kernel.git/drivers/gpu/drm/msm/msm_drv.c, branch linux-rolling-stable

drm/msm: Fix bootup splat with separate_gpu_drm modparam

2025-10-15T10:03:55Z

[ Upstream commit f028bcafb6dfb4c2bb656cbff9e6a66222d3d3d7 ] The drm_gem_for_each_gpuvm_bo() call from lookup_vma() accesses drm_gem_obj.gpuva.list, which is not initialized when the drm driver does not support DRIVER_GEM_GPUVA feature. Enable it for msm_kms drm driver to fix the splat seen when msm.separate_gpu_drm=1 modparam is set: [ 9.506020] Unable to handle kernel paging request at virtual address fffffffffffffff0 [ 9.523160] Mem abort info: [ 9.523161] ESR = 0x0000000096000006 [ 9.523163] EC = 0x25: DABT (current EL), IL = 32 bits [ 9.523165] SET = 0, FnV = 0 [ 9.523166] EA = 0, S1PTW = 0 [ 9.523167] FSC = 0x06: level 2 translation fault [ 9.523169] Data abort info: [ 9.523170] ISV = 0, ISS = 0x00000006, ISS2 = 0x00000000 [ 9.523171] CM = 0, WnR = 0, TnD = 0, TagAccess = 0 [ 9.523172] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 [ 9.523174] swapper pgtable: 4k pages, 48-bit VAs, pgdp=0000000ad370f000 [ 9.523176] [fffffffffffffff0] pgd=0000000000000000, p4d=0000000ad4787403, pud=0000000ad4788403, pmd=0000000000000000 [ 9.523184] Internal error: Oops: 0000000096000006 [#1] SMP [ 9.592968] CPU: 9 UID: 0 PID: 448 Comm: (udev-worker) Not tainted 6.17.0-rc4-assorted-fix-00005-g0e9bb53a2282-dirty #3 PREEMPT [ 9.592970] Hardware name: Qualcomm CRD, BIOS 6.0.240718.BOOT.MXF.2.4-00515-HAMOA-1 07/18/2024 [ 9.592971] pstate: a1400005 (NzCv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--) [ 9.592973] pc : lookup_vma+0x28/0xe0 [msm] [ 9.592996] lr : get_vma_locked+0x2c/0x128 [msm] [ 9.763632] sp : ffff800082dab460 [ 9.763666] Call trace: [ 9.763668] lookup_vma+0x28/0xe0 [msm] (P) [ 9.763688] get_vma_locked+0x2c/0x128 [msm] [ 9.763706] msm_gem_get_and_pin_iova_range+0x68/0x11c [msm] [ 9.763723] msm_gem_get_and_pin_iova+0x18/0x24 [msm] [ 9.763740] msm_fbdev_driver_fbdev_probe+0xd0/0x258 [msm] [ 9.763760] __drm_fb_helper_initial_config_and_unlock+0x288/0x528 [drm_kms_helper] [ 9.763771] drm_fb_helper_initial_config+0x44/0x54 [drm_kms_helper] [ 9.763779] drm_fbdev_client_hotplug+0x84/0xd4 [drm_client_lib] [ 9.763782] drm_client_register+0x58/0x9c [drm] [ 9.763806] drm_fbdev_client_setup+0xe8/0xcf0 [drm_client_lib] [ 9.763809] drm_client_setup+0xb4/0xd8 [drm_client_lib] [ 9.763811] msm_drm_kms_post_init+0x2c/0x3c [msm] [ 9.763830] msm_drm_init+0x1a8/0x22c [msm] [ 9.763848] msm_drm_bind+0x30/0x3c [msm] [ 9.919273] try_to_bring_up_aggregate_device+0x168/0x1d4 [ 9.919283] __component_add+0xa4/0x170 [ 9.919286] component_add+0x14/0x20 [ 9.919288] msm_dp_display_probe_tail+0x4c/0xac [msm] [ 9.919315] msm_dp_auxbus_done_probe+0x14/0x20 [msm] [ 9.919335] dp_aux_ep_probe+0x4c/0xf0 [drm_dp_aux_bus] [ 9.919341] really_probe+0xbc/0x298 [ 9.919345] __driver_probe_device+0x78/0x12c [ 9.919348] driver_probe_device+0x40/0x160 [ 9.919350] __driver_attach+0x94/0x19c [ 9.919353] bus_for_each_dev+0x74/0xd4 [ 9.919355] driver_attach+0x24/0x30 [ 9.919358] bus_add_driver+0xe4/0x208 [ 9.919360] driver_register+0x60/0x128 [ 9.919363] __dp_aux_dp_driver_register+0x24/0x30 [drm_dp_aux_bus] [ 9.919365] atana33xc20_init+0x20/0x1000 [panel_samsung_atna33xc20] [ 9.919370] do_one_initcall+0x6c/0x1b0 [ 9.919374] do_init_module+0x58/0x234 [ 9.919377] load_module+0x19cc/0x1bd4 [ 9.919380] init_module_from_file+0x84/0xc4 [ 9.919382] __arm64_sys_finit_module+0x1b8/0x2cc [ 9.919384] invoke_syscall+0x48/0x110 [ 9.919389] el0_svc_common.constprop.0+0xc8/0xe8 [ 9.919393] do_el0_svc+0x20/0x2c [ 9.919396] el0_svc+0x34/0xf0 [ 9.919401] el0t_64_sync_handler+0xa0/0xe4 [ 9.919403] el0t_64_sync+0x198/0x19c [ 9.919407] Code: eb0000bf 54000480 d100a003 aa0303e2 (f8418c44) [ 9.919410] ---[ end trace 0000000000000000 ]--- Fixes: 217ed15bd399 ("drm/msm: enable separate binding of GPU and display devices") Signed-off-by: Akhil P Oommen Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/672257/ Link: https://lore.kernel.org/r/20250902-assorted-sept-1-v1-1-f3ec9baed513@oss.qualcomm.com Signed-off-by: Dmitry Baryshkov Signed-off-by: Sasha Levin

drm/msm: Take the ioctls away from the KMS-only driver

2025-07-05T16:59:12Z

KMS-only drivers should only allocate dumb buffers. The driver custom ioctls are only meant for the usermode gpu driver (mesa), and not for general consumption, so they don't make sense for standalone KMS drivers. Signed-off-by: Rob Clark Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662598/

drm/msm: Clean up split driver features

2025-07-05T16:59:12Z

Avoid the possibility of missing features between the split and unified drm driver cases by defining DRIVER_FEATURES_GPU / KMS and using those in the drm_driver initializations. Signed-off-by: Rob Clark Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662595/

drm/msm: enable separate binding of GPU and display devices

2025-07-05T14:37:21Z

There are cases when we want to have separate DRM devices for GPU and display pipelines. One example is development, when it is beneficial to be able to bind the GPU driver separately, without the display pipeline (and without the hacks adding "amd,imageon" to the compatible string). Another example is some of Qualcomm platforms, which have two MDSS units, but only one GPU. With current approach it is next to impossible to support this usecase properly, while separate binding allows users to have three DRM devices: two for MDSS units and a single headless GPU. Add kernel param msm.separate_gpu_kms, which if set to true forces creation of separate display and GPU DRM devices. Mesa supports this setup by using the kmsro wrapper. The param is disabled by default, in order to be able to test userspace for the compatibility issues. Simple clients are able to handle this setup automatically. Signed-off-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662590/ [Rob: renamed the modparam to separate_gpu_kms, and add missing DRIVER_GEM_GPUVA] Signed-off-by: Rob Clark

drm/msm: rework binding of Imageon GPUs

2025-07-05T14:13:35Z

Currently the msm driver creates an extra interim platform device for Imageon GPUs. This is not ideal, as the device doesn't have corresponding OF node. If the headless mode is used for newer GPUs, then the msm_use_mmu() function can not detect corresponding IOMMU devices. Also the DRM device (although it's headless) is created with modesetting flags being set. To solve all these issues, rework the way the Imageon devices are bound. Remove the interim device, don't register a component and instead use a cut-down version of the normal functions to probe or remove the driver. Signed-off-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662584/ Signed-off-by: Rob Clark

drm/msm: move helper calls to msm_kms.c

2025-07-05T14:13:35Z

Extract two more KMS-related codepieces to msm_kms.c, removing last pieces of KMS code from msm_drv.c. Reviewed-by: Abhinav Kumar Signed-off-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662574/ Signed-off-by: Rob Clark

drm/msm: move wq handling to KMS code

2025-07-05T14:13:35Z

The global workqueue is only used for vblanks inside KMS code. Move allocation / flushing / deallcation of it to msm_kms.c Signed-off-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/662573/ Signed-off-by: Rob Clark

drm/msm: Defer VMA unmap for fb unpins

2025-07-05T00:48:38Z

With the conversion to drm_gpuvm, we lost the lazy VMA cleanup, which means that fb cleanup/unpin when pageflipping to new scanout buffers immediately unmaps the scanout buffer. This is costly (with tlbinv, it can be 4-6ms for a 1080p scanout buffer, and more for higher resolutions)! To avoid this, introduce a vma_ref, which is incremented whenever userspace has a GEM handle or dma-buf fd. When unpinning if the vm is the kms->vm we defer tearing down the VMA until the vma_ref drops to zero. If the buffer is still part of a flip-chain then userspace will be holding some sort of reference to the BO, either via a GEM handle and/or dma-buf fd. So this avoids unmapping the VMA when there is a strong possibility that it will be needed again. Signed-off-by: Rob Clark Tested-by: Antonino Maniscalco Reviewed-by: Antonino Maniscalco Patchwork: https://patchwork.freedesktop.org/patch/661538/

drm/msm: Bump UAPI version

2025-07-05T00:48:38Z

Bump version to signal to userspace that VM_BIND is supported. Signed-off-by: Rob Clark Signed-off-by: Rob Clark Tested-by: Antonino Maniscalco Reviewed-by: Antonino Maniscalco Patchwork: https://patchwork.freedesktop.org/patch/661535/

drm/msm: Add VM_BIND ioctl

2025-07-05T00:48:38Z

Add a VM_BIND ioctl for binding/unbinding buffers into a VM. This is only supported if userspace has opted in to MSM_PARAM_EN_VM_BIND. Signed-off-by: Rob Clark Signed-off-by: Rob Clark Tested-by: Antonino Maniscalco Reviewed-by: Antonino Maniscalco Patchwork: https://patchwork.freedesktop.org/patch/661524/