From: Andrey Grodzovsky Date: Tue, 20 Jun 2017 14:42:33 +0000 (-0400) Subject: drm/amd/display: Add global lock function. X-Git-Tag: v4.19~2140^2~23^2~402 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=03c8a5feeecf95ca1dfabb5b231016c0e76a187c;p=platform%2Fkernel%2Flinux-rpi.git drm/amd/display: Add global lock function. Signed-off-by: Andrey Grodzovsky Reviewed-by: Harry Wentland Signed-off-by: Alex Deucher --- diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_types.c index 83baa48..db38dd0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_types.c @@ -2890,6 +2890,49 @@ static enum surface_update_type amdgpu_dm_check_surfaces_update_type( return update_type; } +/*` + * Grabs all modesetting locks to serialize against any blocking commits, + * Waits for completion of all non blocking commits. + */ +static void aquire_global_lock( + struct drm_device *dev, + struct drm_atomic_state *state) +{ + struct drm_crtc *crtc; + struct drm_crtc_commit *commit; + long ret; + + /* Adding all modeset locks to aquire_ctx will + * ensure that when the framework release it the + * extra locks we are locking here will get released to + */ + drm_modeset_lock_all_ctx(dev, state->acquire_ctx); + + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + spin_lock(&crtc->commit_lock); + commit = list_first_entry_or_null(&crtc->commit_list, + struct drm_crtc_commit, commit_entry); + if (commit) + drm_crtc_commit_get(commit); + spin_unlock(&crtc->commit_lock); + + if (!commit) + continue; + + /* Make sure all pending HW programming completed and + * page flips done + */ + ret = wait_for_completion_timeout(&commit->hw_done, + 10*HZ); + ret = wait_for_completion_timeout(&commit->flip_done, + 10*HZ); + if (ret == 0) + DRM_ERROR("[CRTC:%d:%s] hw_done timed out\n", + crtc->base.id, crtc->name); + drm_crtc_commit_put(commit); + } +} + int amdgpu_dm_atomic_check(struct drm_device *dev, struct drm_atomic_state *state) { @@ -3146,7 +3189,7 @@ int amdgpu_dm_atomic_check(struct drm_device *dev, dc, set[i].surfaces, set[i].surface_count, - set[i].stream) > UPDATE_TYPE_MED) { + set[i].stream) > UPDATE_TYPE_FAST) { wait_for_prev_commits = true; break; } @@ -3160,25 +3203,14 @@ int amdgpu_dm_atomic_check(struct drm_device *dev, * For full updates case when * removing/adding/updateding streams on once CRTC while flipping * on another CRTC, - * Adding all current active CRTC's states to the atomic commit in - * amdgpu_dm_atomic_check will guarantee that any such full update commit + * acquiring global lock will guarantee that any such full + * update commit * will wait for completion of any outstanding flip using DRMs * synchronization events. */ - if (wait_for_prev_commits) { - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc); - struct drm_crtc_state *crtc_state; - - if (acrtc->stream) { - crtc_state = drm_atomic_get_crtc_state(state, crtc); - if (IS_ERR(crtc_state)) { - ret = PTR_ERR(crtc_state); - break; - } - } - } - } + if (wait_for_prev_commits) + aquire_global_lock(dev, state); + } if (context) {