Date: Tue, 8 Aug 2006 22:01:36 -0700 From: John-Mark Gurney <gurney_j@resnet.uoregon.edu> To: Nenhum _de_Nos <matheusber@gmail.com> Cc: freebsd-sparc64@freebsd.org Subject: Re: panic: pcib: uncorrectable error Message-ID: <20060809050136.GK99774@funkthat.com> In-Reply-To: <4956a5e50608081911n1417ea89r58d148c5484b8313@mail.gmail.com> References: <4956a5e50608081911n1417ea89r58d148c5484b8313@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Nenhum _de_Nos wrote this message on Tue, Aug 08, 2006 at 23:11 -0300:
> i have two sunblades 150 one runs FreeBSD 6.0-p9, the other Solaris 8.
> this freebsd is a nfs server that when in huge nfs activity (such as
> untar a 100MB file) panics as follow:
>
> IOMMU fault virtual address 0xc19c2000
> panic: pcib: uncorrectable error AFAIR 0xcec54708 AFSR 0x4100ff002080000
>
> this machine was running 6.1-STABLE, 6.1-RELEASE and all had the same
> error. i've changed the hds and run FreeBSD in the other sunblade, but
> the very same thing happened.
>
> i've tested memory, as for changing among them both and nothing also
> (i have no other pc133 machine).
>
> solaris runs fine, but i really want FreeBSD on them :)
>
> the hardware is the factory default, just the server has a seagate
> 300GB IDE hd and one unused Realtek Gigabit Card.
>
> if anyone can help :)
Hmmm.. could you try turning off dma on ide?
hw.ata.ata_dma=0
It'd be useful to find out which device was doing dma that caused the
fault, because it's likely overrunning the mapping provided... With
enough munging around in iommu_decode_fault, you could work your way
back to the map/tag that caused the fault, and then figure out which
device owned the tag...
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060809050136.GK99774>