All of lore.kernel.org
 help / color / mirror / Atom feed
From: Boris Brezillon <boris.brezillon@collabora.com>
To: "Boris Brezillon" <boris.brezillon@collabora.com>,
	"Steven Price" <steven.price@arm.com>,
	"Liviu Dudau" <liviu.dudau@arm.com>,
	"Adrián Larumbe" <adrian.larumbe@collabora.com>
Cc: dri-devel@lists.freedesktop.org,
	"David Airlie" <airlied@gmail.com>,
	"Simona Vetter" <simona@ffwll.ch>,
	"Akash Goel" <akash.goel@arm.com>,
	"Rob Clark" <robin.clark@oss.qualcomm.com>,
	"Sean Paul" <sean@poorly.run>,
	"Konrad Dybcio" <konradybcio@kernel.org>,
	"Akhil P Oommen" <akhilpo@oss.qualcomm.com>,
	"Maarten Lankhorst" <maarten.lankhorst@linux.intel.com>,
	"Maxime Ripard" <mripard@kernel.org>,
	"Thomas Zimmermann" <tzimmermann@suse.de>,
	"Dmitry Osipenko" <dmitry.osipenko@collabora.com>,
	"Chris Diamand" <chris.diamand@arm.com>,
	"Danilo Krummrich" <dakr@kernel.org>,
	"Matthew Brost" <matthew.brost@intel.com>,
	"Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	"Alice Ryhl" <aliceryhl@google.com>,
	"Chia-I Wu" <olvaffe@gmail.com>,
	kernel@collabora.com
Subject: Re: [PATCH v7 00/10] drm/panthor: Add a GEM shrinker
Date: Fri, 3 Apr 2026 13:54:04 +0200	[thread overview]
Message-ID: <20260403135404.431fe625@fedora> (raw)
In-Reply-To: <20260401134854.2275433-1-boris.brezillon@collabora.com>

On Wed,  1 Apr 2026 15:48:44 +0200
Boris Brezillon <boris.brezillon@collabora.com> wrote:

> Hello,
> 
> This is an attempt at adding a GEM shrinker to panthor so the system
> can finally reclaim GPU memory.
> 
> This implementation is losely based on the MSM shrinker (which is why
> I added the MSM maintainers in Cc), and it's relying on the drm_gpuvm
> eviction/validation infrastructure.
> 
> I've only done very basic IGT-based [1] and chromium-based (opening
> a lot of tabs on Aquarium until the system starts reclaiming+swapping
> out GPU buffers) testing, but I'm posting this early so I can get
> preliminary feedback on the implementation. If someone knows about
> better tools/ways to test the shrinker, please let me know.
> 
> A few words about some design/implementation choices:
> - No MADVISE support because I want to see if we can live with just
>   transparent reclaim
> - We considered basing this implementation on the generic shrinker work
>   started by Dmitry [2], but
>   1. with the activeness/idleness tracking happening at the VM
>      granularity, having per-BO LRUs would caused a lot of
>      list_move()s that are not really needed (the VM as a whole
>      become active/idle, we can track individual BOs)
>   2. Thomas Zimmermann recently suggested that we should have our
>      own GEM implementation instead of trying to add this extra reclaim
>      complexity to gem-shmem. There are some plans to create a
>      gem-uma (Unified Memory Architecture) lib that would do more
>      than gem-shmem but in a way that doesn't force all its users
>      to pay the overhead (size overhead of the gem object, mostly)
>      for features they don't use. Patch "Part ways with
>      drm_gem_shmem_object" is showing what this component-based lib
>      API could look like if it were to be extracted
> - At the moment we only support swapout, but we could add an
>   extra flag to specify when buffer content doesn't need to be
>   preserved to avoid the swapout/swapin dance. First candidate for
>   this DISCARD_ON_RECLAIM flag would probably be the tiler heap chunks.
> - Reclaim uses _try_lock() all the way because of the various lock order
>   inversions between the reclaim path and submission paths. That means
>   we don't try very hard to reclaim hot GPU buffers, but the locking is
>   such a mess that I don't really see a better option to be honest.
> 
> 
> Changes in v2:
> - No fundamental changes in this v2, since the feedback I got were more
>   focused on bugs than the overall approach. Check the changelog in each
>   patch for more details.
> 
> Changes in v3:
> - Mostly fixes (see the changelog in each patch)
> 
> Changes in v4:
> - Only fixes (see the changelog in each patch)
> 
> Changes in v5:
> - Only fixes (see the changelog in each patch)
> 
> Changes in v6:
> - Fix huge fault handling and various other things reported by Adrian
>   and the review bot (see the changelog in each patch for more)
> 
> Changes in v7:
> - Fix a deadlock reported by lockdep
> - Add a commit to remove an unused field
> 
> Regards,
> 
> Boris
> 
> [1]https://gitlab.freedesktop.org/bbrezillon/igt-gpu-tools/-/commit/fc76934a5579767d2aabe787d40e38a17c3f4ea4
> [2]https://lkml.org/lkml/2024/1/5/665
> 
> Akash Goel (1):
>   drm/panthor: Add a GEM shrinker
> 
> Boris Brezillon (9):
>   drm/gem: Consider GEM object reclaimable if shrinking fails
>   drm/panthor: Remove unused panthor_vm_op_ctx::map::new_vma field
>   drm/panthor: Move panthor_gems_debugfs_init() to panthor_gem.c
>   drm/panthor: Group panthor_kernel_bo_xxx() helpers
>   drm/panthor: Don't call drm_gpuvm_bo_extobj_add() if the object is
>     private
>   drm/panthor: Part ways with drm_gem_shmem_object
>   drm/panthor: Lazily allocate pages on mmap()
>   drm/panthor: Split panthor_vm_prepare_map_op_ctx() to prepare for
>     reclaim
>   drm/panthor: Track the number of mmap on a BO

Queued to drm-misc-next.

> 
>  drivers/gpu/drm/drm_gem.c                |   10 +
>  drivers/gpu/drm/panthor/Kconfig          |    1 -
>  drivers/gpu/drm/panthor/panthor_device.c |   11 +-
>  drivers/gpu/drm/panthor/panthor_device.h |   73 ++
>  drivers/gpu/drm/panthor/panthor_drv.c    |   33 +-
>  drivers/gpu/drm/panthor/panthor_fw.c     |   16 +-
>  drivers/gpu/drm/panthor/panthor_gem.c    | 1453 ++++++++++++++++++----
>  drivers/gpu/drm/panthor/panthor_gem.h    |  136 +-
>  drivers/gpu/drm/panthor/panthor_mmu.c    |  512 ++++++--
>  drivers/gpu/drm/panthor/panthor_mmu.h    |    8 +
>  drivers/gpu/drm/panthor/panthor_sched.c  |    9 +-
>  11 files changed, 1923 insertions(+), 339 deletions(-)
> 


      parent reply	other threads:[~2026-04-03 11:54 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-01 13:48 [PATCH v7 00/10] drm/panthor: Add a GEM shrinker Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 01/10] drm/gem: Consider GEM object reclaimable if shrinking fails Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 02/10] drm/panthor: Remove unused panthor_vm_op_ctx::map::new_vma field Boris Brezillon
2026-04-02 14:31   ` Liviu Dudau
2026-04-01 13:48 ` [PATCH v7 03/10] drm/panthor: Move panthor_gems_debugfs_init() to panthor_gem.c Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 04/10] drm/panthor: Group panthor_kernel_bo_xxx() helpers Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 05/10] drm/panthor: Don't call drm_gpuvm_bo_extobj_add() if the object is private Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 06/10] drm/panthor: Part ways with drm_gem_shmem_object Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 07/10] drm/panthor: Lazily allocate pages on mmap() Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 08/10] drm/panthor: Split panthor_vm_prepare_map_op_ctx() to prepare for reclaim Boris Brezillon
2026-04-01 13:48 ` [PATCH v7 09/10] drm/panthor: Track the number of mmap on a BO Boris Brezillon
2026-04-02 14:33   ` Liviu Dudau
2026-04-01 13:48 ` [PATCH v7 10/10] drm/panthor: Add a GEM shrinker Boris Brezillon
2026-04-03 11:54 ` Boris Brezillon [this message]

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=20260403135404.431fe625@fedora \
    --to=boris.brezillon@collabora.com \
    --cc=adrian.larumbe@collabora.com \
    --cc=airlied@gmail.com \
    --cc=akash.goel@arm.com \
    --cc=akhilpo@oss.qualcomm.com \
    --cc=aliceryhl@google.com \
    --cc=chris.diamand@arm.com \
    --cc=dakr@kernel.org \
    --cc=dmitry.osipenko@collabora.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=kernel@collabora.com \
    --cc=konradybcio@kernel.org \
    --cc=liviu.dudau@arm.com \
    --cc=maarten.lankhorst@linux.intel.com \
    --cc=matthew.brost@intel.com \
    --cc=mripard@kernel.org \
    --cc=olvaffe@gmail.com \
    --cc=robin.clark@oss.qualcomm.com \
    --cc=sean@poorly.run \
    --cc=simona@ffwll.ch \
    --cc=steven.price@arm.com \
    --cc=thomas.hellstrom@linux.intel.com \
    --cc=tzimmermann@suse.de \
    /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.