Re: [PATCH v4 1/2] cifs: don't update i_size in cifs_do_truncate() without a cached handle

On Sat, Aug 22, 2026 at 3:41 AM Frank Sorenson <[email protected]> wrote:
>
> If find_writable_file() returns NULL, cifs_file_flush() returns 0
> without issuing set_file_size, and the outer if (!rc) block sets i_size
> to 0 before the server is told to truncate.  If the subsequent
> cifs_open() fails, the inode is stranded at size 0 while the server
> file is unchanged.
>
> Move netfs_resize_file() and cifs_setsize() inside the if (cfile) arm
> so they only run after a successful set_file_size().  In the no-handle
> else branch, call truncate_inode_pages() to evict stale pages before
> the O_TRUNC open: cifs_invalidate_cache() is fscache-only and does not
> touch the VFS page cache, so without it old pages persist and may be
> served as stale data after a later file extension.
>
> Fixes: 110fee6b9bb5 ("smb: client: fix missing timestamp updates with O_TRUNC")
> Cc: [email protected]
> Cc: Paulo Alcantara <[email protected]>
> Signed-off-by: Frank Sorenson <[email protected]>
> ---
>  fs/smb/client/file.c | 18 ++++++++++++++----
>  1 file changed, 14 insertions(+), 4 deletions(-)
>
> diff --git a/fs/smb/client/file.c b/fs/smb/client/file.c
> index 389083f9ce00..05aeb9d94ffe 100644
> --- a/fs/smb/client/file.c
> +++ b/fs/smb/client/file.c
> @@ -1012,10 +1012,20 @@ static int cifs_do_truncate(const unsigned int xid, struct dentry *dentry)
>                         server = tcon->ses->server;
>                         rc = server->ops->set_file_size(xid, tcon,
>                                                         cfile, 0, false);
> -               }
> -               if (!rc) {
> -                       netfs_resize_file(&cinode->netfs, 0, true);
> -                       cifs_setsize(inode, 0);
> +                       if (!rc) {
> +                               netfs_resize_file(&cinode->netfs, 0, true);
> +                               cifs_setsize(inode, 0);
> +                               cifs_invalidate_cache(inode, 0);
> +                       }
> +               } else {
> +                       /*
> +                        * No cached handle; the server truncates as part of the
> +                        * O_TRUNC open request that follows.  Evict stale pages
> +                        * from the page cache now so they cannot be served as
> +                        * stale data after the file is later extended; i_size is
> +                        * updated from the server's open response.
> +                        */
> +                       truncate_inode_pages(inode->i_mapping, 0);
truncate_inode_pages() expects both ->i_rwsem and ->invalidate_lock to
be held. Please check the comment of truncate_inode_pages().