From: Chris Wilson Date: Mon, 3 Feb 2020 09:41:47 +0000 (+0000) Subject: drm/i915: Hold reference to previous active fence as we queue X-Git-Url: http://git.lede-project.org./?a=commitdiff_plain;h=30ca04e16cbeea7401dabc51e53b92975144e436;p=openwrt%2Fstaging%2Fblogic.git drm/i915: Hold reference to previous active fence as we queue Take a reference to the previous exclusive fence on the i915_active, as we wish to add an await to it in the caller (and so must prevent it from being freed until we have completed that task). Fixes: e3793468b466 ("drm/i915: Use the async worker to avoid reclaim tainting the ggtt->mutex") Signed-off-by: Chris Wilson Cc: Matthew Auld Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20200203094152.4150550-1-chris@chris-wilson.co.uk --- diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c index da58e5d084f4..9ccb931a733e 100644 --- a/drivers/gpu/drm/i915/i915_active.c +++ b/drivers/gpu/drm/i915/i915_active.c @@ -398,9 +398,13 @@ i915_active_set_exclusive(struct i915_active *ref, struct dma_fence *f) /* We expect the caller to manage the exclusive timeline ordering */ GEM_BUG_ON(i915_active_is_idle(ref)); + rcu_read_lock(); prev = __i915_active_fence_set(&ref->excl, f); - if (!prev) + if (prev) + prev = dma_fence_get_rcu(prev); + else atomic_inc(&ref->count); + rcu_read_unlock(); return prev; } diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index e801e28de470..74dc3ba59ce5 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -422,10 +422,12 @@ int i915_vma_bind(struct i915_vma *vma, * execution and not content or object's backing store lifetime. */ prev = i915_active_set_exclusive(&vma->active, &work->base.dma); - if (prev) + if (prev) { __i915_sw_fence_await_dma_fence(&work->base.chain, prev, &work->cb); + dma_fence_put(prev); + } work->base.dma.error = 0; /* enable the queue_work() */