RE: NForce2 pseudoscience stability testing (2.6.0-test11)

[Allen Martin]

> -----Original Message-----
> From: Jesse Allen [mailto:the3dfxdude@xxxxxxxxxxx] 
> Sent: Wednesday, December 03, 2003 6:46 PM
> To: b@xxxxxxxxxxxxx
> Cc: linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: NForce2 pseudoscience stability testing (2.6.0-test11)
> 
> I suppose I could try right now.  I don't have a pci ide with 
> me right now, but I do have pci scsi cards.  But doesn't 
> running with the generic ide driver basically prove the same 
> thing?  APIC & Generic IDE: works, PIC & nForce IDE: works, 
> APIC & nForce IDE: deadlocks.  It's not like we are expecting 
> faulty nforce ide hardware, or are we?

I don't think there's any faulty nForce IDE hardware or we would have heard
about it from windows users (and we haven't).  

The problem with comparing the nForce IDE driver against the generic IDE
driver is that the generic IDE driver won't enable DMA, so the interrupt
rate will be much different.  If there's some interrupt race condition in
APIC mode, disabling DMA may mask it.

Also are people who are having problems using rounded or flat cables?  It's
possible the problem could be related to DMA CRC errors.  Switching to flat
cables can help with that.

-Allen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/