Hello,
i'm trying FreeBSD 7.1 BETA2 in amd64 and i'm having issues with
interrupt
storms. The machine works without problems as long as i don't stress much
the HD. If i have high IO load after a few hours the system starts to have
interrupt storms and never again recovers from them until i restart.
My SATA card is:
atapci0@pci0:0:18:0: class=0x01018f card=0x73271462 chip=0x43801002 rev=0x00
hdr=0x00
vendor = 'ATI Technologies Inc'
device = 'IXP SB600 Serial ATA Controller'
class = mass storage
subclass = ATA
The MotherBoard is MSI K9AG Neo 2[1] and the kernel is
GENERIC without USB and firewire support.
After digging a bit in the source i've find that linux
people have a quirk for this chipset and i've been unable
to find in our code a similar quirk. You can find the quirk
here:
http://fxr.watson.org/fxr/source/drivers/pci/quirks.c?v=linux-2.6#L870
I don't know exactly what does it do, neither i do know
if it have anything to know with my interrupt issues, but
could someone who knows a bit better tell me if that
could fix the issue or if we need that quirk for other
purposes?
Regards.
Thanks in advance.
[1]:
http://global.msi.com.tw/index.php?func=proddesc&prod_no=1241&maincat_no=1
--
La prueba m?s fehaciente de que existe vida inteligente en otros
planetas, es que no han intentado contactar con nosotros.
Victor Balada Diaz wrote:> i'm trying FreeBSD 7.1 BETA2 in amd64 and i'm having issues with interrupt > storms. The machine works without problems as long as i don't stress much > the HD. If i have high IO load after a few hours the system starts to have > interrupt storms and never again recovers from them until i restart. > > My SATA card is: > > atapci0@pci0:0:18:0: class=0x01018f card=0x73271462 chip=0x43801002 rev=0x00 hdr=0x00 > vendor = 'ATI Technologies Inc' > device = 'IXP SB600 Serial ATA Controller' > class = mass storage > subclass = ATA > > The MotherBoard is MSI K9AG Neo 2[1] and the kernel is > GENERIC without USB and firewire support. > > After digging a bit in the source i've find that linux > people have a quirk for this chipset and i've been unable > to find in our code a similar quirk. You can find the quirk > here: > > http://fxr.watson.org/fxr/source/drivers/pci/quirks.c?v=linux-2.6#L870 > > I don't know exactly what does it do, neither i do know > if it have anything to know with my interrupt issues, but > could someone who knows a bit better tell me if that > could fix the issue or if we need that quirk for other > purposes?It seems this quirk only changes PCIR_PROGIF and PCIR_SUBCLASS registers to PCIP_STORAGE_SATA_AHCI_1_0 and PCIS_STORAGE_SATA values. I don't know which magic these changes should do, but I think they aren't needed. Your controller should work in AHCI mode without this quirk. -- WBR, Andrey V. Elsukov