On Tue, Apr 16, 2024 at 12:07:25PM +0200, Thomas Hellström wrote:
> The -ENOSPC failure from ttm_bo_swapout() meant that the lru_lock
> was dropped and simply restarting the iteration meant we'd likely
> hit the same error again on the same resource. Now that we can
> restart the iteration even if the lock was dropped, do that.
>
It is not clear what you describe in this commit message (-ENOSPC ==
-EBUSY + lru_lock dropped) is true (no comments in code).
It does appears to be true after examining ttm_bo_swapout() closely.
Maybe out of scope for the series but would it be possible to add some
kernel doc to ttm_device_swapout stating this?
Patch it self makes sense to me.
Matt
> Cc: Christian König
> Cc: Somalapuram Amaranath
> Cc:
> Signed-off-by: Thomas Hellström
> ---
> drivers/gpu/drm/ttm/ttm_device.c | 21 +
> 1 file changed, 13 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/gpu/drm/ttm/ttm_device.c
> b/drivers/gpu/drm/ttm/ttm_device.c
> index e8a6a1dab669..4a030b4bc848 100644
> --- a/drivers/gpu/drm/ttm/ttm_device.c
> +++ b/drivers/gpu/drm/ttm/ttm_device.c
> @@ -168,15 +168,20 @@ int ttm_device_swapout(struct ttm_device *bdev, struct
> ttm_operation_ctx *ctx,
>
> num_pages = PFN_UP(bo->base.size);
> ret = ttm_bo_swapout(bo, ctx, gfp_flags);
> - /* ttm_bo_swapout has dropped the lru_lock */
> - if (!ret) {
> - ttm_resource_cursor_fini();
> - return num_pages;
> - }
> - if (ret != -EBUSY) {
> - ttm_resource_cursor_fini();
> - return ret;
> + /* Couldn't swap out, and retained the lru_lock */
> + if (ret == -EBUSY)
> + continue;
> + /* Couldn't swap out and dropped the lru_lock */
> + if (ret == -ENOSPC) {
> + spin_lock(>lru_lock);
> + continue;
> }
> + /*
> + * Dropped the lock and either succeeded or
> + * hit an error that forces us to break.
> + */
> + ttm_resource_cursor_fini();
> + return ret ? ret : num_pages;
> }
> }
> ttm_resource_cursor_fini_locked();
> --
> 2.44.0
>
