On Wed, Aug 27, 2014 at 10:06:53AM +0200, Milan Kocian wrote:
> On Sun, Jun 29, 2014 at 12:54:56PM +0200, David Lamparter wrote:
> > On Wed, Jun 04, 2014 at 01:31:17AM +0200, David Lamparter wrote:
> > > On Wed, Jun 04, 2014 at 01:12:59AM +0200, David Lamparter wrote:
> > > > This series fixes several issues in attribute handling, related to 
> > > > malformed
> > > > attributes and/or route-map "set" applications.
> > > > 
> > > > NB: do not apply 3/6 without 2/6 (or duplicating the fix).
> > > > 
> > > > Thanks to Milan Kocian, your information was quite useful and the 
> > > > hacked patch
> > > > fixed half the problem already ;)
> > > > 
> > > > - [1/6] bgpd: fix some bgp_update_main() attribute leaks
> > > > - [2/6] bgpd: remove duplicate route-map extcommunity code
> > > > - [3/6] bgpd: fix double free after extcommunity set (BZ#799)
> > > > - [4/6] bgpd: fix memory leak on malformed attribute
> > > > - [5/6] bgpd: fix IP endianness in debug message
> > > > - [6/6] bgpd: don't send NOTIFY twice for malformed attrs
> > > 
> > > P.S.: this is not going into 0.99.23 because (a) it doesn't fit the
> > > conditions (it's not a regression since 0.99.21), and (b) it hasn't been
> > > reviewed by someone else, and there might well be new bugs in it.
> > > 
> > > However, i'll probably push it onto the stable/0.99.23 branch, which
> > > will pop up after the release.  Meaning we might have a 0.99.23.1 rather
> > > soon.
> > > 
> > > (The rationale is:  if it was broken for a long time, hurrying the fix
> > > and possibly breaking the release in a more horrible way has no
> > > justification, being broken for a little more won't make a huge
> > > difference.)
> > 
> > Pushed onto stable/0.99.23 and merged into master.
> > 
> > -David
> > 
> 
> Hello,
> 
> I am very sorry for the delay. I had overlooked this rest of thread.
> 
> I will test version 0.99.23.1 as soon as possible and send result.
> 
> Many thanks for solving it.
> 
> Best regards,
> 

Hello,

only for the record (again sorry for long delay). The fix works without
problem in our production environment.

Best regards,

-- 
Milan Kocian

_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev

Reply via email to