On 11/30/23 7:31 AM, gregkh@xxxxxxxxxxxxxxxxxxx wrote:
>
> The patch below does not apply to the 5.4-stable tree.
And the 5.4 variant attached.
--
Jens Axboe
From aa4df1efa47241f8b9faab93f9d73eda4aa9ccf3 Mon Sep 17 00:00:00 2001
From: Keith Busch <kbusch@xxxxxxxxxx>
Date: Mon, 20 Nov 2023 14:18:31 -0800
Subject: [PATCH] io_uring: fix off-by one bvec index
commit d6fef34ee4d102be448146f24caf96d7b4a05401 upstream.
If the offset equals the bv_len of the first registered bvec, then the
request does not include any of that first bvec. Skip it so that drivers
don't have to deal with a zero length bvec, which was observed to break
NVMe's PRP list creation.
Cc: stable@xxxxxxxxxxxxxxx
Fixes: bd11b3a391e3 ("io_uring: don't use iov_iter_advance() for fixed buffers")
Signed-off-by: Keith Busch <kbusch@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20231120221831.2646460-1-kbusch@xxxxxxxx
Signed-off-by: Jens Axboe <axboe@xxxxxxxxx>
---
fs/io_uring.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 3683ddeb625a..adc263400471 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -1256,7 +1256,7 @@ static int io_import_fixed(struct io_ring_ctx *ctx, int rw,
*/
const struct bio_vec *bvec = imu->bvec;
- if (offset <= bvec->bv_len) {
+ if (offset < bvec->bv_len) {
iov_iter_advance(iter, offset);
} else {
unsigned long seg_skip;
--
2.42.0
