Shifting reg.bgid << IORING_OFF_PBUF_SHIFT results in a promotion
from __u16 to a 32 bit signed integer, this is then sign extended
to a 64 bit unsigned long on 64 bit architectures. If reg.bgid is
greater than 0x7fff then this leads to a sign extended result where
all the upper 32 bits of mmap_offset are set to 1. Fix this by
casting reg.bgid to the same type as mmap_offset before performing
the shift.

Fixes: ff4afde8a61f ("io_uring/kbuf: use region api for pbuf rings")
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
---
 io_uring/kbuf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/io_uring/kbuf.c b/io_uring/kbuf.c
index e91260a6156b..15e5e6ec5968 100644
--- a/io_uring/kbuf.c
+++ b/io_uring/kbuf.c
@@ -640,7 +640,7 @@ int io_register_pbuf_ring(struct io_ring_ctx *ctx, void __user *arg)
 			return -ENOMEM;
 	}
 
-	mmap_offset = reg.bgid << IORING_OFF_PBUF_SHIFT;
+	mmap_offset = (unsigned long)reg.bgid << IORING_OFF_PBUF_SHIFT;
 	ring_size = flex_array_size(br, bufs, reg.ring_entries);
 
 	memset(&rd, 0, sizeof(rd));
-- 
2.39.5

On Wed, 04 Dec 2024 15:39:23 +0000, Colin Ian King wrote:
> Shifting reg.bgid << IORING_OFF_PBUF_SHIFT results in a promotion
> from __u16 to a 32 bit signed integer, this is then sign extended
> to a 64 bit unsigned long on 64 bit architectures. If reg.bgid is
> greater than 0x7fff then this leads to a sign extended result where
> all the upper 32 bits of mmap_offset are set to 1. Fix this by
> casting reg.bgid to the same type as mmap_offset before performing
> the shift.
> 
> [...]

Applied, thanks!

[1/1] io_uring/kbuf: fix unintentional sign extension on shift of reg.bgid
      commit: e54fb80b57e4534ae91e65ea14cc66b9f21b4b7a

Best regards,
-- 
Jens Axboe