linux/drivers/gpu/drm/drm_atomic.c, branch v4.3

drm/atomic: Fix bookkeeping with TEST_ONLY, v3.

2015-09-01T08:57:06Z

Commit ec9f932ed41622d120de52a5b525e4d77b9ef17e "drm/atomic: Cleanup on error properly in the atomic ioctl." cleaned up some error paths, but didn't fix the TEST_ONLY path. In the check only case plane->fb shouldn't be updated, and the vblank events should be cleared as on failure. Changes since v1: - Fix -EDEADLK handling of vblank events too. - Free state last with CHECK_ONLY. Changes since v2: - Add comment about freeing crtc_state->event with TEST_ONLY. (Daniel Stone) Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Stone Signed-off-by: Jani Nikula

Revert "drm/atomic: Call ww_acquire_done after check phase is complete"

2015-08-25T07:23:36Z

This reverts commit 992cbf19b32900efa17850b9fa0031fd623edd4d. Until we make fbdev layer atomic we can't call this. Requested-by: Maarten Lankhorst Signed-off-by: Dave Airlie

drm/atomic: Call ww_acquire_done after check phase is complete

2015-08-11T10:04:18Z

We want to make sure that no one tries to acquire more locks and states, and ww mutexes provide debug facilities for that. So use them. v2: Only call acquire_done when ->atomic_check was successful to avoid falling over an -EDEADLK (spotted by Maarten). Cc: Rob Clark Cc: Maarten Lankhorst Reviewed-by: Maarten Lankhorst Signed-off-by: Daniel Vetter

drm/atomic: Paper over locking WARN in default_state_clear

2015-08-11T10:04:17Z

In commit 6f75cea66c8dd043ced282016b21a639af176642 Author: Daniel Vetter Date: Wed Nov 19 18:38:07 2014 +0100 drm/atomic: Only destroy connector states with connection mutex held I tried to fix races of atomic commits against connector hot-unplugging. The idea is to ensure lifetimes by holding the connection_mutex long enough. That works for synchronous commits, but not for async ones. For async atomic commit we really need to fix up connector lifetimes for real. But that's a much bigger task, so just add more duct-tape: For cleaning up connector states we currently don't need the connector itself. So NULL it out and remove the locking check. Of course that check was to protect the entire sequence, but the modeset itself should be save since currently DP MST hot-removal does a dpms-off. And that should synchronize with any outstanding async atomic commit. Or at least that's my hope, this is all a giant mess. Reported-by: Maarten Lankhorst Cc: Maarten Lankhorst Reviewed-by: Maarten Lankhorst Signed-off-by: Daniel Vetter

Merge tag 'topic/connector-locking-2015-07-23' of git://anongit.freedesktop.org/drm-intel into drm-next

2015-07-24T04:30:29Z

connector hotplug locking cleanup and fixes to make it save against atomic. Note that because of depencies this is based on top of the drm-intel-next pull, so that one needs to go in before this one. I've also thrown in the mode_group removal on top since it's defunct, never worked really, no one seems to care and the code can be resurrected easily. * tag 'topic/connector-locking-2015-07-23' of git://anongit.freedesktop.org/drm-intel: drm: gc now dead mode_group code drm: Stop filtering according to mode_group in getresources drm: Roll out drm_for_each_{plane,crtc,encoder} drm/cma-helper: Fix locking in drm_fb_cma_debugfs_show drm: Roll out drm_for_each_connector more drm: Amend connector list locking rules drm/radeon: Take all modeset locks for DP MST hotplug drm/i915: Take all modeset locks for DP MST hotplug drm: Check locking in drm_for_each_fb drm/i915: Use drm_for_each_fb in i915_debugfs.c drm: Check locking in drm_for_each_connector drm/fbdev-helper: Grab mode_config.mutex in drm_fb_helper_single_add_all_connectors drm/probe-helper: Grab mode_config.mutex in poll_init/enable drm: Add modeset object iterators drm: Simplify drm_for_each_legacy_plane arguments

drm: Roll out drm_for_each_connector more

2015-07-22T14:25:47Z

Now that we also grab the connection_mutex and so fixed the race with atomic modeset we can use the iterator there too. The other special case is drm_connector_unplug_all which would have a locking inversion with the sysfs store/show functions if we'd grab the mode_config.mutex around the unplug. We could just grab connection_mutex instead, but that's a bit too much a dirty trick for my taste. Also it's only used by udl, which doesn't do any other kind of connector hotplugging, so should be race-free. Hence just stick with a comment for now. Reviewed-by: Maarten Lankhorst Signed-off-by: Daniel Vetter

drm/atomic: Cleanup on error properly in the atomic ioctl.

2015-07-07T06:58:19Z

It's probably allowed to leave old_fb set to garbage when unlocking, but to prevent undefined behavior unset it just in case. Also crtc_state->event could be NULL on memory allocation failure, in which case event_space is increased for no reason. Note: Contains some general simplification of the cleanup code too. Signed-off-by: Maarten Lankhorst [danvet: Add note about the other changes in here. And fix long line while at it.] Signed-off-by: Daniel Vetter

drm/atomic: Update old_fb after setting a property.

2015-07-02T15:00:49Z

This change updates the old_fb pointer only after acquiring the plane lock, if there are no properties the fb cannot have been changed either, so this works out correctly. Found in a discussion with Rob Clark. Cc: Rob Clark Signed-off-by: Maarten Lankhorst Reviewed-by: Rob Clark Signed-off-by: Daniel Vetter

drm/atomic: Extract needs_modeset function

2015-06-19T15:25:01Z

We use the same check already in the atomic core, so might as well make this official. And it's also reused in e.g. i915. Motivated by Maarten's idea to extract a connector_changed state out of mode_changed. Cc: Maarten Lankhorst Reviewed-By: Maarten Lankhorst Signed-off-by: Daniel Vetter

drm/atomic: Add MODE_ID property

2015-05-26T13:50:42Z

Atomic modesetting: now with modesetting support. v2: Moved drm_atomic_set_mode_prop_for_crtc from previous patch; removed state->active fiddling, documented return code. Changed property type to DRM_MODE_PROP_BLOB. Signed-off-by: Daniel Stone Tested-by: Sean Paul Signed-off-by: Daniel Vetter