From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Peter Eisentraut <peter_e(at)gmx(dot)net> |
Cc: | PostgreSQL Development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Removing src/backend from include path might break on HPUX |
Date: | 2000-07-05 17:39:06 |
Message-ID: | 3462.962818746@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> Now that we don't -I...backend anymore, this might break:
> src/include/c.h:
> #ifdef FIXADE
> #if defined(hpux)
> #include "port/hpux/fixade.h" /* for unaligned access fixup */
> #endif /* hpux */
> #endif
> The comments surrounding this make it pretty unclear whether this is still
> necessary. Tom, any idea?
FIXADE is not defined on the normal HPUX build, so it's a non-issue as
far as I know. AFAICT this file dates back to Berkeley days and exists
only to work around HP compiler bugs that existed about ten years ago.
We don't support HPUX <= 9.01 anymore anyway; there are plenty of other
reasons not to.
I have been planning for some time to nuke fixade.h, but haven't got
round to it. Be my guest (and don't forget to zap the reference to it
in src/backend/Makefile's install stuff too. AFAICT we don't need to
make an installed include/port subdirectory at all anymore...)
BTW, there are a couple of other templates that define NOFIXADE, which
seems to be something completely different; don't mistake it as being
the inverse of FIXADE.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | The Hermit Hacker | 2000-07-05 17:42:21 | Re: Re: zlib for pg_dump |
Previous Message | Denis Perchine | 2000-07-05 17:34:18 | Re[2]: Fwd: Re: Fwd: Problem with recv syscall on socket when other side closed connection |