diff options
23 files changed, 193 insertions, 212 deletions
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.c b/drivers/gpu/drm/i915/gem/selftests/mock_context.c index 2b0327cc47c2..fd8babb513e5 100644 --- a/drivers/gpu/drm/i915/gem/selftests/mock_context.c +++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.c @@ -108,44 +108,6 @@ err_ctx: } struct i915_gem_context * -live_context_for_engine(struct intel_engine_cs *engine, struct file *file) -{ - struct i915_gem_engines *engines; - struct i915_gem_context *ctx; - struct intel_sseu null_sseu = {}; - struct intel_context *ce; - - engines = alloc_engines(1); - if (!engines) - return ERR_PTR(-ENOMEM); - - ctx = live_context(engine->i915, file); - if (IS_ERR(ctx)) { - __free_engines(engines, 0); - return ctx; - } - - ce = intel_context_create(engine); - if (IS_ERR(ce)) { - __free_engines(engines, 0); - return ERR_CAST(ce); - } - - intel_context_set_gem(ce, ctx, null_sseu); - engines->engines[0] = ce; - engines->num_engines = 1; - - mutex_lock(&ctx->engines_mutex); - i915_gem_context_set_user_engines(ctx); - engines = rcu_replace_pointer(ctx->engines, engines, 1); - mutex_unlock(&ctx->engines_mutex); - - engines_idle_release(ctx, engines); - - return ctx; -} - -struct i915_gem_context * kernel_context(struct drm_i915_private *i915, struct i915_address_space *vm) { diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.h b/drivers/gpu/drm/i915/gem/selftests/mock_context.h index 7a02fd9b5866..bc8fb37d2d24 100644 --- a/drivers/gpu/drm/i915/gem/selftests/mock_context.h +++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.h @@ -23,9 +23,6 @@ void mock_context_close(struct i915_gem_context *ctx); struct i915_gem_context * live_context(struct drm_i915_private *i915, struct file *file); -struct i915_gem_context * -live_context_for_engine(struct intel_engine_cs *engine, struct file *file); - struct i915_gem_context *kernel_context(struct drm_i915_private *i915, struct i915_address_space *vm); void kernel_context_close(struct i915_gem_context *ctx); diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c index 075657018739..5cd58e0f0dcf 100644 --- a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c +++ b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c @@ -103,8 +103,7 @@ static struct dma_buf *mock_dmabuf(int npages) struct dma_buf *dmabuf; int i; - mock = kmalloc(sizeof(*mock) + npages * sizeof(struct page *), - GFP_KERNEL); + mock = kmalloc(struct_size(mock, pages, npages), GFP_KERNEL); if (!mock) return ERR_PTR(-ENOMEM); diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c index f6b780f893f5..0c723e7c71a2 100644 --- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c @@ -750,7 +750,7 @@ static void swizzle_page(struct page *page) char *vaddr; int i; - vaddr = kmap(page); + vaddr = kmap_local_page(page); for (i = 0; i < PAGE_SIZE; i += 128) { memcpy(temp, &vaddr[i], 64); @@ -758,7 +758,7 @@ static void swizzle_page(struct page *page) memcpy(&vaddr[i + 64], temp, 64); } - kunmap(page); + kunmap_local(vaddr); } /** diff --git a/drivers/gpu/drm/i915/gt/intel_gt_regs.h b/drivers/gpu/drm/i915/gt/intel_gt_regs.h index 6dba65e54cdb..a6e50af44b46 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_regs.h +++ b/drivers/gpu/drm/i915/gt/intel_gt_regs.h @@ -409,6 +409,9 @@ #define GEN7_SO_PRIM_STORAGE_NEEDED(n) _MMIO(0x5240 + (n) * 8) #define GEN7_SO_PRIM_STORAGE_NEEDED_UDW(n) _MMIO(0x5240 + (n) * 8 + 4) +#define GEN8_WM_CHICKEN2 MCR_REG(0x5584) +#define WAIT_ON_DEPTH_STALL_DONE_DISABLE REG_BIT(5) + #define GEN9_WM_CHICKEN3 _MMIO(0x5588) #define GEN9_FACTOR_IN_CLR_VAL_HIZ (1 << 9) diff --git a/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c b/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c index d7784650e4d9..1154cd2b7c34 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c +++ b/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c @@ -464,6 +464,45 @@ static ssize_t slpc_ignore_eff_freq_store(struct kobject *kobj, return err ?: count; } +static ssize_t slpc_power_profile_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buff) +{ + struct intel_gt *gt = intel_gt_sysfs_get_drvdata(kobj, attr->attr.name); + struct intel_guc_slpc *slpc = >->uc.guc.slpc; + + switch (slpc->power_profile) { + case SLPC_POWER_PROFILES_BASE: + return sysfs_emit(buff, "[%s] %s\n", "base", "power_saving"); + case SLPC_POWER_PROFILES_POWER_SAVING: + return sysfs_emit(buff, "%s [%s]\n", "base", "power_saving"); + } + + return sysfs_emit(buff, "%u\n", slpc->power_profile); +} + +static ssize_t slpc_power_profile_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buff, size_t count) +{ + struct intel_gt *gt = intel_gt_sysfs_get_drvdata(kobj, attr->attr.name); + struct intel_guc_slpc *slpc = >->uc.guc.slpc; + char power_saving[] = "power_saving"; + char base[] = "base"; + int err; + u32 val; + + if (!strncmp(buff, power_saving, sizeof(power_saving) - 1)) + val = SLPC_POWER_PROFILES_POWER_SAVING; + else if (!strncmp(buff, base, sizeof(base) - 1)) + val = SLPC_POWER_PROFILES_BASE; + else + return -EINVAL; + + err = intel_guc_slpc_set_power_profile(slpc, val); + return err ?: count; +} + struct intel_gt_bool_throttle_attr { struct attribute attr; ssize_t (*show)(struct kobject *kobj, struct kobj_attribute *attr, @@ -668,6 +707,7 @@ INTEL_GT_ATTR_RO(media_RP0_freq_mhz); INTEL_GT_ATTR_RO(media_RPn_freq_mhz); INTEL_GT_ATTR_RW(slpc_ignore_eff_freq); +INTEL_GT_ATTR_RW(slpc_power_profile); static const struct attribute *media_perf_power_attrs[] = { &attr_media_freq_factor.attr, @@ -864,6 +904,13 @@ void intel_gt_sysfs_pm_init(struct intel_gt *gt, struct kobject *kobj) gt_warn(gt, "failed to create ignore_eff_freq sysfs (%pe)", ERR_PTR(ret)); } + if (intel_uc_uses_guc_slpc(>->uc)) { + ret = sysfs_create_file(kobj, &attr_slpc_power_profile.attr); + if (ret) + gt_warn(gt, "failed to create slpc_power_profile sysfs (%pe)", + ERR_PTR(ret)); + } + if (i915_mmio_reg_valid(intel_gt_perf_limit_reasons_reg(gt))) { ret = sysfs_create_files(kobj, throttle_reason_attrs); if (ret) diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c index 3ee544e7c203..dbdcfe130ad4 100644 --- a/drivers/gpu/drm/i915/gt/intel_reset.c +++ b/drivers/gpu/drm/i915/gt/intel_reset.c @@ -1113,7 +1113,6 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt) * Warn CI about the unrecoverable wedged condition. * Time for a reboot. */ - gt_err(gt, "Unrecoverable wedged condition\n"); add_taint_for_CI(gt->i915, TAINT_WARN); return false; } @@ -1272,10 +1271,8 @@ void intel_gt_reset(struct intel_gt *gt, } ret = resume(gt); - if (ret) { - gt_err(gt, "Failed to resume (%d)\n", ret); + if (ret) goto taint; - } finish: reset_finish(gt, awake); @@ -1641,7 +1638,6 @@ void intel_gt_set_wedged_on_init(struct intel_gt *gt) set_bit(I915_WEDGED_ON_INIT, >->reset.flags); /* Wedged on init is non-recoverable */ - gt_err(gt, "Non-recoverable wedged on init\n"); add_taint_for_CI(gt->i915, TAINT_WARN); } diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c index fa304ea088e4..2cfaedb04876 100644 --- a/drivers/gpu/drm/i915/gt/intel_rps.c +++ b/drivers/gpu/drm/i915/gt/intel_rps.c @@ -1025,6 +1025,10 @@ void intel_rps_boost(struct i915_request *rq) if (rps_uses_slpc(rps)) { slpc = rps_to_slpc(rps); + /* Waitboost should not be done with power saving profile */ + if (slpc->power_profile == SLPC_POWER_PROFILES_POWER_SAVING) + return; + if (slpc->min_freq_softlimit >= slpc->boost_freq) return; diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index 3ea9b06de1be..116683ebe074 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -691,16 +691,17 @@ static void gen12_ctx_workarounds_init(struct intel_engine_cs *engine, struct drm_i915_private *i915 = engine->i915; /* - * Wa_1409142259:tgl,dg1,adl-p + * Wa_1409142259:tgl,dg1,adl-p,adl-n * Wa_1409347922:tgl,dg1,adl-p * Wa_1409252684:tgl,dg1,adl-p * Wa_1409217633:tgl,dg1,adl-p * Wa_1409207793:tgl,dg1,adl-p - * Wa_1409178076:tgl,dg1,adl-p - * Wa_1408979724:tgl,dg1,adl-p - * Wa_14010443199:tgl,rkl,dg1,adl-p - * Wa_14010698770:tgl,rkl,dg1,adl-s,adl-p - * Wa_1409342910:tgl,rkl,dg1,adl-s,adl-p + * Wa_1409178076:tgl,dg1,adl-p,adl-n + * Wa_1408979724:tgl,dg1,adl-p,adl-n + * Wa_14010443199:tgl,rkl,dg1,adl-p,adl-n + * Wa_14010698770:tgl,rkl,dg1,adl-s,adl-p,adl-n + * Wa_1409342910:tgl,rkl,dg1,adl-s,adl-p,adl-n + * Wa_22010465259:tgl,rkl,dg1,adl-s,adl-p,adl-n */ wa_masked_en(wal, GEN11_COMMON_SLICE_CHICKEN3, GEN12_DISABLE_CPS_AWARE_COLOR_PIPE); @@ -741,6 +742,12 @@ static void gen12_ctx_workarounds_init(struct intel_engine_cs *engine, /* Wa_1606376872 */ wa_masked_en(wal, COMMON_SLICE_CHICKEN4, DISABLE_TDC_LOAD_BALANCING_CALC); } + + /* + * This bit must be set to enable performance optimization for fast + * clears. + */ + wa_mcr_write_or(wal, GEN8_WM_CHICKEN2, WAIT_ON_DEPTH_STALL_DONE_DISABLE); } static void dg1_ctx_workarounds_init(struct intel_engine_cs *engine, diff --git a/drivers/gpu/drm/i915/gt/selftest_rps.c b/drivers/gpu/drm/i915/gt/selftest_rps.c index 78c03e6c0861..73bc91c6ea07 100644 --- a/drivers/gpu/drm/i915/gt/selftest_rps.c +++ b/drivers/gpu/drm/i915/gt/selftest_rps.c @@ -477,12 +477,13 @@ int live_rps_control(void *arg) limit, intel_gpu_freq(rps, limit), min, max, ktime_to_ns(min_dt), ktime_to_ns(max_dt)); - if (limit == rps->min_freq) { - pr_err("%s: GPU throttled to minimum!\n", - engine->name); + if (limit != rps->max_freq) { + u32 throttle = intel_uncore_read(gt->uncore, + intel_gt_perf_limit_reasons_reg(gt)); + + pr_warn("%s: GPU throttled with reasons 0x%08x\n", + engine->name, throttle & GT0_PERF_LIMIT_REASONS_MASK); show_pstate_limits(rps); - err = -ENODEV; - break; } if (igt_flush_test(gt->i915)) { @@ -1115,7 +1116,7 @@ static u64 measure_power(struct intel_rps *rps, int *freq) for (i = 0; i < 5; i++) x[i] = __measure_power(5); - *freq = (*freq + intel_rps_read_actual_frequency(rps)) / 2; + *freq = (*freq + read_cagf(rps)) / 2; /* A simple triangle filter for better result stability */ sort(x, 5, sizeof(*x), cmp_u64, NULL); diff --git a/drivers/gpu/drm/i915/gt/selftest_slpc.c b/drivers/gpu/drm/i915/gt/selftest_slpc.c index e218b229681f..e61bb0bad12c 100644 --- a/drivers/gpu/drm/i915/gt/selftest_slpc.c +++ b/drivers/gpu/drm/i915/gt/selftest_slpc.c @@ -95,6 +95,21 @@ static int slpc_restore_freq(struct intel_guc_slpc *slpc, u32 min, u32 max) return 0; } +static u64 slpc_measure_power(struct intel_rps *rps, int *freq) +{ + u64 x[5]; + int i; + + for (i = 0; i < 5; i++) + x[i] = __measure_power(5); + + *freq = (*freq + intel_rps_read_actual_frequency(rps)) / 2; + + /* A simple triangle filter for better result stability */ + sort(x, 5, sizeof(*x), cmp_u64, NULL); + return div_u64(x[1] + 2 * x[2] + x[3], 4); +} + static u64 measure_power_at_freq(struct intel_gt *gt, int *freq, u64 *power) { int err = 0; @@ -103,7 +118,7 @@ static u64 measure_power_at_freq(struct intel_gt *gt, int *freq, u64 *power) if (err) return err; *freq = intel_rps_read_actual_frequency(>->rps); - *power = measure_power(>->rps, freq); + *power = slpc_measure_power(>->rps, freq); return err; } diff --git a/drivers/gpu/drm/i915/gt/shmem_utils.c b/drivers/gpu/drm/i915/gt/shmem_utils.c index bb696b29ee2c..365c4b8b04f4 100644 --- a/drivers/gpu/drm/i915/gt/shmem_utils.c +++ b/drivers/gpu/drm/i915/gt/shmem_utils.c @@ -108,7 +108,7 @@ static int __shmem_rw(struct file *file, loff_t off, if (IS_ERR(page)) return PTR_ERR(page); - vaddr = kmap(page); + vaddr = kmap_local_page(page); if (write) { memcpy(vaddr + offset_in_page(off), ptr, this); set_page_dirty(page); @@ -116,7 +116,7 @@ static int __shmem_rw(struct file *file, loff_t off, memcpy(ptr, vaddr + offset_in_page(off), this); } mark_page_accessed(page); - kunmap(page); + kunmap_local(vaddr); put_page(page); len -= this; @@ -143,11 +143,11 @@ int shmem_read_to_iosys_map(struct file *file, loff_t off, if (IS_ERR(page)) return PTR_ERR(page); - vaddr = kmap(page); + vaddr = kmap_local_page(page); iosys_map_memcpy_to(map, map_off, vaddr + offset_in_page(off), this); mark_page_accessed(page); - kunmap(page); + kunmap_local(vaddr); put_page(page); len -= this; diff --git a/drivers/gpu/drm/i915/gt/uc/abi/guc_actions_slpc_abi.h b/drivers/gpu/drm/i915/gt/uc/abi/guc_actions_slpc_abi.h index c34674e797c6..6de87ae5669e 100644 --- a/drivers/gpu/drm/i915/gt/uc/abi/guc_actions_slpc_abi.h +++ b/drivers/gpu/drm/i915/gt/uc/abi/guc_actions_slpc_abi.h @@ -228,6 +228,11 @@ struct slpc_optimized_strategies { #define SLPC_OPTIMIZED_STRATEGY_COMPUTE REG_BIT(0) +enum slpc_power_profiles { + SLPC_POWER_PROFILES_BASE = 0x0, + SLPC_POWER_PROFILES_POWER_SAVING = 0x1 +}; + /** * DOC: SLPC H2G MESSAGE FORMAT * diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c index fe53e8eccf4b..e7ccfa520df3 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c @@ -259,13 +259,14 @@ static int guc_wait_ucode(struct intel_guc *guc) } else if (delta_ms > 200) { guc_warn(guc, "excessive init time: %lldms! [status = 0x%08X, count = %d, ret = %d]\n", delta_ms, status, count, ret); - guc_warn(guc, "excessive init time: [freq = %dMHz, before = %dMHz, perf_limit_reasons = 0x%08X]\n", - intel_rps_read_actual_frequency(>->rps), before_freq, + guc_warn(guc, "excessive init time: [freq = %dMHz -> %dMHz vs %dMHz, perf_limit_reasons = 0x%08X]\n", + before_freq, intel_rps_read_actual_frequency(>->rps), + intel_rps_get_requested_frequency(>->rps), intel_uncore_read(uncore, intel_gt_perf_limit_reasons_reg(gt))); } else { - guc_dbg(guc, "init took %lldms, freq = %dMHz, before = %dMHz, status = 0x%08X, count = %d, ret = %d\n", - delta_ms, intel_rps_read_actual_frequency(>->rps), - before_freq, status, count, ret); + guc_dbg(guc, "init took %lldms, freq = %dMHz -> %dMHz vs %dMHz, status = 0x%08X, count = %d, ret = %d\n", + delta_ms, before_freq, intel_rps_read_actual_frequency(>->rps), + intel_rps_get_requested_frequency(>->rps), status, count, ret); } return ret; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_hwconfig.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_hwconfig.c index b67a15f74276..868195c33f5b 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_hwconfig.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_hwconfig.c @@ -7,6 +7,7 @@ #include "gt/intel_hwconfig.h" #include "i915_drv.h" #include "i915_memcpy.h" +#include "intel_guc_print.h" /* * GuC has a blob containing hardware configuration information (HWConfig). @@ -42,6 +43,8 @@ static int __guc_action_get_hwconfig(struct intel_guc *guc, }; int ret; + guc_dbg(guc, "Querying HW config table: size = %d, offset = 0x%08X\n", + ggtt_size, ggtt_offset); ret = intel_guc_send_mmio(guc, action, ARRAY_SIZE(action), NULL, 0); if (ret == -ENXIO) return -ENOENT; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.c index 1a0e1a412fdb..d5ee6e5e1443 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.c @@ -15,6 +15,34 @@ #include "gt/intel_gt_regs.h" #include "gt/intel_rps.h" +/** + * DOC: SLPC - Dynamic Frequency management + * + * Single Loop Power Control (SLPC) is a GuC algorithm that manages + * GT frequency based on busyness and how KMD initializes it. SLPC is + * almost completely in control after initialization except for a few + * scenarios mentioned below. + * + * KMD uses the concept of waitboost to ramp frequency to RP0 when there + * are pending submissions for a context. It achieves this by sending GuC a + * request to update the min frequency to RP0. Waitboost is disabled + * when the request retires. + * + * Another form of frequency control happens through per-context hints. + * A context can be marked as low latency during creation. That will ensure + * that SLPC uses an aggressive frequency ramp when that context is active. + * + * Power profiles add another level of control to these mechanisms. + * When power saving profile is chosen, SLPC will use conservative + * thresholds to ramp frequency, thus saving power. KMD will disable + * waitboosts as well, which achieves further power savings. Base profile + * is default and ensures balanced performance for any workload. + * + * Lastly, users have some level of control through sysfs, where min/max + * frequency values can be altered and the use of efficient freq + * can be toggled. + */ + static inline struct intel_guc *slpc_to_guc(struct intel_guc_slpc *slpc) { return container_of(slpc, struct intel_guc, slpc); @@ -265,6 +293,8 @@ int intel_guc_slpc_init(struct intel_guc_slpc *slpc) slpc->num_boosts = 0; slpc->media_ratio_mode = SLPC_MEDIA_RATIO_MODE_DYNAMIC_CONTROL; + slpc->power_profile = SLPC_POWER_PROFILES_BASE; + mutex_init(&slpc->lock); INIT_WORK(&slpc->boost_work, slpc_boost_work); @@ -575,6 +605,34 @@ int intel_guc_slpc_set_media_ratio_mode(struct intel_guc_slpc *slpc, u32 val) return ret; } +int intel_guc_slpc_set_power_profile(struct intel_guc_slpc *slpc, u32 val) +{ + struct drm_i915_private *i915 = slpc_to_i915(slpc); + intel_wakeref_t wakeref; + int ret = 0; + + if (val > SLPC_POWER_PROFILES_POWER_SAVING) + return -EINVAL; + + mutex_lock(&slpc->lock); + wakeref = intel_runtime_pm_get(&i915->runtime_pm); + + ret = slpc_set_param(slpc, + SLPC_PARAM_POWER_PROFILE, + val); + if (ret) + guc_err(slpc_to_guc(slpc), + "Failed to set power profile to %d: %pe\n", + val, ERR_PTR(ret)); + else + slpc->power_profile = val; + + intel_runtime_pm_put(&i915->runtime_pm, wakeref); + mutex_unlock(&slpc->lock); + + return ret; +} + void intel_guc_pm_intrmsk_enable(struct intel_gt *gt) { u32 pm_intrmsk_mbz = 0; @@ -736,6 +794,13 @@ int intel_guc_slpc_enable(struct intel_guc_slpc *slpc) /* Enable SLPC Optimized Strategy for compute */ intel_guc_slpc_set_strategy(slpc, SLPC_OPTIMIZED_STRATEGY_COMPUTE); + /* Set cached value of power_profile */ + ret = intel_guc_slpc_set_power_profile(slpc, slpc->power_profile); + if (unlikely(ret)) { + guc_probe_error(guc, "Failed to set SLPC power profile: %pe\n", ERR_PTR(ret)); + return ret; + } + return 0; } diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.h index 1cb5fd44f05c..fc9f761b4372 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.h @@ -46,5 +46,6 @@ void intel_guc_slpc_boost(struct intel_guc_slpc *slpc); void intel_guc_slpc_dec_waiters(struct intel_guc_slpc *slpc); int intel_guc_slpc_set_ignore_eff_freq(struct intel_guc_slpc *slpc, bool val); int intel_guc_slpc_set_strategy(struct intel_guc_slpc *slpc, u32 val); +int intel_guc_slpc_set_power_profile(struct intel_guc_slpc *slpc, u32 val); #endif diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc_types.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc_types.h index a88651331497..83673b10ac4e 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc_types.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_slpc_types.h @@ -33,6 +33,9 @@ struct intel_guc_slpc { u32 max_freq_softlimit; bool ignore_eff_freq; + /* Base or power saving */ + u32 power_profile; + /* cached media ratio mode */ u32 media_ratio_mode; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_huc.c b/drivers/gpu/drm/i915/gt/uc/intel_huc.c index b3cbf85c00cb..f30c90650b7e 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_huc.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_huc.c @@ -489,13 +489,15 @@ int intel_huc_wait_for_auth_complete(struct intel_huc *huc, if (delta_ms > 50) { huc_warn(huc, "excessive auth time: %lldms! [status = 0x%08X, count = %d, ret = %d]\n", delta_ms, huc->status[type].reg.reg, count, ret); - huc_warn(huc, "excessive auth time: [freq = %dMHz, before = %dMHz, perf_limit_reasons = 0x%08X]\n", - intel_rps_read_actual_frequency(>->rps), before_freq, + huc_warn(huc, "excessive auth time: [freq = %dMHz -> %dMHz vs %dMHz, perf_limit_reasons = 0x%08X]\n", + before_freq, intel_rps_read_actual_frequency(>->rps), + intel_rps_get_requested_frequency(>->rps), intel_uncore_read(uncore, intel_gt_perf_limit_reasons_reg(gt))); } else { - huc_dbg(huc, "auth took %lldms, freq = %dMHz, before = %dMHz, status = 0x%08X, count = %d, ret = %d\n", - delta_ms, intel_rps_read_actual_frequency(>->rps), - before_freq, huc->status[type].reg.reg, count, ret); + huc_dbg(huc, "auth took %lldms, freq = %dMHz -> %dMHz vs %dMHz, status = 0x%08X, count = %d, ret = %d\n", + delta_ms, before_freq, intel_rps_read_actual_frequency(>->rps), + intel_rps_get_requested_frequency(>->rps), + huc->status[type].reg.reg, count, ret); } /* mark the load process as complete even if the wait failed */ diff --git a/drivers/gpu/drm/i915/i915_module.c b/drivers/gpu/drm/i915/i915_module.c index 2f88970cc0a9..5862754c662c 100644 --- a/drivers/gpu/drm/i915/i915_module.c +++ b/drivers/gpu/drm/i915/i915_module.c @@ -71,8 +71,6 @@ static const struct { { .init = i915_vma_resource_module_init, .exit = i915_vma_resource_module_exit }, { .init = i915_mock_selftests }, - { .init = i915_pmu_init, - .exit = i915_pmu_exit }, { .init = i915_pci_register_driver, .exit = i915_pci_unregister_driver }, { .init = i915_perf_sysctl_register, diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c index e55db036be1b..69a109d02116 100644 --- a/drivers/gpu/drm/i915/i915_pmu.c +++ b/drivers/gpu/drm/i915/i915_pmu.c @@ -28,9 +28,6 @@ BIT(I915_SAMPLE_WAIT) | \ BIT(I915_SAMPLE_SEMA)) -static cpumask_t i915_pmu_cpumask; -static unsigned int i915_pmu_target_cpu = -1; - static struct i915_pmu *event_to_pmu(struct perf_event *event) { return container_of(event->pmu, struct i915_pmu, base); @@ -642,10 +639,6 @@ static int i915_pmu_event_init(struct perf_event *event) if (event->cpu < 0) return -EINVAL; - /* only allow running on one cpu at a time */ - if (!cpumask_test_cpu(event->cpu, &i915_pmu_cpumask)) - return -EINVAL; - if (is_engine_event(event)) ret = engine_event_init(event); else @@ -891,11 +884,6 @@ static void i915_pmu_event_del(struct perf_event *event, int flags) i915_pmu_event_stop(event, PERF_EF_UPDATE); } -static int i915_pmu_event_event_idx(struct perf_event *event) -{ - return 0; -} - struct i915_str_attribute { struct device_attribute attr; const char *str; @@ -940,23 +928,6 @@ static ssize_t i915_pmu_event_show(struct device *dev, return sprintf(buf, "config=0x%lx\n", eattr->val); } -static ssize_t cpumask_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - return cpumap_print_to_pagebuf(true, buf, &i915_pmu_cpumask); -} - -static DEVICE_ATTR_RO(cpumask); - -static struct attribute *i915_cpumask_attrs[] = { - &dev_attr_cpumask.attr, - NULL, -}; - -static const struct attribute_group i915_pmu_cpumask_attr_group = { - .attrs = i915_cpumask_attrs, -}; - #define __event(__counter, __name, __unit) \ { \ .counter = (__counter), \ @@ -1173,92 +1144,12 @@ static void free_event_attributes(struct i915_pmu *pmu) pmu->pmu_attr = NULL; } -static int i915_pmu_cpu_online(unsigned int cpu, struct hlist_node *node) -{ - struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node); - - /* Select the first online CPU as a designated reader. */ - if (cpumask_empty(&i915_pmu_cpumask)) - cpumask_set_cpu(cpu, &i915_pmu_cpumask); - - return 0; -} - -static int i915_pmu_cpu_offline(unsigned int cpu, struct hlist_node *node) -{ - struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node); - unsigned int target = i915_pmu_target_cpu; - - /* - * Unregistering an instance generates a CPU offline event which we must - * ignore to avoid incorrectly modifying the shared i915_pmu_cpumask. - */ - if (!pmu->registered) - return 0; - - if (cpumask_test_and_clear_cpu(cpu, &i915_pmu_cpumask)) { - target = cpumask_any_but(topology_sibling_cpumask(cpu), cpu); - - /* Migrate events if there is a valid target */ - if (target < nr_cpu_ids) { - cpumask_set_cpu(target, &i915_pmu_cpumask); - i915_pmu_target_cpu = target; - } - } - - if (target < nr_cpu_ids && target != pmu->cpuhp.cpu) { - perf_pmu_migrate_context(&pmu->base, cpu, target); - pmu->cpuhp.cpu = target; - } - - return 0; -} - -static enum cpuhp_state cpuhp_state = CPUHP_INVALID; - -int i915_pmu_init(void) -{ - int ret; - - ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, - "perf/x86/intel/i915:online", - i915_pmu_cpu_online, - i915_pmu_cpu_offline); - if (ret < 0) - pr_notice("Failed to setup cpuhp state for i915 PMU! (%d)\n", - ret); - else - cpuhp_state = ret; - - return 0; -} - -void i915_pmu_exit(void) -{ - if (cpuhp_state != CPUHP_INVALID) - cpuhp_remove_multi_state(cpuhp_state); -} - -static int i915_pmu_register_cpuhp_state(struct i915_pmu *pmu) -{ - if (cpuhp_state == CPUHP_INVALID) - return -EINVAL; - - return cpuhp_state_add_instance(cpuhp_state, &pmu->cpuhp.node); -} - -static void i915_pmu_unregister_cpuhp_state(struct i915_pmu *pmu) -{ - cpuhp_state_remove_instance(cpuhp_state, &pmu->cpuhp.node); -} - void i915_pmu_register(struct drm_i915_private *i915) { struct i915_pmu *pmu = &i915->pmu; const struct attribute_group *attr_groups[] = { &i915_pmu_format_attr_group, &pmu->events_attr_group, - &i915_pmu_cpumask_attr_group, NULL }; int ret = -ENOMEM; @@ -1266,7 +1157,6 @@ void i915_pmu_register(struct drm_i915_private *i915) spin_lock_init(&pmu->lock); hrtimer_init(&pmu->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); pmu->timer.function = i915_sample; - pmu->cpuhp.cpu = -1; init_rc6(pmu); if (IS_DGFX(i915)) { @@ -1295,28 +1185,22 @@ void i915_pmu_register(struct drm_i915_private *i915) pmu->base.module = THIS_MODULE; pmu->base.task_ctx_nr = perf_invalid_context; + pmu->base.scope = PERF_PMU_SCOPE_SYS_WIDE; pmu->base.event_init = i915_pmu_event_init; pmu->base.add = i915_pmu_event_add; pmu->base.del = i915_pmu_event_del; pmu->base.start = i915_pmu_event_start; pmu->base.stop = i915_pmu_event_stop; pmu->base.read = i915_pmu_event_read; - pmu->base.event_idx = i915_pmu_event_event_idx; ret = perf_pmu_register(&pmu->base, pmu->name, -1); if (ret) goto err_groups; - ret = i915_pmu_register_cpuhp_state(pmu); - if (ret) - goto err_unreg; - pmu->registered = true; return; -err_unreg: - perf_pmu_unregister(&pmu->base); err_groups: kfree(pmu->base.attr_groups); err_attr: @@ -1340,8 +1224,6 @@ void i915_pmu_unregister(struct drm_i915_private *i915) hrtimer_cancel(&pmu->timer); - i915_pmu_unregister_cpuhp_state(pmu); - perf_pmu_unregister(&pmu->base); kfree(pmu->base.attr_groups); if (IS_DGFX(i915)) diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h index 0ec78c2b4f20..5826cc81858c 100644 --- a/drivers/gpu/drm/i915/i915_pmu.h +++ b/drivers/gpu/drm/i915/i915_pmu.h @@ -57,13 +57,6 @@ struct i915_pmu_sample { struct i915_pmu { /** - * @cpuhp: Struct used for CPU hotplug handling. - */ - struct { - struct hlist_node node; - unsigned int cpu; - } cpuhp; - /** * @base: PMU base. */ struct pmu base; @@ -155,15 +148,11 @@ struct i915_pmu { }; #ifdef CONFIG_PERF_EVENTS -int i915_pmu_init(void); -void i915_pmu_exit(void); void i915_pmu_register(struct drm_i915_private *i915); void i915_pmu_unregister(struct drm_i915_private *i915); void i915_pmu_gt_parked(struct intel_gt *gt); void i915_pmu_gt_unparked(struct intel_gt *gt); #else -static inline int i915_pmu_init(void) { return 0; } -static inline void i915_pmu_exit(void) {} static inline void i915_pmu_register(struct drm_i915_private *i915) {} static inline void i915_pmu_unregister(struct drm_i915_private *i915) {} static inline void i915_pmu_gt_parked(struct intel_gt *gt) {} diff --git a/drivers/gpu/drm/i915/selftests/i915_gem.c b/drivers/gpu/drm/i915/selftests/i915_gem.c index e817d233df61..ad650f67114a 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem.c @@ -45,13 +45,15 @@ static void trash_stolen(struct drm_i915_private *i915) struct i915_ggtt *ggtt = to_gt(i915)->ggtt; const u64 slot = ggtt->error_capture.start; const resource_size_t size = resource_size(&i915->dsm.stolen); + struct rnd_state prng; unsigned long page; - u32 prng = 0x12345678; /* XXX: fsck. needs some more thought... */ if (!i915_ggtt_has_aperture(ggtt)) return; + prandom_seed_state(&prng, 0x12345678); + for (page = 0; page < size; page += PAGE_SIZE) { const dma_addr_t dma = i915->dsm.stolen.start + page; u32 __iomem *s; @@ -64,8 +66,7 @@ static void trash_stolen(struct drm_i915_private *i915) s = io_mapping_map_atomic_wc(&ggtt->iomap, slot); for (x = 0; x < PAGE_SIZE / sizeof(u32); x++) { - prng = next_pseudo_random32(prng); - iowrite32(prng, &s[x]); + iowrite32(prandom_u32_state(&prng), &s[x]); } io_mapping_unmap_atomic(s); } |