April 2017 Archives by author
Starting: Tue Apr 4 06:59:10 UTC 2017
Ending: Sat Apr 29 15:04:50 UTC 2017
Messages: 1801
- [PATCH 01/15] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/15] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/15] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/15] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/15] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/15] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/15] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/15] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/15] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/15] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/15] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/15] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/15] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/15] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 15/15] HAX: let's not play with fire just yet...
Matthew Auld
- [PATCH 01/15] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/15] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/15] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/15] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/15] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/15] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/15] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/15] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/15] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/15] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/15] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/15] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/15] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/15] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 15/15] HAX: let's not play with fire just yet...
Matthew Auld
- [PATCH 01/14] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/14] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/14] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/14] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/14] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/14] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/14] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/14] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/14] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/14] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/14] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/14] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/14] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/14] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 01/15] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/15] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/15] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/15] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/15] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/15] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/15] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/15] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/15] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/15] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/15] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/15] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/15] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/15] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 15/15] hmmm
Matthew Auld
- [PATCH 01/16] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/16] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/16] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/16] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/16] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/16] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/16] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/16] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/16] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/16] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/16] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/16] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/16] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/16] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 15/16] hmmm
Matthew Auld
- [PATCH 16/16] maybe...
Matthew Auld
- [PATCH 01/13] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/13] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/13] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/13] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/13] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/13] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/13] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/13] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/13] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/13] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/13] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/13] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 13/13] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 01/14] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/14] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/14] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/14] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/14] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/14] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/14] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/14] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/14] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/14] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/14] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/14] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 13/14] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 14/14] hmm
Matthew Auld
- [PATCH 01/14] drm/i915: add page_size_mask to dev_info
Matthew Auld
- [PATCH 02/14] drm/i915: introduce gtt page size
Matthew Auld
- [PATCH 03/14] drm/i915: consider page size when doing insertion
Matthew Auld
- [PATCH 04/14] drm/i915: use the max sg->length if the dma_map fails
Matthew Auld
- [PATCH 05/14] mm/shmem: tweak the huge-page interface
Matthew Auld
- [PATCH 06/14] drm/i915: enable huge-pages through shmemfs
Matthew Auld
- [PATCH 07/14] drm/i915: pass page_size to insert_entries
Matthew Auld
- [PATCH 08/14] drm/i915: support inserting 64K pages in the ppgtt
Matthew Auld
- [PATCH 09/14] drm/i915: support inserting 2M pages in the ppgtt
Matthew Auld
- [PATCH 10/14] drm/i915: support inserting 1G pages in the ppgtt
Matthew Auld
- [PATCH 11/14] drm/i915: disable GTT cache for huge-pages
Matthew Auld
- [PATCH 12/14] drm/i915/selftests: modify the gtt tests to also exercise huge pages
Matthew Auld
- [PATCH 13/14] drm/i915/debugfs: include some huge-page metrics
Matthew Auld
- [PATCH 14/14] drm/i915: enable platform level support for huge GTT pages
Matthew Auld
- [PATCH 00/10] Enable OA unit for Gen 8 and 9 in i915 perf
Robert Bragg
- [PATCH 01/10] drm/i915/perf: improve invalid OA format debug message
Robert Bragg
- [PATCH 02/10] drm/i915/perf: better pipeline aged/aging tail updates
Robert Bragg
- [PATCH 03/10] drm/i915/perf: rate limit spurious oa report notice
Robert Bragg
- [PATCH 04/10] drm/i915: expose _SLICE_MASK GETPARM
Robert Bragg
- [PATCH 05/10] drm/i915: expose _SUBSLICE_MASK GETPARM
Robert Bragg
- [PATCH 06/10] drm/i915/perf: Add 'render basic' Gen8+ OA unit configs
Robert Bragg
- [PATCH 07/10] drm/i915/perf: Add OA unit support for Gen 8+
Robert Bragg
- [PATCH 08/10] drm/i915/perf: Add more OA configs for BDW, CHV, SKL + BXT
Robert Bragg
- [PATCH 09/10] drm/i915/perf: per-gen timebase for checking sample freq
Robert Bragg
- [PATCH 10/10] drm/i915/perf: remove perf.hook_lock
Robert Bragg
- [PATCH 0/8] drm: Fix/remove a few static checker error
Imre Deak
- [PATCH 1/8] drm/i915/vlv: Fix port B PLL opamp initialization
Imre Deak
- [PATCH 2/8] drm/i915/dp: Check error return during DPCD capability queries
Imre Deak
- [PATCH 3/8] drm/i915/sdvo: Check error return from intel_sdvo_get_value()
Imre Deak
- [PATCH 4/8] drm/i915: Check error return when setting DMA mask
Imre Deak
- [PATCH 5/8] drm/i915: Check error return when converting pipe to connector
Imre Deak
- [PATCH 6/8] drm/i915: Sanitize stolen memory size calculation
Imre Deak
- [PATCH 7/8] drm/i915/lvds: Remove magic from PLL programming
Imre Deak
- [PATCH 8/8] drm: Remove redundant NULL check during atomic plane commit
Imre Deak
- [PATCH] drm/i915: Modify spinlocks while reading pipe crc
Mika Kahola
- [PATCH] drm/i915: Modify spinlocks while reading pipe crc
Mika Kahola
- [PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB
Kamble, Sagar A
- [PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB
Kamble, Sagar A
- [PATCH 1/1] drm/i915: Suspend GuC during GEM Suspend prior to GPU Reset
Sagar Arun Kamble
- [PATCH 1/2] drm/i915: Suspend GuC during GEM Suspend prior to GPU Reset
Sagar Arun Kamble
- [PATCH 2/2] HAX enable guc submission for CI
Sagar Arun Kamble
- [PATCH v8 01/21] drm/i915/debugfs: Create generic string tokenize function and update CRC control parsing
Sagar Arun Kamble
- [PATCH v8 02/21] drm/i915/gen9+: Separate RPS and RC6 handling
Sagar Arun Kamble
- [PATCH v8 03/21] drm/i915/slpc: Add has_slpc capability flag
Sagar Arun Kamble
- [PATCH v8 04/21] drm/i915/slpc: Add enable_slpc module parameter
Sagar Arun Kamble
- [PATCH v8 05/21] drm/i915/slpc: Sanitize GuC version
Sagar Arun Kamble
- [PATCH v8 06/21] drm/i915/slpc: Lay out SLPC init/enable/disable/cleanup helpers
Sagar Arun Kamble
- [PATCH v8 07/21] drm/i915/slpc: Enable SLPC in GuC if supported
Sagar Arun Kamble
- [PATCH v8 08/21] drm/i915/slpc: Add SLPC communication interfaces
Sagar Arun Kamble
- [PATCH v8 09/21] drm/i915/slpc: Add parameter set/unset/get, task control/status functions
Sagar Arun Kamble
- [PATCH v8 10/21] drm/i915/slpc: Allocate/Release/Initialize SLPC shared data
Sagar Arun Kamble
- [PATCH v8 11/21] drm/i915/slpc: Send RESET event to enable SLPC
Sagar Arun Kamble
- [PATCH v8 12/21] drm/i915/slpc: Send SHUTDOWN event
Sagar Arun Kamble
- [PATCH v8 13/21] drm/i915/slpc: Add support for min/max frequency control
Sagar Arun Kamble
- [PATCH v8 14/21] drm/i915/slpc: Add debugfs support to read/write/revert the parameters
Sagar Arun Kamble
- [PATCH v8 15/21] drm/i915/slpc: Add enable/disable controls for SLPC tasks
Sagar Arun Kamble
- [PATCH v8 16/21] drm/i915/slpc: Add i915_slpc_info to debugfs
Sagar Arun Kamble
- [PATCH v8 17/21] drm/i915/slpc: Add SLPC banner to RPS debugfs interfaces.
Sagar Arun Kamble
- [PATCH v8 18/21] drm/i915/slpc: Add SKL SLPC Support
Sagar Arun Kamble
- [PATCH v8 19/21] drm/i915/slpc: Add Broxton SLPC support
Sagar Arun Kamble
- [PATCH v8 20/21] drm/i915/slpc: Add Kabylake SLPC support
Sagar Arun Kamble
- [PATCH v8 21/21] drm/i915/slpc: Enable SLPC, where supported
Sagar Arun Kamble
- [PATCH v2 1/2] mm: Handle prefault for size = 2GB
Sagar Arun Kamble
- [PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB
Sagar Arun Kamble
- [PATCH 0/8] Implement DDB algorithm and WM cleanup
Mahesh Kumar
- [PATCH 1/8] drm/i915/skl+: calculate pixel_rate & relative_data_rate in fixed point
Mahesh Kumar
- [PATCH 2/8] drm/i915/skl+: use linetime latency if ddb size is not available
Mahesh Kumar
- [PATCH 3/8] drm/i915/skl: Fail the flip if no FB for WM calculation
Mahesh Kumar
- [PATCH 4/8] drm/i915/skl+: no need to memset again
Mahesh Kumar
- [PATCH 5/8] drm/i915/skl+: ddb min requirement may exceed allocation
Mahesh Kumar
- [PATCH 6/8] drm/i915/skl+: Watermark calculation cleanup
Mahesh Kumar
- [PATCH 7/8] drm/i915/skl: New ddb allocation algorithm
Mahesh Kumar
- [PATCH 8/8] drm/i915/skl+: consider max supported plane pixel rate while scaling
Mahesh Kumar
- [PATCH 0/8] Implement DDB algorithm and WM cleanup
Mahesh Kumar
- [PATCH 1/8] drm/i915/skl+: calculate pixel_rate & relative_data_rate in fixed point
Mahesh Kumar
- [PATCH 2/8] drm/i915/skl+: use linetime latency if ddb size is not available
Mahesh Kumar
- [PATCH 3/8] drm/i915/skl: Fail the flip if no FB for WM calculation
Mahesh Kumar
- [PATCH 4/8] drm/i915/skl+: no need to memset again
Mahesh Kumar
- [PATCH 5/8] drm/i915/skl+: ddb min requirement may exceed allocation
Mahesh Kumar
- [PATCH 6/8] drm/i915/skl+: Watermark calculation cleanup
Mahesh Kumar
- [PATCH 7/8] drm/i915/skl: New ddb allocation algorithm
Mahesh Kumar
- [PATCH 8/8] drm/i915/skl+: consider max supported plane pixel rate while scaling
Mahesh Kumar
- [PATCH 01/16] drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v2.
Maarten Lankhorst
- [PATCH 02/16] Revert "drm/i915: Lock mode_config.mutex in intel_display_resume."
Maarten Lankhorst
- [PATCH 03/16] drm/i915: Do not use lock all in hsw_trans_edp_pipe_A_crc_wa
Maarten Lankhorst
- [PATCH 04/16] drm/i915: Remove unused members from intel_tv.c
Maarten Lankhorst
- [PATCH 05/16] drm/i915: Convert intel_dp_mst connector properties to atomic.
Maarten Lankhorst
- [PATCH 06/16] drm/i915: Convert intel_crt connector properties to atomic.
Maarten Lankhorst
- [PATCH 07/16] drm/i915: Convert intel_tv connector properties to atomic, v4.
Maarten Lankhorst
- [PATCH 08/16] drm/i915: Convert intel DVO connector to atomic
Maarten Lankhorst
- [PATCH 09/16] drm/i915: Add plumbing for digital connector state
Maarten Lankhorst
- [PATCH 10/16] drm/i915: Convert DSI connector properties to atomic.
Maarten Lankhorst
- [PATCH 11/16] drm/i915: Convert LVDS connector properties to atomic.
Maarten Lankhorst
- [PATCH 12/16] drm/i915: Make intel_dp->has_audio reflect hw state only
Maarten Lankhorst
- [PATCH 13/16] drm/i915: Convert intel_dp properties to atomic.
Maarten Lankhorst
- [PATCH 14/16] drm/i915: Convert intel_hdmi connector properties to atomic
Maarten Lankhorst
- [PATCH 15/16] drm/i915: Handle force_audio correctly in intel_sdvo
Maarten Lankhorst
- [PATCH 16/16] drm/i915: Convert intel_sdvo connector properties to atomic.
Maarten Lankhorst
- [PATCH 1/5] drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v4.
Maarten Lankhorst
- [PATCH 1/5] drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v4.
Maarten Lankhorst
- [PATCH] drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v4.
Maarten Lankhorst
- [PATCH 0/5] Classify the engines in class + instance
Oscar Mateo
- [PATCH 1/5] drm/i915: Classify the engines in class + instance
Oscar Mateo
- [PATCH 2/5] drm/i915: Use the same vfunc for BSD2 ring init
Oscar Mateo
- [PATCH 3/5] drm/i915: Generate the engine name based on the instance number
Oscar Mateo
- [PATCH 4/5] drm/i915: Split the engine info table in two levels, using class + instance
Oscar Mateo
- [PATCH 5/5] drm/i915: Use the engine class to get the context size
Oscar Mateo
- [PATCH 0/5] Classify the engines in class + instance
Oscar Mateo
- [PATCH 1/5] drm/i915: Classify the engines in class + instance
Oscar Mateo
- [PATCH 2/5] drm/i915: Use the same vfunc for BSD2 ring init
Oscar Mateo
- [PATCH 3/5] drm/i915: Generate the engine name based on the instance number
Oscar Mateo
- [PATCH 4/5] drm/i915: Split the engine info table in two levels, using class + instance
Oscar Mateo
- [PATCH 5/5] drm/i915: Use the engine class to get the context size
Oscar Mateo
- [PATCH 0/5] Classify the engines in class + instance
Oscar Mateo
- [PATCH 1/5] drm/i915: Classify the engines in class + instance
Oscar Mateo
- [PATCH 2/5] drm/i915: Use the same vfunc for BSD2 ring init
Oscar Mateo
- [PATCH 3/5] drm/i915: Generate the engine name based on the instance number
Oscar Mateo
- [PATCH 4/5] drm/i915: Split the engine info table in two levels, using class + instance
Oscar Mateo
- [PATCH 5/5] drm/i915: Use the engine class to get the context size
Oscar Mateo
- [PATCH 0/5] Classify the engines in class + instance (v3)
Oscar Mateo
- [PATCH 1/5] drm/i915: Classify the engines in class + instance
Oscar Mateo
- [PATCH 2/5] drm/i915: Use the same vfunc for BSD2 ring init
Oscar Mateo
- [PATCH 3/5] drm/i915: Generate the engine name based on the instance number
Oscar Mateo
- [PATCH 4/5] drm/i915: Split the engine info table in two levels, using class + instance
Oscar Mateo
- [PATCH 5/5] drm/i915: Use the engine class to get the context size
Oscar Mateo
- [PATCH v6 1/5] drm: Add driver-private objects to atomic state
Dhinakaran Pandiyan
- [PATCH v6 2/5] drm/dp: Introduce MST topology state to track available link bandwidth
Dhinakaran Pandiyan
- [PATCH v6 3/5] drm/dp: Add DP MST helpers to atomically find and release vcpi slots
Dhinakaran Pandiyan
- [PATCH v6 4/5] drm: Connector helper function to release resources
Dhinakaran Pandiyan
- [PATCH v6 5/5] drm/dp: Track MST link bandwidth
Dhinakaran Pandiyan
- [PATCH v7 1/4] drm: Add driver-private objects to atomic state
Dhinakaran Pandiyan
- [PATCH v7 2/4] drm/dp: Introduce MST topology state to track available link bandwidth
Dhinakaran Pandiyan
- [PATCH v7 3/4] drm/dp: Add DP MST helpers to atomically find and release vcpi slots
Dhinakaran Pandiyan
- [PATCH v7 4/4] drm/dp: Track MST link bandwidth
Dhinakaran Pandiyan
- ✗ Fi.CI.BAT: warning for series starting with [1/1] drm/i915: Suspend GuC during GEM Suspend prior to GPU Reset
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/2] drm/i915: Suspend GuC during GEM Suspend prior to GPU Reset
Patchwork
- ✗ Fi.CI.BAT: failure for hmm
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [v8,01/21] drm/i915/debugfs: Create generic string tokenize function and update CRC control parsing
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/16] drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v2.
Patchwork
- ✗ Fi.CI.BAT: failure for hmm2
Patchwork
- ✗ Fi.CI.BAT: failure for Classify the engines in class + instance
Patchwork
- ✗ Fi.CI.BAT: failure for Classify the engines in class + instance (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [v6,1/5] drm: Add driver-private objects to atomic state
Patchwork
- ✓ Fi.CI.BAT: success for Classify the engines in class + instance (rev3)
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/24] drm/i915: Capture dmc firmware information before reset
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/23] drm/i915: Capture dmc firmware information before reset
Patchwork
- ✓ Fi.CI.BAT: success for Enable OA unit for Gen 8 and 9 in i915 perf
Patchwork
- ✗ Fi.CI.BAT: failure for Implement DDB algorithm and WM cleanup (rev5)
Patchwork
- ✗ Fi.CI.BAT: failure for Implement DDB algorithm and WM cleanup (rev6)
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/20] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Patchwork
- ✓ Fi.CI.BAT: success for drm/atomic: Acquire connection_mutex lock in drm_helper_probe_single_connector_modes, v4.
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/20] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Patchwork
- ✓ Fi.CI.BAT: success for Classify the engines in class + instance (rev4)
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [v2,1/2] mm: Handle prefault for size = 2GB
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/3] dma-fence: Reserve 0 as a special NO_CONTEXT token
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/3] dma-fence: Reserve 0 as a special NO_CONTEXT token
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/5] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/21] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/26] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/9] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/6] drm/i915: Stop second guessing the caller for intel_uncore_wait_for_register()
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/12] drm/i915: Rename intel_engine_cs.exec_id to uabi_id
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/33] drm/i915: Rename intel_engine_cs.exec_id to uabi_id
Patchwork
- ✗ Fi.CI.BAT: warning for simplify
Patchwork
- ✗ Fi.CI.BAT: warning for simplify (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for simplify (rev3)
Patchwork
- ✗ Fi.CI.BAT: warning for guc run
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/28] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Mark CPU cache as dirty on every transition for CPU writes (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/25] MT: HACK: Fix nouveau compilation fix
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/26] MT: HACK: Fix nouveau compilation fix
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/25] MT: HACK: Fix nouveau compilation fix
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/28] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✗ Fi.CI.BAT: warning for series starting with [01/25] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/27] drm/i915/selftests: Allocate inode/file dynamically
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/27] drm/i915/selftests: Allocate inode/file dynamically
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/3] drm/i915/selftests: Allocate inode/file dynamically
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/28] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [v7,1/4] drm: Add driver-private objects to atomic state
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Modify spinlocks while reading pipe crc
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Modify spinlocks while reading pipe crc (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Modify spinlocks while reading pipe crc (rev2)
Patchwork
- ✗ Fi.CI.BAT: warning for series starting with [01/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/27] ilk-ci-hax
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/27] ilk-ci-hax
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/27] ilk-ci-hax
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/27] ilk-ci-hax
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/28] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/29] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/31] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/15] drm/i915: add page_size_mask to dev_info
Patchwork
- ✗ Fi.CI.BAT: failure for drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/15] drm/i915: add page_size_mask to dev_info
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/2] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/5] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/14] drm/i915: add page_size_mask to dev_info
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/31] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/15] drm/i915: add page_size_mask to dev_info
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/16] drm/i915: add page_size_mask to dev_info
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Modify spinlocks while reading pipe crc (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Modify spinlocks while reading pipe crc (rev2)
Patchwork
- ✓ Fi.CI.BAT: success for test v7 engine-reset
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/13] drm/i915: add page_size_mask to dev_info
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/14] drm/i915: add page_size_mask to dev_info
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/14] drm/i915: add page_size_mask to dev_info
Patchwork
- ✓ Fi.CI.BAT: success for drm/i915: Compute the ring->space slightly less pessimistically
Patchwork
- ✗ Fi.CI.BAT: warning for drm: Fix/remove a few static checker error
Patchwork
- ✓ Fi.CI.BAT: success for test v7 engine-reset (rev2)
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/30] drm/i915: Avoid the branch in computing intel_ring_space()
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [1/3] drm/i915: Avoid the branch in computing intel_ring_space()
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/14] drm/i915: Avoid the branch in computing intel_ring_space()
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/30] drm/i915: Avoid the branch in computing intel_ring_space()
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/12] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✗ Fi.CI.BAT: failure for series starting with [01/36] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✗ Fi.CI.BAT: warning for series starting with [01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- ✓ Fi.CI.BAT: success for series starting with [01/32] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Patchwork
- [PATCH 01/24] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 02/24] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 03/24] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 04/24] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 05/24] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 06/24] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 07/24] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 08/24] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 09/24] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 10/24] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 11/24] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 12/24] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 13/24] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 14/24] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 15/24] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 16/24] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 17/24] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 18/24] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 19/24] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 20/24] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 21/24] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 22/24] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 23/24] HAX enable guc submission for CI
Michel Thierry
- [PATCH 24/24] HAK: extra info needed
Michel Thierry
- [PATCH 01/23] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 02/23] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 03/23] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 04/23] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 05/23] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 06/23] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 07/23] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 08/23] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 09/23] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 10/23] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 11/23] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 12/23] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 13/23] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 14/23] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 15/23] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 16/23] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 17/23] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 18/23] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 19/23] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 20/23] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 21/23] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 22/23] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 23/23] TEST: drm/i915: mutex-less reset engine
Michel Thierry
- [PATCH 01/20] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 02/20] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 03/20] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 04/20] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 05/20] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 06/20] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 07/20] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 08/20] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 09/20] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 10/20] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 11/20] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 12/20] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 13/20] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 14/20] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 15/20] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 16/20] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 17/20] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 18/20] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 19/20] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 20/20] TEST: drm/i915: mutex-less reset engine
Michel Thierry
- [PATCH 01/20] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 02/20] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 03/20] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 04/20] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 05/20] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 06/20] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 07/20] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 08/20] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 09/20] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 10/20] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 11/20] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 12/20] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 13/20] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 14/20] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 15/20] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 16/20] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 17/20] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 18/20] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 19/20] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 20/20] TEST: drm/i915: mutex-less reset engine
Michel Thierry
- [PATCH 01/25] MT: HACK: Fix nouveau compilation fix
Michel Thierry
- [PATCH 02/25] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 03/25] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 04/25] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 05/25] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 06/25] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 07/25] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 08/25] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 09/25] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 10/25] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 11/25] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 12/25] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 13/25] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 14/25] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 15/25] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 16/25] drm/i915/guc: fix mmio whitelist mmio_start offset and add reminder
Michel Thierry
- [PATCH 17/25] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 18/25] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 19/25] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 20/25] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 21/25] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 22/25] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 23/25] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 24/25] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 25/25] SQUASHME: drm/i915: mutex-less reset engine
Michel Thierry
- [PATCH 01/26] MT: HACK: Fix nouveau compilation fix
Michel Thierry
- [PATCH 02/26] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 03/26] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 04/26] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 05/26] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 06/26] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 07/26] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 08/26] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 09/26] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 10/26] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 11/26] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 12/26] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 13/26] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 14/26] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 15/26] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 16/26] drm/i915/guc: fix mmio whitelist mmio_start offset and add reminder
Michel Thierry
- [PATCH 17/26] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 18/26] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 19/26] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 20/26] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 21/26] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 22/26] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 23/26] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 24/26] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 25/26] SQUASHME: drm/i915: mutex-less reset engine
Michel Thierry
- [PATCH 26/26] HAX enable guc submission for CI
Michel Thierry
- [PATCH 01/25] MT: HACK: Fix nouveau compilation fix
Michel Thierry
- [PATCH 02/25] drm/i915/guc: Fix sleep under spinlock during reset
Michel Thierry
- [PATCH 03/25] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 04/25] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 05/25] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 06/25] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 07/25] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 08/25] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 09/25] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 10/25] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 11/25] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 12/25] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 13/25] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 14/25] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 15/25] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 16/25] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 17/25] drm/i915/guc: fix mmio whitelist mmio_start offset and add reminder
Michel Thierry
- [PATCH 18/25] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 19/25] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 20/25] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 21/25] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 22/25] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 23/25] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 24/25] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 25/25] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 00/26] test v7 engine-reset
Michel Thierry
- [PATCH 01/26] drm/i915/guc: Fix sleep under spinlock during reset
Michel Thierry
- [PATCH 02/26] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 03/26] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 04/26] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 05/26] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 06/26] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 07/26] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 08/26] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 09/26] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 10/26] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 11/26] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 12/26] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 13/26] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 14/26] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 15/26] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 16/26] squashme: drm/i915: cancel reset if couldnt find an active request
Michel Thierry
- [PATCH 17/26] drm/i915/guc: fix mmio whitelist mmio_start offset and add reminder
Michel Thierry
- [PATCH 18/26] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 19/26] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 20/26] drm/i915: Print guc request restart after reset
Michel Thierry
- [PATCH 21/26] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 22/26] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 23/26] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 24/26] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 25/26] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 26/26] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH 00/25] test v7 engine-reset
Michel Thierry
- [PATCH 01/25] drm/i915/guc: Fix sleep under spinlock during reset
Michel Thierry
- [PATCH 02/25] drm/i915: Capture dmc firmware information before reset
Michel Thierry
- [PATCH 03/25] drm/i915: Include GuC fw version in error state
Michel Thierry
- [PATCH 04/25] drm/i915: Include HuC fw version in error state
Michel Thierry
- [PATCH 05/25] drm/i915: Fix stale comment about I915_RESET_IN_PROGRESS flag
Michel Thierry
- [PATCH 06/25] drm/i915: Rename gen8_(un)request_engine_reset to gen8_reset_engine_start/cancel
Michel Thierry
- [PATCH 07/25] drm/i915: Update i915.reset to handle engine resets
Michel Thierry
- [PATCH 08/25] drm/i915/tdr: Modify error handler for per engine hang recovery
Michel Thierry
- [PATCH 09/25] drm/i915/tdr: Add support for per engine reset recovery
Michel Thierry
- [PATCH 10/25] drm/i915: Skip reset request if there is one already
Michel Thierry
- [PATCH 11/25] drm/i915/tdr: Add engine reset count to error state
Michel Thierry
- [PATCH 12/25] drm/i915/tdr: Export per-engine reset count info to debugfs
Michel Thierry
- [PATCH 13/25] drm/i915/tdr: Enable Engine reset and recovery support
Michel Thierry
- [PATCH 14/25] drm/i915: Add engine reset count in get-reset-stats ioctl
Michel Thierry
- [PATCH 15/25] drm/i915/selftests: reset engine self tests
Michel Thierry
- [PATCH 16/25] squashme: drm/i915: cancel reset if couldnt find an active request
Michel Thierry
- [PATCH 17/25] drm/i915/guc: fix mmio whitelist mmio_start offset and add reminder
Michel Thierry
- [PATCH 18/25] drm/i915/guc: Provide register list to be saved/restored during engine reset
Michel Thierry
- [PATCH 19/25] drm/i915/guc: Add support for reset engine using GuC commands
Michel Thierry
- [PATCH 20/25] drm/i915: Watchdog timeout: Pass GuC shared data structure during param load
Michel Thierry
- [PATCH 21/25] drm/i915: Watchdog timeout: IRQ handler for gen8+
Michel Thierry
- [PATCH 22/25] drm/i915: Watchdog timeout: Ringbuffer command emission for gen8+
Michel Thierry
- [PATCH 23/25] drm/i915: Watchdog timeout: DRM kernel interface to set the timeout
Michel Thierry
- [PATCH 24/25] drm/i915: Watchdog timeout: Include threshold value in error state
Michel Thierry
- [PATCH 25/25] drm/i915: Watchdog timeout: Export media reset count from GuC to debugfs
Michel Thierry
- [PATCH] simplify
Tvrtko Ursulin
- [PATCH] simplify
Tvrtko Ursulin
- [PATCH] simplify
Tvrtko Ursulin
- [PATCH] guc run
Tvrtko Ursulin
- [PATCH] hmm
Chris Wilson
- [PATCH] hmm2
Chris Wilson
- [PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB
Chris Wilson
- [PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB
Chris Wilson
- [PATCH 1/3] dma-fence: Reserve 0 as a special NO_CONTEXT token
Chris Wilson
- [PATCH 2/3] drm/i915: Mark up clflushes as belonging to an unordered timeline (NO_CONTEXT)
Chris Wilson
- [PATCH 3/3] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 1/3] dma-fence: Reserve 0 as a special NO_CONTEXT token
Chris Wilson
- [PATCH 2/3] drm/i915: Mark up clflushes as belonging to an unordered timeline (NO_CONTEXT)
Chris Wilson
- [PATCH 3/3] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 1/3] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 2/3] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 3/3] common-tl-sync
Chris Wilson
- [PATCH 1/5] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 2/5] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 3/5] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 4/5] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 5/5] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 01/21] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/21] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/21] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/21] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/21] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/21] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/21] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/21] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/21] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/21] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/21] drm/i915: Combine write_domain flushes to a single function
Chris Wilson
- [PATCH 12/21] drm/i915: Treat WC a separate cache domain
Chris Wilson
- [PATCH 13/21] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/21] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 15/21] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 16/21] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 17/21] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 18/21] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 19/21] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 20/21] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 21/21] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 01/26] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/26] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/26] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/26] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/26] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/26] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/26] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/26] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/26] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/26] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/26] drm/i915: Combine write_domain flushes to a single function
Chris Wilson
- [PATCH 12/26] drm/i915: Treat WC a separate cache domain
Chris Wilson
- [PATCH 13/26] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/26] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 15/26] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 16/26] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 17/26] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 18/26] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 19/26] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 20/26] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 21/26] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 22/26] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 23/26] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 24/26] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 25/26] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 26/26] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 1/9] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 2/9] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 3/9] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 4/9] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 5/9] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 6/9] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 7/9] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 8/9] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 9/9] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 1/6] drm/i915: Stop second guessing the caller for intel_uncore_wait_for_register()
Chris Wilson
- [PATCH 2/6] drm/i915: Stop sleeping from inside gen6_bsd_submit_request()
Chris Wilson
- [PATCH 3/6] drm/i915: Acquire uncore.lock over intel_uncore_wait_for_register()
Chris Wilson
- [PATCH 4/6] drm/i915: Use __intel_uncore_wait_for_register_fw for sandybride_pcode_read
Chris Wilson
- [PATCH 5/6] drm/i915: Use safer intel_uncore_wait_for_register in ring-init
Chris Wilson
- [PATCH 6/6] drm/i915: Add might_sleep() check to wait_for()
Chris Wilson
- [PATCH 01/12] drm/i915: Rename intel_engine_cs.exec_id to uabi_id
Chris Wilson
- [PATCH 02/12] drm/i915/execlists: Document runtime pm for intel_lrc_irq_handler()
Chris Wilson
- [PATCH 03/12] drm/i915: Lie and treat all engines as idle if wedged
Chris Wilson
- [PATCH 04/12] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 05/12] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 06/12] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 01/33] drm/i915: Rename intel_engine_cs.exec_id to uabi_id
Chris Wilson
- [PATCH 07/12] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 02/33] drm/i915/execlists: Document runtime pm for intel_lrc_irq_handler()
Chris Wilson
- [PATCH 08/12] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 03/33] drm/i915: Lie and treat all engines as idle if wedged
Chris Wilson
- [PATCH 09/12] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 04/33] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 10/12] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 05/33] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 11/12] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 06/33] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 12/12] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 07/33] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 08/33] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 09/33] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 10/33] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 11/33] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/33] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/33] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 14/33] drm/i915: Combine write_domain flushes to a single function
Chris Wilson
- [PATCH 15/33] drm/i915: Treat WC a separate cache domain
Chris Wilson
- [PATCH 16/33] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 17/33] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 18/33] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 19/33] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 20/33] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/33] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/33] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/33] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/33] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/33] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/33] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/33] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/33] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/33] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/33] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 31/33] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 32/33] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 33/33] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 01/28] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/28] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/28] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/28] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/28] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/28] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/28] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/28] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/28] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/28] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/28] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/28] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/28] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/28] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/28] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/28] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/28] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/28] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/28] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/28] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/28] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/28] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/28] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/28] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/28] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/28] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/28] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/28] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 02/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 03/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 04/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 05/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 06/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 07/29] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 08/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 09/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 10/29] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 11/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 12/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 13/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 14/29] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 15/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/29] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/29] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/29] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 29/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/26] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 03/26] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 04/26] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 05/26] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 06/26] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 07/26] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 08/26] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 09/26] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 10/26] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 11/26] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 12/26] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 13/26] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 14/26] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 15/26] drm/i915: Copy user requested buffers into the error state
Chris Wilson
- [PATCH 16/26] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 17/26] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 18/26] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 19/26] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 20/26] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 21/26] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 22/26] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 23/26] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 24/26] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 25/26] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 26/26] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/28] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/28] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 03/28] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 04/28] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 05/28] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 06/28] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 07/28] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 08/28] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 09/28] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 10/28] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 11/28] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 12/28] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 13/28] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 14/28] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 15/28] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/28] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/28] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/28] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/28] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/28] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/28] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/28] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/28] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/28] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/28] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 26/28] drm/i915: Group all the global context information together
Chris Wilson
- [PATCH 27/28] drm/i915: Allow contexts to be unreferenced locklessly
Chris Wilson
- [PATCH 28/28] drm/i915: Enable rcu-only context lookups
Chris Wilson
- [PATCH 01/25] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/25] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 03/25] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 04/25] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 05/25] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 06/25] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 07/25] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 08/25] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 09/25] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 10/25] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 11/25] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 12/25] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 13/25] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 14/25] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 15/25] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 16/25] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 17/25] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 18/25] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 19/25] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 20/25] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 21/25] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 22/25] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 23/25] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 24/25] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 25/25] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] drm/i915/selftests: Allocate inode/file dynamically
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 05/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 06/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 07/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 08/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 09/27] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 10/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 11/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 12/27] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] drm/i915/selftests: Allocate inode/file dynamically
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 05/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 06/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 07/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 08/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 09/27] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 10/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 11/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 12/27] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 1/3] drm/i915/selftests: Allocate inode/file dynamically
Chris Wilson
- [PATCH 2/3] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 3/3] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 01/28] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/28] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/28] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/28] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 05/28] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 06/28] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 07/28] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 08/28] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 09/28] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 10/28] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 11/28] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/28] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/28] drm/i915: Only report a wakeup if the waiter was truly asleep
Chris Wilson
- [PATCH 14/28] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 15/28] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 16/28] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 17/28] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 18/28] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 19/28] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 20/28] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 21/28] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 22/28] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 23/28] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 24/28] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 25/28] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 26/28] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 27/28] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 28/28] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 05/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 06/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 07/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 08/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 09/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 10/27] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 01/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 05/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 06/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 07/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 08/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 09/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 10/27] drm/i915: Confirm the request is still active before adding it to the await
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/26] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/26] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 04/26] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 05/26] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 06/26] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 07/26] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 08/26] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 09/26] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 10/26] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 11/26] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 12/26] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 13/26] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 14/26] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 15/26] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 16/26] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 17/26] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 18/26] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 19/26] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 20/26] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 21/26] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 22/26] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 23/26] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 24/26] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 25/26] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 26/26] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] ilk-ci-hax
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 04/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 05/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 06/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 07/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 08/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 09/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] ilk-ci-hax
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 04/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 05/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 06/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 07/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 08/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 09/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] ilk-ci-hax
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 04/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 05/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 06/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 07/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 08/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 09/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/27] ilk-ci-hax
Chris Wilson
- [PATCH 02/27] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 03/27] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 04/27] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 05/27] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 06/27] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 07/27] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 08/27] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 09/27] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/27] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/27] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/27] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/27] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 14/27] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 15/27] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 16/27] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 17/27] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 18/27] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 19/27] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 20/27] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 21/27] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 22/27] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 23/27] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 24/27] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 25/27] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 26/27] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 27/27] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/28] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/28] ilk-ci-hax
Chris Wilson
- [PATCH 03/28] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 04/28] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 05/28] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 06/28] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 07/28] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 08/28] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 09/28] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 10/28] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 11/28] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 12/28] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 13/28] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 14/28] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 15/28] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 16/28] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 17/28] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 18/28] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 19/28] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 20/28] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 21/28] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 22/28] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 23/28] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 24/28] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 25/28] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 26/28] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 27/28] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 28/28] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 2/3] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 3/3] ilk-ci-hax
Chris Wilson
- [PATCH 01/29] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/29] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 03/29] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 04/29] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/29] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/29] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/29] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/29] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/29] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/29] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/29] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/29] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/29] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/29] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 15/29] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/29] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/29] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 18/29] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 19/29] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 20/29] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 21/29] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 22/29] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 23/29] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 24/29] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 25/29] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 26/29] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 27/29] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 28/29] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 29/29] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/30] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 03/30] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 04/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 15/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 18/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 19/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 20/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 21/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 22/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 23/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 24/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 25/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 26/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 27/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 28/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 29/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 30/30] nobug
Chris Wilson
- [PATCH 01/31] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/31] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 03/31] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 04/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/31] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/31] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/31] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/31] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/31] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/31] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/31] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/31] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/31] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/31] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 15/31] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/31] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/31] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 18/31] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 19/31] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 20/31] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 21/31] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 22/31] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 23/31] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 24/31] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 25/31] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 26/31] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 27/31] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 28/31] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 29/31] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 30/31] nobug
Chris Wilson
- [PATCH 31/31] ringbuffer
Chris Wilson
- [PATCH 1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 2/3] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 3/3] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 1/3] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 2/3] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 3/3] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/30] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 03/30] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 04/30] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 05/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 06/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 07/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 08/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 09/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 10/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 11/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 12/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 13/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 14/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 15/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 16/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 17/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 18/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 1/2] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 2/2] dbg
Chris Wilson
- [PATCH 1/5] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 2/5] dbg
Chris Wilson
- [PATCH 3/5] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 4/5] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 5/5] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 01/31] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/31] dbg
Chris Wilson
- [PATCH 03/31] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 04/31] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 05/31] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 06/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 07/31] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 08/31] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 09/31] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 10/31] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 11/31] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 12/31] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 13/31] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 14/31] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 15/31] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 16/31] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 17/31] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 18/31] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 19/31] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 20/31] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 21/31] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 22/31] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 23/31] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 24/31] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 25/31] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 26/31] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 27/31] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 28/31] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 29/31] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 30/31] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 31/31] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/30] drm/i915: Differentiate between sw write location into ring and last hw read
Chris Wilson
- [PATCH 02/30] drm/i915: Poison the request before emitting commands
Chris Wilson
- [PATCH 03/30] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 04/30] drm/i915: Include interesting seqno in the missed breadcrumb debug
Chris Wilson
- [PATCH 05/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 06/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 07/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 08/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 09/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 10/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 11/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 12/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 13/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 14/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 15/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 16/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 17/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 18/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH] drm/i915: Compute the ring->space slightly less pessimistically
Chris Wilson
- [PATCH 01/30] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 02/30] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 03/30] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 04/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 15/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/30] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 1/3] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 2/3] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 3/3] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 01/14] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 02/14] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 03/14] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 04/14] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/14] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/14] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/14] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/14] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/14] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/14] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/14] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/14] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/14] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/14] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 01/30] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 02/30] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 03/30] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 04/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 05/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 06/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 13/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 14/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 15/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/30] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/12] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/12] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/12] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/12] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/12] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/12] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/12] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/12] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/12] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 10/12] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 11/12] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 12/12] tl
Chris Wilson
- [PATCH 01/30] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/30] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/30] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/30] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/30] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/30] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/30] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/30] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/30] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/30] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/30] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/30] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/30] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/30] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/30] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/30] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/30] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/30] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/30] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/30] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/30] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/30] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/30] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/30] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/30] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/30] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 27/30] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 28/30] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 29/30] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 30/30] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/31] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/31] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/31] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/31] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/31] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/31] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/31] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/31] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/31] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/31] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/31] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/31] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/31] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/31] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/31] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/31] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/31] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/31] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/31] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/31] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/31] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/31] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/31] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/31] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/31] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 27/31] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 28/31] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 29/31] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 30/31] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 31/31] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/36] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/36] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/36] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/36] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/36] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/36] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/36] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/36] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/36] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/36] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/36] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/36] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/36] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/36] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/36] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/36] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/36] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/36] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/36] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/36] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/36] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/36] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/36] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/36] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/36] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/36] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 27/36] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 28/36] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 29/36] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 30/36] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 31/36] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 32/36] bdw-irq
Chris Wilson
- [PATCH 33/36] drm/i915: Remove the extraneous irqreturn_t from gen8 sub handlers
Chris Wilson
- [PATCH 34/36] drm/i915: Stop inlining the execlists IRQ handler
Chris Wilson
- [PATCH 35/36] drm/i915: Don't force serialisation on marking up execlists irq posted
Chris Wilson
- [PATCH 36/36] drm/i915/execlists: Reduce lock context between schedule/submit_request
Chris Wilson
- [PATCH 01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/31] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/31] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/31] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/31] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/31] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/31] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/31] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/31] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/31] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/31] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/31] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/31] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/31] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/31] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/31] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/31] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/31] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/31] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/31] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/31] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/31] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/31] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/31] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/31] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/31] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 27/31] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 28/31] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 29/31] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 30/31] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 31/31] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/26] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/26] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/26] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/26] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/26] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/26] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/26] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/26] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/26] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/26] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/26] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/26] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/26] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/26] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/26] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/26] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/26] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/26] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/26] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/26] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/26] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/26] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/26] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/26] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/26] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/26] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 01/31] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/31] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/31] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/31] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/31] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/31] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/31] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/31] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/31] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/31] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/31] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/31] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/31] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/31] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/31] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/31] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/31] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/31] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/31] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 20/31] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 21/31] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 22/31] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 23/31] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 24/31] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 25/31] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 26/31] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 27/31] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 28/31] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 29/31] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 30/31] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 31/31] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
- [PATCH 01/32] drm/i915: Mark CPU cache as dirty on every transition for CPU writes
Chris Wilson
- [PATCH 02/32] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
Chris Wilson
- [PATCH 03/32] drm/i915: Avoid the branch in computing intel_ring_space()
Chris Wilson
- [PATCH 04/32] drm/i915: Report the ring->space from intel_ring_update_space()
Chris Wilson
- [PATCH 05/32] drm/i915: Micro-optimise hotpath through intel_ring_begin()
Chris Wilson
- [PATCH 06/32] drm/i915: Mark up clflushes as belonging to an unordered timeline
Chris Wilson
- [PATCH 07/32] drm/i915: Unwrap top level fence-array
Chris Wilson
- [PATCH 08/32] drm/i915: Lift timeline ordering to await_dma_fence
Chris Wilson
- [PATCH 09/32] drm/i915: Squash repeated awaits on the same fence
Chris Wilson
- [PATCH 10/32] drm/i915: Rename intel_timeline.sync_seqno[] to .global_sync[]
Chris Wilson
- [PATCH 11/32] drm/i915: Do not record a successful syncpoint for a dma-await
Chris Wilson
- [PATCH 12/32] drm/i915: Switch the global i915.semaphores check to a local predicate
Chris Wilson
- [PATCH 13/32] drm/i915: Make ptr_unpack_bits() more function-like
Chris Wilson
- [PATCH 14/32] drm/i915: Redefine ptr_pack_bits() and friends
Chris Wilson
- [PATCH 15/32] drm/i915/execlists: Pack the count into the low bits of the port.request
Chris Wilson
- [PATCH 16/32] drm/i915: Don't mark an execlists context-switch when idle
Chris Wilson
- [PATCH 17/32] drm/i915: Use a define for the default priority [0]
Chris Wilson
- [PATCH 18/32] drm/i915: Split execlist priority queue into rbtree + linked list
Chris Wilson
- [PATCH 19/32] drm/i915/execlists: Reduce lock context between schedule/submit_request
Chris Wilson
- [PATCH 20/32] drm/i915: Reinstate reservation_object zapping for batch_pool objects
Chris Wilson
- [PATCH 21/32] drm/i915: Amalgamate execbuffer parameter structures
Chris Wilson
- [PATCH 22/32] drm/i915: Use vma->exec_entry as our double-entry placeholder
Chris Wilson
- [PATCH 23/32] drm/i915: Split vma exec_link/evict_link
Chris Wilson
- [PATCH 24/32] drm/i915: Store a direct lookup from object handle to vma
Chris Wilson
- [PATCH 25/32] drm/i915: Pass vma to relocate entry
Chris Wilson
- [PATCH 26/32] drm/i915: Eliminate lots of iterations over the execobjects array
Chris Wilson
- [PATCH 27/32] drm/i915: Store a persistent reference for an object in the execbuffer cache
Chris Wilson
- [PATCH 28/32] drm/i915: First try the previous execbuffer location
Chris Wilson
- [PATCH 29/32] drm/i915: Wait upon userptr get-user-pages within execbuffer
Chris Wilson
- [PATCH 30/32] drm/i915: Allow execbuffer to use the first object as the batch
Chris Wilson
- [PATCH 31/32] drm/i915: Async GPU relocation processing
Chris Wilson
- [PATCH 32/32] drm/i915/scheduler: Support user-defined priorities
Chris Wilson
Last message date:
Sat Apr 29 15:04:50 UTC 2017
Archived on: Sat Apr 29 15:04:52 UTC 2017
This archive was generated by
Pipermail 0.09 (Mailman edition).