Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Varun Gupta <varun.gupta@intel.com>
To: igt-dev@lists.freedesktop.org
Cc: nishit.sharma@intel.com
Subject: [PATCH i-g-t] tests/intel/xe_madvise: Drop assert on cleanup unbind in atomic subtests
Date: Tue, 12 May 2026 11:40:35 +0530	[thread overview]
Message-ID: <20260512061035.190653-1-varun.gupta@intel.com> (raw)

The __xe_vm_bind_assert() for UNMAP can sporadically fail with EBUSY
because GPU migration to VRAM may still hold pages when the cleanup
unbind is issued after exec completion.  The unbind before
xe_vm_destroy is best-effort cleanup, use __xe_vm_bind() (non-assert)
instead.

Fixes: 44dea557914c ("tests/intel/xe_madvise: Add atomic-device subtest")
Fixes: 2642f23c1dbe ("tests/intel/xe_madvise: Add atomic-global subtest")
Fixes: 6f88532e2fe2 ("tests/intel/xe_madvise: Add atomic-cpu subtest")

Reviewed-by: Nishit Sharma <nishit.sharma@intel.com>
Signed-off-by: Varun Gupta <varun.gupta@intel.com>
---
 tests/intel/xe_madvise.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/tests/intel/xe_madvise.c b/tests/intel/xe_madvise.c
index cdb115d7e..8f88de0de 100644
--- a/tests/intel/xe_madvise.c
+++ b/tests/intel/xe_madvise.c
@@ -872,8 +872,8 @@ static void test_atomic_device(int fd, struct drm_xe_engine_class_instance *eci)
 	igt_assert_eq(data->data, 1);
 
 	xe_exec_queue_destroy(fd, exec_queue);
-	__xe_vm_bind_assert(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
-			    DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0);
+	__xe_vm_bind(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
+		     DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0, 0);
 	free(data);
 	xe_vm_destroy(fd, vm);
 }
@@ -955,8 +955,8 @@ static void test_atomic_global(int fd, struct drm_xe_engine_class_instance *eci)
 	igt_assert_eq(data->data, n_cpu_ops + 1);
 
 	xe_exec_queue_destroy(fd, exec_queue);
-	__xe_vm_bind_assert(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
-			    DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0);
+	__xe_vm_bind(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
+		     DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0, 0);
 	free(data);
 	xe_vm_destroy(fd, vm);
 }
@@ -1028,8 +1028,8 @@ static void test_atomic_cpu(int fd, struct drm_xe_engine_class_instance *eci)
 	igt_assert_eq(data->data, 0);
 
 	xe_exec_queue_destroy(fd, exec_queue);
-	__xe_vm_bind_assert(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
-			    DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0);
+	__xe_vm_bind(fd, vm, 0, 0, 0, 0, 0x1ull << va_bits,
+		     DRM_XE_VM_BIND_OP_UNMAP, 0, NULL, 0, 0, 0, 0);
 	free(data);
 	xe_vm_destroy(fd, vm);
 }
-- 
2.43.0


             reply	other threads:[~2026-05-12  6:11 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-12  6:10 Varun Gupta [this message]
2026-05-12 12:28 ` ✗ i915.CI.BAT: failure for tests/intel/xe_madvise: Drop assert on cleanup unbind in atomic subtests Patchwork
2026-05-12 13:01 ` ✓ Xe.CI.BAT: success " Patchwork
2026-05-12 23:15 ` ✗ Xe.CI.FULL: failure " Patchwork
2026-05-14  3:21   ` Gupta, Varun

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260512061035.190653-1-varun.gupta@intel.com \
    --to=varun.gupta@intel.com \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=nishit.sharma@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox