On Sun, Nov 17, 2024 at 01:19:09PM -0800, John Hubbard wrote:
On 11/17/24 12:33 PM, gregkh@linuxfoundation.org wrote:
The patch below does not apply to the 6.11-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.11.y git checkout FETCH_HEAD git cherry-pick -x 94efde1d15399f5c88e576923db9bcd422d217f2 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to 'stable@vger.kernel.org' --in-reply-to '2024111754-stamina-flyer-1e05@gregkh' --subject-prefix 'PATCH 6.11.y' HEAD^..
It seems that the last hunk didn't apply because it was just too far away, as far as I can tell. I've manually applied it, resulting in the same diffs as the original, and did a quick smoke test (boot and ran mm tests).
Here's the updated version for 6.11.y:
From: John Hubbard jhubbard@nvidia.com Date: Sun, 17 Nov 2024 13:08:00 -0800 Subject: [PATCH] mm/gup: avoid an unnecessary allocation call for FOLL_LONGTERM cases X-NVConfidentiality: public Cc: John Hubbard jhubbard@nvidia.com
commit 53ba78de064b ("mm/gup: introduce check_and_migrate_movable_folios()") created a new constraint on the pin_user_pages*() API family: a potentially large internal allocation must now occur, for FOLL_LONGTERM cases.
A user-visible consequence has now appeared: user space can no longer pin more than 2GB of memory anymore on x86_64. That's because, on a 4KB PAGE_SIZE system, when user space tries to (indirectly, via a device driver that calls pin_user_pages()) pin 2GB, this requires an allocation of a folio pointers array of MAX_PAGE_ORDER size, which is the limit for kmalloc().
In addition to the directly visible effect described above, there is also the problem of adding an unnecessary allocation. The **pages array argument has already been allocated, and there is no need for a redundant **folios array allocation in this case.
Fix this by avoiding the new allocation entirely. This is done by referring to either the original page[i] within **pages, or to the associated folio. Thanks to David Hildenbrand for suggesting this approach and for providing the initial implementation (which I've tested and adjusted slightly) as well.
[jhubbard@nvidia.com]: tweaked the patch to apply to linux-stable/6.11.y [jhubbard@nvidia.com: whitespace tweak, per David] Link: https://lkml.kernel.org/r/131cf9c8-ebc0-4cbb-b722-22fa8527bf3c@nvidia.com [jhubbard@nvidia.com: bypass pofs_get_folio(), per Oscar] Link: https://lkml.kernel.org/r/c1587c7f-9155-45be-bd62-1e36c0dd6923@nvidia.com Link: https://lkml.kernel.org/r/20241105032944.141488-2-jhubbard@nvidia.com Fixes: 53ba78de064b ("mm/gup: introduce check_and_migrate_movable_folios()") Signed-off-by: John Hubbard jhubbard@nvidia.com Suggested-by: David Hildenbrand david@redhat.com Acked-by: David Hildenbrand david@redhat.com Reviewed-by: Oscar Salvador osalvador@suse.de Cc: Vivek Kasireddy vivek.kasireddy@intel.com Cc: Dave Airlie airlied@redhat.com Cc: Gerd Hoffmann kraxel@redhat.com Cc: Matthew Wilcox willy@infradead.org Cc: Christoph Hellwig hch@infradead.org Cc: Jason Gunthorpe jgg@nvidia.com Cc: Peter Xu peterx@redhat.com Cc: Arnd Bergmann arnd@arndb.de Cc: Daniel Vetter daniel.vetter@ffwll.ch Cc: Dongwon Kim dongwon.kim@intel.com Cc: Hugh Dickins hughd@google.com Cc: Junxiao Chang junxiao.chang@intel.com Cc: stable@vger.kernel.org Signed-off-by: Andrew Morton akpm@linux-foundation.org
mm/gup.c | 114 +++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 77 insertions(+), 37 deletions(-)
diff --git a/mm/gup.c b/mm/gup.c index 947881ff5e8f..fd3d7900c24b 100644 --- a/mm/gup.c +++ b/mm/gup.c @@ -2282,20 +2282,57 @@ struct page *get_dump_page(unsigned long addr) #endif /* CONFIG_ELF_CORE */
#ifdef CONFIG_MIGRATION
+/*
- An array of either pages or folios ("pofs"). Although it may seem
tempting to
- avoid this complication, by simply interpreting a list of folios as a
list of
- pages, that approach won't work in the longer term, because eventually
the
- layouts of struct page and struct folio will become completely
different.
- Furthermore, this pof approach avoids excessive page_folio() calls.
Patch is line-wrapped :(
Can you resend it in a format I can apply it in?
thanks,
greg k-h