who (mis)uses vfork?

[mcvax!minster.york.ac.uk!forsyth at uunet.uu.net]

I was surprised to discover that SunOS 4.0 implements vfork!=fork.  Also,
there is apparently some magic in <vfork.h> on Sun4s to make it work on
those machines.  According to a (Sun) USENIX paper ``Virtual Memory
Architecture in SunOS'', there are programs that rely on a child process
being able to store things in the parent's memory after the vfork; they
malfunction if vfork==fork.  Really!  Does anyone have a list of these
cretinous programs?  For various reasons, I'd like to fix them.

Answers to me by mail, please, if possible; I'll summarise later.

ukc!minster!forsyth
forsyth at minster.york.ac.uk

[[ I know of one:  /bin/csh.  At least it used to do that to implement the
builtin "hashstat" command.  Good luck trying to fix it!  It might be
better to just ditch the command.  As for <vfork.h>, it only contains a
"#pragma" to help out the compiler.  --wnl ]]