From owner-freebsd-bugs@FreeBSD.ORG Thu Dec 23 18:37:20 2010 Return-Path: Delivered-To: freebsd-bugs@FreeBSD.org Received: by hub.freebsd.org (Postfix, from userid 1233) id 582931065672; Thu, 23 Dec 2010 18:37:20 +0000 (UTC) Date: Thu, 23 Dec 2010 18:37:20 +0000 From: Alexander Best To: Bruce Cran Message-ID: <20101223183720.GA84742@freebsd.org> References: <201012121110.oBCBABjm057751@freefall.freebsd.org> <20101212111844.7510053e@core.draftnet> <20101212121923.GA61077@freebsd.org> <20101223103220.0000483f@unknown> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20101223103220.0000483f@unknown> Cc: freebsd-bugs@FreeBSD.org Subject: Re: kern/104624: Sound, mouse and keyboard badly interrupted while I/O (disk, net) X-BeenThere: freebsd-bugs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Bug reports List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 23 Dec 2010 18:37:20 -0000 On Thu Dec 23 10, Bruce Cran wrote: > On Sun, 12 Dec 2010 12:19:23 +0000 > Alexander Best wrote: > > > On Sun Dec 12 10, Bruce Cran wrote: > > > On Sun, 12 Dec 2010 11:10:11 GMT > > > Alexander Best wrote: > > > > > > > i can confirm that a similar issue exists for me running a very > > > > recent HEAD snapshot on amd64. when the disk is performing heavy > > > > i/o (make -j8 buildworld e.g.) sound starts stuttering and the > > > > system goes semi-responsive. mostly the mouse input (usb) is > > > > lagging, but sometimes the cursor freezes in it's current > > > > position for 10 seconds or so. > > > > > > There was a discussion recently where we noticed this problem > > > appeared to occur mostly on systems with Intel Core 2 Duo CPUs. > > > I've seen a login paused for over 5 seconds (between typing the > > > password and the motd being displayed) due to an "svn up" on my > > > laptop for example. > > A good way of triggering this seems to be running sysutils/fio with the > configuration: confirmed on my system. is this output right at the beginning normal? fio: this platform does not support process shared mutexes, forcing use of threads. Use the 'thread' option to get rid of this warning. f1: (g=0): rw=write, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 ... f1: (g=0): rw=write, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 fio: this platform does not support process shared mutexes, forcing use of threads. Use the 'thread' option to get rid of this warning. f2: (g=1): rw=randwrite, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 ... f2: (g=1): rw=randwrite, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 fio: this platform does not support process shared mutexes, forcing use of threads. Use the 'thread' option to get rid of this warning. f3: (g=2): rw=read, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 ... f3: (g=2): rw=read, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 fio: this platform does not support process shared mutexes, forcing use of threads. Use the 'thread' option to get rid of this warning. f4: (g=3): rw=randread, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 ... f4: (g=3): rw=randread, bs=4K-4K/4K-4K, ioengine=sync, iodepth=1 Starting 12 threads and 4 processes cheers. alex > > [global] > size=512m > bsrange=4k-4k > numjobs=4 > > [f1] > rw=write > > [f2] > stonewall > rw=randwrite > > [f3] > stonewall > rw=read > > [f4] > stonewall > rw=randread > > It seems that about half-way through even fio stops doing any IO and > doing an "ls" takes a couple of seconds. > > -- > Bruce Cran -- a13x