diff options
Diffstat (limited to 'drivers/gpu/drm/i915/gt')
27 files changed, 243 insertions, 87 deletions
diff --git a/drivers/gpu/drm/i915/gt/intel_engine.h b/drivers/gpu/drm/i915/gt/intel_engine.h index 325da0414d94..f6a98cf1e5a5 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine.h +++ b/drivers/gpu/drm/i915/gt/intel_engine.h @@ -79,6 +79,29 @@ struct lock_class_key; #define ENGINE_WRITE(...) __ENGINE_WRITE_OP(write, __VA_ARGS__) #define ENGINE_WRITE_FW(...) __ENGINE_WRITE_OP(write_fw, __VA_ARGS__) +#define __HAS_ENGINE(engine_mask, id) ((engine_mask) & BIT(id)) +#define HAS_ENGINE(gt, id) __HAS_ENGINE((gt)->info.engine_mask, id) + +#define __ENGINE_INSTANCES_MASK(mask, first, count) ({ \ + unsigned int first__ = (first); \ + unsigned int count__ = (count); \ + ((mask) & GENMASK(first__ + count__ - 1, first__)) >> first__; \ +}) + +#define ENGINE_INSTANCES_MASK(gt, first, count) \ + __ENGINE_INSTANCES_MASK((gt)->info.engine_mask, first, count) + +#define RCS_MASK(gt) \ + ENGINE_INSTANCES_MASK(gt, RCS0, I915_MAX_RCS) +#define BCS_MASK(gt) \ + ENGINE_INSTANCES_MASK(gt, BCS0, I915_MAX_BCS) +#define VDBOX_MASK(gt) \ + ENGINE_INSTANCES_MASK(gt, VCS0, I915_MAX_VCS) +#define VEBOX_MASK(gt) \ + ENGINE_INSTANCES_MASK(gt, VECS0, I915_MAX_VECS) +#define CCS_MASK(gt) \ + ENGINE_INSTANCES_MASK(gt, CCS0, I915_MAX_CCS) + #define GEN6_RING_FAULT_REG_READ(engine__) \ intel_uncore_read((engine__)->uncore, RING_FAULT_REG(engine__)) @@ -355,4 +378,12 @@ u64 intel_clamp_preempt_timeout_ms(struct intel_engine_cs *engine, u64 value); u64 intel_clamp_stop_timeout_ms(struct intel_engine_cs *engine, u64 value); u64 intel_clamp_timeslice_duration_ms(struct intel_engine_cs *engine, u64 value); +#define rb_to_uabi_engine(rb) \ + rb_entry_safe(rb, struct intel_engine_cs, uabi_node) + +#define for_each_uabi_engine(engine__, i915__) \ + for ((engine__) = rb_to_uabi_engine(rb_first(&(i915__)->uabi_engines));\ + (engine__); \ + (engine__) = rb_to_uabi_engine(rb_next(&(engine__)->uabi_node))) + #endif /* _INTEL_RINGBUFFER_H_ */ diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt.c b/drivers/gpu/drm/i915/gt/intel_ggtt.c index f6c59f20832f..46a5aa4ab9c8 100644 --- a/drivers/gpu/drm/i915/gt/intel_ggtt.c +++ b/drivers/gpu/drm/i915/gt/intel_ggtt.c @@ -289,6 +289,14 @@ u64 gen8_ggtt_pte_encode(dma_addr_t addr, return pte; } +static dma_addr_t gen8_ggtt_pte_decode(u64 pte, bool *is_present, bool *is_local) +{ + *is_present = pte & GEN8_PAGE_PRESENT; + *is_local = pte & GEN12_GGTT_PTE_LM; + + return pte & GEN12_GGTT_PTE_ADDR_MASK; +} + static bool should_update_ggtt_with_bind(struct i915_ggtt *ggtt) { struct intel_gt *gt = ggtt->vm.gt; @@ -435,6 +443,11 @@ static void gen8_set_pte(void __iomem *addr, gen8_pte_t pte) writeq(pte, addr); } +static gen8_pte_t gen8_get_pte(void __iomem *addr) +{ + return readq(addr); +} + static void gen8_ggtt_insert_page(struct i915_address_space *vm, dma_addr_t addr, u64 offset, @@ -450,6 +463,16 @@ static void gen8_ggtt_insert_page(struct i915_address_space *vm, ggtt->invalidate(ggtt); } +static dma_addr_t gen8_ggtt_read_entry(struct i915_address_space *vm, + u64 offset, bool *is_present, bool *is_local) +{ + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); + gen8_pte_t __iomem *pte = + (gen8_pte_t __iomem *)ggtt->gsm + offset / I915_GTT_PAGE_SIZE; + + return ggtt->vm.pte_decode(gen8_get_pte(pte), is_present, is_local); +} + static void gen8_ggtt_insert_page_bind(struct i915_address_space *vm, dma_addr_t addr, u64 offset, unsigned int pat_index, u32 flags) @@ -605,6 +628,17 @@ static void gen6_ggtt_insert_page(struct i915_address_space *vm, ggtt->invalidate(ggtt); } +static dma_addr_t gen6_ggtt_read_entry(struct i915_address_space *vm, + u64 offset, + bool *is_present, bool *is_local) +{ + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); + gen6_pte_t __iomem *pte = + (gen6_pte_t __iomem *)ggtt->gsm + offset / I915_GTT_PAGE_SIZE; + + return vm->pte_decode(ioread32(pte), is_present, is_local); +} + /* * Binds an object into the global gtt with the specified cache level. * The object will be accessible to the GPU via commands whose operands @@ -769,6 +803,14 @@ void intel_ggtt_unbind_vma(struct i915_address_space *vm, vm->clear_range(vm, vma_res->start, vma_res->vma_size); } +dma_addr_t intel_ggtt_read_entry(struct i915_address_space *vm, + u64 offset, bool *is_present, bool *is_local) +{ + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); + + return ggtt->vm.read_entry(vm, offset, is_present, is_local); +} + /* * Reserve the top of the GuC address space for firmware images. Addresses * beyond GUC_GGTT_TOP in the GuC address space are inaccessible by GuC, @@ -1245,6 +1287,7 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt) ggtt->vm.scratch_range = gen8_ggtt_clear_range; ggtt->vm.insert_entries = gen8_ggtt_insert_entries; + ggtt->vm.read_entry = gen8_ggtt_read_entry; /* * Serialize GTT updates with aperture access on BXT if VT-d is on, @@ -1291,6 +1334,8 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt) else ggtt->vm.pte_encode = gen8_ggtt_pte_encode; + ggtt->vm.pte_decode = gen8_ggtt_pte_decode; + return ggtt_probe_common(ggtt, size); } @@ -1390,6 +1435,14 @@ static u64 iris_pte_encode(dma_addr_t addr, return pte; } +static dma_addr_t gen6_pte_decode(u64 pte, bool *is_present, bool *is_local) +{ + *is_present = pte & GEN6_PTE_VALID; + *is_local = false; + + return ((pte & 0xff0) << 28) | (pte & ~0xfff); +} + static int gen6_gmch_probe(struct i915_ggtt *ggtt) { struct drm_i915_private *i915 = ggtt->vm.i915; @@ -1428,6 +1481,7 @@ static int gen6_gmch_probe(struct i915_ggtt *ggtt) ggtt->vm.scratch_range = gen6_ggtt_clear_range; ggtt->vm.insert_page = gen6_ggtt_insert_page; ggtt->vm.insert_entries = gen6_ggtt_insert_entries; + ggtt->vm.read_entry = gen6_ggtt_read_entry; ggtt->vm.cleanup = gen6_gmch_remove; ggtt->invalidate = gen6_ggtt_invalidate; @@ -1443,6 +1497,8 @@ static int gen6_gmch_probe(struct i915_ggtt *ggtt) else ggtt->vm.pte_encode = snb_pte_encode; + ggtt->vm.pte_decode = gen6_pte_decode; + ggtt->vm.vma_ops.bind_vma = intel_ggtt_bind_vma; ggtt->vm.vma_ops.unbind_vma = intel_ggtt_unbind_vma; diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c index 0c723e7c71a2..889e61843ff3 100644 --- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c @@ -328,7 +328,7 @@ static bool fence_is_active(const struct i915_fence_reg *fence) static struct i915_fence_reg *fence_find(struct i915_ggtt *ggtt) { - struct intel_display *display = &ggtt->vm.i915->display; + struct intel_display *display = ggtt->vm.i915->display; struct i915_fence_reg *active = NULL; struct i915_fence_reg *fence, *fn; diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_gmch.c b/drivers/gpu/drm/i915/gt/intel_ggtt_gmch.c index 59eed0a0ce90..c5f5f0bdfb2c 100644 --- a/drivers/gpu/drm/i915/gt/intel_ggtt_gmch.c +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_gmch.c @@ -27,6 +27,13 @@ static void gmch_ggtt_insert_page(struct i915_address_space *vm, intel_gmch_gtt_insert_page(addr, offset >> PAGE_SHIFT, flags); } +static dma_addr_t gmch_ggtt_read_entry(struct i915_address_space *vm, + u64 offset, bool *is_present, bool *is_local) +{ + return intel_gmch_gtt_read_entry(offset >> PAGE_SHIFT, + is_present, is_local); +} + static void gmch_ggtt_insert_entries(struct i915_address_space *vm, struct i915_vma_resource *vma_res, unsigned int pat_index, @@ -103,6 +110,7 @@ int intel_ggtt_gmch_probe(struct i915_ggtt *ggtt) ggtt->vm.insert_entries = gmch_ggtt_insert_entries; ggtt->vm.clear_range = gmch_ggtt_clear_range; ggtt->vm.scratch_range = gmch_ggtt_clear_range; + ggtt->vm.read_entry = gmch_ggtt_read_entry; ggtt->vm.cleanup = gmch_ggtt_remove; ggtt->invalidate = gmch_ggtt_invalidate; diff --git a/drivers/gpu/drm/i915/gt/intel_gsc.c b/drivers/gpu/drm/i915/gt/intel_gsc.c index 1e925c75fb08..c43febc862dc 100644 --- a/drivers/gpu/drm/i915/gt/intel_gsc.c +++ b/drivers/gpu/drm/i915/gt/intel_gsc.c @@ -284,7 +284,7 @@ static void gsc_irq_handler(struct intel_gt *gt, unsigned int intf_id) if (gt->gsc.intf[intf_id].irq < 0) return; - ret = generic_handle_irq(gt->gsc.intf[intf_id].irq); + ret = generic_handle_irq_safe(gt->gsc.intf[intf_id].irq); if (ret) gt_err_ratelimited(gt, "error handling GSC irq: %d\n", ret); } diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm.c b/drivers/gpu/drm/i915/gt/intel_gt_pm.c index 3182f19b9837..c7f59d60fac6 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_pm.c +++ b/drivers/gpu/drm/i915/gt/intel_gt_pm.c @@ -6,6 +6,8 @@ #include <linux/string_helpers.h> #include <linux/suspend.h> +#include "display/intel_display_power.h" + #include "i915_drv.h" #include "i915_irq.h" #include "i915_params.h" @@ -70,7 +72,7 @@ static int __gt_unpark(struct intel_wakeref *wf) { struct intel_gt *gt = container_of(wf, typeof(*gt), wakeref); struct drm_i915_private *i915 = gt->i915; - struct intel_display *display = &i915->display; + struct intel_display *display = i915->display; GT_TRACE(gt, "\n"); @@ -104,7 +106,7 @@ static int __gt_park(struct intel_wakeref *wf) struct intel_gt *gt = container_of(wf, typeof(*gt), wakeref); intel_wakeref_t wakeref = fetch_and_zero(>->awake); struct drm_i915_private *i915 = gt->i915; - struct intel_display *display = &i915->display; + struct intel_display *display = i915->display; GT_TRACE(gt, "\n"); diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm_debugfs.c b/drivers/gpu/drm/i915/gt/intel_gt_pm_debugfs.c index b635aa2820d9..87ef85483bae 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_pm_debugfs.c +++ b/drivers/gpu/drm/i915/gt/intel_gt_pm_debugfs.c @@ -22,7 +22,7 @@ #include "intel_rps.h" #include "intel_runtime_pm.h" #include "intel_uncore.h" -#include "vlv_sideband.h" +#include "vlv_iosf_sb.h" void intel_gt_pm_debugfs_forcewake_user_open(struct intel_gt *gt) { @@ -366,9 +366,9 @@ void intel_gt_pm_frequency_dump(struct intel_gt *gt, struct drm_printer *p) drm_printf(p, "SW control enabled: %s\n", str_yes_no((rpmodectl & GEN6_RP_MEDIA_MODE_MASK) == GEN6_RP_MEDIA_SW_MODE)); - vlv_punit_get(i915); - freq_sts = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS); - vlv_punit_put(i915); + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); + freq_sts = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS); + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); drm_printf(p, "PUNIT_REG_GPU_FREQ_STS: 0x%08x\n", freq_sts); diff --git a/drivers/gpu/drm/i915/gt/intel_gt_requests.c b/drivers/gpu/drm/i915/gt/intel_gt_requests.c index d1a382dfaa1d..93298820bee2 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_requests.c +++ b/drivers/gpu/drm/i915/gt/intel_gt_requests.c @@ -250,11 +250,17 @@ void intel_gt_watchdog_work(struct work_struct *work) llist_for_each_entry_safe(rq, rn, first, watchdog.link) { if (!i915_request_completed(rq)) { struct dma_fence *f = &rq->fence; + const char __rcu *timeline; + const char __rcu *driver; + rcu_read_lock(); + driver = dma_fence_driver_name(f); + timeline = dma_fence_timeline_name(f); pr_notice("Fence expiration time out i915-%s:%s:%llx!\n", - f->ops->get_driver_name(f), - f->ops->get_timeline_name(f), + rcu_dereference(driver), + rcu_dereference(timeline), f->seqno); + rcu_read_unlock(); i915_request_cancel(rq, -EINTR); } i915_request_put(rq); diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.c b/drivers/gpu/drm/i915/gt/intel_gtt.c index 30b128b1fde7..afbc5c769308 100644 --- a/drivers/gpu/drm/i915/gt/intel_gtt.c +++ b/drivers/gpu/drm/i915/gt/intel_gtt.c @@ -176,7 +176,6 @@ static void clear_vm_list(struct list_head *list) i915_vma_destroy_locked(vma); i915_gem_object_put(obj); } - } } diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.h b/drivers/gpu/drm/i915/gt/intel_gtt.h index 0a36ea751b63..9d3a3ad567a0 100644 --- a/drivers/gpu/drm/i915/gt/intel_gtt.h +++ b/drivers/gpu/drm/i915/gt/intel_gtt.h @@ -312,6 +312,7 @@ struct i915_address_space { u64 (*pte_encode)(dma_addr_t addr, unsigned int pat_index, u32 flags); /* Create a valid PTE */ + dma_addr_t (*pte_decode)(u64 pte, bool *is_present, bool *is_local); #define PTE_READ_ONLY BIT(0) #define PTE_LM BIT(1) @@ -340,6 +341,8 @@ struct i915_address_space { struct i915_vma_resource *vma_res, unsigned int pat_index, u32 flags); + dma_addr_t (*read_entry)(struct i915_address_space *vm, + u64 offset, bool *is_present, bool *is_local); void (*cleanup)(struct i915_address_space *vm); void (*foreach)(struct i915_address_space *vm, @@ -590,6 +593,9 @@ void intel_ggtt_bind_vma(struct i915_address_space *vm, void intel_ggtt_unbind_vma(struct i915_address_space *vm, struct i915_vma_resource *vma_res); +dma_addr_t intel_ggtt_read_entry(struct i915_address_space *vm, + u64 offset, bool *is_present, bool *is_local); + int i915_ggtt_probe_hw(struct drm_i915_private *i915); int i915_ggtt_init_hw(struct drm_i915_private *i915); int i915_ggtt_enable_hw(struct drm_i915_private *i915); diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 51847a846002..c481b56fa67d 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -751,7 +751,6 @@ static int lrc_ring_indirect_offset(const struct intel_engine_cs *engine) static int lrc_ring_cmd_buf_cctl(const struct intel_engine_cs *engine) { - if (GRAPHICS_VER_FULL(engine->i915) >= IP_VER(12, 55)) /* * Note that the CSFE context has a dummy slot for CMD_BUF_CCTL diff --git a/drivers/gpu/drm/i915/gt/intel_mocs.c b/drivers/gpu/drm/i915/gt/intel_mocs.c index cf41d325712e..5dd8121f4b15 100644 --- a/drivers/gpu/drm/i915/gt/intel_mocs.c +++ b/drivers/gpu/drm/i915/gt/intel_mocs.c @@ -314,7 +314,6 @@ static const struct drm_i915_mocs_entry icl_mocs_table[] = { }; static const struct drm_i915_mocs_entry dg1_mocs_table[] = { - /* UC */ MOCS_ENTRY(1, 0, L3_1_UC), /* WB - L3 */ diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c index dbdcfe130ad4..4a1675dea1c7 100644 --- a/drivers/gpu/drm/i915/gt/intel_reset.c +++ b/drivers/gpu/drm/i915/gt/intel_reset.c @@ -1205,7 +1205,7 @@ void intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask, const char *reason) { - struct intel_display *display = >->i915->display; + struct intel_display *display = gt->i915->display; intel_engine_mask_t awake; int ret; @@ -1423,7 +1423,7 @@ static void intel_gt_reset_global(struct intel_gt *gt, /* Use a watchdog to ensure that our reset completes */ intel_wedge_on_timeout(&w, gt, 60 * HZ) { struct drm_i915_private *i915 = gt->i915; - struct intel_display *display = &i915->display; + struct intel_display *display = i915->display; bool need_display_reset; bool reset_display; @@ -1448,7 +1448,8 @@ static void intel_gt_reset_global(struct intel_gt *gt, kobject_uevent_env(kobj, KOBJ_CHANGE, reset_done_event); else drm_dev_wedged_event(>->i915->drm, - DRM_WEDGE_RECOVERY_REBIND | DRM_WEDGE_RECOVERY_BUS_RESET); + DRM_WEDGE_RECOVERY_REBIND | DRM_WEDGE_RECOVERY_BUS_RESET, + NULL); } /** diff --git a/drivers/gpu/drm/i915/gt/intel_ring_submission.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c index 6e9977b2d180..2a6d79abf25b 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring_submission.c +++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c @@ -365,7 +365,13 @@ static void reset_prepare(struct intel_engine_cs *engine) ENGINE_READ_FW(engine, RING_HEAD), ENGINE_READ_FW(engine, RING_TAIL), ENGINE_READ_FW(engine, RING_START)); - if (!stop_ring(engine)) { + /* + * Sometimes engine head failed to set to zero even after writing into it. + * Use wait_for_atomic() with 20ms delay to let engine resumes from + * correct RING_HEAD. Experimented different values and determined + * that 20ms works best based on testing. + */ + if (wait_for_atomic((!stop_ring(engine) == 0), 20)) { drm_err(&engine->i915->drm, "failed to set %s head to zero " "ctl %08x head %08x tail %08x start %08x\n", @@ -604,7 +610,6 @@ static int ring_context_alloc(struct intel_context *ce) /* One ringbuffer to rule them all */ GEM_BUG_ON(!engine->legacy.ring); ce->ring = engine->legacy.ring; - ce->timeline = intel_timeline_get(engine->legacy.timeline); GEM_BUG_ON(ce->state); if (engine->context_size) { @@ -617,6 +622,8 @@ static int ring_context_alloc(struct intel_context *ce) ce->state = vma; } + ce->timeline = intel_timeline_get(engine->legacy.timeline); + return 0; } diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c index 71ee01d9ef64..0b35fdd461d4 100644 --- a/drivers/gpu/drm/i915/gt/intel_rps.c +++ b/drivers/gpu/drm/i915/gt/intel_rps.c @@ -8,7 +8,7 @@ #include <drm/intel/i915_drm.h> #include "display/intel_display.h" -#include "display/intel_display_irq.h" +#include "display/intel_display_rps.h" #include "i915_drv.h" #include "i915_irq.h" #include "i915_reg.h" @@ -23,7 +23,7 @@ #include "intel_mchbar_regs.h" #include "intel_pcode.h" #include "intel_rps.h" -#include "vlv_sideband.h" +#include "vlv_iosf_sb.h" #include "../../../platform/x86/intel_ips.h" #define BUSY_MAX_EI 20u /* ms */ @@ -74,7 +74,7 @@ static void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val) static void rps_timer(struct timer_list *t) { - struct intel_rps *rps = from_timer(rps, t, timer); + struct intel_rps *rps = timer_container_of(rps, t, timer); struct intel_gt *gt = rps_to_gt(rps); struct intel_engine_cs *engine; ktime_t dt, last, timestamp; @@ -550,6 +550,7 @@ static unsigned int init_emon(struct intel_uncore *uncore) static bool gen5_rps_enable(struct intel_rps *rps) { struct drm_i915_private *i915 = rps_to_i915(rps); + struct intel_display *display = i915->display; struct intel_uncore *uncore = rps_to_uncore(rps); u8 fstart, vstart; u32 rgvmodectl; @@ -607,9 +608,7 @@ static bool gen5_rps_enable(struct intel_rps *rps) rps->ips.last_count2 = intel_uncore_read(uncore, GFXEC); rps->ips.last_time2 = ktime_get_raw_ns(); - spin_lock(&i915->irq_lock); - ilk_enable_display_irq(i915, DE_PCU_EVENT); - spin_unlock(&i915->irq_lock); + ilk_display_rps_enable(display); spin_unlock_irq(&mchdev_lock); @@ -621,14 +620,13 @@ static bool gen5_rps_enable(struct intel_rps *rps) static void gen5_rps_disable(struct intel_rps *rps) { struct drm_i915_private *i915 = rps_to_i915(rps); + struct intel_display *display = i915->display; struct intel_uncore *uncore = rps_to_uncore(rps); u16 rgvswctl; spin_lock_irq(&mchdev_lock); - spin_lock(&i915->irq_lock); - ilk_disable_display_irq(i915, DE_PCU_EVENT); - spin_unlock(&i915->irq_lock); + ilk_display_rps_disable(display); rgvswctl = intel_uncore_read16(uncore, MEMSWCTL); @@ -822,9 +820,9 @@ static int vlv_rps_set(struct intel_rps *rps, u8 val) struct drm_i915_private *i915 = rps_to_i915(rps); int err; - vlv_punit_get(i915); - err = vlv_punit_write(i915, PUNIT_REG_GPU_FREQ_REQ, val); - vlv_punit_put(i915); + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); + err = vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_REQ, val); + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); GT_TRACE(rps_to_gt(rps), "set val:%x, freq:%d\n", val, intel_gpu_freq(rps, val)); @@ -1282,7 +1280,7 @@ static int chv_rps_max_freq(struct intel_rps *rps) struct intel_gt *gt = rps_to_gt(rps); u32 val; - val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMAX_AT_VMAX_FUSE); switch (gt->info.sseu.eu_total) { case 8: @@ -1309,7 +1307,7 @@ static int chv_rps_rpe_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val; - val = vlv_punit_read(i915, PUNIT_GPU_DUTYCYCLE_REG); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_GPU_DUTYCYCLE_REG); val >>= PUNIT_GPU_DUTYCYCLE_RPE_FREQ_SHIFT; return val & PUNIT_GPU_DUTYCYCLE_RPE_FREQ_MASK; @@ -1320,7 +1318,7 @@ static int chv_rps_guar_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val; - val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMAX_AT_VMAX_FUSE); return val & FB_GFX_FREQ_FUSE_MASK; } @@ -1330,7 +1328,7 @@ static u32 chv_rps_min_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val; - val = vlv_punit_read(i915, FB_GFX_FMIN_AT_VMIN_FUSE); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, FB_GFX_FMIN_AT_VMIN_FUSE); val >>= FB_GFX_FMIN_AT_VMIN_FUSE_SHIFT; return val & FB_GFX_FREQ_FUSE_MASK; @@ -1364,14 +1362,14 @@ static bool chv_rps_enable(struct intel_rps *rps) GEN6_PM_RP_DOWN_TIMEOUT); /* Setting Fixed Bias */ - vlv_punit_get(i915); + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | CHV_BIAS_CPU_50_SOC_50; - vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val); + vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, VLV_TURBO_SOC_OVERRIDE, val); - val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS); - vlv_punit_put(i915); + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); /* RPS code assumes GPLL is used */ drm_WARN_ONCE(&i915->drm, (val & GPLLENABLE) == 0, @@ -1389,7 +1387,7 @@ static int vlv_rps_guar_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val, rp1; - val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FREQ_FUSE); rp1 = val & FB_GFX_FGUARANTEED_FREQ_FUSE_MASK; rp1 >>= FB_GFX_FGUARANTEED_FREQ_FUSE_SHIFT; @@ -1402,7 +1400,7 @@ static int vlv_rps_max_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val, rp0; - val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FREQ_FUSE); rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT; /* Clamp to max */ @@ -1416,9 +1414,9 @@ static int vlv_rps_rpe_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val, rpe; - val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_LO); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FMAX_FUSE_LO); rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT; - val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_HI); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_NC, IOSF_NC_FB_GFX_FMAX_FUSE_HI); rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5; return rpe; @@ -1429,7 +1427,7 @@ static int vlv_rps_min_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); u32 val; - val = vlv_punit_read(i915, PUNIT_REG_GPU_LFM) & 0xff; + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_LFM) & 0xff; /* * According to the BYT Punit GPU turbo HAS 1.1.6.3 the minimum value * for the minimum frequency in GPLL mode is 0xc1. Contrary to this on @@ -1465,15 +1463,15 @@ static bool vlv_rps_enable(struct intel_rps *rps) /* WaGsvRC0ResidencyMethod:vlv */ rps->pm_events = GEN6_PM_RP_UP_EI_EXPIRED; - vlv_punit_get(i915); + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); /* Setting Fixed Bias */ val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | VLV_BIAS_CPU_125_SOC_875; - vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val); + vlv_iosf_sb_write(&i915->drm, VLV_IOSF_SB_PUNIT, VLV_TURBO_SOC_OVERRIDE, val); - val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS); + val = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS); - vlv_punit_put(i915); + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); /* RPS code assumes GPLL is used */ drm_WARN_ONCE(&i915->drm, (val & GPLLENABLE) == 0, @@ -1686,7 +1684,7 @@ static void vlv_init_gpll_ref_freq(struct intel_rps *rps) struct drm_i915_private *i915 = rps_to_i915(rps); rps->gpll_ref_freq = - vlv_get_cck_clock(i915, "GPLL ref", + vlv_get_cck_clock(&i915->drm, "GPLL ref", CCK_GPLL_CLOCK_CONTROL, i915->czclk_freq); @@ -1698,7 +1696,7 @@ static void vlv_rps_init(struct intel_rps *rps) { struct drm_i915_private *i915 = rps_to_i915(rps); - vlv_iosf_sb_get(i915, + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT) | BIT(VLV_IOSF_SB_NC) | BIT(VLV_IOSF_SB_CCK)); @@ -1722,7 +1720,7 @@ static void vlv_rps_init(struct intel_rps *rps) drm_dbg(&i915->drm, "min GPU freq: %d MHz (%u)\n", intel_gpu_freq(rps, rps->min_freq), rps->min_freq); - vlv_iosf_sb_put(i915, + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT) | BIT(VLV_IOSF_SB_NC) | BIT(VLV_IOSF_SB_CCK)); @@ -1732,7 +1730,7 @@ static void chv_rps_init(struct intel_rps *rps) { struct drm_i915_private *i915 = rps_to_i915(rps); - vlv_iosf_sb_get(i915, + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT) | BIT(VLV_IOSF_SB_NC) | BIT(VLV_IOSF_SB_CCK)); @@ -1756,7 +1754,7 @@ static void chv_rps_init(struct intel_rps *rps) drm_dbg(&i915->drm, "min GPU freq: %d MHz (%u)\n", intel_gpu_freq(rps, rps->min_freq), rps->min_freq); - vlv_iosf_sb_put(i915, + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT) | BIT(VLV_IOSF_SB_NC) | BIT(VLV_IOSF_SB_CCK)); @@ -2121,9 +2119,9 @@ static u32 __read_cagf(struct intel_rps *rps, bool take_fw) } else if (GRAPHICS_VER(i915) >= 12) { r = GEN12_RPSTAT1; } else if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) { - vlv_punit_get(i915); - freq = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS); - vlv_punit_put(i915); + vlv_iosf_sb_get(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); + freq = vlv_iosf_sb_read(&i915->drm, VLV_IOSF_SB_PUNIT, PUNIT_REG_GPU_FREQ_STS); + vlv_iosf_sb_put(&i915->drm, BIT(VLV_IOSF_SB_PUNIT)); } else if (GRAPHICS_VER(i915) >= 6) { r = GEN6_RPSTAT1; } else { diff --git a/drivers/gpu/drm/i915/gt/intel_rps_types.h b/drivers/gpu/drm/i915/gt/intel_rps_types.h index 5135b90a2a40..ece445109305 100644 --- a/drivers/gpu/drm/i915/gt/intel_rps_types.h +++ b/drivers/gpu/drm/i915/gt/intel_rps_types.h @@ -57,7 +57,7 @@ struct intel_rps { /* * work, interrupts_enabled and pm_iir are protected by - * i915->irq_lock + * gt->irq_lock */ struct timer_list timer; struct work_struct work; diff --git a/drivers/gpu/drm/i915/gt/intel_wopcm.h b/drivers/gpu/drm/i915/gt/intel_wopcm.h index 17d6aa86008a..d2038b6de5e7 100644 --- a/drivers/gpu/drm/i915/gt/intel_wopcm.h +++ b/drivers/gpu/drm/i915/gt/intel_wopcm.h @@ -1,6 +1,5 @@ +/* SPDX-License-Identifier: MIT */ /* - * SPDX-License-Identifier: MIT - * * Copyright © 2017-2018 Intel Corporation */ diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index 116683ebe074..b37e400f74e5 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -156,7 +156,7 @@ static void _wa_add(struct i915_wa_list *wal, const struct i915_wa *wa) if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */ struct i915_wa *list; - list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa), + list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*list), GFP_KERNEL); if (!list) { drm_err(&i915->drm, "No space for workaround init!\n"); diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c index 64315b714743..79741f043f03 100644 --- a/drivers/gpu/drm/i915/gt/mock_engine.c +++ b/drivers/gpu/drm/i915/gt/mock_engine.c @@ -108,7 +108,7 @@ static void advance(struct i915_request *request) static void hw_delay_complete(struct timer_list *t) { - struct mock_engine *engine = from_timer(engine, t, hw_delay); + struct mock_engine *engine = timer_container_of(engine, t, hw_delay); struct i915_request *request; unsigned long flags; diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index 22e750108c5f..23f04f6f8fba 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -7,6 +7,7 @@ #include "gem/i915_gem_internal.h" +#include "i915_drv.h" #include "i915_selftest.h" #include "intel_engine_heartbeat.h" #include "intel_engine_pm.h" @@ -859,6 +860,14 @@ static int live_lrc_timestamp(void *arg) }; /* + * This test was designed to isolate a hardware bug. + * The bug was found and fixed in future generations but + * now the test pollutes our CI on previous generation. + */ + if (GRAPHICS_VER(gt->i915) == 12) + return 0; + + /* * We want to verify that the timestamp is saved and restore across * context switches and is monotonic. * diff --git a/drivers/gpu/drm/i915/gt/selftest_migrate.c b/drivers/gpu/drm/i915/gt/selftest_migrate.c index 401bee030dbc..54bc447efce0 100644 --- a/drivers/gpu/drm/i915/gt/selftest_migrate.c +++ b/drivers/gpu/drm/i915/gt/selftest_migrate.c @@ -537,7 +537,7 @@ struct spinner_timer { static void spinner_kill(struct timer_list *timer) { - struct spinner_timer *st = from_timer(st, timer, timer); + struct spinner_timer *st = timer_container_of(st, timer, timer); igt_spinner_end(&st->spin); pr_info("%s\n", __func__); @@ -661,7 +661,7 @@ static int live_emit_pte_full_ring(void *arg) out_rq: i915_request_add(rq); /* GEM_BUG_ON(rq->reserved_space > ring->space)? */ timer_delete_sync(&st.timer); - destroy_timer_on_stack(&st.timer); + timer_destroy_on_stack(&st.timer); out_unpin: intel_context_unpin(ce); out_put: diff --git a/drivers/gpu/drm/i915/gt/selftest_rc6.c b/drivers/gpu/drm/i915/gt/selftest_rc6.c index 908483ab0bc8..41716ed454b7 100644 --- a/drivers/gpu/drm/i915/gt/selftest_rc6.c +++ b/drivers/gpu/drm/i915/gt/selftest_rc6.c @@ -33,15 +33,22 @@ int live_rc6_manual(void *arg) { struct intel_gt *gt = arg; struct intel_rc6 *rc6 = >->rc6; - u64 rc0_power, rc6_power; + struct intel_rps *rps = >->rps; intel_wakeref_t wakeref; + u64 rc0_sample_energy[2]; + u64 rc6_sample_energy[2]; + u64 sleep_time = 1000; + u32 rc0_freq = 0; + u32 rc6_freq = 0; + u64 rc0_power; + u64 rc6_power; bool has_power; + u64 threshold; ktime_t dt; u64 res[2]; int err = 0; - u32 rc0_freq = 0; - u32 rc6_freq = 0; - struct intel_rps *rps = >->rps; + u64 diff; + /* * Our claim is that we can "encourage" the GPU to enter rc6 at will. @@ -60,14 +67,15 @@ int live_rc6_manual(void *arg) /* Force RC6 off for starters */ __intel_rc6_disable(rc6); - msleep(1); /* wakeup is not immediate, takes about 100us on icl */ + /* wakeup is not immediate, takes about 100us on icl */ + usleep_range(1000, 2000); res[0] = rc6_residency(rc6); dt = ktime_get(); - rc0_power = librapl_energy_uJ(); - msleep(1000); - rc0_power = librapl_energy_uJ() - rc0_power; + rc0_sample_energy[0] = librapl_energy_uJ(); + msleep(sleep_time); + rc0_sample_energy[1] = librapl_energy_uJ() - rc0_sample_energy[0]; dt = ktime_sub(ktime_get(), dt); res[1] = rc6_residency(rc6); rc0_freq = intel_rps_read_actual_frequency_fw(rps); @@ -79,11 +87,12 @@ int live_rc6_manual(void *arg) } if (has_power) { - rc0_power = div64_u64(NSEC_PER_SEC * rc0_power, + rc0_power = div64_u64(NSEC_PER_SEC * rc0_sample_energy[1], ktime_to_ns(dt)); + if (!rc0_power) { if (rc0_freq) - pr_debug("No power measured while in RC0! GPU Freq: %u in RC0\n", + pr_debug("No power measured while in RC0! GPU Freq: %uMHz in RC0\n", rc0_freq); else pr_err("No power and freq measured while in RC0\n"); @@ -98,10 +107,10 @@ int live_rc6_manual(void *arg) res[0] = rc6_residency(rc6); intel_uncore_forcewake_flush(rc6_to_uncore(rc6), FORCEWAKE_ALL); dt = ktime_get(); - rc6_power = librapl_energy_uJ(); - msleep(1000); + rc6_sample_energy[0] = librapl_energy_uJ(); + msleep(sleep_time); rc6_freq = intel_rps_read_actual_frequency_fw(rps); - rc6_power = librapl_energy_uJ() - rc6_power; + rc6_sample_energy[1] = librapl_energy_uJ() - rc6_sample_energy[0]; dt = ktime_sub(ktime_get(), dt); res[1] = rc6_residency(rc6); if (res[1] == res[0]) { @@ -113,13 +122,24 @@ int live_rc6_manual(void *arg) } if (has_power) { - rc6_power = div64_u64(NSEC_PER_SEC * rc6_power, + rc6_power = div64_u64(NSEC_PER_SEC * rc6_sample_energy[1], ktime_to_ns(dt)); - pr_info("GPU consumed %llduW in RC0 and %llduW in RC6\n", + pr_info("GPU consumed %lluuW in RC0 and %lluuW in RC6\n", rc0_power, rc6_power); + if (2 * rc6_power > rc0_power) { - pr_err("GPU leaked energy while in RC6! GPU Freq: %u in RC6 and %u in RC0\n", - rc6_freq, rc0_freq); + pr_err("GPU leaked energy while in RC6!\n" + "GPU Freq: %uMHz in RC6 and %uMHz in RC0\n" + "RC0 energy before & after sleep respectively: %lluuJ %lluuJ\n" + "RC6 energy before & after sleep respectively: %lluuJ %lluuJ\n", + rc6_freq, rc0_freq, rc0_sample_energy[0], rc0_sample_energy[1], + rc6_sample_energy[0], rc6_sample_energy[1]); + + diff = res[1] - res[0]; + threshold = (9 * NSEC_PER_MSEC * sleep_time) / 10; + if (diff < threshold) + pr_err("Did not enter RC6 properly, RC6 start residency=%lluns, RC6 end residency=%lluns\n", + res[0], res[1]); err = -EINVAL; goto out_unlock; } diff --git a/drivers/gpu/drm/i915/gt/selftest_tlb.c b/drivers/gpu/drm/i915/gt/selftest_tlb.c index 3941f2d6fa47..69ed946a39e5 100644 --- a/drivers/gpu/drm/i915/gt/selftest_tlb.c +++ b/drivers/gpu/drm/i915/gt/selftest_tlb.c @@ -143,7 +143,7 @@ pte_tlbinv(struct intel_context *ce, if (ce->engine->class == OTHER_CLASS) msleep(200); else - msleep(10); + usleep_range(10000, 20000); if (va == vb) { if (!i915_request_completed(rq)) { diff --git a/drivers/gpu/drm/i915/gt/uc/intel_gsc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_gsc_fw.c index 5dc0ccd07636..d550eb6edfb8 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_gsc_fw.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_gsc_fw.c @@ -230,7 +230,7 @@ int intel_gsc_fw_get_binary_info(struct intel_uc_fw *gsc_fw, const void *data, s gt_info(gt, "Invalid GSC firmware for MTL/ARL, got %d.%d.%d.%d but need 102.x.x.x", gsc->release.major, gsc->release.minor, gsc->release.patch, gsc->release.build); - return -EINVAL; + return -EINVAL; } if (min_ver.major) { diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c index 9df80c325fc1..f360f020d8f1 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c @@ -313,8 +313,13 @@ static u32 guc_ctl_wa_flags(struct intel_guc *guc) * * The same WA bit is used for both and 22011391025 is applicable to * all DG2. + * + * Platforms post DG2 prevent this issue in hardware by stalling + * submissions. With this flag GuC will schedule as to avoid such + * stalls. */ - if (IS_DG2(gt->i915)) + if (IS_DG2(gt->i915) || + (CCS_MASK(gt) && GRAPHICS_VER_FULL(gt->i915) >= IP_VER(12, 70))) flags |= GUC_WA_DUAL_QUEUE; /* Wa_22011802037: graphics version 11/12 */ diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index f8cb7c630d5b..127316d2c8aa 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -633,7 +633,7 @@ static int guc_submission_send_busy_loop(struct intel_guc *guc, atomic_inc(&guc->outstanding_submission_g2h); ret = intel_guc_send_busy_loop(guc, action, len, g2h_len_dw, loop); - if (ret) + if (ret && g2h_len_dw) atomic_dec(&guc->outstanding_submission_g2h); return ret; @@ -3443,18 +3443,29 @@ static inline int guc_lrc_desc_unpin(struct intel_context *ce) * GuC is active, lets destroy this context, but at this point we can still be racing * with suspend, so we undo everything if the H2G fails in deregister_context so * that GuC reset will find this context during clean up. + * + * There is a race condition where the reset code could have altered + * this context's state and done a wakeref put before we try to + * deregister it here. So check if the context is still set to be + * destroyed before undoing earlier changes, to avoid two wakeref puts + * on the same context. */ ret = deregister_context(ce, ce->guc_id.id); if (ret) { + bool pending_destroyed; spin_lock_irqsave(&ce->guc_state.lock, flags); - set_context_registered(ce); - clr_context_destroyed(ce); + pending_destroyed = context_destroyed(ce); + if (pending_destroyed) { + set_context_registered(ce); + clr_context_destroyed(ce); + } spin_unlock_irqrestore(&ce->guc_state.lock, flags); /* * As gt-pm is awake at function entry, intel_wakeref_put_async merely decrements * the wakeref immediately but per function spec usage call this after unlock. */ - intel_wakeref_put_async(>->wakeref); + if (pending_destroyed) + intel_wakeref_put_async(>->wakeref); } return ret; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c index ec33ad942115..e848a04a80dc 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c @@ -1116,7 +1116,7 @@ static int uc_fw_xfer(struct intel_uc_fw *uc_fw, u32 dst_offset, u32 dma_flags) _MASKED_BIT_ENABLE(dma_flags | START_DMA)); /* Wait for DMA to finish */ - ret = intel_wait_for_register_fw(uncore, DMA_CTRL, START_DMA, 0, 100); + ret = intel_wait_for_register_fw(uncore, DMA_CTRL, START_DMA, 0, 100, NULL); if (ret) gt_err(gt, "DMA for %s fw failed, DMA_CTRL=%u\n", intel_uc_fw_type_repr(uc_fw->type), |