drm/radeon: Avoid double gpu reset by adding a timeout on IB ring tests.

When the radeon driver resets a gpu, it attempts to test whether all the
rings can successfully handle an IB.  If these rings fail to respond, the
process will wait forever.  Another gpu reset can't happen at this point,
as the current reset holds a lock required to do so.  Instead, make all
the IB tests run with a timeout, so the system can attempt to recover
in this case.

While this doesn't fix the underlying issue with card resets failing, it
gives the system a higher chance of recovering.  These timeouts have been
confirmed to help both a Tathi and Hawaii card recover after a gpu reset.

This also adds a new function, radeon_fence_wait_timeout, that behaves like
fence_wait_timeout.  It is used instead of fence_wait_timeout as it continues
to work during a reset.  radeon_fence_wait is changed to be implemented
using this function.

V2:
 - Changed the timeout to 1s, as the default 10s from radeon_wait_timeout was
too long.  A timeout of 100ms was tested and found to be too short.
 - Changed radeon_fence_wait_timeout to behave more like fence_wait_timeout.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Matthew Dawson <matthew@mjdsystems.ca>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

1 files changed, 32 insertions, 8 deletions

diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index 05815c47b246..7ef075acde9c 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -527,7 +527,7 @@ static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
 }
 
 /**
- * radeon_fence_wait - wait for a fence to signal
+ * radeon_fence_wait_timeout - wait for a fence to signal with timeout
  *
  * @fence: radeon fence object
  * @intr: use interruptible sleep
@@ -535,12 +535,15 @@ static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
  * Wait for the requested fence to signal (all asics).
  * @intr selects whether to use interruptable (true) or non-interruptable
  * (false) sleep when waiting for the fence.
- * Returns 0 if the fence has passed, error for all other cases.
+ * @timeout: maximum time to wait, or MAX_SCHEDULE_TIMEOUT for infinite wait
+ * Returns remaining time if the sequence number has passed, 0 when
+ * the wait timeout, or an error for all other cases.
  */
-int radeon_fence_wait(struct radeon_fence *fence, bool intr)
+long radeon_fence_wait_timeout(struct radeon_fence *fence, bool intr, long timeout)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
 	long r;
+	int r_sig;
 
 	/*
 	 * This function should not be called on !radeon fences.
@@ -552,15 +555,36 @@ int radeon_fence_wait(struct radeon_fence *fence, bool intr)
 		return fence_wait(&fence->base, intr);
 
 	seq[fence->ring] = fence->seq;
-	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
-	if (r < 0) {
+	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, timeout);
+	if (r <= 0) {
 		return r;
 	}
 
-	r = fence_signal(&fence->base);
-	if (!r)
+	r_sig = fence_signal(&fence->base);
+	if (!r_sig)
 		FENCE_TRACE(&fence->base, "signaled from fence_wait\n");
-	return 0;
+	return r;
+}
+
+/**
+ * radeon_fence_wait - wait for a fence to signal
+ *
+ * @fence: radeon fence object
+ * @intr: use interruptible sleep
+ *
+ * Wait for the requested fence to signal (all asics).
+ * @intr selects whether to use interruptable (true) or non-interruptable
+ * (false) sleep when waiting for the fence.
+ * Returns 0 if the fence has passed, error for all other cases.
+ */
+int radeon_fence_wait(struct radeon_fence *fence, bool intr)
+{
+	long r = radeon_fence_wait_timeout(fence, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r > 0) {
+		return 0;
+	} else {
+		return r;
+	}
 }
 
 /**