- 03 Apr, 2023 2 commits
-
-
Konrad Dybcio authored
The adreno_load_gpu() path is guarded by an error check on adreno_load_fw(). This function is responsible for loading Qualcomm-only-signed binaries (e.g. SQE and GMU FW for A6XX), but it does not take the vendor-signed ZAP blob into account. By embedding the SQE (and GMU, if necessary) firmware into the initrd/kernel, we can trigger and unfortunate path that would not bail out early and proceed with gpu->hw_init(). That will fail, as the ZAP loader path will not find the firmware and return back to adreno_load_gpu(). This error path involves pm_runtime_put_sync() which then calls idle() instead of suspend(). This is suboptimal, as it means that we're not going through the clean shutdown sequence. With at least A619_holi, this makes the GPU not wake up until it goes through at least one more start-fail-stop cycle. The pm_runtime_put_sync that appears in the error path actually does not guarantee that because of the earlier enabling of runtime autosuspend. Fix that by using pm_runtime_put_sync_suspend to force a clean shutdown. Test cases: 1. All firmware baked into kernel 2. error loading ZAP fw in initrd -> load from rootfs at DE start Both succeed on A619_holi (SM6375) and A630 (SDM845). Fixes: 0d997f95 ("drm/msm/adreno: fix runtime PM imbalance at gpu load") Signed-off-by:
Konrad Dybcio <konrad.dybcio@linaro.org> Reviewed-by:
Johan Hovold <johan+linaro@kernel.org> Patchwork: https://patchwork.freedesktop.org/patch/530001/ Link: https://lore.kernel.org/r/20230330231517.2747024-1-konrad.dybcio@linaro.orgSigned-off-by:
Rob Clark <robdclark@chromium.org>
-
Konrad Dybcio authored
Recently I contributed the switch to OPP API for all Adreno generations. I did however also skip over the fact that GPUs with a GMU don't specify a core clock of any kind in the GPU node. While that didn't break anything, it did introduce unwanted spam in the dmesg: adreno 5000000.gpu: error -ENOENT: _opp_set_clknames: Couldn't find clock with name: core_clk Guard the entire logic so that it's not used with GMU-equipped GPUs. Fixes: 9f251f93 ("drm/msm/adreno: Use OPP for every GPU generation") Signed-off-by:
Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Patchwork: https://patchwork.freedesktop.org/patch/530347/ Link: https://lore.kernel.org/r/20230223-topic-gmuwrapper-v6-1-2034115bb60c@linaro.orgSigned-off-by:
-
- 31 Mar, 2023 2 commits
-
-
Konrad Dybcio authored
Add support for matching QFPROM fuse values to get the correct speed bin on A650 (SM8250) GPUs. Reviewed-by:
-
Konrad Dybcio authored
Add support for matching QFPROM fuse values to get the correct speed bin on A640 (SM8150) GPUs. Reviewed-by:
Akhil P Oommen <quic_akhilpo@quicinc.com> Signed-off-by:
-
- 30 Mar, 2023 1 commit
-
-
Dmitry Baryshkov authored
Use adreno_fault_handler() to implement a5xx_fault_handler(). This enables devcoredump support on a5xx platforms, allowing one to capture the crashed GPU state at the time of context fault. Signed-off-by:
-
- 28 Mar, 2023 12 commits
-
-
Dmitry Baryshkov authored
Split the a6xx_fault_handler() into the generic adreno_fault_handler() and platform-specific parts. The adreno_fault_handler() can further be used by a5xx and hopefully by a4xx (at some point). Signed-off-by:
-
Dmitry Baryshkov authored
The commit e25e92e0 ("drm/msm: devcoredump iommu fault support") enabled SMMU stalling to collect GPU state, but only for a6xx. It tied enabling the stall with tha per-instance pagetables creation. Since that commit SoCs with a5xx also gained support for adreno-smmu-priv. Move stalling into generic code and add corresponding resume_translation calls. Signed-off-by:
-
Rob Clark authored
Drop our custom thing and switch to drm_crtc_next_vblank_start() for calculating the time of the start of the next vblank period. Signed-off-by:
-
Rob Clark authored
Add a way for various userspace waits to signal urgency. Signed-off-by:
-
Rob Clark authored
Track the nearest deadline on a fence timeline and set a timer to expire shortly before to trigger boost if the fence has not yet been signaled. v2: rebase Signed-off-by:
-
Rob Clark authored
This series adds a deadline hint to fences, so realtime deadlines such as vblank can be communicated to the fence signaller for power/ frequency management decisions. This is partially inspired by a trick i915 does, but implemented via dma-fence for a couple of reasons: 1) To continue to be able to use the atomic helpers 2) To support cases where display and gpu are different drivers See https://patchwork.freedesktop.org/series/93035/ This does not yet add any UAPI, although this will be needed in a number of cases: 1) Workloads "ping-ponging" between CPU and GPU, where we don't want the GPU freq governor to interpret time stalled waiting for GPU as "idle" time 2) Cases where the compositor is waiting for fences to be signaled before issuing the atomic ioctl, for example to maintain 60fps cursor updates even when the GPU is not able to maintain that framerate. Signed-off-by:
-
Rob Clark authored
For an atomic commit updating a single CRTC (ie. a pageflip) calculate the next vblank time, and inform the fence(s) of that deadline. v2: Comment typo fix (danvet) v3: If there are multiple CRTCs, consider the time of the soonest vblank Signed-off-by:
Daniel Vetter <daniel.vetter@ffwll.ch>
-
Rob Clark authored
Will be used in the next commit to set a deadline on fences that an atomic update is waiting on. v2: Calculate time at *start* of vblank period, not end v3: Fix kbuild complaints Signed-off-by:
Mario Kleiner <mario.kleiner.de@gmail.com>
-
Rob Clark authored
As the finished fence is the one that is exposed to userspace, and therefore the one that other operations, like atomic update, would block on, we need to propagate the deadline from from the finished fence to the actual hw fence. v2: Split into drm_sched_fence_set_parent() (ckoenig) v3: Ensure a thread calling drm_sched_fence_set_deadline_finished() sees fence->parent set before drm_sched_fence_set_parent() does this test_bit(DMA_FENCE_FLAG_HAS_DEADLINE_BIT). Signed-off-by:Luben Tuikov <luben.tuikov@amd.com>
-
Rob Clark authored
We had all of the internal driver APIs, but not the all important userspace uABI, in the dma-buf doc. Fix that. And re-arrange the comments slightly as otherwise the comments for the ioctl nr defines would not show up. v2: Fix docs build warning coming from newly including the uabi header in the docs build Signed-off-by:Pekka Paalanen <pekka.paalanen@collabora.com>
-
Rob Clark authored
Add a way to set a deadline on remaining resv fences according to the requested usage. Signed-off-by:
Christian König <christian.koenig@amd.com>
-
Rob Clark authored
Propagate the deadline to all the fences in the chain. v2: Use dma_fence_chain_contained [Tvrtko] Signed-off-by:
-
- 25 Mar, 2023 17 commits
-
-
Rob Clark authored
Move the one-time RPMh setup to a6xx_gmu_init(). To get rid of the hack for one-time init vs start, add in an extra a6xx_rpmh_stop() at the end of the init sequence. Signed-off-by:
-
Rob Clark authored
These allocations are only done the first (successful) time through hw_init() so they won't actually happen in the job_run() path. But lockdep doesn't know this. So dis-entangle them from the hw_init() path. Signed-off-by:
-
Rob Clark authored
It is already a no-op, since we've already loaded the fw from adreno_load_gpu(), so drop the redundant call. Signed-off-by:
-
Rob Clark authored
Avoid allocation under idr_lock, to prevent deadlock against the job_free() path (which runs on same thread as job_run(), which makes it also part of the fence-signaling path. Signed-off-by:
-
Rob Clark authored
Needed to idr_preload() which returns with preemption disabled. Signed-off-by:
-
Rob Clark authored
Now that everything that controls which LRU an obj lives in *except* the backing pages is protected by the LRU lock, add a special path to unpin in the job_run() path, where we are assured that we already have backing pages and will not be racing against eviction (because the GEM object's dma_resv contains the fence that will be signaled when the submit/job completes). Signed-off-by:
-
Rob Clark authored
Since the LRU lock is already acquired when moving an obj between LRUs, we can use it to protect pin_count and madv, without any significant change in locking (ie. it just expands the scope of the lock by a hand- ful of instructions). This prepares the way to decrement the pin_count in the job_run() path without needing to hold the obj lock, to avoid a potential deadlock (or rather stall) caused by the fence-signaling path (job_run()) blocking on shrinker/reclaim. (Only a stall because the wait for fence signaling wait_for_idle() is not infinite.) Signed-off-by:
-
Rob Clark authored
Just code-motion. Signed-off-by:
-
Rob Clark authored
Export the locked version or lru's move_tail(). Signed-off-by:
-
Rob Clark authored
vmap'ing is just pinning in disguise. So treat it as such and simplify the LRU tracking. Signed-off-by:
-
Rob Clark authored
We need to use the inuse count to track that a BO is pinned until we have the hw_fence. But we want to remove the obj lock from the job_run() path as this could deadlock against reclaim/shrinker (because it is blocking the hw_fence from eventually being signaled). So split that tracking out into a per-vma lock with narrower scope. Signed-off-by:
-
Rob Clark authored
Stop open coding VMA construction, which will be needed in the next commit. And since the VMA already has a ptr to the adress space, stop passing that around everywhere. (Also, an aspace always has an mmu so we can drop a couple pointless NULL checks.) Signed-off-by:
-
Rob Clark authored
The flags are only accessed (1) when submit is constructed, before enqueuing to gpu sched (ie. when still visible to only the task calling the submit ioctl), (2) here, where we own a reference to the submit and are serialized on the gpu sched thread, and (3) after the submit is retired and last reference is dropped, which is serialized on the submit's reference count. Hence locking is unneeded here. Signed-off-by:
-
Rob Clark authored
Avoid allocating memory in job_run() by pre-allocating the hw_fence. Signed-off-by:
-
Arnd Bergmann authored
Selecting CONFIG_PM_GENERIC_DOMAINS causes a build failure when CONFIG_PM is not enabled: WARNING: unmet direct dependencies detected for PM_GENERIC_DOMAINS Depends on [n]: PM [=n] Selected by [m]: - DRM_MSM [=m] && HAS_IOMEM [=y] && DRM [=m] && (ARCH_QCOM [=y] || SOC_IMX5 || COMPILE_TEST [=y]) && COMMON_CLK [=y] && IOMMU_SUPPORT [=y] && (QCOM_OCMEM [=y] || QCOM_OCMEM [=y]=n) && (QCOM_LLCC [=n] || QCOM_LLCC [=n]=n) && (QCOM_COMMAND_DB [=y] || QCOM_COMMAND_DB [=y]=n) && DEVFREQ_GOV_SIMPLE_ONDEMAND [=y] drivers/base/power/domain.c:654:13: error: use of undeclared identifier 'pm_wq' queue_work(pm_wq, &genpd->power_off_work); ^ drivers/base/power/domain.c:853:26: error: no member named 'ignore_children' in 'struct dev_pm_info' if (!dev || dev->power.ignore_children) ~~~~~~~~~~ ^ Fixes: c11fa120 ("drm/msm/a6xx: Use genpd notifier to ensure cx-gdsc collapse") Signed-off-by:Arnd Bergmann <arnd@arndb.de> Reviewed-by:
Ulf Hansson <ulf.hansson@linaro.org> Patchwork: https://patchwork.freedesktop.org/patch/528597/ Link: https://lore.kernel.org/r/20230324095502.3289094-1-arnd@kernel.orgSigned-off-by:
-
Rob Clark authored
Propagate the deadline to all the fences in the array. Signed-off-by:
-
Rob Clark authored
Add a way to hint to the fence signaler of an upcoming deadline, such as vblank, which the fence waiter would prefer not to miss. This is to aid the fence signaler in making power management decisions, like boosting frequency as the deadline approaches and awareness of missing deadlines so that can be factored in to the frequency scaling. v2: Drop dma_fence::deadline and related logic to filter duplicate deadlines, to avoid increasing dma_fence size. The fence-context implementation will need similar logic to track deadlines of all the fences on the same timeline. [ckoenig] v3: Clarify locking wrt. set_deadline callback v4: Clarify in docs comment that this is a hint v5: Drop DMA_FENCE_FLAG_HAS_DEADLINE_BIT. v6: More docs v7: Fix typo, clarify past deadlines Signed-off-by:Bagas Sanjaya <bagasdotme@gmail.com>
-
- 21 Mar, 2023 1 commit
-
-
Rob Clark authored
It's been a bit overdue. Regen headers to pull in a2xx perfcntr updates, etc. Signed-off-by:
-
- 20 Mar, 2023 5 commits
-
-
Rob Clark authored
The next generated header update will drop the _LO/_HI suffix, now that the userspace tooling properly understands 64b vs 32b regs (and the _LO/ _HI workarounds are getting cleaned up). So convert to using the 64b reg helpers in prep. Signed-off-by:
-
Adam Skladowski authored
Downstream driver appears to not support preemption on A510 target, trying to use one make device slow and fill log with rings related errors. Set num_rings to 1 to disable preemption. Suggested-by:
Adam Skladowski <a39.skl@gmail.com> Reviewed-by:
-
Abhinav Kumar authored
Update the URI for MSM DRM bugs for users to be able to file bugs at a centralized location. Signed-off-by:
Abhinav Kumar <quic_abhinavk@quicinc.com> Acked-by:
Rob Clark <robdclark@gmail.com> Patchwork: https://patchwork.freedesktop.org/patch/525026/ Link: https://lore.kernel.org/r/1677972416-7353-1-git-send-email-quic_abhinavk@quicinc.comSigned-off-by:
-
Johan Hovold authored
Clean up the component ops initialisers which were indented one level too far. Signed-off-by:
-
Johan Hovold authored
The runtime PM status can only be updated while runtime PM is disabled. Drop the bogus pm_runtime_set_active() call that was made after enabling runtime PM and which (incidentally but correctly) left the runtime PM status set to 'suspended'. Fixes: 2c087a33 ("drm/msm/adreno: Load the firmware before bringing up the hardware") Signed-off-by:
-
