diff options
Diffstat (limited to 'arch/arm/kernel')
-rw-r--r-- | arch/arm/kernel/Makefile | 5 | ||||
-rw-r--r-- | arch/arm/kernel/head.S | 1 | ||||
-rw-r--r-- | arch/arm/kernel/hw_breakpoint.c | 3 | ||||
-rw-r--r-- | arch/arm/kernel/perf_event.c | 19 | ||||
-rw-r--r-- | arch/arm/kernel/perf_event_cpu.c | 117 | ||||
-rw-r--r-- | arch/arm/kernel/perf_event_v7.c | 57 | ||||
-rw-r--r-- | arch/arm/kernel/psci.c | 49 | ||||
-rw-r--r-- | arch/arm/kernel/psci_smp.c | 84 | ||||
-rw-r--r-- | arch/arm/kernel/setup.c | 22 | ||||
-rw-r--r-- | arch/arm/kernel/sleep.S | 32 | ||||
-rw-r--r-- | arch/arm/kernel/smp.c | 28 | ||||
-rw-r--r-- | arch/arm/kernel/topology.c | 135 |
12 files changed, 504 insertions, 48 deletions
diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 5f3338eacad..dd9d90ab65d 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -82,6 +82,9 @@ obj-$(CONFIG_DEBUG_LL) += debug.o obj-$(CONFIG_EARLY_PRINTK) += early_printk.o obj-$(CONFIG_ARM_VIRT_EXT) += hyp-stub.o -obj-$(CONFIG_ARM_PSCI) += psci.o +ifeq ($(CONFIG_ARM_PSCI),y) +obj-y += psci.o +obj-$(CONFIG_SMP) += psci_smp.o +endif extra-y := $(head-y) vmlinux.lds diff --git a/arch/arm/kernel/head.S b/arch/arm/kernel/head.S index 8bac553fe21..2725c87fade 100644 --- a/arch/arm/kernel/head.S +++ b/arch/arm/kernel/head.S @@ -342,7 +342,6 @@ __turn_mmu_on_loc: .long __turn_mmu_on_end #if defined(CONFIG_SMP) - __CPUINIT ENTRY(secondary_startup) /* * Common entry point for secondary CPUs. diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index 1fd749ee4a1..1b803117ed9 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -1049,7 +1049,8 @@ static struct notifier_block dbg_cpu_pm_nb = { static void __init pm_init(void) { - cpu_pm_register_notifier(&dbg_cpu_pm_nb); + if (has_ossr) + cpu_pm_register_notifier(&dbg_cpu_pm_nb); } #else static inline void pm_init(void) diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index e19edc6f2d1..3a9d627cab6 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -12,6 +12,7 @@ */ #define pr_fmt(fmt) "hw perfevents: " fmt +#include <linux/cpumask.h> #include <linux/kernel.h> #include <linux/platform_device.h> #include <linux/pm_runtime.h> @@ -86,6 +87,9 @@ armpmu_map_event(struct perf_event *event, return armpmu_map_cache_event(cache_map, config); case PERF_TYPE_RAW: return armpmu_map_raw_event(raw_event_mask, config); + default: + if (event->attr.type >= PERF_TYPE_MAX) + return armpmu_map_raw_event(raw_event_mask, config); } return -ENOENT; @@ -163,6 +167,8 @@ armpmu_stop(struct perf_event *event, int flags) struct arm_pmu *armpmu = to_arm_pmu(event->pmu); struct hw_perf_event *hwc = &event->hw; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; /* * ARM pmu always has to update the counter, so ignore * PERF_EF_UPDATE, see comments in armpmu_start(). @@ -179,6 +185,8 @@ static void armpmu_start(struct perf_event *event, int flags) struct arm_pmu *armpmu = to_arm_pmu(event->pmu); struct hw_perf_event *hwc = &event->hw; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; /* * ARM pmu always has to reprogram the period, so ignore * PERF_EF_RELOAD, see the comment below. @@ -206,6 +214,9 @@ armpmu_del(struct perf_event *event, int flags) struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; + armpmu_stop(event, PERF_EF_UPDATE); hw_events->events[idx] = NULL; clear_bit(idx, hw_events->used_mask); @@ -222,6 +233,10 @@ armpmu_add(struct perf_event *event, int flags) int idx; int err = 0; + /* An event following a process won't be stopped earlier */ + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return 0; + perf_pmu_disable(event->pmu); /* If we don't have a space for the counter then finish early. */ @@ -424,6 +439,10 @@ static int armpmu_event_init(struct perf_event *event) int err = 0; atomic_t *active_events = &armpmu->active_events; + if (event->cpu != -1 && + !cpumask_test_cpu(event->cpu, &armpmu->valid_cpus)) + return -ENOENT; + /* does not support taken branch sampling */ if (has_branch_stack(event)) return -EOPNOTSUPP; diff --git a/arch/arm/kernel/perf_event_cpu.c b/arch/arm/kernel/perf_event_cpu.c index 1f2740e3dbc..0b48a38e3cf 100644 --- a/arch/arm/kernel/perf_event_cpu.c +++ b/arch/arm/kernel/perf_event_cpu.c @@ -19,6 +19,7 @@ #define pr_fmt(fmt) "CPU PMU: " fmt #include <linux/bitmap.h> +#include <linux/cpu_pm.h> #include <linux/export.h> #include <linux/kernel.h> #include <linux/of.h> @@ -31,33 +32,36 @@ #include <asm/pmu.h> /* Set at runtime when we know what CPU type we are. */ -static struct arm_pmu *cpu_pmu; +static DEFINE_PER_CPU(struct arm_pmu *, cpu_pmu); static DEFINE_PER_CPU(struct perf_event * [ARMPMU_MAX_HWEVENTS], hw_events); static DEFINE_PER_CPU(unsigned long [BITS_TO_LONGS(ARMPMU_MAX_HWEVENTS)], used_mask); static DEFINE_PER_CPU(struct pmu_hw_events, cpu_hw_events); +static DEFINE_PER_CPU(struct cpupmu_regs, cpu_pmu_regs); + /* * Despite the names, these two functions are CPU-specific and are used * by the OProfile/perf code. */ const char *perf_pmu_name(void) { - if (!cpu_pmu) + struct arm_pmu *pmu = per_cpu(cpu_pmu, 0); + if (!pmu) return NULL; - return cpu_pmu->name; + return pmu->name; } EXPORT_SYMBOL_GPL(perf_pmu_name); int perf_num_counters(void) { - int max_events = 0; + struct arm_pmu *pmu = per_cpu(cpu_pmu, 0); - if (cpu_pmu != NULL) - max_events = cpu_pmu->num_events; + if (!pmu) + return 0; - return max_events; + return pmu->num_events; } EXPORT_SYMBOL_GPL(perf_num_counters); @@ -75,11 +79,13 @@ static void cpu_pmu_free_irq(struct arm_pmu *cpu_pmu) { int i, irq, irqs; struct platform_device *pmu_device = cpu_pmu->plat_device; + int cpu = -1; irqs = min(pmu_device->num_resources, num_possible_cpus()); for (i = 0; i < irqs; ++i) { - if (!cpumask_test_and_clear_cpu(i, &cpu_pmu->active_irqs)) + cpu = cpumask_next(cpu, &cpu_pmu->valid_cpus); + if (!cpumask_test_and_clear_cpu(cpu, &cpu_pmu->active_irqs)) continue; irq = platform_get_irq(pmu_device, i); if (irq >= 0) @@ -91,6 +97,7 @@ static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) { int i, err, irq, irqs; struct platform_device *pmu_device = cpu_pmu->plat_device; + int cpu = -1; if (!pmu_device) return -ENODEV; @@ -103,6 +110,7 @@ static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) for (i = 0; i < irqs; ++i) { err = 0; + cpu = cpumask_next(cpu, &cpu_pmu->valid_cpus); irq = platform_get_irq(pmu_device, i); if (irq < 0) continue; @@ -112,7 +120,7 @@ static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) * assume that we're running on a uniprocessor machine and * continue. Otherwise, continue without this interrupt. */ - if (irq_set_affinity(irq, cpumask_of(i)) && irqs > 1) { + if (irq_set_affinity(irq, cpumask_of(cpu)) && irqs > 1) { pr_warning("unable to set irq affinity (irq=%d, cpu=%u)\n", irq, i); continue; @@ -126,7 +134,7 @@ static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) return err; } - cpumask_set_cpu(i, &cpu_pmu->active_irqs); + cpumask_set_cpu(cpu, &cpu_pmu->active_irqs); } return 0; @@ -135,7 +143,7 @@ static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) static void cpu_pmu_init(struct arm_pmu *cpu_pmu) { int cpu; - for_each_possible_cpu(cpu) { + for_each_cpu_mask(cpu, cpu_pmu->valid_cpus) { struct pmu_hw_events *events = &per_cpu(cpu_hw_events, cpu); events->events = per_cpu(hw_events, cpu); events->used_mask = per_cpu(used_mask, cpu); @@ -148,7 +156,7 @@ static void cpu_pmu_init(struct arm_pmu *cpu_pmu) /* Ensure the PMU has sane values out of reset. */ if (cpu_pmu->reset) - on_each_cpu(cpu_pmu->reset, cpu_pmu, 1); + on_each_cpu_mask(&cpu_pmu->valid_cpus, cpu_pmu->reset, cpu_pmu, 1); } /* @@ -160,21 +168,46 @@ static void cpu_pmu_init(struct arm_pmu *cpu_pmu) static int __cpuinit cpu_pmu_notify(struct notifier_block *b, unsigned long action, void *hcpu) { + struct arm_pmu *pmu = per_cpu(cpu_pmu, (long)hcpu); + if ((action & ~CPU_TASKS_FROZEN) != CPU_STARTING) return NOTIFY_DONE; - if (cpu_pmu && cpu_pmu->reset) - cpu_pmu->reset(cpu_pmu); + if (pmu && pmu->reset) + pmu->reset(pmu); else return NOTIFY_DONE; return NOTIFY_OK; } +static int cpu_pmu_pm_notify(struct notifier_block *b, + unsigned long action, void *hcpu) +{ + int cpu = smp_processor_id(); + struct arm_pmu *pmu = per_cpu(cpu_pmu, cpu); + struct cpupmu_regs *pmuregs = &per_cpu(cpu_pmu_regs, cpu); + + if (!pmu) + return NOTIFY_DONE; + + if (action == CPU_PM_ENTER && pmu->save_regs) { + pmu->save_regs(pmu, pmuregs); + } else if (action == CPU_PM_EXIT && pmu->restore_regs) { + pmu->restore_regs(pmu, pmuregs); + } + + return NOTIFY_OK; +} + static struct notifier_block __cpuinitdata cpu_pmu_hotplug_notifier = { .notifier_call = cpu_pmu_notify, }; +static struct notifier_block __cpuinitdata cpu_pmu_pm_notifier = { + .notifier_call = cpu_pmu_pm_notify, +}; + /* * PMU platform driver and devicetree bindings. */ @@ -246,6 +279,9 @@ static int probe_current_pmu(struct arm_pmu *pmu) } } + /* assume PMU support all the CPUs in this case */ + cpumask_setall(&pmu->valid_cpus); + put_cpu(); return ret; } @@ -253,15 +289,10 @@ static int probe_current_pmu(struct arm_pmu *pmu) static int cpu_pmu_device_probe(struct platform_device *pdev) { const struct of_device_id *of_id; - int (*init_fn)(struct arm_pmu *); struct device_node *node = pdev->dev.of_node; struct arm_pmu *pmu; - int ret = -ENODEV; - - if (cpu_pmu) { - pr_info("attempt to register multiple PMU devices!"); - return -ENOSPC; - } + int ret = 0; + int cpu; pmu = kzalloc(sizeof(struct arm_pmu), GFP_KERNEL); if (!pmu) { @@ -270,8 +301,28 @@ static int cpu_pmu_device_probe(struct platform_device *pdev) } if (node && (of_id = of_match_node(cpu_pmu_of_device_ids, pdev->dev.of_node))) { - init_fn = of_id->data; - ret = init_fn(pmu); + smp_call_func_t init_fn = (smp_call_func_t)of_id->data; + struct device_node *ncluster; + int cluster = -1; + cpumask_t sibling_mask; + + ncluster = of_parse_phandle(node, "cluster", 0); + if (ncluster) { + int len; + const u32 *hwid; + hwid = of_get_property(ncluster, "reg", &len); + if (hwid && len == 4) + cluster = be32_to_cpup(hwid); + } + /* set sibling mask to all cpu mask if socket is not specified */ + if (cluster == -1 || + cluster_to_logical_mask(cluster, &sibling_mask)) + cpumask_setall(&sibling_mask); + + smp_call_function_any(&sibling_mask, init_fn, pmu, 1); + + /* now set the valid_cpus after init */ + cpumask_copy(&pmu->valid_cpus, &sibling_mask); } else { ret = probe_current_pmu(pmu); } @@ -281,10 +332,12 @@ static int cpu_pmu_device_probe(struct platform_device *pdev) goto out_free; } - cpu_pmu = pmu; - cpu_pmu->plat_device = pdev; - cpu_pmu_init(cpu_pmu); - ret = armpmu_register(cpu_pmu, PERF_TYPE_RAW); + for_each_cpu_mask(cpu, pmu->valid_cpus) + per_cpu(cpu_pmu, cpu) = pmu; + + pmu->plat_device = pdev; + cpu_pmu_init(pmu); + ret = armpmu_register(pmu, -1); if (!ret) return 0; @@ -313,9 +366,17 @@ static int __init register_pmu_driver(void) if (err) return err; + err = cpu_pm_register_notifier(&cpu_pmu_pm_notifier); + if (err) { + unregister_cpu_notifier(&cpu_pmu_hotplug_notifier); + return err; + } + err = platform_driver_register(&cpu_pmu_driver); - if (err) + if (err) { + cpu_pm_unregister_notifier(&cpu_pmu_pm_notifier); unregister_cpu_notifier(&cpu_pmu_hotplug_notifier); + } return err; } diff --git a/arch/arm/kernel/perf_event_v7.c b/arch/arm/kernel/perf_event_v7.c index 039cffb053a..654db5030c3 100644 --- a/arch/arm/kernel/perf_event_v7.c +++ b/arch/arm/kernel/perf_event_v7.c @@ -950,6 +950,51 @@ static void armv7_pmnc_dump_regs(struct arm_pmu *cpu_pmu) } #endif +static void armv7pmu_save_regs(struct arm_pmu *cpu_pmu, + struct cpupmu_regs *regs) +{ + unsigned int cnt; + asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (regs->pmc)); + if (!(regs->pmc & ARMV7_PMNC_E)) + return; + + asm volatile("mrc p15, 0, %0, c9, c12, 1" : "=r" (regs->pmcntenset)); + asm volatile("mrc p15, 0, %0, c9, c14, 0" : "=r" (regs->pmuseren)); + asm volatile("mrc p15, 0, %0, c9, c14, 1" : "=r" (regs->pmintenset)); + asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (regs->pmxevtcnt[0])); + for (cnt = ARMV7_IDX_COUNTER0; + cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { + armv7_pmnc_select_counter(cnt); + asm volatile("mrc p15, 0, %0, c9, c13, 1" + : "=r"(regs->pmxevttype[cnt])); + asm volatile("mrc p15, 0, %0, c9, c13, 2" + : "=r"(regs->pmxevtcnt[cnt])); + } + return; +} + +static void armv7pmu_restore_regs(struct arm_pmu *cpu_pmu, + struct cpupmu_regs *regs) +{ + unsigned int cnt; + if (!(regs->pmc & ARMV7_PMNC_E)) + return; + + asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (regs->pmcntenset)); + asm volatile("mcr p15, 0, %0, c9, c14, 0" : : "r" (regs->pmuseren)); + asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (regs->pmintenset)); + asm volatile("mcr p15, 0, %0, c9, c13, 0" : : "r" (regs->pmxevtcnt[0])); + for (cnt = ARMV7_IDX_COUNTER0; + cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { + armv7_pmnc_select_counter(cnt); + asm volatile("mcr p15, 0, %0, c9, c13, 1" + : : "r"(regs->pmxevttype[cnt])); + asm volatile("mcr p15, 0, %0, c9, c13, 2" + : : "r"(regs->pmxevtcnt[cnt])); + } + asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r" (regs->pmc)); +} + static void armv7pmu_enable_event(struct perf_event *event) { unsigned long flags; @@ -1223,6 +1268,8 @@ static void armv7pmu_init(struct arm_pmu *cpu_pmu) cpu_pmu->start = armv7pmu_start; cpu_pmu->stop = armv7pmu_stop; cpu_pmu->reset = armv7pmu_reset; + cpu_pmu->save_regs = armv7pmu_save_regs; + cpu_pmu->restore_regs = armv7pmu_restore_regs; cpu_pmu->max_period = (1LLU << 32) - 1; }; @@ -1240,7 +1287,7 @@ static u32 armv7_read_num_pmnc_events(void) static int armv7_a8_pmu_init(struct arm_pmu *cpu_pmu) { armv7pmu_init(cpu_pmu); - cpu_pmu->name = "ARMv7 Cortex-A8"; + cpu_pmu->name = "ARMv7_Cortex_A8"; cpu_pmu->map_event = armv7_a8_map_event; cpu_pmu->num_events = armv7_read_num_pmnc_events(); return 0; @@ -1249,7 +1296,7 @@ static int armv7_a8_pmu_init(struct arm_pmu *cpu_pmu) static int armv7_a9_pmu_init(struct arm_pmu *cpu_pmu) { armv7pmu_init(cpu_pmu); - cpu_pmu->name = "ARMv7 Cortex-A9"; + cpu_pmu->name = "ARMv7_Cortex_A9"; cpu_pmu->map_event = armv7_a9_map_event; cpu_pmu->num_events = armv7_read_num_pmnc_events(); return 0; @@ -1258,7 +1305,7 @@ static int armv7_a9_pmu_init(struct arm_pmu *cpu_pmu) static int armv7_a5_pmu_init(struct arm_pmu *cpu_pmu) { armv7pmu_init(cpu_pmu); - cpu_pmu->name = "ARMv7 Cortex-A5"; + cpu_pmu->name = "ARMv7_Cortex_A5"; cpu_pmu->map_event = armv7_a5_map_event; cpu_pmu->num_events = armv7_read_num_pmnc_events(); return 0; @@ -1267,7 +1314,7 @@ static int armv7_a5_pmu_init(struct arm_pmu *cpu_pmu) static int armv7_a15_pmu_init(struct arm_pmu *cpu_pmu) { armv7pmu_init(cpu_pmu); - cpu_pmu->name = "ARMv7 Cortex-A15"; + cpu_pmu->name = "ARMv7_Cortex_A15"; cpu_pmu->map_event = armv7_a15_map_event; cpu_pmu->num_events = armv7_read_num_pmnc_events(); cpu_pmu->set_event_filter = armv7pmu_set_event_filter; @@ -1277,7 +1324,7 @@ static int armv7_a15_pmu_init(struct arm_pmu *cpu_pmu) static int armv7_a7_pmu_init(struct arm_pmu *cpu_pmu) { armv7pmu_init(cpu_pmu); - cpu_pmu->name = "ARMv7 Cortex-A7"; + cpu_pmu->name = "ARMv7_Cortex_A7"; cpu_pmu->map_event = armv7_a7_map_event; cpu_pmu->num_events = armv7_read_num_pmnc_events(); cpu_pmu->set_event_filter = armv7pmu_set_event_filter; diff --git a/arch/arm/kernel/psci.c b/arch/arm/kernel/psci.c index 36531643cc2..0daf4f25228 100644 --- a/arch/arm/kernel/psci.c +++ b/arch/arm/kernel/psci.c @@ -17,6 +17,7 @@ #include <linux/init.h> #include <linux/of.h> +#include <linux/string.h> #include <asm/compiler.h> #include <asm/errno.h> @@ -26,6 +27,11 @@ struct psci_operations psci_ops; +/* Type of psci support. Currently can only be enabled or disabled */ +#define PSCI_SUP_DISABLED 0 +#define PSCI_SUP_ENABLED 1 + +static unsigned int psci; static int (*invoke_psci_fn)(u32, u32, u32, u32); enum psci_function { @@ -42,6 +48,7 @@ static u32 psci_function_id[PSCI_FN_MAX]; #define PSCI_RET_EOPNOTSUPP -1 #define PSCI_RET_EINVAL -2 #define PSCI_RET_EPERM -3 +#define PSCI_RET_EALREADYON -4 static int psci_to_linux_errno(int errno) { @@ -54,6 +61,8 @@ static int psci_to_linux_errno(int errno) return -EINVAL; case PSCI_RET_EPERM: return -EPERM; + case PSCI_RET_EALREADYON: + return -EAGAIN; }; return -EINVAL; @@ -158,15 +167,18 @@ static const struct of_device_id psci_of_match[] __initconst = { {}, }; -static int __init psci_init(void) +void __init psci_init(void) { struct device_node *np; const char *method; u32 id; + if (psci == PSCI_SUP_DISABLED) + return; + np = of_find_matching_node(NULL, psci_of_match); if (!np) - return 0; + return; pr_info("probing function IDs from device-tree\n"); @@ -206,6 +218,35 @@ static int __init psci_init(void) out_put_node: of_node_put(np); - return 0; + return; +} + +int __init psci_probe(void) +{ + struct device_node *np; + int ret = -ENODEV; + + if (psci == PSCI_SUP_ENABLED) { + np = of_find_matching_node(NULL, psci_of_match); + if (np) + ret = 0; + } + + of_node_put(np); + return ret; +} + +static int __init early_psci(char *val) +{ + int ret = 0; + + if (strcmp(val, "enable") == 0) + psci = PSCI_SUP_ENABLED; + else if (strcmp(val, "disable") == 0) + psci = PSCI_SUP_DISABLED; + else + ret = -EINVAL; + + return ret; } -early_initcall(psci_init); +early_param("psci", early_psci); diff --git a/arch/arm/kernel/psci_smp.c b/arch/arm/kernel/psci_smp.c new file mode 100644 index 00000000000..23a11424c56 --- /dev/null +++ b/arch/arm/kernel/psci_smp.c @@ -0,0 +1,84 @@ +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 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. See the + * GNU General Public License for more details. + * + * Copyright (C) 2012 ARM Limited + * + * Author: Will Deacon <will.deacon@arm.com> + */ + +#include <linux/init.h> +#include <linux/irqchip/arm-gic.h> +#include <linux/smp.h> +#include <linux/of.h> + +#include <asm/psci.h> +#include <asm/smp_plat.h> + +/* + * psci_smp assumes that the following is true about PSCI: + * + * cpu_suspend Suspend the execution on a CPU + * @state we don't currently describe affinity levels, so just pass 0. + * @entry_point the first instruction to be executed on return + * returns 0 success, < 0 on failure + * + * cpu_off Power down a CPU + * @state we don't currently describe affinity levels, so just pass 0. + * no return on successful call + * + * cpu_on Power up a CPU + * @cpuid cpuid of target CPU, as from MPIDR + * @entry_point the first instruction to be executed on return + * returns 0 success, < 0 on failure + * + * migrate Migrate the context to a different CPU + * @cpuid cpuid of target CPU, as from MPIDR + * returns 0 success, < 0 on failure + * + */ + +extern void secondary_startup(void); + +static int __cpuinit psci_boot_secondary(unsigned int cpu, + struct task_struct *idle) +{ + if (psci_ops.cpu_on) + return psci_ops.cpu_on(cpu_logical_map(cpu), + __pa(secondary_startup)); + return -ENODEV; +} + +#ifdef CONFIG_HOTPLUG_CPU +void __ref psci_cpu_die(unsigned int cpu) +{ + const struct psci_power_state ps = { + .type = PSCI_POWER_STATE_TYPE_POWER_DOWN, + }; + + if (psci_ops.cpu_off) + psci_ops.cpu_off(ps); + + /* We should never return */ + panic("psci: cpu %d failed to shutdown\n", cpu); +} +#else +#define psci_cpu_die NULL +#endif + +bool __init psci_smp_available(void) +{ + /* is cpu_on available at least? */ + return (psci_ops.cpu_on != NULL); +} + +struct smp_operations __initdata psci_smp_ops = { + .smp_boot_secondary = psci_boot_secondary, + .cpu_die = psci_cpu_die, +}; diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index b4b1d397592..6002fe06b12 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -37,6 +37,7 @@ #include <asm/cputype.h> #include <asm/elf.h> #include <asm/procinfo.h> +#include <asm/psci.h> #include <asm/sections.h> #include <asm/setup.h> #include <asm/smp_plat.h> @@ -261,6 +262,19 @@ static int cpu_has_aliasing_icache(unsigned int arch) int aliasing_icache; unsigned int id_reg, num_sets, line_size; +#ifdef CONFIG_BIG_LITTLE + /* + * We expect a combination of Cortex-A15 and Cortex-A7 cores. + * A7 = VIPT aliasing I-cache + * A15 = PIPT (non-aliasing) I-cache + * To cater for this discrepancy, let's assume aliasing I-cache + * all the time. This means unneeded extra work on the A15 but + * only ptrace is affected which is not performance critical. + */ + if ((read_cpuid_id() & 0xff0ffff0) == 0x410fc0f0) + return 1; +#endif + /* PIPT caches never alias. */ if (icache_is_pipt()) return 0; @@ -796,9 +810,15 @@ void __init setup_arch(char **cmdline_p) unflatten_device_tree(); arm_dt_init_cpu_maps(); + psci_init(); #ifdef CONFIG_SMP if (is_smp()) { - smp_set_ops(mdesc->smp); + if (!mdesc->smp_init || !mdesc->smp_init()) { + if (psci_smp_available()) + smp_set_ops(&psci_smp_ops); + else if (mdesc->smp) + smp_set_ops(mdesc->smp); + } smp_init_cpus(); } #endif diff --git a/arch/arm/kernel/sleep.S b/arch/arm/kernel/sleep.S index 987dcf33415..b5c1e636ed8 100644 --- a/arch/arm/kernel/sleep.S +++ b/arch/arm/kernel/sleep.S @@ -4,6 +4,7 @@ #include <asm/assembler.h> #include <asm/glue-cache.h> #include <asm/glue-proc.h> +#include "entry-header.S" .text /* @@ -30,9 +31,8 @@ ENTRY(__cpu_suspend) mov r2, r5 @ virtual SP ldr r3, =sleep_save_sp #ifdef CONFIG_SMP - ALT_SMP(mrc p15, 0, lr, c0, c0, 5) - ALT_UP(mov lr, #0) - and lr, lr, #15 + get_thread_info r5 + ldr lr, [r5, #TI_CPU] @ cpu logical index add r3, r3, lr, lsl #2 #endif bl __cpu_suspend_save @@ -82,10 +82,13 @@ ENDPROC(cpu_resume_after_mmu) .align ENTRY(cpu_resume) #ifdef CONFIG_SMP + mov r1, #0 @ fall-back logical index for UP + ALT_SMP(mrc p15, 0, r0, c0, c0, 5) + ALT_UP_B(1f) + bic r0, #0xff000000 + bl cpu_logical_index @ return logical index in r1 +1: adr r0, sleep_save_sp - ALT_SMP(mrc p15, 0, r1, c0, c0, 5) - ALT_UP(mov r1, #0) - and r1, r1, #15 ldr r0, [r0, r1, lsl #2] @ stack phys addr #else ldr r0, sleep_save_sp @ stack phys addr @@ -102,3 +105,20 @@ sleep_save_sp: .rept CONFIG_NR_CPUS .long 0 @ preserve stack phys ptr here .endr + +#ifdef CONFIG_SMP +cpu_logical_index: + adr r3, cpu_map_ptr + ldr r2, [r3] + add r3, r3, r2 @ virt_to_phys(__cpu_logical_map) + mov r1, #0 +1: + ldr r2, [r3, r1, lsl #2] + cmp r2, r0 + moveq pc, lr + add r1, r1, #1 + b 1b + +cpu_map_ptr: + .long __cpu_logical_map - . +#endif diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 5919eb451bb..dc2843f337a 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -46,6 +46,9 @@ #include <asm/virt.h> #include <asm/mach/arch.h> +#define CREATE_TRACE_POINTS +#include <trace/events/arm-ipi.h> + /* * as from 2.5, kernels no longer have an init_tasks structure * so we need some other way of telling a new secondary core @@ -57,7 +60,7 @@ struct secondary_data secondary_data; * control for which core is the next to come out of the secondary * boot "holding pen" */ -volatile int __cpuinitdata pen_release = -1; +volatile int pen_release = -1; enum ipi_msg_type { IPI_WAKEUP, @@ -66,6 +69,7 @@ enum ipi_msg_type { IPI_CALL_FUNC, IPI_CALL_FUNC_SINGLE, IPI_CPU_STOP, + IPI_COMPLETION, }; static DECLARE_COMPLETION(cpu_running); @@ -463,6 +467,7 @@ static const char *ipi_types[NR_IPI] = { S(IPI_CALL_FUNC, "Function call interrupts"), S(IPI_CALL_FUNC_SINGLE, "Single function call interrupts"), S(IPI_CPU_STOP, "CPU stop interrupts"), + S(IPI_COMPLETION, "completion interrupts"), }; void show_ipi_list(struct seq_file *p, int prec) @@ -588,6 +593,19 @@ static void ipi_cpu_stop(unsigned int cpu) cpu_relax(); } +static DEFINE_PER_CPU(struct completion *, cpu_completion); + +int register_ipi_completion(struct completion *completion, int cpu) +{ + per_cpu(cpu_completion, cpu) = completion; + return IPI_COMPLETION; +} + +static void ipi_complete(unsigned int cpu) +{ + complete(per_cpu(cpu_completion, cpu)); +} + /* * Main handler for inter-processor interrupts */ @@ -604,6 +622,7 @@ void handle_IPI(int ipinr, struct pt_regs *regs) if (ipinr < NR_IPI) __inc_irq_stat(cpu, ipi_irqs[ipinr]); + trace_arm_ipi_entry(ipinr); switch (ipinr) { case IPI_WAKEUP: break; @@ -638,11 +657,18 @@ void handle_IPI(int ipinr, struct pt_regs *regs) irq_exit(); break; + case IPI_COMPLETION: + irq_enter(); + ipi_complete(cpu); + irq_exit(); + break; + default: printk(KERN_CRIT "CPU%u: Unknown IPI message 0x%x\n", cpu, ipinr); break; } + trace_arm_ipi_exit(ipinr); set_irq_regs(old_regs); } diff --git a/arch/arm/kernel/topology.c b/arch/arm/kernel/topology.c index c5a59546a25..677da58d9e8 100644 --- a/arch/arm/kernel/topology.c +++ b/arch/arm/kernel/topology.c @@ -23,6 +23,7 @@ #include <linux/slab.h> #include <asm/cputype.h> +#include <asm/smp_plat.h> #include <asm/topology.h> /* @@ -289,6 +290,140 @@ void store_cpu_topology(unsigned int cpuid) cpu_topology[cpuid].socket_id, mpidr); } + +#ifdef CONFIG_SCHED_HMP + +static const char * const little_cores[] = { + "arm,cortex-a7", + NULL, +}; + +static bool is_little_cpu(struct device_node *cn) +{ + const char * const *lc; + for (lc = little_cores; *lc; lc++) + if (of_device_is_compatible(cn, *lc)) + return true; + return false; +} + +void __init arch_get_fast_and_slow_cpus(struct cpumask *fast, + struct cpumask *slow) +{ + struct device_node *cn = NULL; + int cpu; + + cpumask_clear(fast); + cpumask_clear(slow); + + /* + * Use the config options if they are given. This helps testing + * HMP scheduling on systems without a big.LITTLE architecture. + */ + if (strlen(CONFIG_HMP_FAST_CPU_MASK) && strlen(CONFIG_HMP_SLOW_CPU_MASK)) { + if (cpulist_parse(CONFIG_HMP_FAST_CPU_MASK, fast)) + WARN(1, "Failed to parse HMP fast cpu mask!\n"); + if (cpulist_parse(CONFIG_HMP_SLOW_CPU_MASK, slow)) + WARN(1, "Failed to parse HMP slow cpu mask!\n"); + return; + } + + /* + * Else, parse device tree for little cores. + */ + while ((cn = of_find_node_by_type(cn, "cpu"))) { + + const u32 *mpidr; + int len; + + mpidr = of_get_property(cn, "reg", &len); + if (!mpidr || len != 4) { + pr_err("* %s missing reg property\n", cn->full_name); + continue; + } + + cpu = get_logical_index(be32_to_cpup(mpidr)); + if (cpu == -EINVAL) { + pr_err("couldn't get logical index for mpidr %x\n", + be32_to_cpup(mpidr)); + break; + } + + if (is_little_cpu(cn)) + cpumask_set_cpu(cpu, slow); + else + cpumask_set_cpu(cpu, fast); + } + + if (!cpumask_empty(fast) && !cpumask_empty(slow)) + return; + + /* + * We didn't find both big and little cores so let's call all cores + * fast as this will keep the system running, with all cores being + * treated equal. + */ + cpumask_setall(fast); + cpumask_clear(slow); +} + +struct cpumask hmp_slow_cpu_mask; + +void __init arch_get_hmp_domains(struct list_head *hmp_domains_list) +{ + struct cpumask hmp_fast_cpu_mask; + struct hmp_domain *domain; + + arch_get_fast_and_slow_cpus(&hmp_fast_cpu_mask, &hmp_slow_cpu_mask); + + /* + * Initialize hmp_domains + * Must be ordered with respect to compute capacity. + * Fastest domain at head of list. + */ + if(!cpumask_empty(&hmp_slow_cpu_mask)) { + domain = (struct hmp_domain *) + kmalloc(sizeof(struct hmp_domain), GFP_KERNEL); + cpumask_copy(&domain->possible_cpus, &hmp_slow_cpu_mask); + cpumask_and(&domain->cpus, cpu_online_mask, &domain->possible_cpus); + list_add(&domain->hmp_domains, hmp_domains_list); + } + domain = (struct hmp_domain *) + kmalloc(sizeof(struct hmp_domain), GFP_KERNEL); + cpumask_copy(&domain->possible_cpus, &hmp_fast_cpu_mask); + cpumask_and(&domain->cpus, cpu_online_mask, &domain->possible_cpus); + list_add(&domain->hmp_domains, hmp_domains_list); +} +#endif /* CONFIG_SCHED_HMP */ + + +/* + * cluster_to_logical_mask - return cpu logical mask of CPUs in a cluster + * @socket_id: cluster HW identifier + * @cluster_mask: the cpumask location to be initialized, modified by the + * function only if return value == 0 + * + * Return: + * + * 0 on success + * -EINVAL if cluster_mask is NULL or there is no record matching socket_id + */ +int cluster_to_logical_mask(unsigned int socket_id, cpumask_t *cluster_mask) +{ + int cpu; + + if (!cluster_mask) + return -EINVAL; + + for_each_online_cpu(cpu) + if (socket_id == topology_physical_package_id(cpu)) { + cpumask_copy(cluster_mask, topology_core_cpumask(cpu)); + return 0; + } + + return -EINVAL; +} + /* * init_cpu_topology is called at boot when only one cpu is running * which prevent simultaneous write access to cpu_topology array |