From 18afa288920475a9f95d70920b0c54b43b5256a7 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 29 Mar 2017 10:47:46 +0100 Subject: [PATCH] Revert "drm/i915: Skip execlists_dequeue() early if the list is empty" MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit This reverts commit 6c943de6686f ("drm/i915: Skip execlists_dequeue() early if the list is empty"). The validity of using READ_ONCE there depends upon having a mb to coordinate the assignment of engine->execlist_first inside submit_request() and checking prior to taking the spinlock in execlists_dequeue(). We wrote "the update to TASKLET_SCHED incurs a memory barrier making this cross-cpu checking safe", but failed to notice that this mb was *conditional* on the execlists being ready, i.e. there wasn't the required mb when it was most necessary! We could install an unconditional memory barrier to fixup the READ_ONCE(): diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 7dd732cb9f57..1ed164b16d44 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -616,6 +616,7 @@ static void execlists_submit_request(struct drm_i915_gem_request *request) if (insert_request(&request->priotree, &engine->execlist_queue)) { engine->execlist_first = &request->priotree.node; + smp_wmb(); if (execlists_elsp_ready(engine)) But we have opted to remove the race as it should be rarely effective, and saves us having to explain the necessary memory barriers which we quite clearly failed at. Reported-and-tested-by: Tvrtko Ursulin Fixes: 6c943de6686f ("drm/i915: Skip execlists_dequeue() early if the list is empty") Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Cc: Joonas Lahtinen Cc: Michał Winiarski Link: http://patchwork.freedesktop.org/patch/msgid/20170329100052.29505-1-chris@chris-wilson.co.uk Reviewed-by: Tvrtko Ursulin --- drivers/gpu/drm/i915/i915_guc_submission.c | 12 ------------ drivers/gpu/drm/i915/intel_lrc.c | 12 ------------ 2 files changed, 24 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c index 6193ad7edcf4..1642fff9cf13 100644 --- a/drivers/gpu/drm/i915/i915_guc_submission.c +++ b/drivers/gpu/drm/i915/i915_guc_submission.c @@ -662,18 +662,6 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine) struct rb_node *rb; bool submit = false; - /* After execlist_first is updated, the tasklet will be rescheduled. - * - * If we are currently running (inside the tasklet) and a third - * party queues a request and so updates engine->execlist_first under - * the spinlock (which we have elided), it will atomically set the - * TASKLET_SCHED flag causing the us to be re-executed and pick up - * the change in state (the update to TASKLET_SCHED incurs a memory - * barrier making this cross-cpu checking safe). - */ - if (!READ_ONCE(engine->execlist_first)) - return false; - spin_lock_irq(&engine->timeline->lock); rb = engine->execlist_first; while (rb) { diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 12cf645793ed..c8f7c631fc1f 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -402,18 +402,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine) struct rb_node *rb; bool submit = false; - /* After execlist_first is updated, the tasklet will be rescheduled. - * - * If we are currently running (inside the tasklet) and a third - * party queues a request and so updates engine->execlist_first under - * the spinlock (which we have elided), it will atomically set the - * TASKLET_SCHED flag causing the us to be re-executed and pick up - * the change in state (the update to TASKLET_SCHED incurs a memory - * barrier making this cross-cpu checking safe). - */ - if (!READ_ONCE(engine->execlist_first)) - return; - last = port->request; if (last) /* WaIdleLiteRestore:bdw,skl -- 2.45.2