cvs commit: src/sys/dev/ata ata-dma.c ata-pci.c

[Soren Schmidt]

It seems John Baldwin wrote:
> 
> On 22-Oct-2003 Alexander Leidinger wrote:
> > On Wed, 22 Oct 2003 07:43:53 -0700 (PDT)
> > John Baldwin <jhb at freebsd.org> wrote:
> > 
> >>   Modified files:        (Branch: RELENG_4)
> >>     sys/dev/ata          ata-dma.c ata-pci.c 
> >>   Log:
> >>   MFC: Add support for the Intel ICH5 SATA controller in both legacy and
> >>   enhanced mode.
> >>   
> >>   Submitted by:   Dan Strick <strick at covad.net>
> >>   Approved by:    re (murray)
> > 
> > Does this actually work? On -current it doesn't work flawlessly with an
> > Asus P4P800 board. Accessing one SATA disk results in a lot of spurious
> > interrupts, and using a vinum stripe over 2 SATA disks results in a
> > kernel panic in the ata code.

For current you need to update to the latest I've just yesterday fixed 
the DMA problem that occoured on the SiI3112, and the fix might help
on other problems as well. And please loose vinum out of the mix here,
it is not helping to find evt problems.

> Yes.  Current likely needs the equivalent fix to its interrupt routine
> that this patch includes to avoid the spurious interrupts.

If we are talking about the same thing, current does NOT need that patch
and its also bad on releng4, it clears the error bits which are read
later on for errors...

-Søren