Hi,
On 2021-10-14 22:46:07 -0400, Tom Lane wrote:
> Thomas Munro <thomas.munro@gmail.com> writes:
> > I suppose you could point SHELL at an unsigned copy of sh (codesign
> > --remove-signature, or something from brew/ports/x) so that GNU make
> > should respect, but I don't know how many other exec("/bin/sh") calls
> > might be hiding around the place (I guess perl calls system()?) and
> > might require some kind of LD_PRELOAD hackery... not much fun.
>
> Yeah. I thought about invoking everything via a small wrapper
> that restores the correct setting of DYLD_LIBRARY_PATH. We could
> perhaps make that work for the invocations of test postmasters
> and psqls from "make" and TAP scripts, but hacking up our code's
> sundry uses of system(3) like that seems like it'd be very messy,
> if feasible at all.
It does sound like using relative rpaths might be the thing we want - and like
they've been available since 10.5 or something.
Is there a reason we're using absolute rpaths on a bunch of platforms, rather
than relative ones, which'd allow relocation?
Greetings,
Andres Freund