intel-xe.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v5 0/3] VF double migration
@ 2025-11-20 14:51 Satyanarayana K V P
  2025-11-20 14:51 ` [PATCH v5 1/3] drm/xe/vf: Enable VF migration only on supported GuC versions Satyanarayana K V P
                   ` (5 more replies)
  0 siblings, 6 replies; 11+ messages in thread
From: Satyanarayana K V P @ 2025-11-20 14:51 UTC (permalink / raw)
  To: intel-xe; +Cc: Satyanarayana K V P

In scenarios involving double migration, the VF KMD may encounter
situations where it is instructed to re-migrate before having the
opportunity to send RESFIX_DONE for the initial migration. This can occur
when the fix-up for the prior migration is still underway, but the VF KMD
is migrated again.

Consequently, this may lead to the possibility of sending two migration
notifications (i.e., pending fix-up for the first migration and a second
notification for the new migration). Upon receiving the first RES_FIX
notification, the GuC will resume VF submission on the GPU, potentially
resulting in undefined behavior, such as system hangs or crashes.

To avoid these hangs, a new VF2GUC action `VF2GUC_RESFIX_START` is
sent along with marker and when GUC receives the same marker with
`VF2GUC_RESFIX_DONE`action, it starts scheduling work loads from VF.

---
V4 -> V5:
- Fixed review comments (Michal W).
- Created new function vf_migration_init_late().
- Fixed minor debug log levels and documentation part.
- Moved complete marker logic to vf_post_migration_resfix_start_marker()
- Updated debugfs entries.

V3 -> V4:
- Gated Save/restore on Guc version 70.54.0
- Enabled RESFIX_START by default.
- Updated RESFIX_DONE documention.

V2 -> V3:
- Fixed review comments (Michal W).
- Updated commit message.
- Fixed CI.BAT issues.
- Added helper function to assert on unsupported GUC versions.
- Added debugfs entries to test VF double migration.

V1 -> V2:
- Squashed "Enable RESFIX start marker only on supported GUC
versions" commit into a single commit. (Matt B)
- Use fault injection for testing VF double  migration feature (Matt B).

Satyanarayana K V P (3):
  drm/xe/vf: Enable VF migration only on supported GuC versions
  drm/xe/vf: Introduce RESFIX start marker support
  drm/xe/vf: Add debugfs entries to test VF double migration

 .../gpu/drm/xe/abi/guc_actions_sriov_abi.h    |  67 ++++++++--
 drivers/gpu/drm/xe/xe_gt_sriov_vf.c           | 122 +++++++++++++++---
 drivers/gpu/drm/xe/xe_gt_sriov_vf_debugfs.c   |   5 +
 drivers/gpu/drm/xe/xe_gt_sriov_vf_types.h     |  13 ++
 drivers/gpu/drm/xe/xe_sriov_vf.c              |  36 +++++-
 5 files changed, 209 insertions(+), 34 deletions(-)

-- 
2.51.0


^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2025-11-21  7:21 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-11-20 14:51 [PATCH v5 0/3] VF double migration Satyanarayana K V P
2025-11-20 14:51 ` [PATCH v5 1/3] drm/xe/vf: Enable VF migration only on supported GuC versions Satyanarayana K V P
2025-11-20 21:18   ` Michal Wajdeczko
2025-11-20 14:52 ` [PATCH v5 2/3] drm/xe/vf: Introduce RESFIX start marker support Satyanarayana K V P
2025-11-20 21:37   ` Michal Wajdeczko
2025-11-20 14:52 ` [PATCH v5 3/3] drm/xe/vf: Add debugfs entries to test VF double migration Satyanarayana K V P
2025-11-20 21:53   ` Michal Wajdeczko
2025-11-21  7:21     ` Matthew Brost
2025-11-20 16:01 ` ✓ CI.KUnit: success for VF double migration (rev5) Patchwork
2025-11-20 16:40 ` ✓ Xe.CI.BAT: " Patchwork
2025-11-20 21:10 ` ✗ Xe.CI.Full: failure " Patchwork

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).