| Message ID | 20250304154818.250757-1-willy@infradead.org |
|---|---|
| State | New |
| Headers | show |
| Series | ceph: Fix error handling in fill_readdir_cache() | expand |
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index c15970fa240f..6ac2bd555e86 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -1870,9 +1870,12 @@ static int fill_readdir_cache(struct inode *dir, struct dentry *dn, ctl->folio = __filemap_get_folio(&dir->i_data, pgoff, fgf, mapping_gfp_mask(&dir->i_data)); - if (!ctl->folio) { + if (IS_ERR(ctl->folio)) { + int err = PTR_ERR(ctl->folio); + + ctl->folio = NULL; ctl->index = -1; - return idx == 0 ? -ENOMEM : 0; + return idx == 0 ? err : 0; } /* reading/filling the cache are serialized by * i_rwsem, no need to use folio lock */
__filemap_get_folio() returns an ERR_PTR, not NULL. There are extensive assumptions that ctl->folio is NULL, not an error pointer, so it seems better to fix this one place rather than change all the places which check ctl->folio. Fixes: baff9740bc8f ("ceph: Convert ceph_readdir_cache_control to store a folio") Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org> Cc: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com> --- fs/ceph/inode.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)