* [PATCH v2] iov_iter: Use iov_offset for length calculation in iov_iter_aligned_bvec
[not found] <CGME20250426080212epcas5p2d0c592a4bd539eed242d5ac31d23b323@epcas5p2.samsung.com>
@ 2025-04-26 7:53 ` Nitesh Shetty
2025-04-28 6:27 ` Anuj Gupta
0 siblings, 1 reply; 2+ messages in thread
From: Nitesh Shetty @ 2025-04-26 7:53 UTC (permalink / raw)
To: Alexander Viro, Andrew Morton
Cc: axboe, gost.dev, nitheshshetty, Nitesh Shetty, linux-kernel
If iov_offset is non-zero, then we need to consider iov_offset in length
calculation, otherwise we might pass smaller IOs such as 512 bytes, in
below scenario[1].
This issue is reproducible using lib-uring test/fixed-seg.c application
with fixed buffer on a 512 LBA formatted device.
Fixes â3639f96f24a121ec9f037981b81daf5a8d60a
[1]
At present we pass the alignment check,
for 512 LBA formatted devices, len_mask = 511
when IO is smaller, i->count = 512
has an offset, i->io_offset = 3584
with bvec values, bvec->bv_offset = 256, bvec->bv_len = 3840.
In short, the first 256 bytes are in the current page,
next 256 bytes are in the second page.
Ideally we expect to fail the IO.
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
---
lib/iov_iter.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/iov_iter.c b/lib/iov_iter.c
index bc9391e55d57..9ce83ab71bac 100644
--- a/lib/iov_iter.c
+++ b/lib/iov_iter.c
@@ -820,7 +820,7 @@ static bool iov_iter_aligned_bvec(const struct iov_iter *i, unsigned addr_mask,
size_t size = i->count;
do {
- size_t len = bvec->bv_len;
+ size_t len = bvec->bv_len - skip;
if (len > size)
len = size;
base-commit: 02ddfb981de88a2c15621115dd7be2431252c568
--
2.43.0
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v2] iov_iter: Use iov_offset for length calculation in iov_iter_aligned_bvec
2025-04-26 7:53 ` [PATCH v2] iov_iter: Use iov_offset for length calculation in iov_iter_aligned_bvec Nitesh Shetty
@ 2025-04-28 6:27 ` Anuj Gupta
0 siblings, 0 replies; 2+ messages in thread
From: Anuj Gupta @ 2025-04-28 6:27 UTC (permalink / raw)
To: Nitesh Shetty
Cc: Alexander Viro, Andrew Morton, axboe, gost.dev, nitheshshetty,
linux-kernel
[-- Attachment #1: Type: text/plain, Size: 563 bytes --]
On Sat, Apr 26, 2025 at 01:23:10PM +0530, Nitesh Shetty wrote:
> If iov_offset is non-zero, then we need to consider iov_offset in length
> calculation, otherwise we might pass smaller IOs such as 512 bytes, in
> below scenario[1].
> This issue is reproducible using lib-uring test/fixed-seg.c application
> with fixed buffer on a 512 LBA formatted device.
> Fixes â3639f96f24a121ec9f037981b81daf5a8d60a
>
This needs a proper fixes tag format
Fixes: <12 characters of the SHA-1 ID> ("commit description")
With that
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
[-- Attachment #2: Type: text/plain, Size: 0 bytes --]
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-04-28 9:07 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <CGME20250426080212epcas5p2d0c592a4bd539eed242d5ac31d23b323@epcas5p2.samsung.com>
2025-04-26 7:53 ` [PATCH v2] iov_iter: Use iov_offset for length calculation in iov_iter_aligned_bvec Nitesh Shetty
2025-04-28 6:27 ` Anuj Gupta
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.