On Tue, 23 Jan 2024 12:20:15 +1000
Dave Airlie <airl...@gmail.com> wrote:
> From: Dave Airlie <airl...@redhat.com>
>
> ttm init was changed to look at numa node, but that needed a device.
>
> Only qxl appears to not hand in a device, so just fix that.
>
> Fixes: b0a7ce53d494 ("drm/ttm: Schedule delayed_delete worker closer")
> Reported-by: Steven Rostedt <rost...@goodmis.org>
My tests seem to get past the crash with this patch and not my "fix".
Tested-by: Steven Rostedt (Google) <rost...@goodmis.org>
-- Steve
> Cc: Rajneesh Bhardwaj <rajneesh.bhard...@amd.com>
> Signed-off-by: Dave Airlie <airl...@redhat.com>
> ---
> drivers/gpu/drm/qxl/qxl_ttm.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
> index 1a82629bce3f..1876419ea237 100644
> --- a/drivers/gpu/drm/qxl/qxl_ttm.c
> +++ b/drivers/gpu/drm/qxl/qxl_ttm.c
> @@ -195,7 +195,7 @@ int qxl_ttm_init(struct qxl_device *qdev)
> int num_io_pages; /* != rom->num_io_pages, we include surface0 */
>
> /* No others user of address space so set it to 0 */
> - r = ttm_device_init(&qdev->mman.bdev, &qxl_bo_driver, NULL,
> + r = ttm_device_init(&qdev->mman.bdev, &qxl_bo_driver, qdev->ddev.dev,
> qdev->ddev.anon_inode->i_mapping,
> qdev->ddev.vma_offset_manager,
> false, false);
