Skip to content

RFC Parallel submission aka multi-bb execbuf

Matthew Brost dri-devel intel-gfx oe-kbuild-all 1 version New

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 RFC Latest42 patches 0 repliesNew

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/42drm/i915/guc: GuC submission squashed into single patchdrivers/gpu/drm/i915/Makefile drivers/gpu/drm/i915/gem/i915_gem_context.c drivers/gpu/drm/i915/gem/i915_gem_mman.c +64 more+5122 -8980 replies
2/42drm/i915/guc: Allow flexible number of context idsdrivers/gpu/drm/i915/gt/uc/intel_guc.h drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+10 -70 replies
3/42drm/i915/guc: Connect the number of guc_ids to debugfsdrivers/gpu/drm/i915/gt/uc/intel_guc_debugfs.c drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+33 -10 replies
4/42drm/i915/guc: Don't return -EAGAIN to user when guc_ids exhausteddrivers/gpu/drm/i915/gt/intel_context_types.h drivers/gpu/drm/i915/gt/uc/intel_guc.h drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +1 more+755 -860 replies
5/42drm/i915/guc: Don't allow requests not ready to consume all guc_idsdrivers/gpu/drm/i915/gt/intel_context_types.h drivers/gpu/drm/i915/gt/intel_reset.c drivers/gpu/drm/i915/gt/uc/intel_guc.h +2 more+65 -30 replies
6/42drm/i915/guc: Introduce guc_submit_engine objectdrivers/gpu/drm/i915/gt/uc/intel_guc.h drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c drivers/gpu/drm/i915/gt/uc/intel_guc_submission_types.h +2 more+409 -2560 replies
7/42drm/i915/guc: Check return of __xa_store when registering a contextdrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+11 -30 replies
8/42drm/i915/guc: Non-static lrc descriptor registration bufferdrivers/gpu/drm/i915/gt/intel_context_types.h drivers/gpu/drm/i915/gt/uc/intel_guc.h drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+152 -640 replies
9/42drm/i915/guc: Take GT PM ref when deregistering contextdrivers/gpu/drm/i915/gt/intel_engine_pm.h drivers/gpu/drm/i915/gt/intel_gt_pm.h drivers/gpu/drm/i915/gt/uc/intel_guc.h +1 more+101 -190 replies
10/42drm/i915: Add GT PM unpark workerdrivers/gpu/drm/i915/gt/intel_gt.c drivers/gpu/drm/i915/gt/intel_gt_pm.c drivers/gpu/drm/i915/gt/intel_gt_pm_unpark_work.c +6 more+99 -50 replies
11/42drm/i915/guc: Take engine PM when a context is pinned with GuC submissiondrivers/gpu/drm/i915/Makefile drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+34 -30 replies
12/42drm/i915/guc: Don't call switch_to_kernel_context with GuC submissiondrivers/gpu/drm/i915/gt/intel_engine_pm.c+4 -00 replies
13/42drm/i915/guc: Selftest for GuC flow controldrivers/gpu/drm/i915/gt/uc/intel_guc.h drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h +6 more+661 -100 replies
14/42drm/i915: Add logical engine mappingdrivers/gpu/drm/i915/gt/intel_engine_cs.c drivers/gpu/drm/i915/gt/intel_engine_types.h drivers/gpu/drm/i915/gt/intel_execlists_submission.c +2 more+56 -290 replies
15/42drm/i915: Expose logical engine instance to userdrivers/gpu/drm/i915/i915_query.c include/uapi/drm/i915_drm.h+9 -10 replies
16/42drm/i915/guc: Introduce context parent-child relationshipdrivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context.h drivers/gpu/drm/i915/gt/intel_context_types.h+59 -00 replies
17/42drm/i915/guc: Implement GuC parent-child context pin / unpin functionsdrivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context.h drivers/gpu/drm/i915/gt/intel_context_types.h +6 more+368 -1080 replies
18/42drm/i915/guc: Add multi-lrc context registrationdrivers/gpu/drm/i915/gt/intel_context_types.h drivers/gpu/drm/i915/gt/intel_lrc.c drivers/gpu/drm/i915/gt/uc/abi/guc_actions_abi.h +2 more+130 -20 replies
19/42drm/i915/guc: Ensure GuC schedule operations do not operate on child contextsdrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+38 -100 replies
20/42drm/i915/guc: Assign contexts in parent-child relationship consecutive guc_idsdrivers/gpu/drm/i915/gt/intel_context.h drivers/gpu/drm/i915/gt/intel_reset.c drivers/gpu/drm/i915/gt/uc/intel_guc.h +2 more+178 -700 replies
21/42drm/i915/guc: Add hang check to GuC submit enginedrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c drivers/gpu/drm/i915/gt/uc/intel_guc_submission_types.h+39 -00 replies
22/42drm/i915/guc: Add guc_child_context_destroydrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+7 -00 replies
23/42drm/i915/guc: Implement multi-lrc submissiondrivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c drivers/gpu/drm/i915/i915_request.h+222 -160 replies
24/42drm/i915/guc: Insert submit fences between requests in parent-child relationshipdrivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context.h drivers/gpu/drm/i915/gt/intel_context_types.h +2 more+105 -270 replies
25/42drm/i915/guc: Implement multi-lrc resetdrivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+47 -190 replies
26/42drm/i915/guc: Update debugfs for GuC multi-lrcdrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c+39 -170 replies
27/42drm/i915: Connect UAPI to GuC multi-lrc interfacedrivers/gpu/drm/i915/gem/i915_gem_context.c drivers/gpu/drm/i915/gem/i915_gem_context_types.h drivers/gpu/drm/i915/gt/intel_context_types.h +6 more+407 -290 replies
28/42drm/i915/guc: Add basic GuC multi-lrc selftestdrivers/gpu/drm/i915/gt/uc/intel_guc_submission.c drivers/gpu/drm/i915/gt/uc/selftest_guc_multi_lrc.c drivers/gpu/drm/i915/selftests/i915_live_selftests.h+170 -00 replies
29/42drm/i915/guc: Implement BB boundary preemption for multi-lrcdrivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context_types.h drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h +1 more+289 -30 replies
30/42i915/drm: Move secure execbuf check to execbuf2drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+21 -140 replies
31/42drm/i915: Move input/exec fence handling to i915_gem_execbuffer2drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+43 -320 replies
32/42drm/i915: Move output fence handling to i915_gem_execbuffer2drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+25 -200 replies
33/42drm/i915: Return output fence from i915_gem_do_execbufferdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+19 -200 replies
34/42drm/i915: Store batch index in struct i915_execbufferdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+9 -100 replies
35/42drm/i915: Allow callers of i915_gem_do_execbuffer to override the batch indexdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+5 -10 replies
36/42drm/i915: Teach execbuf there can be more than one batch in the objects listdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+22 -90 replies
37/42drm/i915: Only track object dependencies on first requestdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+6 -50 replies
38/42drm/i915: Force parallel contexts to use copy engine for relocdrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c+4 -20 replies
39/42drm/i915: Multi-batch execbuffer2drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c drivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context_types.h +2 more+267 -390 replies
40/42drm/i915: Eliminate unnecessary VMA calls for multi-BB submissiondrivers/gpu/drm/i915/gem/i915_gem_execbuffer.c drivers/gpu/drm/i915/gem/selftests/i915_gem_execbuffer.c+81 -560 replies
41/42drm/i915: Enable multi-bb execbufdrivers/gpu/drm/i915/gem/i915_gem_context.c+0 -30 replies
42/42drm/i915/execlists: Parallel submission support for execlistsdrivers/gpu/drm/i915/gem/i915_gem_context.c drivers/gpu/drm/i915/gt/intel_context.c drivers/gpu/drm/i915/gt/intel_context_types.h +1 more+204 -40 replies

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.

Read the v1 thread

Cover letter [RFC PATCH 00/42] Parallel submission aka multi-bb execbuf

As discussed in [1] we are introducing a new parallel submission uAPI for the i915 which allows more than 1 BB to be submitted in an execbuf IOCTL. This is the implemenation for both GuC and execlists.

This series is broken into 5 parts.

1. Basic GuC submission. Patch 1 which the squashed version of series [2].

2. A series of GuC patches which introduces a state machine to deal with flow control conditions gracefully (e.g. don't punt them to the user). These are patches 2-13.

3. Update the GuC backend / connections to uAPI to configure it for parallel submission. These are patches 14-29.

4. Update execbuf IOCTL to accept more than 1 BB in a single IOCTL. These are patches 30-41.

5. A weak execlists implemenation for parallel submission. Patch 42.

Looking for initial feedback all parts except #1.

[1] https://patchwork.freedesktop.org/series/92028/
[2] https://patchwork.freedesktop.org/series/91840/

Matthew Brost (42):
drm/i915/guc: GuC submission squashed into single patch drm/i915/guc: Allow flexible number of context ids drm/i915/guc: Connect the number of guc_ids to debugfs drm/i915/guc: Don't return -EAGAIN to user when guc_ids exhausted drm/i915/guc: Don't allow requests not ready to consume all guc_ids drm/i915/guc: Introduce guc_submit_engine object drm/i915/guc: Check return of __xa_store when registering a context drm/i915/guc: Non-static lrc descriptor registration buffer drm/i915/guc: Take GT PM ref when deregistering context drm/i915: Add GT PM unpark worker
drm/i915/guc: Take engine PM when a context is pinned with GuC
submission
drm/i915/guc: Don't call switch_to_kernel_context with GuC submission drm/i915/guc: Selftest for GuC flow control
drm/i915: Add logical engine mapping
drm/i915: Expose logical engine instance to user drm/i915/guc: Introduce context parent-child relationship drm/i915/guc: Implement GuC parent-child context pin / unpin functions drm/i915/guc: Add multi-lrc context registration drm/i915/guc: Ensure GuC schedule operations do not operate on child
contexts
drm/i915/guc: Assign contexts in parent-child relationship consecutive
guc_ids
drm/i915/guc: Add hang check to GuC submit engine drm/i915/guc: Add guc_child_context_destroy
drm/i915/guc: Implement multi-lrc submission
drm/i915/guc: Insert submit fences between requests in parent-child
relationship
drm/i915/guc: Implement multi-lrc reset
drm/i915/guc: Update debugfs for GuC multi-lrc
drm/i915: Connect UAPI to GuC multi-lrc interface drm/i915/guc: Add basic GuC multi-lrc selftest drm/i915/guc: Implement BB boundary preemption for multi-lrc i915/drm: Move secure execbuf check to execbuf2
drm/i915: Move input/exec fence handling to i915_gem_execbuffer2 drm/i915: Move output fence handling to i915_gem_execbuffer2 drm/i915: Return output fence from i915_gem_do_execbuffer drm/i915: Store batch index in struct i915_execbuffer drm/i915: Allow callers of i915_gem_do_execbuffer to override the
batch index
drm/i915: Teach execbuf there can be more than one batch in the
objects list
drm/i915: Only track object dependencies on first request drm/i915: Force parallel contexts to use copy engine for reloc drm/i915: Multi-batch execbuffer2
drm/i915: Eliminate unnecessary VMA calls for multi-BB submission drm/i915: Enable multi-bb execbuf
drm/i915/execlists: Parallel submission support for execlists

 drivers/gpu/drm/i915/Makefile                 |    2 +
 drivers/gpu/drm/i915/gem/i915_gem_context.c   |  176 +-
 .../gpu/drm/i915/gem/i915_gem_context_types.h |    6 +
 .../gpu/drm/i915/gem/i915_gem_execbuffer.c    |  541 +-
 drivers/gpu/drm/i915/gem/i915_gem_mman.c      |    3 +-
 .../i915/gem/selftests/i915_gem_execbuffer.c  |   14 +-
 drivers/gpu/drm/i915/gt/gen8_engine_cs.c      |    6 +-
 drivers/gpu/drm/i915/gt/intel_breadcrumbs.c   |   44 +-
 drivers/gpu/drm/i915/gt/intel_breadcrumbs.h   |   16 +-
 .../gpu/drm/i915/gt/intel_breadcrumbs_types.h |    7 +
 drivers/gpu/drm/i915/gt/intel_context.c       |  275 +-
 drivers/gpu/drm/i915/gt/intel_context.h       |   82 +-
 drivers/gpu/drm/i915/gt/intel_context_types.h |  115 +-
 drivers/gpu/drm/i915/gt/intel_engine.h        |   67 +-
 drivers/gpu/drm/i915/gt/intel_engine_cs.c     |  242 +-
 .../gpu/drm/i915/gt/intel_engine_heartbeat.c  |   71 +-
 .../gpu/drm/i915/gt/intel_engine_heartbeat.h  |    4 +
 drivers/gpu/drm/i915/gt/intel_engine_pm.c     |    4 +
 drivers/gpu/drm/i915/gt/intel_engine_pm.h     |    5 +
 drivers/gpu/drm/i915/gt/intel_engine_types.h  |   14 +-
 drivers/gpu/drm/i915/gt/intel_engine_user.c   |    4 +
 .../drm/i915/gt/intel_execlists_submission.c  |  313 +-
 .../drm/i915/gt/intel_execlists_submission.h  |    4 -
 drivers/gpu/drm/i915/gt/intel_gt.c            |   24 +
 drivers/gpu/drm/i915/gt/intel_gt.h            |    2 +
 drivers/gpu/drm/i915/gt/intel_gt_pm.c         |   14 +-
 drivers/gpu/drm/i915/gt/intel_gt_pm.h         |   13 +
 .../gpu/drm/i915/gt/intel_gt_pm_unpark_work.c |   35 +
 .../gpu/drm/i915/gt/intel_gt_pm_unpark_work.h |   32 +
 drivers/gpu/drm/i915/gt/intel_gt_requests.c   |   21 +-
 drivers/gpu/drm/i915/gt/intel_gt_requests.h   |    9 +-
 drivers/gpu/drm/i915/gt/intel_gt_types.h      |    3 +
 drivers/gpu/drm/i915/gt/intel_lrc.c           |   23 +-
 drivers/gpu/drm/i915/gt/intel_lrc.h           |    6 +-
 drivers/gpu/drm/i915/gt/intel_lrc_reg.h       |    1 -
 drivers/gpu/drm/i915/gt/intel_reset.c         |   58 +-
 .../gpu/drm/i915/gt/intel_ring_submission.c   |   47 +-
 drivers/gpu/drm/i915/gt/intel_rps.c           |    4 +
 drivers/gpu/drm/i915/gt/intel_workarounds.c   |   46 +-
 .../gpu/drm/i915/gt/intel_workarounds_types.h |    1 +
 drivers/gpu/drm/i915/gt/mock_engine.c         |   38 +-
 drivers/gpu/drm/i915/gt/selftest_context.c    |   10 +
 .../drm/i915/gt/selftest_engine_heartbeat.c   |   22 +
 .../drm/i915/gt/selftest_engine_heartbeat.h   |    2 +
 drivers/gpu/drm/i915/gt/selftest_execlists.c  |   12 +-
 drivers/gpu/drm/i915/gt/selftest_hangcheck.c  |  314 +-
 drivers/gpu/drm/i915/gt/selftest_mocs.c       |   50 +-
 .../gpu/drm/i915/gt/selftest_workarounds.c    |  132 +-
 .../gpu/drm/i915/gt/uc/abi/guc_actions_abi.h  |   16 +
 drivers/gpu/drm/i915/gt/uc/intel_guc.c        |   82 +-
 drivers/gpu/drm/i915/gt/uc/intel_guc.h        |  126 +-
 drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c    |  465 +-
 drivers/gpu/drm/i915/gt/uc/intel_guc_ads.h    |    4 +
 drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c     |  175 +-
 drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h     |   25 +-
 .../gpu/drm/i915/gt/uc/intel_guc_debugfs.c    |   56 +-
 drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h   |   98 +-
 .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 4961 +++++++++++++++--
 .../gpu/drm/i915/gt/uc/intel_guc_submission.h |   20 +-
 .../i915/gt/uc/intel_guc_submission_types.h   |   67 +
 drivers/gpu/drm/i915/gt/uc/intel_uc.c         |  101 +-
 drivers/gpu/drm/i915/gt/uc/intel_uc.h         |   11 +
 .../i915/gt/uc/selftest_guc_flow_control.c    |  581 ++
 .../drm/i915/gt/uc/selftest_guc_multi_lrc.c   |  168 +
 drivers/gpu/drm/i915/i915_debugfs_params.c    |   31 +
 drivers/gpu/drm/i915/i915_gem_evict.c         |    1 +
 drivers/gpu/drm/i915/i915_gpu_error.c         |   25 +-
 drivers/gpu/drm/i915/i915_query.c             |    2 +
 drivers/gpu/drm/i915/i915_reg.h               |    2 +
 drivers/gpu/drm/i915/i915_request.c           |  288 +-
 drivers/gpu/drm/i915/i915_request.h           |   43 +
 drivers/gpu/drm/i915/i915_scheduler.c         |   36 +-
 drivers/gpu/drm/i915/i915_scheduler.h         |   13 +-
 drivers/gpu/drm/i915/i915_scheduler_types.h   |   22 +
 drivers/gpu/drm/i915/i915_trace.h             |  199 +-
 drivers/gpu/drm/i915/i915_vma.c               |   13 +-
 drivers/gpu/drm/i915/i915_vma.h               |   16 +-
 drivers/gpu/drm/i915/intel_wakeref.c          |    5 +
 drivers/gpu/drm/i915/intel_wakeref.h          |    1 +
 .../drm/i915/selftests/i915_live_selftests.h  |    2 +
 drivers/gpu/drm/i915/selftests/i915_request.c |    4 +-
 .../gpu/drm/i915/selftests/igt_flush_test.c   |    2 +-
 .../gpu/drm/i915/selftests/igt_live_test.c    |    2 +-
 .../i915/selftests/intel_scheduler_helpers.c  |  101 +
 .../i915/selftests/intel_scheduler_helpers.h  |   36 +
 .../gpu/drm/i915/selftests/mock_gem_device.c  |    3 +-
 include/uapi/drm/i915_drm.h                   |  135 +-
 87 files changed, 9648 insertions(+), 1224 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/gt/intel_gt_pm_unpark_work.c
 create mode 100644 drivers/gpu/drm/i915/gt/intel_gt_pm_unpark_work.h
 create mode 100644 drivers/gpu/drm/i915/gt/uc/intel_guc_submission_types.h
 create mode 100644 drivers/gpu/drm/i915/gt/uc/selftest_guc_flow_control.c
 create mode 100644 drivers/gpu/drm/i915/gt/uc/selftest_guc_multi_lrc.c
 create mode 100644 drivers/gpu/drm/i915/selftests/intel_scheduler_helpers.c
 create mode 100644 drivers/gpu/drm/i915/selftests/intel_scheduler_helpers.h
Signature · 1 line
2.28.0