[Intel-gfx] [PATCH i-g-t 1/3] tests/gem_userptr_blits: More effectively set pages before invalidation

Janusz Krzysztofik janusz.krzysztofik at linux.intel.com
Wed Mar 4 09:58:39 UTC 2020


It has been observed that mmap-offset-invalidate at wb subtest has never
triggered a lockdep loop complain.  To fix it, don't use the ->domain
field of a mapping type being examined, always set read and write
domains to I915_GEM_DOMAIN_GTT instead.

Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik at linux.intel.com>
---
 tests/i915/gem_userptr_blits.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/i915/gem_userptr_blits.c b/tests/i915/gem_userptr_blits.c
index cd1a3af27..8148d7d76 100644
--- a/tests/i915/gem_userptr_blits.c
+++ b/tests/i915/gem_userptr_blits.c
@@ -821,7 +821,7 @@ static void test_mmap_offset_invalidate(int fd, const struct mmap_offset *t)
 	igt_require_f(map, "mmap-offset banned, lockdep loop prevention\n");
 
 	/* set object pages in order to activate MMU notifier for it */
-	gem_set_domain(fd, handle, t->domain, t->domain);
+	gem_set_domain(fd, handle, I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT);
 
 	/* trigger the notifier */
 	munmap(ptr, PAGE_SIZE);
-- 
2.21.1



More information about the Intel-gfx mailing list