Subject: Re: Suggested fix for NetBSD 1.2 kernels
To: None <tech-kern@NetBSD.ORG>
From: Bruce Allen Ediger <metamatic!>
List: tech-kern
Date: 09/26/1996 20:06:05
On Tue, 17 Sep 1996, Jonathan Stone wrote:

> For reference, a tidier version of the patch, which works better
> on machines with clocks not at 100Hz (i.e., most things other
> than i386es and sparcs), is appended at the end of this message.

I hacked this patch into the vm/vm_pageout.c source of a NetBSD/sparc 1.1
kernel, knowing this was probably a very bad idea.  So, I'm not criticizing
Mr Stone.  I'd just like to ask about a weird anomaly that seemingly arose
from this patch.

The kernel with the modified calls to tsleep() in vm/vm_pageout.c caused
twm to coredump with a bus error most of the time.

Two different twms coredumped: the one from the Sept 15th, 1995 binary
X11R6 distribution that used to be on one of the NetBSD ftp sites.
The 2nd twm, I recompiled from X11R6pl12 sources using the Princeton/Bell
Labs lcc-3.5 compiler.  I've been using both twm executables for a while
now, and I've never noticed anything odd from them.

Nothing else seemed to have troubles.