Date: Mon, 11 May 92 13:51:51 EDT
From: davidf AT algol DOT cs DOT umbc DOT edu (Mr. David W. Flater)
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Subject: HD Problems
Status: O

>>your 'theory' on making lost HD interrupts responsible is a good point, I think.
>>I'll try it next time such a problem occurs. Thanks for summarizing.
>
> Keep in mind that *all* interrupts occurring in protected mode are
> handled by go32 through exception handlers.  Try removing all disk
> cache software (the BIOS interrupt only happens in real mode) and see
> if that fixes the problem.

Under 1.05, I did this and everything else I could think of.  I have not had a
chance to repeat my experiments under 1.06, but from what I hear the problem is
widespread.  Curiously enough, a -small- disk cache seems to make go32 run
smoother.  I think the cache is somehow improving matters in ways other than
simply caching blocks of data.  But that's beside the point.

Somehow a barrage of HD interrupts makes things crash.  From what people are
saying it is probably a BIOS/Hardware timing problem that is not the fault of
go32.  Whether or not it can somehow be fixed in go32 is beyond the scope of my
knowledge.  It doesn't look like there is a general, H/W independent solution,
though.

-------------------------------------------------------------------------------
Mr. David W. Flater      davidf AT cs DOT umbc DOT edu                 "Nothing works."
Disclaimer:  Nobody ever holds my opinions.               "Nothing EVER works!"
-------------------------------------------------------------------------------