drm/radeon: make 64bit fences more robust v3

Only increase the higher 32bits if we really detect a wrap around.

v2: instead of increasing the higher 32bits just use the higher
    32bits from the last emitted fence.
v3: also use last emitted fence value as upper limit.

The intention of this patch is to make fences as robust as
they where before introducing 64bit fences. This is
necessary because on older systems it looks like the fence
value gets corrupted on initialization.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=51344

Should also fix:
https://bugs.freedesktop.org/show_bug.cgi?id=54129
https://bugs.freedesktop.org/show_bug.cgi?id=54662
https://bugzilla.redhat.com/show_bug.cgi?id=846505
https://bugzilla.redhat.com/show_bug.cgi?id=845639

3.5 needs a separate patch due to changes in the
fence code.  Will send that out separately.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/radeon/radeon_fence.c

@@ -131,7 +131,7 @@ int radeon_fence_emit(struct radeon_device *rdev,
  */
 void radeon_fence_process(struct radeon_device *rdev, int ring)
 {
-	uint64_t seq, last_seq;
+	uint64_t seq, last_seq, last_emitted;
 	unsigned count_loop = 0;
 	bool wake = false;
 
@@ -158,13 +158,15 @@ void radeon_fence_process(struct radeon_device *rdev, int ring)
 	 */
 	last_seq = atomic64_read(&rdev->fence_drv[ring].last_seq);
 	do {
+		last_emitted = rdev->fence_drv[ring].sync_seq[ring];
 		seq = radeon_fence_read(rdev, ring);
 		seq |= last_seq & 0xffffffff00000000LL;
 		if (seq < last_seq) {
-			seq += 0x100000000LL;
+			seq &= 0xffffffff;
+			seq |= last_emitted & 0xffffffff00000000LL;
 		}
 
-		if (seq == last_seq) {
+		if (seq <= last_seq || seq > last_emitted) {
 			break;
 		}
 		/* If we loop over we don't want to return without