While not good behaviour, it is, however, established behaviour that we
can punt EAGAIN to userspace if we need to retry the ioctl. When trying
to acquire a mutex, prefer to use EAGAIN to propagate losing the race
so that if it does end up back in userspace, we try again.
Fixes: c81471f5e95c ("drm/i915: Copy across scheduler behaviour flags across submit fences")
Closes: https://gitlab.freedesktop.org/drm/intel/issues/800
Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@xxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/gt/intel_timeline.c | 2 +-
drivers/gpu/drm/i915/i915_request.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline.c b/drivers/gpu/drm/i915/gt/intel_timeline.c
index 038e05a6336c..d71aafb66d6e 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline.c
+++ b/drivers/gpu/drm/i915/gt/intel_timeline.c
@@ -527,7 +527,7 @@ int intel_timeline_read_hwsp(struct i915_request *from,
GEM_BUG_ON(rcu_access_pointer(to->timeline) == tl);
- err = -EBUSY;
+ err = -EAGAIN;
if (mutex_trylock(&tl->mutex)) {
struct intel_timeline_cacheline *cl = from->hwsp_cacheline;
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 51bb8a0812a1..fb8738987aeb 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -783,7 +783,7 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
if (!tl) /* already started or maybe even completed */
return 0;
- fence = ERR_PTR(-EBUSY);
+ fence = ERR_PTR(-EAGAIN);
if (mutex_trylock(&tl->mutex)) {
fence = NULL;
if (!i915_request_started(signal) &&
--
2.24.0
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
