Re: [COMMITTERS] pgsql: Allow Pin/UnpinBuffer to operate in a lockfree manner.

On 2016-04-13 20:09:46 -0400, Tom Lane wrote:
> Andres Freund <andres(at)anarazel(dot)de> writes:
> > On 2016-04-13 11:20:19 -0700, Andres Freund wrote:
> >> Heh, I was wondering the same aftering sending the last email. Will do
> >> that then.
>
> > Pushed. Let's see what pademelon says...
>
> I lit off a run, but it'll be a few hours till we have results...

In hindsight obviously, this isn't sufficient because s_lock reports the
caller's file/line/function, not __FILE__ etc. So I guess we're left
moving initialization to an inline function - will do so tomorrow.

I'm also putting up an animal with clang that uses
CFLAGS='-std=c89 -Wc99-extensions -Werror=c99-extensions'
which actually catches this.

I'd previously tested with gcc and "-Wc90-c99-compat -Werror=c90-c99-compat
-Wno-long-long -Wno-error=long-long -std=c90", but that doesn't catch
that error.

(both find a couple trailing commas in enums, will fix)

- Andres