From 8691e5a8f691cc2a4fda0651e8d307aaba0e7d68 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 6 Jun 2008 11:18:06 +0200 Subject: smp_call_function: get rid of the unused nonatomic/retry argument It's never used and the comments refer to nonatomic and retry interchangably. So get rid of it. Acked-by: Jeremy Fitzhardinge Signed-off-by: Jens Axboe --- drivers/acpi/processor_idle.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/acpi/processor_idle.c') diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 556ee158519..4976e5db2b3 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -1339,7 +1339,7 @@ static void smp_callback(void *v) static int acpi_processor_latency_notify(struct notifier_block *b, unsigned long l, void *v) { - smp_call_function(smp_callback, NULL, 0, 1); + smp_call_function(smp_callback, NULL, 1); return NOTIFY_OK; } -- cgit v1.2.3-70-g09d2 From 706546d02384b64e083bd9130c56eaa599c66038 Mon Sep 17 00:00:00 2001 From: Mike Travis Date: Mon, 9 Jun 2008 16:22:23 -0700 Subject: ACPI: change processors from array to per_cpu variable Change processors from an array sized by NR_CPUS to a per_cpu variable. Signed-off-by: Mike Travis Signed-off-by: Andrew Morton Signed-off-by: Len Brown Signed-off-by: Andi Kleen --- drivers/acpi/processor_core.c | 20 +++++++++----------- drivers/acpi/processor_idle.c | 8 ++++---- drivers/acpi/processor_perflib.c | 18 +++++++++--------- drivers/acpi/processor_throttling.c | 14 +++++++------- include/acpi/processor.h | 2 +- 5 files changed, 30 insertions(+), 32 deletions(-) (limited to 'drivers/acpi/processor_idle.c') diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c index feddc8ed870..9a803f85ccf 100644 --- a/drivers/acpi/processor_core.c +++ b/drivers/acpi/processor_core.c @@ -118,7 +118,7 @@ static const struct file_operations acpi_processor_info_fops = { .release = single_release, }; -struct acpi_processor *processors[NR_CPUS]; +DEFINE_PER_CPU(struct acpi_processor *, processors); struct acpi_processor_errata errata __read_mostly; /* -------------------------------------------------------------------------- @@ -614,7 +614,7 @@ static int acpi_processor_get_info(struct acpi_processor *pr, unsigned has_uid) return 0; } -static void *processor_device_array[NR_CPUS]; +static DEFINE_PER_CPU(void *, processor_device_array); static int __cpuinit acpi_processor_start(struct acpi_device *device) { @@ -638,15 +638,15 @@ static int __cpuinit acpi_processor_start(struct acpi_device *device) * ACPI id of processors can be reported wrongly by the BIOS. * Don't trust it blindly */ - if (processor_device_array[pr->id] != NULL && - processor_device_array[pr->id] != device) { + if (per_cpu(processor_device_array, pr->id) != NULL && + per_cpu(processor_device_array, pr->id) != device) { printk(KERN_WARNING "BIOS reported wrong ACPI id " "for the processor\n"); return -ENODEV; } - processor_device_array[pr->id] = device; + per_cpu(processor_device_array, pr->id) = device; - processors[pr->id] = pr; + per_cpu(processors, pr->id) = pr; result = acpi_processor_add_fs(device); if (result) @@ -753,7 +753,7 @@ static int acpi_cpu_soft_notify(struct notifier_block *nfb, unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; - struct acpi_processor *pr = processors[cpu]; + struct acpi_processor *pr = per_cpu(processors, cpu); if (action == CPU_ONLINE && pr) { acpi_processor_ppc_has_changed(pr); @@ -825,8 +825,8 @@ static int acpi_processor_remove(struct acpi_device *device, int type) pr->cdev = NULL; } - processors[pr->id] = NULL; - processor_device_array[pr->id] = NULL; + per_cpu(processors, pr->id) = NULL; + per_cpu(processor_device_array, pr->id) = NULL; kfree(pr); return 0; @@ -1074,8 +1074,6 @@ static int __init acpi_processor_init(void) { int result = 0; - - memset(&processors, 0, sizeof(processors)); memset(&errata, 0, sizeof(errata)); #ifdef CONFIG_SMP diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 4976e5db2b3..0fc310e7dfd 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -401,7 +401,7 @@ static void acpi_processor_idle(void) */ local_irq_disable(); - pr = processors[smp_processor_id()]; + pr = __get_cpu_var(processors); if (!pr) { local_irq_enable(); return; @@ -1431,7 +1431,7 @@ static int acpi_idle_enter_c1(struct cpuidle_device *dev, struct acpi_processor *pr; struct acpi_processor_cx *cx = cpuidle_get_statedata(state); - pr = processors[smp_processor_id()]; + pr = __get_cpu_var(processors); if (unlikely(!pr)) return 0; @@ -1471,7 +1471,7 @@ static int acpi_idle_enter_simple(struct cpuidle_device *dev, u32 t1, t2; int sleep_ticks = 0; - pr = processors[smp_processor_id()]; + pr = __get_cpu_var(processors); if (unlikely(!pr)) return 0; @@ -1549,7 +1549,7 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev, u32 t1, t2; int sleep_ticks = 0; - pr = processors[smp_processor_id()]; + pr = __get_cpu_var(processors); if (unlikely(!pr)) return 0; diff --git a/drivers/acpi/processor_perflib.c b/drivers/acpi/processor_perflib.c index d80b2d1441a..b4749969c6b 100644 --- a/drivers/acpi/processor_perflib.c +++ b/drivers/acpi/processor_perflib.c @@ -89,7 +89,7 @@ static int acpi_processor_ppc_notifier(struct notifier_block *nb, if (event != CPUFREQ_INCOMPATIBLE) goto out; - pr = processors[policy->cpu]; + pr = per_cpu(processors, policy->cpu); if (!pr || !pr->performance) goto out; @@ -572,7 +572,7 @@ int acpi_processor_preregister_performance( /* Call _PSD for all CPUs */ for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) { /* Look only at processors in ACPI namespace */ continue; @@ -603,7 +603,7 @@ int acpi_processor_preregister_performance( * domain info. */ for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) continue; @@ -624,7 +624,7 @@ int acpi_processor_preregister_performance( cpus_clear(covered_cpus); for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) continue; @@ -651,7 +651,7 @@ int acpi_processor_preregister_performance( if (i == j) continue; - match_pr = processors[j]; + match_pr = per_cpu(processors, j); if (!match_pr) continue; @@ -680,7 +680,7 @@ int acpi_processor_preregister_performance( if (i == j) continue; - match_pr = processors[j]; + match_pr = per_cpu(processors, j); if (!match_pr) continue; @@ -697,7 +697,7 @@ int acpi_processor_preregister_performance( err_ret: for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr || !pr->performance) continue; @@ -728,7 +728,7 @@ acpi_processor_register_performance(struct acpi_processor_performance mutex_lock(&performance_mutex); - pr = processors[cpu]; + pr = per_cpu(processors, cpu); if (!pr) { mutex_unlock(&performance_mutex); return -ENODEV; @@ -766,7 +766,7 @@ acpi_processor_unregister_performance(struct acpi_processor_performance mutex_lock(&performance_mutex); - pr = processors[cpu]; + pr = per_cpu(processors, cpu); if (!pr) { mutex_unlock(&performance_mutex); return; diff --git a/drivers/acpi/processor_throttling.c b/drivers/acpi/processor_throttling.c index bb06738860c..8728782aad3 100644 --- a/drivers/acpi/processor_throttling.c +++ b/drivers/acpi/processor_throttling.c @@ -71,7 +71,7 @@ static int acpi_processor_update_tsd_coord(void) * coordination between all CPUs. */ for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) continue; @@ -93,7 +93,7 @@ static int acpi_processor_update_tsd_coord(void) cpus_clear(covered_cpus); for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) continue; @@ -119,7 +119,7 @@ static int acpi_processor_update_tsd_coord(void) if (i == j) continue; - match_pr = processors[j]; + match_pr = per_cpu(processors, j); if (!match_pr) continue; @@ -152,7 +152,7 @@ static int acpi_processor_update_tsd_coord(void) if (i == j) continue; - match_pr = processors[j]; + match_pr = per_cpu(processors, j); if (!match_pr) continue; @@ -172,7 +172,7 @@ static int acpi_processor_update_tsd_coord(void) err_ret: for_each_possible_cpu(i) { - pr = processors[i]; + pr = per_cpu(processors, i); if (!pr) continue; @@ -214,7 +214,7 @@ static int acpi_processor_throttling_notifier(unsigned long event, void *data) struct acpi_processor_throttling *p_throttling; cpu = p_tstate->cpu; - pr = processors[cpu]; + pr = per_cpu(processors, cpu); if (!pr) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Invalid pr pointer\n")); return 0; @@ -1035,7 +1035,7 @@ int acpi_processor_set_throttling(struct acpi_processor *pr, int state) * cpus. */ for_each_cpu_mask(i, online_throttling_cpus) { - match_pr = processors[i]; + match_pr = per_cpu(processors, i); /* * If the pointer is invalid, we will report the * error message and continue. diff --git a/include/acpi/processor.h b/include/acpi/processor.h index 06ebb6ef72a..3795590e152 100644 --- a/include/acpi/processor.h +++ b/include/acpi/processor.h @@ -255,7 +255,7 @@ extern void acpi_processor_unregister_performance(struct int acpi_processor_notify_smm(struct module *calling_module); /* for communication between multiple parts of the processor kernel module */ -extern struct acpi_processor *processors[NR_CPUS]; +DECLARE_PER_CPU(struct acpi_processor *, processors); extern struct acpi_processor_errata errata; void arch_acpi_processor_init_pdc(struct acpi_processor *pr); -- cgit v1.2.3-70-g09d2 From c1e3b377ad48febba6f91b8ae42c44ee4d4ab45e Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Tue, 24 Jun 2008 17:58:53 +0800 Subject: ACPI: Create "idle=halt" bootparam "idle=halt" limits the idle loop to using the halt instruction. No MWAIT, no IO accesses, no C-states deeper than C1. If something is broken in the idle code, "idle=halt" is a less severe workaround than "idle=poll" which disables all power savings. Signed-off-by: Zhao Yakui Signed-off-by: Len Brown Signed-off-by: Andi Kleen --- Documentation/kernel-parameters.txt | 4 +++- arch/ia64/kernel/process.c | 2 ++ arch/x86/kernel/process.c | 17 ++++++++++++++++- drivers/acpi/processor_idle.c | 22 ++++++++++++++++++++++ include/asm-ia64/processor.h | 1 + include/asm-x86/processor.h | 1 + 6 files changed, 45 insertions(+), 2 deletions(-) (limited to 'drivers/acpi/processor_idle.c') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 312fe77764a..65db7f4711a 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -818,7 +818,7 @@ and is between 256 and 4096 characters. It is defined in the file See Documentation/ide/ide.txt. idle= [X86] - Format: idle=poll or idle=mwait + Format: idle=poll or idle=mwait, idle=halt Poll forces a polling idle loop that can slightly improves the performance of waking up a idle CPU, but will use a lot of power and make the system run hot. Not recommended. @@ -826,6 +826,8 @@ and is between 256 and 4096 characters. It is defined in the file to not use it because it doesn't save as much power as a normal idle loop use the MONITOR/MWAIT idle loop anyways. Performance should be the same as idle=poll. + idle=halt. Halt is forced to be used for CPU idle. + In such case C2/C3 won't be used again. ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem Claim all unknown PCI IDE storage controllers. diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c index fabaf08d9a6..612b3c4a060 100644 --- a/arch/ia64/kernel/process.c +++ b/arch/ia64/kernel/process.c @@ -55,6 +55,8 @@ void (*ia64_mark_idle)(int); unsigned long boot_option_idle_override = 0; EXPORT_SYMBOL(boot_option_idle_override); +unsigned long idle_halt; +EXPORT_SYMBOL(idle_halt); void ia64_do_show_stack (struct unw_frame_info *info, void *arg) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 7dceea94723..7fc72949876 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -7,6 +7,10 @@ #include #include #include +#include + +unsigned long idle_halt; +EXPORT_SYMBOL(idle_halt); struct kmem_cache *task_xstate_cachep; @@ -325,7 +329,18 @@ static int __init idle_setup(char *str) pm_idle = poll_idle; } else if (!strcmp(str, "mwait")) force_mwait = 1; - else + else if (!strcmp(str, "halt")) { + /* + * When the boot option of idle=halt is added, halt is + * forced to be used for CPU idle. In such case CPU C2/C3 + * won't be used again. + * To continue to load the CPU idle driver, don't touch + * the boot_option_idle_override. + */ + pm_idle = default_idle; + idle_halt = 1; + return 0; + } else return -1; boot_option_idle_override = 1; diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 0fc310e7dfd..c75c7ace8c1 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -41,6 +41,7 @@ #include #include #include +#include /* * Include the apic definitions for x86 to have the APIC timer related defines @@ -57,6 +58,7 @@ #include #include +#include #define ACPI_PROCESSOR_COMPONENT 0x01000000 #define ACPI_PROCESSOR_CLASS "processor" @@ -955,6 +957,17 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) } else { continue; } + if (cx.type == ACPI_STATE_C1 && idle_halt) { + /* + * In most cases the C1 space_id obtained from + * _CST object is FIXED_HARDWARE access mode. + * But when the option of idle=halt is added, + * the entry_method type should be changed from + * CSTATE_FFH to CSTATE_HALT. + */ + cx.entry_method = ACPI_CSTATE_HALT; + snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); + } } else { snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI IOPORT 0x%x", cx.address); @@ -1780,6 +1793,15 @@ int __cpuinit acpi_processor_power_init(struct acpi_processor *pr, return 0; if (!first_run) { + if (idle_halt) { + /* + * When the boot option of "idle=halt" is added, halt + * is used for CPU IDLE. + * In such case C2/C3 is meaningless. So the max_cstate + * is set to one. + */ + max_cstate = 1; + } dmi_check_system(processor_power_dmi_table); max_cstate = acpi_processor_cstate_check(max_cstate); if (max_cstate < ACPI_C_STATES_MAX) diff --git a/include/asm-ia64/processor.h b/include/asm-ia64/processor.h index 6aff126fc07..f36e28a5f61 100644 --- a/include/asm-ia64/processor.h +++ b/include/asm-ia64/processor.h @@ -763,6 +763,7 @@ prefetchw (const void *x) #define spin_lock_prefetch(x) prefetchw(x) extern unsigned long boot_option_idle_override; +extern unsigned long idle_halt; #endif /* !__ASSEMBLY__ */ diff --git a/include/asm-x86/processor.h b/include/asm-x86/processor.h index 7f738270459..bc221623248 100644 --- a/include/asm-x86/processor.h +++ b/include/asm-x86/processor.h @@ -727,6 +727,7 @@ extern int force_mwait; extern void select_idle_routine(const struct cpuinfo_x86 *c); extern unsigned long boot_option_idle_override; +extern unsigned long idle_halt; extern void enable_sep_cpu(void); extern int sysenter_setup(void); -- cgit v1.2.3-70-g09d2 From da5e09a1b3e5a9fc0b15a3feb64e921ccc55ba74 Mon Sep 17 00:00:00 2001 From: Zhao Yakui Date: Tue, 24 Jun 2008 18:01:09 +0800 Subject: ACPI : Create "idle=nomwait" bootparam "idle=nomwait" disables the use of the MWAIT instruction from both C1 (C1_FFH) and deeper (C2C3_FFH) C-states. When MWAIT is unavailable, the BIOS and OS generally negotiate to use the HALT instruction for C1, and use IO accesses for deeper C-states. This option is useful for power and performance comparisons, and also to work around BIOS bugs where broken MWAIT support is advertised. http://bugzilla.kernel.org/show_bug.cgi?id=10807 http://bugzilla.kernel.org/show_bug.cgi?id=10914 Signed-off-by: Zhao Yakui Signed-off-by: Li Shaohua Signed-off-by: Len Brown Signed-off-by: Andi Kleen --- Documentation/kernel-parameters.txt | 3 ++- arch/ia64/kernel/process.c | 2 ++ arch/x86/kernel/process.c | 11 +++++++++++ drivers/acpi/processor_core.c | 13 +++++++++++++ drivers/acpi/processor_idle.c | 6 +++++- include/asm-ia64/processor.h | 1 + include/asm-x86/processor.h | 1 + 7 files changed, 35 insertions(+), 2 deletions(-) (limited to 'drivers/acpi/processor_idle.c') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 65db7f4711a..5e497d16fb5 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -818,7 +818,7 @@ and is between 256 and 4096 characters. It is defined in the file See Documentation/ide/ide.txt. idle= [X86] - Format: idle=poll or idle=mwait, idle=halt + Format: idle=poll or idle=mwait, idle=halt, idle=nomwait Poll forces a polling idle loop that can slightly improves the performance of waking up a idle CPU, but will use a lot of power and make the system run hot. Not recommended. @@ -828,6 +828,7 @@ and is between 256 and 4096 characters. It is defined in the file as idle=poll. idle=halt. Halt is forced to be used for CPU idle. In such case C2/C3 won't be used again. + idle=nomwait. Disable mwait for CPU C-states ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem Claim all unknown PCI IDE storage controllers. diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c index 612b3c4a060..3ab8373103e 100644 --- a/arch/ia64/kernel/process.c +++ b/arch/ia64/kernel/process.c @@ -57,6 +57,8 @@ unsigned long boot_option_idle_override = 0; EXPORT_SYMBOL(boot_option_idle_override); unsigned long idle_halt; EXPORT_SYMBOL(idle_halt); +unsigned long idle_nomwait; +EXPORT_SYMBOL(idle_nomwait); void ia64_do_show_stack (struct unw_frame_info *info, void *arg) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 7fc72949876..4d629c62f4f 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -11,6 +11,8 @@ unsigned long idle_halt; EXPORT_SYMBOL(idle_halt); +unsigned long idle_nomwait; +EXPORT_SYMBOL(idle_nomwait); struct kmem_cache *task_xstate_cachep; @@ -340,6 +342,15 @@ static int __init idle_setup(char *str) pm_idle = default_idle; idle_halt = 1; return 0; + } else if (!strcmp(str, "nomwait")) { + /* + * If the boot option of "idle=nomwait" is added, + * it means that mwait will be disabled for CPU C2/C3 + * states. In such case it won't touch the variable + * of boot_option_idle_override. + */ + idle_nomwait = 1; + return 0; } else return -1; diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c index 9a803f85ccf..4e1bb89fd6c 100644 --- a/drivers/acpi/processor_core.c +++ b/drivers/acpi/processor_core.c @@ -265,7 +265,20 @@ static int acpi_processor_set_pdc(struct acpi_processor *pr) if (!pdc_in) return status; + if (idle_nomwait) { + /* + * If mwait is disabled for CPU C-states, the C2C3_FFH access + * mode will be disabled in the parameter of _PDC object. + * Of course C1_FFH access mode will also be disabled. + */ + union acpi_object *obj; + u32 *buffer = NULL; + obj = pdc_in->pointer; + buffer = (u32 *)(obj->buffer.pointer); + buffer[2] &= ~(ACPI_PDC_C_C2C3_FFH | ACPI_PDC_C_C1_FFH); + + } status = acpi_evaluate_object(pr->handle, "_PDC", pdc_in, NULL); if (ACPI_FAILURE(status)) diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index c75c7ace8c1..d592dbb1d12 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -957,13 +957,17 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) } else { continue; } - if (cx.type == ACPI_STATE_C1 && idle_halt) { + if (cx.type == ACPI_STATE_C1 && + (idle_halt || idle_nomwait)) { /* * In most cases the C1 space_id obtained from * _CST object is FIXED_HARDWARE access mode. * But when the option of idle=halt is added, * the entry_method type should be changed from * CSTATE_FFH to CSTATE_HALT. + * When the option of idle=nomwait is added, + * the C1 entry_method type should be + * CSTATE_HALT. */ cx.entry_method = ACPI_CSTATE_HALT; snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT"); diff --git a/include/asm-ia64/processor.h b/include/asm-ia64/processor.h index f36e28a5f61..f88fa054d01 100644 --- a/include/asm-ia64/processor.h +++ b/include/asm-ia64/processor.h @@ -764,6 +764,7 @@ prefetchw (const void *x) extern unsigned long boot_option_idle_override; extern unsigned long idle_halt; +extern unsigned long idle_nomwait; #endif /* !__ASSEMBLY__ */ diff --git a/include/asm-x86/processor.h b/include/asm-x86/processor.h index bc221623248..55402d2ab93 100644 --- a/include/asm-x86/processor.h +++ b/include/asm-x86/processor.h @@ -728,6 +728,7 @@ extern void select_idle_routine(const struct cpuinfo_x86 *c); extern unsigned long boot_option_idle_override; extern unsigned long idle_halt; +extern unsigned long idle_nomwait; extern void enable_sep_cpu(void); extern int sysenter_setup(void); -- cgit v1.2.3-70-g09d2