From owner-freebsd-current@FreeBSD.ORG Thu Jan 12 10:05:33 2012 Return-Path: Delivered-To: freebsd-current@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 13CD61065673; Thu, 12 Jan 2012 10:05:33 +0000 (UTC) (envelope-from lev@FreeBSD.org) Received: from onlyone.friendlyhosting.spb.ru (onlyone.friendlyhosting.spb.ru [IPv6:2a01:4f8:131:60a2::2]) by mx1.freebsd.org (Postfix) with ESMTP id C73978FC18; Thu, 12 Jan 2012 10:05:32 +0000 (UTC) Received: from lion.home.serebryakov.spb.ru (unknown [IPv6:2001:470:923f:1:1d3e:4d27:b4ee:e1e2]) (Authenticated sender: lev@serebryakov.spb.ru) by onlyone.friendlyhosting.spb.ru (Postfix) with ESMTPA id 749434AC2D; Thu, 12 Jan 2012 14:05:31 +0400 (MSK) Date: Thu, 12 Jan 2012 14:05:25 +0400 From: Lev Serebryakov Organization: FreeBSD Project X-Priority: 3 (Normal) Message-ID: <1699441022.20120112140525@serebryakov.spb.ru> To: Andriy Gapon In-Reply-To: <4F0EADE1.9070803@FreeBSD.org> References: <1379921442.20120112133112@serebryakov.spb.ru> <4F0EADE1.9070803@FreeBSD.org> MIME-Version: 1.0 Content-Type: text/plain; charset=windows-1251 Content-Transfer-Encoding: quoted-printable Cc: freebsd-net@FreeBSD.org, freebsd-current@FreeBSD.org, jhb@FreeBSD.org Subject: Re: SCHED_ULE / NetGraph interaction broken somwhere between r227874 and r229818 X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: lev@FreeBSD.org List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 12 Jan 2012 10:05:33 -0000 Hello, Andriy. You wrote 12 =FF=ED=E2=E0=F0=FF 2012 =E3., 13:54:41: >> Switching to 4BSD helps. 4BSD works as usual: all CPU time is >> interrupts and network thread, system is responsive under heaviest load, >> normal operations of DNS, DHCP and hostapd. > How reproducible is this result? 100% > In other words, have you definitely ruled out all other factors besides t= he > scheduler? I have two almost-identical NanoBSD images which differs in one line in = kernel config -- option about scheduler. Worlds are exactly the same, only kerne= ls were rebuilt. Alexander Motin suggests, that switching scheduler could slightly change stack consumption, which triggers switching to ng_queue instead of direct calls. Really, here is diff between "md5" of all files of one and other images: blob# diff ~lev/bsd-image.md5sums ~lev/ule-image.md5sums 74c74 < MD5 (./boot/kernel/kernel) =3D 3bb0dd757628b5065d27ee5e7fc22eb3 --- > MD5 (./boot/kernel/kernel) =3D 5ba379d2c73e1277566f4bbcb618a9f2 618c618 < MD5 (./conf/base/var/log/userlog) =3D a827af82c1f780687706b19c7d94b29e --- > MD5 (./conf/base/var/log/userlog) =3D fc289b66ae6cb23f9b24b694bf12157b 15678c15678 < MD5 (./var/log/userlog) =3D a827af82c1f780687706b19c7d94b29e --- > MD5 (./var/log/userlog) =3D fc289b66ae6cb23f9b24b694bf12157b --=20 // Black Lion AKA Lev Serebryakov