[PATCH v15 0/9] TTM shrinker helpers and xe buffer object shrinker

Fri Jan 10 13:04:53 UTC 2025

On Fri, 2025-01-10 at 10:47 +0100, Christian König wrote:
> Any objections that I push patch #1 and #3 to drm-misc-next?
> 
> They are general cleanups anyway and this way we have them out of the
> way.
> 
> Regards,
> Christian.

Please do. I have to rewrite the cover letter but that's not a big
deal. Also perhaps consider adding a patch to remove the "allow_pools"
functionality.

Also re-raising the question if there is a way we can make TTM
development scale a bit better, with a team of trusted reviewers? 
I am currently working under the assumption that you want to at least
ack everything that goes into TTM?

Thanks,
Thomas

> 
> Am 17.12.24 um 15:58 schrieb Thomas Hellström:
> > This series implements TTM shrinker / eviction helpers and an xe bo
> > shrinker. It builds on a previous series, *and obsoletes that one*.
> > 
> > https://lore.kernel.org/linux-mm/b7491378-defd-4f1c-31e2-29e4c77e2d67@amd.com/T/
> > 
> > Where the comment about layering
> > https://lore.kernel.org/linux-mm/b7491378-defd-4f1c-31e2-29e4c77e2d67@amd.com/T/#ma918844aa8a6efe8768fdcda0c6590d5c93850c9
> > 
> > now addressed, and this version also implements shmem objects for
> > backup
> > rather than direct swap-cache insertions, which was used in the
> > previuos
> > series. It turns out that with per-page backup / shrinking, shmem
> > objects
> > appears to work just as well as direct swap-cache insertions with
> > the
> > added benefit that was introduced in the previous TTM shrinker
> > series to
> > avoid running out of swap entries isn't really needed.
> > 
> > The series earlier consisted of a LRU traversal part and the
> > current part.
> > The LRU traversal part is merged, but is still mentioned in the
> > history
> > below.
> > 
> > Patch 1 balances ttm_resource_cursor_fini() with an init function.
> > It
> > makes patch 5 more straightforward.
> > 
> > Patch 2 introduces a backup implemententaion.
> > 
> > Patch 3 reworks the structure of ttm_pool_alloc in preparation for
> > patch 4.
> > 
> > Patch 4 introduces functionality in the ttm_pool code for page-by-
> > page shrinking
> > and recovery. It avoids having to temporarily allocate a huge
> > amount of
> > memory to be able to shrink a buffer object. It also introduces the
> > possibility to immediately write-back pages if needed.
> > 
> > Patch 5 Adds a simple error injection to the above code to help
> > increase
> > test coverage.
> > 
> > Patch 6 Implements a macro for LRU iteration.
> > 
> > Patch 7 Introduces driver-facing helpers for shrinking.
> > 
> > Patch 8 Implements the xe bo shrinker.
> > 
> > Patch 9 Increases (removes) the XE_PL_TT watermark.
> > 
> > v2:
> > - Squash obsolete revision history in the patch commit messages.
> > - Fix a couple of review comments by Christian
> > - Don't store the mem_type in the TTM managers but in the
> >    resource cursor.
> > - Rename introduced TTM *back_up* function names to *backup*
> > - Add ttm pool recovery fault injection.
> > - Shrinker xe kunit test
> > - Various bugfixes
> > 
> > v3:
> > - Address some review comments from Matthew Brost and Christian
> > König.
> > - Use the restartable LRU walk for TTM swapping and eviction.
> > - Provide a POC drm_exec locking implementation for exhaustive
> >    eviction. (Christian König).
> > 
> > v4:
> > - Remove the RFC exhaustive eviction part. While the path to
> > exhaustive
> >    eviction is pretty clear and demonstrated in v3, there is still
> > some
> >    drm_exec work that needs to be agreed and implemented.
> > - Add shrinker power management. On some hw we need to wake when
> > shrinking.
> > - Fix the lru walker helper for -EALREADY errors.
> > - Add drm/xe: Increase the XE_PL_TT watermark.
> > 
> > v5:
> > - Update also TTM kunit tests
> > - Handle ghost- and zombie objects in the shrinker.
> > - A couple of compile- and UAF fixes reported by Kernel Build Robot
> > and
> >    Dan Carpenter.
> > 
> > v6:
> > - Address review comments from Matthew Brost on the
> >    restartable LRU traversal path.
> > 
> > v7:
> > - Split out TTM restartable LRU traversal path and merge that.
> > - Adapt the review comments on that series.
> > 
> > v8:
> > - Address review comments from Matthew Brost as detailed in the
> >    respective patches.
> > 
> > v9:
> > - Rebase and fix compilation errors
> > 
> > v10:
> > - Use a LRU iteration macro rather than a function with a callback.
> > - Rebasing and cleanups
> > - Address some additional review comments from Matt Brost.
> > - Drop the shrinker selftest. It was already merged as a swapout
> >    self-test.
> > 
> > v11:
> > - Move more core interaction to additional TTM helpers.
> > - Don't back up without __GFP_FS, and don't start writeback without
> > __GFP_IO.
> > - Rebase.
> > 
> > v12:
> > - Fix an indentation flaw.
> > - Rebase
> > 
> > v13:
> > - Remove the backup base-class, and use direct calls for ttm_backup
> >    (Christian König).
> > - Rebase on the ttm_backup changes.
> > - Move shrunken bos from the LRU list to the unevictable list.
> > - Provide an accessor function with sanity checks to set the
> > - ttm_tt::backup field.
> > - Update documentation.
> > 
> > v14:
> > - Update documentation of ttm_backup_bytes_avail().
> > - Work around converting between struct file * and struct ttm-
> > backup *.
> > - Don't set up backup for imported buffers.
> > 
> > v15:
> > - ttm_backup return value change. (Christian König)
> > - ttm_pool restore/alloc implementation rework (Christian König)
> > - Documentation update.
> > - Remove cond_resched (Christian König)
> > 
> > Cc: Somalapuram Amaranath <Amaranath.Somalapuram at amd.com>
> > Cc: Christian König <christian.koenig at amd.com>
> > Cc: Matthew Brost <matthew.brost at intel.com>
> > Cc: Paulo Zanoni <paulo.r.zanoni at intel.com>
> > Cc: <dri-devel at lists.freedesktop.org>
> > 
> > Thomas Hellström (9):
> >    drm/ttm: Balance ttm_resource_cursor_init() and
> >      ttm_resource_cursor_fini()
> >    drm/ttm: Provide a shmem backup implementation
> >    drm/ttm/pool: Restructure the pool allocation code
> >    drm/ttm/pool, drm/ttm/tt: Provide a helper to shrink pages
> >    drm/ttm: Use fault-injection to test error paths
> >    drm/ttm: Add a macro to perform LRU iteration
> >    drm/ttm: Add helpers for shrinking
> >    drm/xe: Add a shrinker for xe bos
> >    drm/xe: Increase the XE_PL_TT watermark
> > 
> >   drivers/gpu/drm/ttm/Makefile         |   2 +-
> >   drivers/gpu/drm/ttm/ttm_backup.c     | 207 ++++++++
> >   drivers/gpu/drm/ttm/ttm_bo.c         |   3 +-
> >   drivers/gpu/drm/ttm/ttm_bo_util.c    | 250 +++++++++-
> >   drivers/gpu/drm/ttm/ttm_pool.c       | 718
> > +++++++++++++++++++++++----
> >   drivers/gpu/drm/ttm/ttm_resource.c   |  35 +-
> >   drivers/gpu/drm/ttm/ttm_tt.c         |  83 ++++
> >   drivers/gpu/drm/xe/Makefile          |   1 +
> >   drivers/gpu/drm/xe/tests/xe_bo.c     |   6 +-
> >   drivers/gpu/drm/xe/xe_bo.c           | 202 +++++++-
> >   drivers/gpu/drm/xe/xe_bo.h           |  36 ++
> >   drivers/gpu/drm/xe/xe_device.c       |   8 +
> >   drivers/gpu/drm/xe/xe_device_types.h |   2 +
> >   drivers/gpu/drm/xe/xe_shrinker.c     | 258 ++++++++++
> >   drivers/gpu/drm/xe/xe_shrinker.h     |  18 +
> >   drivers/gpu/drm/xe/xe_ttm_sys_mgr.c  |   3 +-
> >   include/drm/ttm/ttm_backup.h         |  74 +++
> >   include/drm/ttm/ttm_bo.h             |  93 ++++
> >   include/drm/ttm/ttm_pool.h           |   8 +
> >   include/drm/ttm/ttm_resource.h       |  11 +-
> >   include/drm/ttm/ttm_tt.h             |  69 ++-
> >   21 files changed, 1935 insertions(+), 152 deletions(-)
> >   create mode 100644 drivers/gpu/drm/ttm/ttm_backup.c
> >   create mode 100644 drivers/gpu/drm/xe/xe_shrinker.c
> >   create mode 100644 drivers/gpu/drm/xe/xe_shrinker.h
> >   create mode 100644 include/drm/ttm/ttm_backup.h
> > 
> 