The cases of vmw_user_bo_synccpu_grab failing with -ERESTARTSYS or -EBUSY are handled in vmw_user_bo_synccpu_ioctl by not printing an error message. However, the error value gets discarded, indicating success. This leads to rendering glitches and a reported drm error on the next ioctl call to release the handle.
This patch propagates the error value from vmw_user_synccpu_grab.
Signed-off-by: Philipp Sieweck psi@informatik.uni-kiel.de --- drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c index 31aecc46624b..81fe4dc5e6ab 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c @@ -621,6 +621,8 @@ int vmw_user_bo_synccpu_ioctl(struct drm_device *dev, void *data, (unsigned int) arg->handle); return ret; } + if (unlikely(ret != 0)) + return ret; break; case drm_vmw_synccpu_release: ret = vmw_user_bo_synccpu_release(file_priv,
Hello Philipp,
On 3/13/22 06:06, Philipp Sieweck wrote:
The cases of vmw_user_bo_synccpu_grab failing with -ERESTARTSYS or -EBUSY are handled in vmw_user_bo_synccpu_ioctl by not printing an error message. However, the error value gets discarded, indicating success. This leads to rendering glitches and a reported drm error on the next ioctl call to release the handle.
This patch propagates the error value from vmw_user_synccpu_grab.
Signed-off-by: Philipp Sieweck psi@informatik.uni-kiel.de
drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 2 ++ 1 file changed, 2 insertions(+)
Patch looks good to me.
Reviewed-by: Javier Martinez Canillas javierm@redhat.com
On Sun, 2022-03-13 at 06:06 +0100, Philipp Sieweck wrote:
The cases of vmw_user_bo_synccpu_grab failing with -ERESTARTSYS or - EBUSY are handled in vmw_user_bo_synccpu_ioctl by not printing an error message. However, the error value gets discarded, indicating success. This leads to rendering glitches and a reported drm error on the next ioctl call to release the handle.
This patch propagates the error value from vmw_user_synccpu_grab.
Signed-off-by: Philipp Sieweck psi@informatik.uni-kiel.de
Hi, Philipp.
Thanks for the patch. Some questions below.
drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c index 31aecc46624b..81fe4dc5e6ab 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c @@ -621,6 +621,8 @@ int vmw_user_bo_synccpu_ioctl(struct drm_device *dev, void *data, (unsigned int) arg->handle); return ret; } + if (unlikely(ret != 0)) + return ret; break;
I'd just break apart the condition above rather than have two if ret != 0. What apps do you see glitches in as a result of this? Can you reproduce it?
z
Hi Zack!
I'd just break apart the condition above rather than have two if ret != 0. What apps do you see glitches in as a result of this? Can you reproduce it?
There are many apps I can use to trigger this drm error. It occurs more often on a Wayland-based desktop than on X11. While looking into this, the most reliable way to trigger it was to open the "About" dialog window in KeePassXC, then to select the contributors tab, and then to scroll a bit. The content of the scrolling area is often only partially updated, which makes the text unreadable.
I tested this on a Windows 10 host with an Intel Iris GPU running the latest VMware Workstation 16.2. The virtual machine contains a current Manjaro Linux GNOME desktop with Wayland enabled. 3d acceleration is active. However, this behavior has been there for quite some time, but for some reason it is more disruptive today.
Philipp
dri-devel@lists.freedesktop.org
-
Javier Martinez Canillas -
Philipp Sieweck -
Zack Rusin