kernel/include/drm/drm_atomic.h, branch linux-4.15.y

drm/atomic: Fix memleak on ERESTARTSYS during non-blocking commits

2018-02-28T09:21:38Z

commit 54f809cfbd6b4a43959039f5d33596ed3297ce16 upstream. During a non-blocking commit, it is possible to return before the commit_tail work is queued (-ERESTARTSYS, for example). Since a reference on the crtc commit object is obtained for the pending vblank event when preparing the commit, the above situation will leave us with an extra reference. Therefore, if the commit_tail worker has not consumed the event at the end of a commit, release it's reference. Changes since v1: - Also check for state->event->base.completion being set, to handle the case where stall_checks() fails in setup_crtc_commit(). Changes since v2: - Add a flag to drm_crtc_commit, to prevent dereferencing a freed event. i915 may unreference the state in a worker. Fixes: 24835e442f28 ("drm: reference count event->completion") Cc: # v4.11+ Signed-off-by: Leo (Sunpeng) Li Acked-by: Harry Wentland #v1 Signed-off-by: Maarten Lankhorst Link: https://patchwork.freedesktop.org/patch/msgid/20180117115108.29608-1-maarten.lankhorst@linux.intel.com Reviewed-by: Sean Paul Signed-off-by: Greg Kroah-Hartman

drm/atomic: Make atomic iterators less surprising

2017-10-06T09:06:11Z

Commit 669c9215afea ("drm/atomic: Make async plane update checks work as intended, v2.") assumed incorrectly that if only 1 plane is matched in the loop, the variables will be set to that plane. In reality we reset them to NULL every time a new plane was iterated. This behavior is surprising, so fix this by making the for loops only assign the variables on a match. When we have not added all the planes/crtc/connector to the state, and there's a few NULL ones after the last one we iterated, te assumption is broken that the pointers will hold the values from the last loop iteration, which holds true for all other for_each macros we're using. Except of course the iterator pointer itself, but that one really is entirely internal. Cc: Dmitry Osipenko Fixes: 669c9215afea ("drm/atomic: Make async plane update checks work as intended, v2.") Signed-off-by: Maarten Lankhorst Link: https://patchwork.freedesktop.org/patch/msgid/20170927083532.5756-2-maarten.lankhorst@linux.intel.com Reviewed-by: Daniel Vetter Tested-by: Dmitry Osipenko

drm/atomic: Remove unneeded null check for private objects

2017-10-06T09:05:50Z

It can be seen in drm_atomic_get_private_obj_state() that ptr will never be NULL, so skip the check for that case. Signed-off-by: Maarten Lankhorst Link: https://patchwork.freedesktop.org/patch/msgid/20170927083532.5756-1-maarten.lankhorst@linux.intel.com Reviewed-by: Daniel Vetter

drm/atomic: Remove deprecated accessor macros

2017-09-08T08:39:37Z

Now that the last users have been converted, we can finally get rid of for_each_obj_in_state, we have better macros to replace them with. Signed-off-by: Maarten Lankhorst Cc: Daniel Vetter Cc: Jani Nikula Cc: Sean Paul Cc: David Airlie Link: https://patchwork.freedesktop.org/patch/msgid/20170719143920.25685-8-maarten.lankhorst@linux.intel.com Reviewed-by: Daniel Vetter

drm/atomic: Fix freeing connector/plane state too early by tracking commits, v3.

2017-09-08T08:39:37Z

Currently we neatly track the crtc state, but forget to look at plane/connector state. When doing a nonblocking modeset, immediately followed by a setprop before the modeset completes, the setprop will see the modesets new state as the old state and free it. This has to be solved by waiting for hw_done on the connector, even if it's not assigned to a crtc. When a connector is unbound we take the last crtc commit, and when it stays unbound we create a new fake crtc commit for that gets signaled on hw_done for all the planes/connectors. We wait for it the same way as we do for crtc's, which will make sure we never run into a use-after-free situation. Changes since v1: - Only create a single disable commit. (danvet) - Fix leak in intel_legacy_cursor_update. Changes since v2: - Make reference counting in drm_atomic_helper_setup_commit more obvious. (pinchartl) - Call cleanup_done for fake commit. (danvet) - Add comments to drm_atomic_helper_setup_commit. (danvet, pinchartl) - Add comment to drm_atomic_helper_swap_state. (pinchartl) Signed-off-by: Maarten Lankhorst Testcase: kms_atomic_transition.plane-use-after-nonblocking-unbind* Cc: Laurent Pinchart Link: https://patchwork.freedesktop.org/patch/msgid/20170904104838.23822-6-maarten.lankhorst@linux.intel.com Reviewed-by: Daniel Vetter

drm/atomic: Move drm_crtc_commit to drm_crtc_state, v4.

2017-09-08T08:39:37Z

Most code only cares about the current commit or previous commit. Fortuantely we already have a place to track those. Move it to drm_crtc_state where it belongs. :) The per-crtc commit_list is kept for places where we have to look deeper than the current or previous commit for checking whether to stall on unpin. This is used in drm_atomic_helper_setup_commit and intel_has_pending_fb_unpin. Changes since v1: - Update kerneldoc for drm_crtc.commit_list. (danvet) Changes since v2: - Remove drm_atomic_helper_async_check hunk. (pinchartl) Changes since v3: - Fix use-after-free in drm_atomic_helper_commit_cleanup_done(). Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/20170904150456.31049-1-maarten.lankhorst@linux.intel.com [mlankhorst: preceeding -> preceding (checkpatch)]

drm/atomic: Return commit in drm_crtc_commit_get for better annotation

2017-09-08T08:39:37Z

This will allow code to do x->commit = drm_crtc_commit_get(commit), making it clearer where references are used. Signed-off-by: Maarten Lankhorst Link: https://patchwork.freedesktop.org/patch/msgid/20170904104838.23822-5-maarten.lankhorst@linux.intel.com Reviewed-by: Daniel Vetter

drm/atomic: Fix typo in kerneldoc

2017-09-01T17:51:42Z

The for_each_crtc_in_state() is used to iterate over CRTCs rather than connectors. Reviewed-by: Ville Syrjälä Signed-off-by: Thierry Reding Link: https://patchwork.freedesktop.org/patch/msgid/20170901144042.6023-2-thierry.reding@gmail.com

drm: Nuke drm_atomic_legacy_backoff

2017-08-08T12:49:29Z

Finally all users are gone! Reviewed-by: Maarten Lankhorst Signed-off-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/20170725080122.20548-9-daniel.vetter@ffwll.ch

drm: Nuke drm_atomic_helper_connector_set_property

2017-08-08T12:47:37Z

It's dead code, the core handles all this directly now. This also allows us to unexport drm_atomic_helper_connector_set_property. The only special case is nouveau which used one function for both pre-nv50 legacy modeset code and post-nv50 atomic world instead of 2 vtables. But amounts to exactly the same. What is rather strange here is how few drivers set this up, I suspect the earlier patch to handle properties in the core did end up fixing a pile of possible issues. Signed-off-by: Daniel Vetter Cc: Daniel Vetter Cc: Jani Nikula Cc: Sean Paul Cc: David Airlie Cc: Ben Skeggs Cc: Benjamin Gaignard Cc: Vincent Abriou Cc: Eric Anholt Cc: intel-gfx@lists.freedesktop.org Cc: nouveau@lists.freedesktop.org Reviewed-by: Maarten Lankhorst Link: https://patchwork.freedesktop.org/patch/msgid/20170725080122.20548-7-daniel.vetter@ffwll.ch Acked-by: Vincent Abriou