[PATCH] Yama: handle 32-bit userspace prctl

[PATCH] Yama: handle 32-bit userspace prctl

[Kees Cook]

When running a 64-bit kernel and receiving prctls from a 32-bit
userspace, the "-1" used as an unsigned long will end up being
misdetected. The kernel is looking for 0xffffffffffffffff instead of
0xffffffff. Since prctl lacks a distinct compat interface, Yama needs
to handle this translation itself. As such, support either value as
meaning PR_SET_PTRACER_ANY, to avoid breaking the ABI for 64-bit.

Signed-off-by: Kees Cook <keescook@chromium.org>
Cc: stable@vger.kernel.org
---
 security/yama/yama_lsm.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/security/yama/yama_lsm.c b/security/yama/yama_lsm.c
index 0cc99a3..dcd6178 100644
--- a/security/yama/yama_lsm.c
+++ b/security/yama/yama_lsm.c
@@ -143,7 +143,7 @@ static int yama_task_prctl(int option, unsigned long arg2, unsigned long arg3,
 		if (arg2 == 0) {
 			yama_ptracer_del(NULL, myself);
 			rc = 0;
-		} else if (arg2 == PR_SET_PTRACER_ANY) {
+		} else if (arg2 == PR_SET_PTRACER_ANY || (int)arg2 == -1) {
 			rc = yama_ptracer_add(NULL, myself);
 		} else {
 			struct task_struct *tracer;
-- 
1.7.0.4


-- 
Kees Cook
Chrome OS Security

Re: [PATCH] Yama: handle 32-bit userspace prctl

[Kees Cook]

Hi James,

Any chance this can get into security-next as well? It'll need to go
back to stable, but it's a small enough corner-case that I don't mind
it not landing in 3.6 directly. I'm happy if it gets into 3.6.x stable
at some point.

Thanks!

-Kees

On Mon, Aug 27, 2012 at 11:38 AM, Kees Cook <keescook@chromium.org> wrote:
> When running a 64-bit kernel and receiving prctls from a 32-bit
> userspace, the "-1" used as an unsigned long will end up being
> misdetected. The kernel is looking for 0xffffffffffffffff instead of
> 0xffffffff. Since prctl lacks a distinct compat interface, Yama needs
> to handle this translation itself. As such, support either value as
> meaning PR_SET_PTRACER_ANY, to avoid breaking the ABI for 64-bit.
>
> Signed-off-by: Kees Cook <keescook@chromium.org>
> Cc: stable@vger.kernel.org
> ---
>  security/yama/yama_lsm.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/security/yama/yama_lsm.c b/security/yama/yama_lsm.c
> index 0cc99a3..dcd6178 100644
> --- a/security/yama/yama_lsm.c
> +++ b/security/yama/yama_lsm.c
> @@ -143,7 +143,7 @@ static int yama_task_prctl(int option, unsigned long arg2, unsigned long arg3,
>                 if (arg2 == 0) {
>                         yama_ptracer_del(NULL, myself);
>                         rc = 0;
> -               } else if (arg2 == PR_SET_PTRACER_ANY) {
> +               } else if (arg2 == PR_SET_PTRACER_ANY || (int)arg2 == -1) {
>                         rc = yama_ptracer_add(NULL, myself);
>                 } else {
>                         struct task_struct *tracer;
> --
> 1.7.0.4
>
>
> --
> Kees Cook
> Chrome OS Security



-- 
Kees Cook
Chrome OS Security

Re: [PATCH] Yama: handle 32-bit userspace prctl

[John Johansen]

On 08/27/2012 11:38 AM, Kees Cook wrote:
> When running a 64-bit kernel and receiving prctls from a 32-bit
> userspace, the "-1" used as an unsigned long will end up being
> misdetected. The kernel is looking for 0xffffffffffffffff instead of
> 0xffffffff. Since prctl lacks a distinct compat interface, Yama needs
> to handle this translation itself. As such, support either value as
> meaning PR_SET_PTRACER_ANY, to avoid breaking the ABI for 64-bit.
> 
> Signed-off-by: Kees Cook <keescook@chromium.org>
> Cc: stable@vger.kernel.org
> ---
>  security/yama/yama_lsm.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/security/yama/yama_lsm.c b/security/yama/yama_lsm.c
> index 0cc99a3..dcd6178 100644
> --- a/security/yama/yama_lsm.c
> +++ b/security/yama/yama_lsm.c
> @@ -143,7 +143,7 @@ static int yama_task_prctl(int option, unsigned long arg2, unsigned long arg3,
>  		if (arg2 == 0) {
>  			yama_ptracer_del(NULL, myself);
>  			rc = 0;
> -		} else if (arg2 == PR_SET_PTRACER_ANY) {
> +		} else if (arg2 == PR_SET_PTRACER_ANY || (int)arg2 == -1) {
>  			rc = yama_ptracer_add(NULL, myself);
>  		} else {
>  			struct task_struct *tracer;
> 

yeah this looks good it should at least hit stable

Acked-by: John Johansen <john.johansen@canonical.com>

Re: [PATCH] Yama: handle 32-bit userspace prctl

[James Morris]

On Mon, 27 Aug 2012, Kees Cook wrote:

> When running a 64-bit kernel and receiving prctls from a 32-bit
> userspace, the "-1" used as an unsigned long will end up being
> misdetected. The kernel is looking for 0xffffffffffffffff instead of
> 0xffffffff. Since prctl lacks a distinct compat interface, Yama needs
> to handle this translation itself. As such, support either value as
> meaning PR_SET_PTRACER_ANY, to avoid breaking the ABI for 64-bit.
> 
> Signed-off-by: Kees Cook <keescook@chromium.org>
> Cc: stable@vger.kernel.org

Applied to
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git next