kernel 2.6.6 amd8111 apic bugs

kernel 2.6.6 amd8111 apic bugs

[Dan Hollis]

Sent to linux-ide ml several weeks ago and received no response. Cc'ing to 
linux-kernel in the hopes someone will be able to figure out whats wrong 
with amd8111 apic.

Responses in email please as im not subscribed to the list.

-Dan

---------- Forwarded message ----------
Date: Mon, 14 Jun 2004 17:34:57 -0700 (PDT)
From: Dan Hollis <goemon@anime.net>
To: J. Ryan Earl <heretic@clanhk.org>
Cc: Jens Axboe <axboe@suse.de>, linux-ide@vger.kernel.org
Subject: Re: kernel 2.6.6 amd8111 dma bugs

On Wed, 2 Jun 2004, J. Ryan Earl wrote:
> Jens Axboe wrote:
> >On Fri, May 28 2004, Dan Hollis wrote:
> >>replies to email as i'm not subscribed to the list.
> >>There seem to be regular dma timeouts:
> >>hdc: dma_timer_expiry: dma status == 0x24
> >>hdc: DMA interrupt recovery
> >>hdc: lost interrupt
> >>hda: dma_timer_expiry: dma status == 0x24
> >>hda: DMA interrupt recovery
> >>hda: lost interrupt
> >>Hardware:
> >>Opteron 140, Tyan Tomcat K8S (S2850)
> >Try disabling ACPI (in .config or boot acpi=off iirc)
> I had that problem, but not with ACPI, only when I forced APIC on.  It 
> was on the VIA controller which uses the same driver.

Turns out it was apic and not acpi at all. Booting with ACPI but noapic 
and I no longer get any dma errors.

Is the bug in the linux apic code or a hardware flaw in the opteron cpu? 
Or something else?

-Dan

Re: kernel 2.6.6 amd8111 apic bugs

[Andreas Steinmetz]

Dan Hollis wrote:
>>>>There seem to be regular dma timeouts:
>>>>hdc: dma_timer_expiry: dma status == 0x24
>>>>hdc: DMA interrupt recovery
>>>>hdc: lost interrupt
>>>>hda: dma_timer_expiry: dma status == 0x24
>>>>hda: DMA interrupt recovery
>>>>hda: lost interrupt
>>>>Hardware:
>>>>Opteron 140, Tyan Tomcat K8S (S2850)
>>>
>>>Try disabling ACPI (in .config or boot acpi=off iirc)
>>
>>I had that problem, but not with ACPI, only when I forced APIC on.  It 
>>was on the VIA controller which uses the same driver.
> 
> 
> Turns out it was apic and not acpi at all. Booting with ACPI but noapic 
> and I no longer get any dma errors.
> 
> Is the bug in the linux apic code or a hardware flaw in the opteron cpu? 
> Or something else?
> 

 From the X86-64 patch release notes of Andi Kleen:

Reports that dual Tyan S2885 and S2880 can lock up when multiple IDE 
channels are stressed in parallel. "noapic" or "ideX=serialize" seems to 
work around it. Andre Hedrick thinks it's a generic bug/race in the IDE 
code.

Looks like you have a similar problem on a UP board.
-- 
Andreas Steinmetz                       SPAMmers use robotrap@domdv.de