* [PATCH] fs: Fix uninitialized 'offp' in statmount_string()
@ 2025-10-11 9:13 Zhen Ni
2025-10-13 10:21 ` Jan Kara
2025-10-13 11:41 ` [PATCH v2] " Zhen Ni
0 siblings, 2 replies; 3+ messages in thread
From: Zhen Ni @ 2025-10-11 9:13 UTC (permalink / raw)
To: viro, brauner, jack; +Cc: linux-fsdevel, Zhen Ni, stable
In statmount_string(), most flags assign an output offset pointer (offp)
which is later updated with the string offset. However, the
STATMOUNT_MNT_UIDMAP and STATMOUNT_MNT_GIDMAP cases directly set the
struct fields instead of using offp. This leaves offp uninitialized,
leading to a possible uninitialized dereference when *offp is updated.
Fix it by assigning offp for UIDMAP and GIDMAP as well, keeping the code
path consistent.
Fixes: 37c4a9590e1e ("statmount: allow to retrieve idmappings")
Cc: stable@vger.kernel.org
Signed-off-by: Zhen Ni <zhen.ni@easystack.cn>
---
fs/namespace.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/namespace.c b/fs/namespace.c
index d82910f33dc4..5b5ab2ae238b 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -5454,11 +5454,11 @@ static int statmount_string(struct kstatmount *s, u64 flag)
ret = statmount_sb_source(s, seq);
break;
case STATMOUNT_MNT_UIDMAP:
- sm->mnt_uidmap = start;
+ offp = &sm->mnt_uidmap;
ret = statmount_mnt_uidmap(s, seq);
break;
case STATMOUNT_MNT_GIDMAP:
- sm->mnt_gidmap = start;
+ offp = &sm->mnt_gidmap;
ret = statmount_mnt_gidmap(s, seq);
break;
default:
--
2.20.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] fs: Fix uninitialized 'offp' in statmount_string()
2025-10-11 9:13 [PATCH] fs: Fix uninitialized 'offp' in statmount_string() Zhen Ni
@ 2025-10-13 10:21 ` Jan Kara
2025-10-13 11:41 ` [PATCH v2] " Zhen Ni
1 sibling, 0 replies; 3+ messages in thread
From: Jan Kara @ 2025-10-13 10:21 UTC (permalink / raw)
To: Zhen Ni; +Cc: viro, brauner, jack, linux-fsdevel, stable
On Sat 11-10-25 17:13:53, Zhen Ni wrote:
> In statmount_string(), most flags assign an output offset pointer (offp)
> which is later updated with the string offset. However, the
> STATMOUNT_MNT_UIDMAP and STATMOUNT_MNT_GIDMAP cases directly set the
> struct fields instead of using offp. This leaves offp uninitialized,
> leading to a possible uninitialized dereference when *offp is updated.
>
> Fix it by assigning offp for UIDMAP and GIDMAP as well, keeping the code
> path consistent.
>
> Fixes: 37c4a9590e1e ("statmount: allow to retrieve idmappings")
> Cc: stable@vger.kernel.org
> Signed-off-by: Zhen Ni <zhen.ni@easystack.cn>
The bug happened because of mismerge between commits 37c4a9590e1e and
e52e97f09fb6 so I think we should also add:
Fixes: e52e97f09fb6 ("statmount: let unset strings be empty")
Otherwise the patch looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/namespace.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/namespace.c b/fs/namespace.c
> index d82910f33dc4..5b5ab2ae238b 100644
> --- a/fs/namespace.c
> +++ b/fs/namespace.c
> @@ -5454,11 +5454,11 @@ static int statmount_string(struct kstatmount *s, u64 flag)
> ret = statmount_sb_source(s, seq);
> break;
> case STATMOUNT_MNT_UIDMAP:
> - sm->mnt_uidmap = start;
> + offp = &sm->mnt_uidmap;
> ret = statmount_mnt_uidmap(s, seq);
> break;
> case STATMOUNT_MNT_GIDMAP:
> - sm->mnt_gidmap = start;
> + offp = &sm->mnt_gidmap;
> ret = statmount_mnt_gidmap(s, seq);
> break;
> default:
> --
> 2.20.1
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCH v2] fs: Fix uninitialized 'offp' in statmount_string()
2025-10-11 9:13 [PATCH] fs: Fix uninitialized 'offp' in statmount_string() Zhen Ni
2025-10-13 10:21 ` Jan Kara
@ 2025-10-13 11:41 ` Zhen Ni
1 sibling, 0 replies; 3+ messages in thread
From: Zhen Ni @ 2025-10-13 11:41 UTC (permalink / raw)
To: viro, brauner, jack; +Cc: linux-fsdevel, Zhen Ni, stable
In statmount_string(), most flags assign an output offset pointer (offp)
which is later updated with the string offset. However, the
STATMOUNT_MNT_UIDMAP and STATMOUNT_MNT_GIDMAP cases directly set the
struct fields instead of using offp. This leaves offp uninitialized,
leading to a possible uninitialized dereference when *offp is updated.
Fix it by assigning offp for UIDMAP and GIDMAP as well, keeping the code
path consistent.
Fixes: 37c4a9590e1e ("statmount: allow to retrieve idmappings")
Fixes: e52e97f09fb6 ("statmount: let unset strings be empty")
Cc: stable@vger.kernel.org
Signed-off-by: Zhen Ni <zhen.ni@easystack.cn>
Reviewed-by: Jan Kara <jack@suse.cz>
---
Changes in v2:
- Add Fixes: e52e97f09fb6 ("statmount: let unset strings be empty")
---
fs/namespace.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/namespace.c b/fs/namespace.c
index d82910f33dc4..5b5ab2ae238b 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -5454,11 +5454,11 @@ static int statmount_string(struct kstatmount *s, u64 flag)
ret = statmount_sb_source(s, seq);
break;
case STATMOUNT_MNT_UIDMAP:
- sm->mnt_uidmap = start;
+ offp = &sm->mnt_uidmap;
ret = statmount_mnt_uidmap(s, seq);
break;
case STATMOUNT_MNT_GIDMAP:
- sm->mnt_gidmap = start;
+ offp = &sm->mnt_gidmap;
ret = statmount_mnt_gidmap(s, seq);
break;
default:
--
2.20.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-10-13 17:24 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-10-11 9:13 [PATCH] fs: Fix uninitialized 'offp' in statmount_string() Zhen Ni
2025-10-13 10:21 ` Jan Kara
2025-10-13 11:41 ` [PATCH v2] " Zhen Ni
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).