As author of the vfork strategy ... I'm supportive of the directions in this thread.👍👍
David's patch seems like clear progress (although maybe now that we have configure, we can make the spawn strategy conditional on HAVE_SPAWN_H) vfork is even (!) less in favor than it used to be, so migrating off of it seems good. Just make sure we don't bring back the OOM problems or unacceptable overhead from e.g. an second exec. Create a 100GB heap, then run /bin/true 1000 times. That big internal comment I wrote years ago is suffering from bitrot. IIRC vfork used to call a special vfork system call, NOT clone, years ago.
