linux_dsm_epyc7002/drivers/gpu/drm/i915/gem
Chris Wilson 16e8745967 drm/i915/gt: Move the batch buffer pool from the engine to the gt
Since the introduction of 'soft-rc6', we aim to park the device quickly
and that results in frequent idling of the whole device. Currently upon
idling we free the batch buffer pool, and so this renders the cache
ineffective for many workloads. If we want to have an effective cache of
recently allocated buffers available for reuse, we need to decouple that
cache from the engine powermanagement and make it timer based. As there
is no reason then to keep it within the engine (where it once made
retirement order easier to track), we can move it up the hierarchy to the
owner of the memory allocations.

v2: Hook up to debugfs/drop_caches to clear the cache on demand.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200430111819.10262-2-chris@chris-wilson.co.uk
2020-04-30 19:12:02 +01:00
..
selftests drm/i915/selftests: Add tiled blits selftest 2020-04-30 08:31:12 +01:00
i915_gem_busy.c drm/i915: Align engine->uabi_class/instance with i915_drm.h 2020-01-20 09:13:01 +00:00
i915_gem_clflush.c drm/i915: Hold reference to intel_frontbuffer as we track activity 2019-12-18 12:09:57 +00:00
i915_gem_clflush.h
i915_gem_client_blt.c drm/i915/gt: Move the batch buffer pool from the engine to the gt 2020-04-30 19:12:02 +01:00
i915_gem_client_blt.h
i915_gem_context_types.h drm/i915/gem: use spinlock_t instead of struct spinlock 2020-02-19 15:08:49 +02:00
i915_gem_context.c drm/i915: Only close vma we open 2020-04-24 11:24:45 +01:00
i915_gem_context.h drm/i915/perf: introduce global sseu pinning 2020-03-17 15:27:55 +02:00
i915_gem_dmabuf.c drm/i915: Skip CPU synchronisation on dmabuf attachments 2020-02-11 00:16:15 +00:00
i915_gem_domain.c drm/i915/gem: Remove object_is_locked assertion from unpin_from_display_plane 2020-04-20 16:23:24 +01:00
i915_gem_execbuffer.c drm/i915/gt: Move the batch buffer pool from the engine to the gt 2020-04-30 19:12:02 +01:00
i915_gem_fence.c
i915_gem_internal.c drm/i915: significantly reduce the use of <drm/i915_drm.h> 2020-02-27 08:35:09 +02:00
i915_gem_ioctls.h drm/i915: Introduce DRM_I915_GEM_MMAP_OFFSET 2019-12-04 15:11:44 +00:00
i915_gem_lmem.c drm/i915/gem: Support discontiguous lmem object maps 2020-01-03 11:26:01 +00:00
i915_gem_lmem.h drm/i915/gem: Support discontiguous lmem object maps 2020-01-03 11:26:01 +00:00
i915_gem_mman.c drm/i915/gem: Mark up the racy read of the mmap_singleton 2020-03-11 12:00:02 +00:00
i915_gem_mman.h drm/i915: Introduce DRM_I915_GEM_MMAP_OFFSET 2019-12-04 15:11:44 +00:00
i915_gem_object_blt.c drm/i915/gt: Move the batch buffer pool from the engine to the gt 2020-04-30 19:12:02 +01:00
i915_gem_object_blt.h drm/i915/gt: Move the batch buffer pool from the engine to the gt 2020-04-30 19:12:02 +01:00
i915_gem_object_types.h drm/i915/gem: Drop cached obj->bind_count 2020-04-02 01:17:39 +01:00
i915_gem_object.c drm/i915: Only close vma we open 2020-04-24 11:24:45 +01:00
i915_gem_object.h drm/i915: significantly reduce the use of <drm/i915_drm.h> 2020-02-27 08:35:09 +02:00
i915_gem_pages.c drm/i915/gem: Drop cached obj->bind_count 2020-04-02 01:17:39 +01:00
i915_gem_phys.c drm/i915/phys: unconditionally call release_memory_region 2020-03-06 10:05:17 +00:00
i915_gem_pm.c drm/i915/gem: Make WARN* drm specific where drm_priv ptr is available 2020-01-22 17:52:39 +02:00
i915_gem_pm.h
i915_gem_region.c drm/i915/gem: Single page objects are naturally contiguous 2020-01-02 21:21:17 +00:00
i915_gem_region.h
i915_gem_shmem.c drm/i915/gem: Make WARN* drm specific where drm_priv ptr is available 2020-01-22 17:52:39 +02:00
i915_gem_shrinker.c drm/i915/gem: Drop cached obj->bind_count 2020-04-02 01:17:39 +01:00
i915_gem_shrinker.h
i915_gem_stolen.c drm/i915/stolen: prefer struct drm_device based logging 2020-04-08 13:49:35 +03:00
i915_gem_stolen.h drm/i915/display: Be explicit in handling the preallocated vma 2020-02-05 09:20:39 +00:00
i915_gem_throttle.c
i915_gem_tiling.c drm/i915/gem: Hold obj->vma.lock over for_each_ggtt_vma() 2020-04-22 15:43:56 +01:00
i915_gem_userptr.c drm/i915: significantly reduce the use of <drm/i915_drm.h> 2020-02-27 08:35:09 +02:00
i915_gem_wait.c
i915_gemfs.c
i915_gemfs.h