Re: pgsql: Replace our hacked version of ax_pthread.m4 with latest upstream - Mailing list pgsql-committers

From Tom Lane
Subject Re: pgsql: Replace our hacked version of ax_pthread.m4 with latest upstream
Date
Msg-id 4280.1438545459@sss.pgh.pa.us
Whole thread Raw
In response to Re: pgsql: Replace our hacked version of ax_pthread.m4 with latest upstream  (Heikki Linnakangas <hlinnaka@iki.fi>)
List pgsql-committers
Heikki Linnakangas <hlinnaka@iki.fi> writes:
> I went through the results, and I believe the script is working the way
> we want now. Attached is a trimmed result set, where I have removed all
> the animals where there was no change in the PTHREAD_CFLAGS value chosen
> between the three versions (configlogs-trimmed). I've also attached the
> raw results and the query I used, for the sake of the archives
> (configlogs-raw)

Cool, thanks for doing the legwork on this!

> On some platforms, however, we are now getting duplicate -D options,
> e.g. on dromedary:

>> PTHREAD_CFLAGS='-D_REENTRANT  -D_REENTRANT -D_THREAD_SAFE -D_POSIX_PTHREAD_SEMANTICS'

> That's harmless, but it would be nice to not clutter the cc command
> line. It's happening because we are doing this in configure.in:

>> # Some platforms use these, so just define them.  They can't hurt if they
>> # are not supported.  For example, on Solaris -D_POSIX_PTHREAD_SEMANTICS
>> # enables 5-arg getpwuid_r, among other things.
>> PTHREAD_CFLAGS="$PTHREAD_CFLAGS -D_REENTRANT -D_THREAD_SAFE -D_POSIX_PTHREAD_SEMANTICS"

> I'm tempted to just remove -D_REENTRAND and -D_THREAD_SAFE, because
> ax_pthread.m4 knows about those, and will add them on platforms where
> they are needed. Any objections?

Seems worth a try anyway.

            regards, tom lane


pgsql-committers by date:

Previous
From: Andrew Dunstan
Date:
Subject: Re: Taint mode in PL/Perl
Next
From: Terry Scheingeld
Date:
Subject: Re: Taint mode in PL/Perl