diff options
author | Rafael J. Wysocki <rafael.j.wysocki@intel.com> | 2020-01-23 00:35:50 +0100 |
---|---|---|
committer | Rafael J. Wysocki <rafael.j.wysocki@intel.com> | 2020-01-23 00:35:50 +0100 |
commit | e6cf623ba3f83118b0f75be2cc1675931e21c887 (patch) | |
tree | b1ca19f8e64af0515fcc026327a53ce4a0fe3480 /drivers/idle/intel_idle.c | |
parent | cefb9409ff995fcc98ce44a07b549ba1b827c31b (diff) | |
parent | a3299182216397a0b943d2549d1997f4eba2bdd2 (diff) |
Merge branch 'intel_idle+acpi'
Merge changes updating the ACPI processor driver in order to export
acpi_processor_evaluate_cst() to the code outside of it and adding
ACPI support to the intel_idle driver based on that.
* intel_idle+acpi:
Documentation: admin-guide: PM: Add intel_idle document
intel_idle: Use ACPI _CST on server systems
intel_idle: Add module parameter to prevent ACPI _CST from being used
intel_idle: Allow ACPI _CST to be used for selected known processors
cpuidle: Allow idle states to be disabled by default
intel_idle: Use ACPI _CST for processor models without C-state tables
intel_idle: Refactor intel_idle_cpuidle_driver_init()
ACPI: processor: Export acpi_processor_evaluate_cst()
ACPI: processor: Make ACPI_PROCESSOR_CSTATE depend on ACPI_PROCESSOR
ACPI: processor: Clean up acpi_processor_evaluate_cst()
ACPI: processor: Introduce acpi_processor_evaluate_cst()
ACPI: processor: Export function to claim _CST control
Diffstat (limited to 'drivers/idle/intel_idle.c')
-rw-r--r-- | drivers/idle/intel_idle.c | 349 |
1 files changed, 281 insertions, 68 deletions
diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 75fd2a7b0842..1467490adfc3 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -41,6 +41,7 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt +#include <linux/acpi.h> #include <linux/kernel.h> #include <linux/cpuidle.h> #include <linux/tick.h> @@ -79,6 +80,7 @@ struct idle_cpu { unsigned long auto_demotion_disable_flags; bool byt_auto_demotion_disable_flag; bool disable_promotion_to_c1e; + bool use_acpi; }; static const struct idle_cpu *icpu; @@ -90,6 +92,11 @@ static void intel_idle_s2idle(struct cpuidle_device *dev, static struct cpuidle_state *cpuidle_state_table; /* + * Enable this state by default even if the ACPI _CST does not list it. + */ +#define CPUIDLE_FLAG_ALWAYS_ENABLE BIT(15) + +/* * Set this flag for states where the HW flushes the TLB for us * and so we don't need cross-calls to keep it consistent. * If this flag is set, SW flushes the TLB, so even if the @@ -124,7 +131,7 @@ static struct cpuidle_state nehalem_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -161,7 +168,7 @@ static struct cpuidle_state snb_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -296,7 +303,7 @@ static struct cpuidle_state ivb_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -341,7 +348,7 @@ static struct cpuidle_state ivt_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 80, .enter = &intel_idle, @@ -378,7 +385,7 @@ static struct cpuidle_state ivt_cstates_4s[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 250, .enter = &intel_idle, @@ -415,7 +422,7 @@ static struct cpuidle_state ivt_cstates_8s[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 500, .enter = &intel_idle, @@ -452,7 +459,7 @@ static struct cpuidle_state hsw_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -520,7 +527,7 @@ static struct cpuidle_state bdw_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -589,7 +596,7 @@ static struct cpuidle_state skl_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -658,7 +665,7 @@ static struct cpuidle_state skx_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -808,7 +815,7 @@ static struct cpuidle_state bxt_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -869,7 +876,7 @@ static struct cpuidle_state dnv_cstates[] = { { .name = "C1E", .desc = "MWAIT 0x01", - .flags = MWAIT2flg(0x01), + .flags = MWAIT2flg(0x01) | CPUIDLE_FLAG_ALWAYS_ENABLE, .exit_latency = 10, .target_residency = 20, .enter = &intel_idle, @@ -944,6 +951,22 @@ static void intel_idle_s2idle(struct cpuidle_device *dev, mwait_idle_with_hints(eax, ecx); } +static bool intel_idle_verify_cstate(unsigned int mwait_hint) +{ + unsigned int mwait_cstate = MWAIT_HINT2CSTATE(mwait_hint) + 1; + unsigned int num_substates = (mwait_substates >> mwait_cstate * 4) & + MWAIT_SUBSTATE_MASK; + + /* Ignore the C-state if there are NO sub-states in CPUID for it. */ + if (num_substates == 0) + return false; + + if (mwait_cstate > 2 && !boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) + mark_tsc_unstable("TSC halts in idle states deeper than C2"); + + return true; +} + static void __setup_broadcast_timer(bool on) { if (on) @@ -975,6 +998,13 @@ static const struct idle_cpu idle_cpu_nehalem = { .disable_promotion_to_c1e = true, }; +static const struct idle_cpu idle_cpu_nhx = { + .state_table = nehalem_cstates, + .auto_demotion_disable_flags = NHM_C1_AUTO_DEMOTE | NHM_C3_AUTO_DEMOTE, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_atom = { .state_table = atom_cstates, }; @@ -993,6 +1023,12 @@ static const struct idle_cpu idle_cpu_snb = { .disable_promotion_to_c1e = true, }; +static const struct idle_cpu idle_cpu_snx = { + .state_table = snb_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_byt = { .state_table = byt_cstates, .disable_promotion_to_c1e = true, @@ -1013,6 +1049,7 @@ static const struct idle_cpu idle_cpu_ivb = { static const struct idle_cpu idle_cpu_ivt = { .state_table = ivt_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, }; static const struct idle_cpu idle_cpu_hsw = { @@ -1020,11 +1057,23 @@ static const struct idle_cpu idle_cpu_hsw = { .disable_promotion_to_c1e = true, }; +static const struct idle_cpu idle_cpu_hsx = { + .state_table = hsw_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_bdw = { .state_table = bdw_cstates, .disable_promotion_to_c1e = true, }; +static const struct idle_cpu idle_cpu_bdx = { + .state_table = bdw_cstates, + .disable_promotion_to_c1e = true, + .use_acpi = true, +}; + static const struct idle_cpu idle_cpu_skl = { .state_table = skl_cstates, .disable_promotion_to_c1e = true, @@ -1033,15 +1082,18 @@ static const struct idle_cpu idle_cpu_skl = { static const struct idle_cpu idle_cpu_skx = { .state_table = skx_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, }; static const struct idle_cpu idle_cpu_avn = { .state_table = avn_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, }; static const struct idle_cpu idle_cpu_knl = { .state_table = knl_cstates, + .use_acpi = true, }; static const struct idle_cpu idle_cpu_bxt = { @@ -1052,20 +1104,21 @@ static const struct idle_cpu idle_cpu_bxt = { static const struct idle_cpu idle_cpu_dnv = { .state_table = dnv_cstates, .disable_promotion_to_c1e = true, + .use_acpi = true, }; static const struct x86_cpu_id intel_idle_ids[] __initconst = { - INTEL_CPU_FAM6(NEHALEM_EP, idle_cpu_nehalem), + INTEL_CPU_FAM6(NEHALEM_EP, idle_cpu_nhx), INTEL_CPU_FAM6(NEHALEM, idle_cpu_nehalem), INTEL_CPU_FAM6(NEHALEM_G, idle_cpu_nehalem), INTEL_CPU_FAM6(WESTMERE, idle_cpu_nehalem), - INTEL_CPU_FAM6(WESTMERE_EP, idle_cpu_nehalem), - INTEL_CPU_FAM6(NEHALEM_EX, idle_cpu_nehalem), + INTEL_CPU_FAM6(WESTMERE_EP, idle_cpu_nhx), + INTEL_CPU_FAM6(NEHALEM_EX, idle_cpu_nhx), INTEL_CPU_FAM6(ATOM_BONNELL, idle_cpu_atom), INTEL_CPU_FAM6(ATOM_BONNELL_MID, idle_cpu_lincroft), - INTEL_CPU_FAM6(WESTMERE_EX, idle_cpu_nehalem), + INTEL_CPU_FAM6(WESTMERE_EX, idle_cpu_nhx), INTEL_CPU_FAM6(SANDYBRIDGE, idle_cpu_snb), - INTEL_CPU_FAM6(SANDYBRIDGE_X, idle_cpu_snb), + INTEL_CPU_FAM6(SANDYBRIDGE_X, idle_cpu_snx), INTEL_CPU_FAM6(ATOM_SALTWELL, idle_cpu_atom), INTEL_CPU_FAM6(ATOM_SILVERMONT, idle_cpu_byt), INTEL_CPU_FAM6(ATOM_SILVERMONT_MID, idle_cpu_tangier), @@ -1073,14 +1126,14 @@ static const struct x86_cpu_id intel_idle_ids[] __initconst = { INTEL_CPU_FAM6(IVYBRIDGE, idle_cpu_ivb), INTEL_CPU_FAM6(IVYBRIDGE_X, idle_cpu_ivt), INTEL_CPU_FAM6(HASWELL, idle_cpu_hsw), - INTEL_CPU_FAM6(HASWELL_X, idle_cpu_hsw), + INTEL_CPU_FAM6(HASWELL_X, idle_cpu_hsx), INTEL_CPU_FAM6(HASWELL_L, idle_cpu_hsw), INTEL_CPU_FAM6(HASWELL_G, idle_cpu_hsw), INTEL_CPU_FAM6(ATOM_SILVERMONT_D, idle_cpu_avn), INTEL_CPU_FAM6(BROADWELL, idle_cpu_bdw), INTEL_CPU_FAM6(BROADWELL_G, idle_cpu_bdw), - INTEL_CPU_FAM6(BROADWELL_X, idle_cpu_bdw), - INTEL_CPU_FAM6(BROADWELL_D, idle_cpu_bdw), + INTEL_CPU_FAM6(BROADWELL_X, idle_cpu_bdx), + INTEL_CPU_FAM6(BROADWELL_D, idle_cpu_bdx), INTEL_CPU_FAM6(SKYLAKE_L, idle_cpu_skl), INTEL_CPU_FAM6(SKYLAKE, idle_cpu_skl), INTEL_CPU_FAM6(KABYLAKE_L, idle_cpu_skl), @@ -1095,6 +1148,162 @@ static const struct x86_cpu_id intel_idle_ids[] __initconst = { {} }; +#define INTEL_CPU_FAM6_MWAIT \ + { X86_VENDOR_INTEL, 6, X86_MODEL_ANY, X86_FEATURE_MWAIT, 0 } + +static const struct x86_cpu_id intel_mwait_ids[] __initconst = { + INTEL_CPU_FAM6_MWAIT, + {} +}; + +static bool intel_idle_max_cstate_reached(int cstate) +{ + if (cstate + 1 > max_cstate) { + pr_info("max_cstate %d reached\n", max_cstate); + return true; + } + return false; +} + +#ifdef CONFIG_ACPI_PROCESSOR_CSTATE +#include <acpi/processor.h> + +static bool no_acpi __read_mostly; +module_param(no_acpi, bool, 0444); +MODULE_PARM_DESC(no_acpi, "Do not use ACPI _CST for building the idle states list"); + +static struct acpi_processor_power acpi_state_table; + +/** + * intel_idle_cst_usable - Check if the _CST information can be used. + * + * Check if all of the C-states listed by _CST in the max_cstate range are + * ACPI_CSTATE_FFH, which means that they should be entered via MWAIT. + */ +static bool intel_idle_cst_usable(void) +{ + int cstate, limit; + + limit = min_t(int, min_t(int, CPUIDLE_STATE_MAX, max_cstate + 1), + acpi_state_table.count); + + for (cstate = 1; cstate < limit; cstate++) { + struct acpi_processor_cx *cx = &acpi_state_table.states[cstate]; + + if (cx->entry_method != ACPI_CSTATE_FFH) + return false; + } + + return true; +} + +static bool intel_idle_acpi_cst_extract(void) +{ + unsigned int cpu; + + if (no_acpi) { + pr_debug("Not allowed to use ACPI _CST\n"); + return false; + } + + for_each_possible_cpu(cpu) { + struct acpi_processor *pr = per_cpu(processors, cpu); + + if (!pr) + continue; + + if (acpi_processor_evaluate_cst(pr->handle, cpu, &acpi_state_table)) + continue; + + acpi_state_table.count++; + + if (!intel_idle_cst_usable()) + continue; + + if (!acpi_processor_claim_cst_control()) { + acpi_state_table.count = 0; + return false; + } + + return true; + } + + pr_debug("ACPI _CST not found or not usable\n"); + return false; +} + +static void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) +{ + int cstate, limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count); + + /* + * If limit > 0, intel_idle_cst_usable() has returned 'true', so all of + * the interesting states are ACPI_CSTATE_FFH. + */ + for (cstate = 1; cstate < limit; cstate++) { + struct acpi_processor_cx *cx; + struct cpuidle_state *state; + + if (intel_idle_max_cstate_reached(cstate)) + break; + + cx = &acpi_state_table.states[cstate]; + + state = &drv->states[drv->state_count++]; + + snprintf(state->name, CPUIDLE_NAME_LEN, "C%d_ACPI", cstate); + strlcpy(state->desc, cx->desc, CPUIDLE_DESC_LEN); + state->exit_latency = cx->latency; + /* + * For C1-type C-states use the same number for both the exit + * latency and target residency, because that is the case for + * C1 in the majority of the static C-states tables above. + * For the other types of C-states, however, set the target + * residency to 3 times the exit latency which should lead to + * a reasonable balance between energy-efficiency and + * performance in the majority of interesting cases. + */ + state->target_residency = cx->latency; + if (cx->type > ACPI_STATE_C1) + state->target_residency *= 3; + + state->flags = MWAIT2flg(cx->address); + if (cx->type > ACPI_STATE_C2) + state->flags |= CPUIDLE_FLAG_TLB_FLUSHED; + + state->enter = intel_idle; + state->enter_s2idle = intel_idle_s2idle; + } +} + +static bool intel_idle_off_by_default(u32 mwait_hint) +{ + int cstate, limit; + + /* + * If there are no _CST C-states, do not disable any C-states by + * default. + */ + if (!acpi_state_table.count) + return false; + + limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count); + /* + * If limit > 0, intel_idle_cst_usable() has returned 'true', so all of + * the interesting states are ACPI_CSTATE_FFH. + */ + for (cstate = 1; cstate < limit; cstate++) { + if (acpi_state_table.states[cstate].address == mwait_hint) + return false; + } + return true; +} +#else /* !CONFIG_ACPI_PROCESSOR_CSTATE */ +static inline bool intel_idle_acpi_cst_extract(void) { return false; } +static inline void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) { } +static inline bool intel_idle_off_by_default(u32 mwait_hint) { return false; } +#endif /* !CONFIG_ACPI_PROCESSOR_CSTATE */ + /* * intel_idle_probe() */ @@ -1109,17 +1318,15 @@ static int __init intel_idle_probe(void) } id = x86_match_cpu(intel_idle_ids); - if (!id) { - if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && - boot_cpu_data.x86 == 6) - pr_debug("does not run on family %d model %d\n", - boot_cpu_data.x86, boot_cpu_data.x86_model); - return -ENODEV; - } - - if (!boot_cpu_has(X86_FEATURE_MWAIT)) { - pr_debug("Please enable MWAIT in BIOS SETUP\n"); - return -ENODEV; + if (id) { + if (!boot_cpu_has(X86_FEATURE_MWAIT)) { + pr_debug("Please enable MWAIT in BIOS SETUP\n"); + return -ENODEV; + } + } else { + id = x86_match_cpu(intel_mwait_ids); + if (!id) + return -ENODEV; } if (boot_cpu_data.cpuid_level < CPUID_MWAIT_LEAF) @@ -1135,7 +1342,13 @@ static int __init intel_idle_probe(void) pr_debug("MWAIT substates: 0x%x\n", mwait_substates); icpu = (const struct idle_cpu *)id->driver_data; - cpuidle_state_table = icpu->state_table; + if (icpu) { + cpuidle_state_table = icpu->state_table; + if (icpu->use_acpi) + intel_idle_acpi_cst_extract(); + } else if (!intel_idle_acpi_cst_extract()) { + return -ENODEV; + } pr_debug("v" INTEL_IDLE_VERSION " model 0x%X\n", boot_cpu_data.x86_model); @@ -1317,60 +1530,39 @@ static void intel_idle_state_table_update(void) } } -/* - * intel_idle_cpuidle_driver_init() - * allocate, initialize cpuidle_states - */ -static void __init intel_idle_cpuidle_driver_init(void) +static void intel_idle_init_cstates_icpu(struct cpuidle_driver *drv) { int cstate; - struct cpuidle_driver *drv = &intel_idle_driver; - - intel_idle_state_table_update(); - - cpuidle_poll_state_init(drv); - drv->state_count = 1; for (cstate = 0; cstate < CPUIDLE_STATE_MAX; ++cstate) { - int num_substates, mwait_hint, mwait_cstate; + unsigned int mwait_hint; - if ((cpuidle_state_table[cstate].enter == NULL) && - (cpuidle_state_table[cstate].enter_s2idle == NULL)) + if (intel_idle_max_cstate_reached(cstate)) break; - if (cstate + 1 > max_cstate) { - pr_info("max_cstate %d reached\n", max_cstate); + if (!cpuidle_state_table[cstate].enter && + !cpuidle_state_table[cstate].enter_s2idle) break; - } - mwait_hint = flg2MWAIT(cpuidle_state_table[cstate].flags); - mwait_cstate = MWAIT_HINT2CSTATE(mwait_hint); - - /* number of sub-states for this state in CPUID.MWAIT */ - num_substates = (mwait_substates >> ((mwait_cstate + 1) * 4)) - & MWAIT_SUBSTATE_MASK; - - /* if NO sub-states for this state in CPUID, skip it */ - if (num_substates == 0) - continue; - - /* if state marked as disabled, skip it */ + /* If marked as unusable, skip this state. */ if (cpuidle_state_table[cstate].flags & CPUIDLE_FLAG_UNUSABLE) { pr_debug("state %s is disabled\n", cpuidle_state_table[cstate].name); continue; } + mwait_hint = flg2MWAIT(cpuidle_state_table[cstate].flags); + if (!intel_idle_verify_cstate(mwait_hint)) + continue; - if (((mwait_cstate + 1) > 2) && - !boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) - mark_tsc_unstable("TSC halts in idle" - " states deeper than C2"); + /* Structure copy. */ + drv->states[drv->state_count] = cpuidle_state_table[cstate]; - drv->states[drv->state_count] = /* structure copy */ - cpuidle_state_table[cstate]; + if (icpu->use_acpi && intel_idle_off_by_default(mwait_hint) && + !(cpuidle_state_table[cstate].flags & CPUIDLE_FLAG_ALWAYS_ENABLE)) + drv->states[drv->state_count].flags |= CPUIDLE_FLAG_OFF; - drv->state_count += 1; + drv->state_count++; } if (icpu->byt_auto_demotion_disable_flag) { @@ -1379,6 +1571,24 @@ static void __init intel_idle_cpuidle_driver_init(void) } } +/* + * intel_idle_cpuidle_driver_init() + * allocate, initialize cpuidle_states + */ +static void __init intel_idle_cpuidle_driver_init(void) +{ + struct cpuidle_driver *drv = &intel_idle_driver; + + intel_idle_state_table_update(); + + cpuidle_poll_state_init(drv); + drv->state_count = 1; + + if (icpu) + intel_idle_init_cstates_icpu(drv); + else + intel_idle_init_cstates_acpi(drv); +} /* * intel_idle_cpu_init() @@ -1397,6 +1607,9 @@ static int intel_idle_cpu_init(unsigned int cpu) return -EIO; } + if (!icpu) + return 0; + if (icpu->auto_demotion_disable_flags) auto_demotion_disable(); |