kernel/linux.git/drivers/gpu/drm/rockchip, branch v6.18.40

drm/rockchip: Test for imported buffers with drm_gem_is_imported()

2026-07-24T14:15:43+00:00

[ Upstream commit 9fc0da81916250f343599a4dd259f097196bf0fb ] Instead of testing import_attach for imported GEM buffers, invoke drm_gem_is_imported() to do the test. The test itself does not change. Signed-off-by: Thomas Zimmermann Cc: Sandy Huang Cc: Heiko Stübner Cc: Andy Yan Cc: linux-rockchip@lists.infradead.org Fixes: b57aa47d39e9 ("drm/gem: Test for imported GEM buffers with helper") Closes: https://lore.kernel.org/dri-devel/38d09d34.4354.196379aa560.Coremail.andyshrk@163.com/ Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20260227133113.235940-11-tzimmermann@suse.de Signed-off-by: Sasha Levin

drm/rockchip: dw_dp: Fix null-ptr-deref in dw_dp_remove()

2026-07-24T14:15:43+00:00

[ Upstream commit 9456381d8b60bb7dd42f2f04afe5ee4ce6e0bc12 ] Attempting to access driver data in the platform driver ->remove() callback may lead to a null pointer dereference since there is no guaranty that the component ->bind() callback invoking platform_set_drvdata() was executed. A common scenario is when Rockchip DRM driver didn't manage to run component_bind_all() because of an (unrelated) error causing early return from rockchip_drm_bind(). Drop the unnecessary call to platform_get_drvdata() and, instead, reference the target device structure via platform_device. Fixes: d68ba7bac955 ("drm/rockchip: Add RK3588 DPTX output support") Signed-off-by: Cristian Ciocaltea Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20260310-drm-rk-fixes-v2-3-645ecfb43f49@collabora.com Signed-off-by: Sasha Levin

drm/rockchip: dw_dp: Switch to drmm_kzalloc()

2026-07-24T14:15:43+00:00

[ Upstream commit ed9da8d23020352ad24c528db09b5acdd78b81fd ] Driver makes use of drmm_encoder_init() to initialize the encoder and automatically handle the cleanup by registering drm_encoder_cleanup() with drmm_add_action(). However, the internal structure containing the encoder part gets allocated with devm_kzalloc(), which happens while component_bind_all() is being called from Rockchip DRM driver. The component framework further ensures it is deallocated as part of releasing all the resources claimed during bind, which is triggered from component_unbind_all(). When the reference to the DRM device gets eventually dropped via drm_dev_put() in rockchip_drm_unbind(), drmm_encoder_alloc_release() attempts to access the now released encoder structure, leading to use-after-free. Ensure driver's internal structure is still reachable on encoder cleanup by switching from a device-managed allocation to a drm-managed one. Fixes: d68ba7bac955 ("drm/rockchip: Add RK3588 DPTX output support") Signed-off-by: Cristian Ciocaltea Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20260310-drm-rk-fixes-v2-2-645ecfb43f49@collabora.com Signed-off-by: Sasha Levin

drm/rockchip: dw_hdmi_qp: Fix RK3576 HPD interrupt handling

2026-02-26T22:59:13+00:00

[ Upstream commit 5f7be8afc40c5ccf1be0410514703e50a49532c0 ] The threaded interrupt handler on RK3576 checks HPD IRQ status before deciding to continue with interrupt clearing and unmasking. This is not only redundant, since a similar verification has been already performed by the hard IRQ handler before masking the interrupt, but is also error prone, because it might happen that hardware clears the status register right after the masking operation completes, and before the threaded handler reads its value. The consequence is that HPD IRQ gets never unmasked, which breaks hotplug detection until reloading the driver or rebooting the system. Drop the unnecessary verification of the HPD interrupt status from the threaded interrupt handler. Fixes: 36439120efbd ("drm/rockchip: dw_hdmi_qp: Add basic RK3576 HDMI output support") Signed-off-by: Cristian Ciocaltea Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20260115-dw-hdmi-qp-hpd-v1-1-e59c166eaa65@collabora.com Signed-off-by: Sasha Levin

drm/rockchip: vop2: Only wait for changed layer cfg done when there is pending cfgdone bits

2026-01-23T10:21:14+00:00

[ Upstream commit 7f6721b767e219343cfe9a894f5bd869ff5b9d3a ] The write of cfgdone bits always done at .atomic_flush. When userspace makes plane zpos changes of two crtc within one commit, at the .atomic_begin stage, crtcN will never receive the "layer change cfg done" event of crtcM because crtcM has not yet written "cfgdone". So only wait when there is pending cfgdone bits to avoid long timeout. Fixes: 3e89a8c68354 ("drm/rockchip: vop2: Fix the update of LAYER/PORT select registers when there are multi display output on rk3588/rk3568") Signed-off-by: Andy Yan Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20250718064120.8811-2-andyshrk@163.com Signed-off-by: Sasha Levin

drm/rockchip: vop2: Add delay between poll registers

2026-01-23T10:21:14+00:00

[ Upstream commit 9fae82450d8a5f9c8fa016cd15186e975609b2ac ] According to the implementation of read_poll_timeout_atomic, if the delay time is 0, it will only use a simple loop based on timeout_us to decrement the count. Therefore, the final timeout time will differ significantly from the set timeout time. So, here we set a specific delay time to ensure that the calculation of the timeout duration is accurate. Fixes: 3e89a8c68354 ("drm/rockchip: vop2: Fix the update of LAYER/PORT select registers when there are multi display output on rk3588/rk3568") Signed-off-by: Andy Yan Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20250718064120.8811-1-andyshrk@163.com Signed-off-by: Sasha Levin

drm/bridge: dw-hdmi-qp: Fix spurious IRQ on resume

2026-01-23T10:21:14+00:00

[ Upstream commit 14adddc65340f2034751c95616861c0e888e2bb1 ] After resume from suspend to RAM, the following splash is generated if the HDMI driver is probed (independent of a connected cable): [ 1194.484052] irq 80: nobody cared (try booting with the "irqpoll" option) [ 1194.484074] CPU: 0 UID: 0 PID: 627 Comm: rtcwake Not tainted 6.17.0-rc7-g96f1a11414b3 #1 PREEMPT [ 1194.484082] Hardware name: Rockchip RK3576 EVB V10 Board (DT) [ 1194.484085] Call trace: [ 1194.484087] ... (stripped) [ 1194.484283] handlers: [ 1194.484285] [<00000000bc363dcb>] dw_hdmi_qp_main_hardirq [dw_hdmi_qp] [ 1194.484302] Disabling IRQ #80 Apparently the HDMI IP is losing part of its state while the system is suspended and generates spurious interrupts during resume. The bug has not yet been noticed, as system suspend does not yet work properly on upstream kernel with either the Rockchip RK3588 or RK3576 platform. Fixes: 128a9bf8ace2 ("drm/rockchip: Add basic RK3588 HDMI output support") Signed-off-by: Sebastian Reichel Reviewed-by: Cristian Ciocaltea Signed-off-by: Heiko Stuebner Link: https://patch.msgid.link/20251014-rockchip-hdmi-suspend-fix-v1-1-983fcbf44839@collabora.com Signed-off-by: Sasha Levin

drm/rockchip: vop2: Use OVL_LAYER_SEL configuration instead of use win_mask calculate used layers

2026-01-08T09:17:17+00:00

commit d3fe9aa495854f8d88c69c41a4b31e69424656ad upstream. When there are multiple Video Ports, and only one of them is working (for example, VP1 is working while VP0 is not), in this case, the win_mask of VP0 is 0. However, we have already set the port mux for VP0 according to vp0->nlayers, and at the same time, in the OVL_LAYER_SEL register, there are windows will also be assigned to layers which will map to the inactive VPs. In this situation, vp0->win_mask is zero as it now working, it is more reliable to calculate the used layers based on the configuration of the OVL_LAYER_SEL register. Note: as the configuration of OVL_LAYER_SEL is take effect when the vsync is come, so we use the value backup in vop2->old_layer_sel instead of read OVL_LAYER_SEL directly. Fixes: 3e89a8c68354 ("drm/rockchip: vop2: Fix the update of LAYER/PORT select registers when there are multi display output on rk3588/rk3568") Cc: stable@vger.kernel.org Reported-by: Diederik de Haas Closes: https://bugs.kde.org/show_bug.cgi?id=511274 Signed-off-by: Andy Yan Tested-by: Dang Huynh Tested-by: Diederik de Haas Signed-off-by: Heiko Stuebner Link: https://lore.kernel.org/r/20251112085024.2480111-1-andyshrk@163.com Signed-off-by: Greg Kroah-Hartman

drm/rockchip: Set VOP for the DRM DMA device

2026-01-08T09:17:16+00:00

commit 7d7bb790aced3b1b8550b74e02fdfc001d044bee upstream. Use VOP for DMA operations performed by DRM core. Rockchip DRM driver is backed by a virtual device that isn't IOMMU-capable, while VOP is the actual display controller device backed by IOMMU. Fixes "swiotlb buffer is full" warning messages originated from GEM prime code paths. Note, that backporting is non-trivial as this depends on commit 143ec8d3f9396 ("drm/prime: Support dedicated DMA device for dma-buf imports"), which landed in v6.16 and commit 421be3ee36a4 ("drm/rockchip: Refactor IOMMU initialisation"), which landed in v5.19. Reported-by: Daniel Stone Fixes: 2048e3286f34 ("drm: rockchip: Add basic drm driver") Cc: stable@vger.kernel.org # v6.16+ Reviewed-by: Sebastian Reichel Signed-off-by: Dmitry Osipenko Tested-by: Cristian Ciocaltea Signed-off-by: Heiko Stuebner Link: https://lore.kernel.org/r/20251022161948.199731-1-dmitry.osipenko@collabora.com Signed-off-by: Greg Kroah-Hartman

drm/rockchip: dw_hdmi: use correct SCLIN mask for RK3228

2025-10-16T15:57:50+00:00

In dw_hdmi_rk3228_setup_hpd(), the SCLIN mask incorrectly references the RK3328 variant. This change updates it to the RK3228-specific mask RK3228_HDMI_SCLIN_MSK using FIELD_PREP_WM16, ensuring proper HPD and I2C pin configuration for RK3228. Change: RK3328_HDMI_SCLIN_MSK -> RK3228_HDMI_SCLIN_MSK Fixes: 63df37f3fc71 ("drm/rockchip: dw_hdmi: switch to FIELD_PREP_WM16* macros") Signed-off-by: Alok Tiwari Reviewed-by: Nicolas Frattaroli Signed-off-by: Heiko Stuebner Link: https://lore.kernel.org/r/20251010173143.72733-1-alok.a.tiwari@oracle.com