X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=drivers%2Facpi%2Fprocessor_idle.c;h=80fa43471f484dd68c190d8726086a2e03cc00b3;hb=43bc926fffd92024b46cafaf7350d669ba9ca884;hp=ff64d333e95f58c68ac45f83832e41060ecb9d1c;hpb=cee37fe97739d85991964371c1f3a745c00dd236;p=linux-2.6.git diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index ff64d333e..80fa43471 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -6,6 +6,8 @@ * Copyright (C) 2004 Dominik Brodowski * Copyright (C) 2004 Anil S Keshavamurthy * - Added processor hotplug support + * Copyright (C) 2005 Venkatesh Pallipadi + * - Added support for C3 on SMP * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * @@ -35,6 +37,7 @@ #include #include #include +#include /* need_resched() */ #include #include @@ -46,15 +49,12 @@ #define ACPI_PROCESSOR_CLASS "processor" #define ACPI_PROCESSOR_DRIVER_NAME "ACPI Processor Driver" #define _COMPONENT ACPI_PROCESSOR_COMPONENT -ACPI_MODULE_NAME ("acpi_processor") - +ACPI_MODULE_NAME("acpi_processor") #define ACPI_PROCESSOR_FILE_POWER "power" - #define US_TO_PM_TIMER_TICKS(t) ((t * (PM_TIMER_FREQUENCY/1000)) / 1000) #define C2_OVERHEAD 4 /* 1us (3.579 ticks per us) */ #define C3_OVERHEAD 4 /* 1us (3.579 ticks per us) */ - -static void (*pm_idle_save)(void); +static void (*pm_idle_save) (void); module_param(max_cstate, uint, 0644); static unsigned int nocst = 0; @@ -67,7 +67,8 @@ module_param(nocst, uint, 0000); * 100 HZ: 0x0000000F: 4 jiffies = 40ms * reduce history for more aggressive entry into C3 */ -static unsigned int bm_history = (HZ >= 800 ? 0xFFFFFFFF : ((1U << (HZ / 25)) - 1)); +static unsigned int bm_history = + (HZ >= 800 ? 0xFFFFFFFF : ((1U << (HZ / 25)) - 1)); module_param(bm_history, uint, 0644); /* -------------------------------------------------------------------------- Power Management @@ -79,38 +80,79 @@ module_param(bm_history, uint, 0644); * * To skip this limit, boot/load with a large max_cstate limit. */ -static int no_c2c3(struct dmi_system_id *id) +static int set_max_cstate(struct dmi_system_id *id) { if (max_cstate > ACPI_PROCESSOR_MAX_POWER) return 0; - printk(KERN_NOTICE PREFIX "%s detected - C2,C3 disabled." - " Override with \"processor.max_cstate=%d\"\n", id->ident, - ACPI_PROCESSOR_MAX_POWER + 1); + printk(KERN_NOTICE PREFIX "%s detected - limiting to C%ld max_cstate." + " Override with \"processor.max_cstate=%d\"\n", id->ident, + (long)id->driver_data, ACPI_PROCESSOR_MAX_POWER + 1); - max_cstate = 1; + max_cstate = (long)id->driver_data; return 0; } - - - -static struct dmi_system_id __initdata processor_power_dmi_table[] = { - { no_c2c3, "IBM ThinkPad R40e", { +/* Actually this shouldn't be __cpuinitdata, would be better to fix the + callers to only run once -AK */ +static struct dmi_system_id __cpuinitdata processor_power_dmi_table[] = { + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET60WW")}, (void *)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET43WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET45WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET47WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET50WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET52WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET55WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET56WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET59WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET60WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET61WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), - DMI_MATCH(DMI_BIOS_VERSION,"1SET60WW") }}, - { no_c2c3, "Medion 41700", { + DMI_MATCH(DMI_BIOS_VERSION,"1SET62WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET64WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET65WW") }, (void*)1}, + { set_max_cstate, "IBM ThinkPad R40e", { + DMI_MATCH(DMI_BIOS_VENDOR,"IBM"), + DMI_MATCH(DMI_BIOS_VERSION,"1SET68WW") }, (void*)1}, + { set_max_cstate, "Medion 41700", { DMI_MATCH(DMI_BIOS_VENDOR,"Phoenix Technologies LTD"), - DMI_MATCH(DMI_BIOS_VERSION,"R01-A1J") }}, + DMI_MATCH(DMI_BIOS_VERSION,"R01-A1J")}, (void *)1}, + { set_max_cstate, "Clevo 5600D", { + DMI_MATCH(DMI_BIOS_VENDOR,"Phoenix Technologies LTD"), + DMI_MATCH(DMI_BIOS_VERSION,"SHE845M0.86C.0013.D.0302131307")}, + (void *)2}, {}, }; - -static inline u32 -ticks_elapsed ( - u32 t1, - u32 t2) +static inline u32 ticks_elapsed(u32 t1, u32 t2) { if (t2 >= t1) return (t2 - t1); @@ -120,13 +162,11 @@ ticks_elapsed ( return ((0xFFFFFFFF - t1) + t2); } - static void -acpi_processor_power_activate ( - struct acpi_processor *pr, - struct acpi_processor_cx *new) +acpi_processor_power_activate(struct acpi_processor *pr, + struct acpi_processor_cx *new) { - struct acpi_processor_cx *old; + struct acpi_processor_cx *old; if (!pr || !new) return; @@ -135,15 +175,16 @@ acpi_processor_power_activate ( if (old) old->promotion.count = 0; - new->demotion.count = 0; + new->demotion.count = 0; /* Cleanup from old state. */ if (old) { switch (old->type) { case ACPI_STATE_C3: /* Disable bus master reload */ - if (new->type != ACPI_STATE_C3) - acpi_set_register(ACPI_BITREG_BUS_MASTER_RLD, 0, ACPI_MTX_DO_NOT_LOCK); + if (new->type != ACPI_STATE_C3 && pr->flags.bm_check) + acpi_set_register(ACPI_BITREG_BUS_MASTER_RLD, 0, + ACPI_MTX_DO_NOT_LOCK); break; } } @@ -152,8 +193,9 @@ acpi_processor_power_activate ( switch (new->type) { case ACPI_STATE_C3: /* Enable bus master reload */ - if (old->type != ACPI_STATE_C3) - acpi_set_register(ACPI_BITREG_BUS_MASTER_RLD, 1, ACPI_MTX_DO_NOT_LOCK); + if (old->type != ACPI_STATE_C3 && pr->flags.bm_check) + acpi_set_register(ACPI_BITREG_BUS_MASTER_RLD, 1, + ACPI_MTX_DO_NOT_LOCK); break; } @@ -162,16 +204,26 @@ acpi_processor_power_activate ( return; } +static void acpi_safe_halt(void) +{ + clear_thread_flag(TIF_POLLING_NRFLAG); + smp_mb__after_clear_bit(); + if (!need_resched()) + safe_halt(); + set_thread_flag(TIF_POLLING_NRFLAG); +} -static void acpi_processor_idle (void) +static atomic_t c3_cpu_count; + +static void acpi_processor_idle(void) { - struct acpi_processor *pr = NULL; + struct acpi_processor *pr = NULL; struct acpi_processor_cx *cx = NULL; struct acpi_processor_cx *next_state = NULL; - int sleep_ticks = 0; - u32 t1, t2 = 0; + int sleep_ticks = 0; + u32 t1, t2 = 0; - pr = processors[_smp_processor_id()]; + pr = processors[smp_processor_id()]; if (!pr) return; @@ -191,8 +243,13 @@ static void acpi_processor_idle (void) } cx = pr->power.state; - if (!cx) - goto easy_out; + if (!cx) { + if (pm_idle_save) + pm_idle_save(); + else + acpi_safe_halt(); + return; + } /* * Check BM Activity @@ -201,8 +258,8 @@ static void acpi_processor_idle (void) * for demotion. */ if (pr->flags.bm_check) { - u32 bm_status = 0; - unsigned long diff = jiffies - pr->power.bm_check_timestamp; + u32 bm_status = 0; + unsigned long diff = jiffies - pr->power.bm_check_timestamp; if (diff > 32) diff = 32; @@ -216,11 +273,11 @@ static void acpi_processor_idle (void) } acpi_get_register(ACPI_BITREG_BUS_MASTER_STATUS, - &bm_status, ACPI_MTX_DO_NOT_LOCK); + &bm_status, ACPI_MTX_DO_NOT_LOCK); if (bm_status) { pr->power.bm_activity++; acpi_set_register(ACPI_BITREG_BUS_MASTER_STATUS, - 1, ACPI_MTX_DO_NOT_LOCK); + 1, ACPI_MTX_DO_NOT_LOCK); } /* * PIIX4 Erratum #18: Note that BM_STS doesn't always reflect @@ -229,7 +286,7 @@ static void acpi_processor_idle (void) */ else if (errata.piix4.bmisx) { if ((inb_p(errata.piix4.bmisx + 0x02) & 0x01) - || (inb_p(errata.piix4.bmisx + 0x0A) & 0x01)) + || (inb_p(errata.piix4.bmisx + 0x0A) & 0x01)) pr->power.bm_activity++; } @@ -254,6 +311,17 @@ static void acpi_processor_idle (void) } } +#ifdef CONFIG_HOTPLUG_CPU + /* + * Check for P_LVL2_UP flag before entering C2 and above on + * an SMP system. We do it here instead of doing it at _CST/P_LVL + * detection phase, to work cleanly with logical CPU hotplug. + */ + if ((cx->type != ACPI_STATE_C1) && (num_online_cpus() > 1) && + !pr->flags.has_cst && !acpi_fadt.plvl2_up) + cx = &pr->power.states[ACPI_STATE_C1]; +#endif + cx->usage++; /* @@ -261,6 +329,16 @@ static void acpi_processor_idle (void) * ------ * Invoke the current Cx state to put the processor to sleep. */ + if (cx->type == ACPI_STATE_C2 || cx->type == ACPI_STATE_C3) { + clear_thread_flag(TIF_POLLING_NRFLAG); + smp_mb__after_clear_bit(); + if (need_resched()) { + set_thread_flag(TIF_POLLING_NRFLAG); + local_irq_enable(); + return; + } + } + switch (cx->type) { case ACPI_STATE_C1: @@ -272,9 +350,10 @@ static void acpi_processor_idle (void) if (pm_idle_save) pm_idle_save(); else - safe_halt(); + acpi_safe_halt(); + /* - * TBD: Can't get time duration while in C1, as resumes + * TBD: Can't get time duration while in C1, as resumes * go to an ISR rather than here. Need to instrument * base interrupt handler. */ @@ -292,13 +371,29 @@ static void acpi_processor_idle (void) t2 = inl(acpi_fadt.xpm_tmr_blk.address); /* Re-enable interrupts */ local_irq_enable(); + set_thread_flag(TIF_POLLING_NRFLAG); /* Compute time (ticks) that we were actually asleep */ - sleep_ticks = ticks_elapsed(t1, t2) - cx->latency_ticks - C2_OVERHEAD; + sleep_ticks = + ticks_elapsed(t1, t2) - cx->latency_ticks - C2_OVERHEAD; break; case ACPI_STATE_C3: - /* Disable bus master arbitration */ - acpi_set_register(ACPI_BITREG_ARB_DISABLE, 1, ACPI_MTX_DO_NOT_LOCK); + + if (pr->flags.bm_check) { + if (atomic_inc_return(&c3_cpu_count) == + num_online_cpus()) { + /* + * All CPUs are trying to go to C3 + * Disable bus master arbitration + */ + acpi_set_register(ACPI_BITREG_ARB_DISABLE, 1, + ACPI_MTX_DO_NOT_LOCK); + } + } else { + /* SMP with no shared cache... Invalidate cache */ + ACPI_FLUSH_CPU_CACHE(); + } + /* Get start time (ticks) */ t1 = inl(acpi_fadt.xpm_tmr_blk.address); /* Invoke C3 */ @@ -307,12 +402,19 @@ static void acpi_processor_idle (void) t2 = inl(acpi_fadt.xpm_tmr_blk.address); /* Get end time (ticks) */ t2 = inl(acpi_fadt.xpm_tmr_blk.address); - /* Enable bus master arbitration */ - acpi_set_register(ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_DO_NOT_LOCK); + if (pr->flags.bm_check) { + /* Enable bus master arbitration */ + atomic_dec(&c3_cpu_count); + acpi_set_register(ACPI_BITREG_ARB_DISABLE, 0, + ACPI_MTX_DO_NOT_LOCK); + } + /* Re-enable interrupts */ local_irq_enable(); + set_thread_flag(TIF_POLLING_NRFLAG); /* Compute time (ticks) that we were actually asleep */ - sleep_ticks = ticks_elapsed(t1, t2) - cx->latency_ticks - C3_OVERHEAD; + sleep_ticks = + ticks_elapsed(t1, t2) - cx->latency_ticks - C3_OVERHEAD; break; default: @@ -322,6 +424,15 @@ static void acpi_processor_idle (void) next_state = pr->power.state; +#ifdef CONFIG_HOTPLUG_CPU + /* Don't do promotion/demotion */ + if ((cx->type == ACPI_STATE_C1) && (num_online_cpus() > 1) && + !pr->flags.has_cst && !acpi_fadt.plvl2_up) { + next_state = cx; + goto end; + } +#endif + /* * Promotion? * ---------- @@ -334,15 +445,18 @@ static void acpi_processor_idle (void) ((cx->promotion.state - pr->power.states) <= max_cstate)) { if (sleep_ticks > cx->promotion.threshold.ticks) { cx->promotion.count++; - cx->demotion.count = 0; - if (cx->promotion.count >= cx->promotion.threshold.count) { + cx->demotion.count = 0; + if (cx->promotion.count >= + cx->promotion.threshold.count) { if (pr->flags.bm_check) { - if (!(pr->power.bm_activity & cx->promotion.threshold.bm)) { - next_state = cx->promotion.state; + if (! + (pr->power.bm_activity & cx-> + promotion.threshold.bm)) { + next_state = + cx->promotion.state; goto end; } - } - else { + } else { next_state = cx->promotion.state; goto end; } @@ -367,7 +481,7 @@ static void acpi_processor_idle (void) } } -end: + end: /* * Demote if current state exceeds max_cstate */ @@ -384,22 +498,9 @@ end: */ if (next_state != pr->power.state) acpi_processor_power_activate(pr, next_state); - - return; - - easy_out: - /* do C1 instead of busy loop */ - if (pm_idle_save) - pm_idle_save(); - else - safe_halt(); - return; } - -static int -acpi_processor_set_power_policy ( - struct acpi_processor *pr) +static int acpi_processor_set_power_policy(struct acpi_processor *pr) { unsigned int i; unsigned int state_is_set = 0; @@ -407,7 +508,7 @@ acpi_processor_set_power_policy ( struct acpi_processor_cx *higher = NULL; struct acpi_processor_cx *cx; - ACPI_FUNCTION_TRACE("acpi_processor_set_power_policy"); + ACPI_FUNCTION_TRACE("acpi_processor_set_power_policy"); if (!pr) return_VALUE(-EINVAL); @@ -422,7 +523,7 @@ acpi_processor_set_power_policy ( */ /* startup state */ - for (i=1; i < ACPI_PROCESSOR_MAX_POWER; i++) { + for (i = 1; i < ACPI_PROCESSOR_MAX_POWER; i++) { cx = &pr->power.states[i]; if (!cx->valid) continue; @@ -431,13 +532,13 @@ acpi_processor_set_power_policy ( pr->power.state = cx; state_is_set++; break; - } + } if (!state_is_set) return_VALUE(-ENODEV); /* demotion */ - for (i=1; i < ACPI_PROCESSOR_MAX_POWER; i++) { + for (i = 1; i < ACPI_PROCESSOR_MAX_POWER; i++) { cx = &pr->power.states[i]; if (!cx->valid) continue; @@ -460,7 +561,7 @@ acpi_processor_set_power_policy ( continue; if (higher) { - cx->promotion.state = higher; + cx->promotion.state = higher; cx->promotion.threshold.ticks = cx->latency_ticks; if (cx->type >= ACPI_STATE_C2) cx->promotion.threshold.count = 4; @@ -473,14 +574,11 @@ acpi_processor_set_power_policy ( higher = cx; } - return_VALUE(0); + return_VALUE(0); } - -static int acpi_processor_get_power_info_fadt (struct acpi_processor *pr) +static int acpi_processor_get_power_info_fadt(struct acpi_processor *pr) { - int i; - ACPI_FUNCTION_TRACE("acpi_processor_get_power_info_fadt"); if (!pr) @@ -489,18 +587,18 @@ static int acpi_processor_get_power_info_fadt (struct acpi_processor *pr) if (!pr->pblk) return_VALUE(-ENODEV); - for (i = 0; i < ACPI_PROCESSOR_MAX_POWER; i++) - memset(pr->power.states, 0, sizeof(struct acpi_processor_cx)); - /* if info is obtained from pblk/fadt, type equals state */ - pr->power.states[ACPI_STATE_C1].type = ACPI_STATE_C1; pr->power.states[ACPI_STATE_C2].type = ACPI_STATE_C2; pr->power.states[ACPI_STATE_C3].type = ACPI_STATE_C3; - /* the C0 state only exists as a filler in our array, - * and all processors need to support C1 */ - pr->power.states[ACPI_STATE_C0].valid = 1; - pr->power.states[ACPI_STATE_C1].valid = 1; +#ifndef CONFIG_HOTPLUG_CPU + /* + * Check for P_LVL2_UP flag before entering C2 and above on + * an SMP system. + */ + if ((num_online_cpus() > 1) && !acpi_fadt.plvl2_up) + return_VALUE(-ENODEV); +#endif /* determine C2 and C3 address from pblk */ pr->power.states[ACPI_STATE_C2].address = pr->pblk + 4; @@ -518,38 +616,57 @@ static int acpi_processor_get_power_info_fadt (struct acpi_processor *pr) return_VALUE(0); } +static int acpi_processor_get_power_info_default_c1(struct acpi_processor *pr) +{ + ACPI_FUNCTION_TRACE("acpi_processor_get_power_info_default_c1"); + + /* Zero initialize all the C-states info. */ + memset(pr->power.states, 0, sizeof(pr->power.states)); + + /* set the first C-State to C1 */ + pr->power.states[ACPI_STATE_C1].type = ACPI_STATE_C1; + + /* the C0 state only exists as a filler in our array, + * and all processors need to support C1 */ + pr->power.states[ACPI_STATE_C0].valid = 1; + pr->power.states[ACPI_STATE_C1].valid = 1; + + return_VALUE(0); +} -static int acpi_processor_get_power_info_cst (struct acpi_processor *pr) +static int acpi_processor_get_power_info_cst(struct acpi_processor *pr) { - acpi_status status = 0; - acpi_integer count; - int i; - struct acpi_buffer buffer = {ACPI_ALLOCATE_BUFFER, NULL}; - union acpi_object *cst; + acpi_status status = 0; + acpi_integer count; + int current_count; + int i; + struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; + union acpi_object *cst; ACPI_FUNCTION_TRACE("acpi_processor_get_power_info_cst"); - if (errata.smp) - return_VALUE(-ENODEV); - if (nocst) return_VALUE(-ENODEV); - pr->power.count = 0; - for (i = 0; i < ACPI_PROCESSOR_MAX_POWER; i++) - memset(pr->power.states, 0, sizeof(struct acpi_processor_cx)); + current_count = 1; + + /* Zero initialize C2 onwards and prepare for fresh CST lookup */ + for (i = 2; i < ACPI_PROCESSOR_MAX_POWER; i++) + memset(&(pr->power.states[i]), 0, + sizeof(struct acpi_processor_cx)); status = acpi_evaluate_object(pr->handle, "_CST", NULL, &buffer); if (ACPI_FAILURE(status)) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, "No _CST, giving up\n")); return_VALUE(-ENODEV); - } + } - cst = (union acpi_object *) buffer.pointer; + cst = (union acpi_object *)buffer.pointer; /* There must be at least 2 elements */ if (!cst || (cst->type != ACPI_TYPE_PACKAGE) || cst->package.count < 2) { - ACPI_DEBUG_PRINT((ACPI_DB_ERROR, "not enough elements in _CST\n")); + ACPI_DEBUG_PRINT((ACPI_DB_ERROR, + "not enough elements in _CST\n")); status = -EFAULT; goto end; } @@ -558,18 +675,12 @@ static int acpi_processor_get_power_info_cst (struct acpi_processor *pr) /* Validate number of power states. */ if (count < 1 || count != cst->package.count - 1) { - ACPI_DEBUG_PRINT((ACPI_DB_ERROR, "count given by _CST is not valid\n")); + ACPI_DEBUG_PRINT((ACPI_DB_ERROR, + "count given by _CST is not valid\n")); status = -EFAULT; goto end; } - /* We support up to ACPI_PROCESSOR_MAX_POWER. */ - if (count > ACPI_PROCESSOR_MAX_POWER) { - printk(KERN_WARNING "Limiting number of power states to max (%d)\n", ACPI_PROCESSOR_MAX_POWER); - printk(KERN_WARNING "Please increase ACPI_PROCESSOR_MAX_POWER if needed.\n"); - count = ACPI_PROCESSOR_MAX_POWER; - } - /* Tell driver that at least _CST is supported. */ pr->flags.has_cst = 1; @@ -581,29 +692,29 @@ static int acpi_processor_get_power_info_cst (struct acpi_processor *pr) memset(&cx, 0, sizeof(cx)); - element = (union acpi_object *) &(cst->package.elements[i]); + element = (union acpi_object *)&(cst->package.elements[i]); if (element->type != ACPI_TYPE_PACKAGE) continue; if (element->package.count != 4) continue; - obj = (union acpi_object *) &(element->package.elements[0]); + obj = (union acpi_object *)&(element->package.elements[0]); if (obj->type != ACPI_TYPE_BUFFER) continue; - reg = (struct acpi_power_register *) obj->buffer.pointer; + reg = (struct acpi_power_register *)obj->buffer.pointer; if (reg->space_id != ACPI_ADR_SPACE_SYSTEM_IO && - (reg->space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) + (reg->space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) continue; cx.address = (reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) ? - 0 : reg->address; + 0 : reg->address; /* There should be an easy way to extract an integer... */ - obj = (union acpi_object *) &(element->package.elements[1]); + obj = (union acpi_object *)&(element->package.elements[1]); if (obj->type != ACPI_TYPE_INTEGER) continue; @@ -613,39 +724,51 @@ static int acpi_processor_get_power_info_cst (struct acpi_processor *pr) (reg->space_id != ACPI_ADR_SPACE_SYSTEM_IO)) continue; - if ((cx.type < ACPI_STATE_C1) || - (cx.type > ACPI_STATE_C3)) + if ((cx.type < ACPI_STATE_C2) || (cx.type > ACPI_STATE_C3)) continue; - obj = (union acpi_object *) &(element->package.elements[2]); + obj = (union acpi_object *)&(element->package.elements[2]); if (obj->type != ACPI_TYPE_INTEGER) continue; cx.latency = obj->integer.value; - obj = (union acpi_object *) &(element->package.elements[3]); + obj = (union acpi_object *)&(element->package.elements[3]); if (obj->type != ACPI_TYPE_INTEGER) continue; cx.power = obj->integer.value; - (pr->power.count)++; - memcpy(&(pr->power.states[pr->power.count]), &cx, sizeof(cx)); + current_count++; + memcpy(&(pr->power.states[current_count]), &cx, sizeof(cx)); + + /* + * We support total ACPI_PROCESSOR_MAX_POWER - 1 + * (From 1 through ACPI_PROCESSOR_MAX_POWER - 1) + */ + if (current_count >= (ACPI_PROCESSOR_MAX_POWER - 1)) { + printk(KERN_WARNING + "Limiting number of power states to max (%d)\n", + ACPI_PROCESSOR_MAX_POWER); + printk(KERN_WARNING + "Please increase ACPI_PROCESSOR_MAX_POWER if needed.\n"); + break; + } } - ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Found %d power states\n", pr->power.count)); + ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Found %d power states\n", + current_count)); /* Validate number of power states discovered */ - if (pr->power.count < 2) - status = -ENODEV; + if (current_count < 2) + status = -EFAULT; -end: + end: acpi_os_free(buffer.pointer); return_VALUE(status); } - static void acpi_processor_power_verify_c2(struct acpi_processor_cx *cx) { ACPI_FUNCTION_TRACE("acpi_processor_get_power_verify_c2"); @@ -659,15 +782,7 @@ static void acpi_processor_power_verify_c2(struct acpi_processor_cx *cx) */ else if (cx->latency > ACPI_PROCESSOR_MAX_C2_LATENCY) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "latency too large [%d]\n", - cx->latency)); - return_VOID; - } - - /* We're (currently) only supporting C2 on UP */ - else if (errata.smp) { - ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "C2 not supported in SMP mode\n")); + "latency too large [%d]\n", cx->latency)); return_VOID; } @@ -681,11 +796,11 @@ static void acpi_processor_power_verify_c2(struct acpi_processor_cx *cx) return_VOID; } - -static void acpi_processor_power_verify_c3( - struct acpi_processor *pr, - struct acpi_processor_cx *cx) +static void acpi_processor_power_verify_c3(struct acpi_processor *pr, + struct acpi_processor_cx *cx) { + static int bm_check_flag; + ACPI_FUNCTION_TRACE("acpi_processor_get_power_verify_c3"); if (!cx->address) @@ -697,22 +812,7 @@ static void acpi_processor_power_verify_c3( */ else if (cx->latency > ACPI_PROCESSOR_MAX_C3_LATENCY) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "latency too large [%d]\n", - cx->latency)); - return_VOID; - } - - /* bus mastering control is necessary */ - else if (!pr->flags.bm_control) { - ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "C3 support requires bus mastering control\n")); - return_VOID; - } - - /* We're (currently) only supporting C2 on UP */ - else if (errata.smp) { - ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "C3 not supported in SMP mode\n")); + "latency too large [%d]\n", cx->latency)); return_VOID; } @@ -725,10 +825,41 @@ static void acpi_processor_power_verify_c3( */ else if (errata.piix4.fdma) { ACPI_DEBUG_PRINT((ACPI_DB_INFO, - "C3 not supported on PIIX4 with Type-F DMA\n")); + "C3 not supported on PIIX4 with Type-F DMA\n")); return_VOID; } + /* All the logic here assumes flags.bm_check is same across all CPUs */ + if (!bm_check_flag) { + /* Determine whether bm_check is needed based on CPU */ + acpi_processor_power_init_bm_check(&(pr->flags), pr->id); + bm_check_flag = pr->flags.bm_check; + } else { + pr->flags.bm_check = bm_check_flag; + } + + if (pr->flags.bm_check) { + /* bus mastering control is necessary */ + if (!pr->flags.bm_control) { + ACPI_DEBUG_PRINT((ACPI_DB_INFO, + "C3 support requires bus mastering control\n")); + return_VOID; + } + } else { + /* + * WBINVD should be set in fadt, for C3 state to be + * supported on when bm_check is not required. + */ + if (acpi_fadt.wb_invd != 1) { + ACPI_DEBUG_PRINT((ACPI_DB_INFO, + "Cache invalidation should work properly" + " for C3 to be enabled on SMP systems\n")); + return_VOID; + } + acpi_set_register(ACPI_BITREG_BUS_MASTER_RLD, + 0, ACPI_MTX_DO_NOT_LOCK); + } + /* * Otherwise we've met all of our C3 requirements. * Normalize the C3 latency to expidite policy. Enable @@ -737,18 +868,22 @@ static void acpi_processor_power_verify_c3( */ cx->valid = 1; cx->latency_ticks = US_TO_PM_TIMER_TICKS(cx->latency); - pr->flags.bm_check = 1; return_VOID; } - static int acpi_processor_power_verify(struct acpi_processor *pr) { unsigned int i; unsigned int working = 0; - for (i=1; i < ACPI_PROCESSOR_MAX_POWER; i++) { +#ifdef ARCH_APICTIMER_STOPS_ON_C3 + int timer_broadcast = 0; + cpumask_t mask = cpumask_of_cpu(pr->id); + on_each_cpu(switch_ipi_to_APIC_timer, &mask, 1, 1); +#endif + + for (i = 1; i < ACPI_PROCESSOR_MAX_POWER; i++) { struct acpi_processor_cx *cx = &pr->power.states[i]; switch (cx->type) { @@ -758,10 +893,21 @@ static int acpi_processor_power_verify(struct acpi_processor *pr) case ACPI_STATE_C2: acpi_processor_power_verify_c2(cx); +#ifdef ARCH_APICTIMER_STOPS_ON_C3 + /* Some AMD systems fake C3 as C2, but still + have timer troubles */ + if (cx->valid && + boot_cpu_data.x86_vendor == X86_VENDOR_AMD) + timer_broadcast++; +#endif break; case ACPI_STATE_C3: acpi_processor_power_verify_c3(pr, cx); +#ifdef ARCH_APICTIMER_STOPS_ON_C3 + if (cx->valid) + timer_broadcast++; +#endif break; } @@ -769,11 +915,15 @@ static int acpi_processor_power_verify(struct acpi_processor *pr) working++; } +#ifdef ARCH_APICTIMER_STOPS_ON_C3 + if (timer_broadcast) + on_each_cpu(switch_APIC_timer_to_ipi, &mask, 1, 1); +#endif + return (working); } -static int acpi_processor_get_power_info ( - struct acpi_processor *pr) +static int acpi_processor_get_power_info(struct acpi_processor *pr) { unsigned int i; int result; @@ -783,15 +933,13 @@ static int acpi_processor_get_power_info ( /* NOTE: the idle thread may not be running while calling * this function */ + /* Adding C1 state */ + acpi_processor_get_power_info_default_c1(pr); result = acpi_processor_get_power_info_cst(pr); - if ((result) || (acpi_processor_power_verify(pr) < 2)) { - result = acpi_processor_get_power_info_fadt(pr); - if (result) - return_VALUE(result); + if (result == -ENODEV) + acpi_processor_get_power_info_fadt(pr); - if (acpi_processor_power_verify(pr) < 2) - return_VALUE(-ENODEV); - } + pr->power.count = acpi_processor_power_verify(pr); /* * Set Default Policy @@ -810,26 +958,26 @@ static int acpi_processor_get_power_info ( * CPU as being "idle manageable" */ for (i = 1; i < ACPI_PROCESSOR_MAX_POWER; i++) { - if (pr->power.states[i].valid) + if (pr->power.states[i].valid) { pr->power.count = i; - if ((pr->power.states[i].valid) && - (pr->power.states[i].type >= ACPI_STATE_C2)) - pr->flags.power = 1; + if (pr->power.states[i].type >= ACPI_STATE_C2) + pr->flags.power = 1; + } } return_VALUE(0); } -int acpi_processor_cst_has_changed (struct acpi_processor *pr) +int acpi_processor_cst_has_changed(struct acpi_processor *pr) { - int result = 0; + int result = 0; ACPI_FUNCTION_TRACE("acpi_processor_cst_has_changed"); if (!pr) - return_VALUE(-EINVAL); + return_VALUE(-EINVAL); - if (errata.smp || nocst) { + if (nocst) { return_VALUE(-ENODEV); } @@ -838,7 +986,7 @@ int acpi_processor_cst_has_changed (struct acpi_processor *pr) /* Fall back to the default idle loop */ pm_idle = pm_idle_save; - synchronize_sched(); /* Relies on interrupts forcing exit from idle. */ + synchronize_sched(); /* Relies on interrupts forcing exit from idle. */ pr->flags.power = 0; result = acpi_processor_get_power_info(pr); @@ -852,8 +1000,8 @@ int acpi_processor_cst_has_changed (struct acpi_processor *pr) static int acpi_processor_power_seq_show(struct seq_file *seq, void *offset) { - struct acpi_processor *pr = (struct acpi_processor *)seq->private; - unsigned int i; + struct acpi_processor *pr = (struct acpi_processor *)seq->private; + unsigned int i; ACPI_FUNCTION_TRACE("acpi_processor_power_seq_show"); @@ -861,17 +1009,17 @@ static int acpi_processor_power_seq_show(struct seq_file *seq, void *offset) goto end; seq_printf(seq, "active state: C%zd\n" - "max_cstate: C%d\n" - "bus master activity: %08x\n", - pr->power.state ? pr->power.state - pr->power.states : 0, - max_cstate, - (unsigned)pr->power.bm_activity); + "max_cstate: C%d\n" + "bus master activity: %08x\n", + pr->power.state ? pr->power.state - pr->power.states : 0, + max_cstate, (unsigned)pr->power.bm_activity); seq_puts(seq, "states:\n"); for (i = 1; i <= pr->power.count; i++) { seq_printf(seq, " %cC%d: ", - (&pr->power.states[i] == pr->power.state?'*':' '), i); + (&pr->power.states[i] == + pr->power.state ? '*' : ' '), i); if (!pr->power.states[i].valid) { seq_puts(seq, "\n"); @@ -895,46 +1043,46 @@ static int acpi_processor_power_seq_show(struct seq_file *seq, void *offset) if (pr->power.states[i].promotion.state) seq_printf(seq, "promotion[C%zd] ", - (pr->power.states[i].promotion.state - - pr->power.states)); + (pr->power.states[i].promotion.state - + pr->power.states)); else seq_puts(seq, "promotion[--] "); if (pr->power.states[i].demotion.state) seq_printf(seq, "demotion[C%zd] ", - (pr->power.states[i].demotion.state - - pr->power.states)); + (pr->power.states[i].demotion.state - + pr->power.states)); else seq_puts(seq, "demotion[--] "); seq_printf(seq, "latency[%03d] usage[%08d]\n", - pr->power.states[i].latency, - pr->power.states[i].usage); + pr->power.states[i].latency, + pr->power.states[i].usage); } -end: + end: return_VALUE(0); } static int acpi_processor_power_open_fs(struct inode *inode, struct file *file) { return single_open(file, acpi_processor_power_seq_show, - PDE(inode)->data); + PDE(inode)->data); } static struct file_operations acpi_processor_power_fops = { - .open = acpi_processor_power_open_fs, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, + .open = acpi_processor_power_open_fs, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, }; - -int acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *device) +int acpi_processor_power_init(struct acpi_processor *pr, + struct acpi_device *device) { - acpi_status status = 0; - static int first_run = 0; - struct proc_dir_entry *entry = NULL; + acpi_status status = 0; + static int first_run = 0; + struct proc_dir_entry *entry = NULL; unsigned int i; ACPI_FUNCTION_TRACE("acpi_processor_power_init"); @@ -942,12 +1090,18 @@ int acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *dev if (!first_run) { dmi_check_system(processor_power_dmi_table); if (max_cstate < ACPI_C_STATES_MAX) - printk(KERN_NOTICE "ACPI: processor limited to max C-state %d\n", max_cstate); + printk(KERN_NOTICE + "ACPI: processor limited to max C-state %d\n", + max_cstate); first_run++; } - if (!errata.smp && (pr->id == 0) && acpi_fadt.cst_cnt && !nocst) { - status = acpi_os_write_port(acpi_fadt.smi_cmd, acpi_fadt.cst_cnt, 8); + if (!pr) + return_VALUE(-EINVAL); + + if (acpi_fadt.cst_cnt && !nocst) { + status = + acpi_os_write_port(acpi_fadt.smi_cmd, acpi_fadt.cst_cnt, 8); if (ACPI_FAILURE(status)) { ACPI_DEBUG_PRINT((ACPI_DB_ERROR, "Notifying BIOS of _CST ability failed\n")); @@ -965,7 +1119,8 @@ int acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *dev printk(KERN_INFO PREFIX "CPU%d (power states:", pr->id); for (i = 1; i <= pr->power.count; i++) if (pr->power.states[i].valid) - printk(" C%d[C%d]", i, pr->power.states[i].type); + printk(" C%d[C%d]", i, + pr->power.states[i].type); printk(")\n"); if (pr->id == 0) { @@ -976,11 +1131,11 @@ int acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *dev /* 'power' [R] */ entry = create_proc_entry(ACPI_PROCESSOR_FILE_POWER, - S_IRUGO, acpi_device_dir(device)); + S_IRUGO, acpi_device_dir(device)); if (!entry) ACPI_DEBUG_PRINT((ACPI_DB_ERROR, - "Unable to create '%s' fs entry\n", - ACPI_PROCESSOR_FILE_POWER)); + "Unable to create '%s' fs entry\n", + ACPI_PROCESSOR_FILE_POWER)); else { entry->proc_fops = &acpi_processor_power_fops; entry->data = acpi_driver_data(device); @@ -992,14 +1147,16 @@ int acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *dev return_VALUE(0); } -int acpi_processor_power_exit(struct acpi_processor *pr, struct acpi_device *device) +int acpi_processor_power_exit(struct acpi_processor *pr, + struct acpi_device *device) { ACPI_FUNCTION_TRACE("acpi_processor_power_exit"); pr->flags.power_setup_done = 0; if (acpi_device_dir(device)) - remove_proc_entry(ACPI_PROCESSOR_FILE_POWER,acpi_device_dir(device)); + remove_proc_entry(ACPI_PROCESSOR_FILE_POWER, + acpi_device_dir(device)); /* Unregister the idle handler when processor #0 is removed. */ if (pr->id == 0) {