Re: ospf6d: refactor kernel route message handling

2019-12-11 Thread Remi Locherer
On Wed, Dec 11, 2019 at 04:38:38PM +0100, Denis Fondras wrote: > On Tue, Dec 10, 2019 at 09:51:12PM +0100, Remi Locherer wrote: > > Unfortunately redistribute does not work anymore. > > > > Indeed, simple tests are too simple... > > Here is an updated diff. ok remi@ > > Index: kroute.c >

Re: ripd: memory leak and double free/use-after-free

2019-12-11 Thread Remi Locherer
On Wed, Dec 11, 2019 at 10:11:58AM +0100, Sebastian Benoit wrote: > Remi Locherer(remi.loche...@relo.ch) on 2019.12.10 22:39:32 +0100: > > On Tue, Dec 10, 2019 at 07:05:27PM +0100, Hiltjo Posthuma wrote: > > > Hi, > > > > > > While looking at the code of ripd: > > > > > > I think there are

Re: login_passwd: reject challenge service

2019-12-11 Thread Thomas L.
On Thu, 5 Dec 2019 13:35:40 + "Lindner, Thomas 1. (Nokia - DE/Nuremberg)" wrote: > The (untested) patch below makes login_passwd behave as described in > the manpage. I've now been able to test the patch and login/su/doas/ssh still work as expected. All the other login_* styles in base are

Re: patch: make sdiff tty-aware

2019-12-11 Thread Theo de Raadt
I'm not sure I understand the goal of the signal handler. sdiff is moving forward through the file, only. If you are in a pager, you want to increase the width for the later output not yet visible? the normal way one does that in programs which don't backtrack and re-output, is by restarting the

Re: ospf6d: convert sendto() to sendmsg()

2019-12-11 Thread Claudio Jeker
On Wed, Dec 11, 2019 at 04:52:53PM +0100, Denis Fondras wrote: > Use sendmsg() instead of sendto() like ospfd(8) does. > > Index: database.c > === > RCS file: /cvs/src/usr.sbin/ospf6d/database.c,v > retrieving revision 1.16 > diff -u

re-enable iwm(4) firmware Tx retries

2019-12-11 Thread Stefan Sperling
Back in February we disabled iwm firmware Tx retries at lower rates. This improved MiRA's Tx rate selection because bad rates do actually look bad, rather than being compensated for by firmware retries. The result was increased Tx throughput. The original discussion is here:

Re: ospf6d: convert sendto() to sendmsg()

2019-12-11 Thread Denis Fondras
On Wed, Dec 11, 2019 at 07:06:26PM +0100, Claudio Jeker wrote: > While it makes sense to pass the ibuf like in ospfd I see no reason to > switch to sendmsg(). In ospfd this is done to prepend the IP header but > that seems not needed here. Is there another reason to switch to > sendmsg()? >

Re: ospf6d: convert sendto() to sendmsg()

2019-12-11 Thread Theo de Raadt
Claudio Jeker wrote: > On Wed, Dec 11, 2019 at 04:52:53PM +0100, Denis Fondras wrote: > > Use sendmsg() instead of sendto() like ospfd(8) does. > > > > Index: database.c > > === > > RCS file: /cvs/src/usr.sbin/ospf6d/database.c,v >

Re: ospf6d: convert sendto() to sendmsg()

2019-12-11 Thread Claudio Jeker
On Wed, Dec 11, 2019 at 04:52:53PM +0100, Denis Fondras wrote: > Use sendmsg() instead of sendto() like ospfd(8) does. > > Index: database.c > === > RCS file: /cvs/src/usr.sbin/ospf6d/database.c,v > retrieving revision 1.16 > diff -u

patch: make sdiff tty-aware

2019-12-11 Thread Marc Espie
This patch allows sdiff to auto-detect tty width, by passing -w auto More importantly, sdiff will adjust its variables on the fly for subsequent lines (I don't know if redrawing the current lines when the tty changes is advisable). It's pretty straightforward, tested thru sysmerge. Doesn't seem

ospf6d: convert sendto() to sendmsg()

2019-12-11 Thread Denis Fondras
Use sendmsg() instead of sendto() like ospfd(8) does. Index: database.c === RCS file: /cvs/src/usr.sbin/ospf6d/database.c,v retrieving revision 1.16 diff -u -p -r1.16 database.c --- database.c 10 May 2019 13:50:34 - 1.16

Re: ospf6d: refactor kernel route message handling

2019-12-11 Thread Denis Fondras
On Tue, Dec 10, 2019 at 09:51:12PM +0100, Remi Locherer wrote: > Unfortunately redistribute does not work anymore. > Indeed, simple tests are too simple... Here is an updated diff. Index: kroute.c === RCS file:

Re: ripd: memory leak and double free/use-after-free

2019-12-11 Thread Sebastian Benoit
Remi Locherer(remi.loche...@relo.ch) on 2019.12.10 22:39:32 +0100: > On Tue, Dec 10, 2019 at 07:05:27PM +0100, Hiltjo Posthuma wrote: > > Hi, > > > > While looking at the code of ripd: > > > > I think there are (also) 2 small memleaks in a debug/error path > > (IMSG_REQUEST_ADD and