On 9/14/24 10:51, yangyun wrote:
> This may be a typo. The comment has said shared locks are
> not allowed when this bit is set. If using shared lock, the
> wait in `fuse_file_cached_io_open` may be forever.
>
> Fixes: 205c1d802683 ("fuse: allow parallel dio writes with FUSE_DIRECT_IO_ALLOW_MMAP")
> CC: stable@xxxxxxxxxxxxxxx
> Signed-off-by: yangyun <yangyun50@xxxxxxxxxx>
> ---
> fs/fuse/file.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/fuse/file.c b/fs/fuse/file.c
> index ca553d7a7c9e..e5f6affb0baa 100644
> --- a/fs/fuse/file.c
> +++ b/fs/fuse/file.c
> @@ -1345,7 +1345,7 @@ static bool fuse_dio_wr_exclusive_lock(struct kiocb *iocb, struct iov_iter *from
>
> /* shared locks are not allowed with parallel page cache IO */
> if (test_bit(FUSE_I_CACHE_IO_MODE, &fi->state))
> - return false;
> + return true;
>
> /* Parallel dio beyond EOF is not supported, at least for now. */
> if (fuse_io_past_eof(iocb, from))
Oh, another of these, maybe I should have named this function
'fuse_dio_deny_wr_shared_lock'. The good part is that it will be later
on caught in fuse_dio_lock/fuse_inode_uncached_io_start and then switch
to an exclusive lock. I.e. testing did not reveal it.
Reviewed-by: Bernd Schubert <bschubert@xxxxxxx>
