diff options
Diffstat (limited to 'arch/xtensa/kernel/smp.c')
| -rw-r--r-- | arch/xtensa/kernel/smp.c | 634 |
1 files changed, 634 insertions, 0 deletions
diff --git a/arch/xtensa/kernel/smp.c b/arch/xtensa/kernel/smp.c new file mode 100644 index 000000000..4dc109dd6 --- /dev/null +++ b/arch/xtensa/kernel/smp.c @@ -0,0 +1,634 @@ +/* + * Xtensa SMP support functions. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2008 - 2013 Tensilica Inc. + * + * Chris Zankel <chris@zankel.net> + * Joe Taylor <joe@tensilica.com> + * Pete Delaney <piet@tensilica.com + */ + +#include <linux/cpu.h> +#include <linux/cpumask.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/irqdomain.h> +#include <linux/irq.h> +#include <linux/kdebug.h> +#include <linux/module.h> +#include <linux/sched/mm.h> +#include <linux/sched/hotplug.h> +#include <linux/sched/task_stack.h> +#include <linux/reboot.h> +#include <linux/seq_file.h> +#include <linux/smp.h> +#include <linux/thread_info.h> + +#include <asm/cacheflush.h> +#include <asm/coprocessor.h> +#include <asm/kdebug.h> +#include <asm/mmu_context.h> +#include <asm/mxregs.h> +#include <asm/platform.h> +#include <asm/tlbflush.h> +#include <asm/traps.h> + +#ifdef CONFIG_SMP +# if XCHAL_HAVE_S32C1I == 0 +# error "The S32C1I option is required for SMP." +# endif +#endif + +static void system_invalidate_dcache_range(unsigned long start, + unsigned long size); +static void system_flush_invalidate_dcache_range(unsigned long start, + unsigned long size); + +/* IPI (Inter Process Interrupt) */ + +#define IPI_IRQ 0 + +static irqreturn_t ipi_interrupt(int irq, void *dev_id); + +void ipi_init(void) +{ + unsigned irq = irq_create_mapping(NULL, IPI_IRQ); + if (request_irq(irq, ipi_interrupt, IRQF_PERCPU, "ipi", NULL)) + pr_err("Failed to request irq %u (ipi)\n", irq); +} + +static inline unsigned int get_core_count(void) +{ + /* Bits 18..21 of SYSCFGID contain the core count minus 1. */ + unsigned int syscfgid = get_er(SYSCFGID); + return ((syscfgid >> 18) & 0xf) + 1; +} + +static inline int get_core_id(void) +{ + /* Bits 0...18 of SYSCFGID contain the core id */ + unsigned int core_id = get_er(SYSCFGID); + return core_id & 0x3fff; +} + +void __init smp_prepare_cpus(unsigned int max_cpus) +{ + unsigned i; + + for_each_possible_cpu(i) + set_cpu_present(i, true); +} + +void __init smp_init_cpus(void) +{ + unsigned i; + unsigned int ncpus = get_core_count(); + unsigned int core_id = get_core_id(); + + pr_info("%s: Core Count = %d\n", __func__, ncpus); + pr_info("%s: Core Id = %d\n", __func__, core_id); + + if (ncpus > NR_CPUS) { + ncpus = NR_CPUS; + pr_info("%s: limiting core count by %d\n", __func__, ncpus); + } + + for (i = 0; i < ncpus; ++i) + set_cpu_possible(i, true); +} + +void __init smp_prepare_boot_cpu(void) +{ + unsigned int cpu = smp_processor_id(); + BUG_ON(cpu != 0); + cpu_asid_cache(cpu) = ASID_USER_FIRST; +} + +void __init smp_cpus_done(unsigned int max_cpus) +{ +} + +static int boot_secondary_processors = 1; /* Set with xt-gdb via .xt-gdb */ +static DECLARE_COMPLETION(cpu_running); + +void secondary_start_kernel(void) +{ + struct mm_struct *mm = &init_mm; + unsigned int cpu = smp_processor_id(); + + init_mmu(); + +#ifdef CONFIG_DEBUG_MISC + if (boot_secondary_processors == 0) { + pr_debug("%s: boot_secondary_processors:%d; Hanging cpu:%d\n", + __func__, boot_secondary_processors, cpu); + for (;;) + __asm__ __volatile__ ("waiti " __stringify(LOCKLEVEL)); + } + + pr_debug("%s: boot_secondary_processors:%d; Booting cpu:%d\n", + __func__, boot_secondary_processors, cpu); +#endif + /* Init EXCSAVE1 */ + + secondary_trap_init(); + + /* All kernel threads share the same mm context. */ + + mmget(mm); + mmgrab(mm); + current->active_mm = mm; + cpumask_set_cpu(cpu, mm_cpumask(mm)); + enter_lazy_tlb(mm, current); + + trace_hardirqs_off(); + + calibrate_delay(); + + notify_cpu_starting(cpu); + + secondary_init_irq(); + local_timer_setup(cpu); + + set_cpu_online(cpu, true); + + local_irq_enable(); + + complete(&cpu_running); + + cpu_startup_entry(CPUHP_AP_ONLINE_IDLE); +} + +static void mx_cpu_start(void *p) +{ + unsigned cpu = (unsigned)p; + unsigned long run_stall_mask = get_er(MPSCORE); + + set_er(run_stall_mask & ~(1u << cpu), MPSCORE); + pr_debug("%s: cpu: %d, run_stall_mask: %lx ---> %lx\n", + __func__, cpu, run_stall_mask, get_er(MPSCORE)); +} + +static void mx_cpu_stop(void *p) +{ + unsigned cpu = (unsigned)p; + unsigned long run_stall_mask = get_er(MPSCORE); + + set_er(run_stall_mask | (1u << cpu), MPSCORE); + pr_debug("%s: cpu: %d, run_stall_mask: %lx ---> %lx\n", + __func__, cpu, run_stall_mask, get_er(MPSCORE)); +} + +#ifdef CONFIG_HOTPLUG_CPU +unsigned long cpu_start_id __cacheline_aligned; +#endif +unsigned long cpu_start_ccount; + +static int boot_secondary(unsigned int cpu, struct task_struct *ts) +{ + unsigned long timeout = jiffies + msecs_to_jiffies(1000); + unsigned long ccount; + int i; + +#ifdef CONFIG_HOTPLUG_CPU + WRITE_ONCE(cpu_start_id, cpu); + /* Pairs with the third memw in the cpu_restart */ + mb(); + system_flush_invalidate_dcache_range((unsigned long)&cpu_start_id, + sizeof(cpu_start_id)); +#endif + smp_call_function_single(0, mx_cpu_start, (void *)cpu, 1); + + for (i = 0; i < 2; ++i) { + do + ccount = get_ccount(); + while (!ccount); + + WRITE_ONCE(cpu_start_ccount, ccount); + + do { + /* + * Pairs with the first two memws in the + * .Lboot_secondary. + */ + mb(); + ccount = READ_ONCE(cpu_start_ccount); + } while (ccount && time_before(jiffies, timeout)); + + if (ccount) { + smp_call_function_single(0, mx_cpu_stop, + (void *)cpu, 1); + WRITE_ONCE(cpu_start_ccount, 0); + return -EIO; + } + } + return 0; +} + +int __cpu_up(unsigned int cpu, struct task_struct *idle) +{ + int ret = 0; + + if (cpu_asid_cache(cpu) == 0) + cpu_asid_cache(cpu) = ASID_USER_FIRST; + + start_info.stack = (unsigned long)task_pt_regs(idle); + wmb(); + + pr_debug("%s: Calling wakeup_secondary(cpu:%d, idle:%p, sp: %08lx)\n", + __func__, cpu, idle, start_info.stack); + + init_completion(&cpu_running); + ret = boot_secondary(cpu, idle); + if (ret == 0) { + wait_for_completion_timeout(&cpu_running, + msecs_to_jiffies(1000)); + if (!cpu_online(cpu)) + ret = -EIO; + } + + if (ret) + pr_err("CPU %u failed to boot\n", cpu); + + return ret; +} + +#ifdef CONFIG_HOTPLUG_CPU + +/* + * __cpu_disable runs on the processor to be shutdown. + */ +int __cpu_disable(void) +{ + unsigned int cpu = smp_processor_id(); + + /* + * Take this CPU offline. Once we clear this, we can't return, + * and we must not schedule until we're ready to give up the cpu. + */ + set_cpu_online(cpu, false); + +#if XTENSA_HAVE_COPROCESSORS + /* + * Flush coprocessor contexts that are active on the current CPU. + */ + local_coprocessors_flush_release_all(); +#endif + /* + * OK - migrate IRQs away from this CPU + */ + migrate_irqs(); + + /* + * Flush user cache and TLB mappings, and then remove this CPU + * from the vm mask set of all processes. + */ + local_flush_cache_all(); + local_flush_tlb_all(); + invalidate_page_directory(); + + clear_tasks_mm_cpumask(cpu); + + return 0; +} + +static void platform_cpu_kill(unsigned int cpu) +{ + smp_call_function_single(0, mx_cpu_stop, (void *)cpu, true); +} + +/* + * called on the thread which is asking for a CPU to be shutdown - + * waits until shutdown has completed, or it is timed out. + */ +void __cpu_die(unsigned int cpu) +{ + unsigned long timeout = jiffies + msecs_to_jiffies(1000); + while (time_before(jiffies, timeout)) { + system_invalidate_dcache_range((unsigned long)&cpu_start_id, + sizeof(cpu_start_id)); + /* Pairs with the second memw in the cpu_restart */ + mb(); + if (READ_ONCE(cpu_start_id) == -cpu) { + platform_cpu_kill(cpu); + return; + } + } + pr_err("CPU%u: unable to kill\n", cpu); +} + +void arch_cpu_idle_dead(void) +{ + cpu_die(); +} +/* + * Called from the idle thread for the CPU which has been shutdown. + * + * Note that we disable IRQs here, but do not re-enable them + * before returning to the caller. This is also the behaviour + * of the other hotplug-cpu capable cores, so presumably coming + * out of idle fixes this. + */ +void __ref cpu_die(void) +{ + idle_task_exit(); + local_irq_disable(); + __asm__ __volatile__( + " movi a2, cpu_restart\n" + " jx a2\n"); +} + +#endif /* CONFIG_HOTPLUG_CPU */ + +enum ipi_msg_type { + IPI_RESCHEDULE = 0, + IPI_CALL_FUNC, + IPI_CPU_STOP, + IPI_MAX +}; + +static const struct { + const char *short_text; + const char *long_text; +} ipi_text[] = { + { .short_text = "RES", .long_text = "Rescheduling interrupts" }, + { .short_text = "CAL", .long_text = "Function call interrupts" }, + { .short_text = "DIE", .long_text = "CPU shutdown interrupts" }, +}; + +struct ipi_data { + unsigned long ipi_count[IPI_MAX]; +}; + +static DEFINE_PER_CPU(struct ipi_data, ipi_data); + +static void send_ipi_message(const struct cpumask *callmask, + enum ipi_msg_type msg_id) +{ + int index; + unsigned long mask = 0; + + for_each_cpu(index, callmask) + mask |= 1 << index; + + set_er(mask, MIPISET(msg_id)); +} + +void arch_send_call_function_ipi_mask(const struct cpumask *mask) +{ + send_ipi_message(mask, IPI_CALL_FUNC); +} + +void arch_send_call_function_single_ipi(int cpu) +{ + send_ipi_message(cpumask_of(cpu), IPI_CALL_FUNC); +} + +void smp_send_reschedule(int cpu) +{ + send_ipi_message(cpumask_of(cpu), IPI_RESCHEDULE); +} + +void smp_send_stop(void) +{ + struct cpumask targets; + + cpumask_copy(&targets, cpu_online_mask); + cpumask_clear_cpu(smp_processor_id(), &targets); + send_ipi_message(&targets, IPI_CPU_STOP); +} + +static void ipi_cpu_stop(unsigned int cpu) +{ + set_cpu_online(cpu, false); + machine_halt(); +} + +irqreturn_t ipi_interrupt(int irq, void *dev_id) +{ + unsigned int cpu = smp_processor_id(); + struct ipi_data *ipi = &per_cpu(ipi_data, cpu); + + for (;;) { + unsigned int msg; + + msg = get_er(MIPICAUSE(cpu)); + set_er(msg, MIPICAUSE(cpu)); + + if (!msg) + break; + + if (msg & (1 << IPI_CALL_FUNC)) { + ++ipi->ipi_count[IPI_CALL_FUNC]; + generic_smp_call_function_interrupt(); + } + + if (msg & (1 << IPI_RESCHEDULE)) { + ++ipi->ipi_count[IPI_RESCHEDULE]; + scheduler_ipi(); + } + + if (msg & (1 << IPI_CPU_STOP)) { + ++ipi->ipi_count[IPI_CPU_STOP]; + ipi_cpu_stop(cpu); + } + } + + return IRQ_HANDLED; +} + +void show_ipi_list(struct seq_file *p, int prec) +{ + unsigned int cpu; + unsigned i; + + for (i = 0; i < IPI_MAX; ++i) { + seq_printf(p, "%*s:", prec, ipi_text[i].short_text); + for_each_online_cpu(cpu) + seq_printf(p, " %10lu", + per_cpu(ipi_data, cpu).ipi_count[i]); + seq_printf(p, " %s\n", ipi_text[i].long_text); + } +} + +int setup_profiling_timer(unsigned int multiplier) +{ + pr_debug("setup_profiling_timer %d\n", multiplier); + return 0; +} + +/* TLB flush functions */ + +struct flush_data { + struct vm_area_struct *vma; + unsigned long addr1; + unsigned long addr2; +}; + +static void ipi_flush_tlb_all(void *arg) +{ + local_flush_tlb_all(); +} + +void flush_tlb_all(void) +{ + on_each_cpu(ipi_flush_tlb_all, NULL, 1); +} + +static void ipi_flush_tlb_mm(void *arg) +{ + local_flush_tlb_mm(arg); +} + +void flush_tlb_mm(struct mm_struct *mm) +{ + on_each_cpu(ipi_flush_tlb_mm, mm, 1); +} + +static void ipi_flush_tlb_page(void *arg) +{ + struct flush_data *fd = arg; + local_flush_tlb_page(fd->vma, fd->addr1); +} + +void flush_tlb_page(struct vm_area_struct *vma, unsigned long addr) +{ + struct flush_data fd = { + .vma = vma, + .addr1 = addr, + }; + on_each_cpu(ipi_flush_tlb_page, &fd, 1); +} + +static void ipi_flush_tlb_range(void *arg) +{ + struct flush_data *fd = arg; + local_flush_tlb_range(fd->vma, fd->addr1, fd->addr2); +} + +void flush_tlb_range(struct vm_area_struct *vma, + unsigned long start, unsigned long end) +{ + struct flush_data fd = { + .vma = vma, + .addr1 = start, + .addr2 = end, + }; + on_each_cpu(ipi_flush_tlb_range, &fd, 1); +} + +static void ipi_flush_tlb_kernel_range(void *arg) +{ + struct flush_data *fd = arg; + local_flush_tlb_kernel_range(fd->addr1, fd->addr2); +} + +void flush_tlb_kernel_range(unsigned long start, unsigned long end) +{ + struct flush_data fd = { + .addr1 = start, + .addr2 = end, + }; + on_each_cpu(ipi_flush_tlb_kernel_range, &fd, 1); +} + +/* Cache flush functions */ + +static void ipi_flush_cache_all(void *arg) +{ + local_flush_cache_all(); +} + +void flush_cache_all(void) +{ + on_each_cpu(ipi_flush_cache_all, NULL, 1); +} + +static void ipi_flush_cache_page(void *arg) +{ + struct flush_data *fd = arg; + local_flush_cache_page(fd->vma, fd->addr1, fd->addr2); +} + +void flush_cache_page(struct vm_area_struct *vma, + unsigned long address, unsigned long pfn) +{ + struct flush_data fd = { + .vma = vma, + .addr1 = address, + .addr2 = pfn, + }; + on_each_cpu(ipi_flush_cache_page, &fd, 1); +} + +static void ipi_flush_cache_range(void *arg) +{ + struct flush_data *fd = arg; + local_flush_cache_range(fd->vma, fd->addr1, fd->addr2); +} + +void flush_cache_range(struct vm_area_struct *vma, + unsigned long start, unsigned long end) +{ + struct flush_data fd = { + .vma = vma, + .addr1 = start, + .addr2 = end, + }; + on_each_cpu(ipi_flush_cache_range, &fd, 1); +} + +static void ipi_flush_icache_range(void *arg) +{ + struct flush_data *fd = arg; + local_flush_icache_range(fd->addr1, fd->addr2); +} + +void flush_icache_range(unsigned long start, unsigned long end) +{ + struct flush_data fd = { + .addr1 = start, + .addr2 = end, + }; + on_each_cpu(ipi_flush_icache_range, &fd, 1); +} +EXPORT_SYMBOL(flush_icache_range); + +/* ------------------------------------------------------------------------- */ + +static void ipi_invalidate_dcache_range(void *arg) +{ + struct flush_data *fd = arg; + __invalidate_dcache_range(fd->addr1, fd->addr2); +} + +static void system_invalidate_dcache_range(unsigned long start, + unsigned long size) +{ + struct flush_data fd = { + .addr1 = start, + .addr2 = size, + }; + on_each_cpu(ipi_invalidate_dcache_range, &fd, 1); +} + +static void ipi_flush_invalidate_dcache_range(void *arg) +{ + struct flush_data *fd = arg; + __flush_invalidate_dcache_range(fd->addr1, fd->addr2); +} + +static void system_flush_invalidate_dcache_range(unsigned long start, + unsigned long size) +{ + struct flush_data fd = { + .addr1 = start, + .addr2 = size, + }; + on_each_cpu(ipi_flush_invalidate_dcache_range, &fd, 1); +} |