aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/gpu (follow)
AgeCommit message (Collapse)AuthorFilesLines
2018-12-18drm/amd/display: Raise dispclk value for PolarisRoman Li1-0/+7
[Why] The visual corruption due to low display clock value observed on some systems [How] There was earlier patch for dspclk: 'drm/amd/display: Raise dispclk value for dce_update_clocks' Adding +15% workaround also to to dce112_update_clocks Signed-off-by: Roman Li <Roman.Li@amd.com> Reviewed-by: Leo Li <Sunpeng.Li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amd/display: Skip fast cursor updates for fb changesNicholas Kazlauskas1-0/+10
[Why] The behavior of drm_atomic_helper_cleanup_planes differs depending on whether the commit was asynchronous or not. When it's called from amdgpu_dm_atomic_commit_tail during a typical atomic commit the plane state has been swapped so it calls cleanup_fb on the old plane state. However, in the asynchronous commit codepath the call to drm_atomic_helper_commit also calls dm_plane_helper_cleanup_fb after atomic_async_update has been called. Since the plane state is updated in place and has not been swapped the cleanup_fb call affects the new plane state. This results in a use after free for the given sequence: - Fast update, fb1 pin/ref, fb1 unpin/unref - Fast update, fb2 pin/ref, fb2 unpin/unref - Slow update, fb1 pin/ref, fb2 unpin/unref - Fast update, fb2 pin/ref -> use after free. bug [How] Disallow framebuffer changes in the fast path. Since this includes a NULL framebuffer, this means that only framebuffers that have been previously pin+ref at least once will be used, preventing a use after free. This has a significant throughput reduction for cursor updates where the framebuffer changes. For most desktop usage this isn't a problem, but it does introduce performance regressions for two specific IGT tests: - cursor-vs-flip-toggle - cursor-vs-flip-varying-size Fixes: 2cc751931afc ("drm/amd/display: Add fast path for cursor plane updates") Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Harry Wentland <harry.wentland@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu: correct the return value for error caseEvan Quan3-16/+31
It should not return 0 for error case as '0' is actually a special value for index. Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdkfd: Fix handling of return code of dma_buf_getFelix Kuehling1-2/+2
On errors, dma_buf_get returns a negative error code, rather than NULL. Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu:Improves robustness of SOC15_WAIT_ON_RREGJames Zhu1-2/+7
If register value is updating, reset timeout counter. It improves robustness of SOC15_WAIT_ON_RREG. Signed-off-by: James Zhu <James.Zhu@amd.com> Reviewed-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn:Remove bit 31 for scratch2 to indicate the WA is activeJames Zhu1-2/+2
Remove bit 31 for scratch2 to indicate the Hardware bug work around is active. Signed-off-by: James Zhu <James.Zhu@amd.com> Acked-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn:Scan enc/jpeg fences to init dpg pause new stateJames Zhu1-4/+14
Scan enc/jpeg fences to init dpg pause new state in begin use. It will help set dpg mode to desire state actively. Signed-off-by: James Zhu <James.Zhu@amd.com> Acked-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn:Always check all vcn ring status during dpg mode stopJames Zhu1-8/+16
Always check all vcn ring status during dpg mode stop, it will help identify which vcn ring may cause the issue. Signed-off-by: James Zhu <James.Zhu@amd.com> Acked-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn:Update ring point to JPEG before init JPEG wptrJames Zhu1-1/+2
It is a bug fix. Signed-off-by: James Zhu <James.Zhu@amd.com> Acked-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn:Always gate vcn block during hw finishingJames Zhu1-1/+2
Under Dynamic Power Gate mode, UVD_STATUS needn't be checked. Signed-off-by: James Zhu <James.Zhu@amd.com> Acked-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/amdgpu/vcn: Update vcn.cur_state during suspendJames Zhu1-1/+2
Replace vcn_v1_0_stop with vcn_v1_0_set_powergating_state during suspend, to keep adev->vcn.cur_state update. It will fix VCN S3 hung issue. Signed-off-by: James Zhu <James.Zhu@amd.com> Reviewed-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-18drm/xen-front: Use Xen common shared buffer implementationOleksandr Andrushchenko6-520/+26
Use page directory based shared buffer implementation now available as common code for Xen frontend drivers. Remove flushing of shared buffer on page flip as this workaround needs a proper fix. Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com> Reviewed-by: Noralf Trønnes <noralf@tronnes.org> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
2018-12-18Merge branch 'etnaviv/next' of https://git.pengutronix.de/git/lst/linux into drm-nextDaniel Vetter5-53/+21
Lucas writes: "nothing major this time, mostly some cleanups that were found on the way of reworking the code in preparation for new feature additions." Small conflict in drivers/gpu/drm/etnaviv/etnaviv_drv.c because drm-misc-next also has a patch to switch over to _put() functions. Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> From: Lucas Stach <l.stach@pengutronix.de> Link: https://patchwork.freedesktop.org/patch/msgid/1545130845.5874.23.camel@pengutronix.de
2018-12-18drm/etnaviv: remove lastctx member from gpu structLucas Stach4-12/+1
It only written and we don't infer any useful information from it anymore. Remove it. Signed-off-by: Lucas Stach <l.stach@pengutronix.de> Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2018-12-18drm/etnaviv: replace header include with forward declarationLucas Stach2-3/+4
The etnaviv_gpu header only needs to know about the pointer types, so replace by a forward declaration and only include the headers where needed. Signed-off-by: Lucas Stach <l.stach@pengutronix.de> Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2018-12-18drm/etnaviv: remove unnecessary local irq disableLucas Stach1-13/+5
The only event function that is called from IRQ context is event_free, which is already using atomic bitmap operations, so we can avoid taking the event spinlock in this function completely. As other the other functions still using the event spinlock are all called from normal process context, we can avoid disabling IRQs while holding the spinlock. Signed-off-by: Lucas Stach <l.stach@pengutronix.de> Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2018-12-14drm/amdgpu: WARN once if amdgpu_bo_unpin is called for an unpinned BOMichel Dänzer1-1/+1
It indicates a pin/unpin imbalance bug somewhere. While the bug isn't necessarily in the call chain hitting this, it's at least one part involved. Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: enable Vega20 page queue supportEvan Quan1-2/+1
Page queue is supported on Vega20 with SDMA firmware 123 onwards. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: use different irq ring ID for Vega20 page queuesEvan Quan1-2/+4
Vega20 uses ring id 1 for page queues EOP irq while previous ASICs take ring id 3. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: update the vm invalidation engine layout V2Evan Quan2-22/+41
We need new invalidation engine layout due to new SDMA page queues added. V2: fix coding style and add correct return value Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Oak Zeng <Oak.Zeng@amd.com> Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: increase the MAX ring numberEvan Quan1-1/+1
As two more SDMA page queue rings are added on Vega20. Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Oak Zeng <Oak.Zeng@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: fix NULL fence handling in amdgpu_cs_fence_to_handle_ioctlChristian König1-0/+3
When the fence is already signaled it is perfectly normal to get a NULL fence here. But since we can't export that we need to use a stub fence. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu/psp: Correct and refine the vmr support. (v2)Emily Deng3-15/+16
Currently driver only psp v11 support vmr. v2: squash in unused variable removal (Alex) Signed-off-by: Emily Deng <Emily.Deng@amd.com> Reviewed-by: Xiangliang Yu <Xiangliang.Yu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu/psp_v3_1: Get psp fw version through reading registerEmily Deng1-1/+4
If PSP FW is running already, driver will not load PSP FW again and skip it. So psp fw version is not correct if reading it from FW binary file, need to get right version from register. Signed-off-by: Emily Deng <Emily.Deng@amd.com> Reviewed-by: Xiangliang Yu <Xiangliang.Yu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: access register without KIQEmily Deng1-9/+9
There is no need to access register such as mmSMC_IND_INDEX_11 and mmSMC_IND_DATA_11, PCIE_INDEX, PCIE_DATA through KIQ because they are VF-copy. Signed-off-by: Emily Deng <Emily.Deng@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/amdgpu: kfd_pre_reset outside req_full_gpu cause sriov hangwentalou1-4/+6
XGMI hive put kfd_pre_reset into amdgpu_device_lock_adev, but outside req_full_gpu of sriov. It would make sriov hang during reset. Signed-off-by: Wentao Lou <Wentao.Lou@amd.com> Reviewed-by: Shaoyun Liu <Shaoyun.Liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-14drm/exynos: fimd: Make pixel blend mode configurableChristoph Manszewski1-12/+56
The fimd hardware supports different blend modes. Add pixel blend mode property and make it configurable, by modifying the blend equation. Tested on TRATS2 with Exynos 4412 CPU, on top of linux-next-20181019. Signed-off-by: Christoph Manszewski <c.manszewski@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
2018-12-14drm/exynos: fimd: Make plane alpha configurableChristoph Manszewski1-21/+54
The fimd hardware supports variable plane alpha. Currently planes are opaque, make this configurable. Tested on TRATS2 with Exynos 4412 CPU, on top of linux-next-20181019. Signed-off-by: Christoph Manszewski <c.manszewski@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com>
2018-12-13dma-mapping: bypass indirect calls for dma-directChristoph Hellwig1-1/+1
Avoid expensive indirect calls in the fast path DMA mapping operations by directly calling the dma_direct_* ops if we are using the directly mapped DMA operations. Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Tested-by: Jesper Dangaard Brouer <brouer@redhat.com> Tested-by: Tony Luck <tony.luck@intel.com>
2018-12-14Merge branch 'vmwgfx-fixes-4.20' of git://people.freedesktop.org/~thomash/linux into drm-fixesDave Airlie6-4/+103
One regression fix for avoiding kernel OOM, one cleanup return fix. Signed-off-by: Dave Airlie <airlied@redhat.com> From: Thomas Hellstrom <thellstrom@vmware.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181213122815.10581-1-thellstrom@vmware.com
2018-12-14Merge tag 'vmwgfx-next-2018-12-13' of git://people.freedesktop.org/~thomash/linux into drm-nextDave Airlie2-2/+2
Pull request of 2018-12-13 Two minor fixes for next pull. Signed-off-by: Dave Airlie <airlied@redhat.com> From: Thomas Hellstrom <thellstrom@vmware.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181213130848.3080-1-thellstrom@vmware.com
2018-12-13drm/vmwgfx: Protect from excessive execbuf kernel memory allocations v3Thomas Hellstrom6-2/+103
With the new validation code, a malicious user-space app could potentially submit command streams with enough buffer-object and resource references in them to have the resulting allocated validion nodes and relocations make the kernel run out of GFP_KERNEL memory. Protect from this by having the validation code reserve TTM graphics memory when allocating. Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Deepak Rawat <drawat@vmware.com> --- v2: Removed leftover debug printouts
2018-12-13Merge tag 'drm-msm-next-2018-12-12' of git://people.freedesktop.org/~robclark/linux into drm-nextDave Airlie125-5493/+4566
This time around, seeing some love for some older hw: - a2xx gpu support for apq8060 (hp touchpad) and imx5 (headless gpu-only mode) - a2xx gpummu support (a2xx was pre-iommu) - mdp4 display support for apq8060/touchpad For display/dpu: - a big pile of continuing dpu fixes and cleanups On the gpu side of things: - per-submit statistics and traceevents for better profiling - a6xx crashdump support - decouple get_iova() and page pinning.. so we can unpin from physical memory inactive bo's while using softpin to lower cpu overhead - new interface to set debug names on GEM BOs and debugfs output improvements - additional submit flag to indicate buffers that are used to dump (so $debugfs/rd cmdstream dumping is useful with softpin + state-objects) Signed-off-by: Dave Airlie <airlied@redhat.com> From: Rob Clark <robdclark@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGvVvLPD9_Z4kyfGe98Y--byj6HbxHivEYSgF7Rq7=bFnw@mail.gmail.com
2018-12-13Merge branch 'linux-4.21' of git://github.com/skeggsb/linux into drm-nextDave Airlie119-308/+2450
Mostly just initial support for Turing TU104/TU106 chipsets. Support for TU102 is missing as I don't yet have HW, but it should be trivial to add in later in the merge window (in theory). It's a bit of a rough first pass that'll get improved in future releases as a finish figuring out some of the other HW changes, but it's good enough as it stands for modesetting and suspend/resume etc. Acceleration bring-up is incomplete due to NVIDIA not yet having provided FW images for me to use, though command submission and copy engines are functional already. Signed-off-by: Dave Airlie <airlied@redhat.com> From: Ben Skeggs <skeggsb@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/CACAvsv7KmfcQqZcx+wh_1UKjTovp4PH_5UVMfeyxUu-M9WLZfw@mail.gmail.com
2018-12-13Merge tag 'drm/tegra/for-4.21-rc1' of git://anongit.freedesktop.org/tegra/linux into drm-nextDave Airlie20-39/+940
drm/tegra: Changes for v4.21-rc1 These changes contain a couple of minor fixes for host1x and the Falcon library in Tegra DRM. There are also a couple of missing pieces that finally enable support for host1x, VIC and display on Tegra194. I've also added a patch that enables audio over HDMI using the SOR which has been tested, and works, on both Tegra186 and Tegra194. Signed-off-by: Dave Airlie <airlied@redhat.com> From: Thierry Reding <thierry.reding@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181207134712.32683-1-thierry.reding@gmail.com
2018-12-13Merge branch 'drm-next-4.21' of git://people.freedesktop.org/~agd5f/linux into drm-nextDave Airlie98-865/+1686
[airlied: make etnaviv build again] amdgpu: - DC trace support - More DC documentation - XGMI hive reset support - Rework IH interaction with KFD - Misc fixes and cleanups - Powerplay updates for newer polaris variants - Add cursor plane update fast path - Enable gpu reset by default on CI parts - Fix config with KFD/HSA not enabled amdkfd: - Limit vram overcommit - dmabuf support - Support for doorbell BOs ttm: - Support for simultaneous submissions to multiple engines scheduler: - Add helpers for hw with preemption support Signed-off-by: Dave Airlie <airlied@redhat.com> From: Alex Deucher <alexdeucher@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181207233119.16861-1-alexander.deucher@amd.com
2018-12-13Merge branch 'mediatek-drm-fixes-4.20' of https://github.com/ckhu-mediatek/linux.git-tags into drm-fixesDave Airlie1-4/+7
Single bridge attachment fix. Signed-off-by: Dave Airlie <airlied@redhat.com> From: CK Hu <ck.hu@mediatek.com> Link: https://patchwork.freedesktop.org/patch/msgid/1544407975.18825.3.camel@mtksdaap41
2018-12-13Merge branch 'linux-4.20' of git://github.com/skeggsb/linux into drm-fixesDave Airlie2-11/+25
Three fixes: tegra regression fix display flushing fix mst cleanup fix. Signed-off-by: Dave Airlie <airlied@redhat.com> From: Ben Skeggs <skeggsb@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/CACAvsv7WCPzjQZonk+eS1FgEUKirz-4LOrVpMUVMM=D-GjbVpg@mail.gmail.com
2018-12-13Merge branch 'drm-fixes-4.20' of git://people.freedesktop.org/~agd5f/linux into drm-fixesDave Airlie8-9/+58
Fixes for 4.20: - Stability fixes for new polaris variants (e.g., RX590) - New vega pci ids - Vega20 smu fix - Ctx locking fix Signed-off-by: Dave Airlie <airlied@redhat.com> From: Alex Deucher <alexdeucher@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181212203022.3054-1-alexander.deucher@amd.com
2018-12-13Merge tag 'drm-misc-fixes-2018-12-12' of git://anongit.freedesktop.org/drm/drm-misc into drm-fixesDave Airlie1-6/+0
- rockchip: Revert change causing WARN on shutdown (Brian) Cc: Brian Norris <briannorris@chromium.org> Signed-off-by: Dave Airlie <airlied@redhat.com> From: Sean Paul <sean@poorly.run> Link: https://patchwork.freedesktop.org/patch/msgid/20181212204309.GA150523@art_vandelay
2018-12-13Merge tag 'drm-intel-fixes-2018-12-12-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixesDave Airlie8-97/+94
- Two fixes to avoid GPU hangs (on Braswell and Gen3) Signed-off-by: Dave Airlie <airlied@redhat.com> From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181212134010.GA18900@jlahtine-desk.ger.corp.intel.com
2018-12-13Merge tag 'drm-intel-fixes-2018-12-07' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixesDave Airlie9-204/+430
- Fix for system crash after GPU hang (Bugzilla #107945) - GVT fix for guest graphics corruption (https://github.com/intel/gvt-linux/issues/61) Signed-off-by: Dave Airlie <airlied@redhat.com> From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181207104352.GA18214@jlahtine-desk.ger.corp.intel.com
2018-12-12drm/amd/display: Add fast path for cursor plane updatesNicholas Kazlauskas2-2/+73
[Why] Legacy cursor plane updates from drm helpers go through the full atomic codepath. A high volume of cursor updates through this slow code path can cause subsequent page-flips to skip vblank intervals since each individual update is slow. This problem is particularly noticeable for the compton compositor. [How] A fast path for cursor plane updates is added by using DRM asynchronous commit support provided by async_check and async_update. These don't do a full state/flip_done dependency stall and they don't block other commit work. However, DC still expects itself to be single-threaded for anything that can issue register writes. Screen corruption or hangs can occur if write sequences overlap. Every call that potentially perform register writes needs to be guarded for asynchronous updates to work. The dc_lock mutex was added for this. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106175 Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Acked-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-12drm/amdgpu: drop fclk/gfxclk ratio settingEvan Quan1-1/+1
Since this is not needed any more on the latest SMC firmware. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Feifei Xu <Feifei.Xu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-12drm/amdgpu: Enable GPU recovery by default for CIAndrey Grodzovsky1-0/+2
I retested Bonaire (gfx7 dGPU) and it works fine. Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-12-12drm/vmwgfx: remove redundant return ret statementColin Ian King1-2/+0
The return statement is redundant as there is a return statement immediately before it so we have dead code that can be removed. Also remove the unused declaration of ret. Detected by CoverityScan, CID#1473793 ("Structurally dead code") Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2018-12-12drm/i915: Flush GPU relocs harder for gen3Chris Wilson2-9/+13
Adding an extra MI_STORE_DWORD_IMM to the gpu relocation path for gen3 was good, but still not good enough. To survive 24+ hours under test we needed to perform not one, not two but three extra store-dw. Doing so for each GPU relocation was a little unsightly and since we need to worry about userspace hitting the same issues, we should apply the dummy store-dw into the EMIT_FLUSH. Fixes: 7dd4f6729f92 ("drm/i915: Async GPU relocation processing") References: 7fa28e146994 ("drm/i915: Write GPU relocs harder with gen3") Testcase: igt/gem_tiled_fence_blits # blb/pnv Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181207134037.11848-1-chris@chris-wilson.co.uk (cherry picked from commit a889580c087a9cf91fddb3832ece284174214183) Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
2018-12-12drm/i915: Allocate a common scratch pageChris Wilson7-87/+75
Currently we allocate a scratch page for each engine, but since we only ever write into it for post-sync operations, it is not exposed to userspace nor do we care for coherency. As we then do not care about its contents, we can use one page for all, reducing our allocations and avoid complications by not assuming per-engine isolation. For later use, it simplifies engine initialisation (by removing the allocation that required struct_mutex!) and means that we can always rely on there being a scratch page. v2: Check that we allocated a large enough scratch for I830 w/a Fixes: 06e562e7f515 ("drm/i915/ringbuffer: Delay after EMIT_INVALIDATE for gen4/gen5") # v4.18.20 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108850 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181204141522.13640-1-chris@chris-wilson.co.uk Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: <stable@vger.kernel.org> # v4.18.20+ (cherry picked from commit 5179749925933575a67f9d8f16d0cc204f98a29f) [Joonas: Use new function in gen9_init_indirectctx_bb too] Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
2018-12-12drm/i915/execlists: Apply a full mb before execution for BraswellChris Wilson1-1/+6
Braswell is really picky about having our writes posted to memory before we execute or else the GPU may see stale values. A wmb() is insufficient as it only ensures the writes are visible to other cores, we need a full mb() to ensure the writes are in memory and visible to the GPU. The most frequent failure in flushing before execution is that we see stale PTE values and execute the wrong pages. References: 987abd5c62f9 ("drm/i915/execlists: Force write serialisation into context image vs execution") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: stable@vger.kernel.org Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20181206084431.9805-3-chris@chris-wilson.co.uk (cherry picked from commit 490b8c65b9db45896769e1095e78725775f47b3e) Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
2018-12-12drm/nouveau/kms: Fix memory leak in nv50_mstm_del()Lyude Paul1-0/+1
Noticed this while working on redoing the reference counting scheme in the DP MST helpers. Nouveau doesn't attempt to call drm_dp_mst_topology_mgr_destroy() at all, which leaves it leaking all of the resources for drm_dp_mst_topology_mgr and it's children mstbs+ports. Fixes: f479c0ba4a17 ("drm/nouveau/kms/nv50: initial support for DP 1.2 multi-stream") Signed-off-by: Lyude Paul <lyude@redhat.com> Cc: <stable@vger.kernel.org> # v4.10+ Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.