netdev
[Top] [All Lists]

Re: Oops in arp_rcv, patch

To: "David S. Miller" <davem@xxxxxxxxxx>
Subject: Re: Oops in arp_rcv, patch
From: Jacek Konieczny <jajcus@xxxxxxx>
Date: Fri, 6 Jul 2001 10:35:08 +0200
Cc: netdev@xxxxxxxxxxx
In-reply-to: <15173.230.571263.353935@xxxxxxxxxxxxxxx>
Mail-followup-to: "David S. Miller" <davem@xxxxxxxxxx>, netdev@xxxxxxxxxxx
Sender: owner-netdev@xxxxxxxxxxx
User-agent: Mutt/1.3.18i
On Thu, Jul 05, 2001 at 05:05:58PM -0700, David S. Miller wrote:
> 
> Jacek Konieczny writes:
>  > -                          neigh_release(n);
>  > +                          if (n) neigh_release(n);
>  ...
>  > The buggy code seems unchanged in 2.4.5 kernel.
> 
> It looks perfectly fine to me, all neigh_event_ns() return arguments
> are checked in the current 2.4.x sources, I don't know what you are
> talking about wrt. 2.4.5 since it is fine there too:
> 
>                       n = neigh_event_ns(&arp_tbl, sha, &sip, dev);
>                       if (n) {
> ...
>                               neigh_release(n);
>                       }
> ...
>                               n = neigh_event_ns(&arp_tbl, sha, &sip, dev);
>                               if (n)
>                                       neigh_release(n);
> ...
> 
> What made you think that 2.4.5 still has the bug?  You must have been
> looking at some other source tree.
Or maybe I was looking not good enough. Sorry, for false alarm.
But the bug is really present in 2.2.19 and the patch was made for that
version. At least, after applying that my router started building its
uptime again.

Greets,
        Jacek

<Prev in Thread] Current Thread [Next in Thread>