Re: GDB cvs won't build on AIX 4.1

[Andrew Cagney <ac131313@cygnus.com>]

I just created gdb/390 to track this but then noticed that Peter is an 
AIX maintainer.  Was that an OK?

(Oh and don't forget, no one maintains configure.in everyone gets to 
change it, but only if they dare :-)

enjoy,
Andrew

> I am no autoconf expert, but it looks ok to me.
> It would be nice, if it could make it into the upcoming GDB release.
> 
> 
>> --=-=-=
>> 
>> On Feb  4, 2002, "Peter.Schauer" <Peter.Schauer@Regent.E-Technik.TU-Muenchen.DE> wrote:
>> 
> 
>> > -mminimal-toc is only available with gcc, so we might loose again, if someone
>> > tries to build GDB with the native compiler.
> 
>> 
>> Besides, I'm not convinced the performance loss of -mminimal-toc would
>> be smaller than that of -bbigtoc, so I went with the latter for now.
>> Tested on AIX 4.1.  Ok to install?
>> 
>> 
>> --=-=-=
>> Content-Type: text/x-patch
>> Content-Disposition: inline; filename=gdb-aix-bigtoc.patch
>> 
>> Index: gdb/ChangeLog
>> from  Alexandre Oliva  <aoliva@redhat.com>
>> 
>> * configure.in (gdb_cv_bigtoc): Check for -bbigtoc on AIX.
>> * configure: Rebuilt.
>> 
>> Index: gdb/configure.in
>> ===================================================================
>> RCS file: /cvs/src/src/gdb/configure.in,v
>> retrieving revision 1.82
>> diff -u -p -r1.82 configure.in
>> --- gdb/configure.in 2002/02/24 04:31:13 1.82
>> +++ gdb/configure.in 2002/02/24 13:05:16
>> @@ -511,6 +511,23 @@ AC_MSG_RESULT($gdb_cv_scanf_has_long_dou
>> > AC_FUNC_MMAP
>> > +case ${host_os} in
>> +aix*)
>> +  AC_CACHE_CHECK([for -bbigtoc option], [gdb_cv_bigtoc], [
>> +    SAVE_LDFLAGS=$LDFLAGS
>> +
>> +    case $GCC in
>> +    yes) gdb_cv_bigtoc=-Wl,-bbigtoc ;;
>> +    *) gdb_cv_bigtoc=-bbigtoc ;;
>> +    esac
>> +
>> +    LDFLAGS=$LDFLAGS\ $gdb_cv_bigtoc
>> +    AC_TRY_LINK([], [int i;], [], [gdb_cv_bigtoc=])
>> +  ])
>> +  CONFIG_LDFLAGS="${CONFIG_LDFLAGS} ${gdb_cv_bigtoc}"
>> +  ;;
>> +esac
>> +
>> dnl See if thread_db library is around for Solaris thread debugging.  Note that
>> dnl we must explicitly test for version 1 of the library because version 0
>> dnl (present on Solaris 2.4 or earlier) doesn't have the same API.
>> 
>> --=-=-=
>