Subject: Re: misc kernel problems
To: None <port-mac68k@NetBSD.ORG>
From: Benoit MARTEL <magus@cs.mcgill.ca>
List: port-mac68k
Date: 03/09/1997 10:43:01
On Sun, 9 Mar 1997, Taras Ivanenko wrote:
[SNIP]
> Second, I have a Mac game, 'Chinese Tiles'. I noticed that when I boot
> NetBSD after playing this game the system makes absolutely horrible
> shierk *the first time* it is supposed to beep. Next beeps sound
> right. It seems to me that something is not initialized completely in
> the sound driver. (Hint: when, after the same game, MacOS wants to
> beep, it crashes.) I do not like this sound but it does not seem to
> do any harm to the system.
> 
> 	Taras Ivanenko

I get get a shrieking beep every time consistently now that I recompiled 
with newer sources (tarballs from last week). I dont do anything in MacOS 
but click on the booter.

I run a IIvx and to tell the truth this beep is not really worse then the 
one I used to get. Even since I ran BSD (since IIvx was supported with 
it's own console) the beep I got was very faint and like the sound you 
get when you plug headphones in or out of a stereo. Now for the first 
time, it's loud and a bit more like a "real" beep except that it's 
shrieking (chalk sound).

Is either of these the normal behaviour?

I'll update my sources to -current and see.
I just decided to try and follow -current and since this involves a lot 
of compiles, I was wondering what kind of compile time people are 
getting. On my IIvx with 8Mb and a "recent" HD it took me 5h15 wall clock 
time to compile a generic kernel from scratch. Is this on par with you 
guys? How much does more memory speed it up etc...

Is there any kind of performance stats for MacBSD on different Macs or 
different configurations? If there isn't what about running the "BYTE 
magazine universal UNIX Benchmark"?

If there are no such stats and people agree with BYTE's standard of speed 
I would volonteer to gather the benchmark's output from everybody and put 
up results on my web page. I'll wait a day or two to see the reactions 
and then post again to say whether it's a go (and give a pointer to the 
benchmark and on how to send me stuff).

Ben,

-----------------------------------------------------------------
"Because user errors often produce unpredictable results, the user
should try to avoid them."

IBM MVS/XA System Programming Library.