Commit 3236f57a authored by Chris Wilson's avatar Chris Wilson Committed by Daniel Vetter
Browse files

drm/i915: Use a non-blocking wait for set-to-domain ioctl

The principal use for set-to-domain is for userspace to serialise
operations with a particular buffer, for example to maintain coherency
with a CPU map or to ratelimit its rendering by waiting on all previous
operations before continuing. As such we tend to hold the struct_mutex
for long periods during the synchronisation and so cause contention
issues with other users of the graphics device, even for independent
operations as memory management. An example is the contention between
compiz and X which causes jitter in the display and a drop in peak

The ultimate solution would be a set of fine grained locks and lockless
operations, but an intermediate step is to first attempt the
synchronisation for set-to-domain without holding the mutex. This
introduces a number of race conditions, so we limit it use to the ioctl
periphery where we have no dependent state and can safely complete with
a locked synchronisation afterwards.
Signed-off-by: default avatarChris Wilson <>
Signed-off-by: default avatarDaniel Vetter <>
parent b361237b
......@@ -1133,6 +1133,52 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
return 0;
/* A nonblocking variant of the above wait. This is a highly dangerous routine
* as the object state may change during this call.
static __must_check int
i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
bool readonly)
struct drm_device *dev = obj->;
struct drm_i915_private *dev_priv = dev->dev_private;
struct intel_ring_buffer *ring = obj->ring;
u32 seqno;
int ret;
seqno = readonly ? obj->last_write_seqno : obj->last_read_seqno;
if (seqno == 0)
return 0;
ret = i915_gem_check_wedge(dev_priv, true);
if (ret)
return ret;
ret = i915_gem_check_olr(ring, seqno);
if (ret)
return ret;
ret = __wait_seqno(ring, seqno, true, NULL);
/* Manually manage the write flush as we may have not yet
* retired the buffer.
if (obj->last_write_seqno &&
i915_seqno_passed(seqno, obj->last_write_seqno)) {
obj->last_write_seqno = 0;
obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
return ret;
* Called when user space prepares to use an object with the CPU, either
* through the mmap ioctl's mapping or a GTT mapping.
......@@ -1170,6 +1216,14 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
goto unlock;
/* Try to flush the object off the GPU without holding the lock.
* We will repeat the flush holding the lock in the normal manner
* to catch cases where we are gazumped.
ret = i915_gem_object_wait_rendering__nonblocking(obj, !write_domain);
if (ret)
goto unref;
if (read_domains & I915_GEM_DOMAIN_GTT) {
ret = i915_gem_object_set_to_gtt_domain(obj, write_domain != 0);
......@@ -1183,6 +1237,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
ret = i915_gem_object_set_to_cpu_domain(obj, write_domain != 0);
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment