Re: [PATCH v3 1/1] Fix detection of preadv/pwritev support for OSX.

On Tue, Mar 30, 2021 at 6:37 AM James Hilliard
<james(dot)hilliard1(at)gmail(dot)com> wrote:
> Should it work if I just attach it to the thread like this?

Yes. It automatically tries patches that are attached to threads that
are registered on commitfest.postgresql.org on 4 OSes, and we can see
that it succeeded, and we can inspect the configure output and see
that only the two clang-based systems detected and used the new
unguarded-availability-new flags and used them.

This should be alphabetised better:

HAVE_PREAD => undef,
- HAVE_PREADV => undef,
+ HAVE_DECL_PREADV => 0,
HAVE_PSTAT => undef,

So the question here is really: do we want to support Apple cross-SDK
builds, in our configure scripts? It costs very little to switch from
traditional "does-this-symbol-exist?" tests to testing declarations,
so no objections here.

I doubt people will remember to do this for other new syscall probes
though, so it might be a matter of discussing it case-by-case when a
problem shows up. For example, I recently added another new test,
specifically targeting macOS: pthread_barrier_wait. One day they
might add it to libSystem and we might need to tweak that one
similarly.