NForce2 pseudoscience stability testing (2.6.0-test11)
How is the performance of the generic IDE driver?
My experience that it was almost unusable (kernel 2.4.22, 2.4.23) (fsck taking hours, etc)
Quoted 17 lines
Quoted 4 lines
Quoted 4 lines
On Wed, 2003-12-03 at 01:58, Allen Martin wrote: -----Original Message----- From: Ian Kumlien [mailto:[email protected]] Sent:
Tuesday, December 02, 2003 4:47 PMQuoted 8 lines
Quoted 3 lines
Well, IDE is what i'd blame. My original experience about lost interrupts leads me to ide. Since i never loose interrupts without io-apic.
Can someone who has a system showing this problem try booting from a PCI IDE card to see if it makes any difference? I'd try the experiment here, but I can't reproduce the hanging that's being reported.
Or just to verify boot a kernel without the nvidia/amd ide driver and io-apic enabled.
-- Ian Kumlien
Quoted 4 lines · On Wed, 2003-12-03 at 02:23, [email protected] wrote:
How is the performance of the generic IDE driver?
My experience that it was almost unusable (kernel 2.4.22, 2.4.23) (fsck taking hours, etc)
Thats not the issue... The issue is if the machine survives or crashes. Again, it locks up for quite some time but it always comes back. And you get a "hda lost interrupt" message in dmesg.
If we all have that, and deadlock when using the amd/nvidia driver.. then we know that that might be the fault. The machine still locks for a while so, it's not just the ide, but it might be a good place to start.
Signature · 1 line
Ian Kumlien <[email protected]>