Thread: AIX buildfarm failure
I am seeing buildfarm failures on AIX because stdio.h is being included before pg_config.h (which has the definition of _LARGE_FILES). The problem is stemming from math.h including stdlib.h, which (after several more inclusions) ends up including stdio.h. This is where the fgetpos64 different definitions is coming from. If I move <math.h> down after "postgres.h" in nodeHash.c, the problem goes away. Do we want to consider putting math.h into the standard include set? Or is there a general rule that postgres.h needs to be the first include file (before system headers, etc)? Thanks,-rocco
"Rocco Altier" <RoccoA@Routescape.com> writes:
> If I move <math.h> down after "postgres.h" in nodeHash.c, the problem
> goes away.
Bruce, you broke it. Have you forgotten the fundamental inclusion rule?
postgres.h (or postgres_fe.h, or c.h) first, then system headers, then
our own other headers.
regards, tom lane
Fixed. --------------------------------------------------------------------------- Tom Lane wrote: > "Rocco Altier" <RoccoA@Routescape.com> writes: > > If I move <math.h> down after "postgres.h" in nodeHash.c, the problem > > goes away. > > Bruce, you broke it. Have you forgotten the fundamental inclusion rule? > postgres.h (or postgres_fe.h, or c.h) first, then system headers, then > our own other headers. > > regards, tom lane -- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +
Now it dies on nodeSubplan.c... I am guessing there will be others as well. Perhaps a check to make sure postgres.h is first in the includes can be added to the include checking scripts you have been updating? Thanks,-rocco > -----Original Message----- > From: Bruce Momjian [mailto:bruce@momjian.us] > Sent: Thursday, July 13, 2006 4:14 PM > To: Tom Lane > Cc: Rocco Altier; pgsql-hackers@postgresql.org > Subject: Re: [HACKERS] AIX buildfarm failure > > > > Fixed. > > -------------------------------------------------------------- > ------------- > > Tom Lane wrote: > > "Rocco Altier" <RoccoA@Routescape.com> writes: > > > If I move <math.h> down after "postgres.h" in nodeHash.c, > the problem > > > goes away. > > > > Bruce, you broke it. Have you forgotten the fundamental > inclusion rule? > > postgres.h (or postgres_fe.h, or c.h) first, then system > headers, then > > our own other headers. > > > > regards, tom lane > > -- > Bruce Momjian bruce@momjian.us > EnterpriseDB http://www.enterprisedb.com > > + If your life is a hard drive, Christ can be your backup. + >
Rocco Altier wrote: > Now it dies on nodeSubplan.c... > > I am guessing there will be others as well. I check them all the math.h mentions. --------------------------------------------------------------------------- > > Perhaps a check to make sure postgres.h is first in the includes can be > added to the include checking scripts you have been updating? > > Thanks, > -rocco > > > -----Original Message----- > > From: Bruce Momjian [mailto:bruce@momjian.us] > > Sent: Thursday, July 13, 2006 4:14 PM > > To: Tom Lane > > Cc: Rocco Altier; pgsql-hackers@postgresql.org > > Subject: Re: [HACKERS] AIX buildfarm failure > > > > > > > > Fixed. > > > > -------------------------------------------------------------- > > ------------- > > > > Tom Lane wrote: > > > "Rocco Altier" <RoccoA@Routescape.com> writes: > > > > If I move <math.h> down after "postgres.h" in nodeHash.c, > > the problem > > > > goes away. > > > > > > Bruce, you broke it. Have you forgotten the fundamental > > inclusion rule? > > > postgres.h (or postgres_fe.h, or c.h) first, then system > > headers, then > > > our own other headers. > > > > > > regards, tom lane > > > > -- > > Bruce Momjian bruce@momjian.us > > EnterpriseDB http://www.enterprisedb.com > > > > + If your life is a hard drive, Christ can be your backup. + > > -- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +