2022-04-08 09:26:59

by Xiaomeng Tong

[permalink] [raw]
Subject: [PATCH v3] md: fix an incorrect NULL check in does_sb_need_changing

The bug is here:
if (!rdev)

The list iterator value 'rdev' will *always* be set and non-NULL
by rdev_for_each(), so it is incorrect to assume that the iterator
value will be NULL if the list is empty or no element found.
Otherwise it will bypass the NULL check and lead to invalid memory
access passing the check.

To fix the bug, use a new variable 'iter' as the list iterator,
while using the original variable 'rdev' as a dedicated pointer to
point to the found element.

Cc: [email protected]
Fixes: 2aa82191ac36 ("md-cluster: Perform a lazy update")
Acked-by: Guoqing Jiang <[email protected]>
Signed-off-by: Xiaomeng Tong <[email protected]>
---

changes since v2:
- fix typo (Song Liu)

changes since v1:
- rephrase the subject (Guoqing Jiang)
- add Acked-by: for Guoqing Jiang

v2:https://lore.kernel.org/lkml/[email protected]/
v1:https://lore.kernel.org/lkml/[email protected]/

---
drivers/md/md.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4d38bd7dadd6..7476fc204172 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -2629,14 +2629,16 @@ static void sync_sbs(struct mddev *mddev, int nospares)

static bool does_sb_need_changing(struct mddev *mddev)
{
- struct md_rdev *rdev;
+ struct md_rdev *rdev = NULL, *iter;
struct mdp_superblock_1 *sb;
int role;

/* Find a good rdev */
- rdev_for_each(rdev, mddev)
- if ((rdev->raid_disk >= 0) && !test_bit(Faulty, &rdev->flags))
+ rdev_for_each(iter, mddev)
+ if ((iter->raid_disk >= 0) && !test_bit(Faulty, &iter->flags)) {
+ rdev = iter;
break;
+ }

/* No good device found. */
if (!rdev)
--
2.17.1


2022-04-08 16:50:01

by Goldwyn Rodrigues

[permalink] [raw]
Subject: Re: [PATCH v3] md: fix an incorrect NULL check in does_sb_need_changing

On 16:37 08/04, Xiaomeng Tong wrote:
> The bug is here:
> if (!rdev)
>
> The list iterator value 'rdev' will *always* be set and non-NULL
> by rdev_for_each(), so it is incorrect to assume that the iterator
> value will be NULL if the list is empty or no element found.
> Otherwise it will bypass the NULL check and lead to invalid memory
> access passing the check.
>
> To fix the bug, use a new variable 'iter' as the list iterator,
> while using the original variable 'rdev' as a dedicated pointer to
> point to the found element.
>
> Cc: [email protected]
> Fixes: 2aa82191ac36 ("md-cluster: Perform a lazy update")
> Acked-by: Guoqing Jiang <[email protected]>
> Signed-off-by: Xiaomeng Tong <[email protected]>

Also safeguards from reading sb from a faulty device if all devices are
faulty.

Acked-by: Goldwyn Rodrigues <[email protected]>

> v2:https://lore.kernel.org/lkml/[email protected]/
> v1:https://lore.kernel.org/lkml/[email protected]/
>
> ---
> drivers/md/md.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/md/md.c b/drivers/md/md.c
> index 4d38bd7dadd6..7476fc204172 100644
> --- a/drivers/md/md.c
> +++ b/drivers/md/md.c
> @@ -2629,14 +2629,16 @@ static void sync_sbs(struct mddev *mddev, int nospares)
>
> static bool does_sb_need_changing(struct mddev *mddev)
> {
> - struct md_rdev *rdev;
> + struct md_rdev *rdev = NULL, *iter;
> struct mdp_superblock_1 *sb;
> int role;
>
> /* Find a good rdev */
> - rdev_for_each(rdev, mddev)
> - if ((rdev->raid_disk >= 0) && !test_bit(Faulty, &rdev->flags))
> + rdev_for_each(iter, mddev)
> + if ((iter->raid_disk >= 0) && !test_bit(Faulty, &iter->flags)) {
> + rdev = iter;
> break;
> + }
>
> /* No good device found. */
> if (!rdev)
> --
> 2.17.1
>

--
Goldwyn

2022-04-11 06:13:58

by Song Liu

[permalink] [raw]
Subject: Re: [PATCH v3] md: fix an incorrect NULL check in does_sb_need_changing

On Fri, Apr 8, 2022 at 5:23 AM Goldwyn Rodrigues <[email protected]> wrote:
>
> On 16:37 08/04, Xiaomeng Tong wrote:
> > The bug is here:
> > if (!rdev)
> >
> > The list iterator value 'rdev' will *always* be set and non-NULL
> > by rdev_for_each(), so it is incorrect to assume that the iterator
> > value will be NULL if the list is empty or no element found.
> > Otherwise it will bypass the NULL check and lead to invalid memory
> > access passing the check.
> >
> > To fix the bug, use a new variable 'iter' as the list iterator,
> > while using the original variable 'rdev' as a dedicated pointer to
> > point to the found element.
> >
> > Cc: [email protected]
> > Fixes: 2aa82191ac36 ("md-cluster: Perform a lazy update")
> > Acked-by: Guoqing Jiang <[email protected]>
> > Signed-off-by: Xiaomeng Tong <[email protected]>
>
> Also safeguards from reading sb from a faulty device if all devices are
> faulty.
>
> Acked-by: Goldwyn Rodrigues <[email protected]>

Applied to md-next. Thanks!