Re: Hints on causes of crashes?

From: Johan Klockars <rand_at_cd.chalmers.se>
Date: Tue, 14 Mar 100 14:44:02 +0100 (MET)

> As long as you're using the Nova you don't have to worry about the blitter
> at all, as Nova-VDI never touch it. If the crashes occur in normal GEM-apps
> under Nova-VDI it's almost certainly not the blitter that's the problem.

I'd strike the 'almost' in the last sentence. ;-)

> > > >Leaving the mouse in 1 place is ok but if I keep moving it slightly (say in
> > > >small circles) then every now and then the cursor jumps randomly about 1cm.
>
> This sounds like a hardware-problem to me, but it's possible that there's

Actually, this is very similar to what you can get with fVDI if you don't
let a device driver do the mouse cursor drawing.
The problem then is that the drawing of the cursor takes too long, so the
interrupts will still be turned off when the next mouse package arrives
from the keyboard processor.

I'm not saying that exactly this is your problem (that would be quite
unlikely), but make sure you don't have anything installed that might turn
off interrupts for 'long' intervals.

> > > >When I drag windows around the screen by the menu bar sooner or later they
> > > >start to leave window frame 'droppings' and all window updates cease.
>
> If this happens on the Nova you probably haven't set the Nova-interface to
> "serialized" in the toolkit-driver.

Either that or there's something slightly wrong with the NOVA hardware or the
connection to it.

> > > >When I launch applications (e.g. Imagecopy) I fairly often get a bus error.
>
> Copyback-cache enabled??

Sure sounds like it to me.

-- 
  Chalmers University   | Why are these |  e-mail:   rand_at_cd.chalmers.se
     of Technology      |  .signatures  |            johan_at_rand.thn.htu.se
                        | so hard to do |  WWW/ftp:  rand.thn.htu.se
   Gothenburg, Sweden   |     well?     |            (fVDI, MGIFv5, QLem)
Received on to. april 13 2000 - 15:26:34 CEST

This archive was generated by hypermail 2.3.0 : ti. nov. 03 2015 - 20:07:54 CET