Date: Mon, 14 Mar 2011 14:09:53 -0400 From: Arnaud Lacombe <lacombar@gmail.com> To: =?ISO-8859-1?Q?=D6zkan_KIRIK?= <ozkan.kirik@gmail.com> Cc: freebsd-net@freebsd.org, freebsd-stable@freebsd.org, Jack Vogel <jfvogel@gmail.com> Subject: Re: em0 with latest driver hangs again and again (without "Watchdogtimeout" message!) Message-ID: <AANLkTi=OthKAyHmQs0LOvaWz1t3by%2BteE2GH6-qyK2aA@mail.gmail.com> In-Reply-To: <AANLkTinpcKYx0CyonisdOYWJCoWn_nh4CRMZP4XBGQvH@mail.gmail.com> References: <1975926365.20110223121637@serebryakov.spb.ru> <4D64EC8C.2080007@sentex.net> <1004451940.20110223143607@serebryakov.spb.ru> <AANLkTi=1hQ-tSYtt=KrRi60yi7P9GvVDTG%2Bn8Pz61M_H@mail.gmail.com> <AANLkTinmmqSphSJpMahbvoQ-6-0UnV-O_ESV_R98oyqO@mail.gmail.com> <AANLkTik19U424kaV%2BvcF98s_MzWMvfek90WmBm%2BemO98@mail.gmail.com> <4D6D00C1.1040805@sentex.net> <1416421652.20110301225215@serebryakov.spb.ru> <AANLkTinkEJaqRZ2YxyZqzfZ8a3iZt7H2YscXzU_QiSiS@mail.gmail.com> <1627628072.20110303111046@serebryakov.spb.ru> <AANLkTikKs-a4d79A6LnnXJG_REo9iVNYJRAwiJXRStH1@mail.gmail.com> <1894628540.20110304012554@serebryakov.spb.ru> <31A99DAA7E5F4095B22B9DD57DD0E19E@multiplay.co.uk> <494278763.20110305130320@serebryakov.spb.ru> <AANLkTikBxTmUZaMXMk2t1RTZ5VhU7zZG=FUv2hJUEzmR@mail.gmail.com> <AANLkTinkk8btZLoyNYRb4w68484=zjKqn%2BfAemGjj%2BXL@mail.gmail.com> <AANLkTin6%2BLqe5r4ETscpgMZpZNo92CDAhFgKicTcC7rO@mail.gmail.com> <AANLkTim-HV-7=r-V7fNFuT5HYLGz=j9PrHv1ejJrHQyV@mail.gmail.com> <AANLkTinpcKYx0CyonisdOYWJCoWn_nh4CRMZP4XBGQvH@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi, On Mon, Mar 14, 2011 at 10:17 AM, =D6zkan KIRIK <ozkan.kirik@gmail.com> wro= te: > Hi Arnaud, > > # sysctl kern.ipc.nmbclusters > kern.ipc.nmbclusters: 262144 > Well, then this is not surprising em(4) behave well, knowing that it is unable to handle resource shortage, give it plenty of it and it will be happy. 200MBps is about 18000 1500bytes packet by seconds. I will not trust the quality of any code which need 10x more resources available than needed to do the job. Even 1Gbps should be attainable with only 128k clusters. - Arnaud > On Mon, Mar 14, 2011 at 4:02 PM, Arnaud Lacombe <lacombar@gmail.com> wrot= e: >> Hi, >> >> On Fri, Mar 11, 2011 at 9:57 AM, =D6zkan KIRIK <ozkan.kirik@gmail.com> w= rote: >>> Hi again, >>> >>> I wanna share state of test machine. em-7.2.2 driver runs as kld. No ha= ngs. >>> >> How high is `nmbclusters' ? >> >> Thanks, >> =A0- Arnaud >> >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTi=OthKAyHmQs0LOvaWz1t3by%2BteE2GH6-qyK2aA>