Re: Q:How to build OCI client on AIX 3.2

From: David Hinz <dhinz_at_dna406.dna.mci.com>
Date: 1996/09/19
Message-ID: <3241860C.446B_at_dna406.dna.mci.com>#1/1


In reference to my previous post we are trying to build with Oracle 7.3.2 on AIX. When we build with 7.3.2 on DEC UNIX 3.2c everything builds fine with just the libclntsh.so library.

dave.

David Hinz wrote:
>
> We are trying to build Oracle OCI programs on AIX 3.2 and are running
> into problems getting the executables to link with libclntsh.a. If we
> link with just the $ORACLE_HOME/lib/libcnltsh.a library we get the
> following unresolved symbols:
> .ocom
> .oexfet
> .odefin
> .olog
> .ologof
> .oclose
> .oparse
> .oerhms
> .oopen
> .ocof
>
> By going in and deciphering the clntsh.mk file in
> $ORACLE_HOME/rdbms/lib. I was able to determine that all of the "^o*"
> objects were exported from the libclient.a before it is linked into the
> libclntsh.a library.
>
> Does this mean that the above object files (ocom, olog, ologof) should
> not be used in the shared client environment? Why are the o* objects
> removed?
>
> We were able to get an executable built by linking our program with
> libclntsh.a and libclient.a which then linked in the above list of
> objects. Is this a safe thing to do?
>
> Is there a better way to build OCI client executables using shared
> libraries?
>
> I also noticed that the oracle.mk file uses the oraxlc script to set up
> a bunch of environment variables when compiling the cdemo* demo
> programs. Should we also use oraxlc or can xlc be used safely?
>
> Does Oracle provide any documentation on build OCI programs??????
>
> Thanks for any info.
> dave.
> --
> David Hinz --- MCI Telecommunications Corporation
> E-mail: dhinz_at_dna406.dna.mci.com Pager: 1-800-PAGE-MCI PIN: 172-3148
> Phone: (303) 291-6108 VNET: 636-6108 Fax: (303) 291-6365
 

-- 
David Hinz --- MCI Telecommunications Corporation
E-mail: dhinz_at_dna406.dna.mci.com   Pager: 1-800-PAGE-MCI PIN: 172-3148
Phone: (303) 291-6108   VNET: 636-6108    Fax: (303) 291-6365
Received on Thu Sep 19 1996 - 00:00:00 CEST

Original text of this message