Re: irq19 interrupt storm?
On Wednesday 17 September 2008 05:17:27 pm John Baldwin wrote: > On Wednesday 17 September 2008 11:00:24 am Scott Gasch wrote: > > You're right: atapci1, atapci2, fwohci0 and uhci4 are all sharing the same > > irq (19) while irqs 20, 21, 22 at least seem completely unused. Here's a > > dumb question: how do I fix it? I tried setting "plug and play OS" in the > > BIOS and then using device.hints to push different devices to different > > irqs. But every time I tried a new hint it seemed to be ignored. I was > > trying stuff like: > > > > set hint.atapci.1.irq="20" > > set hint ata.4.irq="20" (ata4 is a channel on atapci1) > > set hint fwhco.0.irq="20" > > etc... > > > > > > I also tried to move the dc driver to a new irq as a test. This was also > > seemingly ignored. > > > > I then tried turning "plug and play OS" off in the BIOS but I don't see > > anywhere to set the IRQs of the onboard SATA controllers via the menus. I'm > > looking for a BIOS upgrade now... any other advice? > > Unfortunately you can't really move PCI IRQs around. You can read about more > of the gritty details here: > > http://people.freebsd.org/~jhb/papers/bsdcan/2007/ > > You might be able to shuffle some IRQs around using 'hw.pciX.Y.INTA.irq' > tunables. Gah, wrong tunables. These devices are on PCI link devices, so you'd need to do something like 'hw.pci.LNKA.irq' (where LNKA is the name of the link device in the ACPI namespace). Verbose boot messages (boot -v) can tell you which link device you PCI devices are using. -- John Baldwin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: irq19 interrupt storm?
On Wednesday 17 September 2008 11:00:24 am Scott Gasch wrote: > You're right: atapci1, atapci2, fwohci0 and uhci4 are all sharing the same > irq (19) while irqs 20, 21, 22 at least seem completely unused. Here's a > dumb question: how do I fix it? I tried setting "plug and play OS" in the > BIOS and then using device.hints to push different devices to different > irqs. But every time I tried a new hint it seemed to be ignored. I was > trying stuff like: > > set hint.atapci.1.irq="20" > set hint ata.4.irq="20" (ata4 is a channel on atapci1) > set hint fwhco.0.irq="20" > etc... > > > I also tried to move the dc driver to a new irq as a test. This was also > seemingly ignored. > > I then tried turning "plug and play OS" off in the BIOS but I don't see > anywhere to set the IRQs of the onboard SATA controllers via the menus. I'm > looking for a BIOS upgrade now... any other advice? Unfortunately you can't really move PCI IRQs around. You can read about more of the gritty details here: http://people.freebsd.org/~jhb/papers/bsdcan/2007/ You might be able to shuffle some IRQs around using 'hw.pciX.Y.INTA.irq' tunables. Probably you have a device driver whose interrupt handler isn't handling some condition. I would suspect ata as it's interrupt handler is rather simplistic with no chipset-specific hooks, and I've seen several reports of interrupt storms with ata(4) recently. > Thx, > Scott > > > On Tue, Sep 16, 2008 at 7:12 AM, Gary Jennejohn > <[EMAIL PROTECTED]>wrote: > > > On Mon, 15 Sep 2008 22:57:38 -0700 > > "Scott Gasch" <[EMAIL PROTECTED]> wrote: > > > > > Hi, > > > > > > I'm running freebsd 7.0-RELEASE-p4 on a 4-core amd64 box. Nearly 100% of > > > 1 cpu is constantly being used handling irq19: uhci4 interrupts. This > > > seems to happen both with and without any USB devices plugged in: > > > > > > vmstat -i > > > interrupt total rate > > > irq1: atkbd0 5 0 > > > irq6: fdc0 1 0 > > > irq17: mskc0 dc0 1180547 18 > > > irq18: skc0 uhci2* 163250699 2512 > > > irq19: uhci4++ 3187989508 49072 > > > > I think the ++ here indicates that two or more devices are sharing this > > interrupt. Try doing "grep irq.*19 /var/run/dmesg.boot" to see which > > ones. One of these devices could be the culprit. > > > > --- > > Gary Jennejohn > > > ___ > [EMAIL PROTECTED] mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-amd64 > To unsubscribe, send any mail to "[EMAIL PROTECTED]" > -- John Baldwin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: irq19 interrupt storm?
You're right: atapci1, atapci2, fwohci0 and uhci4 are all sharing the same irq (19) while irqs 20, 21, 22 at least seem completely unused. Here's a dumb question: how do I fix it? I tried setting "plug and play OS" in the BIOS and then using device.hints to push different devices to different irqs. But every time I tried a new hint it seemed to be ignored. I was trying stuff like: set hint.atapci.1.irq="20" set hint ata.4.irq="20" (ata4 is a channel on atapci1) set hint fwhco.0.irq="20" etc... I also tried to move the dc driver to a new irq as a test. This was also seemingly ignored. I then tried turning "plug and play OS" off in the BIOS but I don't see anywhere to set the IRQs of the onboard SATA controllers via the menus. I'm looking for a BIOS upgrade now... any other advice? Thx, Scott On Tue, Sep 16, 2008 at 7:12 AM, Gary Jennejohn <[EMAIL PROTECTED]>wrote: > On Mon, 15 Sep 2008 22:57:38 -0700 > "Scott Gasch" <[EMAIL PROTECTED]> wrote: > > > Hi, > > > > I'm running freebsd 7.0-RELEASE-p4 on a 4-core amd64 box. Nearly 100% of > > 1 cpu is constantly being used handling irq19: uhci4 interrupts. This > > seems to happen both with and without any USB devices plugged in: > > > > vmstat -i > > interrupt total rate > > irq1: atkbd0 5 0 > > irq6: fdc0 1 0 > > irq17: mskc0 dc0 1180547 18 > > irq18: skc0 uhci2* 163250699 2512 > > irq19: uhci4++ 3187989508 49072 > > I think the ++ here indicates that two or more devices are sharing this > interrupt. Try doing "grep irq.*19 /var/run/dmesg.boot" to see which > ones. One of these devices could be the culprit. > > --- > Gary Jennejohn > ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"