Re: PostmasterContext survives into parallel workers!?

Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> (Also vaguely on the list of things to clean up: can't we make it so
> that bgworkers aren't launched from inside a signal handler? Blech.)

So are other postmaster children, I believe. We could probably try
to rewrite the postmaster to not do useful work in signal handlers,
but rely on a lot of volatile flags set by the handlers. Not convinced
this would be anything but a cosmetic improvement, though. And it
could create new portability problems to replace any that it removed;
we'd have to be *absolutely* certain that the main select() call would
return with EINTR rather than resuming after any interrupt.

regards, tom lane