int mtrr_close(struct inode *ino, struct file *file) +{ + unsigned int *fcount = FILE_FCOUNT(file); + int i, max; + + if (fcount != NULL) { + max = num_var_ranges; + for (i = 0; i < max; ++i) { + while (fcount[i] > 0) { + mtrr_del(i, 0, 0); + --fcount[i]; + } + } + kfree(fcount); + FILE_FCOUNT(file) = NULL; + } + return single_release(ino, file); +} + +static int mtrr_seq_show(struct seq_file *seq, void *offset) +{ + char factor; + int i, max; + mtrr_type type; + unsigned long base, size; + + max = num_var_ranges; + for (i = 0; i < max; i++) { + mtrr_if->get(i, &base, &size, &type); + if (size == 0) { + mtrr_usage_table[i] = 0; + continue; + } + if (size < (0x100000 >> PAGE_SHIFT)) { + /* less than 1MB */ + factor = 'K'; + size <<= PAGE_SHIFT - 10; + } else { + factor = 'M'; + size >>= 20 - PAGE_SHIFT; + } + /* Base can be > 32bit */ + seq_printf(seq, "reg%02i: base=0x%06lx000 (%5luMB), size=%5lu%cB, count=%d: %s\n", + i, base, base >> (20 - PAGE_SHIFT), + size, factor, + mtrr_usage_table[i], mtrr_attrib_to_str(type)); + } + return 0; +} + +static int mtrr_open(struct inode *inode, struct file *file) +{ + if (!mtrr_if) + return -EIO; + if (!mtrr_if->get) + return -ENXIO; + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + return single_open(file, mtrr_seq_show, NULL); +} + +static const struct proc_ops mtrr_proc_ops = { + .proc_open = mtrr_open, + .proc_read = seq_read, + .proc_lseek = seq_lseek, + .proc_write = mtrr_write, + .proc_ioctl = mtrr_ioctl, +#ifdef CONFIG_COMPAT + .proc_compat_ioctl = mtrr_ioctl, +#endif + .proc_release = mtrr_close, +}; + +static int __init mtrr_if_init(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + if ((!cpu_has(c, X86_FEATURE_MTRR)) && + (!cpu_has(c, X86_FEATURE_K6_MTRR)) && + (!cpu_has(c, X86_FEATURE_CYRIX_ARR)) && + (!cpu_has(c, X86_FEATURE_CENTAUR_MCR))) + return -ENODEV; + + proc_create("mtrr", S_IWUSR | S_IRUGO, NULL, &mtrr_proc_ops); + return 0; +} +arch_initcall(mtrr_if_init); +#endif /* CONFIG_PROC_FS */ diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.c b/arch/x86/kernel/cpu/mtrr/mtrr.c new file mode 100644 index 000000000..5f436cb4f --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/mtrr.c @@ -0,0 +1,888 @@ +/* Generic MTRR (Memory Type Range Register) driver. + + Copyright (C) 1997-2000 Richard Gooch + Copyright (c) 2002 Patrick Mochel + + This library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public + License as published by the Free Software Foundation; either + version 2 of the License, or (at your option) any later version. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with this library; if not, write to the Free + Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + + Richard Gooch may be reached by email at rgooch@atnf.csiro.au + The postal address is: + Richard Gooch, c/o ATNF, P. O. Box 76, Epping, N.S.W., 2121, Australia. + + Source: "Pentium Pro Family Developer's Manual, Volume 3: + Operating System Writer's Guide" (Intel document number 242692), + section 11.11.7 + + This was cleaned and made readable by Patrick Mochel <mochel@osdl.org> + on 6-7 March 2002. + Source: Intel Architecture Software Developers Manual, Volume 3: + System Programming Guide; Section 9.11. (1997 edition - PPro). +*/ + +#define DEBUG + +#include <linux/types.h> /* FIXME: kvm_para.h needs this */ + +#include <linux/stop_machine.h> +#include <linux/kvm_para.h> +#include <linux/uaccess.h> +#include <linux/export.h> +#include <linux/mutex.h> +#include <linux/init.h> +#include <linux/sort.h> +#include <linux/cpu.h> +#include <linux/pci.h> +#include <linux/smp.h> +#include <linux/syscore_ops.h> +#include <linux/rcupdate.h> + +#include <asm/cpufeature.h> +#include <asm/e820/api.h> +#include <asm/mtrr.h> +#include <asm/msr.h> +#include <asm/memtype.h> + +#include "mtrr.h" + +/* arch_phys_wc_add returns an MTRR register index plus this offset. */ +#define MTRR_TO_PHYS_WC_OFFSET 1000 + +u32 num_var_ranges; +static bool __mtrr_enabled; + +static bool mtrr_enabled(void) +{ + return __mtrr_enabled; +} + +unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES]; +static DEFINE_MUTEX(mtrr_mutex); + +u64 size_or_mask, size_and_mask; +static bool mtrr_aps_delayed_init; + +static const struct mtrr_ops *mtrr_ops[X86_VENDOR_NUM] __ro_after_init; + +const struct mtrr_ops *mtrr_if; + +static void set_mtrr(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type); + +void __init set_mtrr_ops(const struct mtrr_ops *ops) +{ + if (ops->vendor && ops->vendor < X86_VENDOR_NUM) + mtrr_ops[ops->vendor] = ops; +} + +/* Returns non-zero if we have the write-combining memory type */ +static int have_wrcomb(void) +{ + struct pci_dev *dev; + + dev = pci_get_class(PCI_CLASS_BRIDGE_HOST << 8, NULL); + if (dev != NULL) { + /* + * ServerWorks LE chipsets < rev 6 have problems with + * write-combining. Don't allow it and leave room for other + * chipsets to be tagged + */ + if (dev->vendor == PCI_VENDOR_ID_SERVERWORKS && + dev->device == PCI_DEVICE_ID_SERVERWORKS_LE && + dev->revision <= 5) { + pr_info("Serverworks LE rev < 6 detected. Write-combining disabled.\n"); + pci_dev_put(dev); + return 0; + } + /* + * Intel 450NX errata # 23. Non ascending cacheline evictions to + * write combining memory may resulting in data corruption + */ + if (dev->vendor == PCI_VENDOR_ID_INTEL && + dev->device == PCI_DEVICE_ID_INTEL_82451NX) { + pr_info("Intel 450NX MMC detected. Write-combining disabled.\n"); + pci_dev_put(dev); + return 0; + } + pci_dev_put(dev); + } + return mtrr_if->have_wrcomb ? mtrr_if->have_wrcomb() : 0; +} + +/* This function returns the number of variable MTRRs */ +static void __init set_num_var_ranges(void) +{ + unsigned long config = 0, dummy; + + if (use_intel()) + rdmsr(MSR_MTRRcap, config, dummy); + else if (is_cpu(AMD) || is_cpu(HYGON)) + config = 2; + else if (is_cpu(CYRIX) || is_cpu(CENTAUR)) + config = 8; + + num_var_ranges = config & 0xff; +} + +static void __init init_table(void) +{ + int i, max; + + max = num_var_ranges; + for (i = 0; i < max; i++) + mtrr_usage_table[i] = 1; +} + +struct set_mtrr_data { + unsigned long smp_base; + unsigned long smp_size; + unsigned int smp_reg; + mtrr_type smp_type; +}; + +/** + * mtrr_rendezvous_handler - Work done in the synchronization handler. Executed + * by all the CPUs. + * @info: pointer to mtrr configuration data + * + * Returns nothing. + */ +static int mtrr_rendezvous_handler(void *info) +{ + struct set_mtrr_data *data = info; + + /* + * We use this same function to initialize the mtrrs during boot, + * resume, runtime cpu online and on an explicit request to set a + * specific MTRR. + * + * During boot or suspend, the state of the boot cpu's mtrrs has been + * saved, and we want to replicate that across all the cpus that come + * online (either at the end of boot or resume or during a runtime cpu + * online). If we're doing that, @reg is set to something special and on + * all the cpu's we do mtrr_if->set_all() (On the logical cpu that + * started the boot/resume sequence, this might be a duplicate + * set_all()). + */ + if (data->smp_reg != ~0U) { + mtrr_if->set(data->smp_reg, data->smp_base, + data->smp_size, data->smp_type); + } else if (mtrr_aps_delayed_init || !cpu_online(smp_processor_id())) { + mtrr_if->set_all(); + } + return 0; +} + +static inline int types_compatible(mtrr_type type1, mtrr_type type2) +{ + return type1 == MTRR_TYPE_UNCACHABLE || + type2 == MTRR_TYPE_UNCACHABLE || + (type1 == MTRR_TYPE_WRTHROUGH && type2 == MTRR_TYPE_WRBACK) || + (type1 == MTRR_TYPE_WRBACK && type2 == MTRR_TYPE_WRTHROUGH); +} + +/** + * set_mtrr - update mtrrs on all processors + * @reg: mtrr in question + * @base: mtrr base + * @size: mtrr size + * @type: mtrr type + * + * This is kinda tricky, but fortunately, Intel spelled it out for us cleanly: + * + * 1. Queue work to do the following on all processors: + * 2. Disable Interrupts + * 3. Wait for all procs to do so + * 4. Enter no-fill cache mode + * 5. Flush caches + * 6. Clear PGE bit + * 7. Flush all TLBs + * 8. Disable all range registers + * 9. Update the MTRRs + * 10. Enable all range registers + * 11. Flush all TLBs and caches again + * 12. Enter normal cache mode and reenable caching + * 13. Set PGE + * 14. Wait for buddies to catch up + * 15. Enable interrupts. + * + * What does that mean for us? Well, stop_machine() will ensure that + * the rendezvous handler is started on each CPU. And in lockstep they + * do the state transition of disabling interrupts, updating MTRR's + * (the CPU vendors may each do it differently, so we call mtrr_if->set() + * callback and let them take care of it.) and enabling interrupts. + * + * Note that the mechanism is the same for UP systems, too; all the SMP stuff + * becomes nops. + */ +static void +set_mtrr(unsigned int reg, unsigned long base, unsigned long size, mtrr_type type) +{ + struct set_mtrr_data data = { .smp_reg = reg, + .smp_base = base, + .smp_size = size, + .smp_type = type + }; + + stop_machine(mtrr_rendezvous_handler, &data, cpu_online_mask); +} + +static void set_mtrr_cpuslocked(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type) +{ + struct set_mtrr_data data = { .smp_reg = reg, + .smp_base = base, + .smp_size = size, + .smp_type = type + }; + + stop_machine_cpuslocked(mtrr_rendezvous_handler, &data, cpu_online_mask); +} + +static void set_mtrr_from_inactive_cpu(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type) +{ + struct set_mtrr_data data = { .smp_reg = reg, + .smp_base = base, + .smp_size = size, + .smp_type = type + }; + + stop_machine_from_inactive_cpu(mtrr_rendezvous_handler, &data, + cpu_callout_mask); +} + +/** + * mtrr_add_page - Add a memory type region + * @base: Physical base address of region in pages (in units of 4 kB!) + * @size: Physical size of region in pages (4 kB) + * @type: Type of MTRR desired + * @increment: If this is true do usage counting on the region + * + * Memory type region registers control the caching on newer Intel and + * non Intel processors. This function allows drivers to request an + * MTRR is added. The details and hardware specifics of each processor's + * implementation are hidden from the caller, but nevertheless the + * caller should expect to need to provide a power of two size on an + * equivalent power of two boundary. + * + * If the region cannot be added either because all regions are in use + * or the CPU cannot support it a negative value is returned. On success + * the register number for this entry is returned, but should be treated + * as a cookie only. + * + * On a multiprocessor machine the changes are made to all processors. + * This is required on x86 by the Intel processors. + * + * The available types are + * + * %MTRR_TYPE_UNCACHABLE - No caching + * + * %MTRR_TYPE_WRBACK - Write data back in bursts whenever + * + * %MTRR_TYPE_WRCOMB - Write data back soon but allow bursts + * + * %MTRR_TYPE_WRTHROUGH - Cache reads but not writes + * + * BUGS: Needs a quiet flag for the cases where drivers do not mind + * failures and do not wish system log messages to be sent. + */ +int mtrr_add_page(unsigned long base, unsigned long size, + unsigned int type, bool increment) +{ + unsigned long lbase, lsize; + int i, replace, error; + mtrr_type ltype; + + if (!mtrr_enabled()) + return -ENXIO; + + error = mtrr_if->validate_add_page(base, size, type); + if (error) + return error; + + if (type >= MTRR_NUM_TYPES) { + pr_warn("type: %u invalid\n", type); + return -EINVAL; + } + + /* If the type is WC, check that this processor supports it */ + if ((type == MTRR_TYPE_WRCOMB) && !have_wrcomb()) { + pr_warn("your processor doesn't support write-combining\n"); + return -ENOSYS; + } + + if (!size) { + pr_warn("zero sized request\n"); + return -EINVAL; + } + + if ((base | (base + size - 1)) >> + (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) { + pr_warn("base or size exceeds the MTRR width\n"); + return -EINVAL; + } + + error = -EINVAL; + replace = -1; + + /* No CPU hotplug when we change MTRR entries */ + get_online_cpus(); + + /* Search for existing MTRR */ + mutex_lock(&mtrr_mutex); + for (i = 0; i < num_var_ranges; ++i) { + mtrr_if->get(i, &lbase, &lsize, <ype); + if (!lsize || base > lbase + lsize - 1 || + base + size - 1 < lbase) + continue; + /* + * At this point we know there is some kind of + * overlap/enclosure + */ + if (base < lbase || base + size - 1 > lbase + lsize - 1) { + if (base <= lbase && + base + size - 1 >= lbase + lsize - 1) { + /* New region encloses an existing region */ + if (type == ltype) { + replace = replace == -1 ? i : -2; + continue; + } else if (types_compatible(type, ltype)) + continue; + } + pr_warn("0x%lx000,0x%lx000 overlaps existing 0x%lx000,0x%lx000\n", base, size, lbase, + lsize); + goto out; + } + /* New region is enclosed by an existing region */ + if (ltype != type) { + if (types_compatible(type, ltype)) + continue; + pr_warn("type mismatch for %lx000,%lx000 old: %s new: %s\n", + base, size, mtrr_attrib_to_str(ltype), + mtrr_attrib_to_str(type)); + goto out; + } + if (increment) + ++mtrr_usage_table[i]; + error = i; + goto out; + } + /* Search for an empty MTRR */ + i = mtrr_if->get_free_region(base, size, replace); + if (i >= 0) { + set_mtrr_cpuslocked(i, base, size, type); + if (likely(replace < 0)) { + mtrr_usage_table[i] = 1; + } else { + mtrr_usage_table[i] = mtrr_usage_table[replace]; + if (increment) + mtrr_usage_table[i]++; + if (unlikely(replace != i)) { + set_mtrr_cpuslocked(replace, 0, 0, 0); + mtrr_usage_table[replace] = 0; + } + } + } else { + pr_info("no more MTRRs available\n"); + } + error = i; + out: + mutex_unlock(&mtrr_mutex); + put_online_cpus(); + return error; +} + +static int mtrr_check(unsigned long base, unsigned long size) +{ + if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) { + pr_warn("size and base must be multiples of 4 kiB\n"); + pr_debug("size: 0x%lx base: 0x%lx\n", size, base); + dump_stack(); + return -1; + } + return 0; +} + +/** + * mtrr_add - Add a memory type region + * @base: Physical base address of region + * @size: Physical size of region + * @type: Type of MTRR desired + * @increment: If this is true do usage counting on the region + * + * Memory type region registers control the caching on newer Intel and + * non Intel processors. This function allows drivers to request an + * MTRR is added. The details and hardware specifics of each processor's + * implementation are hidden from the caller, but nevertheless the + * caller should expect to need to provide a power of two size on an + * equivalent power of two boundary. + * + * If the region cannot be added either because all regions are in use + * or the CPU cannot support it a negative value is returned. On success + * the register number for this entry is returned, but should be treated + * as a cookie only. + * + * On a multiprocessor machine the changes are made to all processors. + * This is required on x86 by the Intel processors. + * + * The available types are + * + * %MTRR_TYPE_UNCACHABLE - No caching + * + * %MTRR_TYPE_WRBACK - Write data back in bursts whenever + * + * %MTRR_TYPE_WRCOMB - Write data back soon but allow bursts + * + * %MTRR_TYPE_WRTHROUGH - Cache reads but not writes + * + * BUGS: Needs a quiet flag for the cases where drivers do not mind + * failures and do not wish system log messages to be sent. + */ +int mtrr_add(unsigned long base, unsigned long size, unsigned int type, + bool increment) +{ + if (!mtrr_enabled()) + return -ENODEV; + if (mtrr_check(base, size)) + return -EINVAL; + return mtrr_add_page(base >> PAGE_SHIFT, size >> PAGE_SHIFT, type, + increment); +} + +/** + * mtrr_del_page - delete a memory type region + * @reg: Register returned by mtrr_add + * @base: Physical base address + * @size: Size of region + * + * If register is supplied then base and size are ignored. This is + * how drivers should call it. + * + * Releases an MTRR region. If the usage count drops to zero the + * register is freed and the region returns to default state. + * On success the register is returned, on failure a negative error + * code. + */ +int mtrr_del_page(int reg, unsigned long base, unsigned long size) +{ + int i, max; + mtrr_type ltype; + unsigned long lbase, lsize; + int error = -EINVAL; + + if (!mtrr_enabled()) + return -ENODEV; + + max = num_var_ranges; + /* No CPU hotplug when we change MTRR entries */ + get_online_cpus(); + mutex_lock(&mtrr_mutex); + if (reg < 0) { + /* Search for existing MTRR */ + for (i = 0; i < max; ++i) { + mtrr_if->get(i, &lbase, &lsize, <ype); + if (lbase == base && lsize == size) { + reg = i; + break; + } + } + if (reg < 0) { + pr_debug("no MTRR for %lx000,%lx000 found\n", + base, size); + goto out; + } + } + if (reg >= max) { + pr_warn("register: %d too big\n", reg); + goto out; + } + mtrr_if->get(reg, &lbase, &lsize, <ype); + if (lsize < 1) { + pr_warn("MTRR %d not used\n", reg); + goto out; + } + if (mtrr_usage_table[reg] < 1) { + pr_warn("reg: %d has count=0\n", reg); + goto out; + } + if (--mtrr_usage_table[reg] < 1) + set_mtrr_cpuslocked(reg, 0, 0, 0); + error = reg; + out: + mutex_unlock(&mtrr_mutex); + put_online_cpus(); + return error; +} + +/** + * mtrr_del - delete a memory type region + * @reg: Register returned by mtrr_add + * @base: Physical base address + * @size: Size of region + * + * If register is supplied then base and size are ignored. This is + * how drivers should call it. + * + * Releases an MTRR region. If the usage count drops to zero the + * register is freed and the region returns to default state. + * On success the register is returned, on failure a negative error + * code. + */ +int mtrr_del(int reg, unsigned long base, unsigned long size) +{ + if (!mtrr_enabled()) + return -ENODEV; + if (mtrr_check(base, size)) + return -EINVAL; + return mtrr_del_page(reg, base >> PAGE_SHIFT, size >> PAGE_SHIFT); +} + +/** + * arch_phys_wc_add - add a WC MTRR and handle errors if PAT is unavailable + * @base: Physical base address + * @size: Size of region + * + * If PAT is available, this does nothing. If PAT is unavailable, it + * attempts to add a WC MTRR covering size bytes starting at base and + * logs an error if this fails. + * + * The called should provide a power of two size on an equivalent + * power of two boundary. + * + * Drivers must store the return value to pass to mtrr_del_wc_if_needed, + * but drivers should not try to interpret that return value. + */ +int arch_phys_wc_add(unsigned long base, unsigned long size) +{ + int ret; + + if (pat_enabled() || !mtrr_enabled()) + return 0; /* Success! (We don't need to do anything.) */ + + ret = mtrr_add(base, size, MTRR_TYPE_WRCOMB, true); + if (ret < 0) { + pr_warn("Failed to add WC MTRR for [%p-%p]; performance may suffer.", + (void *)base, (void *)(base + size - 1)); + return ret; + } + return ret + MTRR_TO_PHYS_WC_OFFSET; +} +EXPORT_SYMBOL(arch_phys_wc_add); + +/* + * arch_phys_wc_del - undoes arch_phys_wc_add + * @handle: Return value from arch_phys_wc_add + * + * This cleans up after mtrr_add_wc_if_needed. + * + * The API guarantees that mtrr_del_wc_if_needed(error code) and + * mtrr_del_wc_if_needed(0) do nothing. + */ +void arch_phys_wc_del(int handle) +{ + if (handle >= 1) { + WARN_ON(handle < MTRR_TO_PHYS_WC_OFFSET); + mtrr_del(handle - MTRR_TO_PHYS_WC_OFFSET, 0, 0); + } +} +EXPORT_SYMBOL(arch_phys_wc_del); + +/* + * arch_phys_wc_index - translates arch_phys_wc_add's return value + * @handle: Return value from arch_phys_wc_add + * + * This will turn the return value from arch_phys_wc_add into an mtrr + * index suitable for debugging. + * + * Note: There is no legitimate use for this function, except possibly + * in printk line. Alas there is an illegitimate use in some ancient + * drm ioctls. + */ +int arch_phys_wc_index(int handle) +{ + if (handle < MTRR_TO_PHYS_WC_OFFSET) + return -1; + else + return handle - MTRR_TO_PHYS_WC_OFFSET; +} +EXPORT_SYMBOL_GPL(arch_phys_wc_index); + +/* + * HACK ALERT! + * These should be called implicitly, but we can't yet until all the initcall + * stuff is done... + */ +static void __init init_ifs(void) +{ +#ifndef CONFIG_X86_64 + amd_init_mtrr(); + cyrix_init_mtrr(); + centaur_init_mtrr(); +#endif +} + +/* The suspend/resume methods are only for CPU without MTRR. CPU using generic + * MTRR driver doesn't require this + */ +struct mtrr_value { + mtrr_type ltype; + unsigned long lbase; + unsigned long lsize; +}; + +static struct mtrr_value mtrr_value[MTRR_MAX_VAR_RANGES]; + +static int mtrr_save(void) +{ + int i; + + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &mtrr_value[i].lbase, + &mtrr_value[i].lsize, + &mtrr_value[i].ltype); + } + return 0; +} + +static void mtrr_restore(void) +{ + int i; + + for (i = 0; i < num_var_ranges; i++) { + if (mtrr_value[i].lsize) { + set_mtrr(i, mtrr_value[i].lbase, + mtrr_value[i].lsize, + mtrr_value[i].ltype); + } + } +} + + + +static struct syscore_ops mtrr_syscore_ops = { + .suspend = mtrr_save, + .resume = mtrr_restore, +}; + +int __initdata changed_by_mtrr_cleanup; + +#define SIZE_OR_MASK_BITS(n) (~((1ULL << ((n) - PAGE_SHIFT)) - 1)) +/** + * mtrr_bp_init - initialize mtrrs on the boot CPU + * + * This needs to be called early; before any of the other CPUs are + * initialized (i.e. before smp_init()). + * + */ +void __init mtrr_bp_init(void) +{ + u32 phys_addr; + + init_ifs(); + + phys_addr = 32; + + if (boot_cpu_has(X86_FEATURE_MTRR)) { + mtrr_if = &generic_mtrr_ops; + size_or_mask = SIZE_OR_MASK_BITS(36); + size_and_mask = 0x00f00000; + phys_addr = 36; + + /* + * This is an AMD specific MSR, but we assume(hope?) that + * Intel will implement it too when they extend the address + * bus of the Xeon. + */ + if (cpuid_eax(0x80000000) >= 0x80000008) { + phys_addr = cpuid_eax(0x80000008) & 0xff; + /* CPUID workaround for Intel 0F33/0F34 CPU */ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && + boot_cpu_data.x86 == 0xF && + boot_cpu_data.x86_model == 0x3 && + (boot_cpu_data.x86_stepping == 0x3 || + boot_cpu_data.x86_stepping == 0x4)) + phys_addr = 36; + + size_or_mask = SIZE_OR_MASK_BITS(phys_addr); + size_and_mask = ~size_or_mask & 0xfffff00000ULL; + } else if (boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR && + boot_cpu_data.x86 == 6) { + /* + * VIA C* family have Intel style MTRRs, + * but don't support PAE + */ + size_or_mask = SIZE_OR_MASK_BITS(32); + size_and_mask = 0; + phys_addr = 32; + } + } else { + switch (boot_cpu_data.x86_vendor) { + case X86_VENDOR_AMD: + if (cpu_feature_enabled(X86_FEATURE_K6_MTRR)) { + /* Pre-Athlon (K6) AMD CPU MTRRs */ + mtrr_if = mtrr_ops[X86_VENDOR_AMD]; + size_or_mask = SIZE_OR_MASK_BITS(32); + size_and_mask = 0; + } + break; + case X86_VENDOR_CENTAUR: + if (cpu_feature_enabled(X86_FEATURE_CENTAUR_MCR)) { + mtrr_if = mtrr_ops[X86_VENDOR_CENTAUR]; + size_or_mask = SIZE_OR_MASK_BITS(32); + size_and_mask = 0; + } + break; + case X86_VENDOR_CYRIX: + if (cpu_feature_enabled(X86_FEATURE_CYRIX_ARR)) { + mtrr_if = mtrr_ops[X86_VENDOR_CYRIX]; + size_or_mask = SIZE_OR_MASK_BITS(32); + size_and_mask = 0; + } + break; + default: + break; + } + } + + if (mtrr_if) { + __mtrr_enabled = true; + set_num_var_ranges(); + init_table(); + if (use_intel()) { + /* BIOS may override */ + __mtrr_enabled = get_mtrr_state(); + + if (mtrr_enabled()) + mtrr_bp_pat_init(); + + if (mtrr_cleanup(phys_addr)) { + changed_by_mtrr_cleanup = 1; + mtrr_if->set_all(); + } + } + } + + if (!mtrr_enabled()) { + pr_info("Disabled\n"); + + /* + * PAT initialization relies on MTRR's rendezvous handler. + * Skip PAT init until the handler can initialize both + * features independently. + */ + pat_disable("MTRRs disabled, skipping PAT initialization too."); + } +} + +void mtrr_ap_init(void) +{ + if (!mtrr_enabled()) + return; + + if (!use_intel() || mtrr_aps_delayed_init) + return; + + /* + * Ideally we should hold mtrr_mutex here to avoid mtrr entries + * changed, but this routine will be called in cpu boot time, + * holding the lock breaks it. + * + * This routine is called in two cases: + * + * 1. very earily time of software resume, when there absolutely + * isn't mtrr entry changes; + * + * 2. cpu hotadd time. We let mtrr_add/del_page hold cpuhotplug + * lock to prevent mtrr entry changes + */ + set_mtrr_from_inactive_cpu(~0U, 0, 0, 0); +} + +/** + * Save current fixed-range MTRR state of the first cpu in cpu_online_mask. + */ +void mtrr_save_state(void) +{ + int first_cpu; + + if (!mtrr_enabled()) + return; + + first_cpu = cpumask_first(cpu_online_mask); + smp_call_function_single(first_cpu, mtrr_save_fixed_ranges, NULL, 1); +} + +void set_mtrr_aps_delayed_init(void) +{ + if (!mtrr_enabled()) + return; + if (!use_intel()) + return; + + mtrr_aps_delayed_init = true; +} + +/* + * Delayed MTRR initialization for all AP's + */ +void mtrr_aps_init(void) +{ + if (!use_intel() || !mtrr_enabled()) + return; + + /* + * Check if someone has requested the delay of AP MTRR initialization, + * by doing set_mtrr_aps_delayed_init(), prior to this point. If not, + * then we are done. + */ + if (!mtrr_aps_delayed_init) + return; + + set_mtrr(~0U, 0, 0, 0); + mtrr_aps_delayed_init = false; +} + +void mtrr_bp_restore(void) +{ + if (!use_intel() || !mtrr_enabled()) + return; + + mtrr_if->set_all(); +} + +static int __init mtrr_init_finialize(void) +{ + if (!mtrr_enabled()) + return 0; + + if (use_intel()) { + if (!changed_by_mtrr_cleanup) + mtrr_state_warn(); + return 0; + } + + /* + * The CPU has no MTRR and seems to not support SMP. They have + * specific drivers, we use a tricky method to support + * suspend/resume for them. + * + * TBD: is there any system with such CPU which supports + * suspend/resume? If no, we should remove the code. + */ + register_syscore_ops(&mtrr_syscore_ops); + + return 0; +} +subsys_initcall(mtrr_init_finialize); diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h new file mode 100644 index 000000000..2ac99e561 --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/mtrr.h @@ -0,0 +1,80 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * local MTRR defines. + */ + +#include <linux/types.h> +#include <linux/stddef.h> + +#define MTRR_CHANGE_MASK_FIXED 0x01 +#define MTRR_CHANGE_MASK_VARIABLE 0x02 +#define MTRR_CHANGE_MASK_DEFTYPE 0x04 + +extern unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES]; + +struct mtrr_ops { + u32 vendor; + u32 use_intel_if; + void (*set)(unsigned int reg, unsigned long base, + unsigned long size, mtrr_type type); + void (*set_all)(void); + + void (*get)(unsigned int reg, unsigned long *base, + unsigned long *size, mtrr_type *type); + int (*get_free_region)(unsigned long base, unsigned long size, + int replace_reg); + int (*validate_add_page)(unsigned long base, unsigned long size, + unsigned int type); + int (*have_wrcomb)(void); +}; + +extern int generic_get_free_region(unsigned long base, unsigned long size, + int replace_reg); +extern int generic_validate_add_page(unsigned long base, unsigned long size, + unsigned int type); + +extern const struct mtrr_ops generic_mtrr_ops; + +extern int positive_have_wrcomb(void); + +/* library functions for processor-specific routines */ +struct set_mtrr_context { + unsigned long flags; + unsigned long cr4val; + u32 deftype_lo; + u32 deftype_hi; + u32 ccr3; +}; + +void set_mtrr_done(struct set_mtrr_context *ctxt); +void set_mtrr_cache_disable(struct set_mtrr_context *ctxt); +void set_mtrr_prepare_save(struct set_mtrr_context *ctxt); + +void fill_mtrr_var_range(unsigned int index, + u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi); +bool get_mtrr_state(void); +void mtrr_bp_pat_init(void); + +extern void __init set_mtrr_ops(const struct mtrr_ops *ops); + +extern u64 size_or_mask, size_and_mask; +extern const struct mtrr_ops *mtrr_if; + +#define is_cpu(vnd) (mtrr_if && mtrr_if->vendor == X86_VENDOR_##vnd) +#define use_intel() (mtrr_if && mtrr_if->use_intel_if == 1) + +extern unsigned int num_var_ranges; +extern u64 mtrr_tom2; +extern struct mtrr_state_type mtrr_state; + +void mtrr_state_warn(void); +const char *mtrr_attrib_to_str(int x); +void mtrr_wrmsr(unsigned, unsigned, unsigned); + +/* CPU specific mtrr init functions */ +int amd_init_mtrr(void); +int cyrix_init_mtrr(void); +int centaur_init_mtrr(void); + +extern int changed_by_mtrr_cleanup; +extern int mtrr_cleanup(unsigned address_bits); diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c new file mode 100644 index 000000000..a548d9104 --- /dev/null +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -0,0 +1,171 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * local apic based NMI watchdog for various CPUs. + * + * This file also handles reservation of performance counters for coordination + * with other users (like oprofile). + * + * Note that these events normally don't tick when the CPU idles. This means + * the frequency varies with CPU load. + * + * Original code for K7/P6 written by Keith Owens + * + */ + +#include <linux/percpu.h> +#include <linux/export.h> +#include <linux/kernel.h> +#include <linux/bitops.h> +#include <linux/smp.h> +#include <asm/nmi.h> +#include <linux/kprobes.h> + +#include <asm/apic.h> +#include <asm/perf_event.h> + +/* + * this number is calculated from Intel's MSR_P4_CRU_ESCR5 register and it's + * offset from MSR_P4_BSU_ESCR0. + * + * It will be the max for all platforms (for now) + */ +#define NMI_MAX_COUNTER_BITS 66 + +/* + * perfctr_nmi_owner tracks the ownership of the perfctr registers: + * evtsel_nmi_owner tracks the ownership of the event selection + * - different performance counters/ event selection may be reserved for + * different subsystems this reservation system just tries to coordinate + * things a little + */ +static DECLARE_BITMAP(perfctr_nmi_owner, NMI_MAX_COUNTER_BITS); +static DECLARE_BITMAP(evntsel_nmi_owner, NMI_MAX_COUNTER_BITS); + +/* converts an msr to an appropriate reservation bit */ +static inline unsigned int nmi_perfctr_msr_to_bit(unsigned int msr) +{ + /* returns the bit offset of the performance counter register */ + switch (boot_cpu_data.x86_vendor) { + case X86_VENDOR_HYGON: + case X86_VENDOR_AMD: + if (msr >= MSR_F15H_PERF_CTR) + return (msr - MSR_F15H_PERF_CTR) >> 1; + return msr - MSR_K7_PERFCTR0; + case X86_VENDOR_INTEL: + if (cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) + return msr - MSR_ARCH_PERFMON_PERFCTR0; + + switch (boot_cpu_data.x86) { + case 6: + return msr - MSR_P6_PERFCTR0; + case 11: + return msr - MSR_KNC_PERFCTR0; + case 15: + return msr - MSR_P4_BPU_PERFCTR0; + } + break; + case X86_VENDOR_ZHAOXIN: + case X86_VENDOR_CENTAUR: + return msr - MSR_ARCH_PERFMON_PERFCTR0; + } + return 0; +} + +/* + * converts an msr to an appropriate reservation bit + * returns the bit offset of the event selection register + */ +static inline unsigned int nmi_evntsel_msr_to_bit(unsigned int msr) +{ + /* returns the bit offset of the event selection register */ + switch (boot_cpu_data.x86_vendor) { + case X86_VENDOR_HYGON: + case X86_VENDOR_AMD: + if (msr >= MSR_F15H_PERF_CTL) + return (msr - MSR_F15H_PERF_CTL) >> 1; + return msr - MSR_K7_EVNTSEL0; + case X86_VENDOR_INTEL: + if (cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) + return msr - MSR_ARCH_PERFMON_EVENTSEL0; + + switch (boot_cpu_data.x86) { + case 6: + return msr - MSR_P6_EVNTSEL0; + case 11: + return msr - MSR_KNC_EVNTSEL0; + case 15: + return msr - MSR_P4_BSU_ESCR0; + } + break; + case X86_VENDOR_ZHAOXIN: + case X86_VENDOR_CENTAUR: + return msr - MSR_ARCH_PERFMON_EVENTSEL0; + } + return 0; + +} + +/* checks for a bit availability (hack for oprofile) */ +int avail_to_resrv_perfctr_nmi_bit(unsigned int counter) +{ + BUG_ON(counter > NMI_MAX_COUNTER_BITS); + + return !test_bit(counter, perfctr_nmi_owner); +} +EXPORT_SYMBOL(avail_to_resrv_perfctr_nmi_bit); + +int reserve_perfctr_nmi(unsigned int msr) +{ + unsigned int counter; + + counter = nmi_perfctr_msr_to_bit(msr); + /* register not managed by the allocator? */ + if (counter > NMI_MAX_COUNTER_BITS) + return 1; + + if (!test_and_set_bit(counter, perfctr_nmi_owner)) + return 1; + return 0; +} +EXPORT_SYMBOL(reserve_perfctr_nmi); + +void release_perfctr_nmi(unsigned int msr) +{ + unsigned int counter; + + counter = nmi_perfctr_msr_to_bit(msr); + /* register not managed by the allocator? */ + if (counter > NMI_MAX_COUNTER_BITS) + return; + + clear_bit(counter, perfctr_nmi_owner); +} +EXPORT_SYMBOL(release_perfctr_nmi); + +int reserve_evntsel_nmi(unsigned int msr) +{ + unsigned int counter; + + counter = nmi_evntsel_msr_to_bit(msr); + /* register not managed by the allocator? */ + if (counter > NMI_MAX_COUNTER_BITS) + return 1; + + if (!test_and_set_bit(counter, evntsel_nmi_owner)) + return 1; + return 0; +} +EXPORT_SYMBOL(reserve_evntsel_nmi); + +void release_evntsel_nmi(unsigned int msr) +{ + unsigned int counter; + + counter = nmi_evntsel_msr_to_bit(msr); + /* register not managed by the allocator? */ + if (counter > NMI_MAX_COUNTER_BITS) + return; + + clear_bit(counter, evntsel_nmi_owner); +} +EXPORT_SYMBOL(release_evntsel_nmi); diff --git a/arch/x86/kernel/cpu/powerflags.c b/arch/x86/kernel/cpu/powerflags.c new file mode 100644 index 000000000..fd6ec2aa0 --- /dev/null +++ b/arch/x86/kernel/cpu/powerflags.c @@ -0,0 +1,24 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Strings for the various x86 power flags + * + * This file must not contain any executable code. + */ + +#include <asm/cpufeature.h> + +const char *const x86_power_flags[32] = { + "ts", /* temperature sensor */ + "fid", /* frequency id control */ + "vid", /* voltage id control */ + "ttp", /* thermal trip */ + "tm", /* hardware thermal control */ + "stc", /* software thermal control */ + "100mhzsteps", /* 100 MHz multiplier control */ + "hwpstate", /* hardware P-state control */ + "", /* tsc invariant mapped to constant_tsc */ + "cpb", /* core performance boost */ + "eff_freq_ro", /* Readonly aperf/mperf */ + "proc_feedback", /* processor feedback interface */ + "acc_power", /* accumulated power mechanism */ +}; diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c new file mode 100644 index 000000000..4eec8889b --- /dev/null +++ b/arch/x86/kernel/cpu/proc.c @@ -0,0 +1,182 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/smp.h> +#include <linux/timex.h> +#include <linux/string.h> +#include <linux/seq_file.h> +#include <linux/cpufreq.h> + +#include "cpu.h" + +#ifdef CONFIG_X86_VMX_FEATURE_NAMES +extern const char * const x86_vmx_flags[NVMXINTS*32]; +#endif + +/* + * Get CPU information for use by the procfs. + */ +static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, + unsigned int cpu) +{ +#ifdef CONFIG_SMP + seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); + seq_printf(m, "siblings\t: %d\n", + cpumask_weight(topology_core_cpumask(cpu))); + seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); + seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); + seq_printf(m, "apicid\t\t: %d\n", c->apicid); + seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); +#endif +} + +#ifdef CONFIG_X86_32 +static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c) +{ + seq_printf(m, + "fdiv_bug\t: %s\n" + "f00f_bug\t: %s\n" + "coma_bug\t: %s\n" + "fpu\t\t: %s\n" + "fpu_exception\t: %s\n" + "cpuid level\t: %d\n" + "wp\t\t: yes\n", + boot_cpu_has_bug(X86_BUG_FDIV) ? "yes" : "no", + boot_cpu_has_bug(X86_BUG_F00F) ? "yes" : "no", + boot_cpu_has_bug(X86_BUG_COMA) ? "yes" : "no", + boot_cpu_has(X86_FEATURE_FPU) ? "yes" : "no", + boot_cpu_has(X86_FEATURE_FPU) ? "yes" : "no", + c->cpuid_level); +} +#else +static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c) +{ + seq_printf(m, + "fpu\t\t: yes\n" + "fpu_exception\t: yes\n" + "cpuid level\t: %d\n" + "wp\t\t: yes\n", + c->cpuid_level); +} +#endif + +static int show_cpuinfo(struct seq_file *m, void *v) +{ + struct cpuinfo_x86 *c = v; + unsigned int cpu; + int i; + + cpu = c->cpu_index; + seq_printf(m, "processor\t: %u\n" + "vendor_id\t: %s\n" + "cpu family\t: %d\n" + "model\t\t: %u\n" + "model name\t: %s\n", + cpu, + c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown", + c->x86, + c->x86_model, + c->x86_model_id[0] ? c->x86_model_id : "unknown"); + + if (c->x86_stepping || c->cpuid_level >= 0) + seq_printf(m, "stepping\t: %d\n", c->x86_stepping); + else + seq_puts(m, "stepping\t: unknown\n"); + if (c->microcode) + seq_printf(m, "microcode\t: 0x%x\n", c->microcode); + + if (cpu_has(c, X86_FEATURE_TSC)) { + unsigned int freq = aperfmperf_get_khz(cpu); + + if (!freq) + freq = cpufreq_quick_get(cpu); + if (!freq) + freq = cpu_khz; + seq_printf(m, "cpu MHz\t\t: %u.%03u\n", + freq / 1000, (freq % 1000)); + } + + /* Cache size */ + if (c->x86_cache_size) + seq_printf(m, "cache size\t: %u KB\n", c->x86_cache_size); + + show_cpuinfo_core(m, c, cpu); + show_cpuinfo_misc(m, c); + + seq_puts(m, "flags\t\t:"); + for (i = 0; i < 32*NCAPINTS; i++) + if (cpu_has(c, i) && x86_cap_flags[i] != NULL) + seq_printf(m, " %s", x86_cap_flags[i]); + +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + if (cpu_has(c, X86_FEATURE_VMX) && c->vmx_capability[0]) { + seq_puts(m, "\nvmx flags\t:"); + for (i = 0; i < 32*NVMXINTS; i++) { + if (test_bit(i, (unsigned long *)c->vmx_capability) && + x86_vmx_flags[i] != NULL) + seq_printf(m, " %s", x86_vmx_flags[i]); + } + } +#endif + + seq_puts(m, "\nbugs\t\t:"); + for (i = 0; i < 32*NBUGINTS; i++) { + unsigned int bug_bit = 32*NCAPINTS + i; + + if (cpu_has_bug(c, bug_bit) && x86_bug_flags[i]) + seq_printf(m, " %s", x86_bug_flags[i]); + } + + seq_printf(m, "\nbogomips\t: %lu.%02lu\n", + c->loops_per_jiffy/(500000/HZ), + (c->loops_per_jiffy/(5000/HZ)) % 100); + +#ifdef CONFIG_X86_64 + if (c->x86_tlbsize > 0) + seq_printf(m, "TLB size\t: %d 4K pages\n", c->x86_tlbsize); +#endif + seq_printf(m, "clflush size\t: %u\n", c->x86_clflush_size); + seq_printf(m, "cache_alignment\t: %d\n", c->x86_cache_alignment); + seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n", + c->x86_phys_bits, c->x86_virt_bits); + + seq_puts(m, "power management:"); + for (i = 0; i < 32; i++) { + if (c->x86_power & (1 << i)) { + if (i < ARRAY_SIZE(x86_power_flags) && + x86_power_flags[i]) + seq_printf(m, "%s%s", + x86_power_flags[i][0] ? " " : "", + x86_power_flags[i]); + else + seq_printf(m, " [%d]", i); + } + } + + seq_puts(m, "\n\n"); + + return 0; +} + +static void *c_start(struct seq_file *m, loff_t *pos) +{ + *pos = cpumask_next(*pos - 1, cpu_online_mask); + if ((*pos) < nr_cpu_ids) + return &cpu_data(*pos); + return NULL; +} + +static void *c_next(struct seq_file *m, void *v, loff_t *pos) +{ + (*pos)++; + return c_start(m, pos); +} + +static void c_stop(struct seq_file *m, void *v) +{ +} + +const struct seq_operations cpuinfo_op = { + .start = c_start, + .next = c_next, + .stop = c_stop, + .show = show_cpuinfo, +}; diff --git a/arch/x86/kernel/cpu/rdrand.c b/arch/x86/kernel/cpu/rdrand.c new file mode 100644 index 000000000..c4be62058 --- /dev/null +++ b/arch/x86/kernel/cpu/rdrand.c @@ -0,0 +1,66 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * This file is part of the Linux kernel. + * + * Copyright (c) 2011, Intel Corporation + * Authors: Fenghua Yu <fenghua.yu@intel.com>, + * H. Peter Anvin <hpa@linux.intel.com> + */ + +#include <asm/processor.h> +#include <asm/archrandom.h> +#include <asm/sections.h> + +static int __init x86_rdrand_setup(char *s) +{ + setup_clear_cpu_cap(X86_FEATURE_RDRAND); + setup_clear_cpu_cap(X86_FEATURE_RDSEED); + return 1; +} +__setup("nordrand", x86_rdrand_setup); + +/* + * RDRAND has Built-In-Self-Test (BIST) that runs on every invocation. + * Run the instruction a few times as a sanity check. + * If it fails, it is simple to disable RDRAND here. + */ +#define SANITY_CHECK_LOOPS 8 + +#ifdef CONFIG_ARCH_RANDOM +void x86_init_rdrand(struct cpuinfo_x86 *c) +{ + unsigned int changed = 0; + unsigned long tmp, prev; + int i; + + if (!cpu_has(c, X86_FEATURE_RDRAND)) + return; + + for (i = 0; i < SANITY_CHECK_LOOPS; i++) { + if (!rdrand_long(&tmp)) { + clear_cpu_cap(c, X86_FEATURE_RDRAND); + pr_warn_once("rdrand: disabled\n"); + return; + } + } + + /* + * Stupid sanity-check whether RDRAND does *actually* generate + * some at least random-looking data. + */ + prev = tmp; + for (i = 0; i < SANITY_CHECK_LOOPS; i++) { + if (rdrand_long(&tmp)) { + if (prev != tmp) + changed++; + + prev = tmp; + } + } + + if (WARN_ON_ONCE(!changed)) + pr_emerg( +"RDRAND gives funky smelling output, might consider not using it by booting with \"nordrand\""); + +} +#endif diff --git a/arch/x86/kernel/cpu/resctrl/Makefile b/arch/x86/kernel/cpu/resctrl/Makefile new file mode 100644 index 000000000..4a06c37b9 --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/Makefile @@ -0,0 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-$(CONFIG_X86_CPU_RESCTRL) += core.o rdtgroup.o monitor.o +obj-$(CONFIG_X86_CPU_RESCTRL) += ctrlmondata.o pseudo_lock.o +CFLAGS_pseudo_lock.o = -I$(src) diff --git a/arch/x86/kernel/cpu/resctrl/core.c b/arch/x86/kernel/cpu/resctrl/core.c new file mode 100644 index 000000000..4ccb9039f --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/core.c @@ -0,0 +1,1047 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Resource Director Technology(RDT) + * - Cache Allocation code. + * + * Copyright (C) 2016 Intel Corporation + * + * Authors: + * Fenghua Yu <fenghua.yu@intel.com> + * Tony Luck <tony.luck@intel.com> + * Vikas Shivappa <vikas.shivappa@intel.com> + * + * More information about RDT be found in the Intel (R) x86 Architecture + * Software Developer Manual June 2016, volume 3, section 17.17. + */ + +#define pr_fmt(fmt) "resctrl: " fmt + +#include <linux/slab.h> +#include <linux/err.h> +#include <linux/cacheinfo.h> +#include <linux/cpuhotplug.h> + +#include <asm/intel-family.h> +#include <asm/resctrl.h> +#include "internal.h" + +/* Mutex to protect rdtgroup access. */ +DEFINE_MUTEX(rdtgroup_mutex); + +/* + * The cached resctrl_pqr_state is strictly per CPU and can never be + * updated from a remote CPU. Functions which modify the state + * are called with interrupts disabled and no preemption, which + * is sufficient for the protection. + */ +DEFINE_PER_CPU(struct resctrl_pqr_state, pqr_state); + +/* + * Used to store the max resource name width and max resource data width + * to display the schemata in a tabular format + */ +int max_name_width, max_data_width; + +/* + * Global boolean for rdt_alloc which is true if any + * resource allocation is enabled. + */ +bool rdt_alloc_capable; + +static void +mba_wrmsr_intel(struct rdt_domain *d, struct msr_param *m, + struct rdt_resource *r); +static void +cat_wrmsr(struct rdt_domain *d, struct msr_param *m, struct rdt_resource *r); +static void +mba_wrmsr_amd(struct rdt_domain *d, struct msr_param *m, + struct rdt_resource *r); + +#define domain_init(id) LIST_HEAD_INIT(rdt_resources_all[id].domains) + +struct rdt_resource rdt_resources_all[] = { + [RDT_RESOURCE_L3] = + { + .rid = RDT_RESOURCE_L3, + .name = "L3", + .domains = domain_init(RDT_RESOURCE_L3), + .msr_base = MSR_IA32_L3_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 3, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 1, + .cbm_idx_offset = 0, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_L3DATA] = + { + .rid = RDT_RESOURCE_L3DATA, + .name = "L3DATA", + .domains = domain_init(RDT_RESOURCE_L3DATA), + .msr_base = MSR_IA32_L3_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 3, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 2, + .cbm_idx_offset = 0, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_L3CODE] = + { + .rid = RDT_RESOURCE_L3CODE, + .name = "L3CODE", + .domains = domain_init(RDT_RESOURCE_L3CODE), + .msr_base = MSR_IA32_L3_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 3, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 2, + .cbm_idx_offset = 1, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_L2] = + { + .rid = RDT_RESOURCE_L2, + .name = "L2", + .domains = domain_init(RDT_RESOURCE_L2), + .msr_base = MSR_IA32_L2_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 2, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 1, + .cbm_idx_offset = 0, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_L2DATA] = + { + .rid = RDT_RESOURCE_L2DATA, + .name = "L2DATA", + .domains = domain_init(RDT_RESOURCE_L2DATA), + .msr_base = MSR_IA32_L2_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 2, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 2, + .cbm_idx_offset = 0, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_L2CODE] = + { + .rid = RDT_RESOURCE_L2CODE, + .name = "L2CODE", + .domains = domain_init(RDT_RESOURCE_L2CODE), + .msr_base = MSR_IA32_L2_CBM_BASE, + .msr_update = cat_wrmsr, + .cache_level = 2, + .cache = { + .min_cbm_bits = 1, + .cbm_idx_mult = 2, + .cbm_idx_offset = 1, + }, + .parse_ctrlval = parse_cbm, + .format_str = "%d=%0*x", + .fflags = RFTYPE_RES_CACHE, + }, + [RDT_RESOURCE_MBA] = + { + .rid = RDT_RESOURCE_MBA, + .name = "MB", + .domains = domain_init(RDT_RESOURCE_MBA), + .cache_level = 3, + .parse_ctrlval = parse_bw, + .format_str = "%d=%*u", + .fflags = RFTYPE_RES_MB, + }, +}; + +static unsigned int cbm_idx(struct rdt_resource *r, unsigned int closid) +{ + return closid * r->cache.cbm_idx_mult + r->cache.cbm_idx_offset; +} + +/* + * cache_alloc_hsw_probe() - Have to probe for Intel haswell server CPUs + * as they do not have CPUID enumeration support for Cache allocation. + * The check for Vendor/Family/Model is not enough to guarantee that + * the MSRs won't #GP fault because only the following SKUs support + * CAT: + * Intel(R) Xeon(R) CPU E5-2658 v3 @ 2.20GHz + * Intel(R) Xeon(R) CPU E5-2648L v3 @ 1.80GHz + * Intel(R) Xeon(R) CPU E5-2628L v3 @ 2.00GHz + * Intel(R) Xeon(R) CPU E5-2618L v3 @ 2.30GHz + * Intel(R) Xeon(R) CPU E5-2608L v3 @ 2.00GHz + * Intel(R) Xeon(R) CPU E5-2658A v3 @ 2.20GHz + * + * Probe by trying to write the first of the L3 cach mask registers + * and checking that the bits stick. Max CLOSids is always 4 and max cbm length + * is always 20 on hsw server parts. The minimum cache bitmask length + * allowed for HSW server is always 2 bits. Hardcode all of them. + */ +static inline void cache_alloc_hsw_probe(void) +{ + struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3]; + u32 l, h, max_cbm = BIT_MASK(20) - 1; + + if (wrmsr_safe(MSR_IA32_L3_CBM_BASE, max_cbm, 0)) + return; + + rdmsr(MSR_IA32_L3_CBM_BASE, l, h); + + /* If all the bits were set in MSR, return success */ + if (l != max_cbm) + return; + + r->num_closid = 4; + r->default_ctrl = max_cbm; + r->cache.cbm_len = 20; + r->cache.shareable_bits = 0xc0000; + r->cache.min_cbm_bits = 2; + r->alloc_capable = true; + r->alloc_enabled = true; + + rdt_alloc_capable = true; +} + +bool is_mba_sc(struct rdt_resource *r) +{ + if (!r) + return rdt_resources_all[RDT_RESOURCE_MBA].membw.mba_sc; + + return r->membw.mba_sc; +} + +/* + * rdt_get_mb_table() - get a mapping of bandwidth(b/w) percentage values + * exposed to user interface and the h/w understandable delay values. + * + * The non-linear delay values have the granularity of power of two + * and also the h/w does not guarantee a curve for configured delay + * values vs. actual b/w enforced. + * Hence we need a mapping that is pre calibrated so the user can + * express the memory b/w as a percentage value. + */ +static inline bool rdt_get_mb_table(struct rdt_resource *r) +{ + /* + * There are no Intel SKUs as of now to support non-linear delay. + */ + pr_info("MBA b/w map not implemented for cpu:%d, model:%d", + boot_cpu_data.x86, boot_cpu_data.x86_model); + + return false; +} + +static bool __get_mem_config_intel(struct rdt_resource *r) +{ + union cpuid_0x10_3_eax eax; + union cpuid_0x10_x_edx edx; + u32 ebx, ecx, max_delay; + + cpuid_count(0x00000010, 3, &eax.full, &ebx, &ecx, &edx.full); + r->num_closid = edx.split.cos_max + 1; + max_delay = eax.split.max_delay + 1; + r->default_ctrl = MAX_MBA_BW; + r->membw.arch_needs_linear = true; + if (ecx & MBA_IS_LINEAR) { + r->membw.delay_linear = true; + r->membw.min_bw = MAX_MBA_BW - max_delay; + r->membw.bw_gran = MAX_MBA_BW - max_delay; + } else { + if (!rdt_get_mb_table(r)) + return false; + r->membw.arch_needs_linear = false; + } + r->data_width = 3; + + if (boot_cpu_has(X86_FEATURE_PER_THREAD_MBA)) + r->membw.throttle_mode = THREAD_THROTTLE_PER_THREAD; + else + r->membw.throttle_mode = THREAD_THROTTLE_MAX; + thread_throttle_mode_init(); + + r->alloc_capable = true; + r->alloc_enabled = true; + + return true; +} + +static bool __rdt_get_mem_config_amd(struct rdt_resource *r) +{ + union cpuid_0x10_3_eax eax; + union cpuid_0x10_x_edx edx; + u32 ebx, ecx; + + cpuid_count(0x80000020, 1, &eax.full, &ebx, &ecx, &edx.full); + r->num_closid = edx.split.cos_max + 1; + r->default_ctrl = MAX_MBA_BW_AMD; + + /* AMD does not use delay */ + r->membw.delay_linear = false; + r->membw.arch_needs_linear = false; + + /* + * AMD does not use memory delay throttle model to control + * the allocation like Intel does. + */ + r->membw.throttle_mode = THREAD_THROTTLE_UNDEFINED; + r->membw.min_bw = 0; + r->membw.bw_gran = 1; + /* Max value is 2048, Data width should be 4 in decimal */ + r->data_width = 4; + + r->alloc_capable = true; + r->alloc_enabled = true; + + return true; +} + +static void rdt_get_cache_alloc_cfg(int idx, struct rdt_resource *r) +{ + union cpuid_0x10_1_eax eax; + union cpuid_0x10_x_edx edx; + u32 ebx, ecx; + + cpuid_count(0x00000010, idx, &eax.full, &ebx, &ecx, &edx.full); + r->num_closid = edx.split.cos_max + 1; + r->cache.cbm_len = eax.split.cbm_len + 1; + r->default_ctrl = BIT_MASK(eax.split.cbm_len + 1) - 1; + r->cache.shareable_bits = ebx & r->default_ctrl; + r->data_width = (r->cache.cbm_len + 3) / 4; + r->alloc_capable = true; + r->alloc_enabled = true; +} + +static void rdt_get_cdp_config(int level, int type) +{ + struct rdt_resource *r_l = &rdt_resources_all[level]; + struct rdt_resource *r = &rdt_resources_all[type]; + + r->num_closid = r_l->num_closid / 2; + r->cache.cbm_len = r_l->cache.cbm_len; + r->default_ctrl = r_l->default_ctrl; + r->cache.shareable_bits = r_l->cache.shareable_bits; + r->data_width = (r->cache.cbm_len + 3) / 4; + r->alloc_capable = true; + /* + * By default, CDP is disabled. CDP can be enabled by mount parameter + * "cdp" during resctrl file system mount time. + */ + r->alloc_enabled = false; +} + +static void rdt_get_cdp_l3_config(void) +{ + rdt_get_cdp_config(RDT_RESOURCE_L3, RDT_RESOURCE_L3DATA); + rdt_get_cdp_config(RDT_RESOURCE_L3, RDT_RESOURCE_L3CODE); +} + +static void rdt_get_cdp_l2_config(void) +{ + rdt_get_cdp_config(RDT_RESOURCE_L2, RDT_RESOURCE_L2DATA); + rdt_get_cdp_config(RDT_RESOURCE_L2, RDT_RESOURCE_L2CODE); +} + +static void +mba_wrmsr_amd(struct rdt_domain *d, struct msr_param *m, struct rdt_resource *r) +{ + unsigned int i; + + for (i = m->low; i < m->high; i++) + wrmsrl(r->msr_base + i, d->ctrl_val[i]); +} + +/* + * Map the memory b/w percentage value to delay values + * that can be written to QOS_MSRs. + * There are currently no SKUs which support non linear delay values. + */ +u32 delay_bw_map(unsigned long bw, struct rdt_resource *r) +{ + if (r->membw.delay_linear) + return MAX_MBA_BW - bw; + + pr_warn_once("Non Linear delay-bw map not supported but queried\n"); + return r->default_ctrl; +} + +static void +mba_wrmsr_intel(struct rdt_domain *d, struct msr_param *m, + struct rdt_resource *r) +{ + unsigned int i; + + /* Write the delay values for mba. */ + for (i = m->low; i < m->high; i++) + wrmsrl(r->msr_base + i, delay_bw_map(d->ctrl_val[i], r)); +} + +static void +cat_wrmsr(struct rdt_domain *d, struct msr_param *m, struct rdt_resource *r) +{ + unsigned int i; + + for (i = m->low; i < m->high; i++) + wrmsrl(r->msr_base + cbm_idx(r, i), d->ctrl_val[i]); +} + +struct rdt_domain *get_domain_from_cpu(int cpu, struct rdt_resource *r) +{ + struct rdt_domain *d; + + list_for_each_entry(d, &r->domains, list) { + /* Find the domain that contains this CPU */ + if (cpumask_test_cpu(cpu, &d->cpu_mask)) + return d; + } + + return NULL; +} + +void rdt_ctrl_update(void *arg) +{ + struct msr_param *m = arg; + struct rdt_resource *r = m->res; + int cpu = smp_processor_id(); + struct rdt_domain *d; + + d = get_domain_from_cpu(cpu, r); + if (d) { + r->msr_update(d, m, r); + return; + } + pr_warn_once("cpu %d not found in any domain for resource %s\n", + cpu, r->name); +} + +/* + * rdt_find_domain - Find a domain in a resource that matches input resource id + * + * Search resource r's domain list to find the resource id. If the resource + * id is found in a domain, return the domain. Otherwise, if requested by + * caller, return the first domain whose id is bigger than the input id. + * The domain list is sorted by id in ascending order. + */ +struct rdt_domain *rdt_find_domain(struct rdt_resource *r, int id, + struct list_head **pos) +{ + struct rdt_domain *d; + struct list_head *l; + + if (id < 0) + return ERR_PTR(-ENODEV); + + list_for_each(l, &r->domains) { + d = list_entry(l, struct rdt_domain, list); + /* When id is found, return its domain. */ + if (id == d->id) + return d; + /* Stop searching when finding id's position in sorted list. */ + if (id < d->id) + break; + } + + if (pos) + *pos = l; + + return NULL; +} + +void setup_default_ctrlval(struct rdt_resource *r, u32 *dc, u32 *dm) +{ + int i; + + /* + * Initialize the Control MSRs to having no control. + * For Cache Allocation: Set all bits in cbm + * For Memory Allocation: Set b/w requested to 100% + * and the bandwidth in MBps to U32_MAX + */ + for (i = 0; i < r->num_closid; i++, dc++, dm++) { + *dc = r->default_ctrl; + *dm = MBA_MAX_MBPS; + } +} + +static int domain_setup_ctrlval(struct rdt_resource *r, struct rdt_domain *d) +{ + struct msr_param m; + u32 *dc, *dm; + + dc = kmalloc_array(r->num_closid, sizeof(*d->ctrl_val), GFP_KERNEL); + if (!dc) + return -ENOMEM; + + dm = kmalloc_array(r->num_closid, sizeof(*d->mbps_val), GFP_KERNEL); + if (!dm) { + kfree(dc); + return -ENOMEM; + } + + d->ctrl_val = dc; + d->mbps_val = dm; + setup_default_ctrlval(r, dc, dm); + + m.low = 0; + m.high = r->num_closid; + r->msr_update(d, &m, r); + return 0; +} + +static int domain_setup_mon_state(struct rdt_resource *r, struct rdt_domain *d) +{ + size_t tsize; + + if (is_llc_occupancy_enabled()) { + d->rmid_busy_llc = bitmap_zalloc(r->num_rmid, GFP_KERNEL); + if (!d->rmid_busy_llc) + return -ENOMEM; + INIT_DELAYED_WORK(&d->cqm_limbo, cqm_handle_limbo); + } + if (is_mbm_total_enabled()) { + tsize = sizeof(*d->mbm_total); + d->mbm_total = kcalloc(r->num_rmid, tsize, GFP_KERNEL); + if (!d->mbm_total) { + bitmap_free(d->rmid_busy_llc); + return -ENOMEM; + } + } + if (is_mbm_local_enabled()) { + tsize = sizeof(*d->mbm_local); + d->mbm_local = kcalloc(r->num_rmid, tsize, GFP_KERNEL); + if (!d->mbm_local) { + bitmap_free(d->rmid_busy_llc); + kfree(d->mbm_total); + return -ENOMEM; + } + } + + if (is_mbm_enabled()) { + INIT_DELAYED_WORK(&d->mbm_over, mbm_handle_overflow); + mbm_setup_overflow_handler(d, MBM_OVERFLOW_INTERVAL); + } + + return 0; +} + +/* + * domain_add_cpu - Add a cpu to a resource's domain list. + * + * If an existing domain in the resource r's domain list matches the cpu's + * resource id, add the cpu in the domain. + * + * Otherwise, a new domain is allocated and inserted into the right position + * in the domain list sorted by id in ascending order. + * + * The order in the domain list is visible to users when we print entries + * in the schemata file and schemata input is validated to have the same order + * as this list. + */ +static void domain_add_cpu(int cpu, struct rdt_resource *r) +{ + int id = get_cpu_cacheinfo_id(cpu, r->cache_level); + struct list_head *add_pos = NULL; + struct rdt_domain *d; + + d = rdt_find_domain(r, id, &add_pos); + if (IS_ERR(d)) { + pr_warn("Couldn't find cache id for CPU %d\n", cpu); + return; + } + + if (d) { + cpumask_set_cpu(cpu, &d->cpu_mask); + if (r->cache.arch_has_per_cpu_cfg) + rdt_domain_reconfigure_cdp(r); + return; + } + + d = kzalloc_node(sizeof(*d), GFP_KERNEL, cpu_to_node(cpu)); + if (!d) + return; + + d->id = id; + cpumask_set_cpu(cpu, &d->cpu_mask); + + rdt_domain_reconfigure_cdp(r); + + if (r->alloc_capable && domain_setup_ctrlval(r, d)) { + kfree(d); + return; + } + + if (r->mon_capable && domain_setup_mon_state(r, d)) { + kfree(d->ctrl_val); + kfree(d->mbps_val); + kfree(d); + return; + } + + list_add_tail(&d->list, add_pos); + + /* + * If resctrl is mounted, add + * per domain monitor data directories. + */ + if (static_branch_unlikely(&rdt_mon_enable_key)) + mkdir_mondata_subdir_allrdtgrp(r, d); +} + +static void domain_remove_cpu(int cpu, struct rdt_resource *r) +{ + int id = get_cpu_cacheinfo_id(cpu, r->cache_level); + struct rdt_domain *d; + + d = rdt_find_domain(r, id, NULL); + if (IS_ERR_OR_NULL(d)) { + pr_warn("Couldn't find cache id for CPU %d\n", cpu); + return; + } + + cpumask_clear_cpu(cpu, &d->cpu_mask); + if (cpumask_empty(&d->cpu_mask)) { + /* + * If resctrl is mounted, remove all the + * per domain monitor data directories. + */ + if (static_branch_unlikely(&rdt_mon_enable_key)) + rmdir_mondata_subdir_allrdtgrp(r, d->id); + list_del(&d->list); + if (r->mon_capable && is_mbm_enabled()) + cancel_delayed_work(&d->mbm_over); + if (is_llc_occupancy_enabled() && has_busy_rmid(r, d)) { + /* + * When a package is going down, forcefully + * decrement rmid->ebusy. There is no way to know + * that the L3 was flushed and hence may lead to + * incorrect counts in rare scenarios, but leaving + * the RMID as busy creates RMID leaks if the + * package never comes back. + */ + __check_limbo(d, true); + cancel_delayed_work(&d->cqm_limbo); + } + + /* + * rdt_domain "d" is going to be freed below, so clear + * its pointer from pseudo_lock_region struct. + */ + if (d->plr) + d->plr->d = NULL; + + kfree(d->ctrl_val); + kfree(d->mbps_val); + bitmap_free(d->rmid_busy_llc); + kfree(d->mbm_total); + kfree(d->mbm_local); + kfree(d); + return; + } + + if (r == &rdt_resources_all[RDT_RESOURCE_L3]) { + if (is_mbm_enabled() && cpu == d->mbm_work_cpu) { + cancel_delayed_work(&d->mbm_over); + mbm_setup_overflow_handler(d, 0); + } + if (is_llc_occupancy_enabled() && cpu == d->cqm_work_cpu && + has_busy_rmid(r, d)) { + cancel_delayed_work(&d->cqm_limbo); + cqm_setup_limbo_handler(d, 0); + } + } +} + +static void clear_closid_rmid(int cpu) +{ + struct resctrl_pqr_state *state = this_cpu_ptr(&pqr_state); + + state->default_closid = 0; + state->default_rmid = 0; + state->cur_closid = 0; + state->cur_rmid = 0; + wrmsr(IA32_PQR_ASSOC, 0, 0); +} + +static int resctrl_online_cpu(unsigned int cpu) +{ + struct rdt_resource *r; + + mutex_lock(&rdtgroup_mutex); + for_each_capable_rdt_resource(r) + domain_add_cpu(cpu, r); + /* The cpu is set in default rdtgroup after online. */ + cpumask_set_cpu(cpu, &rdtgroup_default.cpu_mask); + clear_closid_rmid(cpu); + mutex_unlock(&rdtgroup_mutex); + + return 0; +} + +static void clear_childcpus(struct rdtgroup *r, unsigned int cpu) +{ + struct rdtgroup *cr; + + list_for_each_entry(cr, &r->mon.crdtgrp_list, mon.crdtgrp_list) { + if (cpumask_test_and_clear_cpu(cpu, &cr->cpu_mask)) { + break; + } + } +} + +static int resctrl_offline_cpu(unsigned int cpu) +{ + struct rdtgroup *rdtgrp; + struct rdt_resource *r; + + mutex_lock(&rdtgroup_mutex); + for_each_capable_rdt_resource(r) + domain_remove_cpu(cpu, r); + list_for_each_entry(rdtgrp, &rdt_all_groups, rdtgroup_list) { + if (cpumask_test_and_clear_cpu(cpu, &rdtgrp->cpu_mask)) { + clear_childcpus(rdtgrp, cpu); + break; + } + } + clear_closid_rmid(cpu); + mutex_unlock(&rdtgroup_mutex); + + return 0; +} + +/* + * Choose a width for the resource name and resource data based on the + * resource that has widest name and cbm. + */ +static __init void rdt_init_padding(void) +{ + struct rdt_resource *r; + int cl; + + for_each_alloc_capable_rdt_resource(r) { + cl = strlen(r->name); + if (cl > max_name_width) + max_name_width = cl; + + if (r->data_width > max_data_width) + max_data_width = r->data_width; + } +} + +enum { + RDT_FLAG_CMT, + RDT_FLAG_MBM_TOTAL, + RDT_FLAG_MBM_LOCAL, + RDT_FLAG_L3_CAT, + RDT_FLAG_L3_CDP, + RDT_FLAG_L2_CAT, + RDT_FLAG_L2_CDP, + RDT_FLAG_MBA, +}; + +#define RDT_OPT(idx, n, f) \ +[idx] = { \ + .name = n, \ + .flag = f \ +} + +struct rdt_options { + char *name; + int flag; + bool force_off, force_on; +}; + +static struct rdt_options rdt_options[] __initdata = { + RDT_OPT(RDT_FLAG_CMT, "cmt", X86_FEATURE_CQM_OCCUP_LLC), + RDT_OPT(RDT_FLAG_MBM_TOTAL, "mbmtotal", X86_FEATURE_CQM_MBM_TOTAL), + RDT_OPT(RDT_FLAG_MBM_LOCAL, "mbmlocal", X86_FEATURE_CQM_MBM_LOCAL), + RDT_OPT(RDT_FLAG_L3_CAT, "l3cat", X86_FEATURE_CAT_L3), + RDT_OPT(RDT_FLAG_L3_CDP, "l3cdp", X86_FEATURE_CDP_L3), + RDT_OPT(RDT_FLAG_L2_CAT, "l2cat", X86_FEATURE_CAT_L2), + RDT_OPT(RDT_FLAG_L2_CDP, "l2cdp", X86_FEATURE_CDP_L2), + RDT_OPT(RDT_FLAG_MBA, "mba", X86_FEATURE_MBA), +}; +#define NUM_RDT_OPTIONS ARRAY_SIZE(rdt_options) + +static int __init set_rdt_options(char *str) +{ + struct rdt_options *o; + bool force_off; + char *tok; + + if (*str == '=') + str++; + while ((tok = strsep(&str, ",")) != NULL) { + force_off = *tok == '!'; + if (force_off) + tok++; + for (o = rdt_options; o < &rdt_options[NUM_RDT_OPTIONS]; o++) { + if (strcmp(tok, o->name) == 0) { + if (force_off) + o->force_off = true; + else + o->force_on = true; + break; + } + } + } + return 1; +} +__setup("rdt", set_rdt_options); + +static bool __init rdt_cpu_has(int flag) +{ + bool ret = boot_cpu_has(flag); + struct rdt_options *o; + + if (!ret) + return ret; + + for (o = rdt_options; o < &rdt_options[NUM_RDT_OPTIONS]; o++) { + if (flag == o->flag) { + if (o->force_off) + ret = false; + if (o->force_on) + ret = true; + break; + } + } + return ret; +} + +static __init bool get_mem_config(void) +{ + if (!rdt_cpu_has(X86_FEATURE_MBA)) + return false; + + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + return __get_mem_config_intel(&rdt_resources_all[RDT_RESOURCE_MBA]); + else if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) + return __rdt_get_mem_config_amd(&rdt_resources_all[RDT_RESOURCE_MBA]); + + return false; +} + +static __init bool get_rdt_alloc_resources(void) +{ + bool ret = false; + + if (rdt_alloc_capable) + return true; + + if (!boot_cpu_has(X86_FEATURE_RDT_A)) + return false; + + if (rdt_cpu_has(X86_FEATURE_CAT_L3)) { + rdt_get_cache_alloc_cfg(1, &rdt_resources_all[RDT_RESOURCE_L3]); + if (rdt_cpu_has(X86_FEATURE_CDP_L3)) + rdt_get_cdp_l3_config(); + ret = true; + } + if (rdt_cpu_has(X86_FEATURE_CAT_L2)) { + /* CPUID 0x10.2 fields are same format at 0x10.1 */ + rdt_get_cache_alloc_cfg(2, &rdt_resources_all[RDT_RESOURCE_L2]); + if (rdt_cpu_has(X86_FEATURE_CDP_L2)) + rdt_get_cdp_l2_config(); + ret = true; + } + + if (get_mem_config()) + ret = true; + + return ret; +} + +static __init bool get_rdt_mon_resources(void) +{ + if (rdt_cpu_has(X86_FEATURE_CQM_OCCUP_LLC)) + rdt_mon_features |= (1 << QOS_L3_OCCUP_EVENT_ID); + if (rdt_cpu_has(X86_FEATURE_CQM_MBM_TOTAL)) + rdt_mon_features |= (1 << QOS_L3_MBM_TOTAL_EVENT_ID); + if (rdt_cpu_has(X86_FEATURE_CQM_MBM_LOCAL)) + rdt_mon_features |= (1 << QOS_L3_MBM_LOCAL_EVENT_ID); + + if (!rdt_mon_features) + return false; + + return !rdt_get_mon_l3_config(&rdt_resources_all[RDT_RESOURCE_L3]); +} + +static __init void __check_quirks_intel(void) +{ + switch (boot_cpu_data.x86_model) { + case INTEL_FAM6_HASWELL_X: + if (!rdt_options[RDT_FLAG_L3_CAT].force_off) + cache_alloc_hsw_probe(); + break; + case INTEL_FAM6_SKYLAKE_X: + if (boot_cpu_data.x86_stepping <= 4) + set_rdt_options("!cmt,!mbmtotal,!mbmlocal,!l3cat"); + else + set_rdt_options("!l3cat"); + } +} + +static __init void check_quirks(void) +{ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + __check_quirks_intel(); +} + +static __init bool get_rdt_resources(void) +{ + rdt_alloc_capable = get_rdt_alloc_resources(); + rdt_mon_capable = get_rdt_mon_resources(); + + return (rdt_mon_capable || rdt_alloc_capable); +} + +static __init void rdt_init_res_defs_intel(void) +{ + struct rdt_resource *r; + + for_each_rdt_resource(r) { + if (r->rid == RDT_RESOURCE_L3 || + r->rid == RDT_RESOURCE_L3DATA || + r->rid == RDT_RESOURCE_L3CODE || + r->rid == RDT_RESOURCE_L2 || + r->rid == RDT_RESOURCE_L2DATA || + r->rid == RDT_RESOURCE_L2CODE) { + r->cache.arch_has_sparse_bitmaps = false; + r->cache.arch_has_empty_bitmaps = false; + r->cache.arch_has_per_cpu_cfg = false; + } else if (r->rid == RDT_RESOURCE_MBA) { + r->msr_base = MSR_IA32_MBA_THRTL_BASE; + r->msr_update = mba_wrmsr_intel; + } + } +} + +static __init void rdt_init_res_defs_amd(void) +{ + struct rdt_resource *r; + + for_each_rdt_resource(r) { + if (r->rid == RDT_RESOURCE_L3 || + r->rid == RDT_RESOURCE_L3DATA || + r->rid == RDT_RESOURCE_L3CODE || + r->rid == RDT_RESOURCE_L2 || + r->rid == RDT_RESOURCE_L2DATA || + r->rid == RDT_RESOURCE_L2CODE) { + r->cache.arch_has_sparse_bitmaps = true; + r->cache.arch_has_empty_bitmaps = true; + r->cache.arch_has_per_cpu_cfg = true; + } else if (r->rid == RDT_RESOURCE_MBA) { + r->msr_base = MSR_IA32_MBA_BW_BASE; + r->msr_update = mba_wrmsr_amd; + } + } +} + +static __init void rdt_init_res_defs(void) +{ + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + rdt_init_res_defs_intel(); + else if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) + rdt_init_res_defs_amd(); +} + +static enum cpuhp_state rdt_online; + +/* Runs once on the BSP during boot. */ +void resctrl_cpu_detect(struct cpuinfo_x86 *c) +{ + if (!cpu_has(c, X86_FEATURE_CQM_LLC)) { + c->x86_cache_max_rmid = -1; + c->x86_cache_occ_scale = -1; + c->x86_cache_mbm_width_offset = -1; + return; + } + + /* will be overridden if occupancy monitoring exists */ + c->x86_cache_max_rmid = cpuid_ebx(0xf); + + if (cpu_has(c, X86_FEATURE_CQM_OCCUP_LLC) || + cpu_has(c, X86_FEATURE_CQM_MBM_TOTAL) || + cpu_has(c, X86_FEATURE_CQM_MBM_LOCAL)) { + u32 eax, ebx, ecx, edx; + + /* QoS sub-leaf, EAX=0Fh, ECX=1 */ + cpuid_count(0xf, 1, &eax, &ebx, &ecx, &edx); + + c->x86_cache_max_rmid = ecx; + c->x86_cache_occ_scale = ebx; + c->x86_cache_mbm_width_offset = eax & 0xff; + + if (c->x86_vendor == X86_VENDOR_AMD && !c->x86_cache_mbm_width_offset) + c->x86_cache_mbm_width_offset = MBM_CNTR_WIDTH_OFFSET_AMD; + } +} + +static int __init resctrl_late_init(void) +{ + struct rdt_resource *r; + int state, ret; + + /* + * Initialize functions(or definitions) that are different + * between vendors here. + */ + rdt_init_res_defs(); + + check_quirks(); + + if (!get_rdt_resources()) + return -ENODEV; + + rdt_init_padding(); + + state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, + "x86/resctrl/cat:online:", + resctrl_online_cpu, resctrl_offline_cpu); + if (state < 0) + return state; + + ret = rdtgroup_init(); + if (ret) { + cpuhp_remove_state(state); + return ret; + } + rdt_online = state; + + for_each_alloc_capable_rdt_resource(r) + pr_info("%s allocation detected\n", r->name); + + for_each_mon_capable_rdt_resource(r) + pr_info("%s monitoring detected\n", r->name); + + return 0; +} + +late_initcall(resctrl_late_init); + +static void __exit resctrl_exit(void) +{ + cpuhp_remove_state(rdt_online); + rdtgroup_exit(); +} + +__exitcall(resctrl_exit); diff --git a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c new file mode 100644 index 000000000..c877642e8 --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c @@ -0,0 +1,490 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Resource Director Technology(RDT) + * - Cache Allocation code. + * + * Copyright (C) 2016 Intel Corporation + * + * Authors: + * Fenghua Yu <fenghua.yu@intel.com> + * Tony Luck <tony.luck@intel.com> + * + * More information about RDT be found in the Intel (R) x86 Architecture + * Software Developer Manual June 2016, volume 3, section 17.17. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cpu.h> +#include <linux/kernfs.h> +#include <linux/seq_file.h> +#include <linux/slab.h> +#include "internal.h" + +/* + * Check whether MBA bandwidth percentage value is correct. The value is + * checked against the minimum and max bandwidth values specified by the + * hardware. The allocated bandwidth percentage is rounded to the next + * control step available on the hardware. + */ +static bool bw_validate(char *buf, unsigned long *data, struct rdt_resource *r) +{ + unsigned long bw; + int ret; + + /* + * Only linear delay values is supported for current Intel SKUs. + */ + if (!r->membw.delay_linear && r->membw.arch_needs_linear) { + rdt_last_cmd_puts("No support for non-linear MB domains\n"); + return false; + } + + ret = kstrtoul(buf, 10, &bw); + if (ret) { + rdt_last_cmd_printf("Non-decimal digit in MB value %s\n", buf); + return false; + } + + if ((bw < r->membw.min_bw || bw > r->default_ctrl) && + !is_mba_sc(r)) { + rdt_last_cmd_printf("MB value %ld out of range [%d,%d]\n", bw, + r->membw.min_bw, r->default_ctrl); + return false; + } + + *data = roundup(bw, (unsigned long)r->membw.bw_gran); + return true; +} + +int parse_bw(struct rdt_parse_data *data, struct rdt_resource *r, + struct rdt_domain *d) +{ + unsigned long bw_val; + + if (d->have_new_ctrl) { + rdt_last_cmd_printf("Duplicate domain %d\n", d->id); + return -EINVAL; + } + + if (!bw_validate(data->buf, &bw_val, r)) + return -EINVAL; + d->new_ctrl = bw_val; + d->have_new_ctrl = true; + + return 0; +} + +/* + * Check whether a cache bit mask is valid. + * For Intel the SDM says: + * Please note that all (and only) contiguous '1' combinations + * are allowed (e.g. FFFFH, 0FF0H, 003CH, etc.). + * Additionally Haswell requires at least two bits set. + * AMD allows non-contiguous bitmasks. + */ +static bool cbm_validate(char *buf, u32 *data, struct rdt_resource *r) +{ + unsigned long first_bit, zero_bit, val; + unsigned int cbm_len = r->cache.cbm_len; + int ret; + + ret = kstrtoul(buf, 16, &val); + if (ret) { + rdt_last_cmd_printf("Non-hex character in the mask %s\n", buf); + return false; + } + + if ((!r->cache.arch_has_empty_bitmaps && val == 0) || + val > r->default_ctrl) { + rdt_last_cmd_puts("Mask out of range\n"); + return false; + } + + first_bit = find_first_bit(&val, cbm_len); + zero_bit = find_next_zero_bit(&val, cbm_len, first_bit); + + /* Are non-contiguous bitmaps allowed? */ + if (!r->cache.arch_has_sparse_bitmaps && + (find_next_bit(&val, cbm_len, zero_bit) < cbm_len)) { + rdt_last_cmd_printf("The mask %lx has non-consecutive 1-bits\n", val); + return false; + } + + if ((zero_bit - first_bit) < r->cache.min_cbm_bits) { + rdt_last_cmd_printf("Need at least %d bits in the mask\n", + r->cache.min_cbm_bits); + return false; + } + + *data = val; + return true; +} + +/* + * Read one cache bit mask (hex). Check that it is valid for the current + * resource type. + */ +int parse_cbm(struct rdt_parse_data *data, struct rdt_resource *r, + struct rdt_domain *d) +{ + struct rdtgroup *rdtgrp = data->rdtgrp; + u32 cbm_val; + + if (d->have_new_ctrl) { + rdt_last_cmd_printf("Duplicate domain %d\n", d->id); + return -EINVAL; + } + + /* + * Cannot set up more than one pseudo-locked region in a cache + * hierarchy. + */ + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP && + rdtgroup_pseudo_locked_in_hierarchy(d)) { + rdt_last_cmd_puts("Pseudo-locked region in hierarchy\n"); + return -EINVAL; + } + + if (!cbm_validate(data->buf, &cbm_val, r)) + return -EINVAL; + + if ((rdtgrp->mode == RDT_MODE_EXCLUSIVE || + rdtgrp->mode == RDT_MODE_SHAREABLE) && + rdtgroup_cbm_overlaps_pseudo_locked(d, cbm_val)) { + rdt_last_cmd_puts("CBM overlaps with pseudo-locked region\n"); + return -EINVAL; + } + + /* + * The CBM may not overlap with the CBM of another closid if + * either is exclusive. + */ + if (rdtgroup_cbm_overlaps(r, d, cbm_val, rdtgrp->closid, true)) { + rdt_last_cmd_puts("Overlaps with exclusive group\n"); + return -EINVAL; + } + + if (rdtgroup_cbm_overlaps(r, d, cbm_val, rdtgrp->closid, false)) { + if (rdtgrp->mode == RDT_MODE_EXCLUSIVE || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + rdt_last_cmd_puts("Overlaps with other group\n"); + return -EINVAL; + } + } + + d->new_ctrl = cbm_val; + d->have_new_ctrl = true; + + return 0; +} + +/* + * For each domain in this resource we expect to find a series of: + * id=mask + * separated by ";". The "id" is in decimal, and must match one of + * the "id"s for this resource. + */ +static int parse_line(char *line, struct rdt_resource *r, + struct rdtgroup *rdtgrp) +{ + struct rdt_parse_data data; + char *dom = NULL, *id; + struct rdt_domain *d; + unsigned long dom_id; + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP && + r->rid == RDT_RESOURCE_MBA) { + rdt_last_cmd_puts("Cannot pseudo-lock MBA resource\n"); + return -EINVAL; + } + +next: + if (!line || line[0] == '\0') + return 0; + dom = strsep(&line, ";"); + id = strsep(&dom, "="); + if (!dom || kstrtoul(id, 10, &dom_id)) { + rdt_last_cmd_puts("Missing '=' or non-numeric domain\n"); + return -EINVAL; + } + dom = strim(dom); + list_for_each_entry(d, &r->domains, list) { + if (d->id == dom_id) { + data.buf = dom; + data.rdtgrp = rdtgrp; + if (r->parse_ctrlval(&data, r, d)) + return -EINVAL; + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + /* + * In pseudo-locking setup mode and just + * parsed a valid CBM that should be + * pseudo-locked. Only one locked region per + * resource group and domain so just do + * the required initialization for single + * region and return. + */ + rdtgrp->plr->r = r; + rdtgrp->plr->d = d; + rdtgrp->plr->cbm = d->new_ctrl; + d->plr = rdtgrp->plr; + return 0; + } + goto next; + } + } + return -EINVAL; +} + +int update_domains(struct rdt_resource *r, int closid) +{ + struct msr_param msr_param; + cpumask_var_t cpu_mask; + struct rdt_domain *d; + bool mba_sc; + u32 *dc; + int cpu; + + if (!zalloc_cpumask_var(&cpu_mask, GFP_KERNEL)) + return -ENOMEM; + + msr_param.low = closid; + msr_param.high = msr_param.low + 1; + msr_param.res = r; + + mba_sc = is_mba_sc(r); + list_for_each_entry(d, &r->domains, list) { + dc = !mba_sc ? d->ctrl_val : d->mbps_val; + if (d->have_new_ctrl && d->new_ctrl != dc[closid]) { + cpumask_set_cpu(cpumask_any(&d->cpu_mask), cpu_mask); + dc[closid] = d->new_ctrl; + } + } + + /* + * Avoid writing the control msr with control values when + * MBA software controller is enabled + */ + if (cpumask_empty(cpu_mask) || mba_sc) + goto done; + cpu = get_cpu(); + /* Update resource control msr on this CPU if it's in cpu_mask. */ + if (cpumask_test_cpu(cpu, cpu_mask)) + rdt_ctrl_update(&msr_param); + /* Update resource control msr on other CPUs. */ + smp_call_function_many(cpu_mask, rdt_ctrl_update, &msr_param, 1); + put_cpu(); + +done: + free_cpumask_var(cpu_mask); + + return 0; +} + +static int rdtgroup_parse_resource(char *resname, char *tok, + struct rdtgroup *rdtgrp) +{ + struct rdt_resource *r; + + for_each_alloc_enabled_rdt_resource(r) { + if (!strcmp(resname, r->name) && rdtgrp->closid < r->num_closid) + return parse_line(tok, r, rdtgrp); + } + rdt_last_cmd_printf("Unknown or unsupported resource name '%s'\n", resname); + return -EINVAL; +} + +ssize_t rdtgroup_schemata_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct rdtgroup *rdtgrp; + struct rdt_domain *dom; + struct rdt_resource *r; + char *tok, *resname; + int ret = 0; + + /* Valid input requires a trailing newline */ + if (nbytes == 0 || buf[nbytes - 1] != '\n') + return -EINVAL; + buf[nbytes - 1] = '\0'; + + cpus_read_lock(); + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + rdtgroup_kn_unlock(of->kn); + cpus_read_unlock(); + return -ENOENT; + } + rdt_last_cmd_clear(); + + /* + * No changes to pseudo-locked region allowed. It has to be removed + * and re-created instead. + */ + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) { + ret = -EINVAL; + rdt_last_cmd_puts("Resource group is pseudo-locked\n"); + goto out; + } + + for_each_alloc_enabled_rdt_resource(r) { + list_for_each_entry(dom, &r->domains, list) + dom->have_new_ctrl = false; + } + + while ((tok = strsep(&buf, "\n")) != NULL) { + resname = strim(strsep(&tok, ":")); + if (!tok) { + rdt_last_cmd_puts("Missing ':'\n"); + ret = -EINVAL; + goto out; + } + if (tok[0] == '\0') { + rdt_last_cmd_printf("Missing '%s' value\n", resname); + ret = -EINVAL; + goto out; + } + ret = rdtgroup_parse_resource(resname, tok, rdtgrp); + if (ret) + goto out; + } + + for_each_alloc_enabled_rdt_resource(r) { + ret = update_domains(r, rdtgrp->closid); + if (ret) + goto out; + } + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + /* + * If pseudo-locking fails we keep the resource group in + * mode RDT_MODE_PSEUDO_LOCKSETUP with its class of service + * active and updated for just the domain the pseudo-locked + * region was requested for. + */ + ret = rdtgroup_pseudo_lock_create(rdtgrp); + } + +out: + rdtgroup_kn_unlock(of->kn); + cpus_read_unlock(); + return ret ?: nbytes; +} + +static void show_doms(struct seq_file *s, struct rdt_resource *r, int closid) +{ + struct rdt_domain *dom; + bool sep = false; + u32 ctrl_val; + + seq_printf(s, "%*s:", max_name_width, r->name); + list_for_each_entry(dom, &r->domains, list) { + if (sep) + seq_puts(s, ";"); + + ctrl_val = (!is_mba_sc(r) ? dom->ctrl_val[closid] : + dom->mbps_val[closid]); + seq_printf(s, r->format_str, dom->id, max_data_width, + ctrl_val); + sep = true; + } + seq_puts(s, "\n"); +} + +int rdtgroup_schemata_show(struct kernfs_open_file *of, + struct seq_file *s, void *v) +{ + struct rdtgroup *rdtgrp; + struct rdt_resource *r; + int ret = 0; + u32 closid; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (rdtgrp) { + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + for_each_alloc_enabled_rdt_resource(r) + seq_printf(s, "%s:uninitialized\n", r->name); + } else if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) { + if (!rdtgrp->plr->d) { + rdt_last_cmd_clear(); + rdt_last_cmd_puts("Cache domain offline\n"); + ret = -ENODEV; + } else { + seq_printf(s, "%s:%d=%x\n", + rdtgrp->plr->r->name, + rdtgrp->plr->d->id, + rdtgrp->plr->cbm); + } + } else { + closid = rdtgrp->closid; + for_each_alloc_enabled_rdt_resource(r) { + if (closid < r->num_closid) + show_doms(s, r, closid); + } + } + } else { + ret = -ENOENT; + } + rdtgroup_kn_unlock(of->kn); + return ret; +} + +void mon_event_read(struct rmid_read *rr, struct rdt_resource *r, + struct rdt_domain *d, struct rdtgroup *rdtgrp, + int evtid, int first) +{ + /* + * setup the parameters to send to the IPI to read the data. + */ + rr->rgrp = rdtgrp; + rr->evtid = evtid; + rr->r = r; + rr->d = d; + rr->val = 0; + rr->first = first; + + smp_call_function_any(&d->cpu_mask, mon_event_count, rr, 1); +} + +int rdtgroup_mondata_show(struct seq_file *m, void *arg) +{ + struct kernfs_open_file *of = m->private; + u32 resid, evtid, domid; + struct rdtgroup *rdtgrp; + struct rdt_resource *r; + union mon_data_bits md; + struct rdt_domain *d; + struct rmid_read rr; + int ret = 0; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + ret = -ENOENT; + goto out; + } + + md.priv = of->kn->priv; + resid = md.u.rid; + domid = md.u.domid; + evtid = md.u.evtid; + + r = &rdt_resources_all[resid]; + d = rdt_find_domain(r, domid, NULL); + if (IS_ERR_OR_NULL(d)) { + ret = -ENOENT; + goto out; + } + + mon_event_read(&rr, r, d, rdtgrp, evtid, false); + + if (rr.val & RMID_VAL_ERROR) + seq_puts(m, "Error\n"); + else if (rr.val & RMID_VAL_UNAVAIL) + seq_puts(m, "Unavailable\n"); + else + seq_printf(m, "%llu\n", rr.val * r->mon_scale); + +out: + rdtgroup_kn_unlock(of->kn); + return ret; +} diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h new file mode 100644 index 000000000..f65d3c0db --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/internal.h @@ -0,0 +1,632 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_X86_RESCTRL_INTERNAL_H +#define _ASM_X86_RESCTRL_INTERNAL_H + +#include <linux/sched.h> +#include <linux/kernfs.h> +#include <linux/fs_context.h> +#include <linux/jump_label.h> + +#define MSR_IA32_L3_QOS_CFG 0xc81 +#define MSR_IA32_L2_QOS_CFG 0xc82 +#define MSR_IA32_L3_CBM_BASE 0xc90 +#define MSR_IA32_L2_CBM_BASE 0xd10 +#define MSR_IA32_MBA_THRTL_BASE 0xd50 +#define MSR_IA32_MBA_BW_BASE 0xc0000200 + +#define MSR_IA32_QM_CTR 0x0c8e +#define MSR_IA32_QM_EVTSEL 0x0c8d + +#define L3_QOS_CDP_ENABLE 0x01ULL + +#define L2_QOS_CDP_ENABLE 0x01ULL + +/* + * Event IDs are used to program IA32_QM_EVTSEL before reading event + * counter from IA32_QM_CTR + */ +#define QOS_L3_OCCUP_EVENT_ID 0x01 +#define QOS_L3_MBM_TOTAL_EVENT_ID 0x02 +#define QOS_L3_MBM_LOCAL_EVENT_ID 0x03 + +#define CQM_LIMBOCHECK_INTERVAL 1000 + +#define MBM_CNTR_WIDTH_BASE 24 +#define MBM_OVERFLOW_INTERVAL 1000 +#define MAX_MBA_BW 100u +#define MBA_IS_LINEAR 0x4 +#define MBA_MAX_MBPS U32_MAX +#define MAX_MBA_BW_AMD 0x800 +#define MBM_CNTR_WIDTH_OFFSET_AMD 20 + +#define RMID_VAL_ERROR BIT_ULL(63) +#define RMID_VAL_UNAVAIL BIT_ULL(62) +/* + * With the above fields in use 62 bits remain in MSR_IA32_QM_CTR for + * data to be returned. The counter width is discovered from the hardware + * as an offset from MBM_CNTR_WIDTH_BASE. + */ +#define MBM_CNTR_WIDTH_OFFSET_MAX (62 - MBM_CNTR_WIDTH_BASE) + + +struct rdt_fs_context { + struct kernfs_fs_context kfc; + bool enable_cdpl2; + bool enable_cdpl3; + bool enable_mba_mbps; +}; + +static inline struct rdt_fs_context *rdt_fc2context(struct fs_context *fc) +{ + struct kernfs_fs_context *kfc = fc->fs_private; + + return container_of(kfc, struct rdt_fs_context, kfc); +} + +DECLARE_STATIC_KEY_FALSE(rdt_enable_key); +DECLARE_STATIC_KEY_FALSE(rdt_mon_enable_key); + +/** + * struct mon_evt - Entry in the event list of a resource + * @evtid: event id + * @name: name of the event + */ +struct mon_evt { + u32 evtid; + char *name; + struct list_head list; +}; + +/** + * struct mon_data_bits - Monitoring details for each event file + * @rid: Resource id associated with the event file. + * @evtid: Event id associated with the event file + * @domid: The domain to which the event file belongs + */ +union mon_data_bits { + void *priv; + struct { + unsigned int rid : 10; + unsigned int evtid : 8; + unsigned int domid : 14; + } u; +}; + +struct rmid_read { + struct rdtgroup *rgrp; + struct rdt_resource *r; + struct rdt_domain *d; + int evtid; + bool first; + u64 val; +}; + +extern unsigned int resctrl_cqm_threshold; +extern bool rdt_alloc_capable; +extern bool rdt_mon_capable; +extern unsigned int rdt_mon_features; + +enum rdt_group_type { + RDTCTRL_GROUP = 0, + RDTMON_GROUP, + RDT_NUM_GROUP, +}; + +/** + * enum rdtgrp_mode - Mode of a RDT resource group + * @RDT_MODE_SHAREABLE: This resource group allows sharing of its allocations + * @RDT_MODE_EXCLUSIVE: No sharing of this resource group's allocations allowed + * @RDT_MODE_PSEUDO_LOCKSETUP: Resource group will be used for Pseudo-Locking + * @RDT_MODE_PSEUDO_LOCKED: No sharing of this resource group's allocations + * allowed AND the allocations are Cache Pseudo-Locked + * + * The mode of a resource group enables control over the allowed overlap + * between allocations associated with different resource groups (classes + * of service). User is able to modify the mode of a resource group by + * writing to the "mode" resctrl file associated with the resource group. + * + * The "shareable", "exclusive", and "pseudo-locksetup" modes are set by + * writing the appropriate text to the "mode" file. A resource group enters + * "pseudo-locked" mode after the schemata is written while the resource + * group is in "pseudo-locksetup" mode. + */ +enum rdtgrp_mode { + RDT_MODE_SHAREABLE = 0, + RDT_MODE_EXCLUSIVE, + RDT_MODE_PSEUDO_LOCKSETUP, + RDT_MODE_PSEUDO_LOCKED, + + /* Must be last */ + RDT_NUM_MODES, +}; + +/** + * struct mongroup - store mon group's data in resctrl fs. + * @mon_data_kn kernlfs node for the mon_data directory + * @parent: parent rdtgrp + * @crdtgrp_list: child rdtgroup node list + * @rmid: rmid for this rdtgroup + */ +struct mongroup { + struct kernfs_node *mon_data_kn; + struct rdtgroup *parent; + struct list_head crdtgrp_list; + u32 rmid; +}; + +/** + * struct pseudo_lock_region - pseudo-lock region information + * @r: RDT resource to which this pseudo-locked region + * belongs + * @d: RDT domain to which this pseudo-locked region + * belongs + * @cbm: bitmask of the pseudo-locked region + * @lock_thread_wq: waitqueue used to wait on the pseudo-locking thread + * completion + * @thread_done: variable used by waitqueue to test if pseudo-locking + * thread completed + * @cpu: core associated with the cache on which the setup code + * will be run + * @line_size: size of the cache lines + * @size: size of pseudo-locked region in bytes + * @kmem: the kernel memory associated with pseudo-locked region + * @minor: minor number of character device associated with this + * region + * @debugfs_dir: pointer to this region's directory in the debugfs + * filesystem + * @pm_reqs: Power management QoS requests related to this region + */ +struct pseudo_lock_region { + struct rdt_resource *r; + struct rdt_domain *d; + u32 cbm; + wait_queue_head_t lock_thread_wq; + int thread_done; + int cpu; + unsigned int line_size; + unsigned int size; + void *kmem; + unsigned int minor; + struct dentry *debugfs_dir; + struct list_head pm_reqs; +}; + +/** + * struct rdtgroup - store rdtgroup's data in resctrl file system. + * @kn: kernfs node + * @rdtgroup_list: linked list for all rdtgroups + * @closid: closid for this rdtgroup + * @cpu_mask: CPUs assigned to this rdtgroup + * @flags: status bits + * @waitcount: how many cpus expect to find this + * group when they acquire rdtgroup_mutex + * @type: indicates type of this rdtgroup - either + * monitor only or ctrl_mon group + * @mon: mongroup related data + * @mode: mode of resource group + * @plr: pseudo-locked region + */ +struct rdtgroup { + struct kernfs_node *kn; + struct list_head rdtgroup_list; + u32 closid; + struct cpumask cpu_mask; + int flags; + atomic_t waitcount; + enum rdt_group_type type; + struct mongroup mon; + enum rdtgrp_mode mode; + struct pseudo_lock_region *plr; +}; + +/* rdtgroup.flags */ +#define RDT_DELETED 1 + +/* rftype.flags */ +#define RFTYPE_FLAGS_CPUS_LIST 1 + +/* + * Define the file type flags for base and info directories. + */ +#define RFTYPE_INFO BIT(0) +#define RFTYPE_BASE BIT(1) +#define RF_CTRLSHIFT 4 +#define RF_MONSHIFT 5 +#define RF_TOPSHIFT 6 +#define RFTYPE_CTRL BIT(RF_CTRLSHIFT) +#define RFTYPE_MON BIT(RF_MONSHIFT) +#define RFTYPE_TOP BIT(RF_TOPSHIFT) +#define RFTYPE_RES_CACHE BIT(8) +#define RFTYPE_RES_MB BIT(9) +#define RF_CTRL_INFO (RFTYPE_INFO | RFTYPE_CTRL) +#define RF_MON_INFO (RFTYPE_INFO | RFTYPE_MON) +#define RF_TOP_INFO (RFTYPE_INFO | RFTYPE_TOP) +#define RF_CTRL_BASE (RFTYPE_BASE | RFTYPE_CTRL) + +/* List of all resource groups */ +extern struct list_head rdt_all_groups; + +extern int max_name_width, max_data_width; + +int __init rdtgroup_init(void); +void __exit rdtgroup_exit(void); + +/** + * struct rftype - describe each file in the resctrl file system + * @name: File name + * @mode: Access mode + * @kf_ops: File operations + * @flags: File specific RFTYPE_FLAGS_* flags + * @fflags: File specific RF_* or RFTYPE_* flags + * @seq_show: Show content of the file + * @write: Write to the file + */ +struct rftype { + char *name; + umode_t mode; + struct kernfs_ops *kf_ops; + unsigned long flags; + unsigned long fflags; + + int (*seq_show)(struct kernfs_open_file *of, + struct seq_file *sf, void *v); + /* + * write() is the generic write callback which maps directly to + * kernfs write operation and overrides all other operations. + * Maximum write size is determined by ->max_write_len. + */ + ssize_t (*write)(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off); +}; + +/** + * struct mbm_state - status for each MBM counter in each domain + * @chunks: Total data moved (multiply by rdt_group.mon_scale to get bytes) + * @prev_msr Value of IA32_QM_CTR for this RMID last time we read it + * @prev_bw_msr:Value of previous IA32_QM_CTR for bandwidth counting + * @prev_bw The most recent bandwidth in MBps + * @delta_bw Difference between the current and previous bandwidth + * @delta_comp Indicates whether to compute the delta_bw + */ +struct mbm_state { + u64 chunks; + u64 prev_msr; + u64 prev_bw_msr; + u32 prev_bw; + u32 delta_bw; + bool delta_comp; +}; + +/** + * struct rdt_domain - group of cpus sharing an RDT resource + * @list: all instances of this resource + * @id: unique id for this instance + * @cpu_mask: which cpus share this resource + * @rmid_busy_llc: + * bitmap of which limbo RMIDs are above threshold + * @mbm_total: saved state for MBM total bandwidth + * @mbm_local: saved state for MBM local bandwidth + * @mbm_over: worker to periodically read MBM h/w counters + * @cqm_limbo: worker to periodically read CQM h/w counters + * @mbm_work_cpu: + * worker cpu for MBM h/w counters + * @cqm_work_cpu: + * worker cpu for CQM h/w counters + * @ctrl_val: array of cache or mem ctrl values (indexed by CLOSID) + * @mbps_val: When mba_sc is enabled, this holds the bandwidth in MBps + * @new_ctrl: new ctrl value to be loaded + * @have_new_ctrl: did user provide new_ctrl for this domain + * @plr: pseudo-locked region (if any) associated with domain + */ +struct rdt_domain { + struct list_head list; + int id; + struct cpumask cpu_mask; + unsigned long *rmid_busy_llc; + struct mbm_state *mbm_total; + struct mbm_state *mbm_local; + struct delayed_work mbm_over; + struct delayed_work cqm_limbo; + int mbm_work_cpu; + int cqm_work_cpu; + u32 *ctrl_val; + u32 *mbps_val; + u32 new_ctrl; + bool have_new_ctrl; + struct pseudo_lock_region *plr; +}; + +/** + * struct msr_param - set a range of MSRs from a domain + * @res: The resource to use + * @low: Beginning index from base MSR + * @high: End index + */ +struct msr_param { + struct rdt_resource *res; + int low; + int high; +}; + +/** + * struct rdt_cache - Cache allocation related data + * @cbm_len: Length of the cache bit mask + * @min_cbm_bits: Minimum number of consecutive bits to be set + * @cbm_idx_mult: Multiplier of CBM index + * @cbm_idx_offset: Offset of CBM index. CBM index is computed by: + * closid * cbm_idx_multi + cbm_idx_offset + * in a cache bit mask + * @shareable_bits: Bitmask of shareable resource with other + * executing entities + * @arch_has_sparse_bitmaps: True if a bitmap like f00f is valid. + * @arch_has_empty_bitmaps: True if the '0' bitmap is valid. + * @arch_has_per_cpu_cfg: True if QOS_CFG register for this cache + * level has CPU scope. + */ +struct rdt_cache { + unsigned int cbm_len; + unsigned int min_cbm_bits; + unsigned int cbm_idx_mult; + unsigned int cbm_idx_offset; + unsigned int shareable_bits; + bool arch_has_sparse_bitmaps; + bool arch_has_empty_bitmaps; + bool arch_has_per_cpu_cfg; +}; + +/** + * enum membw_throttle_mode - System's memory bandwidth throttling mode + * @THREAD_THROTTLE_UNDEFINED: Not relevant to the system + * @THREAD_THROTTLE_MAX: Memory bandwidth is throttled at the core + * always using smallest bandwidth percentage + * assigned to threads, aka "max throttling" + * @THREAD_THROTTLE_PER_THREAD: Memory bandwidth is throttled at the thread + */ +enum membw_throttle_mode { + THREAD_THROTTLE_UNDEFINED = 0, + THREAD_THROTTLE_MAX, + THREAD_THROTTLE_PER_THREAD, +}; + +/** + * struct rdt_membw - Memory bandwidth allocation related data + * @min_bw: Minimum memory bandwidth percentage user can request + * @bw_gran: Granularity at which the memory bandwidth is allocated + * @delay_linear: True if memory B/W delay is in linear scale + * @arch_needs_linear: True if we can't configure non-linear resources + * @throttle_mode: Bandwidth throttling mode when threads request + * different memory bandwidths + * @mba_sc: True if MBA software controller(mba_sc) is enabled + * @mb_map: Mapping of memory B/W percentage to memory B/W delay + */ +struct rdt_membw { + u32 min_bw; + u32 bw_gran; + u32 delay_linear; + bool arch_needs_linear; + enum membw_throttle_mode throttle_mode; + bool mba_sc; + u32 *mb_map; +}; + +static inline bool is_llc_occupancy_enabled(void) +{ + return (rdt_mon_features & (1 << QOS_L3_OCCUP_EVENT_ID)); +} + +static inline bool is_mbm_total_enabled(void) +{ + return (rdt_mon_features & (1 << QOS_L3_MBM_TOTAL_EVENT_ID)); +} + +static inline bool is_mbm_local_enabled(void) +{ + return (rdt_mon_features & (1 << QOS_L3_MBM_LOCAL_EVENT_ID)); +} + +static inline bool is_mbm_enabled(void) +{ + return (is_mbm_total_enabled() || is_mbm_local_enabled()); +} + +static inline bool is_mbm_event(int e) +{ + return (e >= QOS_L3_MBM_TOTAL_EVENT_ID && + e <= QOS_L3_MBM_LOCAL_EVENT_ID); +} + +struct rdt_parse_data { + struct rdtgroup *rdtgrp; + char *buf; +}; + +/** + * struct rdt_resource - attributes of an RDT resource + * @rid: The index of the resource + * @alloc_enabled: Is allocation enabled on this machine + * @mon_enabled: Is monitoring enabled for this feature + * @alloc_capable: Is allocation available on this machine + * @mon_capable: Is monitor feature available on this machine + * @name: Name to use in "schemata" file + * @num_closid: Number of CLOSIDs available + * @cache_level: Which cache level defines scope of this resource + * @default_ctrl: Specifies default cache cbm or memory B/W percent. + * @msr_base: Base MSR address for CBMs + * @msr_update: Function pointer to update QOS MSRs + * @data_width: Character width of data when displaying + * @domains: All domains for this resource + * @cache: Cache allocation related data + * @format_str: Per resource format string to show domain value + * @parse_ctrlval: Per resource function pointer to parse control values + * @evt_list: List of monitoring events + * @num_rmid: Number of RMIDs available + * @mon_scale: cqm counter * mon_scale = occupancy in bytes + * @fflags: flags to choose base and info files + */ +struct rdt_resource { + int rid; + bool alloc_enabled; + bool mon_enabled; + bool alloc_capable; + bool mon_capable; + char *name; + int num_closid; + int cache_level; + u32 default_ctrl; + unsigned int msr_base; + void (*msr_update) (struct rdt_domain *d, struct msr_param *m, + struct rdt_resource *r); + int data_width; + struct list_head domains; + struct rdt_cache cache; + struct rdt_membw membw; + const char *format_str; + int (*parse_ctrlval)(struct rdt_parse_data *data, + struct rdt_resource *r, + struct rdt_domain *d); + struct list_head evt_list; + int num_rmid; + unsigned int mon_scale; + unsigned int mbm_width; + unsigned long fflags; +}; + +int parse_cbm(struct rdt_parse_data *data, struct rdt_resource *r, + struct rdt_domain *d); +int parse_bw(struct rdt_parse_data *data, struct rdt_resource *r, + struct rdt_domain *d); + +extern struct mutex rdtgroup_mutex; + +extern struct rdt_resource rdt_resources_all[]; +extern struct rdtgroup rdtgroup_default; +DECLARE_STATIC_KEY_FALSE(rdt_alloc_enable_key); + +extern struct dentry *debugfs_resctrl; + +enum { + RDT_RESOURCE_L3, + RDT_RESOURCE_L3DATA, + RDT_RESOURCE_L3CODE, + RDT_RESOURCE_L2, + RDT_RESOURCE_L2DATA, + RDT_RESOURCE_L2CODE, + RDT_RESOURCE_MBA, + + /* Must be the last */ + RDT_NUM_RESOURCES, +}; + +#define for_each_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) + +#define for_each_capable_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) \ + if (r->alloc_capable || r->mon_capable) + +#define for_each_alloc_capable_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) \ + if (r->alloc_capable) + +#define for_each_mon_capable_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) \ + if (r->mon_capable) + +#define for_each_alloc_enabled_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) \ + if (r->alloc_enabled) + +#define for_each_mon_enabled_rdt_resource(r) \ + for (r = rdt_resources_all; r < rdt_resources_all + RDT_NUM_RESOURCES;\ + r++) \ + if (r->mon_enabled) + +/* CPUID.(EAX=10H, ECX=ResID=1).EAX */ +union cpuid_0x10_1_eax { + struct { + unsigned int cbm_len:5; + } split; + unsigned int full; +}; + +/* CPUID.(EAX=10H, ECX=ResID=3).EAX */ +union cpuid_0x10_3_eax { + struct { + unsigned int max_delay:12; + } split; + unsigned int full; +}; + +/* CPUID.(EAX=10H, ECX=ResID).EDX */ +union cpuid_0x10_x_edx { + struct { + unsigned int cos_max:16; + } split; + unsigned int full; +}; + +void rdt_last_cmd_clear(void); +void rdt_last_cmd_puts(const char *s); +void rdt_last_cmd_printf(const char *fmt, ...); + +void rdt_ctrl_update(void *arg); +struct rdtgroup *rdtgroup_kn_lock_live(struct kernfs_node *kn); +void rdtgroup_kn_unlock(struct kernfs_node *kn); +int rdtgroup_kn_mode_restrict(struct rdtgroup *r, const char *name); +int rdtgroup_kn_mode_restore(struct rdtgroup *r, const char *name, + umode_t mask); +struct rdt_domain *rdt_find_domain(struct rdt_resource *r, int id, + struct list_head **pos); +ssize_t rdtgroup_schemata_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off); +int rdtgroup_schemata_show(struct kernfs_open_file *of, + struct seq_file *s, void *v); +bool rdtgroup_cbm_overlaps(struct rdt_resource *r, struct rdt_domain *d, + unsigned long cbm, int closid, bool exclusive); +unsigned int rdtgroup_cbm_to_size(struct rdt_resource *r, struct rdt_domain *d, + unsigned long cbm); +enum rdtgrp_mode rdtgroup_mode_by_closid(int closid); +int rdtgroup_tasks_assigned(struct rdtgroup *r); +int rdtgroup_locksetup_enter(struct rdtgroup *rdtgrp); +int rdtgroup_locksetup_exit(struct rdtgroup *rdtgrp); +bool rdtgroup_cbm_overlaps_pseudo_locked(struct rdt_domain *d, unsigned long cbm); +bool rdtgroup_pseudo_locked_in_hierarchy(struct rdt_domain *d); +int rdt_pseudo_lock_init(void); +void rdt_pseudo_lock_release(void); +int rdtgroup_pseudo_lock_create(struct rdtgroup *rdtgrp); +void rdtgroup_pseudo_lock_remove(struct rdtgroup *rdtgrp); +struct rdt_domain *get_domain_from_cpu(int cpu, struct rdt_resource *r); +int update_domains(struct rdt_resource *r, int closid); +int closids_supported(void); +void closid_free(int closid); +int alloc_rmid(void); +void free_rmid(u32 rmid); +int rdt_get_mon_l3_config(struct rdt_resource *r); +void mon_event_count(void *info); +int rdtgroup_mondata_show(struct seq_file *m, void *arg); +void rmdir_mondata_subdir_allrdtgrp(struct rdt_resource *r, + unsigned int dom_id); +void mkdir_mondata_subdir_allrdtgrp(struct rdt_resource *r, + struct rdt_domain *d); +void mon_event_read(struct rmid_read *rr, struct rdt_resource *r, + struct rdt_domain *d, struct rdtgroup *rdtgrp, + int evtid, int first); +void mbm_setup_overflow_handler(struct rdt_domain *dom, + unsigned long delay_ms); +void mbm_handle_overflow(struct work_struct *work); +bool is_mba_sc(struct rdt_resource *r); +void setup_default_ctrlval(struct rdt_resource *r, u32 *dc, u32 *dm); +u32 delay_bw_map(unsigned long bw, struct rdt_resource *r); +void cqm_setup_limbo_handler(struct rdt_domain *dom, unsigned long delay_ms); +void cqm_handle_limbo(struct work_struct *work); +bool has_busy_rmid(struct rdt_resource *r, struct rdt_domain *d); +void __check_limbo(struct rdt_domain *d, bool force_free); +void rdt_domain_reconfigure_cdp(struct rdt_resource *r); +void __init thread_throttle_mode_init(void); + +#endif /* _ASM_X86_RESCTRL_INTERNAL_H */ diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c new file mode 100644 index 000000000..576f16a50 --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/monitor.c @@ -0,0 +1,649 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Resource Director Technology(RDT) + * - Monitoring code + * + * Copyright (C) 2017 Intel Corporation + * + * Author: + * Vikas Shivappa <vikas.shivappa@intel.com> + * + * This replaces the cqm.c based on perf but we reuse a lot of + * code and datastructures originally from Peter Zijlstra and Matt Fleming. + * + * More information about RDT be found in the Intel (R) x86 Architecture + * Software Developer Manual June 2016, volume 3, section 17.17. + */ + +#include <linux/module.h> +#include <linux/slab.h> +#include <asm/cpu_device_id.h> +#include "internal.h" + +struct rmid_entry { + u32 rmid; + int busy; + struct list_head list; +}; + +/** + * @rmid_free_lru A least recently used list of free RMIDs + * These RMIDs are guaranteed to have an occupancy less than the + * threshold occupancy + */ +static LIST_HEAD(rmid_free_lru); + +/** + * @rmid_limbo_count count of currently unused but (potentially) + * dirty RMIDs. + * This counts RMIDs that no one is currently using but that + * may have a occupancy value > intel_cqm_threshold. User can change + * the threshold occupancy value. + */ +static unsigned int rmid_limbo_count; + +/** + * @rmid_entry - The entry in the limbo and free lists. + */ +static struct rmid_entry *rmid_ptrs; + +/* + * Global boolean for rdt_monitor which is true if any + * resource monitoring is enabled. + */ +bool rdt_mon_capable; + +/* + * Global to indicate which monitoring events are enabled. + */ +unsigned int rdt_mon_features; + +/* + * This is the threshold cache occupancy at which we will consider an + * RMID available for re-allocation. + */ +unsigned int resctrl_cqm_threshold; + +static inline struct rmid_entry *__rmid_entry(u32 rmid) +{ + struct rmid_entry *entry; + + entry = &rmid_ptrs[rmid]; + WARN_ON(entry->rmid != rmid); + + return entry; +} + +static u64 __rmid_read(u32 rmid, u32 eventid) +{ + u64 val; + + /* + * As per the SDM, when IA32_QM_EVTSEL.EvtID (bits 7:0) is configured + * with a valid event code for supported resource type and the bits + * IA32_QM_EVTSEL.RMID (bits 41:32) are configured with valid RMID, + * IA32_QM_CTR.data (bits 61:0) reports the monitored data. + * IA32_QM_CTR.Error (bit 63) and IA32_QM_CTR.Unavailable (bit 62) + * are error bits. + */ + wrmsr(MSR_IA32_QM_EVTSEL, eventid, rmid); + rdmsrl(MSR_IA32_QM_CTR, val); + + return val; +} + +static bool rmid_dirty(struct rmid_entry *entry) +{ + u64 val = __rmid_read(entry->rmid, QOS_L3_OCCUP_EVENT_ID); + + return val >= resctrl_cqm_threshold; +} + +/* + * Check the RMIDs that are marked as busy for this domain. If the + * reported LLC occupancy is below the threshold clear the busy bit and + * decrement the count. If the busy count gets to zero on an RMID, we + * free the RMID + */ +void __check_limbo(struct rdt_domain *d, bool force_free) +{ + struct rmid_entry *entry; + struct rdt_resource *r; + u32 crmid = 1, nrmid; + + r = &rdt_resources_all[RDT_RESOURCE_L3]; + + /* + * Skip RMID 0 and start from RMID 1 and check all the RMIDs that + * are marked as busy for occupancy < threshold. If the occupancy + * is less than the threshold decrement the busy counter of the + * RMID and move it to the free list when the counter reaches 0. + */ + for (;;) { + nrmid = find_next_bit(d->rmid_busy_llc, r->num_rmid, crmid); + if (nrmid >= r->num_rmid) + break; + + entry = __rmid_entry(nrmid); + if (force_free || !rmid_dirty(entry)) { + clear_bit(entry->rmid, d->rmid_busy_llc); + if (!--entry->busy) { + rmid_limbo_count--; + list_add_tail(&entry->list, &rmid_free_lru); + } + } + crmid = nrmid + 1; + } +} + +bool has_busy_rmid(struct rdt_resource *r, struct rdt_domain *d) +{ + return find_first_bit(d->rmid_busy_llc, r->num_rmid) != r->num_rmid; +} + +/* + * As of now the RMIDs allocation is global. + * However we keep track of which packages the RMIDs + * are used to optimize the limbo list management. + */ +int alloc_rmid(void) +{ + struct rmid_entry *entry; + + lockdep_assert_held(&rdtgroup_mutex); + + if (list_empty(&rmid_free_lru)) + return rmid_limbo_count ? -EBUSY : -ENOSPC; + + entry = list_first_entry(&rmid_free_lru, + struct rmid_entry, list); + list_del(&entry->list); + + return entry->rmid; +} + +static void add_rmid_to_limbo(struct rmid_entry *entry) +{ + struct rdt_resource *r; + struct rdt_domain *d; + int cpu; + u64 val; + + r = &rdt_resources_all[RDT_RESOURCE_L3]; + + entry->busy = 0; + cpu = get_cpu(); + list_for_each_entry(d, &r->domains, list) { + if (cpumask_test_cpu(cpu, &d->cpu_mask)) { + val = __rmid_read(entry->rmid, QOS_L3_OCCUP_EVENT_ID); + if (val <= resctrl_cqm_threshold) + continue; + } + + /* + * For the first limbo RMID in the domain, + * setup up the limbo worker. + */ + if (!has_busy_rmid(r, d)) + cqm_setup_limbo_handler(d, CQM_LIMBOCHECK_INTERVAL); + set_bit(entry->rmid, d->rmid_busy_llc); + entry->busy++; + } + put_cpu(); + + if (entry->busy) + rmid_limbo_count++; + else + list_add_tail(&entry->list, &rmid_free_lru); +} + +void free_rmid(u32 rmid) +{ + struct rmid_entry *entry; + + if (!rmid) + return; + + lockdep_assert_held(&rdtgroup_mutex); + + entry = __rmid_entry(rmid); + + if (is_llc_occupancy_enabled()) + add_rmid_to_limbo(entry); + else + list_add_tail(&entry->list, &rmid_free_lru); +} + +static u64 mbm_overflow_count(u64 prev_msr, u64 cur_msr, unsigned int width) +{ + u64 shift = 64 - width, chunks; + + chunks = (cur_msr << shift) - (prev_msr << shift); + return chunks >>= shift; +} + +static u64 __mon_event_count(u32 rmid, struct rmid_read *rr) +{ + struct mbm_state *m; + u64 chunks, tval; + + tval = __rmid_read(rmid, rr->evtid); + if (tval & (RMID_VAL_ERROR | RMID_VAL_UNAVAIL)) { + return tval; + } + switch (rr->evtid) { + case QOS_L3_OCCUP_EVENT_ID: + rr->val += tval; + return 0; + case QOS_L3_MBM_TOTAL_EVENT_ID: + m = &rr->d->mbm_total[rmid]; + break; + case QOS_L3_MBM_LOCAL_EVENT_ID: + m = &rr->d->mbm_local[rmid]; + break; + default: + /* + * Code would never reach here because an invalid + * event id would fail the __rmid_read. + */ + return RMID_VAL_ERROR; + } + + if (rr->first) { + memset(m, 0, sizeof(struct mbm_state)); + m->prev_bw_msr = m->prev_msr = tval; + return 0; + } + + chunks = mbm_overflow_count(m->prev_msr, tval, rr->r->mbm_width); + m->chunks += chunks; + m->prev_msr = tval; + + rr->val += m->chunks; + return 0; +} + +/* + * Supporting function to calculate the memory bandwidth + * and delta bandwidth in MBps. + */ +static void mbm_bw_count(u32 rmid, struct rmid_read *rr) +{ + struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3]; + struct mbm_state *m = &rr->d->mbm_local[rmid]; + u64 tval, cur_bw, chunks; + + tval = __rmid_read(rmid, rr->evtid); + if (tval & (RMID_VAL_ERROR | RMID_VAL_UNAVAIL)) + return; + + chunks = mbm_overflow_count(m->prev_bw_msr, tval, rr->r->mbm_width); + cur_bw = (chunks * r->mon_scale) >> 20; + + if (m->delta_comp) + m->delta_bw = abs(cur_bw - m->prev_bw); + m->delta_comp = false; + m->prev_bw = cur_bw; + m->prev_bw_msr = tval; +} + +/* + * This is called via IPI to read the CQM/MBM counters + * on a domain. + */ +void mon_event_count(void *info) +{ + struct rdtgroup *rdtgrp, *entry; + struct rmid_read *rr = info; + struct list_head *head; + u64 ret_val; + + rdtgrp = rr->rgrp; + + ret_val = __mon_event_count(rdtgrp->mon.rmid, rr); + + /* + * For Ctrl groups read data from child monitor groups and + * add them together. Count events which are read successfully. + * Discard the rmid_read's reporting errors. + */ + head = &rdtgrp->mon.crdtgrp_list; + + if (rdtgrp->type == RDTCTRL_GROUP) { + list_for_each_entry(entry, head, mon.crdtgrp_list) { + if (__mon_event_count(entry->mon.rmid, rr) == 0) + ret_val = 0; + } + } + + /* Report error if none of rmid_reads are successful */ + if (ret_val) + rr->val = ret_val; +} + +/* + * Feedback loop for MBA software controller (mba_sc) + * + * mba_sc is a feedback loop where we periodically read MBM counters and + * adjust the bandwidth percentage values via the IA32_MBA_THRTL_MSRs so + * that: + * + * current bandwdith(cur_bw) < user specified bandwidth(user_bw) + * + * This uses the MBM counters to measure the bandwidth and MBA throttle + * MSRs to control the bandwidth for a particular rdtgrp. It builds on the + * fact that resctrl rdtgroups have both monitoring and control. + * + * The frequency of the checks is 1s and we just tag along the MBM overflow + * timer. Having 1s interval makes the calculation of bandwidth simpler. + * + * Although MBA's goal is to restrict the bandwidth to a maximum, there may + * be a need to increase the bandwidth to avoid uncecessarily restricting + * the L2 <-> L3 traffic. + * + * Since MBA controls the L2 external bandwidth where as MBM measures the + * L3 external bandwidth the following sequence could lead to such a + * situation. + * + * Consider an rdtgroup which had high L3 <-> memory traffic in initial + * phases -> mba_sc kicks in and reduced bandwidth percentage values -> but + * after some time rdtgroup has mostly L2 <-> L3 traffic. + * + * In this case we may restrict the rdtgroup's L2 <-> L3 traffic as its + * throttle MSRs already have low percentage values. To avoid + * unnecessarily restricting such rdtgroups, we also increase the bandwidth. + */ +static void update_mba_bw(struct rdtgroup *rgrp, struct rdt_domain *dom_mbm) +{ + u32 closid, rmid, cur_msr, cur_msr_val, new_msr_val; + struct mbm_state *pmbm_data, *cmbm_data; + u32 cur_bw, delta_bw, user_bw; + struct rdt_resource *r_mba; + struct rdt_domain *dom_mba; + struct list_head *head; + struct rdtgroup *entry; + + if (!is_mbm_local_enabled()) + return; + + r_mba = &rdt_resources_all[RDT_RESOURCE_MBA]; + closid = rgrp->closid; + rmid = rgrp->mon.rmid; + pmbm_data = &dom_mbm->mbm_local[rmid]; + + dom_mba = get_domain_from_cpu(smp_processor_id(), r_mba); + if (!dom_mba) { + pr_warn_once("Failure to get domain for MBA update\n"); + return; + } + + cur_bw = pmbm_data->prev_bw; + user_bw = dom_mba->mbps_val[closid]; + delta_bw = pmbm_data->delta_bw; + cur_msr_val = dom_mba->ctrl_val[closid]; + + /* + * For Ctrl groups read data from child monitor groups. + */ + head = &rgrp->mon.crdtgrp_list; + list_for_each_entry(entry, head, mon.crdtgrp_list) { + cmbm_data = &dom_mbm->mbm_local[entry->mon.rmid]; + cur_bw += cmbm_data->prev_bw; + delta_bw += cmbm_data->delta_bw; + } + + /* + * Scale up/down the bandwidth linearly for the ctrl group. The + * bandwidth step is the bandwidth granularity specified by the + * hardware. + * + * The delta_bw is used when increasing the bandwidth so that we + * dont alternately increase and decrease the control values + * continuously. + * + * For ex: consider cur_bw = 90MBps, user_bw = 100MBps and if + * bandwidth step is 20MBps(> user_bw - cur_bw), we would keep + * switching between 90 and 110 continuously if we only check + * cur_bw < user_bw. + */ + if (cur_msr_val > r_mba->membw.min_bw && user_bw < cur_bw) { + new_msr_val = cur_msr_val - r_mba->membw.bw_gran; + } else if (cur_msr_val < MAX_MBA_BW && + (user_bw > (cur_bw + delta_bw))) { + new_msr_val = cur_msr_val + r_mba->membw.bw_gran; + } else { + return; + } + + cur_msr = r_mba->msr_base + closid; + wrmsrl(cur_msr, delay_bw_map(new_msr_val, r_mba)); + dom_mba->ctrl_val[closid] = new_msr_val; + + /* + * Delta values are updated dynamically package wise for each + * rdtgrp everytime the throttle MSR changes value. + * + * This is because (1)the increase in bandwidth is not perfectly + * linear and only "approximately" linear even when the hardware + * says it is linear.(2)Also since MBA is a core specific + * mechanism, the delta values vary based on number of cores used + * by the rdtgrp. + */ + pmbm_data->delta_comp = true; + list_for_each_entry(entry, head, mon.crdtgrp_list) { + cmbm_data = &dom_mbm->mbm_local[entry->mon.rmid]; + cmbm_data->delta_comp = true; + } +} + +static void mbm_update(struct rdt_resource *r, struct rdt_domain *d, int rmid) +{ + struct rmid_read rr; + + rr.first = false; + rr.r = r; + rr.d = d; + + /* + * This is protected from concurrent reads from user + * as both the user and we hold the global mutex. + */ + if (is_mbm_total_enabled()) { + rr.evtid = QOS_L3_MBM_TOTAL_EVENT_ID; + __mon_event_count(rmid, &rr); + } + if (is_mbm_local_enabled()) { + rr.evtid = QOS_L3_MBM_LOCAL_EVENT_ID; + __mon_event_count(rmid, &rr); + + /* + * Call the MBA software controller only for the + * control groups and when user has enabled + * the software controller explicitly. + */ + if (is_mba_sc(NULL)) + mbm_bw_count(rmid, &rr); + } +} + +/* + * Handler to scan the limbo list and move the RMIDs + * to free list whose occupancy < threshold_occupancy. + */ +void cqm_handle_limbo(struct work_struct *work) +{ + unsigned long delay = msecs_to_jiffies(CQM_LIMBOCHECK_INTERVAL); + int cpu = smp_processor_id(); + struct rdt_resource *r; + struct rdt_domain *d; + + mutex_lock(&rdtgroup_mutex); + + r = &rdt_resources_all[RDT_RESOURCE_L3]; + d = container_of(work, struct rdt_domain, cqm_limbo.work); + + __check_limbo(d, false); + + if (has_busy_rmid(r, d)) + schedule_delayed_work_on(cpu, &d->cqm_limbo, delay); + + mutex_unlock(&rdtgroup_mutex); +} + +void cqm_setup_limbo_handler(struct rdt_domain *dom, unsigned long delay_ms) +{ + unsigned long delay = msecs_to_jiffies(delay_ms); + int cpu; + + cpu = cpumask_any(&dom->cpu_mask); + dom->cqm_work_cpu = cpu; + + schedule_delayed_work_on(cpu, &dom->cqm_limbo, delay); +} + +void mbm_handle_overflow(struct work_struct *work) +{ + unsigned long delay = msecs_to_jiffies(MBM_OVERFLOW_INTERVAL); + struct rdtgroup *prgrp, *crgrp; + int cpu = smp_processor_id(); + struct list_head *head; + struct rdt_resource *r; + struct rdt_domain *d; + + mutex_lock(&rdtgroup_mutex); + + if (!static_branch_likely(&rdt_mon_enable_key)) + goto out_unlock; + + r = &rdt_resources_all[RDT_RESOURCE_L3]; + d = container_of(work, struct rdt_domain, mbm_over.work); + + list_for_each_entry(prgrp, &rdt_all_groups, rdtgroup_list) { + mbm_update(r, d, prgrp->mon.rmid); + + head = &prgrp->mon.crdtgrp_list; + list_for_each_entry(crgrp, head, mon.crdtgrp_list) + mbm_update(r, d, crgrp->mon.rmid); + + if (is_mba_sc(NULL)) + update_mba_bw(prgrp, d); + } + + schedule_delayed_work_on(cpu, &d->mbm_over, delay); + +out_unlock: + mutex_unlock(&rdtgroup_mutex); +} + +void mbm_setup_overflow_handler(struct rdt_domain *dom, unsigned long delay_ms) +{ + unsigned long delay = msecs_to_jiffies(delay_ms); + int cpu; + + if (!static_branch_likely(&rdt_mon_enable_key)) + return; + cpu = cpumask_any(&dom->cpu_mask); + dom->mbm_work_cpu = cpu; + schedule_delayed_work_on(cpu, &dom->mbm_over, delay); +} + +static int dom_data_init(struct rdt_resource *r) +{ + struct rmid_entry *entry = NULL; + int i, nr_rmids; + + nr_rmids = r->num_rmid; + rmid_ptrs = kcalloc(nr_rmids, sizeof(struct rmid_entry), GFP_KERNEL); + if (!rmid_ptrs) + return -ENOMEM; + + for (i = 0; i < nr_rmids; i++) { + entry = &rmid_ptrs[i]; + INIT_LIST_HEAD(&entry->list); + + entry->rmid = i; + list_add_tail(&entry->list, &rmid_free_lru); + } + + /* + * RMID 0 is special and is always allocated. It's used for all + * tasks that are not monitored. + */ + entry = __rmid_entry(0); + list_del(&entry->list); + + return 0; +} + +static struct mon_evt llc_occupancy_event = { + .name = "llc_occupancy", + .evtid = QOS_L3_OCCUP_EVENT_ID, +}; + +static struct mon_evt mbm_total_event = { + .name = "mbm_total_bytes", + .evtid = QOS_L3_MBM_TOTAL_EVENT_ID, +}; + +static struct mon_evt mbm_local_event = { + .name = "mbm_local_bytes", + .evtid = QOS_L3_MBM_LOCAL_EVENT_ID, +}; + +/* + * Initialize the event list for the resource. + * + * Note that MBM events are also part of RDT_RESOURCE_L3 resource + * because as per the SDM the total and local memory bandwidth + * are enumerated as part of L3 monitoring. + */ +static void l3_mon_evt_init(struct rdt_resource *r) +{ + INIT_LIST_HEAD(&r->evt_list); + + if (is_llc_occupancy_enabled()) + list_add_tail(&llc_occupancy_event.list, &r->evt_list); + if (is_mbm_total_enabled()) + list_add_tail(&mbm_total_event.list, &r->evt_list); + if (is_mbm_local_enabled()) + list_add_tail(&mbm_local_event.list, &r->evt_list); +} + +int rdt_get_mon_l3_config(struct rdt_resource *r) +{ + unsigned int mbm_offset = boot_cpu_data.x86_cache_mbm_width_offset; + unsigned int cl_size = boot_cpu_data.x86_cache_size; + int ret; + + r->mon_scale = boot_cpu_data.x86_cache_occ_scale; + r->num_rmid = boot_cpu_data.x86_cache_max_rmid + 1; + r->mbm_width = MBM_CNTR_WIDTH_BASE; + + if (mbm_offset > 0 && mbm_offset <= MBM_CNTR_WIDTH_OFFSET_MAX) + r->mbm_width += mbm_offset; + else if (mbm_offset > MBM_CNTR_WIDTH_OFFSET_MAX) + pr_warn("Ignoring impossible MBM counter offset\n"); + + /* + * A reasonable upper limit on the max threshold is the number + * of lines tagged per RMID if all RMIDs have the same number of + * lines tagged in the LLC. + * + * For a 35MB LLC and 56 RMIDs, this is ~1.8% of the LLC. + */ + resctrl_cqm_threshold = cl_size * 1024 / r->num_rmid; + + /* h/w works in units of "boot_cpu_data.x86_cache_occ_scale" */ + resctrl_cqm_threshold /= r->mon_scale; + + ret = dom_data_init(r); + if (ret) + return ret; + + l3_mon_evt_init(r); + + r->mon_capable = true; + r->mon_enabled = true; + + return 0; +} diff --git a/arch/x86/kernel/cpu/resctrl/pseudo_lock.c b/arch/x86/kernel/cpu/resctrl/pseudo_lock.c new file mode 100644 index 000000000..465dce141 --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/pseudo_lock.c @@ -0,0 +1,1594 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Resource Director Technology (RDT) + * + * Pseudo-locking support built on top of Cache Allocation Technology (CAT) + * + * Copyright (C) 2018 Intel Corporation + * + * Author: Reinette Chatre <reinette.chatre@intel.com> + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cacheinfo.h> +#include <linux/cpu.h> +#include <linux/cpumask.h> +#include <linux/debugfs.h> +#include <linux/kthread.h> +#include <linux/mman.h> +#include <linux/perf_event.h> +#include <linux/pm_qos.h> +#include <linux/slab.h> +#include <linux/uaccess.h> + +#include <asm/cacheflush.h> +#include <asm/intel-family.h> +#include <asm/resctrl.h> +#include <asm/perf_event.h> + +#include "../../events/perf_event.h" /* For X86_CONFIG() */ +#include "internal.h" + +#define CREATE_TRACE_POINTS +#include "pseudo_lock_event.h" + +/* + * The bits needed to disable hardware prefetching varies based on the + * platform. During initialization we will discover which bits to use. + */ +static u64 prefetch_disable_bits; + +/* + * Major number assigned to and shared by all devices exposing + * pseudo-locked regions. + */ +static unsigned int pseudo_lock_major; +static unsigned long pseudo_lock_minor_avail = GENMASK(MINORBITS, 0); +static struct class *pseudo_lock_class; + +/** + * get_prefetch_disable_bits - prefetch disable bits of supported platforms + * + * Capture the list of platforms that have been validated to support + * pseudo-locking. This includes testing to ensure pseudo-locked regions + * with low cache miss rates can be created under variety of load conditions + * as well as that these pseudo-locked regions can maintain their low cache + * miss rates under variety of load conditions for significant lengths of time. + * + * After a platform has been validated to support pseudo-locking its + * hardware prefetch disable bits are included here as they are documented + * in the SDM. + * + * When adding a platform here also add support for its cache events to + * measure_cycles_perf_fn() + * + * Return: + * If platform is supported, the bits to disable hardware prefetchers, 0 + * if platform is not supported. + */ +static u64 get_prefetch_disable_bits(void) +{ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL || + boot_cpu_data.x86 != 6) + return 0; + + switch (boot_cpu_data.x86_model) { + case INTEL_FAM6_BROADWELL_X: + /* + * SDM defines bits of MSR_MISC_FEATURE_CONTROL register + * as: + * 0 L2 Hardware Prefetcher Disable (R/W) + * 1 L2 Adjacent Cache Line Prefetcher Disable (R/W) + * 2 DCU Hardware Prefetcher Disable (R/W) + * 3 DCU IP Prefetcher Disable (R/W) + * 63:4 Reserved + */ + return 0xF; + case INTEL_FAM6_ATOM_GOLDMONT: + case INTEL_FAM6_ATOM_GOLDMONT_PLUS: + /* + * SDM defines bits of MSR_MISC_FEATURE_CONTROL register + * as: + * 0 L2 Hardware Prefetcher Disable (R/W) + * 1 Reserved + * 2 DCU Hardware Prefetcher Disable (R/W) + * 63:3 Reserved + */ + return 0x5; + } + + return 0; +} + +/** + * pseudo_lock_minor_get - Obtain available minor number + * @minor: Pointer to where new minor number will be stored + * + * A bitmask is used to track available minor numbers. Here the next free + * minor number is marked as unavailable and returned. + * + * Return: 0 on success, <0 on failure. + */ +static int pseudo_lock_minor_get(unsigned int *minor) +{ + unsigned long first_bit; + + first_bit = find_first_bit(&pseudo_lock_minor_avail, MINORBITS); + + if (first_bit == MINORBITS) + return -ENOSPC; + + __clear_bit(first_bit, &pseudo_lock_minor_avail); + *minor = first_bit; + + return 0; +} + +/** + * pseudo_lock_minor_release - Return minor number to available + * @minor: The minor number made available + */ +static void pseudo_lock_minor_release(unsigned int minor) +{ + __set_bit(minor, &pseudo_lock_minor_avail); +} + +/** + * region_find_by_minor - Locate a pseudo-lock region by inode minor number + * @minor: The minor number of the device representing pseudo-locked region + * + * When the character device is accessed we need to determine which + * pseudo-locked region it belongs to. This is done by matching the minor + * number of the device to the pseudo-locked region it belongs. + * + * Minor numbers are assigned at the time a pseudo-locked region is associated + * with a cache instance. + * + * Return: On success return pointer to resource group owning the pseudo-locked + * region, NULL on failure. + */ +static struct rdtgroup *region_find_by_minor(unsigned int minor) +{ + struct rdtgroup *rdtgrp, *rdtgrp_match = NULL; + + list_for_each_entry(rdtgrp, &rdt_all_groups, rdtgroup_list) { + if (rdtgrp->plr && rdtgrp->plr->minor == minor) { + rdtgrp_match = rdtgrp; + break; + } + } + return rdtgrp_match; +} + +/** + * pseudo_lock_pm_req - A power management QoS request list entry + * @list: Entry within the @pm_reqs list for a pseudo-locked region + * @req: PM QoS request + */ +struct pseudo_lock_pm_req { + struct list_head list; + struct dev_pm_qos_request req; +}; + +static void pseudo_lock_cstates_relax(struct pseudo_lock_region *plr) +{ + struct pseudo_lock_pm_req *pm_req, *next; + + list_for_each_entry_safe(pm_req, next, &plr->pm_reqs, list) { + dev_pm_qos_remove_request(&pm_req->req); + list_del(&pm_req->list); + kfree(pm_req); + } +} + +/** + * pseudo_lock_cstates_constrain - Restrict cores from entering C6 + * + * To prevent the cache from being affected by power management entering + * C6 has to be avoided. This is accomplished by requesting a latency + * requirement lower than lowest C6 exit latency of all supported + * platforms as found in the cpuidle state tables in the intel_idle driver. + * At this time it is possible to do so with a single latency requirement + * for all supported platforms. + * + * Since Goldmont is supported, which is affected by X86_BUG_MONITOR, + * the ACPI latencies need to be considered while keeping in mind that C2 + * may be set to map to deeper sleep states. In this case the latency + * requirement needs to prevent entering C2 also. + */ +static int pseudo_lock_cstates_constrain(struct pseudo_lock_region *plr) +{ + struct pseudo_lock_pm_req *pm_req; + int cpu; + int ret; + + for_each_cpu(cpu, &plr->d->cpu_mask) { + pm_req = kzalloc(sizeof(*pm_req), GFP_KERNEL); + if (!pm_req) { + rdt_last_cmd_puts("Failure to allocate memory for PM QoS\n"); + ret = -ENOMEM; + goto out_err; + } + ret = dev_pm_qos_add_request(get_cpu_device(cpu), + &pm_req->req, + DEV_PM_QOS_RESUME_LATENCY, + 30); + if (ret < 0) { + rdt_last_cmd_printf("Failed to add latency req CPU%d\n", + cpu); + kfree(pm_req); + ret = -1; + goto out_err; + } + list_add(&pm_req->list, &plr->pm_reqs); + } + + return 0; + +out_err: + pseudo_lock_cstates_relax(plr); + return ret; +} + +/** + * pseudo_lock_region_clear - Reset pseudo-lock region data + * @plr: pseudo-lock region + * + * All content of the pseudo-locked region is reset - any memory allocated + * freed. + * + * Return: void + */ +static void pseudo_lock_region_clear(struct pseudo_lock_region *plr) +{ + plr->size = 0; + plr->line_size = 0; + kfree(plr->kmem); + plr->kmem = NULL; + plr->r = NULL; + if (plr->d) + plr->d->plr = NULL; + plr->d = NULL; + plr->cbm = 0; + plr->debugfs_dir = NULL; +} + +/** + * pseudo_lock_region_init - Initialize pseudo-lock region information + * @plr: pseudo-lock region + * + * Called after user provided a schemata to be pseudo-locked. From the + * schemata the &struct pseudo_lock_region is on entry already initialized + * with the resource, domain, and capacity bitmask. Here the information + * required for pseudo-locking is deduced from this data and &struct + * pseudo_lock_region initialized further. This information includes: + * - size in bytes of the region to be pseudo-locked + * - cache line size to know the stride with which data needs to be accessed + * to be pseudo-locked + * - a cpu associated with the cache instance on which the pseudo-locking + * flow can be executed + * + * Return: 0 on success, <0 on failure. Descriptive error will be written + * to last_cmd_status buffer. + */ +static int pseudo_lock_region_init(struct pseudo_lock_region *plr) +{ + struct cpu_cacheinfo *ci; + int ret; + int i; + + /* Pick the first cpu we find that is associated with the cache. */ + plr->cpu = cpumask_first(&plr->d->cpu_mask); + + if (!cpu_online(plr->cpu)) { + rdt_last_cmd_printf("CPU %u associated with cache not online\n", + plr->cpu); + ret = -ENODEV; + goto out_region; + } + + ci = get_cpu_cacheinfo(plr->cpu); + + plr->size = rdtgroup_cbm_to_size(plr->r, plr->d, plr->cbm); + + for (i = 0; i < ci->num_leaves; i++) { + if (ci->info_list[i].level == plr->r->cache_level) { + plr->line_size = ci->info_list[i].coherency_line_size; + return 0; + } + } + + ret = -1; + rdt_last_cmd_puts("Unable to determine cache line size\n"); +out_region: + pseudo_lock_region_clear(plr); + return ret; +} + +/** + * pseudo_lock_init - Initialize a pseudo-lock region + * @rdtgrp: resource group to which new pseudo-locked region will belong + * + * A pseudo-locked region is associated with a resource group. When this + * association is created the pseudo-locked region is initialized. The + * details of the pseudo-locked region are not known at this time so only + * allocation is done and association established. + * + * Return: 0 on success, <0 on failure + */ +static int pseudo_lock_init(struct rdtgroup *rdtgrp) +{ + struct pseudo_lock_region *plr; + + plr = kzalloc(sizeof(*plr), GFP_KERNEL); + if (!plr) + return -ENOMEM; + + init_waitqueue_head(&plr->lock_thread_wq); + INIT_LIST_HEAD(&plr->pm_reqs); + rdtgrp->plr = plr; + return 0; +} + +/** + * pseudo_lock_region_alloc - Allocate kernel memory that will be pseudo-locked + * @plr: pseudo-lock region + * + * Initialize the details required to set up the pseudo-locked region and + * allocate the contiguous memory that will be pseudo-locked to the cache. + * + * Return: 0 on success, <0 on failure. Descriptive error will be written + * to last_cmd_status buffer. + */ +static int pseudo_lock_region_alloc(struct pseudo_lock_region *plr) +{ + int ret; + + ret = pseudo_lock_region_init(plr); + if (ret < 0) + return ret; + + /* + * We do not yet support contiguous regions larger than + * KMALLOC_MAX_SIZE. + */ + if (plr->size > KMALLOC_MAX_SIZE) { + rdt_last_cmd_puts("Requested region exceeds maximum size\n"); + ret = -E2BIG; + goto out_region; + } + + plr->kmem = kzalloc(plr->size, GFP_KERNEL); + if (!plr->kmem) { + rdt_last_cmd_puts("Unable to allocate memory\n"); + ret = -ENOMEM; + goto out_region; + } + + ret = 0; + goto out; +out_region: + pseudo_lock_region_clear(plr); +out: + return ret; +} + +/** + * pseudo_lock_free - Free a pseudo-locked region + * @rdtgrp: resource group to which pseudo-locked region belonged + * + * The pseudo-locked region's resources have already been released, or not + * yet created at this point. Now it can be freed and disassociated from the + * resource group. + * + * Return: void + */ +static void pseudo_lock_free(struct rdtgroup *rdtgrp) +{ + pseudo_lock_region_clear(rdtgrp->plr); + kfree(rdtgrp->plr); + rdtgrp->plr = NULL; +} + +/** + * pseudo_lock_fn - Load kernel memory into cache + * @_rdtgrp: resource group to which pseudo-lock region belongs + * + * This is the core pseudo-locking flow. + * + * First we ensure that the kernel memory cannot be found in the cache. + * Then, while taking care that there will be as little interference as + * possible, the memory to be loaded is accessed while core is running + * with class of service set to the bitmask of the pseudo-locked region. + * After this is complete no future CAT allocations will be allowed to + * overlap with this bitmask. + * + * Local register variables are utilized to ensure that the memory region + * to be locked is the only memory access made during the critical locking + * loop. + * + * Return: 0. Waiter on waitqueue will be woken on completion. + */ +static int pseudo_lock_fn(void *_rdtgrp) +{ + struct rdtgroup *rdtgrp = _rdtgrp; + struct pseudo_lock_region *plr = rdtgrp->plr; + u32 rmid_p, closid_p; + unsigned long i; + u64 saved_msr; +#ifdef CONFIG_KASAN + /* + * The registers used for local register variables are also used + * when KASAN is active. When KASAN is active we use a regular + * variable to ensure we always use a valid pointer, but the cost + * is that this variable will enter the cache through evicting the + * memory we are trying to lock into the cache. Thus expect lower + * pseudo-locking success rate when KASAN is active. + */ + unsigned int line_size; + unsigned int size; + void *mem_r; +#else + register unsigned int line_size asm("esi"); + register unsigned int size asm("edi"); + register void *mem_r asm(_ASM_BX); +#endif /* CONFIG_KASAN */ + + /* + * Make sure none of the allocated memory is cached. If it is we + * will get a cache hit in below loop from outside of pseudo-locked + * region. + * wbinvd (as opposed to clflush/clflushopt) is required to + * increase likelihood that allocated cache portion will be filled + * with associated memory. + */ + native_wbinvd(); + + /* + * Always called with interrupts enabled. By disabling interrupts + * ensure that we will not be preempted during this critical section. + */ + local_irq_disable(); + + /* + * Call wrmsr and rdmsr as directly as possible to avoid tracing + * clobbering local register variables or affecting cache accesses. + * + * Disable the hardware prefetcher so that when the end of the memory + * being pseudo-locked is reached the hardware will not read beyond + * the buffer and evict pseudo-locked memory read earlier from the + * cache. + */ + saved_msr = __rdmsr(MSR_MISC_FEATURE_CONTROL); + __wrmsr(MSR_MISC_FEATURE_CONTROL, prefetch_disable_bits, 0x0); + closid_p = this_cpu_read(pqr_state.cur_closid); + rmid_p = this_cpu_read(pqr_state.cur_rmid); + mem_r = plr->kmem; + size = plr->size; + line_size = plr->line_size; + /* + * Critical section begin: start by writing the closid associated + * with the capacity bitmask of the cache region being + * pseudo-locked followed by reading of kernel memory to load it + * into the cache. + */ + __wrmsr(IA32_PQR_ASSOC, rmid_p, rdtgrp->closid); + /* + * Cache was flushed earlier. Now access kernel memory to read it + * into cache region associated with just activated plr->closid. + * Loop over data twice: + * - In first loop the cache region is shared with the page walker + * as it populates the paging structure caches (including TLB). + * - In the second loop the paging structure caches are used and + * cache region is populated with the memory being referenced. + */ + for (i = 0; i < size; i += PAGE_SIZE) { + /* + * Add a barrier to prevent speculative execution of this + * loop reading beyond the end of the buffer. + */ + rmb(); + asm volatile("mov (%0,%1,1), %%eax\n\t" + : + : "r" (mem_r), "r" (i) + : "%eax", "memory"); + } + for (i = 0; i < size; i += line_size) { + /* + * Add a barrier to prevent speculative execution of this + * loop reading beyond the end of the buffer. + */ + rmb(); + asm volatile("mov (%0,%1,1), %%eax\n\t" + : + : "r" (mem_r), "r" (i) + : "%eax", "memory"); + } + /* + * Critical section end: restore closid with capacity bitmask that + * does not overlap with pseudo-locked region. + */ + __wrmsr(IA32_PQR_ASSOC, rmid_p, closid_p); + + /* Re-enable the hardware prefetcher(s) */ + wrmsrl(MSR_MISC_FEATURE_CONTROL, saved_msr); + local_irq_enable(); + + plr->thread_done = 1; + wake_up_interruptible(&plr->lock_thread_wq); + return 0; +} + +/** + * rdtgroup_monitor_in_progress - Test if monitoring in progress + * @r: resource group being queried + * + * Return: 1 if monitor groups have been created for this resource + * group, 0 otherwise. + */ +static int rdtgroup_monitor_in_progress(struct rdtgroup *rdtgrp) +{ + return !list_empty(&rdtgrp->mon.crdtgrp_list); +} + +/** + * rdtgroup_locksetup_user_restrict - Restrict user access to group + * @rdtgrp: resource group needing access restricted + * + * A resource group used for cache pseudo-locking cannot have cpus or tasks + * assigned to it. This is communicated to the user by restricting access + * to all the files that can be used to make such changes. + * + * Permissions restored with rdtgroup_locksetup_user_restore() + * + * Return: 0 on success, <0 on failure. If a failure occurs during the + * restriction of access an attempt will be made to restore permissions but + * the state of the mode of these files will be uncertain when a failure + * occurs. + */ +static int rdtgroup_locksetup_user_restrict(struct rdtgroup *rdtgrp) +{ + int ret; + + ret = rdtgroup_kn_mode_restrict(rdtgrp, "tasks"); + if (ret) + return ret; + + ret = rdtgroup_kn_mode_restrict(rdtgrp, "cpus"); + if (ret) + goto err_tasks; + + ret = rdtgroup_kn_mode_restrict(rdtgrp, "cpus_list"); + if (ret) + goto err_cpus; + + if (rdt_mon_capable) { + ret = rdtgroup_kn_mode_restrict(rdtgrp, "mon_groups"); + if (ret) + goto err_cpus_list; + } + + ret = 0; + goto out; + +err_cpus_list: + rdtgroup_kn_mode_restore(rdtgrp, "cpus_list", 0777); +err_cpus: + rdtgroup_kn_mode_restore(rdtgrp, "cpus", 0777); +err_tasks: + rdtgroup_kn_mode_restore(rdtgrp, "tasks", 0777); +out: + return ret; +} + +/** + * rdtgroup_locksetup_user_restore - Restore user access to group + * @rdtgrp: resource group needing access restored + * + * Restore all file access previously removed using + * rdtgroup_locksetup_user_restrict() + * + * Return: 0 on success, <0 on failure. If a failure occurs during the + * restoration of access an attempt will be made to restrict permissions + * again but the state of the mode of these files will be uncertain when + * a failure occurs. + */ +static int rdtgroup_locksetup_user_restore(struct rdtgroup *rdtgrp) +{ + int ret; + + ret = rdtgroup_kn_mode_restore(rdtgrp, "tasks", 0777); + if (ret) + return ret; + + ret = rdtgroup_kn_mode_restore(rdtgrp, "cpus", 0777); + if (ret) + goto err_tasks; + + ret = rdtgroup_kn_mode_restore(rdtgrp, "cpus_list", 0777); + if (ret) + goto err_cpus; + + if (rdt_mon_capable) { + ret = rdtgroup_kn_mode_restore(rdtgrp, "mon_groups", 0777); + if (ret) + goto err_cpus_list; + } + + ret = 0; + goto out; + +err_cpus_list: + rdtgroup_kn_mode_restrict(rdtgrp, "cpus_list"); +err_cpus: + rdtgroup_kn_mode_restrict(rdtgrp, "cpus"); +err_tasks: + rdtgroup_kn_mode_restrict(rdtgrp, "tasks"); +out: + return ret; +} + +/** + * rdtgroup_locksetup_enter - Resource group enters locksetup mode + * @rdtgrp: resource group requested to enter locksetup mode + * + * A resource group enters locksetup mode to reflect that it would be used + * to represent a pseudo-locked region and is in the process of being set + * up to do so. A resource group used for a pseudo-locked region would + * lose the closid associated with it so we cannot allow it to have any + * tasks or cpus assigned nor permit tasks or cpus to be assigned in the + * future. Monitoring of a pseudo-locked region is not allowed either. + * + * The above and more restrictions on a pseudo-locked region are checked + * for and enforced before the resource group enters the locksetup mode. + * + * Returns: 0 if the resource group successfully entered locksetup mode, <0 + * on failure. On failure the last_cmd_status buffer is updated with text to + * communicate details of failure to the user. + */ +int rdtgroup_locksetup_enter(struct rdtgroup *rdtgrp) +{ + int ret; + + /* + * The default resource group can neither be removed nor lose the + * default closid associated with it. + */ + if (rdtgrp == &rdtgroup_default) { + rdt_last_cmd_puts("Cannot pseudo-lock default group\n"); + return -EINVAL; + } + + /* + * Cache Pseudo-locking not supported when CDP is enabled. + * + * Some things to consider if you would like to enable this + * support (using L3 CDP as example): + * - When CDP is enabled two separate resources are exposed, + * L3DATA and L3CODE, but they are actually on the same cache. + * The implication for pseudo-locking is that if a + * pseudo-locked region is created on a domain of one + * resource (eg. L3CODE), then a pseudo-locked region cannot + * be created on that same domain of the other resource + * (eg. L3DATA). This is because the creation of a + * pseudo-locked region involves a call to wbinvd that will + * affect all cache allocations on particular domain. + * - Considering the previous, it may be possible to only + * expose one of the CDP resources to pseudo-locking and + * hide the other. For example, we could consider to only + * expose L3DATA and since the L3 cache is unified it is + * still possible to place instructions there are execute it. + * - If only one region is exposed to pseudo-locking we should + * still keep in mind that availability of a portion of cache + * for pseudo-locking should take into account both resources. + * Similarly, if a pseudo-locked region is created in one + * resource, the portion of cache used by it should be made + * unavailable to all future allocations from both resources. + */ + if (rdt_resources_all[RDT_RESOURCE_L3DATA].alloc_enabled || + rdt_resources_all[RDT_RESOURCE_L2DATA].alloc_enabled) { + rdt_last_cmd_puts("CDP enabled\n"); + return -EINVAL; + } + + /* + * Not knowing the bits to disable prefetching implies that this + * platform does not support Cache Pseudo-Locking. + */ + prefetch_disable_bits = get_prefetch_disable_bits(); + if (prefetch_disable_bits == 0) { + rdt_last_cmd_puts("Pseudo-locking not supported\n"); + return -EINVAL; + } + + if (rdtgroup_monitor_in_progress(rdtgrp)) { + rdt_last_cmd_puts("Monitoring in progress\n"); + return -EINVAL; + } + + if (rdtgroup_tasks_assigned(rdtgrp)) { + rdt_last_cmd_puts("Tasks assigned to resource group\n"); + return -EINVAL; + } + + if (!cpumask_empty(&rdtgrp->cpu_mask)) { + rdt_last_cmd_puts("CPUs assigned to resource group\n"); + return -EINVAL; + } + + if (rdtgroup_locksetup_user_restrict(rdtgrp)) { + rdt_last_cmd_puts("Unable to modify resctrl permissions\n"); + return -EIO; + } + + ret = pseudo_lock_init(rdtgrp); + if (ret) { + rdt_last_cmd_puts("Unable to init pseudo-lock region\n"); + goto out_release; + } + + /* + * If this system is capable of monitoring a rmid would have been + * allocated when the control group was created. This is not needed + * anymore when this group would be used for pseudo-locking. This + * is safe to call on platforms not capable of monitoring. + */ + free_rmid(rdtgrp->mon.rmid); + + ret = 0; + goto out; + +out_release: + rdtgroup_locksetup_user_restore(rdtgrp); +out: + return ret; +} + +/** + * rdtgroup_locksetup_exit - resource group exist locksetup mode + * @rdtgrp: resource group + * + * When a resource group exits locksetup mode the earlier restrictions are + * lifted. + * + * Return: 0 on success, <0 on failure + */ +int rdtgroup_locksetup_exit(struct rdtgroup *rdtgrp) +{ + int ret; + + if (rdt_mon_capable) { + ret = alloc_rmid(); + if (ret < 0) { + rdt_last_cmd_puts("Out of RMIDs\n"); + return ret; + } + rdtgrp->mon.rmid = ret; + } + + ret = rdtgroup_locksetup_user_restore(rdtgrp); + if (ret) { + free_rmid(rdtgrp->mon.rmid); + return ret; + } + + pseudo_lock_free(rdtgrp); + return 0; +} + +/** + * rdtgroup_cbm_overlaps_pseudo_locked - Test if CBM or portion is pseudo-locked + * @d: RDT domain + * @cbm: CBM to test + * + * @d represents a cache instance and @cbm a capacity bitmask that is + * considered for it. Determine if @cbm overlaps with any existing + * pseudo-locked region on @d. + * + * @cbm is unsigned long, even if only 32 bits are used, to make the + * bitmap functions work correctly. + * + * Return: true if @cbm overlaps with pseudo-locked region on @d, false + * otherwise. + */ +bool rdtgroup_cbm_overlaps_pseudo_locked(struct rdt_domain *d, unsigned long cbm) +{ + unsigned int cbm_len; + unsigned long cbm_b; + + if (d->plr) { + cbm_len = d->plr->r->cache.cbm_len; + cbm_b = d->plr->cbm; + if (bitmap_intersects(&cbm, &cbm_b, cbm_len)) + return true; + } + return false; +} + +/** + * rdtgroup_pseudo_locked_in_hierarchy - Pseudo-locked region in cache hierarchy + * @d: RDT domain under test + * + * The setup of a pseudo-locked region affects all cache instances within + * the hierarchy of the region. It is thus essential to know if any + * pseudo-locked regions exist within a cache hierarchy to prevent any + * attempts to create new pseudo-locked regions in the same hierarchy. + * + * Return: true if a pseudo-locked region exists in the hierarchy of @d or + * if it is not possible to test due to memory allocation issue, + * false otherwise. + */ +bool rdtgroup_pseudo_locked_in_hierarchy(struct rdt_domain *d) +{ + cpumask_var_t cpu_with_psl; + struct rdt_resource *r; + struct rdt_domain *d_i; + bool ret = false; + + if (!zalloc_cpumask_var(&cpu_with_psl, GFP_KERNEL)) + return true; + + /* + * First determine which cpus have pseudo-locked regions + * associated with them. + */ + for_each_alloc_enabled_rdt_resource(r) { + list_for_each_entry(d_i, &r->domains, list) { + if (d_i->plr) + cpumask_or(cpu_with_psl, cpu_with_psl, + &d_i->cpu_mask); + } + } + + /* + * Next test if new pseudo-locked region would intersect with + * existing region. + */ + if (cpumask_intersects(&d->cpu_mask, cpu_with_psl)) + ret = true; + + free_cpumask_var(cpu_with_psl); + return ret; +} + +/** + * measure_cycles_lat_fn - Measure cycle latency to read pseudo-locked memory + * @_plr: pseudo-lock region to measure + * + * There is no deterministic way to test if a memory region is cached. One + * way is to measure how long it takes to read the memory, the speed of + * access is a good way to learn how close to the cpu the data was. Even + * more, if the prefetcher is disabled and the memory is read at a stride + * of half the cache line, then a cache miss will be easy to spot since the + * read of the first half would be significantly slower than the read of + * the second half. + * + * Return: 0. Waiter on waitqueue will be woken on completion. + */ +static int measure_cycles_lat_fn(void *_plr) +{ + struct pseudo_lock_region *plr = _plr; + u32 saved_low, saved_high; + unsigned long i; + u64 start, end; + void *mem_r; + + local_irq_disable(); + /* + * Disable hardware prefetchers. + */ + rdmsr(MSR_MISC_FEATURE_CONTROL, saved_low, saved_high); + wrmsr(MSR_MISC_FEATURE_CONTROL, prefetch_disable_bits, 0x0); + mem_r = READ_ONCE(plr->kmem); + /* + * Dummy execute of the time measurement to load the needed + * instructions into the L1 instruction cache. + */ + start = rdtsc_ordered(); + for (i = 0; i < plr->size; i += 32) { + start = rdtsc_ordered(); + asm volatile("mov (%0,%1,1), %%eax\n\t" + : + : "r" (mem_r), "r" (i) + : "%eax", "memory"); + end = rdtsc_ordered(); + trace_pseudo_lock_mem_latency((u32)(end - start)); + } + wrmsr(MSR_MISC_FEATURE_CONTROL, saved_low, saved_high); + local_irq_enable(); + plr->thread_done = 1; + wake_up_interruptible(&plr->lock_thread_wq); + return 0; +} + +/* + * Create a perf_event_attr for the hit and miss perf events that will + * be used during the performance measurement. A perf_event maintains + * a pointer to its perf_event_attr so a unique attribute structure is + * created for each perf_event. + * + * The actual configuration of the event is set right before use in order + * to use the X86_CONFIG macro. + */ +static struct perf_event_attr perf_miss_attr = { + .type = PERF_TYPE_RAW, + .size = sizeof(struct perf_event_attr), + .pinned = 1, + .disabled = 0, + .exclude_user = 1, +}; + +static struct perf_event_attr perf_hit_attr = { + .type = PERF_TYPE_RAW, + .size = sizeof(struct perf_event_attr), + .pinned = 1, + .disabled = 0, + .exclude_user = 1, +}; + +struct residency_counts { + u64 miss_before, hits_before; + u64 miss_after, hits_after; +}; + +static int measure_residency_fn(struct perf_event_attr *miss_attr, + struct perf_event_attr *hit_attr, + struct pseudo_lock_region *plr, + struct residency_counts *counts) +{ + u64 hits_before = 0, hits_after = 0, miss_before = 0, miss_after = 0; + struct perf_event *miss_event, *hit_event; + int hit_pmcnum, miss_pmcnum; + u32 saved_low, saved_high; + unsigned int line_size; + unsigned int size; + unsigned long i; + void *mem_r; + u64 tmp; + + miss_event = perf_event_create_kernel_counter(miss_attr, plr->cpu, + NULL, NULL, NULL); + if (IS_ERR(miss_event)) + goto out; + + hit_event = perf_event_create_kernel_counter(hit_attr, plr->cpu, + NULL, NULL, NULL); + if (IS_ERR(hit_event)) + goto out_miss; + + local_irq_disable(); + /* + * Check any possible error state of events used by performing + * one local read. + */ + if (perf_event_read_local(miss_event, &tmp, NULL, NULL)) { + local_irq_enable(); + goto out_hit; + } + if (perf_event_read_local(hit_event, &tmp, NULL, NULL)) { + local_irq_enable(); + goto out_hit; + } + + /* + * Disable hardware prefetchers. + */ + rdmsr(MSR_MISC_FEATURE_CONTROL, saved_low, saved_high); + wrmsr(MSR_MISC_FEATURE_CONTROL, prefetch_disable_bits, 0x0); + + /* Initialize rest of local variables */ + /* + * Performance event has been validated right before this with + * interrupts disabled - it is thus safe to read the counter index. + */ + miss_pmcnum = x86_perf_rdpmc_index(miss_event); + hit_pmcnum = x86_perf_rdpmc_index(hit_event); + line_size = READ_ONCE(plr->line_size); + mem_r = READ_ONCE(plr->kmem); + size = READ_ONCE(plr->size); + + /* + * Read counter variables twice - first to load the instructions + * used in L1 cache, second to capture accurate value that does not + * include cache misses incurred because of instruction loads. + */ + rdpmcl(hit_pmcnum, hits_before); + rdpmcl(miss_pmcnum, miss_before); + /* + * From SDM: Performing back-to-back fast reads are not guaranteed + * to be monotonic. + * Use LFENCE to ensure all previous instructions are retired + * before proceeding. + */ + rmb(); + rdpmcl(hit_pmcnum, hits_before); + rdpmcl(miss_pmcnum, miss_before); + /* + * Use LFENCE to ensure all previous instructions are retired + * before proceeding. + */ + rmb(); + for (i = 0; i < size; i += line_size) { + /* + * Add a barrier to prevent speculative execution of this + * loop reading beyond the end of the buffer. + */ + rmb(); + asm volatile("mov (%0,%1,1), %%eax\n\t" + : + : "r" (mem_r), "r" (i) + : "%eax", "memory"); + } + /* + * Use LFENCE to ensure all previous instructions are retired + * before proceeding. + */ + rmb(); + rdpmcl(hit_pmcnum, hits_after); + rdpmcl(miss_pmcnum, miss_after); + /* + * Use LFENCE to ensure all previous instructions are retired + * before proceeding. + */ + rmb(); + /* Re-enable hardware prefetchers */ + wrmsr(MSR_MISC_FEATURE_CONTROL, saved_low, saved_high); + local_irq_enable(); +out_hit: + perf_event_release_kernel(hit_event); +out_miss: + perf_event_release_kernel(miss_event); +out: + /* + * All counts will be zero on failure. + */ + counts->miss_before = miss_before; + counts->hits_before = hits_before; + counts->miss_after = miss_after; + counts->hits_after = hits_after; + return 0; +} + +static int measure_l2_residency(void *_plr) +{ + struct pseudo_lock_region *plr = _plr; + struct residency_counts counts = {0}; + + /* + * Non-architectural event for the Goldmont Microarchitecture + * from Intel x86 Architecture Software Developer Manual (SDM): + * MEM_LOAD_UOPS_RETIRED D1H (event number) + * Umask values: + * L2_HIT 02H + * L2_MISS 10H + */ + switch (boot_cpu_data.x86_model) { + case INTEL_FAM6_ATOM_GOLDMONT: + case INTEL_FAM6_ATOM_GOLDMONT_PLUS: + perf_miss_attr.config = X86_CONFIG(.event = 0xd1, + .umask = 0x10); + perf_hit_attr.config = X86_CONFIG(.event = 0xd1, + .umask = 0x2); + break; + default: + goto out; + } + + measure_residency_fn(&perf_miss_attr, &perf_hit_attr, plr, &counts); + /* + * If a failure prevented the measurements from succeeding + * tracepoints will still be written and all counts will be zero. + */ + trace_pseudo_lock_l2(counts.hits_after - counts.hits_before, + counts.miss_after - counts.miss_before); +out: + plr->thread_done = 1; + wake_up_interruptible(&plr->lock_thread_wq); + return 0; +} + +static int measure_l3_residency(void *_plr) +{ + struct pseudo_lock_region *plr = _plr; + struct residency_counts counts = {0}; + + /* + * On Broadwell Microarchitecture the MEM_LOAD_UOPS_RETIRED event + * has two "no fix" errata associated with it: BDM35 and BDM100. On + * this platform the following events are used instead: + * LONGEST_LAT_CACHE 2EH (Documented in SDM) + * REFERENCE 4FH + * MISS 41H + */ + + switch (boot_cpu_data.x86_model) { + case INTEL_FAM6_BROADWELL_X: + /* On BDW the hit event counts references, not hits */ + perf_hit_attr.config = X86_CONFIG(.event = 0x2e, + .umask = 0x4f); + perf_miss_attr.config = X86_CONFIG(.event = 0x2e, + .umask = 0x41); + break; + default: + goto out; + } + + measure_residency_fn(&perf_miss_attr, &perf_hit_attr, plr, &counts); + /* + * If a failure prevented the measurements from succeeding + * tracepoints will still be written and all counts will be zero. + */ + + counts.miss_after -= counts.miss_before; + if (boot_cpu_data.x86_model == INTEL_FAM6_BROADWELL_X) { + /* + * On BDW references and misses are counted, need to adjust. + * Sometimes the "hits" counter is a bit more than the + * references, for example, x references but x + 1 hits. + * To not report invalid hit values in this case we treat + * that as misses equal to references. + */ + /* First compute the number of cache references measured */ + counts.hits_after -= counts.hits_before; + /* Next convert references to cache hits */ + counts.hits_after -= min(counts.miss_after, counts.hits_after); + } else { + counts.hits_after -= counts.hits_before; + } + + trace_pseudo_lock_l3(counts.hits_after, counts.miss_after); +out: + plr->thread_done = 1; + wake_up_interruptible(&plr->lock_thread_wq); + return 0; +} + +/** + * pseudo_lock_measure_cycles - Trigger latency measure to pseudo-locked region + * + * The measurement of latency to access a pseudo-locked region should be + * done from a cpu that is associated with that pseudo-locked region. + * Determine which cpu is associated with this region and start a thread on + * that cpu to perform the measurement, wait for that thread to complete. + * + * Return: 0 on success, <0 on failure + */ +static int pseudo_lock_measure_cycles(struct rdtgroup *rdtgrp, int sel) +{ + struct pseudo_lock_region *plr = rdtgrp->plr; + struct task_struct *thread; + unsigned int cpu; + int ret = -1; + + cpus_read_lock(); + mutex_lock(&rdtgroup_mutex); + + if (rdtgrp->flags & RDT_DELETED) { + ret = -ENODEV; + goto out; + } + + if (!plr->d) { + ret = -ENODEV; + goto out; + } + + plr->thread_done = 0; + cpu = cpumask_first(&plr->d->cpu_mask); + if (!cpu_online(cpu)) { + ret = -ENODEV; + goto out; + } + + plr->cpu = cpu; + + if (sel == 1) + thread = kthread_create_on_node(measure_cycles_lat_fn, plr, + cpu_to_node(cpu), + "pseudo_lock_measure/%u", + cpu); + else if (sel == 2) + thread = kthread_create_on_node(measure_l2_residency, plr, + cpu_to_node(cpu), + "pseudo_lock_measure/%u", + cpu); + else if (sel == 3) + thread = kthread_create_on_node(measure_l3_residency, plr, + cpu_to_node(cpu), + "pseudo_lock_measure/%u", + cpu); + else + goto out; + + if (IS_ERR(thread)) { + ret = PTR_ERR(thread); + goto out; + } + kthread_bind(thread, cpu); + wake_up_process(thread); + + ret = wait_event_interruptible(plr->lock_thread_wq, + plr->thread_done == 1); + if (ret < 0) + goto out; + + ret = 0; + +out: + mutex_unlock(&rdtgroup_mutex); + cpus_read_unlock(); + return ret; +} + +static ssize_t pseudo_lock_measure_trigger(struct file *file, + const char __user *user_buf, + size_t count, loff_t *ppos) +{ + struct rdtgroup *rdtgrp = file->private_data; + size_t buf_size; + char buf[32]; + int ret; + int sel; + + buf_size = min(count, (sizeof(buf) - 1)); + if (copy_from_user(buf, user_buf, buf_size)) + return -EFAULT; + + buf[buf_size] = '\0'; + ret = kstrtoint(buf, 10, &sel); + if (ret == 0) { + if (sel != 1 && sel != 2 && sel != 3) + return -EINVAL; + ret = debugfs_file_get(file->f_path.dentry); + if (ret) + return ret; + ret = pseudo_lock_measure_cycles(rdtgrp, sel); + if (ret == 0) + ret = count; + debugfs_file_put(file->f_path.dentry); + } + + return ret; +} + +static const struct file_operations pseudo_measure_fops = { + .write = pseudo_lock_measure_trigger, + .open = simple_open, + .llseek = default_llseek, +}; + +/** + * rdtgroup_pseudo_lock_create - Create a pseudo-locked region + * @rdtgrp: resource group to which pseudo-lock region belongs + * + * Called when a resource group in the pseudo-locksetup mode receives a + * valid schemata that should be pseudo-locked. Since the resource group is + * in pseudo-locksetup mode the &struct pseudo_lock_region has already been + * allocated and initialized with the essential information. If a failure + * occurs the resource group remains in the pseudo-locksetup mode with the + * &struct pseudo_lock_region associated with it, but cleared from all + * information and ready for the user to re-attempt pseudo-locking by + * writing the schemata again. + * + * Return: 0 if the pseudo-locked region was successfully pseudo-locked, <0 + * on failure. Descriptive error will be written to last_cmd_status buffer. + */ +int rdtgroup_pseudo_lock_create(struct rdtgroup *rdtgrp) +{ + struct pseudo_lock_region *plr = rdtgrp->plr; + struct task_struct *thread; + unsigned int new_minor; + struct device *dev; + int ret; + + ret = pseudo_lock_region_alloc(plr); + if (ret < 0) + return ret; + + ret = pseudo_lock_cstates_constrain(plr); + if (ret < 0) { + ret = -EINVAL; + goto out_region; + } + + plr->thread_done = 0; + + thread = kthread_create_on_node(pseudo_lock_fn, rdtgrp, + cpu_to_node(plr->cpu), + "pseudo_lock/%u", plr->cpu); + if (IS_ERR(thread)) { + ret = PTR_ERR(thread); + rdt_last_cmd_printf("Locking thread returned error %d\n", ret); + goto out_cstates; + } + + kthread_bind(thread, plr->cpu); + wake_up_process(thread); + + ret = wait_event_interruptible(plr->lock_thread_wq, + plr->thread_done == 1); + if (ret < 0) { + /* + * If the thread does not get on the CPU for whatever + * reason and the process which sets up the region is + * interrupted then this will leave the thread in runnable + * state and once it gets on the CPU it will derefence + * the cleared, but not freed, plr struct resulting in an + * empty pseudo-locking loop. + */ + rdt_last_cmd_puts("Locking thread interrupted\n"); + goto out_cstates; + } + + ret = pseudo_lock_minor_get(&new_minor); + if (ret < 0) { + rdt_last_cmd_puts("Unable to obtain a new minor number\n"); + goto out_cstates; + } + + /* + * Unlock access but do not release the reference. The + * pseudo-locked region will still be here on return. + * + * The mutex has to be released temporarily to avoid a potential + * deadlock with the mm->mmap_lock which is obtained in the + * device_create() and debugfs_create_dir() callpath below as well as + * before the mmap() callback is called. + */ + mutex_unlock(&rdtgroup_mutex); + + if (!IS_ERR_OR_NULL(debugfs_resctrl)) { + plr->debugfs_dir = debugfs_create_dir(rdtgrp->kn->name, + debugfs_resctrl); + if (!IS_ERR_OR_NULL(plr->debugfs_dir)) + debugfs_create_file("pseudo_lock_measure", 0200, + plr->debugfs_dir, rdtgrp, + &pseudo_measure_fops); + } + + dev = device_create(pseudo_lock_class, NULL, + MKDEV(pseudo_lock_major, new_minor), + rdtgrp, "%s", rdtgrp->kn->name); + + mutex_lock(&rdtgroup_mutex); + + if (IS_ERR(dev)) { + ret = PTR_ERR(dev); + rdt_last_cmd_printf("Failed to create character device: %d\n", + ret); + goto out_debugfs; + } + + /* We released the mutex - check if group was removed while we did so */ + if (rdtgrp->flags & RDT_DELETED) { + ret = -ENODEV; + goto out_device; + } + + plr->minor = new_minor; + + rdtgrp->mode = RDT_MODE_PSEUDO_LOCKED; + closid_free(rdtgrp->closid); + rdtgroup_kn_mode_restore(rdtgrp, "cpus", 0444); + rdtgroup_kn_mode_restore(rdtgrp, "cpus_list", 0444); + + ret = 0; + goto out; + +out_device: + device_destroy(pseudo_lock_class, MKDEV(pseudo_lock_major, new_minor)); +out_debugfs: + debugfs_remove_recursive(plr->debugfs_dir); + pseudo_lock_minor_release(new_minor); +out_cstates: + pseudo_lock_cstates_relax(plr); +out_region: + pseudo_lock_region_clear(plr); +out: + return ret; +} + +/** + * rdtgroup_pseudo_lock_remove - Remove a pseudo-locked region + * @rdtgrp: resource group to which the pseudo-locked region belongs + * + * The removal of a pseudo-locked region can be initiated when the resource + * group is removed from user space via a "rmdir" from userspace or the + * unmount of the resctrl filesystem. On removal the resource group does + * not go back to pseudo-locksetup mode before it is removed, instead it is + * removed directly. There is thus assymmetry with the creation where the + * &struct pseudo_lock_region is removed here while it was not created in + * rdtgroup_pseudo_lock_create(). + * + * Return: void + */ +void rdtgroup_pseudo_lock_remove(struct rdtgroup *rdtgrp) +{ + struct pseudo_lock_region *plr = rdtgrp->plr; + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + /* + * Default group cannot be a pseudo-locked region so we can + * free closid here. + */ + closid_free(rdtgrp->closid); + goto free; + } + + pseudo_lock_cstates_relax(plr); + debugfs_remove_recursive(rdtgrp->plr->debugfs_dir); + device_destroy(pseudo_lock_class, MKDEV(pseudo_lock_major, plr->minor)); + pseudo_lock_minor_release(plr->minor); + +free: + pseudo_lock_free(rdtgrp); +} + +static int pseudo_lock_dev_open(struct inode *inode, struct file *filp) +{ + struct rdtgroup *rdtgrp; + + mutex_lock(&rdtgroup_mutex); + + rdtgrp = region_find_by_minor(iminor(inode)); + if (!rdtgrp) { + mutex_unlock(&rdtgroup_mutex); + return -ENODEV; + } + + filp->private_data = rdtgrp; + atomic_inc(&rdtgrp->waitcount); + /* Perform a non-seekable open - llseek is not supported */ + filp->f_mode &= ~(FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE); + + mutex_unlock(&rdtgroup_mutex); + + return 0; +} + +static int pseudo_lock_dev_release(struct inode *inode, struct file *filp) +{ + struct rdtgroup *rdtgrp; + + mutex_lock(&rdtgroup_mutex); + rdtgrp = filp->private_data; + WARN_ON(!rdtgrp); + if (!rdtgrp) { + mutex_unlock(&rdtgroup_mutex); + return -ENODEV; + } + filp->private_data = NULL; + atomic_dec(&rdtgrp->waitcount); + mutex_unlock(&rdtgroup_mutex); + return 0; +} + +static int pseudo_lock_dev_mremap(struct vm_area_struct *area) +{ + /* Not supported */ + return -EINVAL; +} + +static const struct vm_operations_struct pseudo_mmap_ops = { + .mremap = pseudo_lock_dev_mremap, +}; + +static int pseudo_lock_dev_mmap(struct file *filp, struct vm_area_struct *vma) +{ + unsigned long vsize = vma->vm_end - vma->vm_start; + unsigned long off = vma->vm_pgoff << PAGE_SHIFT; + struct pseudo_lock_region *plr; + struct rdtgroup *rdtgrp; + unsigned long physical; + unsigned long psize; + + mutex_lock(&rdtgroup_mutex); + + rdtgrp = filp->private_data; + WARN_ON(!rdtgrp); + if (!rdtgrp) { + mutex_unlock(&rdtgroup_mutex); + return -ENODEV; + } + + plr = rdtgrp->plr; + + if (!plr->d) { + mutex_unlock(&rdtgroup_mutex); + return -ENODEV; + } + + /* + * Task is required to run with affinity to the cpus associated + * with the pseudo-locked region. If this is not the case the task + * may be scheduled elsewhere and invalidate entries in the + * pseudo-locked region. + */ + if (!cpumask_subset(current->cpus_ptr, &plr->d->cpu_mask)) { + mutex_unlock(&rdtgroup_mutex); + return -EINVAL; + } + + physical = __pa(plr->kmem) >> PAGE_SHIFT; + psize = plr->size - off; + + if (off > plr->size) { + mutex_unlock(&rdtgroup_mutex); + return -ENOSPC; + } + + /* + * Ensure changes are carried directly to the memory being mapped, + * do not allow copy-on-write mapping. + */ + if (!(vma->vm_flags & VM_SHARED)) { + mutex_unlock(&rdtgroup_mutex); + return -EINVAL; + } + + if (vsize > psize) { + mutex_unlock(&rdtgroup_mutex); + return -ENOSPC; + } + + memset(plr->kmem + off, 0, vsize); + + if (remap_pfn_range(vma, vma->vm_start, physical + vma->vm_pgoff, + vsize, vma->vm_page_prot)) { + mutex_unlock(&rdtgroup_mutex); + return -EAGAIN; + } + vma->vm_ops = &pseudo_mmap_ops; + mutex_unlock(&rdtgroup_mutex); + return 0; +} + +static const struct file_operations pseudo_lock_dev_fops = { + .owner = THIS_MODULE, + .llseek = no_llseek, + .read = NULL, + .write = NULL, + .open = pseudo_lock_dev_open, + .release = pseudo_lock_dev_release, + .mmap = pseudo_lock_dev_mmap, +}; + +static char *pseudo_lock_devnode(struct device *dev, umode_t *mode) +{ + struct rdtgroup *rdtgrp; + + rdtgrp = dev_get_drvdata(dev); + if (mode) + *mode = 0600; + return kasprintf(GFP_KERNEL, "pseudo_lock/%s", rdtgrp->kn->name); +} + +int rdt_pseudo_lock_init(void) +{ + int ret; + + ret = register_chrdev(0, "pseudo_lock", &pseudo_lock_dev_fops); + if (ret < 0) + return ret; + + pseudo_lock_major = ret; + + pseudo_lock_class = class_create(THIS_MODULE, "pseudo_lock"); + if (IS_ERR(pseudo_lock_class)) { + ret = PTR_ERR(pseudo_lock_class); + unregister_chrdev(pseudo_lock_major, "pseudo_lock"); + return ret; + } + + pseudo_lock_class->devnode = pseudo_lock_devnode; + return 0; +} + +void rdt_pseudo_lock_release(void) +{ + class_destroy(pseudo_lock_class); + pseudo_lock_class = NULL; + unregister_chrdev(pseudo_lock_major, "pseudo_lock"); + pseudo_lock_major = 0; +} diff --git a/arch/x86/kernel/cpu/resctrl/pseudo_lock_event.h b/arch/x86/kernel/cpu/resctrl/pseudo_lock_event.h new file mode 100644 index 000000000..428ebbd42 --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/pseudo_lock_event.h @@ -0,0 +1,43 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM resctrl + +#if !defined(_TRACE_PSEUDO_LOCK_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_PSEUDO_LOCK_H + +#include <linux/tracepoint.h> + +TRACE_EVENT(pseudo_lock_mem_latency, + TP_PROTO(u32 latency), + TP_ARGS(latency), + TP_STRUCT__entry(__field(u32, latency)), + TP_fast_assign(__entry->latency = latency), + TP_printk("latency=%u", __entry->latency) + ); + +TRACE_EVENT(pseudo_lock_l2, + TP_PROTO(u64 l2_hits, u64 l2_miss), + TP_ARGS(l2_hits, l2_miss), + TP_STRUCT__entry(__field(u64, l2_hits) + __field(u64, l2_miss)), + TP_fast_assign(__entry->l2_hits = l2_hits; + __entry->l2_miss = l2_miss;), + TP_printk("hits=%llu miss=%llu", + __entry->l2_hits, __entry->l2_miss)); + +TRACE_EVENT(pseudo_lock_l3, + TP_PROTO(u64 l3_hits, u64 l3_miss), + TP_ARGS(l3_hits, l3_miss), + TP_STRUCT__entry(__field(u64, l3_hits) + __field(u64, l3_miss)), + TP_fast_assign(__entry->l3_hits = l3_hits; + __entry->l3_miss = l3_miss;), + TP_printk("hits=%llu miss=%llu", + __entry->l3_hits, __entry->l3_miss)); + +#endif /* _TRACE_PSEUDO_LOCK_H */ + +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE pseudo_lock_event +#include <trace/define_trace.h> diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kernel/cpu/resctrl/rdtgroup.c new file mode 100644 index 000000000..1e73b6fae --- /dev/null +++ b/arch/x86/kernel/cpu/resctrl/rdtgroup.c @@ -0,0 +1,3263 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * User interface for Resource Alloction in Resource Director Technology(RDT) + * + * Copyright (C) 2016 Intel Corporation + * + * Author: Fenghua Yu <fenghua.yu@intel.com> + * + * More information about RDT be found in the Intel (R) x86 Architecture + * Software Developer Manual. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cacheinfo.h> +#include <linux/cpu.h> +#include <linux/debugfs.h> +#include <linux/fs.h> +#include <linux/fs_parser.h> +#include <linux/sysfs.h> +#include <linux/kernfs.h> +#include <linux/seq_buf.h> +#include <linux/seq_file.h> +#include <linux/sched/signal.h> +#include <linux/sched/task.h> +#include <linux/slab.h> +#include <linux/task_work.h> +#include <linux/user_namespace.h> + +#include <uapi/linux/magic.h> + +#include <asm/resctrl.h> +#include "internal.h" + +DEFINE_STATIC_KEY_FALSE(rdt_enable_key); +DEFINE_STATIC_KEY_FALSE(rdt_mon_enable_key); +DEFINE_STATIC_KEY_FALSE(rdt_alloc_enable_key); +static struct kernfs_root *rdt_root; +struct rdtgroup rdtgroup_default; +LIST_HEAD(rdt_all_groups); + +/* Kernel fs node for "info" directory under root */ +static struct kernfs_node *kn_info; + +/* Kernel fs node for "mon_groups" directory under root */ +static struct kernfs_node *kn_mongrp; + +/* Kernel fs node for "mon_data" directory under root */ +static struct kernfs_node *kn_mondata; + +static struct seq_buf last_cmd_status; +static char last_cmd_status_buf[512]; + +struct dentry *debugfs_resctrl; + +void rdt_last_cmd_clear(void) +{ + lockdep_assert_held(&rdtgroup_mutex); + seq_buf_clear(&last_cmd_status); +} + +void rdt_last_cmd_puts(const char *s) +{ + lockdep_assert_held(&rdtgroup_mutex); + seq_buf_puts(&last_cmd_status, s); +} + +void rdt_last_cmd_printf(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + lockdep_assert_held(&rdtgroup_mutex); + seq_buf_vprintf(&last_cmd_status, fmt, ap); + va_end(ap); +} + +/* + * Trivial allocator for CLOSIDs. Since h/w only supports a small number, + * we can keep a bitmap of free CLOSIDs in a single integer. + * + * Using a global CLOSID across all resources has some advantages and + * some drawbacks: + * + We can simply set "current->closid" to assign a task to a resource + * group. + * + Context switch code can avoid extra memory references deciding which + * CLOSID to load into the PQR_ASSOC MSR + * - We give up some options in configuring resource groups across multi-socket + * systems. + * - Our choices on how to configure each resource become progressively more + * limited as the number of resources grows. + */ +static int closid_free_map; +static int closid_free_map_len; + +int closids_supported(void) +{ + return closid_free_map_len; +} + +static void closid_init(void) +{ + struct rdt_resource *r; + int rdt_min_closid = 32; + + /* Compute rdt_min_closid across all resources */ + for_each_alloc_enabled_rdt_resource(r) + rdt_min_closid = min(rdt_min_closid, r->num_closid); + + closid_free_map = BIT_MASK(rdt_min_closid) - 1; + + /* CLOSID 0 is always reserved for the default group */ + closid_free_map &= ~1; + closid_free_map_len = rdt_min_closid; +} + +static int closid_alloc(void) +{ + u32 closid = ffs(closid_free_map); + + if (closid == 0) + return -ENOSPC; + closid--; + closid_free_map &= ~(1 << closid); + + return closid; +} + +void closid_free(int closid) +{ + closid_free_map |= 1 << closid; +} + +/** + * closid_allocated - test if provided closid is in use + * @closid: closid to be tested + * + * Return: true if @closid is currently associated with a resource group, + * false if @closid is free + */ +static bool closid_allocated(unsigned int closid) +{ + return (closid_free_map & (1 << closid)) == 0; +} + +/** + * rdtgroup_mode_by_closid - Return mode of resource group with closid + * @closid: closid if the resource group + * + * Each resource group is associated with a @closid. Here the mode + * of a resource group can be queried by searching for it using its closid. + * + * Return: mode as &enum rdtgrp_mode of resource group with closid @closid + */ +enum rdtgrp_mode rdtgroup_mode_by_closid(int closid) +{ + struct rdtgroup *rdtgrp; + + list_for_each_entry(rdtgrp, &rdt_all_groups, rdtgroup_list) { + if (rdtgrp->closid == closid) + return rdtgrp->mode; + } + + return RDT_NUM_MODES; +} + +static const char * const rdt_mode_str[] = { + [RDT_MODE_SHAREABLE] = "shareable", + [RDT_MODE_EXCLUSIVE] = "exclusive", + [RDT_MODE_PSEUDO_LOCKSETUP] = "pseudo-locksetup", + [RDT_MODE_PSEUDO_LOCKED] = "pseudo-locked", +}; + +/** + * rdtgroup_mode_str - Return the string representation of mode + * @mode: the resource group mode as &enum rdtgroup_mode + * + * Return: string representation of valid mode, "unknown" otherwise + */ +static const char *rdtgroup_mode_str(enum rdtgrp_mode mode) +{ + if (mode < RDT_MODE_SHAREABLE || mode >= RDT_NUM_MODES) + return "unknown"; + + return rdt_mode_str[mode]; +} + +/* set uid and gid of rdtgroup dirs and files to that of the creator */ +static int rdtgroup_kn_set_ugid(struct kernfs_node *kn) +{ + struct iattr iattr = { .ia_valid = ATTR_UID | ATTR_GID, + .ia_uid = current_fsuid(), + .ia_gid = current_fsgid(), }; + + if (uid_eq(iattr.ia_uid, GLOBAL_ROOT_UID) && + gid_eq(iattr.ia_gid, GLOBAL_ROOT_GID)) + return 0; + + return kernfs_setattr(kn, &iattr); +} + +static int rdtgroup_add_file(struct kernfs_node *parent_kn, struct rftype *rft) +{ + struct kernfs_node *kn; + int ret; + + kn = __kernfs_create_file(parent_kn, rft->name, rft->mode, + GLOBAL_ROOT_UID, GLOBAL_ROOT_GID, + 0, rft->kf_ops, rft, NULL, NULL); + if (IS_ERR(kn)) + return PTR_ERR(kn); + + ret = rdtgroup_kn_set_ugid(kn); + if (ret) { + kernfs_remove(kn); + return ret; + } + + return 0; +} + +static int rdtgroup_seqfile_show(struct seq_file *m, void *arg) +{ + struct kernfs_open_file *of = m->private; + struct rftype *rft = of->kn->priv; + + if (rft->seq_show) + return rft->seq_show(of, m, arg); + return 0; +} + +static ssize_t rdtgroup_file_write(struct kernfs_open_file *of, char *buf, + size_t nbytes, loff_t off) +{ + struct rftype *rft = of->kn->priv; + + if (rft->write) + return rft->write(of, buf, nbytes, off); + + return -EINVAL; +} + +static struct kernfs_ops rdtgroup_kf_single_ops = { + .atomic_write_len = PAGE_SIZE, + .write = rdtgroup_file_write, + .seq_show = rdtgroup_seqfile_show, +}; + +static struct kernfs_ops kf_mondata_ops = { + .atomic_write_len = PAGE_SIZE, + .seq_show = rdtgroup_mondata_show, +}; + +static bool is_cpu_list(struct kernfs_open_file *of) +{ + struct rftype *rft = of->kn->priv; + + return rft->flags & RFTYPE_FLAGS_CPUS_LIST; +} + +static int rdtgroup_cpus_show(struct kernfs_open_file *of, + struct seq_file *s, void *v) +{ + struct rdtgroup *rdtgrp; + struct cpumask *mask; + int ret = 0; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + + if (rdtgrp) { + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) { + if (!rdtgrp->plr->d) { + rdt_last_cmd_clear(); + rdt_last_cmd_puts("Cache domain offline\n"); + ret = -ENODEV; + } else { + mask = &rdtgrp->plr->d->cpu_mask; + seq_printf(s, is_cpu_list(of) ? + "%*pbl\n" : "%*pb\n", + cpumask_pr_args(mask)); + } + } else { + seq_printf(s, is_cpu_list(of) ? "%*pbl\n" : "%*pb\n", + cpumask_pr_args(&rdtgrp->cpu_mask)); + } + } else { + ret = -ENOENT; + } + rdtgroup_kn_unlock(of->kn); + + return ret; +} + +/* + * This is safe against resctrl_sched_in() called from __switch_to() + * because __switch_to() is executed with interrupts disabled. A local call + * from update_closid_rmid() is proteced against __switch_to() because + * preemption is disabled. + */ +static void update_cpu_closid_rmid(void *info) +{ + struct rdtgroup *r = info; + + if (r) { + this_cpu_write(pqr_state.default_closid, r->closid); + this_cpu_write(pqr_state.default_rmid, r->mon.rmid); + } + + /* + * We cannot unconditionally write the MSR because the current + * executing task might have its own closid selected. Just reuse + * the context switch code. + */ + resctrl_sched_in(current); +} + +/* + * Update the PGR_ASSOC MSR on all cpus in @cpu_mask, + * + * Per task closids/rmids must have been set up before calling this function. + */ +static void +update_closid_rmid(const struct cpumask *cpu_mask, struct rdtgroup *r) +{ + int cpu = get_cpu(); + + if (cpumask_test_cpu(cpu, cpu_mask)) + update_cpu_closid_rmid(r); + smp_call_function_many(cpu_mask, update_cpu_closid_rmid, r, 1); + put_cpu(); +} + +static int cpus_mon_write(struct rdtgroup *rdtgrp, cpumask_var_t newmask, + cpumask_var_t tmpmask) +{ + struct rdtgroup *prgrp = rdtgrp->mon.parent, *crgrp; + struct list_head *head; + + /* Check whether cpus belong to parent ctrl group */ + cpumask_andnot(tmpmask, newmask, &prgrp->cpu_mask); + if (cpumask_weight(tmpmask)) { + rdt_last_cmd_puts("Can only add CPUs to mongroup that belong to parent\n"); + return -EINVAL; + } + + /* Check whether cpus are dropped from this group */ + cpumask_andnot(tmpmask, &rdtgrp->cpu_mask, newmask); + if (cpumask_weight(tmpmask)) { + /* Give any dropped cpus to parent rdtgroup */ + cpumask_or(&prgrp->cpu_mask, &prgrp->cpu_mask, tmpmask); + update_closid_rmid(tmpmask, prgrp); + } + + /* + * If we added cpus, remove them from previous group that owned them + * and update per-cpu rmid + */ + cpumask_andnot(tmpmask, newmask, &rdtgrp->cpu_mask); + if (cpumask_weight(tmpmask)) { + head = &prgrp->mon.crdtgrp_list; + list_for_each_entry(crgrp, head, mon.crdtgrp_list) { + if (crgrp == rdtgrp) + continue; + cpumask_andnot(&crgrp->cpu_mask, &crgrp->cpu_mask, + tmpmask); + } + update_closid_rmid(tmpmask, rdtgrp); + } + + /* Done pushing/pulling - update this group with new mask */ + cpumask_copy(&rdtgrp->cpu_mask, newmask); + + return 0; +} + +static void cpumask_rdtgrp_clear(struct rdtgroup *r, struct cpumask *m) +{ + struct rdtgroup *crgrp; + + cpumask_andnot(&r->cpu_mask, &r->cpu_mask, m); + /* update the child mon group masks as well*/ + list_for_each_entry(crgrp, &r->mon.crdtgrp_list, mon.crdtgrp_list) + cpumask_and(&crgrp->cpu_mask, &r->cpu_mask, &crgrp->cpu_mask); +} + +static int cpus_ctrl_write(struct rdtgroup *rdtgrp, cpumask_var_t newmask, + cpumask_var_t tmpmask, cpumask_var_t tmpmask1) +{ + struct rdtgroup *r, *crgrp; + struct list_head *head; + + /* Check whether cpus are dropped from this group */ + cpumask_andnot(tmpmask, &rdtgrp->cpu_mask, newmask); + if (cpumask_weight(tmpmask)) { + /* Can't drop from default group */ + if (rdtgrp == &rdtgroup_default) { + rdt_last_cmd_puts("Can't drop CPUs from default group\n"); + return -EINVAL; + } + + /* Give any dropped cpus to rdtgroup_default */ + cpumask_or(&rdtgroup_default.cpu_mask, + &rdtgroup_default.cpu_mask, tmpmask); + update_closid_rmid(tmpmask, &rdtgroup_default); + } + + /* + * If we added cpus, remove them from previous group and + * the prev group's child groups that owned them + * and update per-cpu closid/rmid. + */ + cpumask_andnot(tmpmask, newmask, &rdtgrp->cpu_mask); + if (cpumask_weight(tmpmask)) { + list_for_each_entry(r, &rdt_all_groups, rdtgroup_list) { + if (r == rdtgrp) + continue; + cpumask_and(tmpmask1, &r->cpu_mask, tmpmask); + if (cpumask_weight(tmpmask1)) + cpumask_rdtgrp_clear(r, tmpmask1); + } + update_closid_rmid(tmpmask, rdtgrp); + } + + /* Done pushing/pulling - update this group with new mask */ + cpumask_copy(&rdtgrp->cpu_mask, newmask); + + /* + * Clear child mon group masks since there is a new parent mask + * now and update the rmid for the cpus the child lost. + */ + head = &rdtgrp->mon.crdtgrp_list; + list_for_each_entry(crgrp, head, mon.crdtgrp_list) { + cpumask_and(tmpmask, &rdtgrp->cpu_mask, &crgrp->cpu_mask); + update_closid_rmid(tmpmask, rdtgrp); + cpumask_clear(&crgrp->cpu_mask); + } + + return 0; +} + +static ssize_t rdtgroup_cpus_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + cpumask_var_t tmpmask, newmask, tmpmask1; + struct rdtgroup *rdtgrp; + int ret; + + if (!buf) + return -EINVAL; + + if (!zalloc_cpumask_var(&tmpmask, GFP_KERNEL)) + return -ENOMEM; + if (!zalloc_cpumask_var(&newmask, GFP_KERNEL)) { + free_cpumask_var(tmpmask); + return -ENOMEM; + } + if (!zalloc_cpumask_var(&tmpmask1, GFP_KERNEL)) { + free_cpumask_var(tmpmask); + free_cpumask_var(newmask); + return -ENOMEM; + } + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + ret = -ENOENT; + goto unlock; + } + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + ret = -EINVAL; + rdt_last_cmd_puts("Pseudo-locking in progress\n"); + goto unlock; + } + + if (is_cpu_list(of)) + ret = cpulist_parse(buf, newmask); + else + ret = cpumask_parse(buf, newmask); + + if (ret) { + rdt_last_cmd_puts("Bad CPU list/mask\n"); + goto unlock; + } + + /* check that user didn't specify any offline cpus */ + cpumask_andnot(tmpmask, newmask, cpu_online_mask); + if (cpumask_weight(tmpmask)) { + ret = -EINVAL; + rdt_last_cmd_puts("Can only assign online CPUs\n"); + goto unlock; + } + + if (rdtgrp->type == RDTCTRL_GROUP) + ret = cpus_ctrl_write(rdtgrp, newmask, tmpmask, tmpmask1); + else if (rdtgrp->type == RDTMON_GROUP) + ret = cpus_mon_write(rdtgrp, newmask, tmpmask); + else + ret = -EINVAL; + +unlock: + rdtgroup_kn_unlock(of->kn); + free_cpumask_var(tmpmask); + free_cpumask_var(newmask); + free_cpumask_var(tmpmask1); + + return ret ?: nbytes; +} + +/** + * rdtgroup_remove - the helper to remove resource group safely + * @rdtgrp: resource group to remove + * + * On resource group creation via a mkdir, an extra kernfs_node reference is + * taken to ensure that the rdtgroup structure remains accessible for the + * rdtgroup_kn_unlock() calls where it is removed. + * + * Drop the extra reference here, then free the rdtgroup structure. + * + * Return: void + */ +static void rdtgroup_remove(struct rdtgroup *rdtgrp) +{ + kernfs_put(rdtgrp->kn); + kfree(rdtgrp); +} + +static void _update_task_closid_rmid(void *task) +{ + /* + * If the task is still current on this CPU, update PQR_ASSOC MSR. + * Otherwise, the MSR is updated when the task is scheduled in. + */ + if (task == current) + resctrl_sched_in(task); +} + +static void update_task_closid_rmid(struct task_struct *t) +{ + if (IS_ENABLED(CONFIG_SMP) && task_curr(t)) + smp_call_function_single(task_cpu(t), _update_task_closid_rmid, t, 1); + else + _update_task_closid_rmid(t); +} + +static int __rdtgroup_move_task(struct task_struct *tsk, + struct rdtgroup *rdtgrp) +{ + /* If the task is already in rdtgrp, no need to move the task. */ + if ((rdtgrp->type == RDTCTRL_GROUP && tsk->closid == rdtgrp->closid && + tsk->rmid == rdtgrp->mon.rmid) || + (rdtgrp->type == RDTMON_GROUP && tsk->rmid == rdtgrp->mon.rmid && + tsk->closid == rdtgrp->mon.parent->closid)) + return 0; + + /* + * Set the task's closid/rmid before the PQR_ASSOC MSR can be + * updated by them. + * + * For ctrl_mon groups, move both closid and rmid. + * For monitor groups, can move the tasks only from + * their parent CTRL group. + */ + + if (rdtgrp->type == RDTCTRL_GROUP) { + WRITE_ONCE(tsk->closid, rdtgrp->closid); + WRITE_ONCE(tsk->rmid, rdtgrp->mon.rmid); + } else if (rdtgrp->type == RDTMON_GROUP) { + if (rdtgrp->mon.parent->closid == tsk->closid) { + WRITE_ONCE(tsk->rmid, rdtgrp->mon.rmid); + } else { + rdt_last_cmd_puts("Can't move task to different control group\n"); + return -EINVAL; + } + } + + /* + * Ensure the task's closid and rmid are written before determining if + * the task is current that will decide if it will be interrupted. + * This pairs with the full barrier between the rq->curr update and + * resctrl_sched_in() during context switch. + */ + smp_mb(); + + /* + * By now, the task's closid and rmid are set. If the task is current + * on a CPU, the PQR_ASSOC MSR needs to be updated to make the resource + * group go into effect. If the task is not current, the MSR will be + * updated when the task is scheduled in. + */ + update_task_closid_rmid(tsk); + + return 0; +} + +static bool is_closid_match(struct task_struct *t, struct rdtgroup *r) +{ + return (rdt_alloc_capable && + (r->type == RDTCTRL_GROUP) && (t->closid == r->closid)); +} + +static bool is_rmid_match(struct task_struct *t, struct rdtgroup *r) +{ + return (rdt_mon_capable && + (r->type == RDTMON_GROUP) && (t->rmid == r->mon.rmid)); +} + +/** + * rdtgroup_tasks_assigned - Test if tasks have been assigned to resource group + * @r: Resource group + * + * Return: 1 if tasks have been assigned to @r, 0 otherwise + */ +int rdtgroup_tasks_assigned(struct rdtgroup *r) +{ + struct task_struct *p, *t; + int ret = 0; + + lockdep_assert_held(&rdtgroup_mutex); + + rcu_read_lock(); + for_each_process_thread(p, t) { + if (is_closid_match(t, r) || is_rmid_match(t, r)) { + ret = 1; + break; + } + } + rcu_read_unlock(); + + return ret; +} + +static int rdtgroup_task_write_permission(struct task_struct *task, + struct kernfs_open_file *of) +{ + const struct cred *tcred = get_task_cred(task); + const struct cred *cred = current_cred(); + int ret = 0; + + /* + * Even if we're attaching all tasks in the thread group, we only + * need to check permissions on one of them. + */ + if (!uid_eq(cred->euid, GLOBAL_ROOT_UID) && + !uid_eq(cred->euid, tcred->uid) && + !uid_eq(cred->euid, tcred->suid)) { + rdt_last_cmd_printf("No permission to move task %d\n", task->pid); + ret = -EPERM; + } + + put_cred(tcred); + return ret; +} + +static int rdtgroup_move_task(pid_t pid, struct rdtgroup *rdtgrp, + struct kernfs_open_file *of) +{ + struct task_struct *tsk; + int ret; + + rcu_read_lock(); + if (pid) { + tsk = find_task_by_vpid(pid); + if (!tsk) { + rcu_read_unlock(); + rdt_last_cmd_printf("No task %d\n", pid); + return -ESRCH; + } + } else { + tsk = current; + } + + get_task_struct(tsk); + rcu_read_unlock(); + + ret = rdtgroup_task_write_permission(tsk, of); + if (!ret) + ret = __rdtgroup_move_task(tsk, rdtgrp); + + put_task_struct(tsk); + return ret; +} + +static ssize_t rdtgroup_tasks_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct rdtgroup *rdtgrp; + int ret = 0; + pid_t pid; + + if (kstrtoint(strstrip(buf), 0, &pid) || pid < 0) + return -EINVAL; + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + rdtgroup_kn_unlock(of->kn); + return -ENOENT; + } + rdt_last_cmd_clear(); + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + ret = -EINVAL; + rdt_last_cmd_puts("Pseudo-locking in progress\n"); + goto unlock; + } + + ret = rdtgroup_move_task(pid, rdtgrp, of); + +unlock: + rdtgroup_kn_unlock(of->kn); + + return ret ?: nbytes; +} + +static void show_rdt_tasks(struct rdtgroup *r, struct seq_file *s) +{ + struct task_struct *p, *t; + pid_t pid; + + rcu_read_lock(); + for_each_process_thread(p, t) { + if (is_closid_match(t, r) || is_rmid_match(t, r)) { + pid = task_pid_vnr(t); + if (pid) + seq_printf(s, "%d\n", pid); + } + } + rcu_read_unlock(); +} + +static int rdtgroup_tasks_show(struct kernfs_open_file *of, + struct seq_file *s, void *v) +{ + struct rdtgroup *rdtgrp; + int ret = 0; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (rdtgrp) + show_rdt_tasks(rdtgrp, s); + else + ret = -ENOENT; + rdtgroup_kn_unlock(of->kn); + + return ret; +} + +#ifdef CONFIG_PROC_CPU_RESCTRL + +/* + * A task can only be part of one resctrl control group and of one monitor + * group which is associated to that control group. + * + * 1) res: + * mon: + * + * resctrl is not available. + * + * 2) res:/ + * mon: + * + * Task is part of the root resctrl control group, and it is not associated + * to any monitor group. + * + * 3) res:/ + * mon:mon0 + * + * Task is part of the root resctrl control group and monitor group mon0. + * + * 4) res:group0 + * mon: + * + * Task is part of resctrl control group group0, and it is not associated + * to any monitor group. + * + * 5) res:group0 + * mon:mon1 + * + * Task is part of resctrl control group group0 and monitor group mon1. + */ +int proc_resctrl_show(struct seq_file *s, struct pid_namespace *ns, + struct pid *pid, struct task_struct *tsk) +{ + struct rdtgroup *rdtg; + int ret = 0; + + mutex_lock(&rdtgroup_mutex); + + /* Return empty if resctrl has not been mounted. */ + if (!static_branch_unlikely(&rdt_enable_key)) { + seq_puts(s, "res:\nmon:\n"); + goto unlock; + } + + list_for_each_entry(rdtg, &rdt_all_groups, rdtgroup_list) { + struct rdtgroup *crg; + + /* + * Task information is only relevant for shareable + * and exclusive groups. + */ + if (rdtg->mode != RDT_MODE_SHAREABLE && + rdtg->mode != RDT_MODE_EXCLUSIVE) + continue; + + if (rdtg->closid != tsk->closid) + continue; + + seq_printf(s, "res:%s%s\n", (rdtg == &rdtgroup_default) ? "/" : "", + rdtg->kn->name); + seq_puts(s, "mon:"); + list_for_each_entry(crg, &rdtg->mon.crdtgrp_list, + mon.crdtgrp_list) { + if (tsk->rmid != crg->mon.rmid) + continue; + seq_printf(s, "%s", crg->kn->name); + break; + } + seq_putc(s, '\n'); + goto unlock; + } + /* + * The above search should succeed. Otherwise return + * with an error. + */ + ret = -ENOENT; +unlock: + mutex_unlock(&rdtgroup_mutex); + + return ret; +} +#endif + +static int rdt_last_cmd_status_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + int len; + + mutex_lock(&rdtgroup_mutex); + len = seq_buf_used(&last_cmd_status); + if (len) + seq_printf(seq, "%.*s", len, last_cmd_status_buf); + else + seq_puts(seq, "ok\n"); + mutex_unlock(&rdtgroup_mutex); + return 0; +} + +static int rdt_num_closids_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%d\n", r->num_closid); + return 0; +} + +static int rdt_default_ctrl_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%x\n", r->default_ctrl); + return 0; +} + +static int rdt_min_cbm_bits_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%u\n", r->cache.min_cbm_bits); + return 0; +} + +static int rdt_shareable_bits_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%x\n", r->cache.shareable_bits); + return 0; +} + +/** + * rdt_bit_usage_show - Display current usage of resources + * + * A domain is a shared resource that can now be allocated differently. Here + * we display the current regions of the domain as an annotated bitmask. + * For each domain of this resource its allocation bitmask + * is annotated as below to indicate the current usage of the corresponding bit: + * 0 - currently unused + * X - currently available for sharing and used by software and hardware + * H - currently used by hardware only but available for software use + * S - currently used and shareable by software only + * E - currently used exclusively by one resource group + * P - currently pseudo-locked by one resource group + */ +static int rdt_bit_usage_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + /* + * Use unsigned long even though only 32 bits are used to ensure + * test_bit() is used safely. + */ + unsigned long sw_shareable = 0, hw_shareable = 0; + unsigned long exclusive = 0, pseudo_locked = 0; + struct rdt_domain *dom; + int i, hwb, swb, excl, psl; + enum rdtgrp_mode mode; + bool sep = false; + u32 *ctrl; + + mutex_lock(&rdtgroup_mutex); + hw_shareable = r->cache.shareable_bits; + list_for_each_entry(dom, &r->domains, list) { + if (sep) + seq_putc(seq, ';'); + ctrl = dom->ctrl_val; + sw_shareable = 0; + exclusive = 0; + seq_printf(seq, "%d=", dom->id); + for (i = 0; i < closids_supported(); i++, ctrl++) { + if (!closid_allocated(i)) + continue; + mode = rdtgroup_mode_by_closid(i); + switch (mode) { + case RDT_MODE_SHAREABLE: + sw_shareable |= *ctrl; + break; + case RDT_MODE_EXCLUSIVE: + exclusive |= *ctrl; + break; + case RDT_MODE_PSEUDO_LOCKSETUP: + /* + * RDT_MODE_PSEUDO_LOCKSETUP is possible + * here but not included since the CBM + * associated with this CLOSID in this mode + * is not initialized and no task or cpu can be + * assigned this CLOSID. + */ + break; + case RDT_MODE_PSEUDO_LOCKED: + case RDT_NUM_MODES: + WARN(1, + "invalid mode for closid %d\n", i); + break; + } + } + for (i = r->cache.cbm_len - 1; i >= 0; i--) { + pseudo_locked = dom->plr ? dom->plr->cbm : 0; + hwb = test_bit(i, &hw_shareable); + swb = test_bit(i, &sw_shareable); + excl = test_bit(i, &exclusive); + psl = test_bit(i, &pseudo_locked); + if (hwb && swb) + seq_putc(seq, 'X'); + else if (hwb && !swb) + seq_putc(seq, 'H'); + else if (!hwb && swb) + seq_putc(seq, 'S'); + else if (excl) + seq_putc(seq, 'E'); + else if (psl) + seq_putc(seq, 'P'); + else /* Unused bits remain */ + seq_putc(seq, '0'); + } + sep = true; + } + seq_putc(seq, '\n'); + mutex_unlock(&rdtgroup_mutex); + return 0; +} + +static int rdt_min_bw_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%u\n", r->membw.min_bw); + return 0; +} + +static int rdt_num_rmids_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%d\n", r->num_rmid); + + return 0; +} + +static int rdt_mon_features_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + struct mon_evt *mevt; + + list_for_each_entry(mevt, &r->evt_list, list) + seq_printf(seq, "%s\n", mevt->name); + + return 0; +} + +static int rdt_bw_gran_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%u\n", r->membw.bw_gran); + return 0; +} + +static int rdt_delay_linear_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%u\n", r->membw.delay_linear); + return 0; +} + +static int max_threshold_occ_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + seq_printf(seq, "%u\n", resctrl_cqm_threshold * r->mon_scale); + + return 0; +} + +static int rdt_thread_throttle_mode_show(struct kernfs_open_file *of, + struct seq_file *seq, void *v) +{ + struct rdt_resource *r = of->kn->parent->priv; + + if (r->membw.throttle_mode == THREAD_THROTTLE_PER_THREAD) + seq_puts(seq, "per-thread\n"); + else + seq_puts(seq, "max\n"); + + return 0; +} + +static ssize_t max_threshold_occ_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct rdt_resource *r = of->kn->parent->priv; + unsigned int bytes; + int ret; + + ret = kstrtouint(buf, 0, &bytes); + if (ret) + return ret; + + if (bytes > (boot_cpu_data.x86_cache_size * 1024)) + return -EINVAL; + + resctrl_cqm_threshold = bytes / r->mon_scale; + + return nbytes; +} + +/* + * rdtgroup_mode_show - Display mode of this resource group + */ +static int rdtgroup_mode_show(struct kernfs_open_file *of, + struct seq_file *s, void *v) +{ + struct rdtgroup *rdtgrp; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + rdtgroup_kn_unlock(of->kn); + return -ENOENT; + } + + seq_printf(s, "%s\n", rdtgroup_mode_str(rdtgrp->mode)); + + rdtgroup_kn_unlock(of->kn); + return 0; +} + +/** + * rdt_cdp_peer_get - Retrieve CDP peer if it exists + * @r: RDT resource to which RDT domain @d belongs + * @d: Cache instance for which a CDP peer is requested + * @r_cdp: RDT resource that shares hardware with @r (RDT resource peer) + * Used to return the result. + * @d_cdp: RDT domain that shares hardware with @d (RDT domain peer) + * Used to return the result. + * + * RDT resources are managed independently and by extension the RDT domains + * (RDT resource instances) are managed independently also. The Code and + * Data Prioritization (CDP) RDT resources, while managed independently, + * could refer to the same underlying hardware. For example, + * RDT_RESOURCE_L2CODE and RDT_RESOURCE_L2DATA both refer to the L2 cache. + * + * When provided with an RDT resource @r and an instance of that RDT + * resource @d rdt_cdp_peer_get() will return if there is a peer RDT + * resource and the exact instance that shares the same hardware. + * + * Return: 0 if a CDP peer was found, <0 on error or if no CDP peer exists. + * If a CDP peer was found, @r_cdp will point to the peer RDT resource + * and @d_cdp will point to the peer RDT domain. + */ +static int rdt_cdp_peer_get(struct rdt_resource *r, struct rdt_domain *d, + struct rdt_resource **r_cdp, + struct rdt_domain **d_cdp) +{ + struct rdt_resource *_r_cdp = NULL; + struct rdt_domain *_d_cdp = NULL; + int ret = 0; + + switch (r->rid) { + case RDT_RESOURCE_L3DATA: + _r_cdp = &rdt_resources_all[RDT_RESOURCE_L3CODE]; + break; + case RDT_RESOURCE_L3CODE: + _r_cdp = &rdt_resources_all[RDT_RESOURCE_L3DATA]; + break; + case RDT_RESOURCE_L2DATA: + _r_cdp = &rdt_resources_all[RDT_RESOURCE_L2CODE]; + break; + case RDT_RESOURCE_L2CODE: + _r_cdp = &rdt_resources_all[RDT_RESOURCE_L2DATA]; + break; + default: + ret = -ENOENT; + goto out; + } + + /* + * When a new CPU comes online and CDP is enabled then the new + * RDT domains (if any) associated with both CDP RDT resources + * are added in the same CPU online routine while the + * rdtgroup_mutex is held. It should thus not happen for one + * RDT domain to exist and be associated with its RDT CDP + * resource but there is no RDT domain associated with the + * peer RDT CDP resource. Hence the WARN. + */ + _d_cdp = rdt_find_domain(_r_cdp, d->id, NULL); + if (WARN_ON(IS_ERR_OR_NULL(_d_cdp))) { + _r_cdp = NULL; + _d_cdp = NULL; + ret = -EINVAL; + } + +out: + *r_cdp = _r_cdp; + *d_cdp = _d_cdp; + + return ret; +} + +/** + * __rdtgroup_cbm_overlaps - Does CBM for intended closid overlap with other + * @r: Resource to which domain instance @d belongs. + * @d: The domain instance for which @closid is being tested. + * @cbm: Capacity bitmask being tested. + * @closid: Intended closid for @cbm. + * @exclusive: Only check if overlaps with exclusive resource groups + * + * Checks if provided @cbm intended to be used for @closid on domain + * @d overlaps with any other closids or other hardware usage associated + * with this domain. If @exclusive is true then only overlaps with + * resource groups in exclusive mode will be considered. If @exclusive + * is false then overlaps with any resource group or hardware entities + * will be considered. + * + * @cbm is unsigned long, even if only 32 bits are used, to make the + * bitmap functions work correctly. + * + * Return: false if CBM does not overlap, true if it does. + */ +static bool __rdtgroup_cbm_overlaps(struct rdt_resource *r, struct rdt_domain *d, + unsigned long cbm, int closid, bool exclusive) +{ + enum rdtgrp_mode mode; + unsigned long ctrl_b; + u32 *ctrl; + int i; + + /* Check for any overlap with regions used by hardware directly */ + if (!exclusive) { + ctrl_b = r->cache.shareable_bits; + if (bitmap_intersects(&cbm, &ctrl_b, r->cache.cbm_len)) + return true; + } + + /* Check for overlap with other resource groups */ + ctrl = d->ctrl_val; + for (i = 0; i < closids_supported(); i++, ctrl++) { + ctrl_b = *ctrl; + mode = rdtgroup_mode_by_closid(i); + if (closid_allocated(i) && i != closid && + mode != RDT_MODE_PSEUDO_LOCKSETUP) { + if (bitmap_intersects(&cbm, &ctrl_b, r->cache.cbm_len)) { + if (exclusive) { + if (mode == RDT_MODE_EXCLUSIVE) + return true; + continue; + } + return true; + } + } + } + + return false; +} + +/** + * rdtgroup_cbm_overlaps - Does CBM overlap with other use of hardware + * @r: Resource to which domain instance @d belongs. + * @d: The domain instance for which @closid is being tested. + * @cbm: Capacity bitmask being tested. + * @closid: Intended closid for @cbm. + * @exclusive: Only check if overlaps with exclusive resource groups + * + * Resources that can be allocated using a CBM can use the CBM to control + * the overlap of these allocations. rdtgroup_cmb_overlaps() is the test + * for overlap. Overlap test is not limited to the specific resource for + * which the CBM is intended though - when dealing with CDP resources that + * share the underlying hardware the overlap check should be performed on + * the CDP resource sharing the hardware also. + * + * Refer to description of __rdtgroup_cbm_overlaps() for the details of the + * overlap test. + * + * Return: true if CBM overlap detected, false if there is no overlap + */ +bool rdtgroup_cbm_overlaps(struct rdt_resource *r, struct rdt_domain *d, + unsigned long cbm, int closid, bool exclusive) +{ + struct rdt_resource *r_cdp; + struct rdt_domain *d_cdp; + + if (__rdtgroup_cbm_overlaps(r, d, cbm, closid, exclusive)) + return true; + + if (rdt_cdp_peer_get(r, d, &r_cdp, &d_cdp) < 0) + return false; + + return __rdtgroup_cbm_overlaps(r_cdp, d_cdp, cbm, closid, exclusive); +} + +/** + * rdtgroup_mode_test_exclusive - Test if this resource group can be exclusive + * + * An exclusive resource group implies that there should be no sharing of + * its allocated resources. At the time this group is considered to be + * exclusive this test can determine if its current schemata supports this + * setting by testing for overlap with all other resource groups. + * + * Return: true if resource group can be exclusive, false if there is overlap + * with allocations of other resource groups and thus this resource group + * cannot be exclusive. + */ +static bool rdtgroup_mode_test_exclusive(struct rdtgroup *rdtgrp) +{ + int closid = rdtgrp->closid; + struct rdt_resource *r; + bool has_cache = false; + struct rdt_domain *d; + + for_each_alloc_enabled_rdt_resource(r) { + if (r->rid == RDT_RESOURCE_MBA) + continue; + has_cache = true; + list_for_each_entry(d, &r->domains, list) { + if (rdtgroup_cbm_overlaps(r, d, d->ctrl_val[closid], + rdtgrp->closid, false)) { + rdt_last_cmd_puts("Schemata overlaps\n"); + return false; + } + } + } + + if (!has_cache) { + rdt_last_cmd_puts("Cannot be exclusive without CAT/CDP\n"); + return false; + } + + return true; +} + +/** + * rdtgroup_mode_write - Modify the resource group's mode + * + */ +static ssize_t rdtgroup_mode_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct rdtgroup *rdtgrp; + enum rdtgrp_mode mode; + int ret = 0; + + /* Valid input requires a trailing newline */ + if (nbytes == 0 || buf[nbytes - 1] != '\n') + return -EINVAL; + buf[nbytes - 1] = '\0'; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + rdtgroup_kn_unlock(of->kn); + return -ENOENT; + } + + rdt_last_cmd_clear(); + + mode = rdtgrp->mode; + + if ((!strcmp(buf, "shareable") && mode == RDT_MODE_SHAREABLE) || + (!strcmp(buf, "exclusive") && mode == RDT_MODE_EXCLUSIVE) || + (!strcmp(buf, "pseudo-locksetup") && + mode == RDT_MODE_PSEUDO_LOCKSETUP) || + (!strcmp(buf, "pseudo-locked") && mode == RDT_MODE_PSEUDO_LOCKED)) + goto out; + + if (mode == RDT_MODE_PSEUDO_LOCKED) { + rdt_last_cmd_puts("Cannot change pseudo-locked group\n"); + ret = -EINVAL; + goto out; + } + + if (!strcmp(buf, "shareable")) { + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + ret = rdtgroup_locksetup_exit(rdtgrp); + if (ret) + goto out; + } + rdtgrp->mode = RDT_MODE_SHAREABLE; + } else if (!strcmp(buf, "exclusive")) { + if (!rdtgroup_mode_test_exclusive(rdtgrp)) { + ret = -EINVAL; + goto out; + } + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + ret = rdtgroup_locksetup_exit(rdtgrp); + if (ret) + goto out; + } + rdtgrp->mode = RDT_MODE_EXCLUSIVE; + } else if (!strcmp(buf, "pseudo-locksetup")) { + ret = rdtgroup_locksetup_enter(rdtgrp); + if (ret) + goto out; + rdtgrp->mode = RDT_MODE_PSEUDO_LOCKSETUP; + } else { + rdt_last_cmd_puts("Unknown or unsupported mode\n"); + ret = -EINVAL; + } + +out: + rdtgroup_kn_unlock(of->kn); + return ret ?: nbytes; +} + +/** + * rdtgroup_cbm_to_size - Translate CBM to size in bytes + * @r: RDT resource to which @d belongs. + * @d: RDT domain instance. + * @cbm: bitmask for which the size should be computed. + * + * The bitmask provided associated with the RDT domain instance @d will be + * translated into how many bytes it represents. The size in bytes is + * computed by first dividing the total cache size by the CBM length to + * determine how many bytes each bit in the bitmask represents. The result + * is multiplied with the number of bits set in the bitmask. + * + * @cbm is unsigned long, even if only 32 bits are used to make the + * bitmap functions work correctly. + */ +unsigned int rdtgroup_cbm_to_size(struct rdt_resource *r, + struct rdt_domain *d, unsigned long cbm) +{ + struct cpu_cacheinfo *ci; + unsigned int size = 0; + int num_b, i; + + num_b = bitmap_weight(&cbm, r->cache.cbm_len); + ci = get_cpu_cacheinfo(cpumask_any(&d->cpu_mask)); + for (i = 0; i < ci->num_leaves; i++) { + if (ci->info_list[i].level == r->cache_level) { + size = ci->info_list[i].size / r->cache.cbm_len * num_b; + break; + } + } + + return size; +} + +/** + * rdtgroup_size_show - Display size in bytes of allocated regions + * + * The "size" file mirrors the layout of the "schemata" file, printing the + * size in bytes of each region instead of the capacity bitmask. + * + */ +static int rdtgroup_size_show(struct kernfs_open_file *of, + struct seq_file *s, void *v) +{ + struct rdtgroup *rdtgrp; + struct rdt_resource *r; + struct rdt_domain *d; + unsigned int size; + int ret = 0; + bool sep; + u32 ctrl; + + rdtgrp = rdtgroup_kn_lock_live(of->kn); + if (!rdtgrp) { + rdtgroup_kn_unlock(of->kn); + return -ENOENT; + } + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) { + if (!rdtgrp->plr->d) { + rdt_last_cmd_clear(); + rdt_last_cmd_puts("Cache domain offline\n"); + ret = -ENODEV; + } else { + seq_printf(s, "%*s:", max_name_width, + rdtgrp->plr->r->name); + size = rdtgroup_cbm_to_size(rdtgrp->plr->r, + rdtgrp->plr->d, + rdtgrp->plr->cbm); + seq_printf(s, "%d=%u\n", rdtgrp->plr->d->id, size); + } + goto out; + } + + for_each_alloc_enabled_rdt_resource(r) { + sep = false; + seq_printf(s, "%*s:", max_name_width, r->name); + list_for_each_entry(d, &r->domains, list) { + if (sep) + seq_putc(s, ';'); + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP) { + size = 0; + } else { + ctrl = (!is_mba_sc(r) ? + d->ctrl_val[rdtgrp->closid] : + d->mbps_val[rdtgrp->closid]); + if (r->rid == RDT_RESOURCE_MBA) + size = ctrl; + else + size = rdtgroup_cbm_to_size(r, d, ctrl); + } + seq_printf(s, "%d=%u", d->id, size); + sep = true; + } + seq_putc(s, '\n'); + } + +out: + rdtgroup_kn_unlock(of->kn); + + return ret; +} + +/* rdtgroup information files for one cache resource. */ +static struct rftype res_common_files[] = { + { + .name = "last_cmd_status", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_last_cmd_status_show, + .fflags = RF_TOP_INFO, + }, + { + .name = "num_closids", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_num_closids_show, + .fflags = RF_CTRL_INFO, + }, + { + .name = "mon_features", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_mon_features_show, + .fflags = RF_MON_INFO, + }, + { + .name = "num_rmids", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_num_rmids_show, + .fflags = RF_MON_INFO, + }, + { + .name = "cbm_mask", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_default_ctrl_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_CACHE, + }, + { + .name = "min_cbm_bits", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_min_cbm_bits_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_CACHE, + }, + { + .name = "shareable_bits", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_shareable_bits_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_CACHE, + }, + { + .name = "bit_usage", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_bit_usage_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_CACHE, + }, + { + .name = "min_bandwidth", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_min_bw_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_MB, + }, + { + .name = "bandwidth_gran", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_bw_gran_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_MB, + }, + { + .name = "delay_linear", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_delay_linear_show, + .fflags = RF_CTRL_INFO | RFTYPE_RES_MB, + }, + /* + * Platform specific which (if any) capabilities are provided by + * thread_throttle_mode. Defer "fflags" initialization to platform + * discovery. + */ + { + .name = "thread_throttle_mode", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdt_thread_throttle_mode_show, + }, + { + .name = "max_threshold_occupancy", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = max_threshold_occ_write, + .seq_show = max_threshold_occ_show, + .fflags = RF_MON_INFO | RFTYPE_RES_CACHE, + }, + { + .name = "cpus", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = rdtgroup_cpus_write, + .seq_show = rdtgroup_cpus_show, + .fflags = RFTYPE_BASE, + }, + { + .name = "cpus_list", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = rdtgroup_cpus_write, + .seq_show = rdtgroup_cpus_show, + .flags = RFTYPE_FLAGS_CPUS_LIST, + .fflags = RFTYPE_BASE, + }, + { + .name = "tasks", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = rdtgroup_tasks_write, + .seq_show = rdtgroup_tasks_show, + .fflags = RFTYPE_BASE, + }, + { + .name = "schemata", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = rdtgroup_schemata_write, + .seq_show = rdtgroup_schemata_show, + .fflags = RF_CTRL_BASE, + }, + { + .name = "mode", + .mode = 0644, + .kf_ops = &rdtgroup_kf_single_ops, + .write = rdtgroup_mode_write, + .seq_show = rdtgroup_mode_show, + .fflags = RF_CTRL_BASE, + }, + { + .name = "size", + .mode = 0444, + .kf_ops = &rdtgroup_kf_single_ops, + .seq_show = rdtgroup_size_show, + .fflags = RF_CTRL_BASE, + }, + +}; + +static int rdtgroup_add_files(struct kernfs_node *kn, unsigned long fflags) +{ + struct rftype *rfts, *rft; + int ret, len; + + rfts = res_common_files; + len = ARRAY_SIZE(res_common_files); + + lockdep_assert_held(&rdtgroup_mutex); + + for (rft = rfts; rft < rfts + len; rft++) { + if (rft->fflags && ((fflags & rft->fflags) == rft->fflags)) { + ret = rdtgroup_add_file(kn, rft); + if (ret) + goto error; + } + } + + return 0; +error: + pr_warn("Failed to add %s, err=%d\n", rft->name, ret); + while (--rft >= rfts) { + if ((fflags & rft->fflags) == rft->fflags) + kernfs_remove_by_name(kn, rft->name); + } + return ret; +} + +static struct rftype *rdtgroup_get_rftype_by_name(const char *name) +{ + struct rftype *rfts, *rft; + int len; + + rfts = res_common_files; + len = ARRAY_SIZE(res_common_files); + + for (rft = rfts; rft < rfts + len; rft++) { + if (!strcmp(rft->name, name)) + return rft; + } + + return NULL; +} + +void __init thread_throttle_mode_init(void) +{ + struct rftype *rft; + + rft = rdtgroup_get_rftype_by_name("thread_throttle_mode"); + if (!rft) + return; + + rft->fflags = RF_CTRL_INFO | RFTYPE_RES_MB; +} + +/** + * rdtgroup_kn_mode_restrict - Restrict user access to named resctrl file + * @r: The resource group with which the file is associated. + * @name: Name of the file + * + * The permissions of named resctrl file, directory, or link are modified + * to not allow read, write, or execute by any user. + * + * WARNING: This function is intended to communicate to the user that the + * resctrl file has been locked down - that it is not relevant to the + * particular state the system finds itself in. It should not be relied + * on to protect from user access because after the file's permissions + * are restricted the user can still change the permissions using chmod + * from the command line. + * + * Return: 0 on success, <0 on failure. + */ +int rdtgroup_kn_mode_restrict(struct rdtgroup *r, const char *name) +{ + struct iattr iattr = {.ia_valid = ATTR_MODE,}; + struct kernfs_node *kn; + int ret = 0; + + kn = kernfs_find_and_get_ns(r->kn, name, NULL); + if (!kn) + return -ENOENT; + + switch (kernfs_type(kn)) { + case KERNFS_DIR: + iattr.ia_mode = S_IFDIR; + break; + case KERNFS_FILE: + iattr.ia_mode = S_IFREG; + break; + case KERNFS_LINK: + iattr.ia_mode = S_IFLNK; + break; + } + + ret = kernfs_setattr(kn, &iattr); + kernfs_put(kn); + return ret; +} + +/** + * rdtgroup_kn_mode_restore - Restore user access to named resctrl file + * @r: The resource group with which the file is associated. + * @name: Name of the file + * @mask: Mask of permissions that should be restored + * + * Restore the permissions of the named file. If @name is a directory the + * permissions of its parent will be used. + * + * Return: 0 on success, <0 on failure. + */ +int rdtgroup_kn_mode_restore(struct rdtgroup *r, const char *name, + umode_t mask) +{ + struct iattr iattr = {.ia_valid = ATTR_MODE,}; + struct kernfs_node *kn, *parent; + struct rftype *rfts, *rft; + int ret, len; + + rfts = res_common_files; + len = ARRAY_SIZE(res_common_files); + + for (rft = rfts; rft < rfts + len; rft++) { + if (!strcmp(rft->name, name)) + iattr.ia_mode = rft->mode & mask; + } + + kn = kernfs_find_and_get_ns(r->kn, name, NULL); + if (!kn) + return -ENOENT; + + switch (kernfs_type(kn)) { + case KERNFS_DIR: + parent = kernfs_get_parent(kn); + if (parent) { + iattr.ia_mode |= parent->mode; + kernfs_put(parent); + } + iattr.ia_mode |= S_IFDIR; + break; + case KERNFS_FILE: + iattr.ia_mode |= S_IFREG; + break; + case KERNFS_LINK: + iattr.ia_mode |= S_IFLNK; + break; + } + + ret = kernfs_setattr(kn, &iattr); + kernfs_put(kn); + return ret; +} + +static int rdtgroup_mkdir_info_resdir(struct rdt_resource *r, char *name, + unsigned long fflags) +{ + struct kernfs_node *kn_subdir; + int ret; + + kn_subdir = kernfs_create_dir(kn_info, name, + kn_info->mode, r); + if (IS_ERR(kn_subdir)) + return PTR_ERR(kn_subdir); + + ret = rdtgroup_kn_set_ugid(kn_subdir); + if (ret) + return ret; + + ret = rdtgroup_add_files(kn_subdir, fflags); + if (!ret) + kernfs_activate(kn_subdir); + + return ret; +} + +static int rdtgroup_create_info_dir(struct kernfs_node *parent_kn) +{ + struct rdt_resource *r; + unsigned long fflags; + char name[32]; + int ret; + + /* create the directory */ + kn_info = kernfs_create_dir(parent_kn, "info", parent_kn->mode, NULL); + if (IS_ERR(kn_info)) + return PTR_ERR(kn_info); + + ret = rdtgroup_add_files(kn_info, RF_TOP_INFO); + if (ret) + goto out_destroy; + + for_each_alloc_enabled_rdt_resource(r) { + fflags = r->fflags | RF_CTRL_INFO; + ret = rdtgroup_mkdir_info_resdir(r, r->name, fflags); + if (ret) + goto out_destroy; + } + + for_each_mon_enabled_rdt_resource(r) { + fflags = r->fflags | RF_MON_INFO; + sprintf(name, "%s_MON", r->name); + ret = rdtgroup_mkdir_info_resdir(r, name, fflags); + if (ret) + goto out_destroy; + } + + ret = rdtgroup_kn_set_ugid(kn_info); + if (ret) + goto out_destroy; + + kernfs_activate(kn_info); + + return 0; + +out_destroy: + kernfs_remove(kn_info); + return ret; +} + +static int +mongroup_create_dir(struct kernfs_node *parent_kn, struct rdtgroup *prgrp, + char *name, struct kernfs_node **dest_kn) +{ + struct kernfs_node *kn; + int ret; + + /* create the directory */ + kn = kernfs_create_dir(parent_kn, name, parent_kn->mode, prgrp); + if (IS_ERR(kn)) + return PTR_ERR(kn); + + if (dest_kn) + *dest_kn = kn; + + ret = rdtgroup_kn_set_ugid(kn); + if (ret) + goto out_destroy; + + kernfs_activate(kn); + + return 0; + +out_destroy: + kernfs_remove(kn); + return ret; +} + +static void l3_qos_cfg_update(void *arg) +{ + bool *enable = arg; + + wrmsrl(MSR_IA32_L3_QOS_CFG, *enable ? L3_QOS_CDP_ENABLE : 0ULL); +} + +static void l2_qos_cfg_update(void *arg) +{ + bool *enable = arg; + + wrmsrl(MSR_IA32_L2_QOS_CFG, *enable ? L2_QOS_CDP_ENABLE : 0ULL); +} + +static inline bool is_mba_linear(void) +{ + return rdt_resources_all[RDT_RESOURCE_MBA].membw.delay_linear; +} + +static int set_cache_qos_cfg(int level, bool enable) +{ + void (*update)(void *arg); + struct rdt_resource *r_l; + cpumask_var_t cpu_mask; + struct rdt_domain *d; + int cpu; + + if (level == RDT_RESOURCE_L3) + update = l3_qos_cfg_update; + else if (level == RDT_RESOURCE_L2) + update = l2_qos_cfg_update; + else + return -EINVAL; + + if (!zalloc_cpumask_var(&cpu_mask, GFP_KERNEL)) + return -ENOMEM; + + r_l = &rdt_resources_all[level]; + list_for_each_entry(d, &r_l->domains, list) { + if (r_l->cache.arch_has_per_cpu_cfg) + /* Pick all the CPUs in the domain instance */ + for_each_cpu(cpu, &d->cpu_mask) + cpumask_set_cpu(cpu, cpu_mask); + else + /* Pick one CPU from each domain instance to update MSR */ + cpumask_set_cpu(cpumask_any(&d->cpu_mask), cpu_mask); + } + cpu = get_cpu(); + /* Update QOS_CFG MSR on this cpu if it's in cpu_mask. */ + if (cpumask_test_cpu(cpu, cpu_mask)) + update(&enable); + /* Update QOS_CFG MSR on all other cpus in cpu_mask. */ + smp_call_function_many(cpu_mask, update, &enable, 1); + put_cpu(); + + free_cpumask_var(cpu_mask); + + return 0; +} + +/* Restore the qos cfg state when a domain comes online */ +void rdt_domain_reconfigure_cdp(struct rdt_resource *r) +{ + if (!r->alloc_capable) + return; + + if (r == &rdt_resources_all[RDT_RESOURCE_L2DATA]) + l2_qos_cfg_update(&r->alloc_enabled); + + if (r == &rdt_resources_all[RDT_RESOURCE_L3DATA]) + l3_qos_cfg_update(&r->alloc_enabled); +} + +/* + * Enable or disable the MBA software controller + * which helps user specify bandwidth in MBps. + * MBA software controller is supported only if + * MBM is supported and MBA is in linear scale. + */ +static int set_mba_sc(bool mba_sc) +{ + struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_MBA]; + struct rdt_domain *d; + + if (!is_mbm_enabled() || !is_mba_linear() || + mba_sc == is_mba_sc(r)) + return -EINVAL; + + r->membw.mba_sc = mba_sc; + list_for_each_entry(d, &r->domains, list) + setup_default_ctrlval(r, d->ctrl_val, d->mbps_val); + + return 0; +} + +static int cdp_enable(int level, int data_type, int code_type) +{ + struct rdt_resource *r_ldata = &rdt_resources_all[data_type]; + struct rdt_resource *r_lcode = &rdt_resources_all[code_type]; + struct rdt_resource *r_l = &rdt_resources_all[level]; + int ret; + + if (!r_l->alloc_capable || !r_ldata->alloc_capable || + !r_lcode->alloc_capable) + return -EINVAL; + + ret = set_cache_qos_cfg(level, true); + if (!ret) { + r_l->alloc_enabled = false; + r_ldata->alloc_enabled = true; + r_lcode->alloc_enabled = true; + } + return ret; +} + +static int cdpl3_enable(void) +{ + return cdp_enable(RDT_RESOURCE_L3, RDT_RESOURCE_L3DATA, + RDT_RESOURCE_L3CODE); +} + +static int cdpl2_enable(void) +{ + return cdp_enable(RDT_RESOURCE_L2, RDT_RESOURCE_L2DATA, + RDT_RESOURCE_L2CODE); +} + +static void cdp_disable(int level, int data_type, int code_type) +{ + struct rdt_resource *r = &rdt_resources_all[level]; + + r->alloc_enabled = r->alloc_capable; + + if (rdt_resources_all[data_type].alloc_enabled) { + rdt_resources_all[data_type].alloc_enabled = false; + rdt_resources_all[code_type].alloc_enabled = false; + set_cache_qos_cfg(level, false); + } +} + +static void cdpl3_disable(void) +{ + cdp_disable(RDT_RESOURCE_L3, RDT_RESOURCE_L3DATA, RDT_RESOURCE_L3CODE); +} + +static void cdpl2_disable(void) +{ + cdp_disable(RDT_RESOURCE_L2, RDT_RESOURCE_L2DATA, RDT_RESOURCE_L2CODE); +} + +static void cdp_disable_all(void) +{ + if (rdt_resources_all[RDT_RESOURCE_L3DATA].alloc_enabled) + cdpl3_disable(); + if (rdt_resources_all[RDT_RESOURCE_L2DATA].alloc_enabled) + cdpl2_disable(); +} + +/* + * We don't allow rdtgroup directories to be created anywhere + * except the root directory. Thus when looking for the rdtgroup + * structure for a kernfs node we are either looking at a directory, + * in which case the rdtgroup structure is pointed at by the "priv" + * field, otherwise we have a file, and need only look to the parent + * to find the rdtgroup. + */ +static struct rdtgroup *kernfs_to_rdtgroup(struct kernfs_node *kn) +{ + if (kernfs_type(kn) == KERNFS_DIR) { + /* + * All the resource directories use "kn->priv" + * to point to the "struct rdtgroup" for the + * resource. "info" and its subdirectories don't + * have rdtgroup structures, so return NULL here. + */ + if (kn == kn_info || kn->parent == kn_info) + return NULL; + else + return kn->priv; + } else { + return kn->parent->priv; + } +} + +struct rdtgroup *rdtgroup_kn_lock_live(struct kernfs_node *kn) +{ + struct rdtgroup *rdtgrp = kernfs_to_rdtgroup(kn); + + if (!rdtgrp) + return NULL; + + atomic_inc(&rdtgrp->waitcount); + kernfs_break_active_protection(kn); + + mutex_lock(&rdtgroup_mutex); + + /* Was this group deleted while we waited? */ + if (rdtgrp->flags & RDT_DELETED) + return NULL; + + return rdtgrp; +} + +void rdtgroup_kn_unlock(struct kernfs_node *kn) +{ + struct rdtgroup *rdtgrp = kernfs_to_rdtgroup(kn); + + if (!rdtgrp) + return; + + mutex_unlock(&rdtgroup_mutex); + + if (atomic_dec_and_test(&rdtgrp->waitcount) && + (rdtgrp->flags & RDT_DELETED)) { + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) + rdtgroup_pseudo_lock_remove(rdtgrp); + kernfs_unbreak_active_protection(kn); + rdtgroup_remove(rdtgrp); + } else { + kernfs_unbreak_active_protection(kn); + } +} + +static int mkdir_mondata_all(struct kernfs_node *parent_kn, + struct rdtgroup *prgrp, + struct kernfs_node **mon_data_kn); + +static int rdt_enable_ctx(struct rdt_fs_context *ctx) +{ + int ret = 0; + + if (ctx->enable_cdpl2) + ret = cdpl2_enable(); + + if (!ret && ctx->enable_cdpl3) + ret = cdpl3_enable(); + + if (!ret && ctx->enable_mba_mbps) + ret = set_mba_sc(true); + + return ret; +} + +static int rdt_get_tree(struct fs_context *fc) +{ + struct rdt_fs_context *ctx = rdt_fc2context(fc); + struct rdt_domain *dom; + struct rdt_resource *r; + int ret; + + cpus_read_lock(); + mutex_lock(&rdtgroup_mutex); + /* + * resctrl file system can only be mounted once. + */ + if (static_branch_unlikely(&rdt_enable_key)) { + ret = -EBUSY; + goto out; + } + + ret = rdt_enable_ctx(ctx); + if (ret < 0) + goto out_cdp; + + closid_init(); + + ret = rdtgroup_create_info_dir(rdtgroup_default.kn); + if (ret < 0) + goto out_mba; + + if (rdt_mon_capable) { + ret = mongroup_create_dir(rdtgroup_default.kn, + &rdtgroup_default, "mon_groups", + &kn_mongrp); + if (ret < 0) + goto out_info; + + ret = mkdir_mondata_all(rdtgroup_default.kn, + &rdtgroup_default, &kn_mondata); + if (ret < 0) + goto out_mongrp; + rdtgroup_default.mon.mon_data_kn = kn_mondata; + } + + ret = rdt_pseudo_lock_init(); + if (ret) + goto out_mondata; + + ret = kernfs_get_tree(fc); + if (ret < 0) + goto out_psl; + + if (rdt_alloc_capable) + static_branch_enable_cpuslocked(&rdt_alloc_enable_key); + if (rdt_mon_capable) + static_branch_enable_cpuslocked(&rdt_mon_enable_key); + + if (rdt_alloc_capable || rdt_mon_capable) + static_branch_enable_cpuslocked(&rdt_enable_key); + + if (is_mbm_enabled()) { + r = &rdt_resources_all[RDT_RESOURCE_L3]; + list_for_each_entry(dom, &r->domains, list) + mbm_setup_overflow_handler(dom, MBM_OVERFLOW_INTERVAL); + } + + goto out; + +out_psl: + rdt_pseudo_lock_release(); +out_mondata: + if (rdt_mon_capable) + kernfs_remove(kn_mondata); +out_mongrp: + if (rdt_mon_capable) + kernfs_remove(kn_mongrp); +out_info: + kernfs_remove(kn_info); +out_mba: + if (ctx->enable_mba_mbps) + set_mba_sc(false); +out_cdp: + cdp_disable_all(); +out: + rdt_last_cmd_clear(); + mutex_unlock(&rdtgroup_mutex); + cpus_read_unlock(); + return ret; +} + +enum rdt_param { + Opt_cdp, + Opt_cdpl2, + Opt_mba_mbps, + nr__rdt_params +}; + +static const struct fs_parameter_spec rdt_fs_parameters[] = { + fsparam_flag("cdp", Opt_cdp), + fsparam_flag("cdpl2", Opt_cdpl2), + fsparam_flag("mba_MBps", Opt_mba_mbps), + {} +}; + +static int rdt_parse_param(struct fs_context *fc, struct fs_parameter *param) +{ + struct rdt_fs_context *ctx = rdt_fc2context(fc); + struct fs_parse_result result; + int opt; + + opt = fs_parse(fc, rdt_fs_parameters, param, &result); + if (opt < 0) + return opt; + + switch (opt) { + case Opt_cdp: + ctx->enable_cdpl3 = true; + return 0; + case Opt_cdpl2: + ctx->enable_cdpl2 = true; + return 0; + case Opt_mba_mbps: + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) + return -EINVAL; + ctx->enable_mba_mbps = true; + return 0; + } + + return -EINVAL; +} + +static void rdt_fs_context_free(struct fs_context *fc) +{ + struct rdt_fs_context *ctx = rdt_fc2context(fc); + + kernfs_free_fs_context(fc); + kfree(ctx); +} + +static const struct fs_context_operations rdt_fs_context_ops = { + .free = rdt_fs_context_free, + .parse_param = rdt_parse_param, + .get_tree = rdt_get_tree, +}; + +static int rdt_init_fs_context(struct fs_context *fc) +{ + struct rdt_fs_context *ctx; + + ctx = kzalloc(sizeof(struct rdt_fs_context), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + + ctx->kfc.root = rdt_root; + ctx->kfc.magic = RDTGROUP_SUPER_MAGIC; + fc->fs_private = &ctx->kfc; + fc->ops = &rdt_fs_context_ops; + put_user_ns(fc->user_ns); + fc->user_ns = get_user_ns(&init_user_ns); + fc->global = true; + return 0; +} + +static int reset_all_ctrls(struct rdt_resource *r) +{ + struct msr_param msr_param; + cpumask_var_t cpu_mask; + struct rdt_domain *d; + int i, cpu; + + if (!zalloc_cpumask_var(&cpu_mask, GFP_KERNEL)) + return -ENOMEM; + + msr_param.res = r; + msr_param.low = 0; + msr_param.high = r->num_closid; + + /* + * Disable resource control for this resource by setting all + * CBMs in all domains to the maximum mask value. Pick one CPU + * from each domain to update the MSRs below. + */ + list_for_each_entry(d, &r->domains, list) { + cpumask_set_cpu(cpumask_any(&d->cpu_mask), cpu_mask); + + for (i = 0; i < r->num_closid; i++) + d->ctrl_val[i] = r->default_ctrl; + } + cpu = get_cpu(); + /* Update CBM on this cpu if it's in cpu_mask. */ + if (cpumask_test_cpu(cpu, cpu_mask)) + rdt_ctrl_update(&msr_param); + /* Update CBM on all other cpus in cpu_mask. */ + smp_call_function_many(cpu_mask, rdt_ctrl_update, &msr_param, 1); + put_cpu(); + + free_cpumask_var(cpu_mask); + + return 0; +} + +/* + * Move tasks from one to the other group. If @from is NULL, then all tasks + * in the systems are moved unconditionally (used for teardown). + * + * If @mask is not NULL the cpus on which moved tasks are running are set + * in that mask so the update smp function call is restricted to affected + * cpus. + */ +static void rdt_move_group_tasks(struct rdtgroup *from, struct rdtgroup *to, + struct cpumask *mask) +{ + struct task_struct *p, *t; + + read_lock(&tasklist_lock); + for_each_process_thread(p, t) { + if (!from || is_closid_match(t, from) || + is_rmid_match(t, from)) { + WRITE_ONCE(t->closid, to->closid); + WRITE_ONCE(t->rmid, to->mon.rmid); + + /* + * Order the closid/rmid stores above before the loads + * in task_curr(). This pairs with the full barrier + * between the rq->curr update and resctrl_sched_in() + * during context switch. + */ + smp_mb(); + + /* + * If the task is on a CPU, set the CPU in the mask. + * The detection is inaccurate as tasks might move or + * schedule before the smp function call takes place. + * In such a case the function call is pointless, but + * there is no other side effect. + */ + if (IS_ENABLED(CONFIG_SMP) && mask && task_curr(t)) + cpumask_set_cpu(task_cpu(t), mask); + } + } + read_unlock(&tasklist_lock); +} + +static void free_all_child_rdtgrp(struct rdtgroup *rdtgrp) +{ + struct rdtgroup *sentry, *stmp; + struct list_head *head; + + head = &rdtgrp->mon.crdtgrp_list; + list_for_each_entry_safe(sentry, stmp, head, mon.crdtgrp_list) { + free_rmid(sentry->mon.rmid); + list_del(&sentry->mon.crdtgrp_list); + + if (atomic_read(&sentry->waitcount) != 0) + sentry->flags = RDT_DELETED; + else + rdtgroup_remove(sentry); + } +} + +/* + * Forcibly remove all of subdirectories under root. + */ +static void rmdir_all_sub(void) +{ + struct rdtgroup *rdtgrp, *tmp; + + /* Move all tasks to the default resource group */ + rdt_move_group_tasks(NULL, &rdtgroup_default, NULL); + + list_for_each_entry_safe(rdtgrp, tmp, &rdt_all_groups, rdtgroup_list) { + /* Free any child rmids */ + free_all_child_rdtgrp(rdtgrp); + + /* Remove each rdtgroup other than root */ + if (rdtgrp == &rdtgroup_default) + continue; + + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) + rdtgroup_pseudo_lock_remove(rdtgrp); + + /* + * Give any CPUs back to the default group. We cannot copy + * cpu_online_mask because a CPU might have executed the + * offline callback already, but is still marked online. + */ + cpumask_or(&rdtgroup_default.cpu_mask, + &rdtgroup_default.cpu_mask, &rdtgrp->cpu_mask); + + free_rmid(rdtgrp->mon.rmid); + + kernfs_remove(rdtgrp->kn); + list_del(&rdtgrp->rdtgroup_list); + + if (atomic_read(&rdtgrp->waitcount) != 0) + rdtgrp->flags = RDT_DELETED; + else + rdtgroup_remove(rdtgrp); + } + /* Notify online CPUs to update per cpu storage and PQR_ASSOC MSR */ + update_closid_rmid(cpu_online_mask, &rdtgroup_default); + + kernfs_remove(kn_info); + kernfs_remove(kn_mongrp); + kernfs_remove(kn_mondata); +} + +static void rdt_kill_sb(struct super_block *sb) +{ + struct rdt_resource *r; + + cpus_read_lock(); + mutex_lock(&rdtgroup_mutex); + + set_mba_sc(false); + + /*Put everything back to default values. */ + for_each_alloc_enabled_rdt_resource(r) + reset_all_ctrls(r); + cdp_disable_all(); + rmdir_all_sub(); + rdt_pseudo_lock_release(); + rdtgroup_default.mode = RDT_MODE_SHAREABLE; + static_branch_disable_cpuslocked(&rdt_alloc_enable_key); + static_branch_disable_cpuslocked(&rdt_mon_enable_key); + static_branch_disable_cpuslocked(&rdt_enable_key); + kernfs_kill_sb(sb); + mutex_unlock(&rdtgroup_mutex); + cpus_read_unlock(); +} + +static struct file_system_type rdt_fs_type = { + .name = "resctrl", + .init_fs_context = rdt_init_fs_context, + .parameters = rdt_fs_parameters, + .kill_sb = rdt_kill_sb, +}; + +static int mon_addfile(struct kernfs_node *parent_kn, const char *name, + void *priv) +{ + struct kernfs_node *kn; + int ret = 0; + + kn = __kernfs_create_file(parent_kn, name, 0444, + GLOBAL_ROOT_UID, GLOBAL_ROOT_GID, 0, + &kf_mondata_ops, priv, NULL, NULL); + if (IS_ERR(kn)) + return PTR_ERR(kn); + + ret = rdtgroup_kn_set_ugid(kn); + if (ret) { + kernfs_remove(kn); + return ret; + } + + return ret; +} + +/* + * Remove all subdirectories of mon_data of ctrl_mon groups + * and monitor groups with given domain id. + */ +void rmdir_mondata_subdir_allrdtgrp(struct rdt_resource *r, unsigned int dom_id) +{ + struct rdtgroup *prgrp, *crgrp; + char name[32]; + + if (!r->mon_enabled) + return; + + list_for_each_entry(prgrp, &rdt_all_groups, rdtgroup_list) { + sprintf(name, "mon_%s_%02d", r->name, dom_id); + kernfs_remove_by_name(prgrp->mon.mon_data_kn, name); + + list_for_each_entry(crgrp, &prgrp->mon.crdtgrp_list, mon.crdtgrp_list) + kernfs_remove_by_name(crgrp->mon.mon_data_kn, name); + } +} + +static int mkdir_mondata_subdir(struct kernfs_node *parent_kn, + struct rdt_domain *d, + struct rdt_resource *r, struct rdtgroup *prgrp) +{ + union mon_data_bits priv; + struct kernfs_node *kn; + struct mon_evt *mevt; + struct rmid_read rr; + char name[32]; + int ret; + + sprintf(name, "mon_%s_%02d", r->name, d->id); + /* create the directory */ + kn = kernfs_create_dir(parent_kn, name, parent_kn->mode, prgrp); + if (IS_ERR(kn)) + return PTR_ERR(kn); + + ret = rdtgroup_kn_set_ugid(kn); + if (ret) + goto out_destroy; + + if (WARN_ON(list_empty(&r->evt_list))) { + ret = -EPERM; + goto out_destroy; + } + + priv.u.rid = r->rid; + priv.u.domid = d->id; + list_for_each_entry(mevt, &r->evt_list, list) { + priv.u.evtid = mevt->evtid; + ret = mon_addfile(kn, mevt->name, priv.priv); + if (ret) + goto out_destroy; + + if (is_mbm_event(mevt->evtid)) + mon_event_read(&rr, r, d, prgrp, mevt->evtid, true); + } + kernfs_activate(kn); + return 0; + +out_destroy: + kernfs_remove(kn); + return ret; +} + +/* + * Add all subdirectories of mon_data for "ctrl_mon" groups + * and "monitor" groups with given domain id. + */ +void mkdir_mondata_subdir_allrdtgrp(struct rdt_resource *r, + struct rdt_domain *d) +{ + struct kernfs_node *parent_kn; + struct rdtgroup *prgrp, *crgrp; + struct list_head *head; + + if (!r->mon_enabled) + return; + + list_for_each_entry(prgrp, &rdt_all_groups, rdtgroup_list) { + parent_kn = prgrp->mon.mon_data_kn; + mkdir_mondata_subdir(parent_kn, d, r, prgrp); + + head = &prgrp->mon.crdtgrp_list; + list_for_each_entry(crgrp, head, mon.crdtgrp_list) { + parent_kn = crgrp->mon.mon_data_kn; + mkdir_mondata_subdir(parent_kn, d, r, crgrp); + } + } +} + +static int mkdir_mondata_subdir_alldom(struct kernfs_node *parent_kn, + struct rdt_resource *r, + struct rdtgroup *prgrp) +{ + struct rdt_domain *dom; + int ret; + + list_for_each_entry(dom, &r->domains, list) { + ret = mkdir_mondata_subdir(parent_kn, dom, r, prgrp); + if (ret) + return ret; + } + + return 0; +} + +/* + * This creates a directory mon_data which contains the monitored data. + * + * mon_data has one directory for each domain whic are named + * in the format mon_<domain_name>_<domain_id>. For ex: A mon_data + * with L3 domain looks as below: + * ./mon_data: + * mon_L3_00 + * mon_L3_01 + * mon_L3_02 + * ... + * + * Each domain directory has one file per event: + * ./mon_L3_00/: + * llc_occupancy + * + */ +static int mkdir_mondata_all(struct kernfs_node *parent_kn, + struct rdtgroup *prgrp, + struct kernfs_node **dest_kn) +{ + struct rdt_resource *r; + struct kernfs_node *kn; + int ret; + + /* + * Create the mon_data directory first. + */ + ret = mongroup_create_dir(parent_kn, prgrp, "mon_data", &kn); + if (ret) + return ret; + + if (dest_kn) + *dest_kn = kn; + + /* + * Create the subdirectories for each domain. Note that all events + * in a domain like L3 are grouped into a resource whose domain is L3 + */ + for_each_mon_enabled_rdt_resource(r) { + ret = mkdir_mondata_subdir_alldom(kn, r, prgrp); + if (ret) + goto out_destroy; + } + + return 0; + +out_destroy: + kernfs_remove(kn); + return ret; +} + +/** + * cbm_ensure_valid - Enforce validity on provided CBM + * @_val: Candidate CBM + * @r: RDT resource to which the CBM belongs + * + * The provided CBM represents all cache portions available for use. This + * may be represented by a bitmap that does not consist of contiguous ones + * and thus be an invalid CBM. + * Here the provided CBM is forced to be a valid CBM by only considering + * the first set of contiguous bits as valid and clearing all bits. + * The intention here is to provide a valid default CBM with which a new + * resource group is initialized. The user can follow this with a + * modification to the CBM if the default does not satisfy the + * requirements. + */ +static u32 cbm_ensure_valid(u32 _val, struct rdt_resource *r) +{ + unsigned int cbm_len = r->cache.cbm_len; + unsigned long first_bit, zero_bit; + unsigned long val = _val; + + if (!val) + return 0; + + first_bit = find_first_bit(&val, cbm_len); + zero_bit = find_next_zero_bit(&val, cbm_len, first_bit); + + /* Clear any remaining bits to ensure contiguous region */ + bitmap_clear(&val, zero_bit, cbm_len - zero_bit); + return (u32)val; +} + +/* + * Initialize cache resources per RDT domain + * + * Set the RDT domain up to start off with all usable allocations. That is, + * all shareable and unused bits. All-zero CBM is invalid. + */ +static int __init_one_rdt_domain(struct rdt_domain *d, struct rdt_resource *r, + u32 closid) +{ + struct rdt_resource *r_cdp = NULL; + struct rdt_domain *d_cdp = NULL; + u32 used_b = 0, unused_b = 0; + unsigned long tmp_cbm; + enum rdtgrp_mode mode; + u32 peer_ctl, *ctrl; + int i; + + rdt_cdp_peer_get(r, d, &r_cdp, &d_cdp); + d->have_new_ctrl = false; + d->new_ctrl = r->cache.shareable_bits; + used_b = r->cache.shareable_bits; + ctrl = d->ctrl_val; + for (i = 0; i < closids_supported(); i++, ctrl++) { + if (closid_allocated(i) && i != closid) { + mode = rdtgroup_mode_by_closid(i); + if (mode == RDT_MODE_PSEUDO_LOCKSETUP) + /* + * ctrl values for locksetup aren't relevant + * until the schemata is written, and the mode + * becomes RDT_MODE_PSEUDO_LOCKED. + */ + continue; + /* + * If CDP is active include peer domain's + * usage to ensure there is no overlap + * with an exclusive group. + */ + if (d_cdp) + peer_ctl = d_cdp->ctrl_val[i]; + else + peer_ctl = 0; + used_b |= *ctrl | peer_ctl; + if (mode == RDT_MODE_SHAREABLE) + d->new_ctrl |= *ctrl | peer_ctl; + } + } + if (d->plr && d->plr->cbm > 0) + used_b |= d->plr->cbm; + unused_b = used_b ^ (BIT_MASK(r->cache.cbm_len) - 1); + unused_b &= BIT_MASK(r->cache.cbm_len) - 1; + d->new_ctrl |= unused_b; + /* + * Force the initial CBM to be valid, user can + * modify the CBM based on system availability. + */ + d->new_ctrl = cbm_ensure_valid(d->new_ctrl, r); + /* + * Assign the u32 CBM to an unsigned long to ensure that + * bitmap_weight() does not access out-of-bound memory. + */ + tmp_cbm = d->new_ctrl; + if (bitmap_weight(&tmp_cbm, r->cache.cbm_len) < r->cache.min_cbm_bits) { + rdt_last_cmd_printf("No space on %s:%d\n", r->name, d->id); + return -ENOSPC; + } + d->have_new_ctrl = true; + + return 0; +} + +/* + * Initialize cache resources with default values. + * + * A new RDT group is being created on an allocation capable (CAT) + * supporting system. Set this group up to start off with all usable + * allocations. + * + * If there are no more shareable bits available on any domain then + * the entire allocation will fail. + */ +static int rdtgroup_init_cat(struct rdt_resource *r, u32 closid) +{ + struct rdt_domain *d; + int ret; + + list_for_each_entry(d, &r->domains, list) { + ret = __init_one_rdt_domain(d, r, closid); + if (ret < 0) + return ret; + } + + return 0; +} + +/* Initialize MBA resource with default values. */ +static void rdtgroup_init_mba(struct rdt_resource *r) +{ + struct rdt_domain *d; + + list_for_each_entry(d, &r->domains, list) { + d->new_ctrl = is_mba_sc(r) ? MBA_MAX_MBPS : r->default_ctrl; + d->have_new_ctrl = true; + } +} + +/* Initialize the RDT group's allocations. */ +static int rdtgroup_init_alloc(struct rdtgroup *rdtgrp) +{ + struct rdt_resource *r; + int ret; + + for_each_alloc_enabled_rdt_resource(r) { + if (r->rid == RDT_RESOURCE_MBA) { + rdtgroup_init_mba(r); + } else { + ret = rdtgroup_init_cat(r, rdtgrp->closid); + if (ret < 0) + return ret; + } + + ret = update_domains(r, rdtgrp->closid); + if (ret < 0) { + rdt_last_cmd_puts("Failed to initialize allocations\n"); + return ret; + } + + } + + rdtgrp->mode = RDT_MODE_SHAREABLE; + + return 0; +} + +static int mkdir_rdt_prepare(struct kernfs_node *parent_kn, + const char *name, umode_t mode, + enum rdt_group_type rtype, struct rdtgroup **r) +{ + struct rdtgroup *prdtgrp, *rdtgrp; + struct kernfs_node *kn; + uint files = 0; + int ret; + + prdtgrp = rdtgroup_kn_lock_live(parent_kn); + if (!prdtgrp) { + ret = -ENODEV; + goto out_unlock; + } + + if (rtype == RDTMON_GROUP && + (prdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP || + prdtgrp->mode == RDT_MODE_PSEUDO_LOCKED)) { + ret = -EINVAL; + rdt_last_cmd_puts("Pseudo-locking in progress\n"); + goto out_unlock; + } + + /* allocate the rdtgroup. */ + rdtgrp = kzalloc(sizeof(*rdtgrp), GFP_KERNEL); + if (!rdtgrp) { + ret = -ENOSPC; + rdt_last_cmd_puts("Kernel out of memory\n"); + goto out_unlock; + } + *r = rdtgrp; + rdtgrp->mon.parent = prdtgrp; + rdtgrp->type = rtype; + INIT_LIST_HEAD(&rdtgrp->mon.crdtgrp_list); + + /* kernfs creates the directory for rdtgrp */ + kn = kernfs_create_dir(parent_kn, name, mode, rdtgrp); + if (IS_ERR(kn)) { + ret = PTR_ERR(kn); + rdt_last_cmd_puts("kernfs create error\n"); + goto out_free_rgrp; + } + rdtgrp->kn = kn; + + /* + * kernfs_remove() will drop the reference count on "kn" which + * will free it. But we still need it to stick around for the + * rdtgroup_kn_unlock(kn) call. Take one extra reference here, + * which will be dropped by kernfs_put() in rdtgroup_remove(). + */ + kernfs_get(kn); + + ret = rdtgroup_kn_set_ugid(kn); + if (ret) { + rdt_last_cmd_puts("kernfs perm error\n"); + goto out_destroy; + } + + files = RFTYPE_BASE | BIT(RF_CTRLSHIFT + rtype); + ret = rdtgroup_add_files(kn, files); + if (ret) { + rdt_last_cmd_puts("kernfs fill error\n"); + goto out_destroy; + } + + if (rdt_mon_capable) { + ret = alloc_rmid(); + if (ret < 0) { + rdt_last_cmd_puts("Out of RMIDs\n"); + goto out_destroy; + } + rdtgrp->mon.rmid = ret; + + ret = mkdir_mondata_all(kn, rdtgrp, &rdtgrp->mon.mon_data_kn); + if (ret) { + rdt_last_cmd_puts("kernfs subdir error\n"); + goto out_idfree; + } + } + kernfs_activate(kn); + + /* + * The caller unlocks the parent_kn upon success. + */ + return 0; + +out_idfree: + free_rmid(rdtgrp->mon.rmid); +out_destroy: + kernfs_put(rdtgrp->kn); + kernfs_remove(rdtgrp->kn); +out_free_rgrp: + kfree(rdtgrp); +out_unlock: + rdtgroup_kn_unlock(parent_kn); + return ret; +} + +static void mkdir_rdt_prepare_clean(struct rdtgroup *rgrp) +{ + kernfs_remove(rgrp->kn); + free_rmid(rgrp->mon.rmid); + rdtgroup_remove(rgrp); +} + +/* + * Create a monitor group under "mon_groups" directory of a control + * and monitor group(ctrl_mon). This is a resource group + * to monitor a subset of tasks and cpus in its parent ctrl_mon group. + */ +static int rdtgroup_mkdir_mon(struct kernfs_node *parent_kn, + const char *name, umode_t mode) +{ + struct rdtgroup *rdtgrp, *prgrp; + int ret; + + ret = mkdir_rdt_prepare(parent_kn, name, mode, RDTMON_GROUP, &rdtgrp); + if (ret) + return ret; + + prgrp = rdtgrp->mon.parent; + rdtgrp->closid = prgrp->closid; + + /* + * Add the rdtgrp to the list of rdtgrps the parent + * ctrl_mon group has to track. + */ + list_add_tail(&rdtgrp->mon.crdtgrp_list, &prgrp->mon.crdtgrp_list); + + rdtgroup_kn_unlock(parent_kn); + return ret; +} + +/* + * These are rdtgroups created under the root directory. Can be used + * to allocate and monitor resources. + */ +static int rdtgroup_mkdir_ctrl_mon(struct kernfs_node *parent_kn, + const char *name, umode_t mode) +{ + struct rdtgroup *rdtgrp; + struct kernfs_node *kn; + u32 closid; + int ret; + + ret = mkdir_rdt_prepare(parent_kn, name, mode, RDTCTRL_GROUP, &rdtgrp); + if (ret) + return ret; + + kn = rdtgrp->kn; + ret = closid_alloc(); + if (ret < 0) { + rdt_last_cmd_puts("Out of CLOSIDs\n"); + goto out_common_fail; + } + closid = ret; + ret = 0; + + rdtgrp->closid = closid; + ret = rdtgroup_init_alloc(rdtgrp); + if (ret < 0) + goto out_id_free; + + list_add(&rdtgrp->rdtgroup_list, &rdt_all_groups); + + if (rdt_mon_capable) { + /* + * Create an empty mon_groups directory to hold the subset + * of tasks and cpus to monitor. + */ + ret = mongroup_create_dir(kn, rdtgrp, "mon_groups", NULL); + if (ret) { + rdt_last_cmd_puts("kernfs subdir error\n"); + goto out_del_list; + } + } + + goto out_unlock; + +out_del_list: + list_del(&rdtgrp->rdtgroup_list); +out_id_free: + closid_free(closid); +out_common_fail: + mkdir_rdt_prepare_clean(rdtgrp); +out_unlock: + rdtgroup_kn_unlock(parent_kn); + return ret; +} + +/* + * We allow creating mon groups only with in a directory called "mon_groups" + * which is present in every ctrl_mon group. Check if this is a valid + * "mon_groups" directory. + * + * 1. The directory should be named "mon_groups". + * 2. The mon group itself should "not" be named "mon_groups". + * This makes sure "mon_groups" directory always has a ctrl_mon group + * as parent. + */ +static bool is_mon_groups(struct kernfs_node *kn, const char *name) +{ + return (!strcmp(kn->name, "mon_groups") && + strcmp(name, "mon_groups")); +} + +static int rdtgroup_mkdir(struct kernfs_node *parent_kn, const char *name, + umode_t mode) +{ + /* Do not accept '\n' to avoid unparsable situation. */ + if (strchr(name, '\n')) + return -EINVAL; + + /* + * If the parent directory is the root directory and RDT + * allocation is supported, add a control and monitoring + * subdirectory + */ + if (rdt_alloc_capable && parent_kn == rdtgroup_default.kn) + return rdtgroup_mkdir_ctrl_mon(parent_kn, name, mode); + + /* + * If RDT monitoring is supported and the parent directory is a valid + * "mon_groups" directory, add a monitoring subdirectory. + */ + if (rdt_mon_capable && is_mon_groups(parent_kn, name)) + return rdtgroup_mkdir_mon(parent_kn, name, mode); + + return -EPERM; +} + +static int rdtgroup_rmdir_mon(struct kernfs_node *kn, struct rdtgroup *rdtgrp, + cpumask_var_t tmpmask) +{ + struct rdtgroup *prdtgrp = rdtgrp->mon.parent; + int cpu; + + /* Give any tasks back to the parent group */ + rdt_move_group_tasks(rdtgrp, prdtgrp, tmpmask); + + /* Update per cpu rmid of the moved CPUs first */ + for_each_cpu(cpu, &rdtgrp->cpu_mask) + per_cpu(pqr_state.default_rmid, cpu) = prdtgrp->mon.rmid; + /* + * Update the MSR on moved CPUs and CPUs which have moved + * task running on them. + */ + cpumask_or(tmpmask, tmpmask, &rdtgrp->cpu_mask); + update_closid_rmid(tmpmask, NULL); + + rdtgrp->flags = RDT_DELETED; + free_rmid(rdtgrp->mon.rmid); + + /* + * Remove the rdtgrp from the parent ctrl_mon group's list + */ + WARN_ON(list_empty(&prdtgrp->mon.crdtgrp_list)); + list_del(&rdtgrp->mon.crdtgrp_list); + + kernfs_remove(rdtgrp->kn); + + return 0; +} + +static int rdtgroup_ctrl_remove(struct kernfs_node *kn, + struct rdtgroup *rdtgrp) +{ + rdtgrp->flags = RDT_DELETED; + list_del(&rdtgrp->rdtgroup_list); + + kernfs_remove(rdtgrp->kn); + return 0; +} + +static int rdtgroup_rmdir_ctrl(struct kernfs_node *kn, struct rdtgroup *rdtgrp, + cpumask_var_t tmpmask) +{ + int cpu; + + /* Give any tasks back to the default group */ + rdt_move_group_tasks(rdtgrp, &rdtgroup_default, tmpmask); + + /* Give any CPUs back to the default group */ + cpumask_or(&rdtgroup_default.cpu_mask, + &rdtgroup_default.cpu_mask, &rdtgrp->cpu_mask); + + /* Update per cpu closid and rmid of the moved CPUs first */ + for_each_cpu(cpu, &rdtgrp->cpu_mask) { + per_cpu(pqr_state.default_closid, cpu) = rdtgroup_default.closid; + per_cpu(pqr_state.default_rmid, cpu) = rdtgroup_default.mon.rmid; + } + + /* + * Update the MSR on moved CPUs and CPUs which have moved + * task running on them. + */ + cpumask_or(tmpmask, tmpmask, &rdtgrp->cpu_mask); + update_closid_rmid(tmpmask, NULL); + + closid_free(rdtgrp->closid); + free_rmid(rdtgrp->mon.rmid); + + rdtgroup_ctrl_remove(kn, rdtgrp); + + /* + * Free all the child monitor group rmids. + */ + free_all_child_rdtgrp(rdtgrp); + + return 0; +} + +static int rdtgroup_rmdir(struct kernfs_node *kn) +{ + struct kernfs_node *parent_kn = kn->parent; + struct rdtgroup *rdtgrp; + cpumask_var_t tmpmask; + int ret = 0; + + if (!zalloc_cpumask_var(&tmpmask, GFP_KERNEL)) + return -ENOMEM; + + rdtgrp = rdtgroup_kn_lock_live(kn); + if (!rdtgrp) { + ret = -EPERM; + goto out; + } + + /* + * If the rdtgroup is a ctrl_mon group and parent directory + * is the root directory, remove the ctrl_mon group. + * + * If the rdtgroup is a mon group and parent directory + * is a valid "mon_groups" directory, remove the mon group. + */ + if (rdtgrp->type == RDTCTRL_GROUP && parent_kn == rdtgroup_default.kn && + rdtgrp != &rdtgroup_default) { + if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP || + rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) { + ret = rdtgroup_ctrl_remove(kn, rdtgrp); + } else { + ret = rdtgroup_rmdir_ctrl(kn, rdtgrp, tmpmask); + } + } else if (rdtgrp->type == RDTMON_GROUP && + is_mon_groups(parent_kn, kn->name)) { + ret = rdtgroup_rmdir_mon(kn, rdtgrp, tmpmask); + } else { + ret = -EPERM; + } + +out: + rdtgroup_kn_unlock(kn); + free_cpumask_var(tmpmask); + return ret; +} + +static int rdtgroup_show_options(struct seq_file *seq, struct kernfs_root *kf) +{ + if (rdt_resources_all[RDT_RESOURCE_L3DATA].alloc_enabled) + seq_puts(seq, ",cdp"); + + if (rdt_resources_all[RDT_RESOURCE_L2DATA].alloc_enabled) + seq_puts(seq, ",cdpl2"); + + if (is_mba_sc(&rdt_resources_all[RDT_RESOURCE_MBA])) + seq_puts(seq, ",mba_MBps"); + + return 0; +} + +static struct kernfs_syscall_ops rdtgroup_kf_syscall_ops = { + .mkdir = rdtgroup_mkdir, + .rmdir = rdtgroup_rmdir, + .show_options = rdtgroup_show_options, +}; + +static int __init rdtgroup_setup_root(void) +{ + int ret; + + rdt_root = kernfs_create_root(&rdtgroup_kf_syscall_ops, + KERNFS_ROOT_CREATE_DEACTIVATED | + KERNFS_ROOT_EXTRA_OPEN_PERM_CHECK, + &rdtgroup_default); + if (IS_ERR(rdt_root)) + return PTR_ERR(rdt_root); + + mutex_lock(&rdtgroup_mutex); + + rdtgroup_default.closid = 0; + rdtgroup_default.mon.rmid = 0; + rdtgroup_default.type = RDTCTRL_GROUP; + INIT_LIST_HEAD(&rdtgroup_default.mon.crdtgrp_list); + + list_add(&rdtgroup_default.rdtgroup_list, &rdt_all_groups); + + ret = rdtgroup_add_files(rdt_root->kn, RF_CTRL_BASE); + if (ret) { + kernfs_destroy_root(rdt_root); + goto out; + } + + rdtgroup_default.kn = rdt_root->kn; + kernfs_activate(rdtgroup_default.kn); + +out: + mutex_unlock(&rdtgroup_mutex); + + return ret; +} + +/* + * rdtgroup_init - rdtgroup initialization + * + * Setup resctrl file system including set up root, create mount point, + * register rdtgroup filesystem, and initialize files under root directory. + * + * Return: 0 on success or -errno + */ +int __init rdtgroup_init(void) +{ + int ret = 0; + + seq_buf_init(&last_cmd_status, last_cmd_status_buf, + sizeof(last_cmd_status_buf)); + + ret = rdtgroup_setup_root(); + if (ret) + return ret; + + ret = sysfs_create_mount_point(fs_kobj, "resctrl"); + if (ret) + goto cleanup_root; + + ret = register_filesystem(&rdt_fs_type); + if (ret) + goto cleanup_mountpoint; + + /* + * Adding the resctrl debugfs directory here may not be ideal since + * it would let the resctrl debugfs directory appear on the debugfs + * filesystem before the resctrl filesystem is mounted. + * It may also be ok since that would enable debugging of RDT before + * resctrl is mounted. + * The reason why the debugfs directory is created here and not in + * rdt_get_tree() is because rdt_get_tree() takes rdtgroup_mutex and + * during the debugfs directory creation also &sb->s_type->i_mutex_key + * (the lockdep class of inode->i_rwsem). Other filesystem + * interactions (eg. SyS_getdents) have the lock ordering: + * &sb->s_type->i_mutex_key --> &mm->mmap_lock + * During mmap(), called with &mm->mmap_lock, the rdtgroup_mutex + * is taken, thus creating dependency: + * &mm->mmap_lock --> rdtgroup_mutex for the latter that can cause + * issues considering the other two lock dependencies. + * By creating the debugfs directory here we avoid a dependency + * that may cause deadlock (even though file operations cannot + * occur until the filesystem is mounted, but I do not know how to + * tell lockdep that). + */ + debugfs_resctrl = debugfs_create_dir("resctrl", NULL); + + return 0; + +cleanup_mountpoint: + sysfs_remove_mount_point(fs_kobj, "resctrl"); +cleanup_root: + kernfs_destroy_root(rdt_root); + + return ret; +} + +void __exit rdtgroup_exit(void) +{ + debugfs_remove_recursive(debugfs_resctrl); + unregister_filesystem(&rdt_fs_type); + sysfs_remove_mount_point(fs_kobj, "resctrl"); + kernfs_destroy_root(rdt_root); +} diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattered.c new file mode 100644 index 000000000..f1cd1b6fb --- /dev/null +++ b/arch/x86/kernel/cpu/scattered.c @@ -0,0 +1,68 @@ +/* + * Routines to identify additional cpu features that are scattered in + * cpuid space. + */ +#include <linux/cpu.h> + +#include <asm/memtype.h> +#include <asm/apic.h> +#include <asm/processor.h> + +#include "cpu.h" + +struct cpuid_bit { + u16 feature; + u8 reg; + u8 bit; + u32 level; + u32 sub_leaf; +}; + +/* + * Please keep the leaf sorted by cpuid_bit.level for faster search. + * X86_FEATURE_MBA is supported by both Intel and AMD. But the CPUID + * levels are different and there is a separate entry for each. + */ +static const struct cpuid_bit cpuid_bits[] = { + { X86_FEATURE_APERFMPERF, CPUID_ECX, 0, 0x00000006, 0 }, + { X86_FEATURE_EPB, CPUID_ECX, 3, 0x00000006, 0 }, + { X86_FEATURE_RRSBA_CTRL, CPUID_EDX, 2, 0x00000007, 2 }, + { X86_FEATURE_CQM_LLC, CPUID_EDX, 1, 0x0000000f, 0 }, + { X86_FEATURE_CQM_OCCUP_LLC, CPUID_EDX, 0, 0x0000000f, 1 }, + { X86_FEATURE_CQM_MBM_TOTAL, CPUID_EDX, 1, 0x0000000f, 1 }, + { X86_FEATURE_CQM_MBM_LOCAL, CPUID_EDX, 2, 0x0000000f, 1 }, + { X86_FEATURE_CAT_L3, CPUID_EBX, 1, 0x00000010, 0 }, + { X86_FEATURE_CAT_L2, CPUID_EBX, 2, 0x00000010, 0 }, + { X86_FEATURE_CDP_L3, CPUID_ECX, 2, 0x00000010, 1 }, + { X86_FEATURE_CDP_L2, CPUID_ECX, 2, 0x00000010, 2 }, + { X86_FEATURE_MBA, CPUID_EBX, 3, 0x00000010, 0 }, + { X86_FEATURE_PER_THREAD_MBA, CPUID_ECX, 0, 0x00000010, 3 }, + { X86_FEATURE_HW_PSTATE, CPUID_EDX, 7, 0x80000007, 0 }, + { X86_FEATURE_CPB, CPUID_EDX, 9, 0x80000007, 0 }, + { X86_FEATURE_PROC_FEEDBACK, CPUID_EDX, 11, 0x80000007, 0 }, + { X86_FEATURE_MBA, CPUID_EBX, 6, 0x80000008, 0 }, + { 0, 0, 0, 0, 0 } +}; + +void init_scattered_cpuid_features(struct cpuinfo_x86 *c) +{ + u32 max_level; + u32 regs[4]; + const struct cpuid_bit *cb; + + for (cb = cpuid_bits; cb->feature; cb++) { + + /* Verify that the level is valid */ + max_level = cpuid_eax(cb->level & 0xffff0000); + if (max_level < cb->level || + max_level > (cb->level | 0xffff)) + continue; + + cpuid_count(cb->level, cb->sub_leaf, ®s[CPUID_EAX], + ®s[CPUID_EBX], ®s[CPUID_ECX], + ®s[CPUID_EDX]); + + if (regs[cb->reg] & (1 << cb->bit)) + set_cpu_cap(c, cb->feature); + } +} diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c new file mode 100644 index 000000000..58d17c01d --- /dev/null +++ b/arch/x86/kernel/cpu/topology.c @@ -0,0 +1,168 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Check for extended topology enumeration cpuid leaf 0xb and if it + * exists, use it for populating initial_apicid and cpu topology + * detection. + */ + +#include <linux/cpu.h> +#include <asm/apic.h> +#include <asm/memtype.h> +#include <asm/processor.h> + +#include "cpu.h" + +/* leaf 0xb SMT level */ +#define SMT_LEVEL 0 + +/* extended topology sub-leaf types */ +#define INVALID_TYPE 0 +#define SMT_TYPE 1 +#define CORE_TYPE 2 +#define DIE_TYPE 5 + +#define LEAFB_SUBTYPE(ecx) (((ecx) >> 8) & 0xff) +#define BITS_SHIFT_NEXT_LEVEL(eax) ((eax) & 0x1f) +#define LEVEL_MAX_SIBLINGS(ebx) ((ebx) & 0xffff) + +unsigned int __max_die_per_package __read_mostly = 1; +EXPORT_SYMBOL(__max_die_per_package); + +#ifdef CONFIG_SMP +/* + * Check if given CPUID extended toplogy "leaf" is implemented + */ +static int check_extended_topology_leaf(int leaf) +{ + unsigned int eax, ebx, ecx, edx; + + cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + + if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE)) + return -1; + + return 0; +} +/* + * Return best CPUID Extended Toplogy Leaf supported + */ +static int detect_extended_topology_leaf(struct cpuinfo_x86 *c) +{ + if (c->cpuid_level >= 0x1f) { + if (check_extended_topology_leaf(0x1f) == 0) + return 0x1f; + } + + if (c->cpuid_level >= 0xb) { + if (check_extended_topology_leaf(0xb) == 0) + return 0xb; + } + + return -1; +} +#endif + +int detect_extended_topology_early(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int eax, ebx, ecx, edx; + int leaf; + + leaf = detect_extended_topology_leaf(c); + if (leaf < 0) + return -1; + + set_cpu_cap(c, X86_FEATURE_XTOPOLOGY); + + cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + /* + * initial apic id, which also represents 32-bit extended x2apic id. + */ + c->initial_apicid = edx; + smp_num_siblings = max_t(int, smp_num_siblings, LEVEL_MAX_SIBLINGS(ebx)); +#endif + return 0; +} + +/* + * Check for extended topology enumeration cpuid leaf, and if it + * exists, use it for populating initial_apicid and cpu topology + * detection. + */ +int detect_extended_topology(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int eax, ebx, ecx, edx, sub_index; + unsigned int ht_mask_width, core_plus_mask_width, die_plus_mask_width; + unsigned int core_select_mask, core_level_siblings; + unsigned int die_select_mask, die_level_siblings; + unsigned int pkg_mask_width; + bool die_level_present = false; + int leaf; + + leaf = detect_extended_topology_leaf(c); + if (leaf < 0) + return -1; + + /* + * Populate HT related information from sub-leaf level 0. + */ + cpuid_count(leaf, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + c->initial_apicid = edx; + core_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + smp_num_siblings = max_t(int, smp_num_siblings, LEVEL_MAX_SIBLINGS(ebx)); + core_plus_mask_width = ht_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + die_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + pkg_mask_width = die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + + sub_index = 1; + while (true) { + cpuid_count(leaf, sub_index, &eax, &ebx, &ecx, &edx); + + /* + * Check for the Core type in the implemented sub leaves. + */ + if (LEAFB_SUBTYPE(ecx) == CORE_TYPE) { + core_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + core_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + die_level_siblings = core_level_siblings; + die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + } + if (LEAFB_SUBTYPE(ecx) == DIE_TYPE) { + die_level_present = true; + die_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + die_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + } + + if (LEAFB_SUBTYPE(ecx) != INVALID_TYPE) + pkg_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + else + break; + + sub_index++; + } + + core_select_mask = (~(-1 << pkg_mask_width)) >> ht_mask_width; + die_select_mask = (~(-1 << die_plus_mask_width)) >> + core_plus_mask_width; + + c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, + ht_mask_width) & core_select_mask; + + if (die_level_present) { + c->cpu_die_id = apic->phys_pkg_id(c->initial_apicid, + core_plus_mask_width) & die_select_mask; + } + + c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, + pkg_mask_width); + /* + * Reinit the apicid, now that we have extended initial_apicid. + */ + c->apicid = apic->phys_pkg_id(c->initial_apicid, 0); + + c->x86_max_cores = (core_level_siblings / smp_num_siblings); + __max_die_per_package = (die_level_siblings / core_level_siblings); +#endif + return 0; +} diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c new file mode 100644 index 000000000..42c939827 --- /dev/null +++ b/arch/x86/kernel/cpu/transmeta.c @@ -0,0 +1,111 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/sched/clock.h> +#include <linux/mm.h> +#include <asm/cpufeature.h> +#include <asm/msr.h> +#include "cpu.h" + +static void early_init_transmeta(struct cpuinfo_x86 *c) +{ + u32 xlvl; + + /* Transmeta-defined flags: level 0x80860001 */ + xlvl = cpuid_eax(0x80860000); + if ((xlvl & 0xffff0000) == 0x80860000) { + if (xlvl >= 0x80860001) + c->x86_capability[CPUID_8086_0001_EDX] = cpuid_edx(0x80860001); + } +} + +static void init_transmeta(struct cpuinfo_x86 *c) +{ + unsigned int cap_mask, uk, max, dummy; + unsigned int cms_rev1, cms_rev2; + unsigned int cpu_rev, cpu_freq = 0, cpu_flags, new_cpu_rev; + char cpu_info[65]; + + early_init_transmeta(c); + + cpu_detect_cache_sizes(c); + + /* Print CMS and CPU revision */ + max = cpuid_eax(0x80860000); + cpu_rev = 0; + if (max >= 0x80860001) { + cpuid(0x80860001, &dummy, &cpu_rev, &cpu_freq, &cpu_flags); + if (cpu_rev != 0x02000000) { + pr_info("CPU: Processor revision %u.%u.%u.%u, %u MHz\n", + (cpu_rev >> 24) & 0xff, + (cpu_rev >> 16) & 0xff, + (cpu_rev >> 8) & 0xff, + cpu_rev & 0xff, + cpu_freq); + } + } + if (max >= 0x80860002) { + cpuid(0x80860002, &new_cpu_rev, &cms_rev1, &cms_rev2, &dummy); + if (cpu_rev == 0x02000000) { + pr_info("CPU: Processor revision %08X, %u MHz\n", + new_cpu_rev, cpu_freq); + } + pr_info("CPU: Code Morphing Software revision %u.%u.%u-%u-%u\n", + (cms_rev1 >> 24) & 0xff, + (cms_rev1 >> 16) & 0xff, + (cms_rev1 >> 8) & 0xff, + cms_rev1 & 0xff, + cms_rev2); + } + if (max >= 0x80860006) { + cpuid(0x80860003, + (void *)&cpu_info[0], + (void *)&cpu_info[4], + (void *)&cpu_info[8], + (void *)&cpu_info[12]); + cpuid(0x80860004, + (void *)&cpu_info[16], + (void *)&cpu_info[20], + (void *)&cpu_info[24], + (void *)&cpu_info[28]); + cpuid(0x80860005, + (void *)&cpu_info[32], + (void *)&cpu_info[36], + (void *)&cpu_info[40], + (void *)&cpu_info[44]); + cpuid(0x80860006, + (void *)&cpu_info[48], + (void *)&cpu_info[52], + (void *)&cpu_info[56], + (void *)&cpu_info[60]); + cpu_info[64] = '\0'; + pr_info("CPU: %s\n", cpu_info); + } + + /* Unhide possibly hidden capability flags */ + rdmsr(0x80860004, cap_mask, uk); + wrmsr(0x80860004, ~0, uk); + c->x86_capability[CPUID_1_EDX] = cpuid_edx(0x00000001); + wrmsr(0x80860004, cap_mask, uk); + + /* All Transmeta CPUs have a constant TSC */ + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + +#ifdef CONFIG_SYSCTL + /* + * randomize_va_space slows us down enormously; + * it probably triggers retranslation of x86->native bytecode + */ + randomize_va_space = 0; +#endif +} + +static const struct cpu_dev transmeta_cpu_dev = { + .c_vendor = "Transmeta", + .c_ident = { "GenuineTMx86", "TransmetaCPU" }, + .c_early_init = early_init_transmeta, + .c_init = init_transmeta, + .c_x86_vendor = X86_VENDOR_TRANSMETA, +}; + +cpu_dev_register(transmeta_cpu_dev); diff --git a/arch/x86/kernel/cpu/tsx.c b/arch/x86/kernel/cpu/tsx.c new file mode 100644 index 000000000..da06bbb5d --- /dev/null +++ b/arch/x86/kernel/cpu/tsx.c @@ -0,0 +1,139 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Transactional Synchronization Extensions (TSX) control. + * + * Copyright (C) 2019 Intel Corporation + * + * Author: + * Pawan Gupta <pawan.kumar.gupta@linux.intel.com> + */ + +#include <linux/cpufeature.h> + +#include <asm/cmdline.h> + +#include "cpu.h" + +#undef pr_fmt +#define pr_fmt(fmt) "tsx: " fmt + +enum tsx_ctrl_states tsx_ctrl_state __ro_after_init = TSX_CTRL_NOT_SUPPORTED; + +void tsx_disable(void) +{ + u64 tsx; + + rdmsrl(MSR_IA32_TSX_CTRL, tsx); + + /* Force all transactions to immediately abort */ + tsx |= TSX_CTRL_RTM_DISABLE; + + /* + * Ensure TSX support is not enumerated in CPUID. + * This is visible to userspace and will ensure they + * do not waste resources trying TSX transactions that + * will always abort. + */ + tsx |= TSX_CTRL_CPUID_CLEAR; + + wrmsrl(MSR_IA32_TSX_CTRL, tsx); +} + +void tsx_enable(void) +{ + u64 tsx; + + rdmsrl(MSR_IA32_TSX_CTRL, tsx); + + /* Enable the RTM feature in the cpu */ + tsx &= ~TSX_CTRL_RTM_DISABLE; + + /* + * Ensure TSX support is enumerated in CPUID. + * This is visible to userspace and will ensure they + * can enumerate and use the TSX feature. + */ + tsx &= ~TSX_CTRL_CPUID_CLEAR; + + wrmsrl(MSR_IA32_TSX_CTRL, tsx); +} + +static enum tsx_ctrl_states x86_get_tsx_auto_mode(void) +{ + if (boot_cpu_has_bug(X86_BUG_TAA)) + return TSX_CTRL_DISABLE; + + return TSX_CTRL_ENABLE; +} + +void __init tsx_init(void) +{ + char arg[5] = {}; + int ret; + + /* + * TSX is controlled via MSR_IA32_TSX_CTRL. However, support for this + * MSR is enumerated by ARCH_CAP_TSX_MSR bit in MSR_IA32_ARCH_CAPABILITIES. + * + * TSX control (aka MSR_IA32_TSX_CTRL) is only available after a + * microcode update on CPUs that have their MSR_IA32_ARCH_CAPABILITIES + * bit MDS_NO=1. CPUs with MDS_NO=0 are not planned to get + * MSR_IA32_TSX_CTRL support even after a microcode update. Thus, + * tsx= cmdline requests will do nothing on CPUs without + * MSR_IA32_TSX_CTRL support. + */ + if (!(x86_read_arch_cap_msr() & ARCH_CAP_TSX_CTRL_MSR)) + return; + + setup_force_cpu_cap(X86_FEATURE_MSR_TSX_CTRL); + + ret = cmdline_find_option(boot_command_line, "tsx", arg, sizeof(arg)); + if (ret >= 0) { + if (!strcmp(arg, "on")) { + tsx_ctrl_state = TSX_CTRL_ENABLE; + } else if (!strcmp(arg, "off")) { + tsx_ctrl_state = TSX_CTRL_DISABLE; + } else if (!strcmp(arg, "auto")) { + tsx_ctrl_state = x86_get_tsx_auto_mode(); + } else { + tsx_ctrl_state = TSX_CTRL_DISABLE; + pr_err("invalid option, defaulting to off\n"); + } + } else { + /* tsx= not provided */ + if (IS_ENABLED(CONFIG_X86_INTEL_TSX_MODE_AUTO)) + tsx_ctrl_state = x86_get_tsx_auto_mode(); + else if (IS_ENABLED(CONFIG_X86_INTEL_TSX_MODE_OFF)) + tsx_ctrl_state = TSX_CTRL_DISABLE; + else + tsx_ctrl_state = TSX_CTRL_ENABLE; + } + + if (tsx_ctrl_state == TSX_CTRL_DISABLE) { + tsx_disable(); + + /* + * tsx_disable() will change the state of the RTM and HLE CPUID + * bits. Clear them here since they are now expected to be not + * set. + */ + setup_clear_cpu_cap(X86_FEATURE_RTM); + setup_clear_cpu_cap(X86_FEATURE_HLE); + } else if (tsx_ctrl_state == TSX_CTRL_ENABLE) { + + /* + * HW defaults TSX to be enabled at bootup. + * We may still need the TSX enable support + * during init for special cases like + * kexec after TSX is disabled. + */ + tsx_enable(); + + /* + * tsx_enable() will change the state of the RTM and HLE CPUID + * bits. Force them here since they are now expected to be set. + */ + setup_force_cpu_cap(X86_FEATURE_RTM); + setup_force_cpu_cap(X86_FEATURE_HLE); + } +} diff --git a/arch/x86/kernel/cpu/umc.c b/arch/x86/kernel/cpu/umc.c new file mode 100644 index 000000000..65a58a390 --- /dev/null +++ b/arch/x86/kernel/cpu/umc.c @@ -0,0 +1,26 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/kernel.h> +#include <asm/processor.h> +#include "cpu.h" + +/* + * UMC chips appear to be only either 386 or 486, + * so no special init takes place. + */ + +static const struct cpu_dev umc_cpu_dev = { + .c_vendor = "UMC", + .c_ident = { "UMC UMC UMC" }, + .legacy_models = { + { .family = 4, .model_names = + { + [1] = "U5D", + [2] = "U5S", + } + }, + }, + .c_x86_vendor = X86_VENDOR_UMC, +}; + +cpu_dev_register(umc_cpu_dev); + diff --git a/arch/x86/kernel/cpu/umwait.c b/arch/x86/kernel/cpu/umwait.c new file mode 100644 index 000000000..ec8064c0a --- /dev/null +++ b/arch/x86/kernel/cpu/umwait.c @@ -0,0 +1,238 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/syscore_ops.h> +#include <linux/suspend.h> +#include <linux/cpu.h> + +#include <asm/msr.h> +#include <asm/mwait.h> + +#define UMWAIT_C02_ENABLE 0 + +#define UMWAIT_CTRL_VAL(max_time, c02_disable) \ + (((max_time) & MSR_IA32_UMWAIT_CONTROL_TIME_MASK) | \ + ((c02_disable) & MSR_IA32_UMWAIT_CONTROL_C02_DISABLE)) + +/* + * Cache IA32_UMWAIT_CONTROL MSR. This is a systemwide control. By default, + * umwait max time is 100000 in TSC-quanta and C0.2 is enabled + */ +static u32 umwait_control_cached = UMWAIT_CTRL_VAL(100000, UMWAIT_C02_ENABLE); + +/* + * Cache the original IA32_UMWAIT_CONTROL MSR value which is configured by + * hardware or BIOS before kernel boot. + */ +static u32 orig_umwait_control_cached __ro_after_init; + +/* + * Serialize access to umwait_control_cached and IA32_UMWAIT_CONTROL MSR in + * the sysfs write functions. + */ +static DEFINE_MUTEX(umwait_lock); + +static void umwait_update_control_msr(void * unused) +{ + lockdep_assert_irqs_disabled(); + wrmsr(MSR_IA32_UMWAIT_CONTROL, READ_ONCE(umwait_control_cached), 0); +} + +/* + * The CPU hotplug callback sets the control MSR to the global control + * value. + * + * Disable interrupts so the read of umwait_control_cached and the WRMSR + * are protected against a concurrent sysfs write. Otherwise the sysfs + * write could update the cached value after it had been read on this CPU + * and issue the IPI before the old value had been written. The IPI would + * interrupt, write the new value and after return from IPI the previous + * value would be written by this CPU. + * + * With interrupts disabled the upcoming CPU either sees the new control + * value or the IPI is updating this CPU to the new control value after + * interrupts have been reenabled. + */ +static int umwait_cpu_online(unsigned int cpu) +{ + local_irq_disable(); + umwait_update_control_msr(NULL); + local_irq_enable(); + return 0; +} + +/* + * The CPU hotplug callback sets the control MSR to the original control + * value. + */ +static int umwait_cpu_offline(unsigned int cpu) +{ + /* + * This code is protected by the CPU hotplug already and + * orig_umwait_control_cached is never changed after it caches + * the original control MSR value in umwait_init(). So there + * is no race condition here. + */ + wrmsr(MSR_IA32_UMWAIT_CONTROL, orig_umwait_control_cached, 0); + + return 0; +} + +/* + * On resume, restore IA32_UMWAIT_CONTROL MSR on the boot processor which + * is the only active CPU at this time. The MSR is set up on the APs via the + * CPU hotplug callback. + * + * This function is invoked on resume from suspend and hibernation. On + * resume from suspend the restore should be not required, but we neither + * trust the firmware nor does it matter if the same value is written + * again. + */ +static void umwait_syscore_resume(void) +{ + umwait_update_control_msr(NULL); +} + +static struct syscore_ops umwait_syscore_ops = { + .resume = umwait_syscore_resume, +}; + +/* sysfs interface */ + +/* + * When bit 0 in IA32_UMWAIT_CONTROL MSR is 1, C0.2 is disabled. + * Otherwise, C0.2 is enabled. + */ +static inline bool umwait_ctrl_c02_enabled(u32 ctrl) +{ + return !(ctrl & MSR_IA32_UMWAIT_CONTROL_C02_DISABLE); +} + +static inline u32 umwait_ctrl_max_time(u32 ctrl) +{ + return ctrl & MSR_IA32_UMWAIT_CONTROL_TIME_MASK; +} + +static inline void umwait_update_control(u32 maxtime, bool c02_enable) +{ + u32 ctrl = maxtime & MSR_IA32_UMWAIT_CONTROL_TIME_MASK; + + if (!c02_enable) + ctrl |= MSR_IA32_UMWAIT_CONTROL_C02_DISABLE; + + WRITE_ONCE(umwait_control_cached, ctrl); + /* Propagate to all CPUs */ + on_each_cpu(umwait_update_control_msr, NULL, 1); +} + +static ssize_t +enable_c02_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + u32 ctrl = READ_ONCE(umwait_control_cached); + + return sprintf(buf, "%d\n", umwait_ctrl_c02_enabled(ctrl)); +} + +static ssize_t enable_c02_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + bool c02_enable; + u32 ctrl; + int ret; + + ret = kstrtobool(buf, &c02_enable); + if (ret) + return ret; + + mutex_lock(&umwait_lock); + + ctrl = READ_ONCE(umwait_control_cached); + if (c02_enable != umwait_ctrl_c02_enabled(ctrl)) + umwait_update_control(ctrl, c02_enable); + + mutex_unlock(&umwait_lock); + + return count; +} +static DEVICE_ATTR_RW(enable_c02); + +static ssize_t +max_time_show(struct device *kobj, struct device_attribute *attr, char *buf) +{ + u32 ctrl = READ_ONCE(umwait_control_cached); + + return sprintf(buf, "%u\n", umwait_ctrl_max_time(ctrl)); +} + +static ssize_t max_time_store(struct device *kobj, + struct device_attribute *attr, + const char *buf, size_t count) +{ + u32 max_time, ctrl; + int ret; + + ret = kstrtou32(buf, 0, &max_time); + if (ret) + return ret; + + /* bits[1:0] must be zero */ + if (max_time & ~MSR_IA32_UMWAIT_CONTROL_TIME_MASK) + return -EINVAL; + + mutex_lock(&umwait_lock); + + ctrl = READ_ONCE(umwait_control_cached); + if (max_time != umwait_ctrl_max_time(ctrl)) + umwait_update_control(max_time, umwait_ctrl_c02_enabled(ctrl)); + + mutex_unlock(&umwait_lock); + + return count; +} +static DEVICE_ATTR_RW(max_time); + +static struct attribute *umwait_attrs[] = { + &dev_attr_enable_c02.attr, + &dev_attr_max_time.attr, + NULL +}; + +static struct attribute_group umwait_attr_group = { + .attrs = umwait_attrs, + .name = "umwait_control", +}; + +static int __init umwait_init(void) +{ + struct device *dev; + int ret; + + if (!boot_cpu_has(X86_FEATURE_WAITPKG)) + return -ENODEV; + + /* + * Cache the original control MSR value before the control MSR is + * changed. This is the only place where orig_umwait_control_cached + * is modified. + */ + rdmsrl(MSR_IA32_UMWAIT_CONTROL, orig_umwait_control_cached); + + ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "umwait:online", + umwait_cpu_online, umwait_cpu_offline); + if (ret < 0) { + /* + * On failure, the control MSR on all CPUs has the + * original control value. + */ + return ret; + } + + register_syscore_ops(&umwait_syscore_ops); + + /* + * Add umwait control interface. Ignore failure, so at least the + * default values are set up in case the machine manages to boot. + */ + dev = cpu_subsys.dev_root; + return sysfs_create_group(&dev->kobj, &umwait_attr_group); +} +device_initcall(umwait_init); diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c new file mode 100644 index 000000000..924571fe5 --- /dev/null +++ b/arch/x86/kernel/cpu/vmware.c @@ -0,0 +1,525 @@ +/* + * VMware Detection code. + * + * Copyright (C) 2008, VMware, Inc. + * Author : Alok N Kataria <akataria@vmware.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ + +#include <linux/dmi.h> +#include <linux/init.h> +#include <linux/export.h> +#include <linux/clocksource.h> +#include <linux/cpu.h> +#include <linux/reboot.h> +#include <asm/div64.h> +#include <asm/x86_init.h> +#include <asm/hypervisor.h> +#include <asm/timer.h> +#include <asm/apic.h> +#include <asm/vmware.h> +#include <asm/svm.h> + +#undef pr_fmt +#define pr_fmt(fmt) "vmware: " fmt + +#define CPUID_VMWARE_INFO_LEAF 0x40000000 +#define CPUID_VMWARE_FEATURES_LEAF 0x40000010 +#define CPUID_VMWARE_FEATURES_ECX_VMMCALL BIT(0) +#define CPUID_VMWARE_FEATURES_ECX_VMCALL BIT(1) + +#define VMWARE_HYPERVISOR_MAGIC 0x564D5868 + +#define VMWARE_CMD_GETVERSION 10 +#define VMWARE_CMD_GETHZ 45 +#define VMWARE_CMD_GETVCPU_INFO 68 +#define VMWARE_CMD_LEGACY_X2APIC 3 +#define VMWARE_CMD_VCPU_RESERVED 31 +#define VMWARE_CMD_STEALCLOCK 91 + +#define STEALCLOCK_NOT_AVAILABLE (-1) +#define STEALCLOCK_DISABLED 0 +#define STEALCLOCK_ENABLED 1 + +#define VMWARE_PORT(cmd, eax, ebx, ecx, edx) \ + __asm__("inl (%%dx), %%eax" : \ + "=a"(eax), "=c"(ecx), "=d"(edx), "=b"(ebx) : \ + "a"(VMWARE_HYPERVISOR_MAGIC), \ + "c"(VMWARE_CMD_##cmd), \ + "d"(VMWARE_HYPERVISOR_PORT), "b"(UINT_MAX) : \ + "memory") + +#define VMWARE_VMCALL(cmd, eax, ebx, ecx, edx) \ + __asm__("vmcall" : \ + "=a"(eax), "=c"(ecx), "=d"(edx), "=b"(ebx) : \ + "a"(VMWARE_HYPERVISOR_MAGIC), \ + "c"(VMWARE_CMD_##cmd), \ + "d"(0), "b"(UINT_MAX) : \ + "memory") + +#define VMWARE_VMMCALL(cmd, eax, ebx, ecx, edx) \ + __asm__("vmmcall" : \ + "=a"(eax), "=c"(ecx), "=d"(edx), "=b"(ebx) : \ + "a"(VMWARE_HYPERVISOR_MAGIC), \ + "c"(VMWARE_CMD_##cmd), \ + "d"(0), "b"(UINT_MAX) : \ + "memory") + +#define VMWARE_CMD(cmd, eax, ebx, ecx, edx) do { \ + switch (vmware_hypercall_mode) { \ + case CPUID_VMWARE_FEATURES_ECX_VMCALL: \ + VMWARE_VMCALL(cmd, eax, ebx, ecx, edx); \ + break; \ + case CPUID_VMWARE_FEATURES_ECX_VMMCALL: \ + VMWARE_VMMCALL(cmd, eax, ebx, ecx, edx); \ + break; \ + default: \ + VMWARE_PORT(cmd, eax, ebx, ecx, edx); \ + break; \ + } \ + } while (0) + +struct vmware_steal_time { + union { + uint64_t clock; /* stolen time counter in units of vtsc */ + struct { + /* only for little-endian */ + uint32_t clock_low; + uint32_t clock_high; + }; + }; + uint64_t reserved[7]; +}; + +static unsigned long vmware_tsc_khz __ro_after_init; +static u8 vmware_hypercall_mode __ro_after_init; + +static inline int __vmware_platform(void) +{ + uint32_t eax, ebx, ecx, edx; + VMWARE_CMD(GETVERSION, eax, ebx, ecx, edx); + return eax != (uint32_t)-1 && ebx == VMWARE_HYPERVISOR_MAGIC; +} + +static unsigned long vmware_get_tsc_khz(void) +{ + return vmware_tsc_khz; +} + +#ifdef CONFIG_PARAVIRT +static struct cyc2ns_data vmware_cyc2ns __ro_after_init; +static bool vmw_sched_clock __initdata = true; +static DEFINE_PER_CPU_DECRYPTED(struct vmware_steal_time, vmw_steal_time) __aligned(64); +static bool has_steal_clock; +static bool steal_acc __initdata = true; /* steal time accounting */ + +static __init int setup_vmw_sched_clock(char *s) +{ + vmw_sched_clock = false; + return 0; +} +early_param("no-vmw-sched-clock", setup_vmw_sched_clock); + +static __init int parse_no_stealacc(char *arg) +{ + steal_acc = false; + return 0; +} +early_param("no-steal-acc", parse_no_stealacc); + +static unsigned long long notrace vmware_sched_clock(void) +{ + unsigned long long ns; + + ns = mul_u64_u32_shr(rdtsc(), vmware_cyc2ns.cyc2ns_mul, + vmware_cyc2ns.cyc2ns_shift); + ns -= vmware_cyc2ns.cyc2ns_offset; + return ns; +} + +static void __init vmware_cyc2ns_setup(void) +{ + struct cyc2ns_data *d = &vmware_cyc2ns; + unsigned long long tsc_now = rdtsc(); + + clocks_calc_mult_shift(&d->cyc2ns_mul, &d->cyc2ns_shift, + vmware_tsc_khz, NSEC_PER_MSEC, 0); + d->cyc2ns_offset = mul_u64_u32_shr(tsc_now, d->cyc2ns_mul, + d->cyc2ns_shift); + + pr_info("using clock offset of %llu ns\n", d->cyc2ns_offset); +} + +static int vmware_cmd_stealclock(uint32_t arg1, uint32_t arg2) +{ + uint32_t result, info; + + asm volatile (VMWARE_HYPERCALL : + "=a"(result), + "=c"(info) : + "a"(VMWARE_HYPERVISOR_MAGIC), + "b"(0), + "c"(VMWARE_CMD_STEALCLOCK), + "d"(0), + "S"(arg1), + "D"(arg2) : + "memory"); + return result; +} + +static bool stealclock_enable(phys_addr_t pa) +{ + return vmware_cmd_stealclock(upper_32_bits(pa), + lower_32_bits(pa)) == STEALCLOCK_ENABLED; +} + +static int __stealclock_disable(void) +{ + return vmware_cmd_stealclock(0, 1); +} + +static void stealclock_disable(void) +{ + __stealclock_disable(); +} + +static bool vmware_is_stealclock_available(void) +{ + return __stealclock_disable() != STEALCLOCK_NOT_AVAILABLE; +} + +/** + * vmware_steal_clock() - read the per-cpu steal clock + * @cpu: the cpu number whose steal clock we want to read + * + * The function reads the steal clock if we are on a 64-bit system, otherwise + * reads it in parts, checking that the high part didn't change in the + * meantime. + * + * Return: + * The steal clock reading in ns. + */ +static uint64_t vmware_steal_clock(int cpu) +{ + struct vmware_steal_time *steal = &per_cpu(vmw_steal_time, cpu); + uint64_t clock; + + if (IS_ENABLED(CONFIG_64BIT)) + clock = READ_ONCE(steal->clock); + else { + uint32_t initial_high, low, high; + + do { + initial_high = READ_ONCE(steal->clock_high); + /* Do not reorder initial_high and high readings */ + virt_rmb(); + low = READ_ONCE(steal->clock_low); + /* Keep low reading in between */ + virt_rmb(); + high = READ_ONCE(steal->clock_high); + } while (initial_high != high); + + clock = ((uint64_t)high << 32) | low; + } + + return mul_u64_u32_shr(clock, vmware_cyc2ns.cyc2ns_mul, + vmware_cyc2ns.cyc2ns_shift); +} + +static void vmware_register_steal_time(void) +{ + int cpu = smp_processor_id(); + struct vmware_steal_time *st = &per_cpu(vmw_steal_time, cpu); + + if (!has_steal_clock) + return; + + if (!stealclock_enable(slow_virt_to_phys(st))) { + has_steal_clock = false; + return; + } + + pr_info("vmware-stealtime: cpu %d, pa %llx\n", + cpu, (unsigned long long) slow_virt_to_phys(st)); +} + +static void vmware_disable_steal_time(void) +{ + if (!has_steal_clock) + return; + + stealclock_disable(); +} + +static void vmware_guest_cpu_init(void) +{ + if (has_steal_clock) + vmware_register_steal_time(); +} + +static void vmware_pv_guest_cpu_reboot(void *unused) +{ + vmware_disable_steal_time(); +} + +static int vmware_pv_reboot_notify(struct notifier_block *nb, + unsigned long code, void *unused) +{ + if (code == SYS_RESTART) + on_each_cpu(vmware_pv_guest_cpu_reboot, NULL, 1); + return NOTIFY_DONE; +} + +static struct notifier_block vmware_pv_reboot_nb = { + .notifier_call = vmware_pv_reboot_notify, +}; + +#ifdef CONFIG_SMP +static void __init vmware_smp_prepare_boot_cpu(void) +{ + vmware_guest_cpu_init(); + native_smp_prepare_boot_cpu(); +} + +static int vmware_cpu_online(unsigned int cpu) +{ + local_irq_disable(); + vmware_guest_cpu_init(); + local_irq_enable(); + return 0; +} + +static int vmware_cpu_down_prepare(unsigned int cpu) +{ + local_irq_disable(); + vmware_disable_steal_time(); + local_irq_enable(); + return 0; +} +#endif + +static __init int activate_jump_labels(void) +{ + if (has_steal_clock) { + static_key_slow_inc(¶virt_steal_enabled); + if (steal_acc) + static_key_slow_inc(¶virt_steal_rq_enabled); + } + + return 0; +} +arch_initcall(activate_jump_labels); + +static void __init vmware_paravirt_ops_setup(void) +{ + pv_info.name = "VMware hypervisor"; + pv_ops.cpu.io_delay = paravirt_nop; + + if (vmware_tsc_khz == 0) + return; + + vmware_cyc2ns_setup(); + + if (vmw_sched_clock) + pv_ops.time.sched_clock = vmware_sched_clock; + + if (vmware_is_stealclock_available()) { + has_steal_clock = true; + pv_ops.time.steal_clock = vmware_steal_clock; + + /* We use reboot notifier only to disable steal clock */ + register_reboot_notifier(&vmware_pv_reboot_nb); + +#ifdef CONFIG_SMP + smp_ops.smp_prepare_boot_cpu = + vmware_smp_prepare_boot_cpu; + if (cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN, + "x86/vmware:online", + vmware_cpu_online, + vmware_cpu_down_prepare) < 0) + pr_err("vmware_guest: Failed to install cpu hotplug callbacks\n"); +#else + vmware_guest_cpu_init(); +#endif + } +} +#else +#define vmware_paravirt_ops_setup() do {} while (0) +#endif + +/* + * VMware hypervisor takes care of exporting a reliable TSC to the guest. + * Still, due to timing difference when running on virtual cpus, the TSC can + * be marked as unstable in some cases. For example, the TSC sync check at + * bootup can fail due to a marginal offset between vcpus' TSCs (though the + * TSCs do not drift from each other). Also, the ACPI PM timer clocksource + * is not suitable as a watchdog when running on a hypervisor because the + * kernel may miss a wrap of the counter if the vcpu is descheduled for a + * long time. To skip these checks at runtime we set these capability bits, + * so that the kernel could just trust the hypervisor with providing a + * reliable virtual TSC that is suitable for timekeeping. + */ +static void __init vmware_set_capabilities(void) +{ + setup_force_cpu_cap(X86_FEATURE_CONSTANT_TSC); + setup_force_cpu_cap(X86_FEATURE_TSC_RELIABLE); + if (vmware_hypercall_mode == CPUID_VMWARE_FEATURES_ECX_VMCALL) + setup_force_cpu_cap(X86_FEATURE_VMCALL); + else if (vmware_hypercall_mode == CPUID_VMWARE_FEATURES_ECX_VMMCALL) + setup_force_cpu_cap(X86_FEATURE_VMW_VMMCALL); +} + +static void __init vmware_platform_setup(void) +{ + uint32_t eax, ebx, ecx, edx; + uint64_t lpj, tsc_khz; + + VMWARE_CMD(GETHZ, eax, ebx, ecx, edx); + + if (ebx != UINT_MAX) { + lpj = tsc_khz = eax | (((uint64_t)ebx) << 32); + do_div(tsc_khz, 1000); + WARN_ON(tsc_khz >> 32); + pr_info("TSC freq read from hypervisor : %lu.%03lu MHz\n", + (unsigned long) tsc_khz / 1000, + (unsigned long) tsc_khz % 1000); + + if (!preset_lpj) { + do_div(lpj, HZ); + preset_lpj = lpj; + } + + vmware_tsc_khz = tsc_khz; + x86_platform.calibrate_tsc = vmware_get_tsc_khz; + x86_platform.calibrate_cpu = vmware_get_tsc_khz; + +#ifdef CONFIG_X86_LOCAL_APIC + /* Skip lapic calibration since we know the bus frequency. */ + lapic_timer_period = ecx / HZ; + pr_info("Host bus clock speed read from hypervisor : %u Hz\n", + ecx); +#endif + } else { + pr_warn("Failed to get TSC freq from the hypervisor\n"); + } + + vmware_paravirt_ops_setup(); + +#ifdef CONFIG_X86_IO_APIC + no_timer_check = 1; +#endif + + vmware_set_capabilities(); +} + +static u8 __init vmware_select_hypercall(void) +{ + int eax, ebx, ecx, edx; + + cpuid(CPUID_VMWARE_FEATURES_LEAF, &eax, &ebx, &ecx, &edx); + return (ecx & (CPUID_VMWARE_FEATURES_ECX_VMMCALL | + CPUID_VMWARE_FEATURES_ECX_VMCALL)); +} + +/* + * While checking the dmi string information, just checking the product + * serial key should be enough, as this will always have a VMware + * specific string when running under VMware hypervisor. + * If !boot_cpu_has(X86_FEATURE_HYPERVISOR), vmware_hypercall_mode + * intentionally defaults to 0. + */ +static uint32_t __init vmware_platform(void) +{ + if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) { + unsigned int eax; + unsigned int hyper_vendor_id[3]; + + cpuid(CPUID_VMWARE_INFO_LEAF, &eax, &hyper_vendor_id[0], + &hyper_vendor_id[1], &hyper_vendor_id[2]); + if (!memcmp(hyper_vendor_id, "VMwareVMware", 12)) { + if (eax >= CPUID_VMWARE_FEATURES_LEAF) + vmware_hypercall_mode = + vmware_select_hypercall(); + + pr_info("hypercall mode: 0x%02x\n", + (unsigned int) vmware_hypercall_mode); + + return CPUID_VMWARE_INFO_LEAF; + } + } else if (dmi_available && dmi_name_in_serial("VMware") && + __vmware_platform()) + return 1; + + return 0; +} + +/* Checks if hypervisor supports x2apic without VT-D interrupt remapping. */ +static bool __init vmware_legacy_x2apic_available(void) +{ + uint32_t eax, ebx, ecx, edx; + VMWARE_CMD(GETVCPU_INFO, eax, ebx, ecx, edx); + return (eax & (1 << VMWARE_CMD_VCPU_RESERVED)) == 0 && + (eax & (1 << VMWARE_CMD_LEGACY_X2APIC)) != 0; +} + +#ifdef CONFIG_AMD_MEM_ENCRYPT +static void vmware_sev_es_hcall_prepare(struct ghcb *ghcb, + struct pt_regs *regs) +{ + /* Copy VMWARE specific Hypercall parameters to the GHCB */ + ghcb_set_rip(ghcb, regs->ip); + ghcb_set_rbx(ghcb, regs->bx); + ghcb_set_rcx(ghcb, regs->cx); + ghcb_set_rdx(ghcb, regs->dx); + ghcb_set_rsi(ghcb, regs->si); + ghcb_set_rdi(ghcb, regs->di); + ghcb_set_rbp(ghcb, regs->bp); +} + +static bool vmware_sev_es_hcall_finish(struct ghcb *ghcb, struct pt_regs *regs) +{ + if (!(ghcb_rbx_is_valid(ghcb) && + ghcb_rcx_is_valid(ghcb) && + ghcb_rdx_is_valid(ghcb) && + ghcb_rsi_is_valid(ghcb) && + ghcb_rdi_is_valid(ghcb) && + ghcb_rbp_is_valid(ghcb))) + return false; + + regs->bx = ghcb->save.rbx; + regs->cx = ghcb->save.rcx; + regs->dx = ghcb->save.rdx; + regs->si = ghcb->save.rsi; + regs->di = ghcb->save.rdi; + regs->bp = ghcb->save.rbp; + + return true; +} +#endif + +const __initconst struct hypervisor_x86 x86_hyper_vmware = { + .name = "VMware", + .detect = vmware_platform, + .type = X86_HYPER_VMWARE, + .init.init_platform = vmware_platform_setup, + .init.x2apic_available = vmware_legacy_x2apic_available, +#ifdef CONFIG_AMD_MEM_ENCRYPT + .runtime.sev_es_hcall_prepare = vmware_sev_es_hcall_prepare, + .runtime.sev_es_hcall_finish = vmware_sev_es_hcall_finish, +#endif +}; diff --git a/arch/x86/kernel/cpu/zhaoxin.c b/arch/x86/kernel/cpu/zhaoxin.c new file mode 100644 index 000000000..05fa4ef63 --- /dev/null +++ b/arch/x86/kernel/cpu/zhaoxin.c @@ -0,0 +1,133 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/sched.h> +#include <linux/sched/clock.h> + +#include <asm/cpu.h> +#include <asm/cpufeature.h> + +#include "cpu.h" + +#define MSR_ZHAOXIN_FCR57 0x00001257 + +#define ACE_PRESENT (1 << 6) +#define ACE_ENABLED (1 << 7) +#define ACE_FCR (1 << 7) /* MSR_ZHAOXIN_FCR */ + +#define RNG_PRESENT (1 << 2) +#define RNG_ENABLED (1 << 3) +#define RNG_ENABLE (1 << 8) /* MSR_ZHAOXIN_RNG */ + +static void init_zhaoxin_cap(struct cpuinfo_x86 *c) +{ + u32 lo, hi; + + /* Test for Extended Feature Flags presence */ + if (cpuid_eax(0xC0000000) >= 0xC0000001) { + u32 tmp = cpuid_edx(0xC0000001); + + /* Enable ACE unit, if present and disabled */ + if ((tmp & (ACE_PRESENT | ACE_ENABLED)) == ACE_PRESENT) { + rdmsr(MSR_ZHAOXIN_FCR57, lo, hi); + /* Enable ACE unit */ + lo |= ACE_FCR; + wrmsr(MSR_ZHAOXIN_FCR57, lo, hi); + pr_info("CPU: Enabled ACE h/w crypto\n"); + } + + /* Enable RNG unit, if present and disabled */ + if ((tmp & (RNG_PRESENT | RNG_ENABLED)) == RNG_PRESENT) { + rdmsr(MSR_ZHAOXIN_FCR57, lo, hi); + /* Enable RNG unit */ + lo |= RNG_ENABLE; + wrmsr(MSR_ZHAOXIN_FCR57, lo, hi); + pr_info("CPU: Enabled h/w RNG\n"); + } + + /* + * Store Extended Feature Flags as word 5 of the CPU + * capability bit array + */ + c->x86_capability[CPUID_C000_0001_EDX] = cpuid_edx(0xC0000001); + } + + if (c->x86 >= 0x6) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); +} + +static void early_init_zhaoxin(struct cpuinfo_x86 *c) +{ + if (c->x86 >= 0x6) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +#endif + if (c->x86_power & (1 << 8)) { + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + } + + if (c->cpuid_level >= 0x00000001) { + u32 eax, ebx, ecx, edx; + + cpuid(0x00000001, &eax, &ebx, &ecx, &edx); + /* + * If HTT (EDX[28]) is set EBX[16:23] contain the number of + * apicids which are reserved per package. Store the resulting + * shift value for the package management code. + */ + if (edx & (1U << 28)) + c->x86_coreid_bits = get_count_order((ebx >> 16) & 0xff); + } + +} + +static void init_zhaoxin(struct cpuinfo_x86 *c) +{ + early_init_zhaoxin(c); + init_intel_cacheinfo(c); + detect_num_cpu_cores(c); +#ifdef CONFIG_X86_32 + detect_ht(c); +#endif + + if (c->cpuid_level > 9) { + unsigned int eax = cpuid_eax(10); + + /* + * Check for version and the number of counters + * Version(eax[7:0]) can't be 0; + * Counters(eax[15:8]) should be greater than 1; + */ + if ((eax & 0xff) && (((eax >> 8) & 0xff) > 1)) + set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); + } + + if (c->x86 >= 0x6) + init_zhaoxin_cap(c); +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); +#endif + + init_ia32_feat_ctl(c); +} + +#ifdef CONFIG_X86_32 +static unsigned int +zhaoxin_size_cache(struct cpuinfo_x86 *c, unsigned int size) +{ + return size; +} +#endif + +static const struct cpu_dev zhaoxin_cpu_dev = { + .c_vendor = "zhaoxin", + .c_ident = { " Shanghai " }, + .c_early_init = early_init_zhaoxin, + .c_init = init_zhaoxin, +#ifdef CONFIG_X86_32 + .legacy_cache_size = zhaoxin_size_cache, +#endif + .c_x86_vendor = X86_VENDOR_ZHAOXIN, +}; + +cpu_dev_register(zhaoxin_cpu_dev); |