leandrof/linux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
2889caa923
linux/drivers/gpu/drm/i915
History
Chris Wilson 2889caa923 drm/i915: Eliminate lots of iterations over the execobjects array 
The major scaling bottleneck in execbuffer is the processing of the
execobjects. Creating an auxiliary list is inefficient when compared to
using the execobject array we already have allocated.

Reservation is then split into phases. As we lookup up the VMA, we
try and bind it back into active location. Only if that fails, do we add
it to the unbound list for phase 2. In phase 2, we try and add all those
objects that could not fit into their previous location, with fallback
to retrying all objects and evicting the VM in case of severe
fragmentation. (This is the same as before, except that phase 1 is now
done inline with looking up the VMA to avoid an iteration over the
execobject array. In the ideal case, we eliminate the separate reservation
phase). During the reservation phase, we only evict from the VM between
passes (rather than currently as we try to fit every new VMA). In
testing with Unreal Engine's Atlantis demo which stresses the eviction
logic on gen7 class hardware, this speed up the framerate by a factor of
2.

The second loop amalgamation is between move_to_gpu and move_to_active.
As we always submit the request, even if incomplete, we can use the
current request to track active VMA as we perform the flushes and
synchronisation required.

The next big advancement is to avoid copying back to the user any
execobjects and relocations that are not changed.

v2: Add a Theory of Operation spiel.
v3: Fall back to slow relocations in preparation for flushing userptrs.
v4: Document struct members, factor out eb_validate_vma(), add a few
more comments to explain some magic and hide other magic behind macros.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
2017-06-16 16:54:05 +01:00
..
gvt Merge tag 'gvt-next-2017-06-08' of https://github.com/01org/gvt-linux into drm-intel-next-queued 2017-06-16 10:03:01 +03:00
selftests drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
dvo_ch7xxx.c drm/i915/dvo: fix debug logging on unknown DID 2017-06-01 15:53:03 +03:00
dvo_ch7017.c drm/i915/lvds: Remove magic from PLL programming 2017-05-10 13:47:55 +03:00
dvo_ivch.c
dvo_ns2501.c
dvo_sil164.c
dvo_tfp410.c
dvo.h
i915_cmd_parser.c drm/i915: Redefine ptr_pack_bits() and friends 2017-05-17 13:38:04 +01:00
i915_debugfs.c drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_drv.c drm/i915: expose _SUBSLICE_MASK GETPARM 2017-06-14 12:31:57 -07:00
i915_drv.h drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
i915_gem_batch_pool.c drm/i915: Reinstate reservation_object zapping for batch_pool objects 2017-06-14 14:06:22 +01:00
i915_gem_batch_pool.h
i915_gem_clflush.c drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty 2017-06-16 14:52:27 +01:00
i915_gem_clflush.h drm/i915: Mark up clflushes as belonging to an unordered timeline 2017-05-03 11:08:45 +01:00
i915_gem_context.c drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_gem_context.h drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_gem_dmabuf.c drm/i915: Implement dma_buf_ops->kmap 2017-05-03 23:15:02 +01:00
i915_gem_evict.c drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
i915_gem_execbuffer.c drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
i915_gem_fence_reg.c
i915_gem_fence_reg.h
i915_gem_gtt.c drm/i915/guc: Assert that we switch between known ggtt->invalidate functions 2017-06-01 14:17:41 +01:00
i915_gem_gtt.h drm/i915: return the correct usable aperture size under gvt environment 2017-06-02 14:28:46 +01:00
i915_gem_internal.c drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty 2017-06-16 14:52:27 +01:00
i915_gem_object.h drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_gem_render_state.c
i915_gem_render_state.h
i915_gem_request.c drm/i915: Remove the spin-request during execbuf await_request 2017-06-08 12:33:09 +01:00
i915_gem_request.h Merge tag 'drm-intel-next-2017-05-29' of git://anongit.freedesktop.org/git/drm-intel into drm-next 2017-05-30 15:25:28 +10:00
i915_gem_shrinker.c drm/i915: Spin for struct_mutex inside shrinker 2017-06-14 10:55:11 +01:00
i915_gem_stolen.c drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty 2017-06-16 14:52:27 +01:00
i915_gem_tiling.c drm/i915: Fix logical inversion for gen4 quirking 2017-06-01 14:27:44 +01:00
i915_gem_timeline.c drm/i915: Squash repeated awaits on the same fence 2017-05-03 11:08:48 +01:00
i915_gem_timeline.h drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[] 2017-05-03 11:08:52 +01:00
i915_gem_userptr.c drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty 2017-06-16 14:52:27 +01:00
i915_gem.c drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_gem.h drm/i915: Squash repeated awaits on the same fence 2017-05-03 11:08:48 +01:00
i915_gpu_error.c drm/i915/execlists: Pack the count into the low bits of the port.request 2017-05-17 13:38:06 +01:00
i915_guc_reg.h
i915_guc_submission.c drm/i915/guc: Fix doorbell id selection 2017-05-31 10:34:03 +03:00
i915_ioc32.c
i915_irq.c drm/i915/cnp: Introduce Cannonpoint PCH. 2017-06-02 13:56:19 -07:00
i915_memcpy.c
i915_mm.c
i915_oa_bdw.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_bdw.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_oa_bxt.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_bxt.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_oa_chv.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_chv.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_oa_glk.c drm/i915/perf: add GLK support 2017-06-14 12:31:58 -07:00
i915_oa_glk.h drm/i915/perf: add GLK support 2017-06-14 12:31:58 -07:00
i915_oa_hsw.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_hsw.h drm/i915/perf: rework mux configurations queries 2017-06-14 12:31:57 -07:00
i915_oa_kblgt2.c drm/i915/perf: add KBL support 2017-06-14 12:31:58 -07:00
i915_oa_kblgt2.h drm/i915/perf: add KBL support 2017-06-14 12:31:58 -07:00
i915_oa_kblgt3.c drm/i915/perf: add KBL support 2017-06-14 12:31:58 -07:00
i915_oa_kblgt3.h drm/i915/perf: add KBL support 2017-06-14 12:31:58 -07:00
i915_oa_sklgt2.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_sklgt2.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_oa_sklgt3.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_sklgt3.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_oa_sklgt4.c drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT 2017-06-14 12:31:57 -07:00
i915_oa_sklgt4.h drm/i915/perf: Add 'render basic' Gen8+ OA unit configs 2017-06-14 12:31:57 -07:00
i915_params.c
i915_params.h
i915_pci.c drm/i915/DMC/CNL: Load DMC on CNL 2017-06-12 09:44:34 -07:00
i915_perf.c drm/i915/perf: add GLK support 2017-06-14 12:31:58 -07:00
i915_pvinfo.h drm/i915: Fix GVT-g PVINFO version compatibility check 2017-06-12 14:39:24 +03:00
i915_reg.h drm/i915/perf: Add OA unit support for Gen 8+ 2017-06-14 12:31:57 -07:00
i915_selftest.h
i915_suspend.c
i915_sw_fence.c drm/i915: Import the kfence selftests for i915_sw_fence 2017-05-17 13:38:02 +01:00
i915_sw_fence.h drm/i915: Remove kref from i915_sw_fence 2017-05-17 13:38:01 +01:00
i915_syncmap.c drm/i915: Squash repeated awaits on the same fence 2017-05-03 11:08:48 +01:00
i915_syncmap.h drm/i915: Squash repeated awaits on the same fence 2017-05-03 11:08:48 +01:00
i915_sysfs.c drm/i915: Do not leak dev_priv->l3_parity.remap_info[] 2017-05-02 12:09:13 +03:00
i915_trace_points.c
i915_trace.h drm/i915: Add g4x watermark tracepoint 2017-05-10 16:48:32 +03:00
i915_utils.h drm/i915: Store a direct lookup from object handle to vma 2017-06-16 16:54:04 +01:00
i915_vgpu.c drm/i915: Fix GVT-g PVINFO version compatibility check 2017-06-12 14:39:24 +03:00
i915_vgpu.h
i915_vma.c drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
i915_vma.h drm/i915: Eliminate lots of iterations over the execobjects array 2017-06-16 16:54:05 +01:00
intel_acpi.c
intel_atomic_plane.c Merge tag 'drm-intel-next-2017-05-29' of git://anongit.freedesktop.org/git/drm-intel into drm-next 2017-05-30 15:25:28 +10:00
intel_atomic.c drm/i915/cnl: Fix Cannonlake scaler mode programing. 2017-06-12 09:45:55 -07:00
intel_audio.c drm/i915: Reorganize intel_lpe_audio_notify() arguments 2017-05-03 16:20:48 +03:00
intel_bios.c
intel_bios.h
intel_breadcrumbs.c drm/i915: Skip adding the request to the signal tree is complete 2017-06-08 12:33:08 +01:00
intel_cdclk.c drm/i915/cnl: Allow dynamic cdclk changes on CNL 2017-06-12 09:41:33 -07:00
intel_color.c
intel_crt.c drm/i915: Convert intel_crt connector properties to atomic. 2017-04-12 10:53:22 +02:00
intel_csr.c drm/i915: Use HAS_CSR instead of gen number on DMC load. 2017-06-12 09:45:30 -07:00
intel_ddi.c drm/i915/cfl: Basic DDI plumbing for Coffee Lake. 2017-06-12 09:48:58 -07:00
intel_device_info.c drm/i915/cfl: Introduce Coffee Lake platform definition. 2017-06-08 09:55:08 -07:00
intel_display.c drm/i915: Remove pipe A quirk remnants 2017-06-15 15:38:27 +03:00
intel_dp_aux_backlight.c drm/i915: Don't enable backlight at setup time. 2017-06-15 16:11:08 +03:00
intel_dp_link_training.c drm/i915: Implement Link Rate fallback on Link training failure 2017-04-13 21:57:37 +03:00
intel_dp_mst.c Merge tag 'drm-intel-next-2017-05-29' of git://anongit.freedesktop.org/git/drm-intel into drm-next 2017-05-30 15:25:28 +10:00
intel_dp.c drm/i915/cnl: Implement voltage swing sequence. 2017-06-12 09:44:00 -07:00
intel_dpio_phy.c
intel_dpll_mgr.c drm/i915/cnl: make function cnl_ddi_dp_set_dpll_hw_state static 2017-06-15 15:53:56 +03:00
intel_dpll_mgr.h drm/i915/cnl: Initialize PLLs 2017-06-12 09:42:18 -07:00
intel_drv.h drm/i915: Add i830 "pipes power well" 2017-06-15 15:35:38 +03:00
intel_dsi_dcs_backlight.c drm/i915: Pass atomic state to backlight enable/disable/set callbacks. 2017-06-12 16:06:28 +02:00
intel_dsi_pll.c
intel_dsi_vbt.c drm/i915/glk: Calculate high/low switch count for GLK 2017-05-15 18:29:46 +03:00
intel_dsi.c drm/i915/glk: Add cold boot sequence for GLK DSI 2017-06-15 23:03:57 +03:00
intel_dsi.h
intel_dvo.c drm/i915: Convert intel DVO connector to atomic 2017-04-12 10:53:29 +02:00
intel_engine_cs.c drm/i915: Check the ring is empty when declaring the engines are idle 2017-06-01 14:22:16 +01:00
intel_fbc.c drm/i915: Fix 90/270 rotated coordinates for FBC 2017-06-06 19:31:28 +03:00
intel_fbdev.c drm: Add DRM_MODE_ROTATE_ and DRM_MODE_REFLECT_ to UAPI 2017-05-22 09:49:48 +02:00
intel_fifo_underrun.c drm/i915/cnl: Enable fifo underrun for Cannonlake. 2017-06-12 09:46:08 -07:00
intel_frontbuffer.c
intel_frontbuffer.h
intel_guc_ct.c drm/i915/guc: Introduce buffer based cmd transport 2017-05-26 13:26:53 +01:00
intel_guc_ct.h drm/i915/guc: Introduce buffer based cmd transport 2017-05-26 13:26:53 +01:00
intel_guc_fwif.h drm/i915/guc: Introduce buffer based cmd transport 2017-05-26 13:26:53 +01:00
intel_guc_loader.c drm/i915/guc: Load GuC on Coffee Lake 2017-06-09 11:56:53 -07:00
intel_guc_log.c drm/i915: Treat WC a separate cache domain 2017-04-12 12:35:17 +01:00
intel_gvt.c drm/i915/gvt: Return -EIO if host GuC submission is enabled when loading GVT-g 2017-05-30 16:00:07 +03:00
intel_gvt.h drm/i915/gvt: Add gvt options sanitize function 2017-05-30 15:59:47 +03:00
intel_hangcheck.c drm/i915: Avoid format string expansion from engine names 2017-04-19 15:49:27 +03:00
intel_hdmi.c drm/i915/cnp: add CNP gmbus support 2017-06-02 13:59:32 -07:00
intel_hotplug.c drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v4. 2017-04-06 21:29:23 +02:00
intel_huc.c drm/i915/huc: Load HuC on Coffee Lake 2017-06-09 11:57:16 -07:00
intel_i2c.c drm/i915/cnp: add CNP gmbus support 2017-06-02 13:59:32 -07:00
intel_lpe_audio.c drm/i915: Fix runtime PM for LPE audio 2017-05-15 14:44:11 +03:00
intel_lrc.c drm/i915/perf: Add OA unit support for Gen 8+ 2017-06-14 12:31:57 -07:00
intel_lrc.h drm/i915: Sanitize engine context sizes 2017-04-28 12:11:59 +03:00
intel_lspcon.c
intel_lvds.c drm/i915: Pass crtc_state and connector state to backlight enable/disable functions 2017-06-12 16:05:45 +02:00
intel_mocs.c drm/i915/cnl: Cannonlake has same MOCS table than Skylake. 2017-06-07 07:29:51 -07:00
intel_mocs.h
intel_modes.c
intel_opregion.c drm/i915: Pass connector state to intel_panel_set_backlight_acpi 2017-06-12 16:06:10 +02:00
intel_overlay.c drm/i915: Remove pipe A quirk remnants 2017-06-15 15:38:27 +03:00
intel_panel.c drm/i915: Pass atomic state to backlight enable/disable/set callbacks. 2017-06-12 16:06:28 +02:00
intel_pipe_crc.c drm/i915: use memdup_user_nul 2017-05-08 09:28:39 +02:00
intel_pm.c Revert "drm/i915/skl: New ddb allocation algorithm" 2017-06-15 14:01:14 -07:00
intel_psr.c drm/i915/psr: disable psr2 for resolution greater than 32X20 2017-05-30 10:11:41 -07:00
intel_renderstate_gen6.c
intel_renderstate_gen7.c
intel_renderstate_gen8.c
intel_renderstate_gen9.c
intel_renderstate.h
intel_ringbuffer.c drm/i915: Micro-optimise hotpath through intel_ring_begin() 2017-05-04 15:40:38 +01:00
intel_ringbuffer.h drm/i915: Split execlist priority queue into rbtree + linked list 2017-05-17 13:38:09 +01:00
intel_runtime_pm.c drm/i915: Add i830 "pipes power well" 2017-06-15 15:35:38 +03:00
intel_sdvo_regs.h
intel_sdvo.c drm/i915: Actually attach the tv_format property to the SDVO connector 2017-06-16 15:46:57 +03:00
intel_sideband.c
intel_sprite.c drm/i915: Workaround VLV/CHV DSI scanline counter hardware fail 2017-06-06 17:58:42 +03:00
intel_tv.c drm/i915: Convert intel_tv connector properties to atomic, v5. 2017-04-12 10:53:22 +02:00
intel_uc.c drm/i915/guc: Clear enable_guc_loading in case of init failure 2017-06-08 12:21:19 +03:00
intel_uc.h drm/i915/guc: capture GuC logs if FW fails to load 2017-05-26 13:59:56 +01:00
intel_uncore.c drm/i915: Remove decoupled MMIO code 2017-05-30 09:59:41 +03:00
intel_uncore.h drm/i915: Keep the forcewake timer alive for 1ms past the most recent use 2017-05-26 15:58:21 +01:00
intel_vbt_defs.h
Kconfig
Kconfig.debug Merge tag 'drm-intel-next-2017-05-29' of git://anongit.freedesktop.org/git/drm-intel into drm-next 2017-05-30 15:25:28 +10:00
Makefile drm/i915/perf: add GLK support 2017-06-14 12:31:58 -07:00