[Intel-gfx] [PATCH i-g-t] i915/gem_exec_schedule: Switch bach to gem_set_domain()

Chris Wilson chris at chris-wilson.co.uk
Sun Feb 17 18:16:19 UTC 2019


The write hazard lies extend also to the cache-dirty tracking; as we
purposefully do not tell the kernel we are writing to the bo, it fails
to note the CPU cache as dirty and so the gem_read() may not
sufficiently flush the caches prior to reading back from the GPU.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Antonio Argenziano <antonio.argenziano at intel.com>
---
 tests/i915/gem_exec_schedule.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tests/i915/gem_exec_schedule.c b/tests/i915/gem_exec_schedule.c
index 59102b6bc..a9383000a 100644
--- a/tests/i915/gem_exec_schedule.c
+++ b/tests/i915/gem_exec_schedule.c
@@ -54,7 +54,8 @@ uint32_t __sync_read_u32(int fd, uint32_t handle, uint64_t offset)
 {
 	uint32_t value;
 
-	gem_sync(fd, handle); /* No write hazard lies! */
+	gem_set_domain(fd, handle, /* No write hazard lies! */
+		       I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT);
 	gem_read(fd, handle, offset, &value, sizeof(value));
 
 	return value;
@@ -63,7 +64,8 @@ uint32_t __sync_read_u32(int fd, uint32_t handle, uint64_t offset)
 static inline
 void __sync_read_u32_count(int fd, uint32_t handle, uint32_t *dst, uint64_t size)
 {
-	gem_sync(fd, handle); /* No write hazard lies! */
+	gem_set_domain(fd, handle, /* No write hazard lies! */
+		       I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT);
 	gem_read(fd, handle, 0, dst, size);
 }
 
-- 
2.20.1



More information about the Intel-gfx mailing list