block_dev: don't update file access position for sync direct IO

For sync direct IO, generic_file_direct_write/generic_file_read_iter
will update file access position. Don't duplicate the update in
.direct_IO. This cause my raid array can't assemble.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Jens Axboe <redacted>
Signed-off-by: Shaohua Li <redacted>
---
 fs/block_dev.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/fs/block_dev.c b/fs/block_dev.c
index 95acbd2..1eb7dcf 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c

@@ -264,7 +264,6 @@ __blkdev_direct_IO_simple(struct kiocb *iocb, struct iov_iter *iter,
 
 	if (unlikely(bio.bi_error))
 		return bio.bi_error;
-	iocb->ki_pos += ret;
 	return ret;
 }
 

@@ -411,10 +410,8 @@ __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, int nr_pages)
 	__set_current_state(TASK_RUNNING);
 
 	ret = dio->bio.bi_error;
-	if (likely(!ret)) {
+	if (likely(!ret))
 		ret = dio->size;
-		iocb->ki_pos += ret;
-	}
 
 	bio_put(&dio->bio);
 	return ret;

-- 
2.9.3

On 12/13/2016 08:07 PM, Shaohua Li wrote:

For sync direct IO, generic_file_direct_write/generic_file_read_iter
will update file access position. Don't duplicate the update in
.direct_IO. This cause my raid array can't assemble.

That's not great... Thanks, added.

-- 
Jens Axboe

On Tue, Dec 13, 2016 at 09:08:18PM -0700, Jens Axboe wrote:

That's not great... Thanks, added.

Ooops, yeah - we're still going through ->direct_IO for block devices.
I'll take a stab at removing that, as it's just a pointless indirect
call.