Обсуждение: Problems compiling 6.5.3 on solaris 2.6

I'm trying to get postgresql 6.5.3 working on solaris 2.6 with gcc, and it's
just not cooperating.

First of all, during the configure process, I see:

--(snip!)--
checking for ANSI C header files... grep: can't open conftest.out
no
checking for sys/wait.h that is POSIX.1 compatible... yes
checking for arpa/inet.h... grep: can't open conftest.out
yes
checking for crypt.h... grep: can't open conftest.out
yes
checking for dld.h... grep: can't open conftest.out
yes
--(snip!)--

..and a whole bunch more (endian.h, float.h, getopt.h, that whole block
down to pwd.h), which scares me.

Then, further on, it gets down to the index stuff, and I get this:

--(snip!)--
gcc -I../../../include -I../../../backend    -Wall -Wmissing-prototypes -I..
/..    -c -o istrat.o istrat.c
istrat.c: In function `OperatorRelationFillScanKeyEntry':
istrat.c:492: `F_OIDEQ' undeclared (first use in this function)
istrat.c:492: (Each undeclared identifier is reported only once
istrat.c:492: for each function it appears in.)
istrat.c: In function `IndexSupportInitialize':
istrat.c:558: `F_OIDEQ' undeclared (first use in this function)
gmake[3]: *** [istrat.o] Error 1
gmake[3]: Leaving directory `/tmp/postgresql-6.5.3/src/backend/access/index'
gmake[2]: *** [submake] Error 2
gmake[2]: Leaving directory `/tmp/postgresql-6.5.3/src/backend/access'
gmake[1]: *** [access.dir] Error 2
gmake[1]: Leaving directory `/tmp/postgresql-6.5.3/src/backend'
gmake: *** [all] Error 2
--(snip!)--

Any ideas what should be defining F_OIDEQ?

I appreciate any help you can offer.

Thanks,

--
Ben Reed
breed@ops.com
Onion Peel Solutions

Ben Reed writes:

> I'm trying to get postgresql 6.5.3 working on solaris 2.6 with gcc,
> and it's just not cooperating.

Free advice: you really don't want to get it working either. You instead
want to try 7.0.2. ;-)

> checking for ANSI C header files... grep: can't open conftest.out
> no

I gotta hand it to those guys: very informative error message. The shell
code looks like this:

| ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
| { (eval echo configure:3273: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
| ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`

which should surely create a conftest.out file.

> gcc -I../../../include -I../../../backend    -Wall -Wmissing-prototypes -I..
> /..    -c -o istrat.o istrat.c
> istrat.c: In function `OperatorRelationFillScanKeyEntry':
> istrat.c:492: `F_OIDEQ' undeclared (first use in this function)

> Any ideas what should be defining F_OIDEQ?

Yup, it should be in backend/utils/fmgr.h. That file is created from a
bunch of other files, and I don't think this is the first time Solaris awk
or similar would have balked. If you want to investigate further you could
go into backend/utils and try to run the commands by hand somehow, but it
seems to me that your whole system is somehow in a shaky
state. (PATH? shared libraries? umask?, ...)


--
Peter Eisentraut                  Sernanders väg 10:115
peter_e@gmx.net                   75262 Uppsala
http://yi.org/peter-e/            Sweden