On Sat, 6 May 2023 10:34:31 -0700 Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > On Sat, May 6, 2023 at 10:10 AM Linus Torvalds > <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > > > > .. in the meantime, I did end up applying your patch. > > Final (?) note on this: I not only applied your patch, but went > looking for any other missed cases. > > And found one in fs/nfs/dir.c, which like ext4 had grown a new use of > __filemap_get_folio(). > > But unlike ext4 it hadn't been caught in linux-next (or I had just > missed it) and so I hadn't caught it in my merge either. > > I hope that's the last one. > > I grepped for all these __filemap_get_folio() cases when I did the MM > merge that brought in that change (and did the ext4 merge fixup), but > then the nfs pull happened later and I didn't think to check for new > cases... > > A current grep seems to say that it's all good. But we had all missed > the second check in filemap_fault(), so... > We have a related afs fix which I plan to send over later today: From: Christoph Hellwig <hch@xxxxxx> Subject: afs: fix the afs_dir_get_folio return value Date: Wed, 3 May 2023 17:45:26 +0200 Keep returning NULL on failure instead of letting an ERR_PTR escape to callers that don't expect it. Link: https://lkml.kernel.org/r/20230503154526.1223095-2-hch@xxxxxx Fixes: 66dabbb65d67 ("mm: return an ERR_PTR from __filemap_get_folio") Signed-off-by: Christoph Hellwig <hch@xxxxxx> Reported-by: Jan Kara <jack@xxxxxxx> Reviewed-by: Jan Kara <jack@xxxxxxx> Reviewed-by: David Howells <dhowells@xxxxxxxxxx> Tested-by: David Howells <dhowells@xxxxxxxxxx> Cc: Marc Dionne <marc.dionne@xxxxxxxxxxxx> Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- fs/afs/dir_edit.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) --- a/fs/afs/dir_edit.c~afs-fix-the-afs_dir_get_folio-return-value +++ a/fs/afs/dir_edit.c @@ -115,11 +115,12 @@ static struct folio *afs_dir_get_folio(s folio = __filemap_get_folio(mapping, index, FGP_LOCK | FGP_ACCESSED | FGP_CREAT, mapping->gfp_mask); - if (IS_ERR(folio)) + if (IS_ERR(folio)) { clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); - else if (folio && !folio_test_private(folio)) + return NULL; + } + if (!folio_test_private(folio)) folio_attach_private(folio, (void *)1); - return folio; } _
