From patchwork Wed Apr 20 06:56:34 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Trinabh Gupta X-Patchwork-Id: 720841 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by demeter1.kernel.org (8.14.4/8.14.3) with ESMTP id p3K6ugTu014402 for ; Wed, 20 Apr 2011 06:56:54 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753160Ab1DTG4o (ORCPT ); Wed, 20 Apr 2011 02:56:44 -0400 Received: from e28smtp02.in.ibm.com ([122.248.162.2]:44594 "EHLO e28smtp02.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752990Ab1DTG4n (ORCPT ); Wed, 20 Apr 2011 02:56:43 -0400 Received: from d28relay03.in.ibm.com (d28relay03.in.ibm.com [9.184.220.60]) by e28smtp02.in.ibm.com (8.14.4/8.13.1) with ESMTP id p3K6uctR022186; Wed, 20 Apr 2011 12:26:38 +0530 Received: from d28av01.in.ibm.com (d28av01.in.ibm.com [9.184.220.63]) by d28relay03.in.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id p3K6ucOk4100334; Wed, 20 Apr 2011 12:26:38 +0530 Received: from d28av01.in.ibm.com (loopback [127.0.0.1]) by d28av01.in.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id p3K6ubEo004126; Wed, 20 Apr 2011 12:26:38 +0530 Received: from tringupt.in.ibm.com ([9.79.234.201]) by d28av01.in.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id p3K6uYdQ003938; Wed, 20 Apr 2011 12:26:34 +0530 From: Trinabh Gupta Subject: [RFC PATCH V3 4/4] cpuidle: Single/Global registration of idle states To: arjan@linux.intel.com, peterz@infradead.org, lenb@kernel.org, venki@google.com, ak@linux.intel.com, len.brown@intel.com Cc: linux-kernel@vger.kernel.org, linux-pm@lists.linux-foundation.org, linux-sh@vger.kernel.org, linux-arm-kernel@lists.infradead.org, davinci-linux-open-source@linux.davincidsp.com, linux-omap@vger.kernel.org, linux-acpi@vger.kernel.org Date: Wed, 20 Apr 2011 12:26:34 +0530 Message-ID: <20110420065608.332.30043.stgit@tringupt.in.ibm.com> In-Reply-To: <20110420065445.332.13688.stgit@tringupt.in.ibm.com> References: <20110420065445.332.13688.stgit@tringupt.in.ibm.com> User-Agent: StGIT/0.14.3 MIME-Version: 1.0 Sender: linux-acpi-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-acpi@vger.kernel.org X-Greylist: IP, sender and recipient auto-whitelisted, not delayed by milter-greylist-4.2.6 (demeter1.kernel.org [140.211.167.41]); Wed, 20 Apr 2011 06:56:55 +0000 (UTC) With this patch there is single copy of cpuidle_states structure instead of per-cpu. The statistics needed on per-cpu basis by the governor are kept per-cpu. This simplifies the cpuidle subsystem as state registration is done by single cpu only. Having single copy of cpuidle_states saves memory. Rare case of asymmetric C-states can be handled within the cpuidle driverand architectures such as POWER do not have asymmetric C-states. To Do: 1. Handle the case when idle states may change at run time and acpi_processor_cst_has_changed() routine is called. 2. Handle acpi_processor_power_exit() correctly i.e. ensure unregistration of cpuidle driver since registration is now moved inside acpi_processor_power_init(). Signed-off-by: Trinabh Gupta --- arch/arm/mach-at91/cpuidle.c | 31 +++++---- arch/arm/mach-davinci/cpuidle.c | 39 ++++++----- arch/arm/mach-kirkwood/cpuidle.c | 30 +++++--- arch/arm/mach-omap2/cpuidle34xx.c | 97 +++++++++++++++++++++------ arch/sh/kernel/cpu/shmobile/cpuidle.c | 18 +++-- drivers/acpi/processor_driver.c | 18 +---- drivers/acpi/processor_idle.c | 117 +++++++++++++++++++++++++++------ drivers/cpuidle/cpuidle.c | 46 ++++--------- drivers/cpuidle/driver.c | 25 +++++++ drivers/cpuidle/governors/ladder.c | 26 ++++--- drivers/cpuidle/governors/menu.c | 20 +++--- drivers/cpuidle/sysfs.c | 3 + drivers/idle/intel_idle.c | 80 +++++++++++++++++------ include/linux/cpuidle.h | 21 ++++-- 14 files changed, 378 insertions(+), 193 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/arch/arm/mach-at91/cpuidle.c b/arch/arm/mach-at91/cpuidle.c index ed38e3c..039b378 100644 --- a/arch/arm/mach-at91/cpuidle.c +++ b/arch/arm/mach-at91/cpuidle.c @@ -33,6 +33,7 @@ static struct cpuidle_driver at91_idle_driver = { /* Actual code that puts the SoC in different idle states */ static int at91_enter_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct timeval before, after; @@ -68,27 +69,29 @@ static int at91_enter_idle(struct cpuidle_device *dev, static int at91_init_cpuidle(void) { struct cpuidle_device *device; - - cpuidle_register_driver(&at91_idle_driver); + struct cpuidle_drive *driver = &at91_idle_driver; device = &per_cpu(at91_cpuidle_device, smp_processor_id()); device->state_count = AT91_MAX_STATES; + driver->state_count = AT91_MAX_STATES; /* Wait for interrupt state */ - device->states[0].enter = at91_enter_idle; - device->states[0].exit_latency = 1; - device->states[0].target_residency = 10000; - device->states[0].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[0].name, "WFI"); - strcpy(device->states[0].desc, "Wait for interrupt"); + driver->states[0].enter = at91_enter_idle; + driver->states[0].exit_latency = 1; + driver->states[0].target_residency = 10000; + driver->states[0].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[0].name, "WFI"); + strcpy(driver->states[0].desc, "Wait for interrupt"); /* Wait for interrupt and RAM self refresh state */ - device->states[1].enter = at91_enter_idle; - device->states[1].exit_latency = 10; - device->states[1].target_residency = 10000; - device->states[1].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[1].name, "RAM_SR"); - strcpy(device->states[1].desc, "WFI and RAM Self Refresh"); + driver->states[1].enter = at91_enter_idle; + driver->states[1].exit_latency = 10; + driver->states[1].target_residency = 10000; + driver->states[1].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[1].name, "RAM_SR"); + strcpy(driver->states[1].desc, "WFI and RAM Self Refresh"); + + cpuidle_register_driver(&at91_idle_driver); if (cpuidle_register_device(device)) { printk(KERN_ERR "at91_init_cpuidle: Failed registering\n"); diff --git a/arch/arm/mach-davinci/cpuidle.c b/arch/arm/mach-davinci/cpuidle.c index e3aebe6..ba46a83 100644 --- a/arch/arm/mach-davinci/cpuidle.c +++ b/arch/arm/mach-davinci/cpuidle.c @@ -78,6 +78,7 @@ static struct davinci_ops davinci_states[DAVINCI_CPUIDLE_MAX_STATES] = { /* Actual code that puts the SoC in different idle states */ static int davinci_enter_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct cpuidle_state_usage *state_usage = &dev->states_usage[index]; @@ -112,6 +113,7 @@ static int __init davinci_cpuidle_probe(struct platform_device *pdev) { int ret; struct cpuidle_device *device; + strcut cpuidle_driver *driver = &davinci_idle_driver; struct davinci_cpuidle_config *pdata = pdev->dev.platform_data; device = &per_cpu(davinci_cpuidle_device, smp_processor_id()); @@ -123,32 +125,33 @@ static int __init davinci_cpuidle_probe(struct platform_device *pdev) ddr2_reg_base = pdata->ddr2_ctlr_base; - ret = cpuidle_register_driver(&davinci_idle_driver); - if (ret) { - dev_err(&pdev->dev, "failed to register driver\n"); - return ret; - } - /* Wait for interrupt state */ - device->states[0].enter = davinci_enter_idle; - device->states[0].exit_latency = 1; - device->states[0].target_residency = 10000; - device->states[0].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[0].name, "WFI"); - strcpy(device->states[0].desc, "Wait for interrupt"); + driver->states[0].enter = davinci_enter_idle; + driver->states[0].exit_latency = 1; + driver->states[0].target_residency = 10000; + driver->states[0].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[0].name, "WFI"); + strcpy(driver->states[0].desc, "Wait for interrupt"); /* Wait for interrupt and DDR self refresh state */ - device->states[1].enter = davinci_enter_idle; - device->states[1].exit_latency = 10; - device->states[1].target_residency = 10000; - device->states[1].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[1].name, "DDR SR"); - strcpy(device->states[1].desc, "WFI and DDR Self Refresh"); + driver->states[1].enter = davinci_enter_idle; + driver->states[1].exit_latency = 10; + driver->states[1].target_residency = 10000; + driver->states[1].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[1].name, "DDR SR"); + strcpy(driver->states[1].desc, "WFI and DDR Self Refresh"); if (pdata->ddr2_pdown) davinci_states[1].flags |= DAVINCI_CPUIDLE_FLAGS_DDR2_PWDN; cpuidle_set_statedata(&device->states_usage[1], &davinci_states[1]); device->state_count = DAVINCI_CPUIDLE_MAX_STATES; + driver->state_count = DAVINCI_CPUIDLE_MAX_STATES; + + ret = cpuidle_register_driver(&davinci_idle_driver); + if (ret) { + dev_err(&pdev->dev, "failed to register driver\n"); + return ret; + } ret = cpuidle_register_device(device); if (ret) { diff --git a/arch/arm/mach-kirkwood/cpuidle.c b/arch/arm/mach-kirkwood/cpuidle.c index d135a41..f04f6b1 100644 --- a/arch/arm/mach-kirkwood/cpuidle.c +++ b/arch/arm/mach-kirkwood/cpuidle.c @@ -32,6 +32,7 @@ static DEFINE_PER_CPU(struct cpuidle_device, kirkwood_cpuidle_device); /* Actual code that puts the SoC in different idle states */ static int kirkwood_enter_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct timeval before, after; @@ -70,28 +71,29 @@ static int kirkwood_enter_idle(struct cpuidle_device *dev, static int kirkwood_init_cpuidle(void) { struct cpuidle_device *device; - - cpuidle_register_driver(&kirkwood_idle_driver); + struct cpuidle_driver *driver = &kirkwood_idle_driver; device = &per_cpu(kirkwood_cpuidle_device, smp_processor_id()); device->state_count = KIRKWOOD_MAX_STATES; + driver->state_count = KIRKWOOD_MAX_STATES; /* Wait for interrupt state */ - device->states[0].enter = kirkwood_enter_idle; - device->states[0].exit_latency = 1; - device->states[0].target_residency = 10000; - device->states[0].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[0].name, "WFI"); - strcpy(device->states[0].desc, "Wait for interrupt"); + driver->states[0].enter = kirkwood_enter_idle; + driver->states[0].exit_latency = 1; + driver->states[0].target_residency = 10000; + driver->states[0].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[0].name, "WFI"); + strcpy(driver->states[0].desc, "Wait for interrupt"); /* Wait for interrupt and DDR self refresh state */ - device->states[1].enter = kirkwood_enter_idle; - device->states[1].exit_latency = 10; - device->states[1].target_residency = 10000; - device->states[1].flags = CPUIDLE_FLAG_TIME_VALID; - strcpy(device->states[1].name, "DDR SR"); - strcpy(device->states[1].desc, "WFI and DDR Self Refresh"); + driver->states[1].enter = kirkwood_enter_idle; + driver->states[1].exit_latency = 10; + driver->states[1].target_residency = 10000; + driver->states[1].flags = CPUIDLE_FLAG_TIME_VALID; + strcpy(driver->states[1].name, "DDR SR"); + strcpy(driver->states[1].desc, "WFI and DDR Self Refresh"); + cpuidle_register_driver(&kirkwood_idle_driver); if (cpuidle_register_device(device)) { printk(KERN_ERR "kirkwood_init_cpuidle: Failed registering\n"); return -EIO; diff --git a/arch/arm/mach-omap2/cpuidle34xx.c b/arch/arm/mach-omap2/cpuidle34xx.c index 4282420..6641574 100644 --- a/arch/arm/mach-omap2/cpuidle34xx.c +++ b/arch/arm/mach-omap2/cpuidle34xx.c @@ -114,12 +114,14 @@ static int _cpuidle_deny_idle(struct powerdomain *pwrdm, /** * omap3_enter_idle - Programs OMAP3 to enter the specified state * @dev: cpuidle device + * @drv: cpuidle driver * @index: the index of state to be entered * * Called from the CPUidle framework to program the device to the * specified target state selected by the governor. */ static int omap3_enter_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct omap3_processor_cx *cx = @@ -175,6 +177,7 @@ return_sleep_time: /** * next_valid_state - Find next valid c-state * @dev: cpuidle device + * @drv: cpuidle driver * @index: Index of currently selected c-state * * If the state corresponding to index is valid, index is returned back @@ -182,9 +185,10 @@ return_sleep_time: * still valid (as defined in omap3_power_states[]) and returns its index. */ static int next_valid_state(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { - struct cpuidle_state *curr = &dev->states[index]; + struct cpuidle_state *curr = &drv->states[index]; struct cpuidle_state_usage *curr_usage = &dev->states_usage[index]; struct cpuidle_state *next = NULL; struct omap3_processor_cx *cx; @@ -202,8 +206,8 @@ static int next_valid_state(struct cpuidle_device *dev, * Reach the current state starting at highest C-state */ for (; idx >= OMAP3_STATE_C1; idx--) { - if (&dev->states[idx] == curr) { - next = &dev->states[idx]; + if (&drv->states[idx] == curr) { + next = &drv->states[idx]; next_index = idx; break; } @@ -224,7 +228,7 @@ static int next_valid_state(struct cpuidle_device *dev, cx = cpuidle_get_statedata(&dev->states_usage[idx]); if (cx->valid) { - next = &dev->states[idx]; + next = &drv->states[idx]; next_index = idx; break; } @@ -241,6 +245,7 @@ static int next_valid_state(struct cpuidle_device *dev, /** * omap3_enter_idle_bm - Checks for any bus activity * @dev: cpuidle device + * @drv: cpuidle driver * @index: array index of target state to be programmed * * Used for C states with CPUIDLE_FLAG_CHECK_BM flag set. This @@ -248,18 +253,19 @@ static int next_valid_state(struct cpuidle_device *dev, * device to the specified or a safer state. */ static int omap3_enter_idle_bm(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { - struct cpuidle_state *state = &dev->states[index]; - int new_state_idx = next_valid_state(dev, index); + struct cpuidle_state *state = &drv->states[index]; + int new_state_idx = next_valid_state(dev, drv, index); u32 core_next_state, per_next_state = 0, per_saved_state = 0; u32 cam_state; struct omap3_processor_cx *cx; int ret; if ((state->flags & CPUIDLE_FLAG_CHECK_BM) && omap3_idle_bm_check()) { - BUG_ON(dev->safe_state_index < 0); - new_state_idx = dev->safe_state_index; + BUG_ON(drv->safe_state_index < 0); + new_state_idx = drv->safe_state_index; goto select_state; } @@ -280,7 +286,7 @@ static int omap3_enter_idle_bm(struct cpuidle_device *dev, */ cam_state = pwrdm_read_pwrst(cam_pd); if (cam_state == PWRDM_POWER_ON) { - new_state_idx = dev->safe_state_index; + new_state_idx = drv->safe_state_index; goto select_state; } @@ -298,7 +304,7 @@ static int omap3_enter_idle_bm(struct cpuidle_device *dev, pwrdm_set_next_pwrst(per_pd, per_next_state); select_state: - ret = omap3_enter_idle(dev, new_state_idx); + ret = omap3_enter_idle(dev, drv, new_state_idx); /* Restore original PER state if it was modified */ if (per_next_state != per_saved_state) @@ -497,18 +503,15 @@ struct cpuidle_driver omap3_idle_driver = { }; /** - * omap3_idle_init - Init routine for OMAP3 idle + * omap3_cpuidle_driver_init * - * Registers the OMAP3 specific cpuidle driver with the cpuidle - * framework with the valid set of states. + * Sets up and registers omap3 idle driver */ -int __init omap3_idle_init(void) +static int omap3_cpuidle_driver_init(void) { - int i, count = 0; + int i, retval, count = 0; struct omap3_processor_cx *cx; struct cpuidle_state *state; - struct cpuidle_state_usage *state_usage; - struct cpuidle_device *dev; mpu_pd = pwrdm_lookup("mpu_pwrdm"); core_pd = pwrdm_lookup("core_pwrdm"); @@ -516,18 +519,13 @@ int __init omap3_idle_init(void) cam_pd = pwrdm_lookup("cam_pwrdm"); omap_init_power_states(); - cpuidle_register_driver(&omap3_idle_driver); - - dev = &per_cpu(omap3_idle_dev, smp_processor_id()); for (i = OMAP3_STATE_C1; i < OMAP3_MAX_STATES; i++) { cx = &omap3_power_states[i]; - state = &dev->states[count]; - state_usage = &dev->states_usage[count]; + state = &drv->states[count]; if (!cx->valid) continue; - cpuidle_set_statedata(state_usage, cx); state->exit_latency = cx->sleep_latency + cx->wakeup_latency; state->target_residency = cx->threshold; state->flags = cx->flags; @@ -542,13 +540,64 @@ int __init omap3_idle_init(void) if (!count) return -EINVAL; - dev->state_count = count; + drv->state_count = count; if (enable_off_mode) omap3_cpuidle_update_states(PWRDM_POWER_OFF, PWRDM_POWER_OFF); else omap3_cpuidle_update_states(PWRDM_POWER_RET, PWRDM_POWER_RET); + retval = cpuidle_register_driver(&omap3_idle_driver); + if (retval) { + printk(KERN_ERR "%s: CPUidle register driver failed\n", + __func__); + return retval; + } + + return 0; +} + +/** + * omap3_idle_init + * + * Registers the OMAP3 specific cpuidle driver with the cpuidle + * framework with the valid set of states, registers per cpu + * cpuidle device. + */ +int __init omap3_idle_init(void) +{ + int i, retval, count = 0; + static int cpuidle_drv_init; + struct omap3_processor_cx *cx; + struct cpuidle_state_usage *state_usage; + struct cpuidle_device *dev; + + if (!cpuidle_drv_init) { + retval = omap3_cpuidle_driver_init(); + if (retval) { + printk(KERN_ERR "%s: CPUidle register driver failed\n", + __func__); + return retval; + } + cpuidle_drv_init++; + } + + dev = &per_cpu(omap3_idle_dev, smp_processor_id()); + + for (i = OMAP3_STATE_C1; i < OMAP3_MAX_STATES; i++) { + cx = &omap3_power_states[i]; + state_usage = &dev->states_usage[count]; + + if (!cx->valid) + continue; + cpuidle_set_statedata(state_usage, cx); + count++; + } + + if (!count) + return -EINVAL; + dev->state_count = count; + if (cpuidle_register_device(dev)) { printk(KERN_ERR "%s: CPUidle register device failed\n", __func__); diff --git a/arch/sh/kernel/cpu/shmobile/cpuidle.c b/arch/sh/kernel/cpu/shmobile/cpuidle.c index 2340d62..b689f17 100644 --- a/arch/sh/kernel/cpu/shmobile/cpuidle.c +++ b/arch/sh/kernel/cpu/shmobile/cpuidle.c @@ -25,6 +25,7 @@ static unsigned long cpuidle_mode[] = { }; static int cpuidle_sleep_enter(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { unsigned long allowed_mode = arch_hwblk_sleep_mode(); @@ -67,19 +68,19 @@ static struct cpuidle_driver cpuidle_driver = { void sh_mobile_setup_cpuidle(void) { struct cpuidle_device *dev = &cpuidle_dev; + struct cpuidle_driver *drv = &cpuidle_driver; struct cpuidle_state *state; int i; - cpuidle_register_driver(&cpuidle_driver); for (i = 0; i < CPUIDLE_STATE_MAX; i++) { - dev->states[i].name[0] = '\0'; - dev->states[i].desc[0] = '\0'; + drv->states[i].name[0] = '\0'; + drv->states[i].desc[0] = '\0'; } i = CPUIDLE_DRIVER_STATE_START; - state = &dev->states[i++]; + state = &drv->states[i++]; snprintf(state->name, CPUIDLE_NAME_LEN, "C1"); strncpy(state->desc, "SuperH Sleep Mode", CPUIDLE_DESC_LEN); state->exit_latency = 1; @@ -89,10 +90,10 @@ void sh_mobile_setup_cpuidle(void) state->flags |= CPUIDLE_FLAG_TIME_VALID; state->enter = cpuidle_sleep_enter; - dev->safe_state_index = i-1; + drv->safe_state_index = i-1; if (sh_mobile_sleep_supported & SUSP_SH_SF) { - state = &dev->states[i++]; + state = &drv->states[i++]; snprintf(state->name, CPUIDLE_NAME_LEN, "C2"); strncpy(state->desc, "SuperH Sleep Mode [SF]", CPUIDLE_DESC_LEN); @@ -105,7 +106,7 @@ void sh_mobile_setup_cpuidle(void) } if (sh_mobile_sleep_supported & SUSP_SH_STANDBY) { - state = &dev->states[i++]; + state = &drv->states[i++]; snprintf(state->name, CPUIDLE_NAME_LEN, "C3"); strncpy(state->desc, "SuperH Mobile Standby Mode [SF]", CPUIDLE_DESC_LEN); @@ -117,7 +118,10 @@ void sh_mobile_setup_cpuidle(void) state->enter = cpuidle_sleep_enter; } + drv->state_count = i; dev->state_count = i; + cpuidle_register_driver(&cpuidle_driver); + cpuidle_register_device(dev); } diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index a4e0f1b..91464b4 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -503,8 +503,7 @@ static int __cpuinit acpi_processor_add(struct acpi_device *device) acpi_processor_get_throttling_info(pr); acpi_processor_get_limit_info(pr); - - if (cpuidle_get_driver() == &acpi_idle_driver) + if (!cpuidle_get_driver() || cpuidle_get_driver() == &acpi_idle_driver) acpi_processor_power_init(pr, device); pr->cdev = thermal_cooling_device_register("Processor", device, @@ -800,17 +799,9 @@ static int __init acpi_processor_init(void) memset(&errata, 0, sizeof(errata)); - if (!cpuidle_register_driver(&acpi_idle_driver)) { - printk(KERN_DEBUG "ACPI: %s registered with cpuidle\n", - acpi_idle_driver.name); - } else { - printk(KERN_DEBUG "ACPI: acpi_idle yielding to %s\n", - cpuidle_get_driver()->name); - } - result = acpi_bus_register_driver(&acpi_processor_driver); if (result < 0) - goto out_cpuidle; + return result; acpi_processor_install_hotplug_notify(); @@ -821,11 +812,6 @@ static int __init acpi_processor_init(void) acpi_processor_throttling_init(); return 0; - -out_cpuidle: - cpuidle_unregister_driver(&acpi_idle_driver); - - return result; } static void __exit acpi_processor_exit(void) diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index bd29363..505355d 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -741,11 +741,13 @@ static inline void acpi_idle_do_entry(struct acpi_processor_cx *cx) /** * acpi_idle_enter_c1 - enters an ACPI C1 state-type * @dev: the target CPU + * @drv: cpuidle driver containing cpuidle state info * @index: index of target state * * This is equivalent to the HALT instruction. */ -static int acpi_idle_enter_c1(struct cpuidle_device *dev, int index) +static int acpi_idle_enter_c1(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { ktime_t kt1, kt2; s64 idle_time; @@ -789,9 +791,11 @@ static int acpi_idle_enter_c1(struct cpuidle_device *dev, int index) /** * acpi_idle_enter_simple - enters an ACPI state without BM handling * @dev: the target CPU + * @drv: cpuidle driver with cpuidle state information * @index: the index of suggested state */ -static int acpi_idle_enter_simple(struct cpuidle_device *dev, int index) +static int acpi_idle_enter_simple(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct acpi_processor *pr; struct cpuidle_state_usage *state_usage = &dev->states_usage[index]; @@ -873,11 +877,13 @@ static DEFINE_SPINLOCK(c3_lock); /** * acpi_idle_enter_bm - enters C3 with proper BM handling * @dev: the target CPU + * @drv: cpuidle driver containing state data * @index: the index of suggested state * * If BM is detected, the deepest non-C3 idle state is entered instead. */ -static int acpi_idle_enter_bm(struct cpuidle_device *dev, int index) +static int acpi_idle_enter_bm(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { struct acpi_processor *pr; struct cpuidle_state_usage *state_usage = &dev->states_usage[index]; @@ -900,9 +906,9 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev, int index) } if (!cx->bm_sts_skip && acpi_idle_bm_check()) { - if (dev->safe_state_index >= 0) { - return dev->states[dev->safe_state_index].enter(dev, - dev->safe_state_index); + if (drv->safe_state_index >= 0) { + return drv->states[drv->safe_state_index].enter(dev, + drv, drv->safe_state_index); } else { local_irq_disable(); acpi_safe_halt(); @@ -999,14 +1005,15 @@ struct cpuidle_driver acpi_idle_driver = { }; /** - * acpi_processor_setup_cpuidle - prepares and configures CPUIDLE + * acpi_processor_setup_cpuidle_cx - prepares and configures CPUIDLE + * device i.e. per-cpu data + * * @pr: the ACPI processor */ -static int acpi_processor_setup_cpuidle(struct acpi_processor *pr) +static int acpi_processor_setup_cpuidle_cx(struct acpi_processor *pr) { int i, count = CPUIDLE_DRIVER_STATE_START; struct acpi_processor_cx *cx; - struct cpuidle_state *state; struct cpuidle_state_usage *state_usage; struct cpuidle_device *dev = &pr->power.dev; @@ -1018,18 +1025,12 @@ static int acpi_processor_setup_cpuidle(struct acpi_processor *pr) } dev->cpu = pr->id; - dev->safe_state_index = -1; - for (i = 0; i < CPUIDLE_STATE_MAX; i++) { - dev->states[i].name[0] = '\0'; - dev->states[i].desc[0] = '\0'; - } if (max_cstate == 0) max_cstate = 1; for (i = 1; i < ACPI_PROCESSOR_MAX_POWER && i <= max_cstate; i++) { cx = &pr->power.states[i]; - state = &dev->states[count]; state_usage = &dev->states_usage[count]; if (!cx->valid) @@ -1041,8 +1042,61 @@ static int acpi_processor_setup_cpuidle(struct acpi_processor *pr) !(acpi_gbl_FADT.flags & ACPI_FADT_C2_MP_SUPPORTED)) continue; #endif + cpuidle_set_statedata(state_usage, cx); + count++; + if (count == CPUIDLE_STATE_MAX) + break; + } + + dev->state_count = count; + + if (!count) + return -EINVAL; + + return 0; +} + +/** + * acpi_processor_setup_cpuidle states- prepares and configures cpuidle + * global state data i.e. idle routines + * + * @pr: the ACPI processor + */ +static int acpi_processor_setup_cpuidle_states(struct acpi_processor *pr) +{ + int i, count = CPUIDLE_DRIVER_STATE_START; + struct acpi_processor_cx *cx; + struct cpuidle_state *state; + struct cpuidle_driver *drv = &acpi_idle_driver; + + if (!pr->flags.power_setup_done) + return -EINVAL; + + if (pr->flags.power == 0) { + return -EINVAL; + } + + drv->safe_state_index = -1; + + if (max_cstate == 0) + max_cstate = 1; + + for (i = 1; i < ACPI_PROCESSOR_MAX_POWER && i <= max_cstate; i++) { + cx = &pr->power.states[i]; + + if (!cx->valid) + continue; + +#ifdef CONFIG_HOTPLUG_CPU + if ((cx->type != ACPI_STATE_C1) && (num_online_cpus() > 1) && + !pr->flags.has_cst && + !(acpi_gbl_FADT.flags & ACPI_FADT_C2_MP_SUPPORTED)) + continue; +#endif + + state = &drv->states[count]; snprintf(state->name, CPUIDLE_NAME_LEN, "C%d", i); strncpy(state->desc, cx->desc, CPUIDLE_DESC_LEN); state->exit_latency = cx->latency; @@ -1055,13 +1109,13 @@ static int acpi_processor_setup_cpuidle(struct acpi_processor *pr) state->flags |= CPUIDLE_FLAG_TIME_VALID; state->enter = acpi_idle_enter_c1; - dev->safe_state_index = count; + drv->safe_state_index = count; break; case ACPI_STATE_C2: state->flags |= CPUIDLE_FLAG_TIME_VALID; state->enter = acpi_idle_enter_simple; - dev->safe_state_index = count; + drv->safe_state_index = count; break; case ACPI_STATE_C3: @@ -1077,7 +1131,7 @@ static int acpi_processor_setup_cpuidle(struct acpi_processor *pr) break; } - dev->state_count = count; + drv->state_count = count; if (!count) return -EINVAL; @@ -1106,7 +1160,15 @@ int acpi_processor_cst_has_changed(struct acpi_processor *pr) cpuidle_disable_device(&pr->power.dev); acpi_processor_get_power_info(pr); if (pr->flags.power) { - acpi_processor_setup_cpuidle(pr); + /* + * To Do: Currently state data within driver + * is not updated. So change this to also update + * actual state data i.e. what this routine is + * meant for. Maybe complete unregistration and + * re-registration. + * + */ + acpi_processor_setup_cpuidle_cx(pr); ret = cpuidle_enable_device(&pr->power.dev); } cpuidle_resume_and_unlock(); @@ -1118,7 +1180,7 @@ int __cpuinit acpi_processor_power_init(struct acpi_processor *pr, struct acpi_device *device) { acpi_status status = 0; - static int first_run; + static int first_run, acpi_processor_registered; if (disabled_by_idle_boot_param()) return 0; @@ -1154,7 +1216,15 @@ int __cpuinit acpi_processor_power_init(struct acpi_processor *pr, * platforms that only support C1. */ if (pr->flags.power) { - acpi_processor_setup_cpuidle(pr); + if (!acpi_processor_registered) { + acpi_processor_setup_cpuidle_states(pr); + if (!cpuidle_register_driver(&acpi_idle_driver)) { + printk(KERN_DEBUG "ACPI: %s registered with cpuidle\n", + acpi_idle_driver.name); + acpi_processor_registered = 1; + } + } + acpi_processor_setup_cpuidle_cx(pr); if (cpuidle_register_device(&pr->power.dev)) return -EIO; } @@ -1168,6 +1238,11 @@ int acpi_processor_power_exit(struct acpi_processor *pr, return 0; cpuidle_unregister_device(&pr->power.dev); + /* We will have to have unregister driver as well here + * since we move register_driver to power_init. Maybe + * just do an unregister everytime; which will be successful + * only during the last call. + */ pr->flags.power_setup_done = 0; return 0; diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index 5d6f98d..845d3ef 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -50,6 +50,7 @@ static int __cpuidle_register_device(struct cpuidle_device *dev); static void cpuidle_idle_call(void) { struct cpuidle_device *dev = __this_cpu_read(cpuidle_devices); + struct cpuidle_driver *drv = cpuidle_get_driver(); struct cpuidle_state *target_state; int next_state, entered_state; @@ -76,19 +77,19 @@ static void cpuidle_idle_call(void) #endif /* ask the governor for the next state */ - next_state = cpuidle_curr_governor->select(dev); + next_state = cpuidle_curr_governor->select(drv, dev); if (need_resched()) { local_irq_enable(); return; } - target_state = &dev->states[next_state]; + target_state = &drv->states[next_state]; /* Is using next_state here correct?? */ trace_power_start(POWER_CSTATE, next_state, dev->cpu); trace_cpu_idle(next_state, dev->cpu); - entered_state = target_state->enter(dev, next_state); + entered_state = target_state->enter(dev, drv, next_state); trace_power_end(dev->cpu); trace_cpu_idle(PWR_EVENT_EXIT, dev->cpu); @@ -144,7 +145,8 @@ void cpuidle_resume_and_unlock(void) EXPORT_SYMBOL_GPL(cpuidle_resume_and_unlock); #ifdef CONFIG_ARCH_HAS_CPU_RELAX -static int poll_idle(struct cpuidle_device *dev, int index) +static int poll_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { ktime_t t1, t2; s64 diff; @@ -167,12 +169,9 @@ static int poll_idle(struct cpuidle_device *dev, int index) return index; } -static void poll_idle_init(struct cpuidle_device *dev) +static void poll_idle_init(struct cpuidle_driver *drv) { - struct cpuidle_state *state = &dev->states[0]; - struct cpuidle_state_usage *state_usage = &dev->states_usage[0]; - - cpuidle_set_statedata(state_usage, NULL); + struct cpuidle_state *state = &drv->states[0]; snprintf(state->name, CPUIDLE_NAME_LEN, "POLL"); snprintf(state->desc, CPUIDLE_DESC_LEN, "CPUIDLE CORE POLL IDLE"); @@ -183,7 +182,7 @@ static void poll_idle_init(struct cpuidle_device *dev) state->enter = poll_idle; } #else -static void poll_idle_init(struct cpuidle_device *dev) {} +static void poll_idle_init(struct cpuidle_driver *drv) {} #endif /* CONFIG_ARCH_HAS_CPU_RELAX */ /** @@ -210,13 +209,14 @@ int cpuidle_enable_device(struct cpuidle_device *dev) return ret; } - poll_idle_init(dev); + poll_idle_init(cpuidle_get_driver()); + cpuidle_set_statedata(&dev->states_usage[0], NULL); if ((ret = cpuidle_add_state_sysfs(dev))) return ret; if (cpuidle_curr_governor->enable && - (ret = cpuidle_curr_governor->enable(dev))) + (ret = cpuidle_curr_governor->enable(cpuidle_get_driver(), dev))) goto fail_sysfs; for (i = 0; i < dev->state_count; i++) { @@ -257,7 +257,7 @@ void cpuidle_disable_device(struct cpuidle_device *dev) dev->enabled = 0; if (cpuidle_curr_governor->disable) - cpuidle_curr_governor->disable(dev); + cpuidle_curr_governor->disable(cpuidle_get_driver(), dev); cpuidle_remove_state_sysfs(dev); enabled_devices--; @@ -285,26 +285,6 @@ static int __cpuidle_register_device(struct cpuidle_device *dev) init_completion(&dev->kobj_unregister); - /* - * cpuidle driver should set the dev->power_specified bit - * before registering the device if the driver provides - * power_usage numbers. - * - * For those devices whose ->power_specified is not set, - * we fill in power_usage with decreasing values as the - * cpuidle code has an implicit assumption that state Cn - * uses less power than C(n-1). - * - * With CONFIG_ARCH_HAS_CPU_RELAX, C0 is already assigned - * an power value of -1. So we use -2, -3, etc, for other - * c-states. - */ - if (!dev->power_specified) { - int i; - for (i = CPUIDLE_DRIVER_STATE_START; i < dev->state_count; i++) - dev->states[i].power_usage = -1 - i; - } - per_cpu(cpuidle_devices, dev->cpu) = dev; list_add(&dev->device_list, &cpuidle_detected_devices); if ((ret = cpuidle_add_sysfs(sys_dev))) { diff --git a/drivers/cpuidle/driver.c b/drivers/cpuidle/driver.c index fd1601e..4b04129 100644 --- a/drivers/cpuidle/driver.c +++ b/drivers/cpuidle/driver.c @@ -17,6 +17,30 @@ static struct cpuidle_driver *cpuidle_curr_driver; DEFINE_SPINLOCK(cpuidle_driver_lock); +static void __cpuidle_register_driver(struct cpuidle_driver *drv) +{ + /* + * cpuidle driver should set the drv->power_specified bit + * before registering if the driver provides + * power_usage numbers. + * + * If power_specified is not set, + * we fill in power_usage with decreasing values as the + * cpuidle code has an implicit assumption that state Cn + * uses less power than C(n-1). + * + * With CONFIG_ARCH_HAS_CPU_RELAX, C0 is already assigned + * an power value of -1. So we use -2, -3, etc, for other + * c-states. + */ + if (!drv->power_specified) { + int i; + for (i = CPUIDLE_DRIVER_STATE_START; i < drv->state_count; i++) + drv->states[i].power_usage = -1 - i; + } +} + + /** * cpuidle_register_driver - registers a driver * @drv: the driver @@ -31,6 +55,7 @@ int cpuidle_register_driver(struct cpuidle_driver *drv) spin_unlock(&cpuidle_driver_lock); return -EBUSY; } + __cpuidle_register_driver(drv); cpuidle_curr_driver = drv; spin_unlock(&cpuidle_driver_lock); diff --git a/drivers/cpuidle/governors/ladder.c b/drivers/cpuidle/governors/ladder.c index 6a686a7..988d813 100644 --- a/drivers/cpuidle/governors/ladder.c +++ b/drivers/cpuidle/governors/ladder.c @@ -60,9 +60,11 @@ static inline void ladder_do_selection(struct ladder_device *ldev, /** * ladder_select_state - selects the next state to enter + * @drv: cpuidle driver * @dev: the CPU */ -static int ladder_select_state(struct cpuidle_device *dev) +static int ladder_select_state(struct cpuidle_driver *drv, + struct cpuidle_device *dev) { struct ladder_device *ldev = &__get_cpu_var(ladder_devices); struct ladder_device_state *last_state; @@ -77,15 +79,15 @@ static int ladder_select_state(struct cpuidle_device *dev) last_state = &ldev->states[last_idx]; - if (dev->states[last_idx].flags & CPUIDLE_FLAG_TIME_VALID) - last_residency = cpuidle_get_last_residency(dev) - dev->states[last_idx].exit_latency; + if (drv->states[last_idx].flags & CPUIDLE_FLAG_TIME_VALID) + last_residency = cpuidle_get_last_residency(dev) - drv->states[last_idx].exit_latency; else last_residency = last_state->threshold.promotion_time + 1; /* consider promotion */ - if (last_idx < dev->state_count - 1 && + if (last_idx < drv->state_count - 1 && last_residency > last_state->threshold.promotion_time && - dev->states[last_idx + 1].exit_latency <= latency_req) { + drv->states[last_idx + 1].exit_latency <= latency_req) { last_state->stats.promotion_count++; last_state->stats.demotion_count = 0; if (last_state->stats.promotion_count >= last_state->threshold.promotion_count) { @@ -96,11 +98,11 @@ static int ladder_select_state(struct cpuidle_device *dev) /* consider demotion */ if (last_idx > CPUIDLE_DRIVER_STATE_START && - dev->states[last_idx].exit_latency > latency_req) { + drv->states[last_idx].exit_latency > latency_req) { int i; for (i = last_idx - 1; i > CPUIDLE_DRIVER_STATE_START; i--) { - if (dev->states[i].exit_latency <= latency_req) + if (drv->states[i].exit_latency <= latency_req) break; } ladder_do_selection(ldev, last_idx, i); @@ -123,9 +125,11 @@ static int ladder_select_state(struct cpuidle_device *dev) /** * ladder_enable_device - setup for the governor + * @drv: cpuidle driver * @dev: the CPU */ -static int ladder_enable_device(struct cpuidle_device *dev) +static int ladder_enable_device(struct cpuidle_driver *drv, + struct cpuidle_device *dev) { int i; struct ladder_device *ldev = &per_cpu(ladder_devices, dev->cpu); @@ -134,8 +138,8 @@ static int ladder_enable_device(struct cpuidle_device *dev) ldev->last_state_idx = CPUIDLE_DRIVER_STATE_START; - for (i = 0; i < dev->state_count; i++) { - state = &dev->states[i]; + for (i = 0; i < drv->state_count; i++) { + state = &drv->states[i]; lstate = &ldev->states[i]; lstate->stats.promotion_count = 0; @@ -144,7 +148,7 @@ static int ladder_enable_device(struct cpuidle_device *dev) lstate->threshold.promotion_count = PROMOTION_COUNT; lstate->threshold.demotion_count = DEMOTION_COUNT; - if (i < dev->state_count - 1) + if (i < drv->state_count - 1) lstate->threshold.promotion_time = state->exit_latency; if (i > 0) lstate->threshold.demotion_time = state->exit_latency; diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c index 40b5630..968149b 100644 --- a/drivers/cpuidle/governors/menu.c +++ b/drivers/cpuidle/governors/menu.c @@ -182,7 +182,7 @@ static inline int performance_multiplier(void) static DEFINE_PER_CPU(struct menu_device, menu_devices); -static void menu_update(struct cpuidle_device *dev); +static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev); /* This implements DIV_ROUND_CLOSEST but avoids 64 bit division */ static u64 div_round64(u64 dividend, u32 divisor) @@ -228,9 +228,10 @@ static void detect_repeating_patterns(struct menu_device *data) /** * menu_select - selects the next idle state to enter + * @drv: cpuidle driver containing state data * @dev: the CPU */ -static int menu_select(struct cpuidle_device *dev) +static int menu_select(struct cpuidle_driver *drv, struct cpuidle_device *dev) { struct menu_device *data = &__get_cpu_var(menu_devices); int latency_req = pm_qos_request(PM_QOS_CPU_DMA_LATENCY); @@ -239,7 +240,7 @@ static int menu_select(struct cpuidle_device *dev) int multiplier; if (data->needs_update) { - menu_update(dev); + menu_update(drv, dev); data->needs_update = 0; } @@ -283,8 +284,8 @@ static int menu_select(struct cpuidle_device *dev) * Find the idle state with the lowest power while satisfying * our constraints. */ - for (i = CPUIDLE_DRIVER_STATE_START; i < dev->state_count; i++) { - struct cpuidle_state *s = &dev->states[i]; + for (i = CPUIDLE_DRIVER_STATE_START; i < drv->state_count; i++) { + struct cpuidle_state *s = &drv->states[i]; if (s->target_residency > data->predicted_us) continue; @@ -321,14 +322,15 @@ static void menu_reflect(struct cpuidle_device *dev, int index) /** * menu_update - attempts to guess what happened after entry + * @drv: cpuidle driver containing state data * @dev: the CPU */ -static void menu_update(struct cpuidle_device *dev) +static void menu_update(struct cpuidle_driver *drv, struct cpuidle_device *dev) { struct menu_device *data = &__get_cpu_var(menu_devices); int last_idx = data->last_state_idx; unsigned int last_idle_us = cpuidle_get_last_residency(dev); - struct cpuidle_state *target = &dev->states[last_idx]; + struct cpuidle_state *target = &drv->states[last_idx]; unsigned int measured_us; u64 new_factor; @@ -382,9 +384,11 @@ static void menu_update(struct cpuidle_device *dev) /** * menu_enable_device - scans a CPU's states and does setup + * @drv: cpuidle driver * @dev: the CPU */ -static int menu_enable_device(struct cpuidle_device *dev) +static int menu_enable_device(struct cpuidle_driver *drv, + struct cpuidle_device *dev) { struct menu_device *data = &per_cpu(menu_devices, dev->cpu); diff --git a/drivers/cpuidle/sysfs.c b/drivers/cpuidle/sysfs.c index 09c9c77..90be2ad 100644 --- a/drivers/cpuidle/sysfs.c +++ b/drivers/cpuidle/sysfs.c @@ -318,13 +318,14 @@ int cpuidle_add_state_sysfs(struct cpuidle_device *device) { int i, ret = -ENOMEM; struct cpuidle_state_kobj *kobj; + struct cpuidle_driver *drv = cpuidle_get_driver(); /* state statistics */ for (i = 0; i < device->state_count; i++) { kobj = kzalloc(sizeof(struct cpuidle_state_kobj), GFP_KERNEL); if (!kobj) goto error_state; - kobj->state = &device->states[i]; + kobj->state = &drv->states[i]; kobj->state_usage = &device->states_usage[i]; init_completion(&kobj->kobj_unregister); diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 4f92d96..3bcd8c2 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -81,7 +81,8 @@ static unsigned int mwait_substates; static unsigned int lapic_timer_reliable_states = (1 << 1); /* Default to only C1 */ static struct cpuidle_device __percpu *intel_idle_cpuidle_devices; -static int intel_idle(struct cpuidle_device *dev, int index); +static int intel_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index); static struct cpuidle_state *cpuidle_state_table; @@ -229,13 +230,15 @@ static int get_driver_data(int cstate) /** * intel_idle * @dev: cpuidle_device + * @drv: cpuidle driver * @index: index of cpuidle state * */ -static int intel_idle(struct cpuidle_device *dev, int index) +static int intel_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { unsigned long ecx = 1; /* break on interrupt flag */ - struct cpuidle_state *state = &dev->states[index]; + struct cpuidle_state *state = &drv->states[index]; struct cpuidle_state_usage *state_usage = &dev->states_usage[index]; unsigned long eax = (unsigned long)cpuidle_get_statedata(state_usage); unsigned int cstate; @@ -424,6 +427,60 @@ static void intel_idle_cpuidle_devices_uninit(void) return; } /* + * intel_idle_cpuidle_driver_init() + * allocate, initialize cpuidle_states + */ +static int intel_idle_cpuidle_driver_init(void) +{ + int cstate; + struct cpuidle_driver *drv = &intel_idle_driver; + + drv->state_count = 1; + + for (cstate = 1; cstate < MWAIT_MAX_NUM_CSTATES; ++cstate) { + int num_substates; + + if (cstate > max_cstate) { + printk(PREFIX "max_cstate %d reached\n", + max_cstate); + break; + } + + /* does the state exist in CPUID.MWAIT? */ + num_substates = (mwait_substates >> ((cstate) * 4)) + & MWAIT_SUBSTATE_MASK; + if (num_substates == 0) + continue; + /* is the state not enabled? */ + if (cpuidle_state_table[cstate].enter == NULL) { + /* does the driver not know about the state? */ + if (*cpuidle_state_table[cstate].name == '\0') + pr_debug(PREFIX "unaware of model 0x%x" + " MWAIT %d please" + " contact lenb@kernel.org", + boot_cpu_data.x86_model, cstate); + continue; + } + + if ((cstate > 2) && + !boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) + mark_tsc_unstable("TSC halts in idle" + " states deeper than C2"); + + drv->states[drv->state_count] = /* structure copy */ + cpuidle_state_table[cstate]; + + drv->state_count += 1; + } + + if (auto_demotion_disable_flags) + smp_call_function(auto_demotion_disable, NULL, 1); + + return 0; +} + + +/* * intel_idle_cpuidle_devices_init() * allocate, initialize, register cpuidle_devices */ @@ -457,23 +514,9 @@ static int intel_idle_cpuidle_devices_init(void) continue; /* is the state not enabled? */ if (cpuidle_state_table[cstate].enter == NULL) { - /* does the driver not know about the state? */ - if (*cpuidle_state_table[cstate].name == '\0') - pr_debug(PREFIX "unaware of model 0x%x" - " MWAIT %d please" - " contact lenb@kernel.org", - boot_cpu_data.x86_model, cstate); continue; } - if ((cstate > 2) && - !boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) - mark_tsc_unstable("TSC halts in idle" - " states deeper than C2"); - - dev->states[dev->state_count] = /* structure copy */ - cpuidle_state_table[cstate]; - dev->states_usage[dev->state_count].driver_data = (void *)get_driver_data(cstate); @@ -488,8 +531,6 @@ static int intel_idle_cpuidle_devices_init(void) return -EIO; } } - if (auto_demotion_disable_flags) - smp_call_function(auto_demotion_disable, NULL, 1); return 0; } @@ -507,6 +548,7 @@ static int __init intel_idle_init(void) if (retval) return retval; + intel_idle_cpuidle_driver_init(); retval = cpuidle_register_driver(&intel_idle_driver); if (retval) { printk(KERN_DEBUG PREFIX "intel_idle yielding to %s", diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 5a1a238..c70fb8c 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -22,6 +22,7 @@ #define CPUIDLE_DESC_LEN 32 struct cpuidle_device; +struct cpuidle_driver; /**************************** @@ -45,6 +46,7 @@ struct cpuidle_state { unsigned int target_residency; /* in US */ int (*enter) (struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index); }; @@ -83,20 +85,17 @@ struct cpuidle_state_kobj { struct cpuidle_device { unsigned int registered:1; unsigned int enabled:1; - unsigned int power_specified:1; unsigned int cpu; int last_residency; int state_count; - struct cpuidle_state states[CPUIDLE_STATE_MAX]; struct cpuidle_state_usage states_usage[CPUIDLE_STATE_MAX]; struct cpuidle_state_kobj *kobjs[CPUIDLE_STATE_MAX]; - struct list_head device_list; + struct list_head device_list; struct kobject kobj; struct completion kobj_unregister; void *governor_data; - int safe_state_index; }; DECLARE_PER_CPU(struct cpuidle_device *, cpuidle_devices); @@ -120,6 +119,11 @@ static inline int cpuidle_get_last_residency(struct cpuidle_device *dev) struct cpuidle_driver { char name[CPUIDLE_NAME_LEN]; struct module *owner; + + unsigned int power_specified:1; + struct cpuidle_state states[CPUIDLE_STATE_MAX]; + int state_count; + int safe_state_index; }; #ifdef CONFIG_CPU_IDLE @@ -162,10 +166,13 @@ struct cpuidle_governor { struct list_head governor_list; unsigned int rating; - int (*enable) (struct cpuidle_device *dev); - void (*disable) (struct cpuidle_device *dev); + int (*enable) (struct cpuidle_driver *drv, + struct cpuidle_device *dev); + void (*disable) (struct cpuidle_driver *drv, + struct cpuidle_device *dev); - int (*select) (struct cpuidle_device *dev); + int (*select) (struct cpuidle_driver *drv, + struct cpuidle_device *dev); void (*reflect) (struct cpuidle_device *dev, int index); struct module *owner;