> I was trying to compile up cvs today using the Aug. 1 binary snapshot.
> It compiles ok after some tweaking but would just go into an infinite
> loop when run.

This has been discussed quite a bit on the mailing lists.  To summarize:
NetBSD's sigismember() macro was POSIX compliant, an implementation is
permited to not detect errors; CVS is relying on non-portable behavior; we
got tired of definding NetBSD's behavior each time someone tried to
compile CVS; so we gave up and removed the sigaddset, sigdelset and 
sigismember macros.  I don't remember the exact date they were removed, 
but they will be gone when you upgrade your source.

If you are interested, you can look up the details in the NetBSD mailing 
list archives.

	--jtc