Thank you for your contribution! Sashiko AI review found no new issues in this patch, but noticed 2 related pre-existing issues it does not introduce: - [High] drm/loongson: live state mutation in lsdc_cursor_plane_atomic_async_check() [1] - [Medium] drm/loongson: incomplete state copy in lsdc_cursor_plane_atomic_async_update() [2] --
commit 7aa4a6090dacda1cd98d23702b18772083dad3be Author: Ville Syrjälä <[email protected]> drm/atomic: s/async/bypass/ This commit renames the atomic plane async update helpers, hooks, and state flags to "bypass" across the DRM core and display drivers. [ ... ] > diff --git a/drivers/gpu/drm/loongson/lsdc_plane.c > b/drivers/gpu/drm/loongson/lsdc_plane.c > index c425288ba4909..59f85072273a6 100644 > --- a/drivers/gpu/drm/loongson/lsdc_plane.c > +++ b/drivers/gpu/drm/loongson/lsdc_plane.c > @@ -172,9 +172,9 @@ static const struct drm_plane_helper_funcs > lsdc_primary_helper_funcs = { > .atomic_disable = lsdc_primary_atomic_disable, > }; > > -static int lsdc_cursor_plane_atomic_async_check(struct drm_plane *plane, > - struct drm_atomic_commit *state, > - bool flip) > +static int lsdc_cursor_plane_atomic_bypass_check(struct drm_plane *plane, > + struct drm_atomic_commit > *state, > + bool flip) > { > struct drm_plane_state *new_state; > struct drm_crtc_state *crtc_state; [Severity: High] This problem wasn't introduced by this patch, but should lsdc_cursor_plane_atomic_bypass_check() pass new_state instead of plane->state to drm_atomic_helper_check_plane_state() [1]? drivers/gpu/drm/loongson/lsdc_plane.c:lsdc_cursor_plane_atomic_bypass_check() { ... return drm_atomic_helper_check_plane_state(plane->state, crtc_state, DRM_PLANE_NO_SCALING, DRM_PLANE_NO_SCALING, true, true); } drm_atomic_helper_check_plane_state() updates src, dst, and visible on the passed-in state, so passing plane->state mutates the live committed state during the check phase, racing with concurrent readers and corrupting active state if the check fails or runs as TEST_ONLY. > @@ -219,8 +219,8 @@ static int lsdc_cursor_plane_atomic_async_check(struct > drm_plane *plane, > true, true); > } > > -static void lsdc_cursor_plane_atomic_async_update(struct drm_plane *plane, > - struct drm_atomic_commit > *state) > +static void lsdc_cursor_plane_atomic_bypass_update(struct drm_plane *plane, > + struct drm_atomic_commit > *state) > { > struct lsdc_cursor *cursor = to_lsdc_cursor(plane); > const struct lsdc_cursor_plane_ops *ops = cursor->ops; [Severity: Medium] This problem wasn't introduced by this patch, but does lsdc_cursor_plane_atomic_bypass_update() leave plane->state with stale clipping rectangles and visibility [2]? drivers/gpu/drm/loongson/lsdc_plane.c:lsdc_cursor_plane_atomic_bypass_update() { ... plane->state->src_x = new_state->src_x; plane->state->src_y = new_state->src_y; plane->state->src_h = new_state->src_h; plane->state->src_w = new_state->src_w; swap(plane->state->fb, new_state->fb); ... } The in-place state update copies the integer coordinates and swaps fb, but never copies src, dst, or visible from new_state into plane->state, leaving them out of sync with the updated coordinates once the check phase stops mutating plane->state directly. [1] https://sashiko.dev/#/bug/linux-6b50e65f-5d91-4154-8ead-a7e41b3bb981 [2] https://sashiko.dev/#/bug/linux-39745fec-ddf4-4f7a-a3b2-188ca2cfda2a -- Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
