Messages in this thread | | | Date | Fri, 11 Jul 2008 16:24:57 -0700 | From | Stephen Hemminger <> | Subject | Re: [PATCH] bridge: send correct MTU value in PMTU |
| |
On Fri, 11 Jul 2008 19:19:35 +0200 Simon Wunderlich <simon.wunderlich@s2003.tu-chemnitz.de> wrote:
> On Fri, Jul 11, 2008 at 08:34:23AM -0700, Stephen Hemminger wrote: > > On Fri, 11 Jul 2008 14:47:35 +0200 > > Simon Wunderlich <simon.wunderlich@s2003.tu-chemnitz.de> wrote: > > > > > [...] > > > > > > port should always be non-null so the existing fake_rtable can > > just go away no? > > I wondered about that too. But since bridge_parent() is also checking port > for NULL, and the code calling this is carefully checking these cases, i > assumed that this is quite possible. If you can assert that port is > non-NULL, we can remove this, but i have to admit that i don't get the > whole picture to be sure about this.
bridge_parent can only be NULL if the bridge is process of deletion (via RCU). So if bridge_parent() returns NULL, code should not bother with further processing and return (dropping skb if this function does that).
| |