I will continue to argue that testing out-of-the-tree is wrong, and
instead the regression tests should test the installed binaries.
Less than 10 regression tests are going this way. I think it is
a problematic fad.
Moritz Buhl wrote:
> Hi,
> I noticed that the iked regress test fails if I ha
Hi,
I noticed that the iked regress test fails if I have an obj directory.
The following patch adresses this.
mbuhl
Index: regress/sbin/iked/parser/Makefile
===
RCS file: /mount/openbsd/cvs//src/regress/sbin/iked/parser/Makefile,v
re
On Sat, Jun 22, 2019 at 11:16:59AM -0600, Theo de Raadt wrote:
> I can give a few crazy examples: ld, cc, ksh. I'll say again, there
> surely are cases where it is pointless making usage be complete, because
> the compleness can be harmful. Is this one? Maybe...
Fair point, although these tools
On Sat, Jun 22, 2019 at 11:05:00AM -0600, Theo de Raadt wrote:
> I'm not happy with usage messages which fill half a screen. There has
> to be some threshold where we include very little information, and force
> people to the manual page instead. The many *ctl programs stradle this
> threshold an
Theo de Raadt wrote:
> I can give a few crazy examples: ld, cc, ksh. I'll say again, there
> surely are cases where it is pointless making usage be complete, because
> the compleness can be harmful. Is this one? Maybe...
I've been burned a few times by bgpctl having such a long usage.
# bgpct
Klemens Nanni wrote:
> On Sat, Jun 22, 2019 at 07:00:53PM +0200, Mark Kettenis wrote:
> > Not sure such a long list is actually useful, but sure.
> I'd argue it is complete and consistent. Both `man -h ldomctl' and
> `ldomctl [-h]' would lack information, so there was no way of getting
> a quick
On Sat, Jun 22, 2019 at 07:00:53PM +0200, Mark Kettenis wrote:
> Not sure such a long list is actually useful, but sure.
I'd argue it is complete and consistent. Both `man -h ldomctl' and
`ldomctl [-h]' would lack information, so there was no way of getting
a quick overview without reading the ent
Klemens Nanni wrote:
> ldomctl(8) describes much more commands than the poor usage:
>
> $ ldomctl
> usage: ldomctl start|stop|panic domain
> ldomctl status [domain]
>
> Doing as vmctl already does, this diff turns it into
>
> usage: ldomctl command [argument ...
> Date: Sat, 22 Jun 2019 18:35:27 +0200
> From: Klemens Nanni
>
> ldomctl(8) describes much more commands than the poor usage:
>
> $ ldomctl
> usage: ldomctl start|stop|panic domain
> ldomctl status [domain]
>
> Doing as vmctl already does, this diff turns it into
>
>
ldomctl(8) describes much more commands than the poor usage:
$ ldomctl
usage: ldomctl start|stop|panic domain
ldomctl status [domain]
Doing as vmctl already does, this diff turns it into
usage: ldomctl command [argument ...]
ldomctl delete
Reading the code to understand it's usage of timers, I think we can do
better here.
Instead of masking the difference between lower and upper bound to yield
a random summand that fits, instruct the API to limit their result
accordingly. 0x01fe = 510 = 810 - 300.
arc4random_uniform(upper_bound) r
On Sat, Jun 22, 2019 at 11:29:42AM +0200, Klemens Nanni wrote:
> OK?
OK semarie@
> Index: net/if_spppsubr.c
> ===
> RCS file: /cvs/src/sys/net/if_spppsubr.c,v
> retrieving revision 1.174
> diff -u -p -r1.174 if_spppsubr.c
> --- net/
OK?
Index: net/if_spppsubr.c
===
RCS file: /cvs/src/sys/net/if_spppsubr.c,v
retrieving revision 1.174
diff -u -p -r1.174 if_spppsubr.c
--- net/if_spppsubr.c 19 Feb 2018 08:59:52 - 1.174
+++ net/if_spppsubr.c 22 Jun 2019 0
Once again I broke mrt table dumps a bit. This time by not dumping the
community data anymore. Add this back by adding the needed code in
rde_community.c and some other minor adjustments.
With this the just commited regress test passes again :)
--
:wq Claudio
Index: mrt.c
===
On 22.06., Theo Buehler wrote:
> On Fri, Jun 21, 2019 at 01:28:03PM +0200, Reyk Floeter wrote:
> > On Thu, Jun 20, 2019 at 07:58:10PM +0200, Bruno Flueckiger wrote:
> > > Hi,
> > >
> > > The patch below adds OCSP stapling to the TLS server in relayd(8). The
> > > OCSP response is read from a binary
15 matches
Mail list logo
