linux/drivers/gpu/drm/lima, branch v5.7

Merge v5.6-rc2 into drm-misc-next

2020-02-17T09:34:34Z

Lyude needs some patches in 5.6-rc2 and we didn't bring drm-misc-next forward yet, so it looks like a good occasion. Signed-off-by: Maxime Ripard

drm/lima: fix recovering from PLBU out of memory

2020-02-17T02:45:57Z

It looks like on PLBU_OUT_OF_MEM interrupt we need to resume from where we stopped, i.e. new PLBU heap start is old end. Also update end address in GP frame to grow heap on 2nd and subsequent out of memory interrupts. Fixes: 2081e8dcf1ee ("drm/lima: recover task by enlarging heap buffer") Signed-off-by: Vasily Khoruzhick Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200215035026.3180698-1-anarsoul@gmail.com

drm/lima: increase driver version to 1.1

2020-01-27T14:01:20Z

Increase driver version for mesa driver to identify the support of new heap buffer interface. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-6-yuq825@gmail.com

drm/lima: recover task by enlarging heap buffer

2020-01-27T14:01:15Z

Increase heap buffer backup memory when GP receive PLBU out of memory interrupt, then resume the task. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-5-yuq825@gmail.com

drm/lima: support heap buffer creation

2020-01-27T14:01:09Z

heap buffer is used as output of GP and input of PP for Mali Utgard GPU. Size of heap buffer depends on the task so is a runtime variable. Previously we just create a large enough buffer as heap buffer. Now we add a heap buffer type to be able to increase the backup memory dynamically when GP fail due to lack of heap memory. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-4-yuq825@gmail.com

drm/lima: add lima_vm_map_bo

2020-01-27T14:00:48Z

For dynamically mapping added backup memory of lima_bo to vm. This is a preparation for adding heap buffer support. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-3-yuq825@gmail.com

drm/lima: update register info

2020-01-27T14:00:12Z

From Mali r10p0 kernel driver source code. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-2-yuq825@gmail.com

Merge tag 'amd-drm-next-5.6-2020-01-09' of git://people.freedesktop.org/~agd5f/linux into drm-next

2020-01-13T06:49:02Z

amd-drm-next-5.6-2020-01-09: amdgpu: - Enable DCN support on POWER - Enable GFXOFF for Raven1 refresh - Clean up MM engine idle handlers - HDMI 2.0 audio fixes - Fixes for some 10 bpc EDP panels - Watermark fixes for renoir - SR-IOV fixes - Runtime pm robustness fixes - Arcturus VCN fixes - RAS fixes - BACO fixes for Arcturus - Stable pstate fixes for swSMU - HDCP fixes - PSP cleanup - HDMI fixes - Misc cleanups amdkfd: - Spread interrupt work across cores to reduce latency - Topology fixes for APUs - GPU reset improvements UAPI: - Enable DRIVER_SYNCOBJ_TIMELINE for vulkan - Return better error values for kfd process ioctl Signed-off-by: Dave Airlie From: Alex Deucher Link: https://patchwork.freedesktop.org/patch/msgid/20200109230338.8022-1-alexander.deucher@amd.com

drm/lima: use drm_sched_fault for error task handling

2020-01-09T01:27:38Z

drm_sched_job_timedout works with drm_sched_stop as a pair, so we'd better use the drm_sched_fault helper to make the error and timeout handling go the same path. This also fixes application hang when task error. Reviewed-by: Vasily Khoruzhick Tested-by: Andreas Baierl Signed-off-by: Qiang Yu Link: https://patchwork.freedesktop.org/patch/msgid/20200101103831.22429-1-yuq825@gmail.com

drm/scheduler: rework entity creation

2019-12-18T21:09:12Z

Entity currently keeps a copy of run_queue list and modify it in drm_sched_entity_set_priority(). Entities shouldn't modify run_queue list. Use drm_gpu_scheduler list instead of drm_sched_rq list in drm_sched_entity struct. In this way we can select a runqueue based on entity/ctx's priority for a drm scheduler. Signed-off-by: Nirmoy Das Reviewed-by: Christian König Signed-off-by: Alex Deucher