2021-02-05 17:33:03

by Colin King

Subject: [PATCH] fs/jfs: fix potential integer overflow on shift of a int

From: Colin Ian King <[email protected]>

The left shift of int 32 bit integer constant 1 is evaluated using 32 bit
arithmetic and then assigned to a signed 64 bit integer. In the case where
l2nb is 32 or more this can lead to an overflow. Avoid this by shifting
using the BIT_ULL macro instead.

Addresses-Coverity: ("Uninitentional integer overflow")
Fixes: b40c2e665cd5 ("fs/jfs: TRIM support for JFS Filesystem")
Signed-off-by: Colin Ian King <[email protected]>
---
fs/jfs/jfs_dmap.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c
index 94b7c1cb5ceb..47dbca7e52e0 100644
--- a/fs/jfs/jfs_dmap.c
+++ b/fs/jfs/jfs_dmap.c
@@ -1656,7 +1656,7 @@ s64 dbDiscardAG(struct inode *ip, int agno, s64 minlen)
} else if (rc == -ENOSPC) {
/* search for next smaller log2 block */
l2nb = BLKSTOL2(nblocks) - 1;
- nblocks = 1 << l2nb;
+ nblocks = BIT_ULL(l2nb);
} else {
/* Trim any already allocated blocks */
jfs_error(bmp->db_ipbmap->i_sb, "-EIO\n");
--
2.29.2

2021-02-09 16:56:28

by Dave Kleikamp

Subject: Re: [PATCH] fs/jfs: fix potential integer overflow on shift of a int

On 2/5/21 11:11 AM, Colin King wrote:
> From: Colin Ian King <[email protected]>
>
> The left shift of int 32 bit integer constant 1 is evaluated using 32 bit
> arithmetic and then assigned to a signed 64 bit integer. In the case where
> l2nb is 32 or more this can lead to an overflow. Avoid this by shifting
> using the BIT_ULL macro instead.
>
> Addresses-Coverity: ("Uninitentional integer overflow")
> Fixes: b40c2e665cd5 ("fs/jfs: TRIM support for JFS Filesystem")
> Signed-off-by: Colin Ian King <[email protected]>
> ---
> fs/jfs/jfs_dmap.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c
> index 94b7c1cb5ceb..47dbca7e52e0 100644
> --- a/fs/jfs/jfs_dmap.c
> +++ b/fs/jfs/jfs_dmap.c
> @@ -1656,7 +1656,7 @@ s64 dbDiscardAG(struct inode *ip, int agno, s64 minlen)
> } else if (rc == -ENOSPC) {
> /* search for next smaller log2 block */
> l2nb = BLKSTOL2(nblocks) - 1;
> - nblocks = 1 << l2nb;
> + nblocks = BIT_ULL(l2nb);

I'm not sure I like the use of this macro here. It seems to imply a bit
flag of some sort. I think it would be clearer to use

nblocks = 1ULL << l2nb;

Maybe 1LL rather than 1ULL since nblocks is s64.

> } else {
> /* Trim any already allocated blocks */
> jfs_error(bmp->db_ipbmap->i_sb, "-EIO\n");
>