This is a note to let you know that I've just added the patch titled drm/vc4: Fix non-blocking commit getting stuck forever to the 5.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: drm-vc4-fix-non-blocking-commit-getting-stuck-forever.patch and it can be found in the queue-5.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. >From 0c250c150c74a90db298bf2a8bcd0a1dabed2e2f Mon Sep 17 00:00:00 2001 From: Maxime Ripard <maxime@xxxxxxxxxx> Date: Mon, 25 Oct 2021 16:11:06 +0200 Subject: drm/vc4: Fix non-blocking commit getting stuck forever From: Maxime Ripard <maxime@xxxxxxxxxx> commit 0c250c150c74a90db298bf2a8bcd0a1dabed2e2f upstream. In some situation, we can end up being stuck on a non-blocking that went through properly. The situation that seems to trigger it reliably is to first start a non-blocking commit, and then right after, and before we had any vblank interrupt), start a blocking commit. This will lead to the first commit workqueue to be scheduled, setup the display, while the second commit is waiting for the first one to be completed. The vblank interrupt will then be raised, vc4_crtc_handle_vblank() will run and will compare the active dlist in the HVS channel to the one associated with the crtc->state. However, at that point, the second commit is waiting using drm_atomic_helper_wait_for_dependencies that occurs after drm_atomic_helper_swap_state has been called, so crtc->state points to the second commit state. vc4_crtc_handle_vblank() will compare the two dlist addresses and since they don't match will ignore the interrupt. The vblank event will never be reported, and the first and second commit will wait for the first commit completion until they timeout. The underlying reason is that it was never safe to do so. Indeed, accessing the ->state pointer access synchronization is based on ownership guarantees that can only occur within the functions and hooks defined as part of the KMS framework, and obviously the irq handler isn't one of them. The rework to move to generic helpers only uncovered the underlying issue. However, since the code path between drm_atomic_helper_wait_for_dependencies() and drm_atomic_helper_wait_for_vblanks() is serialised and we can't get two commits in that path at the same time, we can work around this issue by setting a variable associated to struct drm_crtc to the dlist we expect, and then using it from the vc4_crtc_handle_vblank() function. Since that state is shared with the modesetting path, we also need to introduce a spinlock to protect the code shared between the interrupt handler and the modesetting path, protecting only our new variable for now. Link: https://lore.kernel.org/all/YWgteNaNeaS9uWDe@phenom.ffwll.local/ Link: https://lore.kernel.org/r/20211025141113.702757-3-maxime@xxxxxxxxxx Fixes: 56d1fe0979dc ("drm/vc4: Make pageflip completion handling more robust.") Acked-by: Daniel Vetter <daniel.vetter@xxxxxxxx> Signed-off-by: Maxime Ripard <maxime@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- drivers/gpu/drm/vc4/vc4_crtc.c | 5 ++++- drivers/gpu/drm/vc4/vc4_drv.h | 14 ++++++++++++++ drivers/gpu/drm/vc4/vc4_hvs.c | 7 +++++-- 3 files changed, 23 insertions(+), 3 deletions(-) --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -713,8 +713,9 @@ static void vc4_crtc_handle_page_flip(st unsigned long flags; spin_lock_irqsave(&dev->event_lock, flags); + spin_lock(&vc4_crtc->irq_lock); if (vc4_crtc->event && - (vc4_state->mm.start == HVS_READ(SCALER_DISPLACTX(chan)) || + (vc4_crtc->current_dlist == HVS_READ(SCALER_DISPLACTX(chan)) || vc4_crtc->feeds_txp)) { drm_crtc_send_vblank_event(crtc, vc4_crtc->event); vc4_crtc->event = NULL; @@ -728,6 +729,7 @@ static void vc4_crtc_handle_page_flip(st */ vc4_hvs_unmask_underrun(dev, chan); } + spin_unlock(&vc4_crtc->irq_lock); spin_unlock_irqrestore(&dev->event_lock, flags); } @@ -1127,6 +1129,7 @@ int vc4_crtc_init(struct drm_device *drm return PTR_ERR(primary_plane); } + spin_lock_init(&vc4_crtc->irq_lock); drm_crtc_init_with_planes(drm, crtc, primary_plane, NULL, crtc_funcs, NULL); drm_crtc_helper_add(crtc, crtc_helper_funcs); --- a/drivers/gpu/drm/vc4/vc4_drv.h +++ b/drivers/gpu/drm/vc4/vc4_drv.h @@ -500,6 +500,20 @@ struct vc4_crtc { * @feeds_txp: True if the CRTC feeds our writeback controller. */ bool feeds_txp; + + /** + * @irq_lock: Spinlock protecting the resources shared between + * the atomic code and our vblank handler. + */ + spinlock_t irq_lock; + + /** + * @current_dlist: Start offset of the display list currently + * set in the HVS for that CRTC. Protected by @irq_lock, and + * copied in vc4_hvs_update_dlist() for the CRTC interrupt + * handler to have access to that value. + */ + unsigned int current_dlist; }; static inline struct vc4_crtc * --- a/drivers/gpu/drm/vc4/vc4_hvs.c +++ b/drivers/gpu/drm/vc4/vc4_hvs.c @@ -365,10 +365,9 @@ static void vc4_hvs_update_dlist(struct struct vc4_dev *vc4 = to_vc4_dev(dev); struct vc4_crtc *vc4_crtc = to_vc4_crtc(crtc); struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(crtc->state); + unsigned long flags; if (crtc->state->event) { - unsigned long flags; - crtc->state->event->pipe = drm_crtc_index(crtc); WARN_ON(drm_crtc_vblank_get(crtc) != 0); @@ -388,6 +387,10 @@ static void vc4_hvs_update_dlist(struct HVS_WRITE(SCALER_DISPLISTX(vc4_state->assigned_channel), vc4_state->mm.start); } + + spin_lock_irqsave(&vc4_crtc->irq_lock, flags); + vc4_crtc->current_dlist = vc4_state->mm.start; + spin_unlock_irqrestore(&vc4_crtc->irq_lock, flags); } void vc4_hvs_atomic_enable(struct drm_crtc *crtc, Patches currently in stable-queue which might be from maxime@xxxxxxxxxx are queue-5.15/clk-bcm-2835-pick-the-closest-clock-rate.patch queue-5.15/drm-vc4-hdmi-move-the-hsm-clock-enable-to-runtime_pm.patch queue-5.15/drm-vc4-hdmi-enable-the-scrambler-on-reconnection.patch queue-5.15/drm-vc4-hdmi-set-a-default-hsm-rate.patch queue-5.15/drm-vc4-fix-non-blocking-commit-getting-stuck-forever.patch queue-5.15/drm-vc4-crtc-make-sure-the-hdmi-controller-is-powere.patch queue-5.15/drm-vc4-crtc-drop-feed_txp-from-state.patch queue-5.15/drm-vc4-hdmi-make-sure-the-controller-is-powered-in-.patch queue-5.15/clk-bcm-2835-remove-rounding-up-the-dividers.patch queue-5.15/drm-vc4-crtc-copy-assigned-channel-to-the-crtc.patch queue-5.15/drm-vc4-hdmi-make-sure-the-device-is-powered-with-cec.patch queue-5.15/drm-vc4-hdmi-make-sure-the-controller-is-powered-up-.patch queue-5.15/drm-vc4-hdmi-rework-the-pre_crtc_configure-error-han.patch