summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/amd/amdgpu
Commit message (Collapse)AuthorAgeFilesLines
* drm/amdgpu: clean up UVD instance handling v2Christian König2018-07-192-67/+64
| | | | | | | | | | | | | | The whole handle, filp and entity handling is superfluous here. We should have reviewed that more thoughtfully. It looks like somebody just made the code instance aware without knowing the background. v2: fix one more missed case in amdgpu_uvd_suspend Reviewed-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove superflous UVD encode entityChristian König2018-07-193-27/+0
| | | | | | | | | Not sure what that was every used for, but now it is completely unused. Reviewed-by: Leo Liu <leo.liu@amd.com> Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu/display: Replace CONFIG_DRM_AMD_DC_DCN1_0 with CONFIG_X86Michel Dänzer2018-07-191-1/+1
| | | | | | | | | Allowing CONFIG_DRM_AMD_DC_DCN1_0 to be disabled on X86 was an opportunity for display with Raven Ridge accidentally not working. Reviewed-by: Harry Wentland <harry.wentland@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu/pm: Remove VLA usageKees Cook2018-07-181-58/+42
| | | | | | | | | | | In the quest to remove all stack VLA usage from the kernel[1], this uses the maximum sane buffer size and removes copy/paste code. [1] https://lkml.kernel.org/r/CA+55aFzCG-zNmZwX4A2FQpadafLfEzK6CC=qPXydAacU1RqZWA@mail.gmail.com Reviewed-by: Rex Zhu <rezhu@amd.com> Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: simplify the bo reference on amdgpu_bo_updateHuang Rui2018-07-181-4/+3
| | | | | | | | | | | BO ptr already be initialized at definition, we needn't use the complicated reference. v2: fix typo at subject line Signed-off-by: Huang Rui <ray.huang@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove internal/unused kernel module parametersSonny Jiang2018-07-182-15/+0
| | | | | | | | Remove internal/unused kernel module parameters Signed-off-by: Sonny Jiang <sonny.jiang@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: change ring priority after pushing the job (v2)Christian König2018-07-182-3/+11
| | | | | | | | | | | Pushing a job can change the ring assignment of an entity. v2: squash in: "drm/amdgpu: fix job priority handling" (Christian) 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>
* drm/amdgpu: allow for more flexible priority handlingChristian König2018-07-181-1/+2
| | | | | | | | | Allow to call amdgpu_ring_priority_get() after pushing the ring to the scheduler. 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>
* drm/amdgpu/gfx9: Update golden settings for vg10.Feifei Xu2018-07-181-1/+12
| | | | | | | | Add some UTCL registers' golden settings. Signed-off-by: Feifei Xu <Feifei.Xu@amd.com> Tested-by: Kevin Wang <Kevin1.Wang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: minor cleanup in amdgpu_job.cChristian König2018-07-171-7/+4
| | | | | | | | | | Remove superflous NULL check, fix coding style a bit, shorten error messages. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove job->adev (v2)Christian König2018-07-172-8/+11
| | | | | | | | | | | We can get that from the ring. v2: squash in "drm/amdgpu: always initialize job->base.sched" (Alex) Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: add amdgpu_job_submit_direct helperChristian König2018-07-168-70/+48
| | | | | | | | | Make sure that we properly initialize at least the sched member. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove job->ringChristian König2018-07-167-28/+29
| | | | | | | | | We can easily get that from the scheduler. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove ring parameter from amdgpu_job_submitChristian König2018-07-169-21/+20
| | | | | | | | | We know the ring through the entity anyway. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: remove fence context from the jobChristian König2018-07-164-4/+1
| | | | | | | | | Can be obtained directly from the fence as well. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: cleanup job headerChristian König2018-07-162-45/+75
| | | | | | | | | | Move job related defines, structure and function declarations to amdgpu_job.h Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Allow to create BO lists in CS ioctl v3Andrey Grodzovsky2018-07-164-40/+107
| | | | | | | | | | | | | | | | | | | | | | | This change is to support MESA performace optimization. Modify CS IOCTL to allow its input as command buffer and an array of buffer handles to create a temporay bo list and then destroy it when IOCTL completes. This saves on calling for BO_LIST create and destry IOCTLs in MESA and by this improves performance. v2: Avoid inserting the temp list into idr struct. v3: Remove idr alloation from amdgpu_bo_list_create. Remove useless argument from amdgpu_cs_parser_fini Minor cosmetic stuff. v4: Revert amdgpu_bo_list_destroy back to static Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu/vi: fix mixed up state in smu clockgating setupAlex Deucher2018-07-131-6/+6
| | | | | | | | Use the PP_STATE_SUPPORT_* rather than AMD_CG_SUPPORT_* when communicating with the SMU. Reviewed-by: Huang Rui <ray.huang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Warn and update pin_size values when destroying a pinned BOMichel Dänzer2018-07-131-7/+25
| | | | | | | | | | | | This shouldn't happen, but if it does, we'll get a backtrace of the caller, and update the pin_size values as needed. v2: * Check bo->pin_count instead of placement flags (Christian König) Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Make pin_size values atomicMichel Dänzer2018-07-134-21/+23
| | | | | | | | | | | | | | Concurrent execution of the non-atomic arithmetic could result in completely bogus values. v2: * Rebased on v2 of the previous patch Cc: stable@vger.kernel.org Bugzilla: https://bugs.freedesktop.org/106872 Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Keep track of amount of pinned CPU visible VRAMMichel Dänzer2018-07-135-19/+14
| | | | | | | | | | | | | | | Instead of CPU invisible VRAM. Preparation for the following, no functional change intended. v2: * Also change amdgpu_vram_mgr_bo_invisible_size to amdgpu_vram_mgr_bo_visible_size, allowing further simplification (Christian König) Cc: stable@vger.kernel.org Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/scheduler: modify args of drm_sched_entity_initNayan Deshmukh2018-07-137-14/+11
| | | | | | | | | | replace run queue by a list of run queues and remove the sched arg as that is part of run queue itself Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com> Reviewed-by: Christian König <christian.koenig@amd.com> Acked-by: Eric Anholt <eric@anholt.net> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: fix TTM move entity init orderChristian König2018-07-131-16/+21
| | | | | | | | | | | | We are initializing the entity before the scheduler is actually initialized. This can lead to all kind of problem, but especially NULL pointer deref because of Nayan's scheduler work. Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amd: Use newly added interrupt source defs for SOC15.Andrey Grodzovsky2018-07-136-12/+24
| | | | | | | Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amd: Use newly added interrupt source defs for VI v3.Andrey Grodzovsky2018-07-1311-23/+41
| | | | | | | | | | v2: Rebase v3: Use defines for CP_SQ and CP_ECC_ERROR interrupts. Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amd/powerplay: add vega12 SMU gfxoff support v3Evan Quan2018-07-131-0/+5
| | | | | | | | | | | Export apis for enabling/disabling SMU gfxoff support. v2: fit the latest gfxoff support framework v3: add feature_mask control Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Huang Rui <ray.huang at amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: reduce the idle period that RLC has to wait before request CGCGEvan Quan2018-07-131-4/+7
| | | | | | | | | Gfxoff feature may depends on the CGCG(on vega12, that's the case). This change will help to enable gfxoff feature more frequently. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: no touch for the reserved bit of RLC_CGTT_MGCG_OVERRIDEEvan Quan2018-07-131-4/+11
| | | | | | | | On vega12, the bit0 of RLC_CGTT_MGCG_OVERRIDE is reserved. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: drop mmRLC_PG_CNTL clear v2Evan Quan2018-07-131-3/+0
| | | | | | | | | | | | SMU owns this register so the driver should not set it to avoid breaking gfxoff. v2: update description Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Alex Deucher <alexander.deucher at amd.com> Reviewed-by: Huang Rui <ray.huang at amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: correct rlc save restore list initialization for v2_1Evan Quan2018-07-131-6/+12
| | | | | | | | | The save restore list initialization does not have to be pg guarded. And for some asic(e.g. Vega12), it does not have cntl/gpm/srm lists. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Huang Rui <ray.huang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: init CSIB regardless of rlc version and pg statusEvan Quan2018-07-131-1/+2
| | | | | | | | | CSIB init has no relation with rlc version and pg status. It should be needed regardless of them. Signed-off-by: Evan Quan <evan.quan@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: pin the csb buffer on hw init v2Evan Quan2018-07-131-0/+40
| | | | | | | | | | | | Without this pin, the csb buffer will be filled with inconsistent data after S3 resume. And that will causes gfx hang on gfxoff exit since this csb will be executed then. v2: fit amdgpu_bo_pin change(take one less argument) Signed-off-by: Evan Quan <evan.quan@amd.com> Reviewed-by: Huang Rui <ray.huang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: get VCN start to process in the dpm disabled caseLeo Liu2018-07-101-1/+1
| | | | | | | | | Fixes: 22cc6c5e19 (drm/amdgpu: Add runtime VCN PG support) Signed-off-by: Leo Liu <leo.liu@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: move cache window setup after power and clock resumeLeo Liu2018-07-101-2/+2
| | | | | | | | | To make register read/write reliable Signed-off-by: Leo Liu <leo.liu@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Take VCN jpeg ring into account in idle work handlerLeo Liu2018-07-101-0/+2
| | | | | | | | | VCN won't get power off when only jpeg active Signed-off-by: Leo Liu <leo.liu@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Verify root PD is mapped into kernel address space (v4)Andrey Grodzovsky2018-07-101-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | Problem: When PD/PT update made by CPU root PD was not yet mapped causing page fault. Fix: Verify root PD is mapped into CPU address space. v2: Make sure that we add the root PD to the relocated list since then it's get mapped into CPU address space bt default in amdgpu_vm_update_directories. v3: Drop change to not move kernel type BOs to evicted list. v4: Remove redundant bo move to relocated list. Link: https://bugs.freedesktop.org/show_bug.cgi?id=107065 Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Present amdgpu_task_info in VM_FAULTS.Andrey Grodzovsky2018-07-103-5/+18
| | | | | | | | | | | Extract and present the reposnsible process and thread when VM_FAULT happens. v2: Use getter and setter functions. Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Acked-by: Jim Qu <Jim.Qu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Add support for logging process info in amdgpu_vm.Andrey Grodzovsky2018-07-102-0/+55
| | | | | | | | | | | | | Add process and thread names and pids and a function to extract this info from relevant amdgpu_vm. v2: Add documentation and fix identation. v3: Add getter and setter functions for amdgpu_task_info. Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Acked-by: Jim Qu <Jim.Qu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Add CLK IP base offsetRex Zhu2018-07-102-0/+2
| | | | | | | | | so we can read/write the registers in CLK domain through RREG32/WREG32_SOC15 Reviewed-by: Evan Quan <evan.quan@amd.com> Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: Reserve VM root shared fence slot for command submission (v3)Michel Dänzer2018-07-101-0/+4
| | | | | | | | | | | | | | | | | | Without this, there could not be enough slots, which could trigger the BUG_ON in reservation_object_add_shared_fence. v2: * Jump to the error label instead of returning directly (Jerry Zhang) v3: * Reserve slots for command submission after VM updates (Christian König) Cc: stable@vger.kernel.org Bugzilla: https://bugs.freedesktop.org/106418 Reported-by: mikhail.v.gavrilov@gmail.com Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: fix kmap error handling for bo creationsJunwei Zhang2018-07-101-1/+1
| | | | | | | | kmap happens after bo pin, so unpin is required on error Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: allocate gart memory when it's required (v3)Junwei Zhang2018-07-106-8/+46
| | | | | | | | | | | | | Instead of calling gart address space on every bo pin, allocates it on demand v2: fix error handling v3: drop the change on amdgpu_amdkfd_gpuvm.c, not needed. Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Felix Kuehling <Felix.Kuehling@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: separate gpu address from bo pinJunwei Zhang2018-07-1016-56/+48
| | | | | | | | | It could be got by amdgpu_bo_gpu_offset() if need Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
* Merge branch 'drm-next-4.19' of git://people.freedesktop.org/~agd5f/linux ↵Dave Airlie2018-07-1044-404/+866
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | into drm-next More features for 4.19: - Use core pcie functionality rather than duplicating our own for pcie gens and lanes - Scheduler function naming cleanups - More documentation - Reworked DC/Powerplay interfaces to improve power savings - Initial stutter mode support for RV (power feature) - Vega12 powerplay updates - GFXOFF fixes - Misc fixes Signed-off-by: Dave Airlie <airlied@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180705221447.2807-1-alexander.deucher@amd.com
| * drm/amdgpu: update documentation for amdgpu_drv.cSonny Jiang2018-07-051-7/+225
| | | | | | | | | | | | | | | | Signed-off-by: Sonny Jiang <sonny.jiang@amd.com> Acked-by: Junwei Zhang <Jerry.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| * drm/amd: Remove errors from sphinx documentationDarren Powell2018-07-052-10/+15
| | | | | | | | | | | | | | | | | | Eliminating the warnings produced by sphinx when processing the sphinx comments in amdgpu_device.c & amdgpu_mn.c Signed-off-by: Darren Powell <darren.powell@amd.com> Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| * drm/amdgpu: use pcie functions for link width and speedAlex Deucher2018-07-054-31/+65
| | | | | | | | | | | | | | | | | | | | | | Use the newly exported pci functions to get the link width and speed rather than using the drm duplicated versions. Also query the GPU link caps directly rather than hardcoding them. Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| * drm/amdgpu: switch firmware path for SI partsAlex Deucher2018-07-053-46/+46
| | | | | | | | | | | | | | | | | | Use separate firmware path for amdgpu to avoid conflicts with radeon on SI parts. Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| * drm/amdgpu: switch firmware path for CIK parts (v2)Alex Deucher2018-07-057-73/+70
| | | | | | | | | | | | | | | | | | | | | | Use separate firmware path for amdgpu to avoid conflicts with radeon on CIK parts. v2: squash in logic simplification (Alex) Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| * drm/amdgpu: fix user fence write race conditionNicolai Hähnle2018-07-051-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The buffer object backing the user fence is reserved using the non-user fence, i.e., as soon as the non-user fence is signaled, the user fence buffer object can be moved or even destroyed. Therefore, emit the user fence first. Both fences have the same cache invalidation behavior, so this should have no user-visible effect. Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
OpenPOWER on IntegriCloud