Bob,
We do not have interrupt assignment problems for the HCA that I'm aware of
using the A8N32-SLI Deluxe MB. We do have message signaled interrupts
enabled as well (Roland's suggestion to Jason). Output follows.
Steve W.
<syslog at boot, int assignment for HCA>
Feb 16 07:50:26 localhost kernel: ACPI: PCI interrupt 0000:07:00.0[A] -> GSI
17 (level, low) -> IRQ 225
<dump of interrupt assignments/counts>
[EMAIL PROTECTED] ~]# cat /proc/interrupts
CPU0
0: 55263251 IO-APIC-edge timer
1: 16 IO-APIC-edge i8042
8: 0 IO-APIC-edge rtc
9: 0 IO-APIC-level acpi
12: 402 IO-APIC-edge i8042
177: 10162 IO-APIC-level libata
185: 0 IO-APIC-level libata, ohci_hcd
193: 5559144 IO-APIC-level ehci_hcd, eth0
201: 0 IO-APIC-level NVidia CK804
209: 18286320 IO-APIC-level xxxxxx
217: 19671760 IO-APIC-level xxxxxx
225: 33752396 IO-APIC-level InfiniHost_III_Lx0
NMI: 2101
LOC: 55255250
ERR: 0
MIS: 0
________________________________________
From: Bob Pearson [mailto:[EMAIL PROTECTED]
Sent: Thursday, February 16, 2006 8:45 AM
To: 'Bill Boas'; [email protected]
Cc: 'Steve Welch'; 'Jim Mott'
Subject: RE: [openib-general] Low cost MBs compatible with Mellanox PCIE
Cards
Jim has worked with the A8N-SLI board. We have worked with a newer version
called A8N32-SLI Deluxe which has much better PCIe support. Ours has only
run the gen1 stack so I have no idea what it does with the gen2 stack. As
for the BIOS issue mentioned perhaps Steve W could take a look and see if
the comment from Jason makes any sense but my impression is that everything
was just fine.
________________________________________
From: Bill Boas [mailto:[EMAIL PROTECTED]
Sent: Wednesday, February 15, 2006 10:56 PM
To: Bob Pearson
Subject: Fwd: [openib-general] Low cost MBs compatible with Mellanox PCIE
Cards
Do we have something to contribute to this thread?
---------- Forwarded message ----------
From: Roland Dreier <[EMAIL PROTECTED]>
Date: Feb 15, 2006 4:52 PM
Subject: Re: [openib-general] Low cost MBs compatible with Mellanox PCIE
Cards
To: Jason Gunthorpe <[EMAIL PROTECTED]>
Cc: [email protected]
Jason> Unfortunately the board's BIOS's ACPI tables are totally
Jason> broken and the BIOS assigns every interrupt source in the
Jason> system to IRQ 5 :< At least it does work in full APIC mode
Jason> which means it might be possible to get MSI working if the
Jason> nvidia bridge isn't broken.
I've not seen any problems with MSI/MSI-X with nforce4 and PCIe HCAs
on both my Asus A8N-SLI and HP DL145G2 systems.
Just building a kernel with CONFIG_PCI_MSI enabled might help, since
it changes the way the kernel numbers interrupts.
- R.
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit
http://openib.org/mailman/listinfo/openib-general
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
