Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tue, 01 Mar 2005 12:21:07 -0800, "Kevin Oberman" <[EMAIL PROTECTED]> wrote: > Please DON'T top post to any FreeBSD list! Who was top-posting? I certainly wasn't! I hate that moronic practice with a vengeance. Only mildly less annoying is people writing to the list and Cc:'ing the message to multiple list members... > This is starting to sound like it might be an interrupt routing issue > and interrupts from the disk are sharing an IRQ with something > else. "Something else" is generating interrupts that are never getting > delivered, but the disk interrupts are waking the appropriate driver to > allow things to proceed. Whatever it was it would appear to be solved now anyway. - -- G. Stewart - [EMAIL PROTECTED] Don't worry about people stealing your ideas. If your ideas are any good, you'll have to ram them down people's throats. -- Howard Aiken -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (FreeBSD) iD8DBQFCJOxmK5oiGLo9AcYRAv1kAJ0Ry39oiEczi8aOrkuBtK3yobt/LwCg21Rf bvbB4VIDXaC/3tfzJlNRaiM= =GUoN -END PGP SIGNATURE- ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
<-Original Message-> From: Kevin Oberman Sent: 3/2/2005 10:45:32 AM To: [EMAIL PROTECTED] Cc: freebsd-stable@freebsd.org Subject: Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD > Date: Tue, 1 Mar 2005 16:36:11 +0100 > From: Godwin Stewart > Sender: [EMAIL PROTECTED] > > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Tue, 1 Mar 2005 02:32:02 -0500 (EST), Jeff Roberson > wrote: > > > Is the process that does the FFT in kernel, niced, or rtprio'd? > > last pid: 93131; load averages: 0.96, 0.49, 0.24 up 0+05:18:20 15:29:47 > 48 processes: 2 running, 46 sleeping > CPU states: 99.6% user, 0.0% nice, 0.4% system, 0.0% interrupt, 0.0% id> le > Mem: 174M Active, 94M Inact, 86M Wired, 14M Cache, 48M Buf, 2564K Free > Swap: 743M Total, 180K Used, 743M Free > > PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND > 93124 godwin 121 0 33288K 23836K RUN 2:22 13.72% 13.72% audacity > > WCPU and CPU climb to about 20%, drop to 0 during disk I/O, then start > climbing again. > > Note that this is with SCHED_4BSD. I'd need to recompile a kernel to get > similar information for ULE/PREEMPTION. > > fx: goes off and compiles a new kernel and comes back when it's done... > > Looks like I spoke too soon. The system is now perfectly stable and usable > with ULE. > > The problems I was having with ULE were on 5.3-STABLE. I'm now on 5.4-PRE. > Were there significant changes to the kernel in between? > Stable is a dynamic thing. Stable as of what date? Lots of fixes to ULE, APIC, ACPI and other things have made it to the kernel at some point in the life of 5.3-Stable. Several have made it rather recently. Comparing a dmesg from when it was not working with one now might be instructive. (They can be found in /var/log/messages[.n.bz2].) Take a look at the details of the device probes before disks are mounted in particular, for changes. > > Can you give me any information on the means by which you transfer data > > from a cassette to your pc? > > Straightforward audio connection from the amp's line out to the sound > card's line in. The FFT filtering isn't performed on the fly BTW. I use > audacity to grab the audio and then work on it after it's "in the box". Really sounds more and more like interrupts were not getting properly delivered. Normally the disk IRQs are not shared, but it really looks like something was broken here for your BIOS. (And it appears to have been fixed!) -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]" . ___Join our internet solutions at http://www.palsign.com";>http://www.palsign.com ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
> Date: Tue, 1 Mar 2005 16:36:11 +0100 > From: Godwin Stewart <[EMAIL PROTECTED]> > Sender: [EMAIL PROTECTED] > > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Tue, 1 Mar 2005 02:32:02 -0500 (EST), Jeff Roberson > <[EMAIL PROTECTED]> wrote: > > > Is the process that does the FFT in kernel, niced, or rtprio'd? > > last pid: 93131; load averages: 0.96, 0.49, 0.24 up 0+05:18:20 15:29:47 > 48 processes: 2 running, 46 sleeping > CPU states: 99.6% user, 0.0% nice, 0.4% system, 0.0% interrupt, 0.0% id> > le > Mem: 174M Active, 94M Inact, 86M Wired, 14M Cache, 48M Buf, 2564K Free > Swap: 743M Total, 180K Used, 743M Free > > PID USERNAME PRI NICE SIZERES STATETIME WCPUCPU COMMAND > 93124 godwin 1210 33288K 23836K RUN 2:22 13.72% 13.72% audacity > > WCPU and CPU climb to about 20%, drop to 0 during disk I/O, then start > climbing again. > > Note that this is with SCHED_4BSD. I'd need to recompile a kernel to get > similar information for ULE/PREEMPTION. > > fx: goes off and compiles a new kernel and comes back when it's done... > > Looks like I spoke too soon. The system is now perfectly stable and usable > with ULE. > > The problems I was having with ULE were on 5.3-STABLE. I'm now on 5.4-PRE. > Were there significant changes to the kernel in between? > Stable is a dynamic thing. Stable as of what date? Lots of fixes to ULE, APIC, ACPI and other things have made it to the kernel at some point in the life of 5.3-Stable. Several have made it rather recently. Comparing a dmesg from when it was not working with one now might be instructive. (They can be found in /var/log/messages[.n.bz2].) Take a look at the details of the device probes before disks are mounted in particular, for changes. > > Can you give me any information on the means by which you transfer data > > from a cassette to your pc? > > Straightforward audio connection from the amp's line out to the sound > card's line in. The FFT filtering isn't performed on the fly BTW. I use > audacity to grab the audio and then work on it after it's "in the box". Really sounds more and more like interrupts were not getting properly delivered. Normally the disk IRQs are not shared, but it really looks like something was broken here for your BIOS. (And it appears to have been fixed!) -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
> Date: Tue, 1 Mar 2005 15:24:58 +0100 > From: Godwin Stewart <[EMAIL PROTECTED]> > Sender: [EMAIL PROTECTED] > > On Mon, 28 Feb 2005 16:21:39 -0800, "Kevin Oberman" <[EMAIL PROTECTED]> wrote: > > > I'd say something is very wrong on your systems and I'd ALMOST bet it's > > ata related. Maybe ATA-MkIII would help things out. > > Possibly, altho' I doubt it given that the only short periods during which > the machine is responsive are *during* disk I/O. Please DON'T top post to any FreeBSD list! This is starting to sound like it might be an interrupt routing issue and interrupts from the disk are sharing an IRQ with something else. "Something else" is generating interrupts that are never getting delivered, but the disk interrupts are waking the appropriate driver to allow things to proceed. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tue, 1 Mar 2005 02:32:02 -0500 (EST), Jeff Roberson <[EMAIL PROTECTED]> wrote: > Is the process that does the FFT in kernel, niced, or rtprio'd? last pid: 93131; load averages: 0.96, 0.49, 0.24 up 0+05:18:20 15:29:47 48 processes: 2 running, 46 sleeping CPU states: 99.6% user, 0.0% nice, 0.4% system, 0.0% interrupt, 0.0% idle Mem: 174M Active, 94M Inact, 86M Wired, 14M Cache, 48M Buf, 2564K Free Swap: 743M Total, 180K Used, 743M Free PID USERNAME PRI NICE SIZERES STATETIME WCPUCPU COMMAND 93124 godwin 1210 33288K 23836K RUN 2:22 13.72% 13.72% audacity WCPU and CPU climb to about 20%, drop to 0 during disk I/O, then start climbing again. Note that this is with SCHED_4BSD. I'd need to recompile a kernel to get similar information for ULE/PREEMPTION. fx: goes off and compiles a new kernel and comes back when it's done... Looks like I spoke too soon. The system is now perfectly stable and usable with ULE. The problems I was having with ULE were on 5.3-STABLE. I'm now on 5.4-PRE. Were there significant changes to the kernel in between? > Can you give me any information on the means by which you transfer data > from a cassette to your pc? Straightforward audio connection from the amp's line out to the sound card's line in. The FFT filtering isn't performed on the fly BTW. I use audacity to grab the audio and then work on it after it's "in the box". - -- G. Stewart - [EMAIL PROTECTED] All those who believe in telekinesis, raise my hand. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (FreeBSD) iD8DBQFCJIvrK5oiGLo9AcYRAtGOAKDOj+vzbIR8r8/Ei9Yzo/9abPKmDACdHcIY pcHHJuEsBDYQzJhRkleWzDQ= =OXpK -END PGP SIGNATURE- ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Mon, 28 Feb 2005 16:21:39 -0800, "Kevin Oberman" <[EMAIL PROTECTED]> wrote: > I'd say something is very wrong on your systems and I'd ALMOST bet it's > ata related. Maybe ATA-MkIII would help things out. Possibly, altho' I doubt it given that the only short periods during which the machine is responsive are *during* disk I/O. - -- G. Stewart - [EMAIL PROTECTED] There are three types of people in this world: - Those who can count - Those who can't -- Walter Dnes in NANAE, 2003-JUL-26. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (FreeBSD) iD8DBQFCJHs6K5oiGLo9AcYRAumZAJ9ksutDEL/7SgsdSZ5nwJ80rbyXfQCggA6r dSgi7FihFYRqyUKvbpKlxQc= =rUEF -END PGP SIGNATURE- ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
On Mon, 28 Feb 2005, Kevin Oberman wrote: > > Date: Sun, 27 Feb 2005 21:36:23 +0100 > > From: Godwin Stewart <[EMAIL PROTECTED]> > > Sender: [EMAIL PROTECTED] > > > > -BEGIN PGP SIGNED MESSAGE- > > Hash: SHA1 > > > > On Sat, 26 Feb 2005 21:49:00 +0100, Michael Nottebrock > > <[EMAIL PROTECTED]> wrote: > > > > > I've been using that for a long time now, since Xorg 6.8.1 breaks vt- > > > switching for me. > > > > Well, I decided to bite the bullet and upgraded to Xorg 6.8.1 anyway. It > > didn't break vt-switching for me, thankfully. Other than the core keyboard > > driver being "kbd" instead of "Keyboard" now, which threw me off for a > > couple of minutes, all went well. It seems to be stable enough. Cross > > fingers, touch wood etc. > > > > I also took advantage of the latest cvsup to 5.4-PRE and ensuing recompile > > to revert to SCHED_4BSD from SCHED_ULE and PREEMPTION in the kernel. The > > difference is staggering. > > > > One of the things I've been doing is to record some of my old cassettes > > (you know, those old plastic things with 2 holes and a tape inside :) onto > > CD. Applying a FFT filter to 50 minutes of audio takes between 10 and 15 > > minutes on this machine (P-III/550, 384MB) depending on the complexity of > > the filter. During this time, with SCHED_ULE and PREEMTION, the machine is > > unusable. It freezes hard for periods of 10-12 seconds and then when it > > unfreezes (while doing disk i/o apparently) the keys you typed turn up in > > the wrong order. Is the process that does the FFT in kernel, niced, or rtprio'd? Can you give me any information on the means by which you transfer data from a cassette to your pc? > > > > However, now that I've reverted to SCHED_4BSD, the machine remains > > perfectly snappy while performing the FFT filter, which doesn't happen > > perceptibly slower. > > > > It could be that I misread things entirely (wouldn't be the first time), > > but wasn't SCHED_ULE's purpose to *improve* the responsiveness of the > > machine when under load? The results I'm getting here are, errmm... > > slightly different... Old hardware maybe? > > This is VERY odd. What you saw with ULE is what I (and most people) saw > with 4BSD. I got very tired of the short pauses I was getting unde 4BSD > on my 5-Stable laptop and was very pleased to get back to ULE a few weeks > ago when I moved it to 6-Current. > > I'd say something is very wrong on your systems and I'd ALMOST bet it's > ata related. Maybe ATA-MkIII would help things out. > -- > R. Kevin Oberman, Network Engineer > Energy Sciences Network (ESnet) > Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) > E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 > ___ > freebsd-stable@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > To unsubscribe, send any mail to "[EMAIL PROTECTED]" > ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
> Date: Sun, 27 Feb 2005 21:36:23 +0100 > From: Godwin Stewart <[EMAIL PROTECTED]> > Sender: [EMAIL PROTECTED] > > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Sat, 26 Feb 2005 21:49:00 +0100, Michael Nottebrock > <[EMAIL PROTECTED]> wrote: > > > I've been using that for a long time now, since Xorg 6.8.1 breaks vt- > > switching for me. > > Well, I decided to bite the bullet and upgraded to Xorg 6.8.1 anyway. It > didn't break vt-switching for me, thankfully. Other than the core keyboard > driver being "kbd" instead of "Keyboard" now, which threw me off for a > couple of minutes, all went well. It seems to be stable enough. Cross > fingers, touch wood etc. > > I also took advantage of the latest cvsup to 5.4-PRE and ensuing recompile > to revert to SCHED_4BSD from SCHED_ULE and PREEMPTION in the kernel. The > difference is staggering. > > One of the things I've been doing is to record some of my old cassettes > (you know, those old plastic things with 2 holes and a tape inside :) onto > CD. Applying a FFT filter to 50 minutes of audio takes between 10 and 15 > minutes on this machine (P-III/550, 384MB) depending on the complexity of > the filter. During this time, with SCHED_ULE and PREEMTION, the machine is > unusable. It freezes hard for periods of 10-12 seconds and then when it > unfreezes (while doing disk i/o apparently) the keys you typed turn up in > the wrong order. > > However, now that I've reverted to SCHED_4BSD, the machine remains > perfectly snappy while performing the FFT filter, which doesn't happen > perceptibly slower. > > It could be that I misread things entirely (wouldn't be the first time), > but wasn't SCHED_ULE's purpose to *improve* the responsiveness of the > machine when under load? The results I'm getting here are, errmm... > slightly different... Old hardware maybe? This is VERY odd. What you saw with ULE is what I (and most people) saw with 4BSD. I got very tired of the short pauses I was getting unde 4BSD on my 5-Stable laptop and was very pleased to get back to ULE a few weeks ago when I moved it to 6-Current. I'd say something is very wrong on your systems and I'd ALMOST bet it's ata related. Maybe ATA-MkIII would help things out. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: [EMAIL PROTECTED] Phone: +1 510 486-8634 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Xorg 6.8.1 and SCHED_ULE vs. SCHED_4BSD
Godwin Stewart napisaÅ(a): On Sat, 26 Feb 2005 21:49:00 +0100, Michael Nottebrock <[EMAIL PROTECTED]> wrote: I've been using that for a long time now, since Xorg 6.8.1 breaks vt- switching for me. One of the things I've been doing is to record some of my old cassettes (you know, those old plastic things with 2 holes and a tape inside :) onto CD. Applying a FFT filter to 50 minutes of audio takes between 10 and 15 minutes on this machine (P-III/550, 384MB) depending on the complexity of the filter. During this time, with SCHED_ULE and PREEMTION, the machine is unusable. It freezes hard for periods of 10-12 seconds and then when it unfreezes (while doing disk i/o apparently) the keys you typed turn up in the wrong order. I currently run 5.4-PRE with ULE and PREEMPTION, on a similar machine (pIII-733 192RAM), also 6.8.1, and i do have to say that ULE has improved responsiveness /alongside kern.hz=800/ incredibly, with none-whatsoever speed degradation (actually my compilations seem to run faster, although that is only a mere hunch not yet backed up by any benchmarking). The only time i might encounter problems, is with the lack of ram, and a lot of disk swap usage, or during untarring of big distfiles (yet it is still giving me better response than with 4BSD, which was utterly terrible :) I honestly would have to say, great job on the ULE, if mere fixing of the possible disk i/o lock ups were to be commited its much better than 4BSD. However, now that I've reverted to SCHED_4BSD, the machine remains perfectly snappy while performing the FFT filter, which doesn't happen perceptibly slower. Perhaphs you can try with different kern.hz settings? -- Mateusz JÄdrasik < [EMAIL PROTECTED] > ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
