FAILED: patch "[PATCH] drm/amdgpu: keep the eviction fence until the last handle is" failed to apply to 6.18-stable tree

The patch below does not apply to the 6.18-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.18.y
git checkout FETCH_HEAD
git cherry-pick -x 3c44cbcec5eb370713bbd2c315ed7c34d8ccec2b
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2026102117-gaming-valid-6a28@gregkh' --subject-prefix 'PATCH 6.18.y' 'HEAD^..'

Possible dependencies:

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 3c44cbcec5eb370713bbd2c315ed7c34d8ccec2b Mon Sep 17 00:00:00 2001
From: Jesse Zhang <[email protected]>
Date: Mon, 12 Oct 2026 12:24:03 +0800
Subject: [PATCH] drm/amdgpu: keep the eviction fence until the last handle is
closed
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

amdgpu_gem_object_close() detaches the user queue eviction fence from the BO before it checks bo_va->ref_count, so closing any handle of a BO removes the fence, even when the file still has another handle to the BO and the bo_va with its GPU mappings stays. The BO can then be evicted without suspending the user queues of the file, which keep accessing it through mappings that are only marked for update.

Seen on navi31 with a SDMA user queue: a BO opened twice in the same file (GEM_FLINK + GEM_OPEN) loses its eviction fence when the second handle is closed (traced attach/detach of the fence):

attach bo=ffff8c78a6c84c00
GEM_CLOSE handle=7
detach bo=ffff8c78a6c84c00

while handle 6 and the mapping are still in use.

Only detach the fence when the last reference to the bo_va goes away, matching amdgpu_gem_object_open(), which attaches it only when the bo_va is created. With this the fence stays until the file is closed.

index ac7c8bb591eb..afd8502a2bc4 100644
@@ -337,11 +337,14 @@ static void amdgpu_gem_object_close(struct drm_gem_object *obj,
 			goto out_unlock;
 	}
 
+	bo_va = amdgpu_vm_bo_find(vm, bo);
+	/* the fence is attached once per bo_va, so keep it for other handles */
+	if (bo_va && --bo_va->ref_count)
+		goto out_unlock;
+
 	if (!amdgpu_vm_is_bo_always_valid(vm, bo))
 		amdgpu_evf_mgr_detach_fence(&fpriv->evf_mgr, bo);
-
-	bo_va = amdgpu_vm_bo_find(vm, bo);
-	if (!bo_va || --bo_va->ref_count)
+	if (!bo_va)
 		goto out_unlock;
 
 	amdgpu_vm_bo_del(adev, bo_va);

From: Jesse Zhang <[email protected]>

[ Upstream commit 3c44cbcec5eb370713bbd2c315ed7c34d8ccec2b ]

amdgpu_gem_object_close() detaches the user queue eviction fence from the BO before it checks bo_va->ref_count, so closing any handle of a BO removes the fence, even when the file still has another handle to the BO and the bo_va with its GPU mappings stays. The BO can then be evicted without suspending the user queues of the file, which keep accessing it through mappings that are only marked for update.

Seen on navi31 with a SDMA user queue: a BO opened twice in the same file (GEM_FLINK + GEM_OPEN) loses its eviction fence when the second handle is closed (traced attach/detach of the fence):

attach bo=ffff8c78a6c84c00
GEM_CLOSE handle=7
detach bo=ffff8c78a6c84c00

while handle 6 and the mapping are still in use.

Only detach the fence when the last reference to the bo_va goes away, matching amdgpu_gem_object_open(), which attaches it only when the bo_va is created. With this the fence stays until the file is closed.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

index 002d848d377f7..3b03157faa7e8 100644
@@ -325,11 +325,14 @@ static void amdgpu_gem_object_close(struct drm_gem_object *obj,
 			goto out_unlock;
 	}
 
+	bo_va = amdgpu_vm_bo_find(vm, bo);
+	/* the fence is attached once per bo_va, so keep it for other handles */
+	if (bo_va && --bo_va->ref_count)
+		goto out_unlock;
+
 	if (!amdgpu_vm_is_bo_always_valid(vm, bo))
 		amdgpu_eviction_fence_detach(&fpriv->evf_mgr, bo);
-
-	bo_va = amdgpu_vm_bo_find(vm, bo);
-	if (!bo_va || --bo_va->ref_count)
+	if (!bo_va)
 		goto out_unlock;
 
 	amdgpu_vm_bo_del(adev, bo_va);

2.53.0