Jason Thorpe writes:
> I've been hacking on this for a few days now, and have it working well 
> enough that I might as well spill the beans...
> 
> I have COMPAT_HPUX pretty much basically working in my private source 
> tree.  There still several problems with it (I had to #if 0 out a bit of 
> code to make it compile :-), but it's enough to run demand-paged (both 
> statically and dynamically linked) HP-UX 9.x binaries "O.K.".  I've been 
> using an ls(1) and a little "hello, world" program .. ls(1) even gets 
> passwd/group stuff from the YP server correctly.
> 
> I tried a telnet(1) ... it .. kinda works .. behaves somewhat stangely, 
> but it's dynamically linked, and was a good test of that.  I suspect the 
> strange behavior is either socket or signal related; when ls(1) talks to 
> the YP server, it takes a _long_ time, and telnet is complaining of 
> sleep() and select(), so, we'll see.
> 

I've noticed the "ls" problem on a pc running NetBSD. Its a problem that sits
inside the getgrent() function. Take the + out of /etc/group and watching
things dramatically speed up.

Ron 

--
U of MN - CS, ACM SysAdmin 	| Life is a simulation of reality.	
email:roskens@cs.umn.edu	| http://www.cs.umn.edu/users/roskens/