Re: ac17-18 OS freeze

From: Andre Hedrick (andre@linux-ide.org)
Date: Sat Jun 17 2000 - 00:15:27 EST


Chris I spoke with Adam today in person about this ISR/INTR deadlock.

I asked him to consider a quick inline ISR test and check.

Please explain to Adam in full detail the problem.

I know where my problems are now, it is all chipset specific setup code.

On Fri, 16 Jun 2000, Christopher Zimmerman wrote:

> Andre Hedrick wrote:
>
> > On Thu, 15 Jun 2000, Christopher Zimmerman wrote:
> >
> > > This deadlock has been around since 2.4.0-test1. However, sine 2.4.0-test1
> > > there has been a large improvement. I have to try much harder now to kill my
> > > boxes. As for the IRQ state I'm not sure I've never checked. Since my test
> > > machines are fairly stable I'm actually using them to do some massive data
> > > processing. Once all my processes have completed I'll try the cat /dev/hda >
> > > /dev/null & thing and monitor the IRQ state. Actually come to think of it
> > > how do you find out the IRQ state for a particular device?
> >
> > Interesting that is when I began getting nailed for deadlocks also...
> >
> > The driver has to export it to a procfs file for reading.
> > Currently PDC2026x cards and CMD64x cards do this in ATA.
> >
> > Andre Hedrick
> > The Linux ATA/IDE guy
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.rutgers.edu
> > Please read the FAQ at http://www.tux.org/lkml/
>
> Okay I've done some more testing today and have narrowed the problem to the
> 3ware and adaptec aic-7896 cards. I ran my test on a box that has an integrated
> Intel 82371AB ide controller and a 4-port 3ware card installed. I ran multiple
> cats of the hda > /dev/null and 5 concurrent bonnies without a hitch. The
> deadlock only seems to arise on systems with both aic7xxx and 3ware cards
> installed.
>
> -zim
>
> Christopher Zimmerman
> AltaVista Company
> Index Build Engineering
>

Andre Hedrick
The Linux ATA/IDE guy

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



This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:14 EST