mbox series

[SRU,Jammy/oem-5.17,Focal/oem-5.14,0/1]

Message ID 20230201132258.37913-1-andrei.gherzan@canonical.com
Headers show
Series [1/1] xfrm: xfrm_policy: fix a possible double xfrm_pols_put() in xfrm_bundle_lookup() | expand

Message

Andrei Gherzan Feb. 1, 2023, 1:22 p.m. UTC 
[Impact]
There are circumstances where a double drop of a refcount can happen in
xfrm_expand_policies - net/xfrm/xfrm_policy.c .

[Potential Regression]
No regression is expected.

[Test]
The change has been applied cleanly on both series and tested with:
- clean build
- boot test in a clean VM for each series

Hangyu Hua (1):
  xfrm: xfrm_policy: fix a possible double xfrm_pols_put() in
    xfrm_bundle_lookup()

 net/xfrm/xfrm_policy.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

Comments

Stefan Bader Feb. 1, 2023, 2:53 p.m. UTC | #1 
On 01.02.23 14:22, Andrei Gherzan wrote:
> [Impact]
> There are circumstances where a double drop of a refcount can happen in
> xfrm_expand_policies - net/xfrm/xfrm_policy.c .
> 
> [Potential Regression]
> No regression is expected.
> 
> [Test]
> The change has been applied cleanly on both series and tested with:
> - clean build
> - boot test in a clean VM for each series
> 
> Hangyu Hua (1):
>    xfrm: xfrm_policy: fix a possible double xfrm_pols_put() in
>      xfrm_bundle_lookup()
> 
>   net/xfrm/xfrm_policy.c | 5 ++++-
>   1 file changed, 4 insertions(+), 1 deletion(-)
> 
This time to the list... the cover email subject also seems incomplete. So v2 
would be appreciated.

-Stefan