summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915
AgeCommit message (Collapse)AuthorFilesLines
2021-03-24drm/i915/selftests: Prepare client blit for obj->mm.lock removal.Maarten Lankhorst1-4/+4
Straightforward conversion, just convert a bunch of calls to unlocked versions. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-43-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915/selftests: Prepare huge_pages testcases for obj->mm.lock removal.Maarten Lankhorst1-7/+21
Straightforward conversion, just convert a bunch of calls to unlocked versions. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-42-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Use a single page table lock for each gtt.Maarten Lankhorst5-3/+56
We may create page table objects on the fly, but we may need to wait with the ww lock held. Instead of waiting on a freed obj lock, ensure we have the same lock for each object to keep -EDEADLK working. This ensures that i915_vma_pin_ww can lock the page tables when required. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-41-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Fix ww locking in shmem_create_from_objectMaarten Lankhorst1-1/+1
Quick fix, just use the unlocked version. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-40-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add missing ww lock in intel_dsb_prepare.Maarten Lankhorst1-1/+1
Because of the long lifetime of the mapping, we cannot wrap this in a simple limited ww lock. Just use the unlocked version of pin_map, because we'll likely release the mapping a lot later, in a different thread. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-39-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add ww locking to dma-buf ops, v2.Maarten Lankhorst1-28/+34
vmap is using pin_pages, but needs to use ww locking, add pin_pages_unlocked to correctly lock the mapping. Also add ww locking to begin/end cpu access. Changes since v1: - Fix i915_gem_map_dma_buf by using pin_pages_unlocked(). Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-38-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Lock ww in ucode objects correctlyMaarten Lankhorst6-5/+27
In the ucode functions, the calls are done before userspace runs, when debugging using debugfs, or when creating semi-permanent mappings; we can safely use the unlocked versions that does the ww dance for us. Because there is no pin_pages_unlocked yet, add it as convenience function. This removes possible lockdep splats about missing resv lock for ucode. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-37-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Increase ww locking for perf.Maarten Lankhorst1-13/+43
We need to lock a few more objects, some temporarily, add ww lock where needed. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-36-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add ww locking around vm_access()Maarten Lankhorst1-2/+22
i915_gem_object_pin_map potentially needs a ww context, so ensure we have one we can revoke. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-35-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add igt_spinner_pin() to allow for ww locking around spinner.Maarten Lankhorst2-46/+95
By default, we assume that it's called inside igt_create_request to keep existing selftests working, but allow for manual pinning when passing a ww context. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-34-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Prepare for obj->mm.lock removal, v2.Thomas Hellström3-7/+33
Stolen objects need to lock, and we may call put_pages when refcount drops to 0, ensure all calls are handled correctly. Changes since v1: - Rebase on top of upstream changes. Idea-from: Thomas Hellström <thomas.hellstrom@intel.com> Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-33-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Fix workarounds selftest, part 1Maarten Lankhorst3-27/+64
pin_map needs the ww lock, so ensure we pin both before submission. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Again pick older version just to side-step conflicts.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210128162612.927917-32-maarten.lankhorst@linux.intel.com Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-32-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Fix pread/pwrite to work with new locking rules.Maarten Lankhorst4-211/+114
We are removing obj->mm.lock, and need to take the reservation lock before we can pin pages. Move the pinning pages into the helper, and merge gtt pwrite/pread preparation and cleanup paths. The fence lock is also removed; it will conflict with fence annotations, because of memory allocations done when pagefaulting inside copy_*_user. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Pick the older version to avoid the conflicts] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210128162612.927917-31-maarten.lankhorst@linux.intel.com Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-31-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Defer pin calls in buffer pool until first use by caller.Maarten Lankhorst5-26/+35
We need to take the obj lock to pin pages, so wait until the callers have done so, before making the object unshrinkable. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-30-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Take obj lock around set_domain ioctlMaarten Lankhorst1-8/+10
We need to lock the object to move it to the correct domain, add the missing lock. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Pick version from an older patch series.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210128162612.927917-29-maarten.lankhorst@linux.intel.com Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-29-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Make __engine_unpark() compatible with ww locking.Maarten Lankhorst1-0/+4
Take the ww lock around engine_unpark. Because of the many many places where rpm is used, I chose the safest option and used a trylock to opportunistically take this lock for __engine_unpark. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-28-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Make lrc_init_wa_ctx compatible with ww locking, v3.Maarten Lankhorst1-11/+38
Make creation separate from pinning, in order to take the lock only once, and pin the mapping with the lock held. Changes since v1: - Rebase on top of upstream changes. Changes since v2: - Fully clear wa_ctx on error. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-27-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Take reservation lock around i915_vma_pin.Maarten Lankhorst4-14/+26
We previously complained when ww == NULL. This function is now only used in selftests to pin an object, and ww locking is now fixed. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Resolve conflict because we don't have a set-domain refactor, see https://lore.kernel.org/intel-gfx/20210203090205.25818-8-chris@chris-wilson.co.uk/ The really worrying thing here is that the above patch had a change in arguments for i915_gem_object_set_to_gtt_domain(), without any explanation. I decided to just faithfully apply Maarten's change but not the argument change which was in Maarten's context diff.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-26-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Move pinning to inside engine_wa_list_verify()Maarten Lankhorst7-10/+33
This should be done as part of the ww loop, in order to remove a i915_vma_pin that needs ww held. Now only i915_ggtt_pin() callers remaining. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-25-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add object locking to vm_fault_cpuMaarten Lankhorst1-0/+4
Take a simple lock so we hold ww around (un)pin_pages as needed. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-24-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Pass ww ctx to intel_pin_to_display_planeMaarten Lankhorst8-64/+86
Instead of multiple lockings, lock the object once, and perform the ww dance around attach_phys and pin_pages. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-23-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Rework clflush to work correctly without obj->mm.lock.Maarten Lankhorst1-8/+7
Pin in the caller, not in the work itself. This should also work better for dma-fence annotations. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-22-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Handle ww locking in init_status_pageMaarten Lankhorst1-13/+24
Try to pin to ggtt first, and use a full ww loop to handle eviction correctly. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-21-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Make ring submission compatible with obj->mm.lock removal, v2.Maarten Lankhorst1-66/+118
We map the initial context during first pin. This allows us to remove pin_map from state allocation, which saves us a few retry loops. We won't need this until first pin anyway. intel_ring_submission_setup() is also reworked slightly to do all pinning in a single ww loop. Changes since v1: - Handle -EDEADLK backoff in intel_ring_submission_setup() better. - Handle smatch errors reported by Dan and testbot. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reported-by: kernel test robot <lkp@intel.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-20-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Populate logical context during first pin.Maarten Lankhorst1-3/+23
This allows us to remove pin_map from state allocation, which saves us a few retry loops. We won't need this until first pin, anyway. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Resolve context conflict because we don't have the i915_scheduler.c extraction from the below patches set: https://lore.kernel.org/intel-gfx/20210203165259.13087-6-chris@chris-wilson.co.uk/] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-19-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Flatten obj->mm.lockMaarten Lankhorst6-47/+27
With userptr fixed, there is no need for all separate lockdep classes now, and we can remove all lockdep tricks used. A trylock in the shrinker is all we need now to flatten the locking hierarchy. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: Resolve conflict because we don't have the patch from Chris to rebrand i915_gem_shrinker_taints_mutex to fs_reclaim_taints_mutex. It's not a bad idea, but if we do it, it should be moved to the right header. See https://lore.kernel.org/intel-gfx/20210202154318.19246-1-chris@chris-wilson.co.uk/] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-18-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Fix userptr so we do not have to worry about obj->mm.lock, v7.Maarten Lankhorst8-584/+395
Instead of doing what we do currently, which will never work with PROVE_LOCKING, do the same as AMD does, and something similar to relocation slowpath. When all locks are dropped, we acquire the pages for pinning. When the locks are taken, we transfer those pages in .get_pages() to the bo. As a final check before installing the fences, we ensure that the mmu notifier was not called; if it is, we return -EAGAIN to userspace to signal it has to start over. Changes since v1: - Unbinding is done in submit_init only. submit_begin() removed. - MMU_NOTFIER -> MMU_NOTIFIER Changes since v2: - Make i915->mm.notifier a spinlock. Changes since v3: - Add WARN_ON if there are any page references left, should have been 0. - Return 0 on success in submit_init(), bug from spinlock conversion. - Release pvec outside of notifier_lock (Thomas). Changes since v4: - Mention why we're clearing eb->[i + 1].vma in the code. (Thomas) - Actually check all invalidations in eb_move_to_gpu. (Thomas) - Do not wait when process is exiting to fix gem_ctx_persistence.userptr. Changes since v5: - Clarify why check on PF_EXITING is (temporarily) required. Changes since v6: - Ensure userptr validity is checked in set_domain through a special path. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Acked-by: Dave Airlie <airlied@redhat.com> [danvet: s/kfree/kvfree/ in i915_gem_object_userptr_drop_ref in the previous review round, but which got lost. The other open questions around page refcount are imo better discussed in a separate series, with amdgpu folks involved]. Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-17-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Make compilation of userptr code depend on MMU_NOTIFIER.Maarten Lankhorst5-37/+31
Now that unsynchronized mappings are removed, the only time userptr works is when the MMU notifier is enabled. Put all of the userptr code behind a mmu notifier ifdef. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-16-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Reject UNSYNCHRONIZED for userptr, v2.Maarten Lankhorst1-8/+2
We should not allow this any more, as it will break with the new userptr implementation, it could still be made to work, but there's no point in doing so. Inspection of the beignet opencl driver shows that it's only used when normal userptr is not available, which means for new kernels you will need CONFIG_I915_USERPTR. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Acked-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Acked-by: Jason Ekstrand <jason@jlekstrand.net> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-15-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Reject more ioctls for userptr, v2.Maarten Lankhorst4-4/+19
There are a couple of ioctl's related to tiling and cache placement, that make no sense for userptr, reject those: - i915_gem_set_tiling_ioctl() Tiling should always be linear for userptr. Changing placement will fail with -ENXIO. - i915_gem_set_caching_ioctl() Userptr memory should always be cached. Changing caching mode will fail with -ENXIO. - i915_gem_set_domain_ioctl() Still temporarily allowed to work as intended, it's used to check userptr validity. With the reworked userptr code, it will keep working for this usecase. This plus the previous changes have been tested against beignet by using its own unit tests, and intel-video-compute by using piglit's opencl tests. Changes since v1: - set_domain was apparently used in iris for checking userptr validity, keep it working as intended. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Acked-by: Jason Ekstrand <jason@jlekstrand.net> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-14-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: No longer allow exporting userptr through dma-bufMaarten Lankhorst1-3/+2
It doesn't make sense to export a memory address, we will prevent allowing access this way to different address spaces when we rework userptr handling, so best to explicitly disable it. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Acked-by: Jason Ekstrand <jason@jlekstrand.net> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-13-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Disable userptr pread/pwrite support.Maarten Lankhorst2-0/+25
Userptr should not need the kernel for a userspace memcpy, userspace needs to call memcpy directly. Specifically, disable i915_gem_pwrite_ioctl() and i915_gem_pread_ioctl(). Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Acked-by: Jason Ekstrand <jason@jlekstrand.net> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-12-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Fix the GT fence revocation runtime PM logicImre Deak3-6/+41
To optimize some task deferring it until runtime resume unless someone holds a runtime PM reference (because in this case the task can be done w/o the overhead of runtime resume), we have to use the runtime PM get-if-active logic: If the runtime PM usage count is 0 (and so get-if-in-use would return false) the runtime suspend handler is not necessarily called yet (it could be just pending), so the device is not necessarily powered down, and so the runtime resume handler is not guaranteed to be called. The fence revocation depends on the above deferral, so add a get-if-active helper and use it during fence revocation. v2: - Add code comment explaining the fence reg programming deferral logic to i915_vma_revoke_fence(). (Chris) - Add Cc: stable and Fixes: tags. (Chris) - Fix the function docbook comment. Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: <stable@vger.kernel.org> # v4.12+ Fixes: 181df2d458f3 ("drm/i915: Take rpm wakelock for releasing the fence on unbind") Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210322204223.919936-1-imre.deak@intel.com (cherry picked from commit 9d58aa46291d4d696bb1eac3436d3118f7bf2573) Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2021-03-24drm/i915: make lockdep slightly happier about execbuf.Maarten Lankhorst4-19/+36
As soon as we install fences, we should stop allocating memory in order to prevent any potential deadlocks. This is required later on, when we start adding support for dma-fence annotations. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-11-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Convert i915_gem_object_attach_phys() to ww locking, v2.Maarten Lankhorst3-10/+22
Simple adding of i915_gem_object_lock, we may start to pass ww to get_pages() in the future, but that won't be the case here; We override shmem's get_pages() handling by calling i915_gem_object_get_pages_phys(), no ww is needed. Changes since v1: - Call shmem put pages directly, the callback would go down the phys free path. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-10-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Rework struct phys attachment handlingMaarten Lankhorst4-60/+78
Instead of creating a separate object type, we make changes to the shmem type, to clear struct page backing. This will allow us to ensure we never run into a race when we exchange obj->ops with other function pointers. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-9-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Move HAS_STRUCT_PAGE to obj->flagsMaarten Lankhorst21-52/+63
We want to remove the changing of ops structure for attaching phys pages, so we need to kill off HAS_STRUCT_PAGE from ops->flags, and put it in the bo. This will remove a potential race of dereferencing the wrong obj->ops without ww mutex held. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> [danvet: apply with wiggle] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-8-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add gem object locking to madvise.Maarten Lankhorst1-1/+7
Doesn't need the full ww lock, only checking if pages are bound. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> #irc Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-7-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Ensure we hold the object mutex in pin correctly.Maarten Lankhorst3-2/+14
Currently we have a lot of places where we hold the gem object lock, but haven't yet been converted to the ww dance. Complain loudly about those places. i915_vma_pin shouldn't have the obj lock held, so we can do a ww dance, while i915_vma_pin_ww should. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> #irc Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-6-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Add missing -EDEADLK handling to execbuf pinning, v2.Maarten Lankhorst1-11/+24
i915_vma_pin may fail with -EDEADLK when we start locking page tables, so ensure we handle this correctly. Changes since v1: - Drop -EDEADLK todo, this commit handles it. - Change eb_pin_vma from sort-of-bool + -EDEADLK to a proper int. (Matt) Cc: Matthew Brost <matthew.brost@intel.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-5-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Move cmd parser pinning to execbufferMaarten Lankhorst8-80/+142
We need to get rid of allocations in the cmd parser, because it needs to be called from a signaling context, first move all pinning to execbuf, where we already hold all locks. Allocate jump_whitelist in the execbuffer, and add annotations around intel_engine_cmd_parser(), to ensure we only call the command parser without allocating any memory, or taking any locks we're not supposed to. Because i915_gem_object_get_page() may also allocate memory, add a path to i915_gem_object_get_sg() that prevents memory allocations, and walk the sg list manually. It should be similarly fast. This has the added benefit of being able to catch all memory allocation errors before the point of no return, and return -ENOMEM safely to the execbuf submitter. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Acked-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-4-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Pin timeline map after first timeline pin, v4.Maarten Lankhorst5-44/+83
We're starting to require the reservation lock for pinning, so wait until we have that. Update the selftests to handle this correctly, and ensure pin is called in live_hwsp_rollover_user() and mock_hwsp_freelist(). Changes since v1: - Fix NULL + XX arithmatic, use casts. (kbuild) Changes since v2: - Clear entire cacheline when pinning. Changes since v3: - CACHELINE_BYTES -> TIMELINE_SEQNO_BYTES. (jekstrand) Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reported-by: kernel test robot <lkp@intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-3-maarten.lankhorst@linux.intel.com
2021-03-24drm/i915: Do not share hwsp across contexts any more, v8.Maarten Lankhorst11-417/+178
Instead of sharing pages with breadcrumbs, give each timeline a single page. This allows unrelated timelines not to share locks any more during command submission. As an additional benefit, seqno wraparound no longer requires i915_vma_pin, which means we no longer need to worry about a potential -EDEADLK at a point where we are ready to submit. Changes since v1: - Fix erroneous i915_vma_acquire that should be a i915_vma_release (ickle). - Extra check for completion in intel_read_hwsp(). Changes since v2: - Fix inconsistent indent in hwsp_alloc() (kbuild) - memset entire cacheline to 0. Changes since v3: - Do same in intel_timeline_reset_seqno(), and clflush for good measure. Changes since v4: - Use refcounting on timeline, instead of relying on i915_active. - Fix waiting on kernel requests. Changes since v5: - Bump amount of slots to maximum (256), for best wraparounds. - Add hwsp_offset to i915_request to fix potential wraparound hang. - Ensure timeline wrap test works with the changes. - Assign hwsp in intel_timeline_read_hwsp() within the rcu lock to fix a hang. Changes since v6: - Rename i915_request_active_offset to i915_request_active_seqno(), and elaborate the function. (tvrtko) Changes since v7: - Move hunk to where it belongs. (jekstrand) - Replace CACHELINE_BYTES with TIMELINE_SEQNO_BYTES. (jekstrand) Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Thomas Hellström <thomas.hellstrom@intel.com> #v1 Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323155059.628690-2-maarten.lankhorst@linux.intel.com
2021-03-24i915_vma: Rename vma_lookup to i915_vma_lookupLiam Howlett1-2/+2
Use i915 prefix to avoid name collision with future vma_lookup() in mm. Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com> Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20210323134208.3077275-1-Liam.Howlett@Oracle.com
2021-03-24drm/i915: Fix the GT fence revocation runtime PM logicImre Deak3-6/+41
To optimize some task deferring it until runtime resume unless someone holds a runtime PM reference (because in this case the task can be done w/o the overhead of runtime resume), we have to use the runtime PM get-if-active logic: If the runtime PM usage count is 0 (and so get-if-in-use would return false) the runtime suspend handler is not necessarily called yet (it could be just pending), so the device is not necessarily powered down, and so the runtime resume handler is not guaranteed to be called. The fence revocation depends on the above deferral, so add a get-if-active helper and use it during fence revocation. v2: - Add code comment explaining the fence reg programming deferral logic to i915_vma_revoke_fence(). (Chris) - Add Cc: stable and Fixes: tags. (Chris) - Fix the function docbook comment. Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: <stable@vger.kernel.org> # v4.12+ Fixes: 181df2d458f3 ("drm/i915: Take rpm wakelock for releasing the fence on unbind") Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210322204223.919936-1-imre.deak@intel.com
2021-03-24drm/i915/display: Simplify GLK display version testsMatt Roper18-84/+71
GLK has always been a bit of a special case since it reports INTEL_GEN() as 9, but has version 10 display IP. Now we can properly represent the display version as 10 and simplify the display generation tests throughout the display code. Aside from manually adding the version to the glk_info structure, the rest of this patch is generated with a Coccinelle semantic patch. Note that we also need to switch any code that matches gen10 today but *not* GLK to be CNL-specific: @@ expression dev_priv; @@ - DISPLAY_VER(dev_priv) > 9 + DISPLAY_VER(dev_priv) >= 10 @@ expression dev_priv, E; @@ ( - DISPLAY_VER(dev_priv) >= 10 && E + (DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv)) && E | - DISPLAY_VER(dev_priv) >= 10 + DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv) | - IS_DISPLAY_RANGE(dev_priv, 10, E) + IS_DISPLAY_RANGE(dev_priv, 11, E) || IS_CANNONLAKE(dev_priv) ) @@ expression dev_priv, E, E2; @@ ( - (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv)) + IS_DISPLAY_VER(dev_priv, 10) | - E || IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv) + E || IS_DISPLAY_VER(dev_priv, 10) | - (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) + IS_DISPLAY_VER(dev_priv, 10) | - IS_GEMINILAKE(dev_priv) || E || IS_CANNONLAKE(dev_priv) + E || IS_DISPLAY_VER(dev_priv, 10) | - E || IS_GEMINILAKE(dev_priv) || E2 || IS_CANNONLAKE(dev_priv) + E || E2 || IS_DISPLAY_VER(dev_priv, 10) | - (IS_DISPLAY_VER(dev_priv, 10) || IS_GEMINILAKE(dev_priv)) + IS_DISPLAY_VER(dev_priv, 10) | - (IS_GEMINILAKE(dev_priv) || IS_DISPLAY_VER(dev_priv, 10)) + IS_DISPLAY_VER(dev_priv, 10) ) @@ expression dev_priv; @@ - (IS_DISPLAY_VER(dev_priv, 9) && !IS_GEMINILAKE(dev_priv)) + IS_DISPLAY_VER(dev_priv, 9) @@ expression dev_priv; @@ ( - !(DISPLAY_VER(dev_priv) >= 11 || IS_DISPLAY_VER(dev_priv, 10)) + DISPLAY_VER(dev_priv) < 10 | - (DISPLAY_VER(dev_priv) >= 11 || IS_DISPLAY_VER(dev_priv, 10)) + DISPLAY_VER(dev_priv) >= 10 ) @@ expression dev_priv, E; @@ - E || DISPLAY_VER(dev_priv) >= 11 || IS_DISPLAY_VER(dev_priv, 10) + E || DISPLAY_VER(dev_priv) >= 10 @@ expression dev_priv, E; @@ - (IS_DISPLAY_RANGE(dev_priv, 11, E) || IS_DISPLAY_VER(dev_priv, 10)) + IS_DISPLAY_RANGE(dev_priv, 10, E) @@ expression dev_priv; @@ ( - DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv) || IS_GEN9_LP(dev_priv) + DISPLAY_VER(dev_priv) >= 10 || IS_GEN9_LP(dev_priv) | - IS_GEN9_LP(dev_priv) || DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv) + IS_GEN9_LP(dev_priv) || DISPLAY_VER(dev_priv) >= 10 ) @@ expression dev_priv, E; @@ - !(DISPLAY_VER(dev_priv) >= E) + DISPLAY_VER(dev_priv) < E v2: - Convert gen10 conditions that don't include GLK into CNL conditions. (Ville) v3: - Rework coccinelle rules so that "ver>=10" turns into "ver>=11||is_cnl." (Ville) v3.1: - Manually re-add the ".display.version = 10" to glk_info after regenerating patch via Coccinelle. v4: - Also apply cocci rules to intel_pm.c and i915_irq.c! (CI) Cc: Ville Syrjälä <ville.syrjala@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210322233840.4056851-1-matthew.d.roper@intel.com
2021-03-24drm/i915: Convert INTEL_GEN() to DISPLAY_VER() as appropriate in i915_irq.cMatt Roper1-33/+33
Convert the display-specific usage of INTEL_GEN, while leaving the non-display usage as-is for now. In the near-future we'll probably want to think about moving display interrupt handling to its own file under the display/ directory. v2: - Use new IS_DISPLAY_VER() macro. Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210320044245.3920043-6-matthew.d.roper@intel.com
2021-03-24drm/i915: Convert INTEL_GEN() to DISPLAY_VER() as appropriate in intel_pm.cMatt Roper1-66/+66
Although most of the code in this file is display-related (watermarks), there's some functions that are not (e.g., clock gating). Thus we need to do the conversions to DISPLAY_VER() manually here rather than using Coccinelle. In the near-future we'll probably want to think about moving watermark logic out of intel_pm.c and into watermark-specific files under the display/ directory. v2: - Use new IS_DISPLAY_VER macro where appropriate. Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210320044245.3920043-5-matthew.d.roper@intel.com
2021-03-24drm/i915/display: Eliminate most usage of INTEL_GEN()Matt Roper43-567/+568
Use Coccinelle to convert most of the usage of INTEL_GEN() and IS_GEN() in the display code to use DISPLAY_VER() comparisons instead. The following semantic patch was used: @@ expression dev_priv, E; @@ - INTEL_GEN(dev_priv) == E + IS_DISPLAY_VER(dev_priv, E) @@ expression dev_priv; @@ - INTEL_GEN(dev_priv) + DISPLAY_VER(dev_priv) @@ expression dev_priv; expression E; @@ - IS_GEN(dev_priv, E) + IS_DISPLAY_VER(dev_priv, E) @@ expression dev_priv; expression from, until; @@ - IS_GEN_RANGE(dev_priv, from, until) + IS_DISPLAY_RANGE(dev_priv, from, until) There are still some display-related uses of INTEL_GEN() in intel_pm.c (watermark code) and i915_irq.c. Those will be updated separately. v2: - Use new IS_DISPLAY_RANGE and IS_DISPLAY_VER helpers. (Jani) Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210320044245.3920043-4-matthew.d.roper@intel.com
2021-03-24drm/i915: Add DISPLAY_VER() and related macrosMatt Roper3-1/+8
Although we've long referred to platforms by a single "GEN" number, the hardware teams have recommended that we stop doing this since the various component IP blocks are going to start using independent number schemes with varying cadence. To support this, hardware platforms a bit down the road are going to start providing MMIO registers that the driver can read to obtain the "graphics version," "media version," and "display version" without needing to do a PCI ID -> platform -> version translation. Although our current platforms don't yet expose these registers (and the next couple we release probably won't have them yet either), the hardware teams would still like to see us move to this independent numbering scheme now in preparation. For i915 that means we should try to eliminate all usage of INTEL_GEN() throughout our code and instead replace it with separate GRAPHICS_VER(), MEDIA_VER(), and DISPLAY_VER() constructs in the code. For old platforms, these will all usually give the same value for each IP block (aside from a few special cases like GLK which we can no more accurately represent as graphics=9 + display=10), but future platforms will have more flexibility to bump IP version numbers independently. The upcoming ADL-P platform will have a display version of 13 and a graphics version of 12, so let's just the first step of breaking out DISPLAY_VER(), but leaving the rest of INTEL_GEN() untouched for now. For now we'll automatically derive the display version from the platform's INTEL_GEN() value except in cases where an alternative display version is explicitly provided in the device info structure. We also add some helper macros IS_DISPLAY_VER(i915, ver) and IS_DISPLAY_RANGE(i915, from, until) that match the behavior of the existing gen-based macros. However unlike IS_GEN(), we will implement those macros with direct comparisons rather than trying to maintain a mask to help compiler optimization. In practice the optimization winds up not being used in very many places (since the vast majority of our platform checks are of the form "gen >= x") so there is pretty minimal size reduction in the final driver binary[1]. We're also likely going to need to extend these version numbers to non-integer major.minor values at some point in the future, so the mask approach won't work at all once we get to platforms like that. [1] The results before/after the next patch in this series, which switches our code over to the new display macros: $ size i915.ko.{orig,new} text data bss dec hex filename 2940291 102944 5384 3048619 2e84ab i915.ko.orig 2940723 102956 5384 3049063 2e8667 i915.ko.new v2: - Move version into device info's display sub-struct. (Jani) - Add extra parentheses to macros. (Jani) - Note the lack of genmask optimization in the display-based macros and give size data. (Lucas) Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210320044245.3920043-3-matthew.d.roper@intel.com
generated by cgit v1.2.3 (git 2.39.0) at 2025-09-16 04:42:55 +0300