Re: [PULL] drm-xe-next

[Jani Nikula <jani.nikula@linux.intel.com>]

On Fri, 25 Oct 2024, Thomas Hellström <thomas.hellstrom@linux.intel.com> wrote:
> On Thu, 2024-10-24 at 19:22 +0000, Matthew Brost wrote:
>> On Thu, Oct 24, 2024 at 07:52:11PM +0200, Thomas Hellstrom wrote:
>> > Hi, Dave & Simona,
>> > 
>> > This week's drm-xe-next PR
>> > 
>> > Thanks,
>> > Thomas
>> > 
>> > 
>> > drm-xe-next-2024-10-24:
>> > UAPI Changes:
>> > - Define and parse OA sync properties (Ashutosh)
>> > 
>> > Driver Changes:
>> > - Add caller info to xe_gt_reset_async (Nirmoy)
>> > - A large forcewake rework / cleanup (Himal)
>> > - A g2h response timeout fix (Badal)
>> > - A PTL workaround (Vinay)
>> > - Handle unreliable MMIO reads during forcewake (Shuicheng)
>> > - Ufence user-space access fixes (Nirmoy)
>> > - Annotate flexible arrays (Matthew Brost)
>> > - Enable GuC lite restore (Fei)
>> > - Prevent GuC register capture on VF (Zhanjun)
>> > - Show VFs VRAM / LMEM provisioning summary over debugfs (Michal)
>> > - Parallel queues fix on GT reset (Nirmoy)
>> > - Move reference grabbing to a job's dma-fence (Matt Brost)
>> > - Mark a number of local workqueues WQ_MEM_RECLAIM (Matt Brost)
>> 
>> This breaks CI [1] - my mistake. Maybe omit these in this weeks PR.

How did this pass CI and get merged in the first place?!?

It's now botching unrelated pre-merge testing all over the place,
e.g. [3] and [4].

BR,
Jani.


[3] https://lore.kernel.org/r/172981565466.1330037.6238046952250769671@2413ebb6fbb6
[4] https://lore.kernel.org/r/172981849964.1330038.16133455483045565936@2413ebb6fbb6


>> 
>> We need [2] merged to fix this. Waiting on an RB but I'd like to get
>> all of this in 6.12.
>> 
>> Matt
>> 
>> [1]
>> https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-140135v2/bat-lnl-1/igt@xe_exec_fault_mode@twice-invalid-fault.html
>> [2] https://patchwork.freedesktop.org/series/140406/
>
> So this CI failure is a warning only and IMHO for drm-xe-next (6.13)
> it's not catastrophic. There might be a window in the bisect history
> where this warning appears. It's perhaps more important for -fixes,
> though.
>
> If we need to wait for the scheduler patch going into drm-misc-next /
> drm-next/ backmerge we'd hold off this branch for too long I fear.
>
> @Dave, @Sima 
> If you feel differently please skip this PR for this week and we'll
> work to get the scheduler patch merged asap.
>
> Thanks,
> Thomas
>
>
>> 
>> > - OA synchronization support (Ashutosh)
>> > 
>> > The following changes since commit
>> > 2eb460ab9f4bc5b575f52568d17936da0af681d8:
>> > 
>> >   drm/xe: Enlarge the invalidation timeout from 150 to 500 (2024-
>> > 10-16 16:11:10 +0100)
>> > 
>> > are available in the Git repository at:
>> > 
>> >   https://gitlab.freedesktop.org/drm/xe/kernel.git tags/drm-xe-
>> > next-2024-10-24
>> > 
>> > for you to fetch changes up to
>> > 85d3f9e84e0628c412b69aa99b63654dfa08ad68:
>> > 
>> >   drm/xe/oa: Allow only certain property changes from config (2024-
>> > 10-23 12:42:20 -0700)
>> > 
>> > ----------------------------------------------------------------
>> > UAPI Changes:
>> > - Define and parse OA sync properties (Ashutosh)
>> > 
>> > Driver Changes:
>> > - Add caller info to xe_gt_reset_async (Nirmoy)
>> > - A large forcewake rework / cleanup (Himal)
>> > - A g2h response timeout fix (Badal)
>> > - A PTL workaround (Vinay)
>> > - Handle unreliable MMIO reads during forcewake (Shuicheng)
>> > - Ufence user-space access fixes (Nirmoy)
>> > - Annotate flexible arrays (Matthew Brost)
>> > - Enable GuC lite restore (Fei)
>> > - Prevent GuC register capture on VF (Zhanjun)
>> > - Show VFs VRAM / LMEM provisioning summary over debugfs (Michal)
>> > - Parallel queues fix on GT reset (Nirmoy)
>> > - Move reference grabbing to a job's dma-fence (Matt Brost)
>> > - Mark a number of local workqueues WQ_MEM_RECLAIM (Matt Brost)
>> > - OA synchronization support (Ashutosh)
>> > 
>> > ----------------------------------------------------------------
>> > Ashutosh Dixit (7):
>> >       drm/xe/oa: Separate batch submission from waiting for
>> > completion
>> >       drm/xe/oa/uapi: Define and parse OA sync properties
>> >       drm/xe/oa: Add input fence dependencies
>> >       drm/xe/oa: Signal output fences
>> >       drm/xe/oa: Move functions up so they can be reused for config
>> > ioctl
>> >       drm/xe/oa: Add syncs support to OA config ioctl
>> >       drm/xe/oa: Allow only certain property changes from config
>> > 
>> > Badal Nilawar (1):
>> >       drm/xe/guc/ct: Flush g2h worker in case of g2h response
>> > timeout
>> > 
>> > Fei Yang (1):
>> >       drm/xe: enable lite restore
>> > 
>> > Himal Prasad Ghimiray (26):
>> >       drm/xe: Add member initialized_domains to xe_force_wake()
>> >       drm/xe/forcewake: Change awake_domain datatype
>> >       drm/xe/forcewake: Add a helper xe_force_wake_ref_has_domain()
>> >       drm/xe: Error handling in xe_force_wake_get()
>> >       drm/xe: Modify xe_force_wake_put to handle _get returned mask
>> >       drm/xe/device: Update handling of xe_force_wake_get return
>> >       drm/xe/hdcp: Update handling of xe_force_wake_get return
>> >       drm/xe/gsc: Update handling of xe_force_wake_get return
>> >       drm/xe/gt: Update handling of xe_force_wake_get return
>> >       drm/xe/xe_gt_idle: Update handling of xe_force_wake_get
>> > return
>> >       drm/xe/devcoredump: Update handling of xe_force_wake_get
>> > return
>> >       drm/xe/tests/mocs: Update xe_force_wake_get() return handling
>> >       drm/xe/mocs: Update handling of xe_force_wake_get return
>> >       drm/xe/xe_drm_client: Update handling of xe_force_wake_get
>> > return
>> >       drm/xe/xe_gt_debugfs: Update handling of xe_force_wake_get
>> > return
>> >       drm/xe/guc: Update handling of xe_force_wake_get return
>> >       drm/xe/huc: Update handling of xe_force_wake_get return
>> >       drm/xe/oa: Handle force_wake_get failure in
>> > xe_oa_stream_init()
>> >       drm/xe/pat: Update handling of xe_force_wake_get return
>> >       drm/xe/gt_tlb_invalidation_ggtt: Update handling of
>> > xe_force_wake_get return
>> >       drm/xe/xe_reg_sr: Update handling of xe_force_wake_get return
>> >       drm/xe/query: Update handling of xe_force_wake_get return
>> >       drm/xe/vram: Update handling of xe_force_wake_get return
>> >       drm/xe: forcewake debugfs open fails on xe_forcewake_get
>> > failure
>> >       drm/xe: Ensure __must_check for xe_force_wake_get() return
>> >       drm/xe: Change return type to void for xe_force_wake_put
>> > 
>> > Matthew Brost (5):
>> >       drm/xe: Use __counted_by for flexible arrays
>> >       drm/xe: Take ref to job's fence in arm
>> >       drm/xe: Mark GGTT work queue with WQ_MEM_RECLAIM
>> >       drm/xe: Mark G2H work queue with WQ_MEM_RECLAIM
>> >       drm/xe: Mark GT work queue with WQ_MEM_RECLAIM
>> > 
>> > Michal Wajdeczko (1):
>> >       drm/xe/pf: Show VFs LMEM provisioning summary over debugfs
>> > 
>> > Nirmoy Das (4):
>> >       drm/xe: Add caller info to xe_gt_reset_async
>> >       drm/xe/ufence: Prefetch ufence addr to catch bogus address
>> >       drm/xe/ufence: Warn if mmget_not_zero() fails
>> >       drm/xe: Don't restart parallel queues multiple times on GT
>> > reset
>> > 
>> > Shuicheng Lin (1):
>> >       drm/xe: Handle unreliable MMIO reads during forcewake
>> > 
>> > Vinay Belgaumkar (1):
>> >       drm/xe/ptl: Apply Wa_14022866841
>> > 
>> > Zhanjun Dong (1):
>> >       drm/xe/guc: Prevent GuC register capture running on VF
>> > 
>> >  drivers/gpu/drm/xe/abi/guc_klvs_abi.h       |   1 +
>> >  drivers/gpu/drm/xe/display/xe_hdcp_gsc.c    |   6 +-
>> >  drivers/gpu/drm/xe/tests/xe_mocs.c          |  18 +-
>> >  drivers/gpu/drm/xe/xe_debugfs.c             |  27 +-
>> >  drivers/gpu/drm/xe/xe_devcoredump.c         |  14 +-
>> >  drivers/gpu/drm/xe/xe_device.c              |  25 +-
>> >  drivers/gpu/drm/xe/xe_drm_client.c          |   8 +-
>> >  drivers/gpu/drm/xe/xe_exec_queue_types.h    |   2 +-
>> >  drivers/gpu/drm/xe/xe_execlist.c            |   2 +-
>> >  drivers/gpu/drm/xe/xe_force_wake.c          | 134 ++++--
>> >  drivers/gpu/drm/xe/xe_force_wake.h          |  23 +-
>> >  drivers/gpu/drm/xe/xe_force_wake_types.h    |   6 +-
>> >  drivers/gpu/drm/xe/xe_ggtt.c                |   2 +-
>> >  drivers/gpu/drm/xe/xe_gsc.c                 |  23 +-
>> >  drivers/gpu/drm/xe/xe_gsc_proxy.c           |   9 +-
>> >  drivers/gpu/drm/xe/xe_gt.c                  | 110 +++--
>> >  drivers/gpu/drm/xe/xe_gt_debugfs.c          |  13 +-
>> >  drivers/gpu/drm/xe/xe_gt_idle.c             |  26 +-
>> >  drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c  |  35 ++
>> >  drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h  |   1 +
>> >  drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c |   5 +
>> >  drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c |   5 +-
>> >  drivers/gpu/drm/xe/xe_guc.c                 |  15 +-
>> >  drivers/gpu/drm/xe/xe_guc_ads.c             |   5 +
>> >  drivers/gpu/drm/xe/xe_guc_capture.c         |   8 +-
>> >  drivers/gpu/drm/xe/xe_guc_ct.c              |  20 +-
>> >  drivers/gpu/drm/xe/xe_guc_fwif.h            |   1 +
>> >  drivers/gpu/drm/xe/xe_guc_log.c             |   9 +-
>> >  drivers/gpu/drm/xe/xe_guc_pc.c              |  50 +-
>> >  drivers/gpu/drm/xe/xe_guc_submit.c          |  29 +-
>> >  drivers/gpu/drm/xe/xe_huc.c                 |   8 +-
>> >  drivers/gpu/drm/xe/xe_mocs.c                |  14 +-
>> >  drivers/gpu/drm/xe/xe_oa.c                  | 678
>> > +++++++++++++++++++---------
>> >  drivers/gpu/drm/xe/xe_oa_types.h            |  12 +
>> >  drivers/gpu/drm/xe/xe_pat.c                 |  65 ++-
>> >  drivers/gpu/drm/xe/xe_query.c               |  10 +-
>> >  drivers/gpu/drm/xe/xe_reg_sr.c              |  24 +-
>> >  drivers/gpu/drm/xe/xe_sched_job.c           |   2 +-
>> >  drivers/gpu/drm/xe/xe_sched_job_types.h     |   3 +-
>> >  drivers/gpu/drm/xe/xe_sync.c                |   5 +-
>> >  drivers/gpu/drm/xe/xe_vram.c                |  12 +-
>> >  drivers/gpu/drm/xe/xe_wa_oob.rules          |   2 +
>> >  include/uapi/drm/xe_drm.h                   |  17 +
>> >  43 files changed, 997 insertions(+), 487 deletions(-)
>

-- 
Jani Nikula, Intel