Hi,
i did some performace tests in current with and without your diff.
There is no difference in performance.
I will try to do performance tests with current on a regular base now.
2016-10-05 10:15 GMT+02:00, Martin Pieuchot :
> On 10/04/16 16:44, Martin Pieuchot wrote:
>> On 10/03/16 16:43, Ma
Hi Martin -
As soon as machine boots and first packet hits the VPN, it hits the
assertion:
rw_status(&netlock) == RW_WRITE
(hand-typed)
TID PID UID PRFLAGS PFLAGS CPU COMMAND
*78967 78967 0 0x14000 0x200 1 crypto
__assert at0x25
ip_output() at 0x843
ipsp_proces
> > panic: rw_enter: netlock locking against myself
>
> Your tree is not up-to-date, you're missing sys/net/route.c r1.332.
My bad.
With updated sources I have tested this for few minutes
and I don't encountered any kernel panic.
CVS sync, wget to download base system from mirror,
browsing web wi
On 10/05/16 20:52, Lampshade wrote:
panic: rw_enter: netlock locking against myself
Your tree is not up-to-date, you're missing sys/net/route.c r1.332.
rt_timer_timer() is now called from the softclock thread.
panic: rw_enter: netlock locking against myself
TIDPID UIDPRFLAGS
12705 12705 1041 0x23
*62630 62630 1000 0x32
PFLAGS CPU COMMAND
0 1chrome
0 0 Xorg
panic()
rw_enter()
rw_enter_write()
rw_enter_timer()
timeout_run()
softlock()
softintr_dispatch(
On 10/04/16 16:44, Martin Pieuchot wrote:
On 10/03/16 16:43, Martin Pieuchot wrote:
Diff below introduces a single write lock that will be used to serialize
access to ip_output().
This lock will be then split in multiple readers and writers to allow
multiple forwarding paths to run in parallel
On Tue, Oct 04, 2016 at 04:44:29PM +0200, Martin Pieuchot wrote:
> On 10/03/16 16:43, Martin Pieuchot wrote:
> > Diff below introduces a single write lock that will be used to serialize
> > access to ip_output().
> >
> > This lock will be then split in multiple readers and writers to allow
> > mul
On 10/03/16 16:43, Martin Pieuchot wrote:
Diff below introduces a single write lock that will be used to serialize
access to ip_output().
This lock will be then split in multiple readers and writers to allow
multiple forwarding paths to run in parallel of each others but still
serialized with th
Diff below introduces a single write lock that will be used to serialize
access to ip_output().
This lock will be then split in multiple readers and writers to allow
multiple forwarding paths to run in parallel of each others but still
serialized with the socket layer.
I'm currently looking for p