From 6b481ab0e6855fb30e2923c51f62f1662d1cda7e Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Fri, 7 Feb 2025 11:25:31 +1000 Subject: [PATCH 1/7] drm/nouveau: select FW caching nouveau tries to load some firmware during suspend that it loaded earlier, but with fw caching disabled it hangs suspend, so just rely on FW cache enabling instead of working around it in the driver. Fixes: 176fdcbddfd2 ("drm/nouveau/gsp/r535: add support for booting GSP-RM") Signed-off-by: Dave Airlie Signed-off-by: Danilo Krummrich Link: https://patchwork.freedesktop.org/patch/msgid/20250207012531.621369-1-airlied@gmail.com --- drivers/gpu/drm/nouveau/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/nouveau/Kconfig b/drivers/gpu/drm/nouveau/Kconfig index ce840300578d..1050a4617fc1 100644 --- a/drivers/gpu/drm/nouveau/Kconfig +++ b/drivers/gpu/drm/nouveau/Kconfig @@ -4,6 +4,7 @@ config DRM_NOUVEAU depends on DRM && PCI && MMU select IOMMU_API select FW_LOADER + select FW_CACHE if PM_SLEEP select DRM_CLIENT_SELECTION select DRM_DISPLAY_DP_HELPER select DRM_DISPLAY_HDMI_HELPER From df1a1ed5e1bdd9cc13148e0e5549f5ebcf76cf13 Mon Sep 17 00:00:00 2001 From: Brendan King Date: Wed, 26 Feb 2025 15:42:19 +0000 Subject: [PATCH 2/7] drm/imagination: avoid deadlock on fence release Do scheduler queue fence release processing on a workqueue, rather than in the release function itself. Fixes deadlock issues such as the following: [ 607.400437] ============================================ [ 607.405755] WARNING: possible recursive locking detected [ 607.415500] -------------------------------------------- [ 607.420817] weston:zfq0/24149 is trying to acquire lock: [ 607.426131] ffff000017d041a0 (reservation_ww_class_mutex){+.+.}-{3:3}, at: pvr_gem_object_vunmap+0x40/0xc0 [powervr] [ 607.436728] but task is already holding lock: [ 607.442554] ffff000017d105a0 (reservation_ww_class_mutex){+.+.}-{3:3}, at: dma_buf_ioctl+0x250/0x554 [ 607.451727] other info that might help us debug this: [ 607.458245] Possible unsafe locking scenario: [ 607.464155] CPU0 [ 607.466601] ---- [ 607.469044] lock(reservation_ww_class_mutex); [ 607.473584] lock(reservation_ww_class_mutex); [ 607.478114] *** DEADLOCK *** Cc: stable@vger.kernel.org Fixes: eaf01ee5ba28 ("drm/imagination: Implement job submission and scheduling") Signed-off-by: Brendan King Reviewed-by: Matt Coster Link: https://patchwork.freedesktop.org/patch/msgid/20250226-fence-release-deadlock-v2-1-6fed2fc1fe88@imgtec.com Signed-off-by: Matt Coster --- drivers/gpu/drm/imagination/pvr_queue.c | 13 +++++++++++-- drivers/gpu/drm/imagination/pvr_queue.h | 4 ++++ 2 files changed, 15 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_queue.c b/drivers/gpu/drm/imagination/pvr_queue.c index c4f08432882b..f3f1c5212df7 100644 --- a/drivers/gpu/drm/imagination/pvr_queue.c +++ b/drivers/gpu/drm/imagination/pvr_queue.c @@ -109,12 +109,20 @@ pvr_queue_fence_get_driver_name(struct dma_fence *f) return PVR_DRIVER_NAME; } +static void pvr_queue_fence_release_work(struct work_struct *w) +{ + struct pvr_queue_fence *fence = container_of(w, struct pvr_queue_fence, release_work); + + pvr_context_put(fence->queue->ctx); + dma_fence_free(&fence->base); +} + static void pvr_queue_fence_release(struct dma_fence *f) { struct pvr_queue_fence *fence = container_of(f, struct pvr_queue_fence, base); + struct pvr_device *pvr_dev = fence->queue->ctx->pvr_dev; - pvr_context_put(fence->queue->ctx); - dma_fence_free(f); + queue_work(pvr_dev->sched_wq, &fence->release_work); } static const char * @@ -268,6 +276,7 @@ pvr_queue_fence_init(struct dma_fence *f, pvr_context_get(queue->ctx); fence->queue = queue; + INIT_WORK(&fence->release_work, pvr_queue_fence_release_work); dma_fence_init(&fence->base, fence_ops, &fence_ctx->lock, fence_ctx->id, atomic_inc_return(&fence_ctx->seqno)); diff --git a/drivers/gpu/drm/imagination/pvr_queue.h b/drivers/gpu/drm/imagination/pvr_queue.h index e06ced69302f..93fe9ac9f58c 100644 --- a/drivers/gpu/drm/imagination/pvr_queue.h +++ b/drivers/gpu/drm/imagination/pvr_queue.h @@ -5,6 +5,7 @@ #define PVR_QUEUE_H #include +#include #include "pvr_cccb.h" #include "pvr_device.h" @@ -63,6 +64,9 @@ struct pvr_queue_fence { /** @queue: Queue that created this fence. */ struct pvr_queue *queue; + + /** @release_work: Fence release work structure. */ + struct work_struct release_work; }; /** From a5c4c3ba95a52d66315acdfbaba9bd82ed39c250 Mon Sep 17 00:00:00 2001 From: Brendan King Date: Wed, 26 Feb 2025 15:43:06 +0000 Subject: [PATCH 3/7] drm/imagination: Hold drm_gem_gpuva lock for unmap Avoid a warning from drm_gem_gpuva_assert_lock_held in drm_gpuva_unlink. The Imagination driver uses the GEM object reservation lock to protect the gpuva list, but the GEM object was not always known in the code paths that ended up calling drm_gpuva_unlink. When the GEM object isn't known, it is found by calling drm_gpuva_find to lookup the object associated with a given virtual address range, or by calling drm_gpuva_find_first when removing all mappings. Cc: stable@vger.kernel.org Fixes: 4bc736f890ce ("drm/imagination: vm: make use of GPUVM's drm_exec helper") Signed-off-by: Brendan King Reviewed-by: Matt Coster Link: https://patchwork.freedesktop.org/patch/msgid/20250226-hold-drm_gem_gpuva-lock-for-unmap-v2-1-3fdacded227f@imgtec.com Signed-off-by: Matt Coster --- drivers/gpu/drm/imagination/pvr_fw_meta.c | 6 +- drivers/gpu/drm/imagination/pvr_vm.c | 134 +++++++++++++++++----- drivers/gpu/drm/imagination/pvr_vm.h | 3 + 3 files changed, 115 insertions(+), 28 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_fw_meta.c b/drivers/gpu/drm/imagination/pvr_fw_meta.c index c39beb70c317..6d13864851fc 100644 --- a/drivers/gpu/drm/imagination/pvr_fw_meta.c +++ b/drivers/gpu/drm/imagination/pvr_fw_meta.c @@ -527,8 +527,10 @@ pvr_meta_vm_map(struct pvr_device *pvr_dev, struct pvr_fw_object *fw_obj) static void pvr_meta_vm_unmap(struct pvr_device *pvr_dev, struct pvr_fw_object *fw_obj) { - pvr_vm_unmap(pvr_dev->kernel_vm_ctx, fw_obj->fw_mm_node.start, - fw_obj->fw_mm_node.size); + struct pvr_gem_object *pvr_obj = fw_obj->gem; + + pvr_vm_unmap_obj(pvr_dev->kernel_vm_ctx, pvr_obj, + fw_obj->fw_mm_node.start, fw_obj->fw_mm_node.size); } static bool diff --git a/drivers/gpu/drm/imagination/pvr_vm.c b/drivers/gpu/drm/imagination/pvr_vm.c index 363f885a7098..2896fa7501b1 100644 --- a/drivers/gpu/drm/imagination/pvr_vm.c +++ b/drivers/gpu/drm/imagination/pvr_vm.c @@ -293,8 +293,9 @@ pvr_vm_bind_op_map_init(struct pvr_vm_bind_op *bind_op, static int pvr_vm_bind_op_unmap_init(struct pvr_vm_bind_op *bind_op, - struct pvr_vm_context *vm_ctx, u64 device_addr, - u64 size) + struct pvr_vm_context *vm_ctx, + struct pvr_gem_object *pvr_obj, + u64 device_addr, u64 size) { int err; @@ -318,6 +319,7 @@ pvr_vm_bind_op_unmap_init(struct pvr_vm_bind_op *bind_op, goto err_bind_op_fini; } + bind_op->pvr_obj = pvr_obj; bind_op->vm_ctx = vm_ctx; bind_op->device_addr = device_addr; bind_op->size = size; @@ -597,20 +599,6 @@ pvr_vm_create_context(struct pvr_device *pvr_dev, bool is_userspace_context) return ERR_PTR(err); } -/** - * pvr_vm_unmap_all() - Unmap all mappings associated with a VM context. - * @vm_ctx: Target VM context. - * - * This function ensures that no mappings are left dangling by unmapping them - * all in order of ascending device-virtual address. - */ -void -pvr_vm_unmap_all(struct pvr_vm_context *vm_ctx) -{ - WARN_ON(pvr_vm_unmap(vm_ctx, vm_ctx->gpuvm_mgr.mm_start, - vm_ctx->gpuvm_mgr.mm_range)); -} - /** * pvr_vm_context_release() - Teardown a VM context. * @ref_count: Pointer to reference counter of the VM context. @@ -703,11 +691,7 @@ pvr_vm_lock_extra(struct drm_gpuvm_exec *vm_exec) struct pvr_vm_bind_op *bind_op = vm_exec->extra.priv; struct pvr_gem_object *pvr_obj = bind_op->pvr_obj; - /* Unmap operations don't have an object to lock. */ - if (!pvr_obj) - return 0; - - /* Acquire lock on the GEM being mapped. */ + /* Acquire lock on the GEM object being mapped/unmapped. */ return drm_exec_lock_obj(&vm_exec->exec, gem_from_pvr_gem(pvr_obj)); } @@ -772,8 +756,10 @@ pvr_vm_map(struct pvr_vm_context *vm_ctx, struct pvr_gem_object *pvr_obj, } /** - * pvr_vm_unmap() - Unmap an already mapped section of device-virtual memory. + * pvr_vm_unmap_obj_locked() - Unmap an already mapped section of device-virtual + * memory. * @vm_ctx: Target VM context. + * @pvr_obj: Target PowerVR memory object. * @device_addr: Virtual device address at the start of the target mapping. * @size: Size of the target mapping. * @@ -784,9 +770,13 @@ pvr_vm_map(struct pvr_vm_context *vm_ctx, struct pvr_gem_object *pvr_obj, * * Any error encountered while performing internal operations required to * destroy the mapping (returned from pvr_vm_gpuva_unmap or * pvr_vm_gpuva_remap). + * + * The vm_ctx->lock must be held when calling this function. */ -int -pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size) +static int +pvr_vm_unmap_obj_locked(struct pvr_vm_context *vm_ctx, + struct pvr_gem_object *pvr_obj, + u64 device_addr, u64 size) { struct pvr_vm_bind_op bind_op = {0}; struct drm_gpuvm_exec vm_exec = { @@ -799,11 +789,13 @@ pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size) }, }; - int err = pvr_vm_bind_op_unmap_init(&bind_op, vm_ctx, device_addr, - size); + int err = pvr_vm_bind_op_unmap_init(&bind_op, vm_ctx, pvr_obj, + device_addr, size); if (err) return err; + pvr_gem_object_get(pvr_obj); + err = drm_gpuvm_exec_lock(&vm_exec); if (err) goto err_cleanup; @@ -818,6 +810,96 @@ pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size) return err; } +/** + * pvr_vm_unmap_obj() - Unmap an already mapped section of device-virtual + * memory. + * @vm_ctx: Target VM context. + * @pvr_obj: Target PowerVR memory object. + * @device_addr: Virtual device address at the start of the target mapping. + * @size: Size of the target mapping. + * + * Return: + * * 0 on success, + * * Any error encountered by pvr_vm_unmap_obj_locked. + */ +int +pvr_vm_unmap_obj(struct pvr_vm_context *vm_ctx, struct pvr_gem_object *pvr_obj, + u64 device_addr, u64 size) +{ + int err; + + mutex_lock(&vm_ctx->lock); + err = pvr_vm_unmap_obj_locked(vm_ctx, pvr_obj, device_addr, size); + mutex_unlock(&vm_ctx->lock); + + return err; +} + +/** + * pvr_vm_unmap() - Unmap an already mapped section of device-virtual memory. + * @vm_ctx: Target VM context. + * @device_addr: Virtual device address at the start of the target mapping. + * @size: Size of the target mapping. + * + * Return: + * * 0 on success, + * * Any error encountered by drm_gpuva_find, + * * Any error encountered by pvr_vm_unmap_obj_locked. + */ +int +pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size) +{ + struct pvr_gem_object *pvr_obj; + struct drm_gpuva *va; + int err; + + mutex_lock(&vm_ctx->lock); + + va = drm_gpuva_find(&vm_ctx->gpuvm_mgr, device_addr, size); + if (va) { + pvr_obj = gem_to_pvr_gem(va->gem.obj); + err = pvr_vm_unmap_obj_locked(vm_ctx, pvr_obj, + va->va.addr, va->va.range); + } else { + err = -ENOENT; + } + + mutex_unlock(&vm_ctx->lock); + + return err; +} + +/** + * pvr_vm_unmap_all() - Unmap all mappings associated with a VM context. + * @vm_ctx: Target VM context. + * + * This function ensures that no mappings are left dangling by unmapping them + * all in order of ascending device-virtual address. + */ +void +pvr_vm_unmap_all(struct pvr_vm_context *vm_ctx) +{ + mutex_lock(&vm_ctx->lock); + + for (;;) { + struct pvr_gem_object *pvr_obj; + struct drm_gpuva *va; + + va = drm_gpuva_find_first(&vm_ctx->gpuvm_mgr, + vm_ctx->gpuvm_mgr.mm_start, + vm_ctx->gpuvm_mgr.mm_range); + if (!va) + break; + + pvr_obj = gem_to_pvr_gem(va->gem.obj); + + WARN_ON(pvr_vm_unmap_obj_locked(vm_ctx, pvr_obj, + va->va.addr, va->va.range)); + } + + mutex_unlock(&vm_ctx->lock); +} + /* Static data areas are determined by firmware. */ static const struct drm_pvr_static_data_area static_data_areas[] = { { diff --git a/drivers/gpu/drm/imagination/pvr_vm.h b/drivers/gpu/drm/imagination/pvr_vm.h index 79406243617c..b0528dffa7f1 100644 --- a/drivers/gpu/drm/imagination/pvr_vm.h +++ b/drivers/gpu/drm/imagination/pvr_vm.h @@ -38,6 +38,9 @@ struct pvr_vm_context *pvr_vm_create_context(struct pvr_device *pvr_dev, int pvr_vm_map(struct pvr_vm_context *vm_ctx, struct pvr_gem_object *pvr_obj, u64 pvr_obj_offset, u64 device_addr, u64 size); +int pvr_vm_unmap_obj(struct pvr_vm_context *vm_ctx, + struct pvr_gem_object *pvr_obj, + u64 device_addr, u64 size); int pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size); void pvr_vm_unmap_all(struct pvr_vm_context *vm_ctx); From 68c3de7f707e8a70e0a6d8087cf0fe4a3d5dbfb0 Mon Sep 17 00:00:00 2001 From: Brendan King Date: Wed, 26 Feb 2025 15:43:54 +0000 Subject: [PATCH 4/7] drm/imagination: only init job done fences once Ensure job done fences are only initialised once. This fixes a memory manager not clean warning from drm_mm_takedown on module unload. Cc: stable@vger.kernel.org Fixes: eaf01ee5ba28 ("drm/imagination: Implement job submission and scheduling") Signed-off-by: Brendan King Reviewed-by: Matt Coster Link: https://patchwork.freedesktop.org/patch/msgid/20250226-init-done-fences-once-v2-1-c1b2f556b329@imgtec.com Signed-off-by: Matt Coster --- drivers/gpu/drm/imagination/pvr_queue.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_queue.c b/drivers/gpu/drm/imagination/pvr_queue.c index f3f1c5212df7..43411be930a2 100644 --- a/drivers/gpu/drm/imagination/pvr_queue.c +++ b/drivers/gpu/drm/imagination/pvr_queue.c @@ -313,8 +313,9 @@ pvr_queue_cccb_fence_init(struct dma_fence *fence, struct pvr_queue *queue) static void pvr_queue_job_fence_init(struct dma_fence *fence, struct pvr_queue *queue) { - pvr_queue_fence_init(fence, queue, &pvr_queue_job_fence_ops, - &queue->job_fence_ctx); + if (!fence->ops) + pvr_queue_fence_init(fence, queue, &pvr_queue_job_fence_ops, + &queue->job_fence_ctx); } /** From 1d2eabb6616433ccaa13927811bdfa205e91ba60 Mon Sep 17 00:00:00 2001 From: Alessio Belle Date: Fri, 21 Feb 2025 10:49:35 +0000 Subject: [PATCH 5/7] drm/imagination: Fix timestamps in firmware traces When firmware traces are enabled, the firmware dumps 48-bit timestamps for each trace as two 32-bit values, highest 32 bits (of which only 16 useful) first. The driver was reassembling them the other way round i.e. interpreting the first value in memory as the lowest 32 bits, and the second value as the highest 32 bits (then truncated to 16 bits). Due to this, firmware trace dumps showed very large timestamps even for traces recorded shortly after GPU boot. The timestamps in these dumps would also sometimes jump backwards because of the truncation. Example trace dumped after loading the powervr module and enabling firmware traces, where each line is commented with the timestamp value in hexadecimal to better show both issues: [93540092739584] : Host Sync Partition marker: 1 // 0x551300000000 [28419798597632] : GPU units deinit // 0x19d900000000 [28548647616512] : GPU deinit // 0x19f700000000 Update logic to reassemble the timestamps halves in the correct order. Fixes: cb56cd610866 ("drm/imagination: Add firmware trace to debugfs") Signed-off-by: Alessio Belle Reviewed-by: Matt Coster Link: https://patchwork.freedesktop.org/patch/msgid/20250221-fix-fw-trace-timestamps-v1-1-dba4aeb030ca@imgtec.com Signed-off-by: Matt Coster --- drivers/gpu/drm/imagination/pvr_fw_trace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_fw_trace.c b/drivers/gpu/drm/imagination/pvr_fw_trace.c index 73707daa4e52..5dbb636d7d4f 100644 --- a/drivers/gpu/drm/imagination/pvr_fw_trace.c +++ b/drivers/gpu/drm/imagination/pvr_fw_trace.c @@ -333,8 +333,8 @@ static int fw_trace_seq_show(struct seq_file *s, void *v) if (sf_id == ROGUE_FW_SF_LAST) return -EINVAL; - timestamp = read_fw_trace(trace_seq_data, 1) | - ((u64)read_fw_trace(trace_seq_data, 2) << 32); + timestamp = ((u64)read_fw_trace(trace_seq_data, 1) << 32) | + read_fw_trace(trace_seq_data, 2); timestamp = (timestamp & ~ROGUE_FWT_TIMESTAMP_TIME_CLRMSK) >> ROGUE_FWT_TIMESTAMP_TIME_SHIFT; From 23e0832d6d7be2d3c713f9390c060b6f1c48bf36 Mon Sep 17 00:00:00 2001 From: Philipp Stanner Date: Tue, 18 Feb 2025 13:41:50 +0100 Subject: [PATCH 6/7] drm/sched: Fix preprocessor guard When writing the header guard for gpu_scheduler_trace.h, a typo, apparently, occurred. Fix the typo and document the scope of the guard. Fixes: 353da3c520b4 ("drm/amdgpu: add tracepoint for scheduler (v2)") Reviewed-by: Tvrtko Ursulin Signed-off-by: Philipp Stanner Link: https://patchwork.freedesktop.org/patch/msgid/20250218124149.118002-2-phasta@kernel.org --- drivers/gpu/drm/scheduler/gpu_scheduler_trace.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h index c75302ca3427..f56e77e7f6d0 100644 --- a/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h +++ b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h @@ -21,7 +21,7 @@ * */ -#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ) +#if !defined(_GPU_SCHED_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ) #define _GPU_SCHED_TRACE_H_ #include @@ -106,7 +106,7 @@ TRACE_EVENT(drm_sched_job_wait_dep, __entry->seqno) ); -#endif +#endif /* _GPU_SCHED_TRACE_H_ */ /* This part must be outside protection */ #undef TRACE_INCLUDE_PATH From 80da96d735094ea22985ced98bc57fe3a4422921 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 4 Mar 2025 14:41:57 +0100 Subject: [PATCH 7/7] drm/bochs: Fix DPMS regression The recent rewrite with the use of regular atomic helpers broke the DPMS unblanking on X11. Fix it by moving the call of bochs_hw_blank(false) from CRTC mode_set_nofb() to atomic_enable(). Fixes: 2037174993c8 ("drm/bochs: Use regular atomic helpers") Link: https://bugzilla.suse.com/show_bug.cgi?id=1238209 Signed-off-by: Takashi Iwai Reviewed-by: Thomas Zimmermann Signed-off-by: Thomas Zimmermann Link: https://patchwork.freedesktop.org/patch/msgid/20250304134203.20534-1-tiwai@suse.de --- drivers/gpu/drm/tiny/bochs.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/tiny/bochs.c b/drivers/gpu/drm/tiny/bochs.c index c67e1f906785..8706763af8fb 100644 --- a/drivers/gpu/drm/tiny/bochs.c +++ b/drivers/gpu/drm/tiny/bochs.c @@ -335,8 +335,6 @@ static void bochs_hw_setmode(struct bochs_device *bochs, struct drm_display_mode bochs->xres, bochs->yres, bochs->bpp, bochs->yres_virtual); - bochs_hw_blank(bochs, false); - bochs_dispi_write(bochs, VBE_DISPI_INDEX_ENABLE, 0); bochs_dispi_write(bochs, VBE_DISPI_INDEX_BPP, bochs->bpp); bochs_dispi_write(bochs, VBE_DISPI_INDEX_XRES, bochs->xres); @@ -506,6 +504,9 @@ static int bochs_crtc_helper_atomic_check(struct drm_crtc *crtc, static void bochs_crtc_helper_atomic_enable(struct drm_crtc *crtc, struct drm_atomic_state *state) { + struct bochs_device *bochs = to_bochs_device(crtc->dev); + + bochs_hw_blank(bochs, false); } static void bochs_crtc_helper_atomic_disable(struct drm_crtc *crtc,