From d72d908b5619896fe7832b0c4c408e664ffe4e69 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 4 Aug 2016 07:52:31 +0100 Subject: [PATCH] drm/i915: Mark up i915_gem_active for locking annotation The future annotations will track the locking used for access to ensure that it is always sufficient. We make the preparations now to present the API ahead and to make sure that GCC can eliminate the unused parameter. Before: 6298417 3619610 696320 10614347 a1f64b vmlinux After: 6298417 3619610 696320 10614347 a1f64b vmlinux (with i915 builtin) Signed-off-by: Chris Wilson Reviewed-by: Joonas Lahtinen Link: http://patchwork.freedesktop.org/patch/msgid/1470293567-10811-12-git-send-email-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_debugfs.c | 12 ++++-- drivers/gpu/drm/i915/i915_gem.c | 49 +++++++++++++++++-------- drivers/gpu/drm/i915/i915_gem_fence.c | 3 +- drivers/gpu/drm/i915/i915_gem_request.h | 36 +++++++++++------- drivers/gpu/drm/i915/i915_gem_tiling.c | 3 +- drivers/gpu/drm/i915/i915_gem_userptr.c | 3 +- drivers/gpu/drm/i915/i915_gpu_error.c | 29 ++++++++++++--- drivers/gpu/drm/i915/intel_display.c | 12 ++++-- 8 files changed, 101 insertions(+), 46 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 24ff7f4bc412..c595cc8169cb 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -155,10 +155,13 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj) obj->base.write_domain); for_each_engine_id(engine, dev_priv, id) seq_printf(m, "%x ", - i915_gem_active_get_seqno(&obj->last_read[id])); + i915_gem_active_get_seqno(&obj->last_read[id], + &obj->base.dev->struct_mutex)); seq_printf(m, "] %x %x%s%s%s", - i915_gem_active_get_seqno(&obj->last_write), - i915_gem_active_get_seqno(&obj->last_fence), + i915_gem_active_get_seqno(&obj->last_write, + &obj->base.dev->struct_mutex), + i915_gem_active_get_seqno(&obj->last_fence, + &obj->base.dev->struct_mutex), i915_cache_level_str(to_i915(obj->base.dev), obj->cache_level), obj->dirty ? " dirty" : "", obj->madv == I915_MADV_DONTNEED ? " purgeable" : ""); @@ -196,7 +199,8 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj) seq_printf(m, " (%s mappable)", s); } - engine = i915_gem_active_get_engine(&obj->last_write); + engine = i915_gem_active_get_engine(&obj->last_write, + &obj->base.dev->struct_mutex); if (engine) seq_printf(m, " (%s)", engine->name); diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 0c158c79f640..2b6199c53c80 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1354,21 +1354,24 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj, int ret, i; if (readonly) { - request = i915_gem_active_peek(&obj->last_write); + request = i915_gem_active_peek(&obj->last_write, + &obj->base.dev->struct_mutex); if (request) { ret = i915_wait_request(request); if (ret) return ret; i = request->engine->id; - if (i915_gem_active_peek(&obj->last_read[i]) == request) + if (i915_gem_active_peek(&obj->last_read[i], + &obj->base.dev->struct_mutex) == request) i915_gem_object_retire__read(obj, i); else i915_gem_object_retire__write(obj); } } else { for (i = 0; i < I915_NUM_ENGINES; i++) { - request = i915_gem_active_peek(&obj->last_read[i]); + request = i915_gem_active_peek(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (!request) continue; @@ -1400,9 +1403,11 @@ i915_gem_object_retire_request(struct drm_i915_gem_object *obj, { int idx = req->engine->id; - if (i915_gem_active_peek(&obj->last_read[idx]) == req) + if (i915_gem_active_peek(&obj->last_read[idx], + &obj->base.dev->struct_mutex) == req) i915_gem_object_retire__read(obj, idx); - else if (i915_gem_active_peek(&obj->last_write) == req) + else if (i915_gem_active_peek(&obj->last_write, + &obj->base.dev->struct_mutex) == req) i915_gem_object_retire__write(obj); if (!i915_reset_in_progress(&req->i915->gpu_error)) @@ -1431,7 +1436,8 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj, if (readonly) { struct drm_i915_gem_request *req; - req = i915_gem_active_get(&obj->last_write); + req = i915_gem_active_get(&obj->last_write, + &obj->base.dev->struct_mutex); if (req == NULL) return 0; @@ -1440,7 +1446,8 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj, for (i = 0; i < I915_NUM_ENGINES; i++) { struct drm_i915_gem_request *req; - req = i915_gem_active_get(&obj->last_read[i]); + req = i915_gem_active_get(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req == NULL) continue; @@ -2387,7 +2394,9 @@ static void i915_gem_object_retire__write(struct drm_i915_gem_object *obj) { GEM_BUG_ON(!i915_gem_active_isset(&obj->last_write)); - GEM_BUG_ON(!(obj->active & intel_engine_flag(i915_gem_active_get_engine(&obj->last_write)))); + GEM_BUG_ON(!(obj->active & + intel_engine_flag(i915_gem_active_get_engine(&obj->last_write, + &obj->base.dev->struct_mutex)))); i915_gem_active_set(&obj->last_write, NULL); intel_fb_obj_flush(obj, true, ORIGIN_CS); @@ -2405,7 +2414,8 @@ i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx) list_del_init(&obj->engine_list[idx]); i915_gem_active_set(&obj->last_read[idx], NULL); - engine = i915_gem_active_get_engine(&obj->last_write); + engine = i915_gem_active_get_engine(&obj->last_write, + &obj->base.dev->struct_mutex); if (engine && engine->id == idx) i915_gem_object_retire__write(obj); @@ -2626,7 +2636,8 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *engine) struct drm_i915_gem_object, engine_list[engine->id]); - if (!list_empty(&i915_gem_active_peek(&obj->last_read[engine->id])->list)) + if (!list_empty(&i915_gem_active_peek(&obj->last_read[engine->id], + &obj->base.dev->struct_mutex)->list)) break; i915_gem_object_retire__read(obj, engine->id); @@ -2759,7 +2770,8 @@ i915_gem_object_flush_active(struct drm_i915_gem_object *obj) for (i = 0; i < I915_NUM_ENGINES; i++) { struct drm_i915_gem_request *req; - req = i915_gem_active_peek(&obj->last_read[i]); + req = i915_gem_active_peek(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req == NULL) continue; @@ -2837,7 +2849,8 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) for (i = 0; i < I915_NUM_ENGINES; i++) { struct drm_i915_gem_request *req; - req = i915_gem_active_get(&obj->last_read[i]); + req = i915_gem_active_get(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req) requests[n++] = req; } @@ -2932,14 +2945,16 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj, if (readonly) { struct drm_i915_gem_request *req; - req = i915_gem_active_peek(&obj->last_write); + req = i915_gem_active_peek(&obj->last_write, + &obj->base.dev->struct_mutex); if (req) requests[n++] = req; } else { for (i = 0; i < I915_NUM_ENGINES; i++) { struct drm_i915_gem_request *req; - req = i915_gem_active_peek(&obj->last_read[i]); + req = i915_gem_active_peek(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req) requests[n++] = req; } @@ -4038,11 +4053,13 @@ i915_gem_busy_ioctl(struct drm_device *dev, void *data, int i; for (i = 0; i < I915_NUM_ENGINES; i++) { - req = i915_gem_active_peek(&obj->last_read[i]); + req = i915_gem_active_peek(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req) args->busy |= 1 << (16 + req->engine->exec_id); } - req = i915_gem_active_peek(&obj->last_write); + req = i915_gem_active_peek(&obj->last_write, + &obj->base.dev->struct_mutex); if (req) args->busy |= req->engine->exec_id; } diff --git a/drivers/gpu/drm/i915/i915_gem_fence.c b/drivers/gpu/drm/i915/i915_gem_fence.c index 9fdbd66128a6..a4ec4fecbea0 100644 --- a/drivers/gpu/drm/i915/i915_gem_fence.c +++ b/drivers/gpu/drm/i915/i915_gem_fence.c @@ -263,7 +263,8 @@ i915_gem_object_wait_fence(struct drm_i915_gem_object *obj) { int ret; - ret = i915_gem_active_wait(&obj->last_fence); + ret = i915_gem_active_wait(&obj->last_fence, + &obj->base.dev->struct_mutex); if (ret) return ret; diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h index e13834e2e3e0..5f8d94cb5f7d 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.h +++ b/drivers/gpu/drm/i915/i915_gem_request.h @@ -296,6 +296,12 @@ i915_gem_active_set(struct i915_gem_active *active, i915_gem_request_assign(&active->request, request); } +static inline struct drm_i915_gem_request * +__i915_gem_active_peek(const struct i915_gem_active *active) +{ + return active->request; +} + /** * i915_gem_active_peek - report the request being monitored * @active - the active tracker @@ -305,7 +311,7 @@ i915_gem_active_set(struct i915_gem_active *active, * caller must hold struct_mutex. */ static inline struct drm_i915_gem_request * -i915_gem_active_peek(const struct i915_gem_active *active) +i915_gem_active_peek(const struct i915_gem_active *active, struct mutex *mutex) { return active->request; } @@ -318,11 +324,11 @@ i915_gem_active_peek(const struct i915_gem_active *active) * if the active tracker is idle. The caller must hold struct_mutex. */ static inline struct drm_i915_gem_request * -i915_gem_active_get(const struct i915_gem_active *active) +i915_gem_active_get(const struct i915_gem_active *active, struct mutex *mutex) { struct drm_i915_gem_request *request; - request = i915_gem_active_peek(active); + request = i915_gem_active_peek(active, mutex); if (!request || i915_gem_request_completed(request)) return NULL; @@ -352,11 +358,12 @@ i915_gem_active_isset(const struct i915_gem_active *active) * the caller to hold struct_mutex (but that can be relaxed if desired). */ static inline bool -i915_gem_active_is_idle(const struct i915_gem_active *active) +i915_gem_active_is_idle(const struct i915_gem_active *active, + struct mutex *mutex) { struct drm_i915_gem_request *request; - request = i915_gem_active_peek(active); + request = i915_gem_active_peek(active, mutex); if (!request || i915_gem_request_completed(request)) return true; @@ -372,11 +379,11 @@ i915_gem_active_is_idle(const struct i915_gem_active *active) * retired first, see i915_gem_active_retire(). */ static inline int __must_check -i915_gem_active_wait(const struct i915_gem_active *active) +i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex) { struct drm_i915_gem_request *request; - request = i915_gem_active_peek(active); + request = i915_gem_active_peek(active, mutex); if (!request) return 0; @@ -393,9 +400,10 @@ i915_gem_active_wait(const struct i915_gem_active *active) * tracker is idle, the function returns immediately. */ static inline int __must_check -i915_gem_active_retire(const struct i915_gem_active *active) +i915_gem_active_retire(const struct i915_gem_active *active, + struct mutex *mutex) { - return i915_gem_active_wait(active); + return i915_gem_active_wait(active, mutex); } /* Convenience functions for peeking at state inside active's request whilst @@ -403,15 +411,17 @@ i915_gem_active_retire(const struct i915_gem_active *active) */ static inline uint32_t -i915_gem_active_get_seqno(const struct i915_gem_active *active) +i915_gem_active_get_seqno(const struct i915_gem_active *active, + struct mutex *mutex) { - return i915_gem_request_get_seqno(i915_gem_active_peek(active)); + return i915_gem_request_get_seqno(i915_gem_active_peek(active, mutex)); } static inline struct intel_engine_cs * -i915_gem_active_get_engine(const struct i915_gem_active *active) +i915_gem_active_get_engine(const struct i915_gem_active *active, + struct mutex *mutex) { - return i915_gem_request_get_engine(i915_gem_active_peek(active)); + return i915_gem_request_get_engine(i915_gem_active_peek(active, mutex)); } #define for_each_active(mask, idx) \ diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c index 8cef2d6b291a..fa2eb4a4f212 100644 --- a/drivers/gpu/drm/i915/i915_gem_tiling.c +++ b/drivers/gpu/drm/i915/i915_gem_tiling.c @@ -242,7 +242,8 @@ i915_gem_set_tiling(struct drm_device *dev, void *data, } obj->fence_dirty = - !i915_gem_active_is_idle(&obj->last_fence) || + !i915_gem_active_is_idle(&obj->last_fence, + &dev->struct_mutex) || obj->fence_reg != I915_FENCE_REG_NONE; obj->tiling_mode = args->tiling_mode; diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c index 00ab5e9d2eb7..e57521dbddc6 100644 --- a/drivers/gpu/drm/i915/i915_gem_userptr.c +++ b/drivers/gpu/drm/i915/i915_gem_userptr.c @@ -74,7 +74,8 @@ static void wait_rendering(struct drm_i915_gem_object *obj) for (i = 0; i < I915_NUM_ENGINES; i++) { struct drm_i915_gem_request *req; - req = i915_gem_active_get(&obj->last_read[i]); + req = i915_gem_active_get(&obj->last_read[i], + &obj->base.dev->struct_mutex); if (req) requests[n++] = req; } diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 585fe2bdadd5..0d2882ae7c89 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -742,18 +742,38 @@ unwind: #define i915_error_ggtt_object_create(dev_priv, src) \ i915_error_object_create((dev_priv), (src), &(dev_priv)->ggtt.base) +/* The error capture is special as tries to run underneath the normal + * locking rules - so we use the raw version of the i915_gem_active lookup. + */ +static inline uint32_t +__active_get_seqno(struct i915_gem_active *active) +{ + return i915_gem_request_get_seqno(__i915_gem_active_peek(active)); +} + +static inline int +__active_get_engine_id(struct i915_gem_active *active) +{ + struct intel_engine_cs *engine; + + engine = i915_gem_request_get_engine(__i915_gem_active_peek(active)); + return engine ? engine->id : -1; +} + static void capture_bo(struct drm_i915_error_buffer *err, struct i915_vma *vma) { struct drm_i915_gem_object *obj = vma->obj; - struct intel_engine_cs *engine; int i; err->size = obj->base.size; err->name = obj->base.name; + for (i = 0; i < I915_NUM_ENGINES; i++) - err->rseqno[i] = i915_gem_active_get_seqno(&obj->last_read[i]); - err->wseqno = i915_gem_active_get_seqno(&obj->last_write); + err->rseqno[i] = __active_get_seqno(&obj->last_read[i]); + err->wseqno = __active_get_seqno(&obj->last_write); + err->engine = __active_get_engine_id(&obj->last_write); + err->gtt_offset = vma->node.start; err->read_domains = obj->base.read_domains; err->write_domain = obj->base.write_domain; @@ -766,9 +786,6 @@ static void capture_bo(struct drm_i915_error_buffer *err, err->purgeable = obj->madv != I915_MADV_WILLNEED; err->userptr = obj->userptr.mm != NULL; err->cache_level = obj->cache_level; - - engine = i915_gem_active_get_engine(&obj->last_write); - err->engine = engine ? engine->id : -1; } static u32 capture_active_bo(struct drm_i915_error_buffer *err, diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 8c03d13d494d..d54a3ea56536 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11370,7 +11370,8 @@ static bool use_mmio_flip(struct intel_engine_cs *engine, if (resv && !reservation_object_test_signaled_rcu(resv, false)) return true; - return engine != i915_gem_active_get_engine(&obj->last_write); + return engine != i915_gem_active_get_engine(&obj->last_write, + &obj->base.dev->struct_mutex); } static void skl_do_mmio_flip(struct intel_crtc *intel_crtc, @@ -11673,7 +11674,8 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, } else if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) { engine = &dev_priv->engine[BCS]; } else if (INTEL_INFO(dev)->gen >= 7) { - engine = i915_gem_active_get_engine(&obj->last_write); + engine = i915_gem_active_get_engine(&obj->last_write, + &obj->base.dev->struct_mutex); if (engine == NULL || engine->id != RCS) engine = &dev_priv->engine[BCS]; } else { @@ -11694,7 +11696,8 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, if (mmio_flip) { INIT_WORK(&work->mmio_work, intel_mmio_flip_work_func); - work->flip_queued_req = i915_gem_active_get(&obj->last_write); + work->flip_queued_req = i915_gem_active_get(&obj->last_write, + &obj->base.dev->struct_mutex); schedule_work(&work->mmio_work); } else { request = i915_gem_request_alloc(engine, engine->last_context); @@ -14038,7 +14041,8 @@ intel_prepare_plane_fb(struct drm_plane *plane, if (ret == 0) { to_intel_plane_state(new_state)->wait_req = - i915_gem_active_get(&obj->last_write); + i915_gem_active_get(&obj->last_write, + &obj->base.dev->struct_mutex); } return ret; -- 2.20.1