NFSv4: fix out path in __nfs4_get_acl_uncached

[ Upstream commit f4e89f1a6dab4c063fc1e823cc9dddc408ff40cf ]

Another highly rare error case when a page allocating loop (inside
__nfs4_get_acl_uncached, this time) is not properly unwound on error.
Since pages array is allocated being uninitialized, need to free only
lower array indices. NULL checks were useful before commit 62a1573fcf84
("NFSv4 fix acl retrieval over krb5i/krb5p mounts") when the array had
been initialized to zero on stack.

Found by Linux Verification Center (linuxtesting.org).

Fixes: 62a1573fcf84 ("NFSv4 fix acl retrieval over krb5i/krb5p mounts")
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 177cb7b..d673836 100644 (file)
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -5995,9 +5995,8 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf,
 out_ok:
        ret = res.acl_len;
 out_free:
-       for (i = 0; i < npages; i++)
-               if (pages[i])
-                       __free_page(pages[i]);
+       while (--i >= 0)
+               __free_page(pages[i]);
        if (res.acl_scratch)
                __free_page(res.acl_scratch);
        kfree(pages);