"disable ata" in kernel configuration not working?

[Doug White]

On Tue, 3 Jun 2003, Andy Farkas wrote:

> On Mon, 2 Jun 2003, Vladislav V. Zhuk wrote:
>
> > It's one of well known demonstration of bugs new ATA driver.
> > This BUG is very strong and don't resolve in kernel.conf.
> > I wrote about this 3 times, but Soren Schmidt <sos at FreeBSD.org>
> > is very busy for resolving this problem more than 1 year
> > or don't read stable@ and e-mail.
>
> Yes, the new ata driver now "owns" irq 14 & 15, although other devices
> seem to work ok if set to those irqs. On one of my boxes, I don't even
> have an ata0 or ata1 yet dmesg says:

Shared interrupts are OK in PCI.

> ...
> FreeBSD 5.1-BETA #0: Mon May 26 01:54:37 EST 2003
> ...
> ahc0: <Adaptec 274X SCSI adapter> at 0x4c00-0x4cff, irq 11 (level)
> ...
> atapci0: <Promise PDC20246 UDMA33 controller>
> port 0xd400-0xd41f,[...] irq 14 at device 5.0 on pci0

show us 'pciconf -lv' output. You probably actually do have a promise IDE
controller on your system, you just can't use it/connect to it.

Have you tried disabling it in the BIOS?

> Q for all: is systat(1) the only way to see interrupt activity?

vmstat -i

If your system is SMP, mptable(8) might be interesting too. If your board
is an Intel server board, you can pull the technical product
specification, which shows you the interrupt map.

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org