From 9f0fc191371843c4fc000a226b0a26b6c059aacd Mon Sep 17 00:00:00 2001 From: Daniel Baumann Date: Sat, 18 May 2024 19:40:19 +0200 Subject: Merging upstream version 6.7.7. Signed-off-by: Daniel Baumann --- arch/s390/kernel/Makefile | 2 +- arch/s390/kernel/crash_dump.c | 6 +- arch/s390/kernel/ctlreg.c | 121 +++++++++++++++++++++++++++ arch/s390/kernel/debug.c | 1 - arch/s390/kernel/diag.c | 3 +- arch/s390/kernel/early.c | 23 ++---- arch/s390/kernel/guarded_storage.c | 6 +- arch/s390/kernel/ipl.c | 2 +- arch/s390/kernel/irq.c | 4 +- arch/s390/kernel/kprobes.c | 21 +++-- arch/s390/kernel/machine_kexec.c | 6 +- arch/s390/kernel/nmi.c | 24 +++--- arch/s390/kernel/perf_cpum_cf.c | 2 +- arch/s390/kernel/perf_event.c | 41 ++++++++++ arch/s390/kernel/perf_pai_crypto.c | 149 +++++++++++++++++++++++++--------- arch/s390/kernel/perf_pai_ext.c | 11 +-- arch/s390/kernel/ptrace.c | 53 ++++++------ arch/s390/kernel/setup.c | 27 +++--- arch/s390/kernel/smp.c | 134 ++++-------------------------- arch/s390/kernel/stacktrace.c | 43 ++++++++++ arch/s390/kernel/syscalls/syscall.tbl | 6 +- arch/s390/kernel/time.c | 4 +- arch/s390/kernel/topology.c | 3 +- arch/s390/kernel/vdso32/Makefile | 10 --- arch/s390/kernel/vdso64/Makefile | 10 --- 25 files changed, 431 insertions(+), 281 deletions(-) create mode 100644 arch/s390/kernel/ctlreg.c (limited to 'arch/s390/kernel') diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 0df2b88cc0..353def9397 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -37,7 +37,7 @@ CFLAGS_unwind_bc.o += -fno-optimize-sibling-calls obj-y := head64.o traps.o time.o process.o earlypgm.o early.o setup.o idle.o vtime.o obj-y += processor.o syscall.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o obj-y += debug.o irq.o ipl.o dis.o diag.o vdso.o cpufeature.o -obj-y += sysinfo.o lgr.o os_info.o +obj-y += sysinfo.o lgr.o os_info.o ctlreg.o obj-y += runtime_instr.o cache.o fpu.o dumpstack.o guarded_storage.o sthyi.o obj-y += entry.o reipl.o kdebugfs.o alternative.o obj-y += nospec-branch.o ipl_vmparm.o machine_kexec_reloc.o unwind_bc.o diff --git a/arch/s390/kernel/crash_dump.c b/arch/s390/kernel/crash_dump.c index 7af69948b2..514feadd4c 100644 --- a/arch/s390/kernel/crash_dump.c +++ b/arch/s390/kernel/crash_dump.c @@ -498,7 +498,7 @@ static int get_mem_chunk_cnt(void) /* * Initialize ELF loads (new kernel) */ -static void loads_init(Elf64_Phdr *phdr, u64 loads_offset) +static void loads_init(Elf64_Phdr *phdr) { phys_addr_t start, end; u64 idx; @@ -507,7 +507,7 @@ static void loads_init(Elf64_Phdr *phdr, u64 loads_offset) phdr->p_filesz = end - start; phdr->p_type = PT_LOAD; phdr->p_offset = start; - phdr->p_vaddr = start; + phdr->p_vaddr = (unsigned long)__va(start); phdr->p_paddr = start; phdr->p_memsz = end - start; phdr->p_flags = PF_R | PF_W | PF_X; @@ -612,7 +612,7 @@ int elfcorehdr_alloc(unsigned long long *addr, unsigned long long *size) ptr = notes_init(phdr_notes, ptr, ((unsigned long) hdr) + hdr_off); /* Init loads */ hdr_off = PTR_DIFF(ptr, hdr); - loads_init(phdr_loads, hdr_off); + loads_init(phdr_loads); *addr = (unsigned long long) hdr; *size = (unsigned long long) hdr_off; BUG_ON(elfcorehdr_size > alloc_size); diff --git a/arch/s390/kernel/ctlreg.c b/arch/s390/kernel/ctlreg.c new file mode 100644 index 0000000000..8cc26cf2c6 --- /dev/null +++ b/arch/s390/kernel/ctlreg.c @@ -0,0 +1,121 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright IBM Corp. 1999, 2023 + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * ctl_lock guards access to global control register contents which + * are kept in the control register save area within absolute lowcore + * at physical address zero. + */ +static DEFINE_SPINLOCK(system_ctl_lock); + +void system_ctlreg_lock(void) + __acquires(&system_ctl_lock) +{ + spin_lock(&system_ctl_lock); +} + +void system_ctlreg_unlock(void) + __releases(&system_ctl_lock) +{ + spin_unlock(&system_ctl_lock); +} + +static bool system_ctlreg_area_init __ro_after_init; + +void __init system_ctlreg_init_save_area(struct lowcore *lc) +{ + struct lowcore *abs_lc; + + abs_lc = get_abs_lowcore(); + __local_ctl_store(0, 15, lc->cregs_save_area); + __local_ctl_store(0, 15, abs_lc->cregs_save_area); + put_abs_lowcore(abs_lc); + system_ctlreg_area_init = true; +} + +struct ctlreg_parms { + unsigned long andval; + unsigned long orval; + unsigned long val; + int request; + int cr; +}; + +static void ctlreg_callback(void *info) +{ + struct ctlreg_parms *pp = info; + struct ctlreg regs[16]; + + __local_ctl_store(0, 15, regs); + if (pp->request == CTLREG_LOAD) { + regs[pp->cr].val = pp->val; + } else { + regs[pp->cr].val &= pp->andval; + regs[pp->cr].val |= pp->orval; + } + __local_ctl_load(0, 15, regs); +} + +static void system_ctlreg_update(void *info) +{ + unsigned long flags; + + if (system_state == SYSTEM_BOOTING) { + /* + * For very early calls do not call on_each_cpu() + * since not everything might be setup. + */ + local_irq_save(flags); + ctlreg_callback(info); + local_irq_restore(flags); + } else { + on_each_cpu(ctlreg_callback, info, 1); + } +} + +void system_ctlreg_modify(unsigned int cr, unsigned long data, int request) +{ + struct ctlreg_parms pp = { .cr = cr, .request = request, }; + struct lowcore *abs_lc; + + switch (request) { + case CTLREG_SET_BIT: + pp.orval = 1UL << data; + pp.andval = -1UL; + break; + case CTLREG_CLEAR_BIT: + pp.orval = 0; + pp.andval = ~(1UL << data); + break; + case CTLREG_LOAD: + pp.val = data; + break; + } + if (system_ctlreg_area_init) { + system_ctlreg_lock(); + abs_lc = get_abs_lowcore(); + if (request == CTLREG_LOAD) { + abs_lc->cregs_save_area[cr].val = pp.val; + } else { + abs_lc->cregs_save_area[cr].val &= pp.andval; + abs_lc->cregs_save_area[cr].val |= pp.orval; + } + put_abs_lowcore(abs_lc); + system_ctlreg_update(&pp); + system_ctlreg_unlock(); + } else { + system_ctlreg_update(&pp); + } +} +EXPORT_SYMBOL(system_ctlreg_modify); diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index a85e0c3e70..85328a0ef3 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -978,7 +978,6 @@ static struct ctl_table s390dbf_table[] = { .mode = S_IRUGO | S_IWUSR, .proc_handler = s390dbf_procactive, }, - { } }; static struct ctl_table_header *s390dbf_sysctl_header; diff --git a/arch/s390/kernel/diag.c b/arch/s390/kernel/diag.c index f9f06cd8fc..92fdc35f02 100644 --- a/arch/s390/kernel/diag.c +++ b/arch/s390/kernel/diag.c @@ -245,6 +245,7 @@ EXPORT_SYMBOL(diag8c); int diag224(void *ptr) { + unsigned long addr = __pa(ptr); int rc = -EOPNOTSUPP; diag_stat_inc(DIAG_STAT_X224); @@ -253,7 +254,7 @@ int diag224(void *ptr) "0: lhi %0,0x0\n" "1:\n" EX_TABLE(0b,1b) - : "+d" (rc) :"d" (0), "d" (ptr) : "memory"); + : "+d" (rc) :"d" (0), "d" (addr) : "memory"); return rc; } EXPORT_SYMBOL(diag224); diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 442ce0489e..eb43e5922a 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -46,6 +46,7 @@ decompressor_handled_param(vmalloc); decompressor_handled_param(dfltcc); decompressor_handled_param(facilities); decompressor_handled_param(nokaslr); +decompressor_handled_param(cmma); #if IS_ENABLED(CONFIG_KVM) decompressor_handled_param(prot_virt); #endif @@ -216,7 +217,7 @@ static __init void detect_machine_facilities(void) { if (test_facility(8)) { S390_lowcore.machine_flags |= MACHINE_FLAG_EDAT1; - __ctl_set_bit(0, 23); + system_ctl_set_bit(0, CR0_EDAT_BIT); } if (test_facility(78)) S390_lowcore.machine_flags |= MACHINE_FLAG_EDAT2; @@ -224,13 +225,13 @@ static __init void detect_machine_facilities(void) S390_lowcore.machine_flags |= MACHINE_FLAG_IDTE; if (test_facility(50) && test_facility(73)) { S390_lowcore.machine_flags |= MACHINE_FLAG_TE; - __ctl_set_bit(0, 55); + system_ctl_set_bit(0, CR0_TRANSACTIONAL_EXECUTION_BIT); } if (test_facility(51)) S390_lowcore.machine_flags |= MACHINE_FLAG_TLB_LC; if (test_facility(129)) { S390_lowcore.machine_flags |= MACHINE_FLAG_VX; - __ctl_set_bit(0, 17); + system_ctl_set_bit(0, CR0_VECTOR_BIT); } if (test_facility(130)) S390_lowcore.machine_flags |= MACHINE_FLAG_NX; @@ -240,7 +241,7 @@ static __init void detect_machine_facilities(void) /* Enabled signed clock comparator comparisons */ S390_lowcore.machine_flags |= MACHINE_FLAG_SCC; clock_comparator_max = -1ULL >> 1; - __ctl_set_bit(0, 53); + system_ctl_set_bit(0, CR0_CLOCK_COMPARATOR_SIGN_BIT); } if (IS_ENABLED(CONFIG_PCI) && test_facility(153)) { S390_lowcore.machine_flags |= MACHINE_FLAG_PCI_MIO; @@ -258,15 +259,9 @@ static inline void save_vector_registers(void) #endif } -static inline void setup_control_registers(void) +static inline void setup_low_address_protection(void) { - unsigned long reg; - - __ctl_store(reg, 0, 0); - reg |= CR0_LOW_ADDRESS_PROTECTION; - reg |= CR0_EMERGENCY_SIGNAL_SUBMASK; - reg |= CR0_EXTERNAL_CALL_SUBMASK; - __ctl_load(reg, 0, 0); + system_ctl_set_bit(0, CR0_LOW_ADDRESS_PROTECTION_BIT); } static inline void setup_access_registers(void) @@ -279,7 +274,7 @@ static inline void setup_access_registers(void) static int __init disable_vector_extension(char *str) { S390_lowcore.machine_flags &= ~MACHINE_FLAG_VX; - __ctl_clear_bit(0, 17); + system_ctl_clear_bit(0, CR0_VECTOR_BIT); return 0; } early_param("novx", disable_vector_extension); @@ -314,7 +309,7 @@ void __init startup_init(void) save_vector_registers(); setup_topology(); sclp_early_detect(); - setup_control_registers(); + setup_low_address_protection(); setup_access_registers(); lockdep_on(); } diff --git a/arch/s390/kernel/guarded_storage.c b/arch/s390/kernel/guarded_storage.c index d14dd1c2e5..0b68168d95 100644 --- a/arch/s390/kernel/guarded_storage.c +++ b/arch/s390/kernel/guarded_storage.c @@ -28,7 +28,7 @@ static int gs_enable(void) return -ENOMEM; gs_cb->gsd = 25; preempt_disable(); - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); load_gs_cb(gs_cb); current->thread.gs_cb = gs_cb; preempt_enable(); @@ -42,7 +42,7 @@ static int gs_disable(void) preempt_disable(); kfree(current->thread.gs_cb); current->thread.gs_cb = NULL; - __ctl_clear_bit(2, 4); + local_ctl_clear_bit(2, CR2_GUARDED_STORAGE_BIT); preempt_enable(); } return 0; @@ -84,7 +84,7 @@ void gs_load_bc_cb(struct pt_regs *regs) if (gs_cb) { kfree(current->thread.gs_cb); current->thread.gs_bc_cb = NULL; - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); load_gs_cb(gs_cb); current->thread.gs_cb = gs_cb; } diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 8d0b95c173..ba75f6bee7 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -2382,7 +2382,7 @@ void s390_reset_system(void) set_prefix(0); /* Disable lowcore protection */ - __ctl_clear_bit(0, 28); + local_ctl_clear_bit(0, CR0_LOW_ADDRESS_PROTECTION_BIT); diag_amode31_ops.diag308_reset(); } diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index b020ff17d2..6f71b0ce10 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -385,7 +385,7 @@ void irq_subclass_register(enum irq_subclass subclass) { spin_lock(&irq_subclass_lock); if (!irq_subclass_refcount[subclass]) - ctl_set_bit(0, subclass); + system_ctl_set_bit(0, subclass); irq_subclass_refcount[subclass]++; spin_unlock(&irq_subclass_lock); } @@ -396,7 +396,7 @@ void irq_subclass_unregister(enum irq_subclass subclass) spin_lock(&irq_subclass_lock); irq_subclass_refcount[subclass]--; if (!irq_subclass_refcount[subclass]) - ctl_clear_bit(0, subclass); + system_ctl_clear_bit(0, subclass); spin_unlock(&irq_subclass_lock); } EXPORT_SYMBOL(irq_subclass_unregister); diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index d4b863ed0a..f0cf20d4b3 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -224,20 +224,27 @@ static void enable_singlestep(struct kprobe_ctlblk *kcb, struct pt_regs *regs, unsigned long ip) { - struct per_regs per_kprobe; + union { + struct ctlreg regs[3]; + struct { + struct ctlreg control; + struct ctlreg start; + struct ctlreg end; + }; + } per_kprobe; /* Set up the PER control registers %cr9-%cr11 */ - per_kprobe.control = PER_EVENT_IFETCH; - per_kprobe.start = ip; - per_kprobe.end = ip; + per_kprobe.control.val = PER_EVENT_IFETCH; + per_kprobe.start.val = ip; + per_kprobe.end.val = ip; /* Save control regs and psw mask */ - __ctl_store(kcb->kprobe_saved_ctl, 9, 11); + __local_ctl_store(9, 11, kcb->kprobe_saved_ctl); kcb->kprobe_saved_imask = regs->psw.mask & (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT); /* Set PER control regs, turns on single step for the given address */ - __ctl_load(per_kprobe, 9, 11); + __local_ctl_load(9, 11, per_kprobe.regs); regs->psw.mask |= PSW_MASK_PER; regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT); regs->psw.addr = ip; @@ -249,7 +256,7 @@ static void disable_singlestep(struct kprobe_ctlblk *kcb, unsigned long ip) { /* Restore control regs and psw mask, set new psw address */ - __ctl_load(kcb->kprobe_saved_ctl, 9, 11); + __local_ctl_load(9, 11, kcb->kprobe_saved_ctl); regs->psw.mask &= ~PSW_MASK_PER; regs->psw.mask |= kcb->kprobe_saved_imask; regs->psw.addr = ip; diff --git a/arch/s390/kernel/machine_kexec.c b/arch/s390/kernel/machine_kexec.c index ce65fc0167..bb0d4d68fc 100644 --- a/arch/s390/kernel/machine_kexec.c +++ b/arch/s390/kernel/machine_kexec.c @@ -94,12 +94,12 @@ static noinline void __machine_kdump(void *image) if (MACHINE_HAS_VX) save_vx_regs((__vector128 *) mcesa->vector_save_area); if (MACHINE_HAS_GS) { - __ctl_store(cr2_old.val, 2, 2); + local_ctl_store(2, &cr2_old.reg); cr2_new = cr2_old; cr2_new.gse = 1; - __ctl_load(cr2_new.val, 2, 2); + local_ctl_load(2, &cr2_new.reg); save_gs_cb((struct gs_cb *) mcesa->guarded_storage_save_area); - __ctl_load(cr2_old.val, 2, 2); + local_ctl_load(2, &cr2_old.reg); } /* * To create a good backchain for this CPU in the dump store_status diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c index 38ec048752..0daf0f1cdf 100644 --- a/arch/s390/kernel/nmi.c +++ b/arch/s390/kernel/nmi.c @@ -22,13 +22,13 @@ #include #include #include +#include #include #include #include #include #include #include -#include #include #include #include @@ -131,10 +131,10 @@ static notrace void s390_handle_damage(void) * Disable low address protection and make machine check new PSW a * disabled wait PSW. Any additional machine check cannot be handled. */ - __ctl_store(cr0.val, 0, 0); + local_ctl_store(0, &cr0.reg); cr0_new = cr0; cr0_new.lap = 0; - __ctl_load(cr0_new.val, 0, 0); + local_ctl_load(0, &cr0_new.reg); psw_save = S390_lowcore.mcck_new_psw; psw_bits(S390_lowcore.mcck_new_psw).io = 0; psw_bits(S390_lowcore.mcck_new_psw).ext = 0; @@ -146,7 +146,7 @@ static notrace void s390_handle_damage(void) * values. This makes possible system dump analysis easier. */ S390_lowcore.mcck_new_psw = psw_save; - __ctl_load(cr0.val, 0, 0); + local_ctl_load(0, &cr0.reg); disabled_wait(); while (1); } @@ -185,7 +185,7 @@ void s390_handle_mcck(void) static int mchchk_wng_posted = 0; /* Use single cpu clear, as we cannot handle smp here. */ - __ctl_clear_bit(14, 24); /* Disable WARNING MCH */ + local_ctl_clear_bit(14, CR14_WARNING_SUBMASK_BIT); if (xchg(&mchchk_wng_posted, 1) == 0) kill_cad_pid(SIGPWR, 1); } @@ -269,9 +269,9 @@ static int notrace s390_validate_registers(union mci mci) */ if (!mci.vr && !test_cpu_flag(CIF_MCCK_GUEST)) kill_task = 1; - cr0.val = S390_lowcore.cregs_save_area[0]; + cr0.reg = S390_lowcore.cregs_save_area[0]; cr0.afp = cr0.vx = 1; - __ctl_load(cr0.val, 0, 0); + local_ctl_load(0, &cr0.reg); asm volatile( " la 1,%0\n" " VLM 0,15,0,1\n" @@ -279,7 +279,7 @@ static int notrace s390_validate_registers(union mci mci) : : "Q" (*(struct vx_array *)mcesa->vector_save_area) : "1"); - __ctl_load(S390_lowcore.cregs_save_area[0], 0, 0); + local_ctl_load(0, &S390_lowcore.cregs_save_area[0]); } /* Validate access registers */ asm volatile( @@ -290,7 +290,7 @@ static int notrace s390_validate_registers(union mci mci) if (!mci.ar) kill_task = 1; /* Validate guarded storage registers */ - cr2.val = S390_lowcore.cregs_save_area[2]; + cr2.reg = S390_lowcore.cregs_save_area[2]; if (cr2.gse) { if (!mci.gs) { /* @@ -505,9 +505,9 @@ NOKPROBE_SYMBOL(s390_do_machine_check); static int __init machine_check_init(void) { - ctl_set_bit(14, 25); /* enable external damage MCH */ - ctl_set_bit(14, 27); /* enable system recovery MCH */ - ctl_set_bit(14, 24); /* enable warning MCH */ + system_ctl_set_bit(14, CR14_EXTERNAL_DAMAGE_SUBMASK_BIT); + system_ctl_set_bit(14, CR14_RECOVERY_SUBMASK_BIT); + system_ctl_set_bit(14, CR14_WARNING_SUBMASK_BIT); return 0; } early_initcall(machine_check_init); diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c index 850c11ea63..41ed6e0f0a 100644 --- a/arch/s390/kernel/perf_cpum_cf.c +++ b/arch/s390/kernel/perf_cpum_cf.c @@ -1193,7 +1193,7 @@ static int __init cpumf_pmu_init(void) * Clear bit 15 of cr0 to unauthorize problem-state to * extract measurement counters */ - ctl_clear_bit(0, 48); + system_ctl_clear_bit(0, CR0_CPUMF_EXTRACTION_AUTH_BIT); /* register handler for measurement-alert interruptions */ rc = register_external_irq(EXT_IRQ_MEASURE_ALERT, diff --git a/arch/s390/kernel/perf_event.c b/arch/s390/kernel/perf_event.c index c27321cb09..dfa77da2fd 100644 --- a/arch/s390/kernel/perf_event.c +++ b/arch/s390/kernel/perf_event.c @@ -15,7 +15,10 @@ #include #include #include +#include +#include #include +#include #include #include #include @@ -212,6 +215,44 @@ void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, } } +void perf_callchain_user(struct perf_callchain_entry_ctx *entry, + struct pt_regs *regs) +{ + struct stack_frame_user __user *sf; + unsigned long ip, sp; + bool first = true; + + if (is_compat_task()) + return; + perf_callchain_store(entry, instruction_pointer(regs)); + sf = (void __user *)user_stack_pointer(regs); + pagefault_disable(); + while (entry->nr < entry->max_stack) { + if (__get_user(sp, &sf->back_chain)) + break; + if (__get_user(ip, &sf->gprs[8])) + break; + if (ip & 0x1) { + /* + * If the instruction address is invalid, and this + * is the first stack frame, assume r14 has not + * been written to the stack yet. Otherwise exit. + */ + if (first && !(regs->gprs[14] & 0x1)) + ip = regs->gprs[14]; + else + break; + } + perf_callchain_store(entry, ip); + /* Sanity check: ABI requires SP to be aligned 8 bytes. */ + if (!sp || sp & 0x7) + break; + sf = (void __user *)sp; + first = false; + } + pagefault_enable(); +} + /* Perf definitions for PMU event attributes in sysfs */ ssize_t cpumf_events_sysfs_show(struct device *dev, struct device_attribute *attr, char *page) diff --git a/arch/s390/kernel/perf_pai_crypto.c b/arch/s390/kernel/perf_pai_crypto.c index fe7d1774de..39a91b0043 100644 --- a/arch/s390/kernel/perf_pai_crypto.c +++ b/arch/s390/kernel/perf_pai_crypto.c @@ -16,8 +16,7 @@ #include #include #include - -#include +#include #include #include @@ -41,7 +40,43 @@ struct paicrypt_map { struct perf_event *event; /* Perf event for sampling */ }; -static DEFINE_PER_CPU(struct paicrypt_map, paicrypt_map); +struct paicrypt_mapptr { + struct paicrypt_map *mapptr; +}; + +static struct paicrypt_root { /* Anchor to per CPU data */ + refcount_t refcnt; /* Overall active events */ + struct paicrypt_mapptr __percpu *mapptr; +} paicrypt_root; + +/* Free per CPU data when the last event is removed. */ +static void paicrypt_root_free(void) +{ + if (refcount_dec_and_test(&paicrypt_root.refcnt)) { + free_percpu(paicrypt_root.mapptr); + paicrypt_root.mapptr = NULL; + } + debug_sprintf_event(cfm_dbg, 5, "%s root.refcount %d\n", __func__, + refcount_read(&paicrypt_root.refcnt)); +} + +/* + * On initialization of first event also allocate per CPU data dynamically. + * Start with an array of pointers, the array size is the maximum number of + * CPUs possible, which might be larger than the number of CPUs currently + * online. + */ +static int paicrypt_root_alloc(void) +{ + if (!refcount_inc_not_zero(&paicrypt_root.refcnt)) { + /* The memory is already zeroed. */ + paicrypt_root.mapptr = alloc_percpu(struct paicrypt_mapptr); + if (!paicrypt_root.mapptr) + return -ENOMEM; + refcount_set(&paicrypt_root.refcnt, 1); + } + return 0; +} /* Release the PMU if event is the last perf event */ static DEFINE_MUTEX(pai_reserve_mutex); @@ -51,7 +86,9 @@ static DEFINE_MUTEX(pai_reserve_mutex); */ static void paicrypt_event_destroy(struct perf_event *event) { - struct paicrypt_map *cpump = per_cpu_ptr(&paicrypt_map, event->cpu); + struct paicrypt_mapptr *mp = per_cpu_ptr(paicrypt_root.mapptr, + event->cpu); + struct paicrypt_map *cpump = mp->mapptr; cpump->event = NULL; static_branch_dec(&pai_key); @@ -66,11 +103,11 @@ static void paicrypt_event_destroy(struct perf_event *event) __func__, (unsigned long)cpump->page, cpump->save); free_page((unsigned long)cpump->page); - cpump->page = NULL; kvfree(cpump->save); - cpump->save = NULL; - cpump->mode = PAI_MODE_NONE; + kfree(cpump); + mp->mapptr = NULL; } + paicrypt_root_free(); mutex_unlock(&pai_reserve_mutex); } @@ -86,7 +123,8 @@ static u64 paicrypt_getctr(struct paicrypt_map *cpump, int nr, bool kernel) */ static u64 paicrypt_getdata(struct perf_event *event, bool kernel) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; u64 sum = 0; int i; @@ -132,11 +170,31 @@ static u64 paicrypt_getall(struct perf_event *event) * * Allocate the memory for the event. */ -static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) +static struct paicrypt_map *paicrypt_busy(struct perf_event *event) { - int rc = 0; + struct perf_event_attr *a = &event->attr; + struct paicrypt_map *cpump = NULL; + struct paicrypt_mapptr *mp; + int rc; mutex_lock(&pai_reserve_mutex); + + /* Allocate root node */ + rc = paicrypt_root_alloc(); + if (rc) + goto unlock; + + /* Allocate node for this event */ + mp = per_cpu_ptr(paicrypt_root.mapptr, event->cpu); + cpump = mp->mapptr; + if (!cpump) { /* Paicrypt_map allocated? */ + cpump = kzalloc(sizeof(*cpump), GFP_KERNEL); + if (!cpump) { + rc = -ENOMEM; + goto free_root; + } + } + if (a->sample_period) { /* Sampling requested */ if (cpump->mode != PAI_MODE_NONE) rc = -EBUSY; /* ... sampling/counting active */ @@ -144,8 +202,15 @@ static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) if (cpump->mode == PAI_MODE_SAMPLING) rc = -EBUSY; /* ... and sampling active */ } + /* + * This error case triggers when there is a conflict: + * Either sampling requested and counting already active, or visa + * versa. Therefore the struct paicrypto_map for this CPU is + * needed or the error could not have occurred. Only adjust root + * node refcount. + */ if (rc) - goto unlock; + goto free_root; /* Allocate memory for counter page and counter extraction. * Only the first counting event has to allocate a page. @@ -158,30 +223,36 @@ static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) rc = -ENOMEM; cpump->page = (unsigned long *)get_zeroed_page(GFP_KERNEL); if (!cpump->page) - goto unlock; + goto free_paicrypt_map; cpump->save = kvmalloc_array(paicrypt_cnt + 1, sizeof(struct pai_userdata), GFP_KERNEL); if (!cpump->save) { free_page((unsigned long)cpump->page); cpump->page = NULL; - goto unlock; + goto free_paicrypt_map; } + + /* Set mode and reference count */ rc = 0; refcount_set(&cpump->refcnt, 1); - -unlock: - /* If rc is non-zero, do not set mode and reference count */ - if (!rc) { - cpump->mode = a->sample_period ? PAI_MODE_SAMPLING - : PAI_MODE_COUNTING; - } + cpump->mode = a->sample_period ? PAI_MODE_SAMPLING : PAI_MODE_COUNTING; + mp->mapptr = cpump; debug_sprintf_event(cfm_dbg, 5, "%s sample_period %#llx users %d" " mode %d refcnt %u page %#lx save %p rc %d\n", __func__, a->sample_period, cpump->active_events, cpump->mode, refcount_read(&cpump->refcnt), (unsigned long)cpump->page, cpump->save, rc); + goto unlock; + +free_paicrypt_map: + kfree(cpump); + mp->mapptr = NULL; +free_root: + paicrypt_root_free(); + +unlock: mutex_unlock(&pai_reserve_mutex); - return rc; + return rc ? ERR_PTR(rc) : cpump; } /* Might be called on different CPU than the one the event is intended for. */ @@ -189,7 +260,6 @@ static int paicrypt_event_init(struct perf_event *event) { struct perf_event_attr *a = &event->attr; struct paicrypt_map *cpump; - int rc; /* PAI crypto PMU registered as PERF_TYPE_RAW, check event type */ if (a->type != PERF_TYPE_RAW && event->pmu->type != a->type) @@ -199,24 +269,16 @@ static int paicrypt_event_init(struct perf_event *event) a->config > PAI_CRYPTO_BASE + paicrypt_cnt) return -EINVAL; /* Allow only CPU wide operation, no process context for now. */ - if (event->hw.target || event->cpu == -1) + if ((event->attach_state & PERF_ATTACH_TASK) || event->cpu == -1) return -ENOENT; /* Allow only CRYPTO_ALL for sampling. */ if (a->sample_period && a->config != PAI_CRYPTO_BASE) return -EINVAL; - cpump = per_cpu_ptr(&paicrypt_map, event->cpu); - rc = paicrypt_busy(a, cpump); - if (rc) - return rc; + cpump = paicrypt_busy(event); + if (IS_ERR(cpump)) + return PTR_ERR(cpump); - /* Event initialization sets last_tag to 0. When later on the events - * are deleted and re-added, do not reset the event count value to zero. - * Events are added, deleted and re-added when 2 or more events - * are active at the same time. - */ - event->hw.last_tag = 0; - cpump->event = event; event->destroy = paicrypt_event_destroy; if (a->sample_period) { @@ -250,23 +312,28 @@ static void paicrypt_start(struct perf_event *event, int flags) { u64 sum; + /* Event initialization sets last_tag to 0. When later on the events + * are deleted and re-added, do not reset the event count value to zero. + * Events are added, deleted and re-added when 2 or more events + * are active at the same time. + */ if (!event->hw.last_tag) { event->hw.last_tag = 1; sum = paicrypt_getall(event); /* Get current value */ - local64_set(&event->count, 0); local64_set(&event->hw.prev_count, sum); } } static int paicrypt_add(struct perf_event *event, int flags) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; unsigned long ccd; if (++cpump->active_events == 1) { ccd = virt_to_phys(cpump->page) | PAI_CRYPTO_KERNEL_OFFSET; WRITE_ONCE(S390_lowcore.ccd, ccd); - __ctl_set_bit(0, 50); + local_ctl_set_bit(0, CR0_CRYPTOGRAPHY_COUNTER_BIT); } cpump->event = event; if (flags & PERF_EF_START && !event->attr.sample_period) { @@ -287,7 +354,8 @@ static void paicrypt_stop(struct perf_event *event, int flags) static void paicrypt_del(struct perf_event *event, int flags) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; if (event->attr.sample_period) perf_sched_cb_dec(event->pmu); @@ -295,7 +363,7 @@ static void paicrypt_del(struct perf_event *event, int flags) /* Only counting needs to read counter */ paicrypt_stop(event, PERF_EF_UPDATE); if (--cpump->active_events == 0) { - __ctl_clear_bit(0, 50); + local_ctl_clear_bit(0, CR0_CRYPTOGRAPHY_COUNTER_BIT); WRITE_ONCE(S390_lowcore.ccd, 0); } } @@ -329,7 +397,8 @@ static size_t paicrypt_copy(struct pai_userdata *userdata, static int paicrypt_push_sample(void) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; struct perf_event *event = cpump->event; struct perf_sample_data data; struct perf_raw_record raw; diff --git a/arch/s390/kernel/perf_pai_ext.c b/arch/s390/kernel/perf_pai_ext.c index c57c1a2032..e7013a2e89 100644 --- a/arch/s390/kernel/perf_pai_ext.c +++ b/arch/s390/kernel/perf_pai_ext.c @@ -17,8 +17,7 @@ #include #include #include - -#include +#include #include #include @@ -249,7 +248,7 @@ static int paiext_event_init(struct perf_event *event) if (rc) return rc; /* Allow only CPU wide operation, no process context for now. */ - if (event->hw.target || event->cpu == -1) + if ((event->attach_state & PERF_ATTACH_TASK) || event->cpu == -1) return -ENOENT; /* Allow only event NNPA_ALL for sampling. */ if (a->sample_period && a->config != PAI_NNPA_BASE) @@ -261,7 +260,6 @@ static int paiext_event_init(struct perf_event *event) rc = paiext_alloc(a, event); if (rc) return rc; - event->hw.last_tag = 0; event->destroy = paiext_event_destroy; if (a->sample_period) { @@ -327,7 +325,6 @@ static void paiext_start(struct perf_event *event, int flags) event->hw.last_tag = 1; sum = paiext_getall(event); /* Get current value */ local64_set(&event->hw.prev_count, sum); - local64_set(&event->count, 0); } static int paiext_add(struct perf_event *event, int flags) @@ -340,7 +337,7 @@ static int paiext_add(struct perf_event *event, int flags) S390_lowcore.aicd = virt_to_phys(cpump->paiext_cb); pcb->acc = virt_to_phys(cpump->area) | 0x1; /* Enable CPU instruction lookup for PAIE1 control block */ - __ctl_set_bit(0, 49); + local_ctl_set_bit(0, CR0_PAI_EXTENSION_BIT); debug_sprintf_event(paiext_dbg, 4, "%s 1508 %llx acc %llx\n", __func__, S390_lowcore.aicd, pcb->acc); } @@ -376,7 +373,7 @@ static void paiext_del(struct perf_event *event, int flags) } if (--cpump->active_events == 0) { /* Disable CPU instruction lookup for PAIE1 control block */ - __ctl_clear_bit(0, 49); + local_ctl_clear_bit(0, CR0_PAI_EXTENSION_BIT); pcb->acc = 0; S390_lowcore.aicd = 0; debug_sprintf_event(paiext_dbg, 4, "%s 1508 %llx acc %llx\n", diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index ea244a73ef..c7ed302a6b 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -41,13 +41,20 @@ void update_cr_regs(struct task_struct *task) { struct pt_regs *regs = task_pt_regs(task); struct thread_struct *thread = &task->thread; - struct per_regs old, new; union ctlreg0 cr0_old, cr0_new; union ctlreg2 cr2_old, cr2_new; int cr0_changed, cr2_changed; - - __ctl_store(cr0_old.val, 0, 0); - __ctl_store(cr2_old.val, 2, 2); + union { + struct ctlreg regs[3]; + struct { + struct ctlreg control; + struct ctlreg start; + struct ctlreg end; + }; + } old, new; + + local_ctl_store(0, &cr0_old.reg); + local_ctl_store(2, &cr2_old.reg); cr0_new = cr0_old; cr2_new = cr2_old; /* Take care of the enable/disable of transactional execution. */ @@ -75,38 +82,38 @@ void update_cr_regs(struct task_struct *task) cr0_changed = cr0_new.val != cr0_old.val; cr2_changed = cr2_new.val != cr2_old.val; if (cr0_changed) - __ctl_load(cr0_new.val, 0, 0); + local_ctl_load(0, &cr0_new.reg); if (cr2_changed) - __ctl_load(cr2_new.val, 2, 2); + local_ctl_load(2, &cr2_new.reg); /* Copy user specified PER registers */ - new.control = thread->per_user.control; - new.start = thread->per_user.start; - new.end = thread->per_user.end; + new.control.val = thread->per_user.control; + new.start.val = thread->per_user.start; + new.end.val = thread->per_user.end; /* merge TIF_SINGLE_STEP into user specified PER registers. */ if (test_tsk_thread_flag(task, TIF_SINGLE_STEP) || test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) { if (test_tsk_thread_flag(task, TIF_BLOCK_STEP)) - new.control |= PER_EVENT_BRANCH; + new.control.val |= PER_EVENT_BRANCH; else - new.control |= PER_EVENT_IFETCH; - new.control |= PER_CONTROL_SUSPENSION; - new.control |= PER_EVENT_TRANSACTION_END; + new.control.val |= PER_EVENT_IFETCH; + new.control.val |= PER_CONTROL_SUSPENSION; + new.control.val |= PER_EVENT_TRANSACTION_END; if (test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) - new.control |= PER_EVENT_IFETCH; - new.start = 0; - new.end = -1UL; + new.control.val |= PER_EVENT_IFETCH; + new.start.val = 0; + new.end.val = -1UL; } /* Take care of the PER enablement bit in the PSW. */ - if (!(new.control & PER_EVENT_MASK)) { + if (!(new.control.val & PER_EVENT_MASK)) { regs->psw.mask &= ~PSW_MASK_PER; return; } regs->psw.mask |= PSW_MASK_PER; - __ctl_store(old, 9, 11); + __local_ctl_store(9, 11, old.regs); if (memcmp(&new, &old, sizeof(struct per_regs)) != 0) - __ctl_load(new, 9, 11); + __local_ctl_load(9, 11, new.regs); } void user_enable_single_step(struct task_struct *task) @@ -385,6 +392,7 @@ static int __poke_user(struct task_struct *child, addr_t addr, addr_t data) /* * floating point control reg. is in the thread structure */ + save_fpu_regs(); if ((unsigned int) data != 0 || test_fp_ctl(data >> (BITS_PER_LONG - 32))) return -EINVAL; @@ -741,6 +749,7 @@ static int __poke_user_compat(struct task_struct *child, /* * floating point control reg. is in the thread structure */ + save_fpu_regs(); if (test_fp_ctl(tmp)) return -EINVAL; child->thread.fpu.fpc = data; @@ -904,9 +913,7 @@ static int s390_fpregs_set(struct task_struct *target, int rc = 0; freg_t fprs[__NUM_FPRS]; - if (target == current) - save_fpu_regs(); - + save_fpu_regs(); if (MACHINE_HAS_VX) convert_vx_to_fp(fprs, target->thread.fpu.vxrs); else @@ -1107,7 +1114,7 @@ static int s390_gs_cb_set(struct task_struct *target, target->thread.gs_cb = data; *target->thread.gs_cb = gs_cb; if (target == current) { - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); restore_gs_cb(target->thread.gs_cb); } preempt_enable(); diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index de6ad0fb23..5701356f4f 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -305,7 +305,7 @@ static void __init setup_zfcpdump(void) return; if (oldmem_data.start) return; - strcat(boot_command_line, " cio_ignore=all,!ipldev,!condev"); + strlcat(boot_command_line, " cio_ignore=all,!ipldev,!condev", COMMAND_LINE_SIZE); console_loglevel = 2; } #else @@ -381,12 +381,6 @@ void stack_free(unsigned long stack) #endif } -void __init __noreturn arch_call_rest_init(void) -{ - smp_reinit_ipl_cpu(); - rest_init(); -} - static unsigned long __init stack_alloc_early(void) { unsigned long stack; @@ -455,7 +449,6 @@ static void __init setup_lowcore(void) lc->restart_fn = (unsigned long) do_restart; lc->restart_data = 0; lc->restart_source = -1U; - __ctl_store(lc->cregs_save_area, 0, 15); lc->spinlock_lockval = arch_spin_lockval(0); lc->spinlock_index = 0; arch_spin_lock_setup(0); @@ -465,6 +458,7 @@ static void __init setup_lowcore(void) lc->kernel_asce = S390_lowcore.kernel_asce; lc->user_asce = S390_lowcore.user_asce; + system_ctlreg_init_save_area(lc); abs_lc = get_abs_lowcore(); abs_lc->restart_stack = lc->restart_stack; abs_lc->restart_fn = lc->restart_fn; @@ -472,7 +466,6 @@ static void __init setup_lowcore(void) abs_lc->restart_source = lc->restart_source; abs_lc->restart_psw = lc->restart_psw; abs_lc->restart_flags = RESTART_FLAG_CTLREGS; - memcpy(abs_lc->cregs_save_area, lc->cregs_save_area, sizeof(abs_lc->cregs_save_area)); abs_lc->program_new_psw = lc->program_new_psw; abs_lc->mcesad = lc->mcesad; put_abs_lowcore(abs_lc); @@ -625,8 +618,8 @@ static void __init reserve_crashkernel(void) phys_addr_t low, high; int rc; - rc = parse_crashkernel(boot_command_line, ident_map_size, &crash_size, - &crash_base); + rc = parse_crashkernel(boot_command_line, ident_map_size, + &crash_size, &crash_base, NULL, NULL); crash_base = ALIGN(crash_base, KEXEC_CRASH_MEM_ALIGN); crash_size = ALIGN(crash_size, KEXEC_CRASH_MEM_ALIGN); @@ -797,15 +790,15 @@ static void __init setup_cr(void) __ctl_duct[4] = (unsigned long)__ctl_duald; /* Update control registers CR2, CR5 and CR15 */ - __ctl_store(cr2.val, 2, 2); - __ctl_store(cr5.val, 5, 5); - __ctl_store(cr15.val, 15, 15); + local_ctl_store(2, &cr2.reg); + local_ctl_store(5, &cr5.reg); + local_ctl_store(15, &cr15.reg); cr2.ducto = (unsigned long)__ctl_duct >> 6; cr5.pasteo = (unsigned long)__ctl_duct >> 6; cr15.lsea = (unsigned long)__ctl_linkage_stack >> 3; - __ctl_load(cr2.val, 2, 2); - __ctl_load(cr5.val, 5, 5); - __ctl_load(cr15.val, 15, 15); + system_ctl_load(2, &cr2.reg); + system_ctl_load(5, &cr5.reg); + system_ctl_load(15, &cr15.reg); } /* diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index a4edb7ea66..f7fcfff09a 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include #include @@ -567,54 +568,6 @@ void arch_irq_work_raise(void) } #endif -/* - * parameter area for the set/clear control bit callbacks - */ -struct ec_creg_mask_parms { - unsigned long orval; - unsigned long andval; - int cr; -}; - -/* - * callback for setting/clearing control bits - */ -static void smp_ctl_bit_callback(void *info) -{ - struct ec_creg_mask_parms *pp = info; - unsigned long cregs[16]; - - __ctl_store(cregs, 0, 15); - cregs[pp->cr] = (cregs[pp->cr] & pp->andval) | pp->orval; - __ctl_load(cregs, 0, 15); -} - -static DEFINE_SPINLOCK(ctl_lock); - -void smp_ctl_set_clear_bit(int cr, int bit, bool set) -{ - struct ec_creg_mask_parms parms = { .cr = cr, }; - struct lowcore *abs_lc; - u64 ctlreg; - - if (set) { - parms.orval = 1UL << bit; - parms.andval = -1UL; - } else { - parms.orval = 0; - parms.andval = ~(1UL << bit); - } - spin_lock(&ctl_lock); - abs_lc = get_abs_lowcore(); - ctlreg = abs_lc->cregs_save_area[cr]; - ctlreg = (ctlreg & parms.andval) | parms.orval; - abs_lc->cregs_save_area[cr] = ctlreg; - put_abs_lowcore(abs_lc); - on_each_cpu(smp_ctl_bit_callback, &parms, 1); - spin_unlock(&ctl_lock); -} -EXPORT_SYMBOL(smp_ctl_set_clear_bit); - #ifdef CONFIG_CRASH_DUMP int smp_store_status(int cpu) @@ -898,7 +851,7 @@ static void smp_start_secondary(void *cpuvoid) S390_lowcore.restart_flags = 0; restore_access_regs(S390_lowcore.access_regs_save_area); cpu_init(); - rcu_cpu_starting(cpu); + rcutree_report_cpu_starting(cpu); init_cpu_timer(); vtime_init(); vdso_getcpu_init(); @@ -935,14 +888,14 @@ int __cpu_up(unsigned int cpu, struct task_struct *tidle) * Make sure global control register contents do not change * until new CPU has initialized control registers. */ - spin_lock(&ctl_lock); + system_ctlreg_lock(); pcpu_prepare_secondary(pcpu, cpu); pcpu_attach_task(pcpu, tidle); pcpu_start_fn(pcpu, smp_start_secondary, NULL); /* Wait until cpu puts itself in the online & active maps */ while (!cpu_online(cpu)) cpu_relax(); - spin_unlock(&ctl_lock); + system_ctlreg_unlock(); return 0; } @@ -957,7 +910,7 @@ early_param("possible_cpus", _setup_possible_cpus); int __cpu_disable(void) { - unsigned long cregs[16]; + struct ctlreg cregs[16]; int cpu; /* Handle possible pending IPIs */ @@ -969,11 +922,11 @@ int __cpu_disable(void) /* Disable pseudo page faults on this cpu. */ pfault_fini(); /* Disable interrupt sources via control register. */ - __ctl_store(cregs, 0, 15); - cregs[0] &= ~0x0000ee70UL; /* disable all external interrupts */ - cregs[6] &= ~0xff000000UL; /* disable all I/O interrupts */ - cregs[14] &= ~0x1f000000UL; /* disable most machine checks */ - __ctl_load(cregs, 0, 15); + __local_ctl_store(0, 15, cregs); + cregs[0].val &= ~0x0000ee70UL; /* disable all external interrupts */ + cregs[6].val &= ~0xff000000UL; /* disable all I/O interrupts */ + cregs[14].val &= ~0x1f000000UL; /* disable most machine checks */ + __local_ctl_load(0, 15, cregs); clear_cpu_flag(CIF_NOHZ_DELAY); return 0; } @@ -1013,12 +966,12 @@ void __init smp_fill_possible_mask(void) void __init smp_prepare_cpus(unsigned int max_cpus) { - /* request the 0x1201 emergency signal external interrupt */ if (register_external_irq(EXT_IRQ_EMERGENCY_SIG, do_ext_call_interrupt)) panic("Couldn't request external interrupt 0x1201"); - /* request the 0x1202 external call external interrupt */ + system_ctl_set_bit(0, 14); if (register_external_irq(EXT_IRQ_EXTERNAL_CALL, do_ext_call_interrupt)) panic("Couldn't request external interrupt 0x1202"); + system_ctl_set_bit(0, 13); } void __init smp_prepare_boot_cpu(void) @@ -1076,11 +1029,9 @@ static ssize_t cpu_configure_store(struct device *dev, cpus_read_lock(); mutex_lock(&smp_cpu_state_mutex); rc = -EBUSY; - /* disallow configuration changes of online cpus and cpu 0 */ + /* disallow configuration changes of online cpus */ cpu = dev->id; cpu = smp_get_base_cpu(cpu); - if (cpu == 0) - goto out; for (i = 0; i <= smp_cpu_mtid; i++) if (cpu_online(cpu + i)) goto out; @@ -1180,7 +1131,7 @@ static int smp_add_present_cpu(int cpu) return -ENOMEM; per_cpu(cpu_device, cpu) = c; s = &c->dev; - c->hotpluggable = 1; + c->hotpluggable = !!cpu; rc = register_cpu(c, cpu); if (rc) goto out; @@ -1258,60 +1209,3 @@ out: return rc; } subsys_initcall(s390_smp_init); - -static __always_inline void set_new_lowcore(struct lowcore *lc) -{ - union register_pair dst, src; - u32 pfx; - - src.even = (unsigned long) &S390_lowcore; - src.odd = sizeof(S390_lowcore); - dst.even = (unsigned long) lc; - dst.odd = sizeof(*lc); - pfx = __pa(lc); - - asm volatile( - " mvcl %[dst],%[src]\n" - " spx %[pfx]\n" - : [dst] "+&d" (dst.pair), [src] "+&d" (src.pair) - : [pfx] "Q" (pfx) - : "memory", "cc"); -} - -int __init smp_reinit_ipl_cpu(void) -{ - unsigned long async_stack, nodat_stack, mcck_stack; - struct lowcore *lc, *lc_ipl; - unsigned long flags, cr0; - u64 mcesad; - - lc_ipl = lowcore_ptr[0]; - lc = (struct lowcore *) __get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER); - nodat_stack = __get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER); - async_stack = stack_alloc(); - mcck_stack = stack_alloc(); - if (!lc || !nodat_stack || !async_stack || !mcck_stack || nmi_alloc_mcesa(&mcesad)) - panic("Couldn't allocate memory"); - - local_irq_save(flags); - local_mcck_disable(); - set_new_lowcore(lc); - S390_lowcore.nodat_stack = nodat_stack + STACK_INIT_OFFSET; - S390_lowcore.async_stack = async_stack + STACK_INIT_OFFSET; - S390_lowcore.mcck_stack = mcck_stack + STACK_INIT_OFFSET; - __ctl_store(cr0, 0, 0); - __ctl_clear_bit(0, 28); /* disable lowcore protection */ - S390_lowcore.mcesad = mcesad; - __ctl_load(cr0, 0, 0); - if (abs_lowcore_map(0, lc, false)) - panic("Couldn't remap absolute lowcore"); - lowcore_ptr[0] = lc; - local_mcck_enable(); - local_irq_restore(flags); - - memblock_free_late(__pa(lc_ipl->mcck_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl->async_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl->nodat_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl), sizeof(*lc_ipl)); - return 0; -} diff --git a/arch/s390/kernel/stacktrace.c b/arch/s390/kernel/stacktrace.c index 0787010139..94f440e383 100644 --- a/arch/s390/kernel/stacktrace.c +++ b/arch/s390/kernel/stacktrace.c @@ -6,9 +6,12 @@ */ #include +#include +#include #include #include #include +#include void arch_stack_walk(stack_trace_consume_fn consume_entry, void *cookie, struct task_struct *task, struct pt_regs *regs) @@ -58,3 +61,43 @@ int arch_stack_walk_reliable(stack_trace_consume_fn consume_entry, return -EINVAL; return 0; } + +void arch_stack_walk_user(stack_trace_consume_fn consume_entry, void *cookie, + const struct pt_regs *regs) +{ + struct stack_frame_user __user *sf; + unsigned long ip, sp; + bool first = true; + + if (is_compat_task()) + return; + if (!consume_entry(cookie, instruction_pointer(regs))) + return; + sf = (void __user *)user_stack_pointer(regs); + pagefault_disable(); + while (1) { + if (__get_user(sp, &sf->back_chain)) + break; + if (__get_user(ip, &sf->gprs[8])) + break; + if (ip & 0x1) { + /* + * If the instruction address is invalid, and this + * is the first stack frame, assume r14 has not + * been written to the stack yet. Otherwise exit. + */ + if (first && !(regs->gprs[14] & 0x1)) + ip = regs->gprs[14]; + else + break; + } + if (!consume_entry(cookie, ip)) + break; + /* Sanity check: ABI requires SP to be aligned 8 bytes. */ + if (!sp || sp & 0x7) + break; + sf = (void __user *)sp; + first = false; + } + pagefault_enable(); +} diff --git a/arch/s390/kernel/syscalls/syscall.tbl b/arch/s390/kernel/syscalls/syscall.tbl index 0122cc1569..86fec9b080 100644 --- a/arch/s390/kernel/syscalls/syscall.tbl +++ b/arch/s390/kernel/syscalls/syscall.tbl @@ -100,7 +100,7 @@ 106 common stat sys_newstat compat_sys_newstat 107 common lstat sys_newlstat compat_sys_newlstat 108 common fstat sys_newfstat compat_sys_newfstat -110 common lookup_dcookie sys_lookup_dcookie compat_sys_lookup_dcookie +110 common lookup_dcookie - - 111 common vhangup sys_vhangup sys_vhangup 112 common idle - - 114 common wait4 sys_wait4 compat_sys_wait4 @@ -455,3 +455,7 @@ 450 common set_mempolicy_home_node sys_set_mempolicy_home_node sys_set_mempolicy_home_node 451 common cachestat sys_cachestat sys_cachestat 452 common fchmodat2 sys_fchmodat2 sys_fchmodat2 +453 common map_shadow_stack sys_map_shadow_stack sys_map_shadow_stack +454 common futex_wake sys_futex_wake sys_futex_wake +455 common futex_wait sys_futex_wait sys_futex_wait +456 common futex_requeue sys_futex_requeue sys_futex_requeue diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index d34d3548c0..14abad953c 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -173,10 +173,10 @@ void init_cpu_timer(void) clockevents_register_device(cd); /* Enable clock comparator timer interrupt. */ - __ctl_set_bit(0,11); + local_ctl_set_bit(0, CR0_CLOCK_COMPARATOR_SUBMASK_BIT); /* Always allow the timing alert external interrupt. */ - __ctl_set_bit(0, 4); + local_ctl_set_bit(0, CR0_ETR_SUBMASK_BIT); } static void clock_comparator_interrupt(struct ext_code ext_code, diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index 68adf1de88..89e91b8ce8 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -522,7 +522,7 @@ static struct sched_domain_topology_level s390_topology[] = { { cpu_coregroup_mask, cpu_core_flags, SD_INIT_NAME(MC) }, { cpu_book_mask, SD_INIT_NAME(BOOK) }, { cpu_drawer_mask, SD_INIT_NAME(DRAWER) }, - { cpu_cpu_mask, SD_INIT_NAME(DIE) }, + { cpu_cpu_mask, SD_INIT_NAME(PKG) }, { NULL, }, }; @@ -636,7 +636,6 @@ static struct ctl_table topology_ctl_table[] = { .mode = 0644, .proc_handler = topology_ctl_handler, }, - { }, }; static int __init topology_init(void) diff --git a/arch/s390/kernel/vdso32/Makefile b/arch/s390/kernel/vdso32/Makefile index 23e868b79a..caec7db6f9 100644 --- a/arch/s390/kernel/vdso32/Makefile +++ b/arch/s390/kernel/vdso32/Makefile @@ -61,16 +61,6 @@ quiet_cmd_vdso32as = VDSO32A $@ quiet_cmd_vdso32cc = VDSO32C $@ cmd_vdso32cc = $(CC) $(c_flags) -c -o $@ $< -# install commands for the unstripped file -quiet_cmd_vdso_install = INSTALL $@ - cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@ - -vdso32.so: $(obj)/vdso32.so.dbg - @mkdir -p $(MODLIB)/vdso - $(call cmd,vdso_install) - -vdso_install: vdso32.so - # Generate VDSO offsets using helper script gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh quiet_cmd_vdsosym = VDSOSYM $@ diff --git a/arch/s390/kernel/vdso64/Makefile b/arch/s390/kernel/vdso64/Makefile index fc1c6ff817..e3c9085f8f 100644 --- a/arch/s390/kernel/vdso64/Makefile +++ b/arch/s390/kernel/vdso64/Makefile @@ -70,16 +70,6 @@ quiet_cmd_vdso64as = VDSO64A $@ quiet_cmd_vdso64cc = VDSO64C $@ cmd_vdso64cc = $(CC) $(c_flags) -c -o $@ $< -# install commands for the unstripped file -quiet_cmd_vdso_install = INSTALL $@ - cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@ - -vdso64.so: $(obj)/vdso64.so.dbg - @mkdir -p $(MODLIB)/vdso - $(call cmd,vdso_install) - -vdso_install: vdso64.so - # Generate VDSO offsets using helper script gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh quiet_cmd_vdsosym = VDSOSYM $@ -- cgit v1.2.3