Re: svn commit: r309143 - head/sys/libkern

2016-12-10 Thread Mateusz Guzik
On Wed, Dec 07, 2016 at 11:41:21AM -0800, Gleb Smirnoff wrote: > On Fri, Nov 25, 2016 at 05:00:10PM +0100, Mateusz Guzik wrote: > M> On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote: > M> > Author: fabient > M> > Date: Fri Nov 25 13:49:33 2016 > M> > New Revision: 309143 > M> > URL: ht

Re: svn commit: r309143 - head/sys/libkern

2016-12-07 Thread Gleb Smirnoff
On Fri, Nov 25, 2016 at 05:00:10PM +0100, Mateusz Guzik wrote: M> On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote: M> > Author: fabient M> > Date: Fri Nov 25 13:49:33 2016 M> > New Revision: 309143 M> > URL: https://svnweb.freebsd.org/changeset/base/309143 M> > M> > Log: M> > In a

Re: svn commit: r309143 - head/sys/libkern

2016-11-25 Thread Oleg Bulyzhin
On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote: > + if (reseed || atomic_cmpset_int(&arc4rand_iniseed_state, > + ARC4_ENTR_HAVE, ARC4_ENTR_SEED)) { > + ARC4_FOREACH(arc4) > + arc4_randomstir(arc4); > + } How about eliminat

Re: svn commit: r309143 - head/sys/libkern

2016-11-25 Thread Fabien Thomas
> Le 25 nov. 2016 à 17:00, Mateusz Guzik a écrit : > > On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote: >> Author: fabient >> Date: Fri Nov 25 13:49:33 2016 >> New Revision: 309143 >> URL: https://svnweb.freebsd.org/changeset/base/309143 >> >> Log: >> In a dual processor system (

Re: svn commit: r309143 - head/sys/libkern

2016-11-25 Thread Mateusz Guzik
On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote: > Author: fabient > Date: Fri Nov 25 13:49:33 2016 > New Revision: 309143 > URL: https://svnweb.freebsd.org/changeset/base/309143 > > Log: > In a dual processor system (2*6 cores) during IPSec throughput tests, > we see a lot of co

svn commit: r309143 - head/sys/libkern

2016-11-25 Thread Fabien Thomas
Author: fabient Date: Fri Nov 25 13:49:33 2016 New Revision: 309143 URL: https://svnweb.freebsd.org/changeset/base/309143 Log: In a dual processor system (2*6 cores) during IPSec throughput tests, we see a lot of contention on the arc4 lock, used to generate the IV of the ESP output packets.