Date: Sat, 19 Mar 2005 09:37:30 -0600 From: Kevin Kinsey <kdk@daleco.biz> To: Brian Curnow <brianccurnow@yahoo.com> Cc: questions@freebsd.org Subject: Re: calcru: negative runtime Message-ID: <423C473A.2070107@daleco.biz> In-Reply-To: <20050319040537.57051.qmail@web53505.mail.yahoo.com> References: <20050319040537.57051.qmail@web53505.mail.yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Brian Curnow wrote: >Thanks for the reply. > >I saw a few other posts that talked about that >setting. My list is bit different, I have TSC, i8254 >and ACPI-safe. When I boot it picks ACPI-safe. > >I've tried the other options but they didn't seem to >make any difference. > > > How did you try them? >I did notice the the kern.timecounter.tick seems to be >wrong. When booting up the tick is reported as 10 but >the seting is 1. Unfortunately kern.timecounter.tick >is read only. > >Any other ideas? > > > In this message: http://lists.freebsd.org/pipermail/freebsd-stable/2004-November/009417.html someone suggests rebuilding the kernel. That was on a box with an AMD K6 type CPU. IIRC, I fixed these issues by running timecounter as i8254. I'm sure that could be set in /etc/sysctl.conf; I don't know if it could be in /boot/loader.conf or not; however, the earlier it was done, sounds like it might be better. I'm sure you also noted that some folks did a "work around" by booting with ACPI disabled. That could definitely be set in loader.conf, but it doesn't sound ideal. However, depending on your motherboard, it might be about the only way to do this, unless you can find and patch the underlying code issue, (if indeed there is one ... IIRC, it just that the chips are a little "buggy", the developers claim), and besides, we newbs generally aren't that proficient in C ;-) As you might guess, we're way over my head now. I hope you can get it going to suit you .... Good luck, Kevin Kinsey >--- Kevin Kinsey <kdk@daleco.biz> wrote: > > >>Brian Curnow wrote: >> >> >> >>> <>I am new to BSD and just downloaded BSD 5.3-RELEASE. >> >>> <>I saw a few other posts on this issue but I didn't see any solutions. >> >>> <> >>> I am getting excessive calcru: negative runtime >>> messages in the syslog (and the console). >>> >>> According to the FAQ the solution is to use sysctl and >> >>> <>set kern.timecounter.method=1, however this param does >> >>>not exist in 5.3. >>> >>>Is there a solution to this issue? >>> >>>Thanks! >>> >>>Brian Curnow >>> >>> >>> >>> >>$sysctl -a | grep timeco >>kern.timecounter.stepwarnings: 0 >>kern.timecounter.nbinuptime: 4005177648 >>kern.timecounter.nnanouptime: 0 >>kern.timecounter.nmicrouptime: 1108115 >>kern.timecounter.nbintime: 1813993971 >>kern.timecounter.nnanotime: 207315298 >>kern.timecounter.nmicrotime: 1606545859 >>kern.timecounter.ngetbinuptime: 0 >>kern.timecounter.ngetnanouptime: 115004537 >>kern.timecounter.ngetmicrouptime: 3989727579 >>kern.timecounter.ngetbintime: 0 >>kern.timecounter.ngetnanotime: 526 >>kern.timecounter.ngetmicrotime: 201307619 >>kern.timecounter.nsetclock: 7 >>kern.timecounter.hardware: ACPI-fast >>kern.timecounter.choice: TSC(800) ACPI-fast(1000) >>i8254(0) dummy(-1000000) >>kern.timecounter.tick: 1 >> >>I'm not the expert here, but I think I've >>experience. >> >>Examine the options your system has for "kern.timecounter.choice" >>(you can see above how I did it) and experiment until you find (if you >>find) a setting that seems to fix the problem. >> >>Certain chipsets seem to have issues, but I couldn't say which ones. >>I have, however, fixed a couple of boxes by picking a different value >>for this sysctl. >> >>It's likely that the FAQ hasn't yet been completely transformed >>into a "post-5.3-is-now-stable" format, since people were warned >>that 5.0-5.2 weren't "production releases". If you find that this >>helps you, write back and I'll try sometime Real Soon Now(tm) >>to send a patch to the FDP folks to address that issue.... >> >>HTH, >> >>Kevin Kinsey >> >> >>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?423C473A.2070107>