If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com --- drivers/vdpa/vdpa_user/vduse_dev.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/drivers/vdpa/vdpa_user/vduse_dev.c b/drivers/vdpa/vdpa_user/vduse_dev.c index 41c0b29739f1..35dceee3ed56 100644 --- a/drivers/vdpa/vdpa_user/vduse_dev.c +++ b/drivers/vdpa/vdpa_user/vduse_dev.c @@ -673,10 +673,15 @@ static void vduse_vdpa_get_config(struct vdpa_device *vdpa, unsigned int offset, { struct vduse_dev *dev = vdpa_to_vduse(vdpa);
- if (offset > dev->config_size || - len > dev->config_size - offset) + /* Initialize the buffer in case of partial copy. */ + memset(buf, 0, len); + + if (offset > dev->config_size) return;
+ if (len > dev->config_size - offset) + len = dev->config_size - offset; + memcpy(buf, dev->config + offset, len); }
On Mon, Aug 29, 2022 at 09:34:24AM +0200, Maxime Coquelin wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Please no blank line above the Acked-by: line here if possible.
thanks,
greg k-h
On 8/29/22 09:48, Greg KH wrote:
On Mon, Aug 29, 2022 at 09:34:24AM +0200, Maxime Coquelin wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Please no blank line above the Acked-by: line here if possible.
Sure.
Jason, do you prefer I post a new revision with this single change or you will handle it while applying? Either way is fine to me.
Thanks, Maxime
thanks,
greg k-h
On Wed, Aug 31, 2022 at 05:01:11PM +0200, Maxime Coquelin wrote:
On 8/29/22 09:48, Greg KH wrote:
On Mon, Aug 29, 2022 at 09:34:24AM +0200, Maxime Coquelin wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Please no blank line above the Acked-by: line here if possible.
Sure.
Jason, do you prefer I post a new revision with this single change or you will handle it while applying? Either way is fine to me.
Thanks, Maxime
Repost pls, easier.
thanks,
greg k-h
On Wed, Aug 31, 2022 at 05:01:11PM +0200, Maxime Coquelin wrote:
On 8/29/22 09:48, Greg KH wrote:
On Mon, Aug 29, 2022 at 09:34:24AM +0200, Maxime Coquelin wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Please no blank line above the Acked-by: line here if possible.
Sure.
Jason, do you prefer I post a new revision with this single change or you will handle it while applying? Either way is fine to me.
Thanks, Maxime
I queue these normally.
thanks,
greg k-h
On 8/31/22 17:46, Michael S. Tsirkin wrote:
On Wed, Aug 31, 2022 at 05:01:11PM +0200, Maxime Coquelin wrote:
On 8/29/22 09:48, Greg KH wrote:
On Mon, Aug 29, 2022 at 09:34:24AM +0200, Maxime Coquelin wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Please no blank line above the Acked-by: line here if possible.
Sure.
Jason, do you prefer I post a new revision with this single change or you will handle it while applying? Either way is fine to me.
Thanks, Maxime
I queue these normally.
Ok, I'm preparing the V2.
Thanks, Maxime
thanks,
greg k-h
On Mon, Aug 29, 2022 at 3:34 PM Maxime Coquelin maxime.coquelin@redhat.com wrote:
If the VDUSE application provides a smaller config space than the driver expects, the driver may use uninitialized memory from the stack.
This patch prevents it by initializing the buffer passed by the driver to store the config value.
This fix addresses CVE-2022-2308.
Cc: xieyongji@bytedance.com Cc: stable@vger.kernel.org # v5.15+ Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Acked-by: Jason Wang jasowang@redhat.com Signed-off-by: Maxime Coquelin maxime.coquelin@redhat.com
Reviewed-by: Xie Yongji xieyongji@bytedance.com
Thanks, Yongji
linux-stable-mirror@lists.linaro.org