2009-11-04 10:50:09

by Eric W. Biederman

Subject: [PATCH] sysfs: Don't leak secdata when a sysfs_dirent is freed.

While refreshing my sysfs patches I noticed a leak in the secdata
implementation. We don't free the secdata when we free the
sysfs dirent.

This is a bug in 2.6.32-rc5 that we really should close.

Signed-off-by: Eric W. Biederman <[email protected]>
---

fs/sysfs/dir.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
index 5fad489..e020183 100644
--- a/fs/sysfs/dir.c
+++ b/fs/sysfs/dir.c
@@ -21,6 +21,7 @@
#include <linux/completion.h>
#include <linux/mutex.h>
#include <linux/slab.h>
+#include <linux/security.h>
#include "sysfs.h"

DEFINE_MUTEX(sysfs_mutex);
@@ -285,6 +286,9 @@ void release_sysfs_dirent(struct sysfs_dirent * sd)
sysfs_put(sd->s_symlink.target_sd);
if (sysfs_type(sd) & SYSFS_COPY_NAME)
kfree(sd->s_name);
+ if (sd->s_iattr && sd->s_iattr->ia_secdata)
+ security_release_secctx(sd->s_iattr->ia_secdata,
+ sd->s_iattr->ia_secdata_len);
kfree(sd->s_iattr);
sysfs_free_ino(sd->s_ino);
kmem_cache_free(sysfs_dir_cachep, sd);
--
1.6.5.2.143.g8cc62

2009-11-04 14:09:20

by Serge E. Hallyn

Subject: Re: [PATCH] sysfs: Don't leak secdata when a sysfs_dirent is freed.

Quoting Eric W. Biederman ([email protected]):
>
> While refreshing my sysfs patches I noticed a leak in the secdata
> implementation. We don't free the secdata when we free the
> sysfs dirent.
>
> This is a bug in 2.6.32-rc5 that we really should close.
>
> Signed-off-by: Eric W. Biederman <[email protected]>

I'm surprised no kmemleak or anything has found this - but I guess
not many sites hand-set custom selinux contexts on sysfs?

Acked-by: Serge Hallyn <[email protected]>

> ---
>
> fs/sysfs/dir.c | 4 ++++
> 1 files changed, 4 insertions(+), 0 deletions(-)
>
> diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
> index 5fad489..e020183 100644
> --- a/fs/sysfs/dir.c
> +++ b/fs/sysfs/dir.c
> @@ -21,6 +21,7 @@
> #include <linux/completion.h>
> #include <linux/mutex.h>
> #include <linux/slab.h>
> +#include <linux/security.h>
> #include "sysfs.h"
>
> DEFINE_MUTEX(sysfs_mutex);
> @@ -285,6 +286,9 @@ void release_sysfs_dirent(struct sysfs_dirent * sd)
> sysfs_put(sd->s_symlink.target_sd);
> if (sysfs_type(sd) & SYSFS_COPY_NAME)
> kfree(sd->s_name);
> + if (sd->s_iattr && sd->s_iattr->ia_secdata)
> + security_release_secctx(sd->s_iattr->ia_secdata,
> + sd->s_iattr->ia_secdata_len);
> kfree(sd->s_iattr);
> sysfs_free_ino(sd->s_ino);
> kmem_cache_free(sysfs_dir_cachep, sd);
> --
> 1.6.5.2.143.g8cc62
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html

2009-11-04 14:26:22

by Eric W. Biederman

Subject: Re: [PATCH] sysfs: Don't leak secdata when a sysfs_dirent is freed.

"Serge E. Hallyn" <[email protected]> writes:

> Quoting Eric W. Biederman ([email protected]):
>>
>> While refreshing my sysfs patches I noticed a leak in the secdata
>> implementation. We don't free the secdata when we free the
>> sysfs dirent.
>>
>> This is a bug in 2.6.32-rc5 that we really should close.
>>
>> Signed-off-by: Eric W. Biederman <[email protected]>
>
> I'm surprised no kmemleak or anything has found this - but I guess
> not many sites hand-set custom selinux contexts on sysfs?

I don't expect so since you couldn't customize it in 2.6.31.
This is brand new code, and this corner case was just overlooked
in the review.

Eric


> Acked-by: Serge Hallyn <[email protected]>
>
>> ---
>>
>> fs/sysfs/dir.c | 4 ++++
>> 1 files changed, 4 insertions(+), 0 deletions(-)
>>
>> diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
>> index 5fad489..e020183 100644
>> --- a/fs/sysfs/dir.c
>> +++ b/fs/sysfs/dir.c
>> @@ -21,6 +21,7 @@
>> #include <linux/completion.h>
>> #include <linux/mutex.h>
>> #include <linux/slab.h>
>> +#include <linux/security.h>
>> #include "sysfs.h"
>>
>> DEFINE_MUTEX(sysfs_mutex);
>> @@ -285,6 +286,9 @@ void release_sysfs_dirent(struct sysfs_dirent * sd)
>> sysfs_put(sd->s_symlink.target_sd);
>> if (sysfs_type(sd) & SYSFS_COPY_NAME)
>> kfree(sd->s_name);
>> + if (sd->s_iattr && sd->s_iattr->ia_secdata)
>> + security_release_secctx(sd->s_iattr->ia_secdata,
>> + sd->s_iattr->ia_secdata_len);
>> kfree(sd->s_iattr);
>> sysfs_free_ino(sd->s_ino);
>> kmem_cache_free(sysfs_dir_cachep, sd);
>> --
>> 1.6.5.2.143.g8cc62
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
>> the body of a message to [email protected]
>> More majordomo info at http://vger.kernel.org/majordomo-info.html

2009-11-04 21:41:18

by James Morris

Subject: [GIT] sysfs: Don't leak secdata when a sysfs_dirent is freed.

Hi Linus,

This fixes a memory leak introduced in 2.6.32-rc5, please pull.


The following changes since commit 91d3f9bacdb4950d2f79fe2ba296aa249f60d06c:
Linus Torvalds (1):
Merge branch 'for-linus' of git://git.kernel.org/.../anholt/drm-intel

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6 for-linus

Eric W. Biederman (1):
sysfs: Don't leak secdata when a sysfs_dirent is freed.

fs/sysfs/dir.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)


--
James Morris
<[email protected]>