PowerStack interrupt problem

PowerStack interrupt problem

[Geert Uytterhoeven]

	Hi,

We're having interrupt problems with a Motorola PowerStack running 2.4.9 (from
MontaVista rsync).

Symptoms are:
  - We receive lots of PS/2 mouse interrupts (irq 12), a few hundreds or
    thousands for each mouse move
  - The number of BAD interrupts increases when moving the mouse, a few tens
    for each mouse move
  - Sometimes the machine locks up (no ping response) after heavy mouse
    movement
  - The PS/2 keyboard is fine

Anyone with a suggestion? Any other PowerStack users with similar problems?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: PowerStack interrupt?problem

[Tom Rini]

On Sun, Sep 30, 2001 at 01:23:46PM +0200, Geert Uytterhoeven wrote:

> We're having interrupt problems with a Motorola PowerStack running 2.4.9 (from
> MontaVista rsync).

Goody, which? :)  I've got a 604/100 right here.

> Symptoms are:
>   - We receive lots of PS/2 mouse interrupts (irq 12), a few hundreds or
>     thousands for each mouse move
>   - The number of BAD interrupts increases when moving the mouse, a few tens
>     for each mouse move
>   - Sometimes the machine locks up (no ping response) after heavy mouse
>     movement
>   - The PS/2 keyboard is fine

Hmm.  I sort-of wonder if the PS/2 changes a while back goofed something up.
My kbd is fine too, and I'll try and throw a mouse on later this week.
Can you grab the bk version and go back to 2.4.8, .7, et al and see if
it works?  I think some PS/2 stuff was done in maybe .5 or .6..

--
Tom Rini (TR1265)
http://gate.crashing.org/~trini/

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: PowerStack interrupt?problem

[Geert Uytterhoeven]

On Sun, 30 Sep 2001, Tom Rini wrote:
> On Sun, Sep 30, 2001 at 01:23:46PM +0200, Geert Uytterhoeven wrote:
> > We're having interrupt problems with a Motorola PowerStack running 2.4.9 (from
> > MontaVista rsync).
>
> Goody, which? :)  I've got a 604/100 right here.

The one that looks like a HiFi, with 133 MHz 604.

> > Symptoms are:
> >   - We receive lots of PS/2 mouse interrupts (irq 12), a few hundreds or
> >     thousands for each mouse move
> >   - The number of BAD interrupts increases when moving the mouse, a few tens
> >     for each mouse move
> >   - Sometimes the machine locks up (no ping response) after heavy mouse
> >     movement
> >   - The PS/2 keyboard is fine
>
> Hmm.  I sort-of wonder if the PS/2 changes a while back goofed something up.
> My kbd is fine too, and I'll try and throw a mouse on later this week.
> Can you grab the bk version and go back to 2.4.8, .7, et al and see if
> it works?  I think some PS/2 stuff was done in maybe .5 or .6..

I cannot test that right now (I don't own the machine, it was sitting next to
me at the Oldenburg Linux/non-ia32 meeting), but the owner promised to join
the mailing list, so I hope he'll do.

Perhaps the problem is related to the PS/2 mouse problem on my LongTrail. I've
always used a serial mouse, because Linux locks up if I insert the PS/2 mouse
(while Linux is running).

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: PowerStack interrupt?problem

[Nils Krueger]

> > Goody, which? :)  I've got a 604/100 right here.
>
> The one that looks like a HiFi, with 133 MHz 604.

It's called PowerStack Series E.

> > Hmm.  I sort-of wonder if the PS/2 changes a while back goofed something
up.
> > My kbd is fine too, and I'll try and throw a mouse on later this week.
> > Can you grab the bk version and go back to 2.4.8, .7, et al and see if
> > it works?  I think some PS/2 stuff was done in maybe .5 or .6..

I had the same problem with 2.4.5, .7-pre3 and .8-pre3. I haven't watched
the interrupts then, but it showed the same symptoms. (keyboard, mouse,
network, scsi are stalled but graphic card is still working)

This night I compiled 2.4.10-pre12 (from MontaVista rsync) and still the
same. (Thank you, Geert, the video-mode kernel parameter works fine; fbcon
gives me 1024x768 directly after bootup now)

One keypress or release of a mouse button causes about 20 interrupts and
sometimes one or two BAD interrupts. (BTW, it's the same on my so far most
stable 2.2.17) Should be normally 4 interrupts (and of course no BAD one)?

> but the owner promised to join
> the mailing list, so I hope he'll do.

ACK

Nilz.

--
http://www.hoschi.de/nilz   (Nilz' Homepage)


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: PowerStack interrupt?problem

[Tom Rini]

On Mon, Oct 01, 2001 at 01:40:19PM +0200, Nils Krueger wrote:
> > > Goody, which? :)  I've got a 604/100 right here.
> >
> > The one that looks like a HiFi, with 133 MHz 604.
>
> It's called PowerStack Series E.

Same thing I've got.

> > > Hmm.  I sort-of wonder if the PS/2 changes a while back goofed something
> up.
> > > My kbd is fine too, and I'll try and throw a mouse on later this week.
> > > Can you grab the bk version and go back to 2.4.8, .7, et al and see if
> > > it works?  I think some PS/2 stuff was done in maybe .5 or .6..
>
> I had the same problem with 2.4.5, .7-pre3 and .8-pre3. I haven't watched
> the interrupts then, but it showed the same symptoms. (keyboard, mouse,
> network, scsi are stalled but graphic card is still working)

Okay, I just tried 2.4.17-pre8 (from _devel) on my PowerStack Series E,
and the ps2 mouse seems to be working with gpm (gpm -m /dev/psaux -t
ps2).

Although it is worth noting that if i did:
while `true`;do cat /proc/interrupts ; sleep 2;done
Interrupts seemed to stop eventually.  It also takes ~100 interrupts for
a tiny mouse movement, which I'm not sure if is right or wrong..

> One keypress or release of a mouse button causes about 20 interrupts and
> sometimes one or two BAD interrupts. (BTW, it's the same on my so far most
> stable 2.2.17) Should be normally 4 interrupts (and of course no BAD one)?

Odd.  It seems to take 3 interrupts/keypress here (6 for up-arrow,
enter).

--
Tom Rini (TR1265)
http://gate.crashing.org/~trini/

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/