Re: regress: bgpd: config: Fix attribute ordering

2020-03-06 Thread Claudio Jeker
On Fri, Mar 06, 2020 at 06:51:16PM +0100, Sebastian Benoit wrote: > I dont see that here. > Sure that you have an up-to-date tree? > And no diff in there? I think the problem is big-endian vs little-endian. I need to think about how and if this should be fixed. > Klemens Nanni(k...@openbsd.org)

Re: regress: bgpd: config: Fix attribute ordering

2020-03-06 Thread Klemens Nanni
On Fri, Mar 06, 2020 at 06:51:16PM +0100, Sebastian Benoit wrote: > I dont see that here. > Sure that you have an up-to-date tree? > And no diff in there? I run regress on sparc64 with a -CURRENT /usr/src tree; latest bgpd is installed from source, but the machine is currently running a snapshot f

Re: regress: bgpd: config: Fix attribute ordering

2020-03-06 Thread Sebastian Benoit
I dont see that here. Sure that you have an up-to-date tree? And no diff in there? Klemens Nanni(k...@openbsd.org) on 2020.03.05 23:39:20 +0100: > > I ran bgpd to test diffs and stumbled across what looks like simple > disorder in the config checks. > > bgpd must have changed in how it orders a

regress: bgpd: config: Fix attribute ordering

2020-03-05 Thread Klemens Nanni
I ran bgpd to test diffs and stumbled across what looks like simple disorder in the config checks. bgpd must have changed in how it orders attributes within `set { ... }' blocks; breaking the sets into multiple lines and diffing line-wise instead of word-wise shows that the printed config indee