* [PATCH] ext4: fix integer overflow when calculating commit interval
@ 2019-08-26 14:35 zhangyi (F)
2019-08-26 15:30 ` Jan Kara
0 siblings, 1 reply; 3+ messages in thread
From: zhangyi (F) @ 2019-08-26 14:35 UTC (permalink / raw)
To: linux-ext4; +Cc: tytso, jack, adilger.kernel, yi.zhang
If user specify a large enough value of "commit=" option, it may trigger
signed integer overflow which may lead to sbi->s_commit_interval becomes
a large or small value, zero in particular.
UBSAN: Undefined behaviour in ../fs/ext4/super.c:1592:31
signed integer overflow:
536870912 * 1000 cannot be represented in type 'int'
[...]
Call trace:
[...]
[<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166
[<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197
[<ffffff9008a2d95c>] __ubsan_handle_mul_overflow+0x4c/0x68 lib/ubsan.c:218
[<ffffff90086d070c>] handle_mount_opt fs/ext4/super.c:1592 [inline]
[<ffffff90086d070c>] parse_options+0x1724/0x1a40 fs/ext4/super.c:1773
[<ffffff90086d51c4>] ext4_remount+0x2ec/0x14a0 fs/ext4/super.c:4834
[...]
Although it is not a big deal, still silence the UBSAN by limit the
input value.
Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
---
fs/ext4/super.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 4079605d437a..7310facffa9d 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1874,6 +1874,13 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
} else if (token == Opt_commit) {
if (arg == 0)
arg = JBD2_DEFAULT_MAX_COMMIT_AGE;
+ else if (arg > INT_MAX / HZ) {
+ ext4_msg(sb, KERN_ERR,
+ "Invalid commit interval %d, "
+ "must be smaller than %d",
+ arg, INT_MAX / HZ);
+ return -1;
+ }
sbi->s_commit_interval = HZ * arg;
} else if (token == Opt_debug_want_extra_isize) {
sbi->s_want_extra_isize = arg;
--
2.23.0.rc2.8.gff66981
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] ext4: fix integer overflow when calculating commit interval
2019-08-26 14:35 [PATCH] ext4: fix integer overflow when calculating commit interval zhangyi (F)
@ 2019-08-26 15:30 ` Jan Kara
2019-08-28 15:25 ` Theodore Y. Ts'o
0 siblings, 1 reply; 3+ messages in thread
From: Jan Kara @ 2019-08-26 15:30 UTC (permalink / raw)
To: zhangyi (F); +Cc: linux-ext4, tytso, jack, adilger.kernel
On Mon 26-08-19 22:35:47, zhangyi (F) wrote:
> If user specify a large enough value of "commit=" option, it may trigger
> signed integer overflow which may lead to sbi->s_commit_interval becomes
> a large or small value, zero in particular.
>
> UBSAN: Undefined behaviour in ../fs/ext4/super.c:1592:31
> signed integer overflow:
> 536870912 * 1000 cannot be represented in type 'int'
> [...]
> Call trace:
> [...]
> [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166
> [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197
> [<ffffff9008a2d95c>] __ubsan_handle_mul_overflow+0x4c/0x68 lib/ubsan.c:218
> [<ffffff90086d070c>] handle_mount_opt fs/ext4/super.c:1592 [inline]
> [<ffffff90086d070c>] parse_options+0x1724/0x1a40 fs/ext4/super.c:1773
> [<ffffff90086d51c4>] ext4_remount+0x2ec/0x14a0 fs/ext4/super.c:4834
> [...]
>
> Although it is not a big deal, still silence the UBSAN by limit the
> input value.
>
> Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
Looks good to me. You can add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/ext4/super.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 4079605d437a..7310facffa9d 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -1874,6 +1874,13 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
> } else if (token == Opt_commit) {
> if (arg == 0)
> arg = JBD2_DEFAULT_MAX_COMMIT_AGE;
> + else if (arg > INT_MAX / HZ) {
> + ext4_msg(sb, KERN_ERR,
> + "Invalid commit interval %d, "
> + "must be smaller than %d",
> + arg, INT_MAX / HZ);
> + return -1;
> + }
> sbi->s_commit_interval = HZ * arg;
> } else if (token == Opt_debug_want_extra_isize) {
> sbi->s_want_extra_isize = arg;
> --
> 2.23.0.rc2.8.gff66981
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] ext4: fix integer overflow when calculating commit interval
2019-08-26 15:30 ` Jan Kara
@ 2019-08-28 15:25 ` Theodore Y. Ts'o
0 siblings, 0 replies; 3+ messages in thread
From: Theodore Y. Ts'o @ 2019-08-28 15:25 UTC (permalink / raw)
To: Jan Kara; +Cc: zhangyi (F), linux-ext4, adilger.kernel
On Mon, Aug 26, 2019 at 05:30:14PM +0200, Jan Kara wrote:
> On Mon 26-08-19 22:35:47, zhangyi (F) wrote:
> > If user specify a large enough value of "commit=" option, it may trigger
> > signed integer overflow which may lead to sbi->s_commit_interval becomes
> > a large or small value, zero in particular.
> >
> > UBSAN: Undefined behaviour in ../fs/ext4/super.c:1592:31
> > signed integer overflow:
> > 536870912 * 1000 cannot be represented in type 'int'
> > [...]
> > Call trace:
> > [...]
> > [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166
> > [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197
> > [<ffffff9008a2d95c>] __ubsan_handle_mul_overflow+0x4c/0x68 lib/ubsan.c:218
> > [<ffffff90086d070c>] handle_mount_opt fs/ext4/super.c:1592 [inline]
> > [<ffffff90086d070c>] parse_options+0x1724/0x1a40 fs/ext4/super.c:1773
> > [<ffffff90086d51c4>] ext4_remount+0x2ec/0x14a0 fs/ext4/super.c:4834
> > [...]
> >
> > Although it is not a big deal, still silence the UBSAN by limit the
> > input value.
> >
> > Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
>
> Looks good to me. You can add:
>
> Reviewed-by: Jan Kara <jack@suse.cz>
Thanks, applied.
- Ted
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-08-28 15:26 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-08-26 14:35 [PATCH] ext4: fix integer overflow when calculating commit interval zhangyi (F)
2019-08-26 15:30 ` Jan Kara
2019-08-28 15:25 ` Theodore Y. Ts'o
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).