All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
To: intel-xe@lists.freedesktop.org
Cc: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	himal.prasad.ghimiray@intel.com,
	"Matthew Brost" <matthew.brost@intel.com>,
	"Matthew Auld" <matthew.auld@intel.com>
Subject: [PATCH v4 0/5] drm/xe: xe-only patches from the multi-device GPUSVM series
Date: Wed, 26 Mar 2025 09:05:46 +0100	[thread overview]
Message-ID: <20250326080551.40201-1-thomas.hellstrom@linux.intel.com> (raw)

A collection of misc patches that are prerequisites for the multi-device GPUSVM
series but that are xe-only and can be merged separately.

Patch 1
- Improved conditional compilation for the Xe GPUSVM support
Patch 2
- Fix a pontential UAF
Patch 3
- Adds a bo remove callback. In addition to GPUSVM this will improve
  on our device unbind support.
Patch 4
- A small fix for multi-gpu-svm support
Patch 5
- A change in implicit UAPI and needed for multi-gpu-svm support.

v2:
- Address review comments for Patch 3 (Matthew Auld)
- Fix system memory atomic access in Patch 5 (Intel CI)
v3:
- Fix compilation errors on i386 with Patch 1.
- Don't introduce an additional pinned.external_evicted list
  in Patch 3 (Matthew Auld)
- Avoid the UAPI change in Patch 5.
v4:
- Rework the pinned kernel bo dma unmapping in Patch 5.

Thomas Hellström (5):
  drm/xe: Introduce CONFIG_DRM_XE_GPUSVM
  drm/xe/svm: Fix a potential bo UAF
  drm/xe/bo: Add a bo remove callback
  drm/xe/migrate: Allow xe_migrate_vram() also on non-pagefault capable
    devices
  drm/xe: Make the PT code handle placement per PTE rather than per vma
    / range

Thomas Hellström (5):
  drm/xe: Introduce CONFIG_DRM_XE_GPUSVM
  drm/xe/svm: Fix a potential bo UAF
  drm/xe/bo: Add a bo remove callback
  drm/xe/migrate: Allow xe_migrate_vram() also on non-pagefault capable
    devices
  drm/xe: Make the PT code handle placement per PTE rather than per vma
    / range

 drivers/gpu/drm/xe/Kconfig           |  16 +++-
 drivers/gpu/drm/xe/Makefile          |   2 +-
 drivers/gpu/drm/xe/xe_bo.c           |  66 +++++++++++---
 drivers/gpu/drm/xe/xe_bo.h           |   2 +
 drivers/gpu/drm/xe/xe_bo_evict.c     |  84 +++++++++++++++++-
 drivers/gpu/drm/xe/xe_bo_evict.h     |   3 +
 drivers/gpu/drm/xe/xe_device.c       |  10 ++-
 drivers/gpu/drm/xe/xe_device_types.h |  10 ++-
 drivers/gpu/drm/xe/xe_migrate.c      |   8 +-
 drivers/gpu/drm/xe/xe_pt.c           | 126 ++++++++++++++-------------
 drivers/gpu/drm/xe/xe_query.c        |   2 +-
 drivers/gpu/drm/xe/xe_svm.c          |  22 ++++-
 drivers/gpu/drm/xe/xe_svm.h          |  73 +++++++++++-----
 drivers/gpu/drm/xe/xe_vm.c           |   2 +-
 14 files changed, 316 insertions(+), 110 deletions(-)

-- 
2.48.1


             reply	other threads:[~2025-03-26  8:06 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-03-26  8:05 Thomas Hellström [this message]
2025-03-26  8:05 ` [PATCH v4 1/5] drm/xe: Introduce CONFIG_DRM_XE_GPUSVM Thomas Hellström
2025-03-26  8:05 ` [PATCH v4 2/5] drm/xe/svm: Fix a potential bo UAF Thomas Hellström
2025-03-26  9:31   ` Ghimiray, Himal Prasad
2025-03-26  9:42     ` Thomas Hellström
2025-03-26  8:05 ` [PATCH v4 3/5] drm/xe/bo: Add a bo remove callback Thomas Hellström
2025-03-26  9:07   ` Matthew Auld
2025-03-26  8:05 ` [PATCH v4 4/5] drm/xe/migrate: Allow xe_migrate_vram() also on non-pagefault capable devices Thomas Hellström
2025-03-26  8:05 ` [PATCH v4 5/5] drm/xe: Make the PT code handle placement per PTE rather than per vma / range Thomas Hellström
2025-03-26  9:48   ` Ghimiray, Himal Prasad
2025-03-31 15:55   ` Lucas De Marchi
2025-04-01 11:55     ` Thomas Hellström
2025-04-01 14:04       ` Lucas De Marchi
2025-03-26  8:12 ` ✓ CI.Patch_applied: success for drm/xe: xe-only patches from the multi-device GPUSVM series (rev6) Patchwork
2025-03-26  8:12 ` ✗ CI.checkpatch: warning " Patchwork
2025-03-26  8:14 ` ✓ CI.KUnit: success " Patchwork
2025-03-26  8:30 ` ✓ CI.Build: " Patchwork
2025-03-26  8:32 ` ✓ CI.Hooks: " Patchwork
2025-03-26  8:34 ` ✓ CI.checksparse: " Patchwork
2025-03-26  8:55 ` ✓ Xe.CI.BAT: " Patchwork
2025-03-26 22:32 ` ✗ Xe.CI.Full: failure " Patchwork
2025-03-27  9:00 ` Patchwork
2025-03-27  9:41 ` Patchwork
2025-03-27 10:38   ` Thomas Hellström

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250326080551.40201-1-thomas.hellstrom@linux.intel.com \
    --to=thomas.hellstrom@linux.intel.com \
    --cc=himal.prasad.ghimiray@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.auld@intel.com \
    --cc=matthew.brost@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.