diff options
| author | Christian König <ckoenig.leichtzumerken@gmail.com> | 2024-04-15 15:48:21 +0200 |
|---|---|---|
| committer | Greg Kroah-Hartman <gregkh@linuxfoundation.org> | 2024-04-27 17:13:03 +0200 |
| commit | d77bfc8dc723cd7e28ba79f47fc581bfd6c2fafd (patch) | |
| tree | cb8392b53e96f20f4ba958ffa4b96d44b89529ed /arch | |
| parent | 2246736ab97e9c1fff1f805f6f79eba911ccccf0 (diff) | |
| download | linux-stable-d77bfc8dc723cd7e28ba79f47fc581bfd6c2fafd.tar.gz linux-stable-d77bfc8dc723cd7e28ba79f47fc581bfd6c2fafd.tar.bz2 linux-stable-d77bfc8dc723cd7e28ba79f47fc581bfd6c2fafd.zip | |
drm/ttm: stop pooling cached NUMA pages v2
commit b6976f323a8687cc0d55bc92c2086fd934324ed5 upstream.
We only pool write combined and uncached allocations because they
require extra overhead on allocation and release.
If we also pool cached NUMA it not only means some extra unnecessary
overhead, but also that under memory pressure it can happen that
pages from the wrong NUMA node enters the pool and are re-used
over and over again.
This can lead to performance reduction after running into memory
pressure.
v2: restructure and cleanup the code a bit from the internal hack to
test this.
Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: 4482d3c94d7f ("drm/ttm: add NUMA node id to the pool")
CC: stable@vger.kernel.org
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240415134821.1919-1-christian.koenig@amd.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'arch')
0 files changed, 0 insertions, 0 deletions