The new boot speed is really nice to watch.
I'm having one problem in that the irq table seems wrong now.
Has anyone brought HDAMA up under the new tree, booted linux, and seen interrupts work on the ethernet devices? Just checking. I am going to take a closer look tomorrow.
ron
Ron,
As soon as Eric gets our local repository synced with the public tree I'll build the HDAMA here. I have a new rev of the HDAMA from Rioworks, so I'll test it here and let you know how it turns out.
--jason--
"Ronald G. Minnich" rminnich@lanl.gov writes:
The new boot speed is really nice to watch.
I'm having one problem in that the irq table seems wrong now.
I have not any problems. pirq or mptable?
Nothing should have changed with respect to interrupts. I think there was a very modest tweak to mptable.c
Eric
On Tue, 2004-10-26 at 11:31, Eric W. Biederman wrote:
"Ronald G. Minnich" rminnich@lanl.gov writes:
The new boot speed is really nice to watch.
I'm having one problem in that the irq table seems wrong now.
I have not any problems. pirq or mptable?
Nothing should have changed with respect to interrupts. I think there was a very modest tweak to mptable.c
There was no problem actually. Ron forgot that the stupid old kernel can't do irq routing on K8.
Ollie
On Mon, 25 Oct 2004, Ronald G. Minnich wrote:
I'm having one problem in that the irq table seems wrong now.
nope, I was wrong, the table is fine.
2.6 even now, in 32-bit mode, can't handle the 0x746b interrupt router ... here is a patch.
ron --- linux-2.6.8.1-boot/arch/i386/pci/irq.c.org 2004-10-26 10:23:40.654334364 -0600 +++ linux-2.6.8.1-boot/arch/i386/pci/irq.c 2004-10-26 10:26:16.002263267 -0600 @@ -613,6 +613,9 @@ case PCI_DEVICE_ID_AMD_VIPER_7443: r->name = "AMD768"; break; + case 0x746b: /*PCI_DEVICE_ID_AMD_VIPER_746B*/ + r->name = "AMD746B"; + break; default: return 0; }