I have no idea what trampoline stacks do, but our threads do fiddle with
the stack of course. Can you first connect your venus to our testserver;
if there are threading problems, they will likely also show up in Venus. 

There are some small test programs for the threads package in
lib-src/mlwp. Try those too. 

Also since there are RVM problems, running the RVM basher would be a good
idea.  This is a much simpler program  than the server.

FYI, we are leaving RH4.2 behind, we don't have enough people to keep up
with old stuff at the moment. 

I noticed you are filled with hate; just relax, it's supposed to be fun,
otherwise don't do it. 

- Peter -

On Fri, 17 Apr 1998 thoth_at_purplefrog.com wrote:

> 
>   At the moment I have no leads to go on with coda, other than diving into
> the source.  I'm getting too old for that.  However, I had an idea that maybe
> my kernel wasn't compatible.
> 
>  frop:20 $ uname -a
> Linux frop 2.0.33 #9 Tue Apr 14 14:56:36 EDT 1998 i586
> 
>   This also includes the "non-executable stack" patch with trampoline
> emulation.  Could this be causing problems?  Should I try disabling that
> option?  (I'd rather not.  The internet is a VERY hostile place these days. 
> The ORTGE web server logs 10-30 netbios probes a day, each consisting of 1-20
> packets.)  Should I try one of the experimental/broken 2.1.x kernels?
> 
>   I'm running a 5.3.12 libc
> 
> lrwxrwxrwx   1 root     root           24 Apr  8 12:20 /usr/lib/libc.so -> ../../lib/libc.so.5.3.12
> 
>   I've already had abominable luck upgrading to 5.4.44 (totally fucked up
> name resolution) and refuse to upgrade to RedHat 5.0 with glibc until they
> eliminate a few hundred more buffer overflows.  keeerist.
> 
>   I am filled with hate, but it's not all your fault.
> 
> -- 
> Bob Forsman                                   thoth_at_gainesville.fl.us
>            http://www.gainesville.fl.us/~thoth/
>