[Raw Msg Headers][Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Autoconf with zmailer2.99.34 and redhat/linux 3.0.3
> Hi,
>
> I'm just trying to install zmailer 2.99.34 on a rh linux 3.0.3/
> kernel 2.0.8 box. It seems the configure script has some bugs:
>
> [....]
> checking for sys/time.h... yes
> checking whether time.h and sys/time.h may both be included... yes
> checking whether struct tm is in sys/time.h or time.h... time.h
> checking for tm_zone in struct tm... no
> checking for tzname... yes
> checking for tm_gmtoff in struct tm... no
> checking for altzone... no
> Aiee, autoconfig did not recognize timezone mechanism! Time-zone offset calculation may give wrong results
>
> That is wrong IMHO. I'm sure that libc 5.2.18 has a TZ
> mechanism.
Yes it does, tests just did not find it at that time.
At the moment I am makeing tests at a diverse set of systems
to verify that the autoconfiguration produces workable system,
and have had success with a bit more bleeding-edge Linux
(my workstation, that is..) which with stock 2.99.34 did
give same complaint.
My success score is now:
- SPARC Solaris 2.4, gcc 2.7.0
- SPARC Solaris 2.5, gcc 2.7.2 / SparcWorks-cc
- SPARC SunOS 4.1, gcc 2.7.0 / bundled cc
- DEC OSF/1 3.2D, cc / cc -migrate / gcc 2.7.2
- SGI IRIX 6.1 (64 bit), bundled cc
- Linux 2.0.11/libc5.4.2, gcc 2.7.2
Yet to try:
- 32-bit SGI machines (IRIX 6.2 -- just got the id there)
- AIX 3.2.5, AIX 4.1 (At a 24 CPU SP4 cluster)
- Cray C94 (ok, that is a joke, but if I really want it,
I can get it..)
> checking for st_blocks in struct stat... yes
> checking for st_blksize in struct stat... yes
> checking for gdbm.h... yes
> checking for ndbm.h... yes
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^
> See this
Hmm...
> checking for sdbm.h... no
> checking for db.h... yes
> checking for dbm.h... no
> Located gdbm_open() routine at gdbm library
>
> ..and then this error message:
>
> configure: error: No <ndbm.h>, but tried to link-test dbm_open() ??
>
> db-devel 1.85-3 is correctly installed. /usr/include/ndbmh.h
> exists and is world-readable.
Ah, editorial error at configure.in
> -Andi
> --
> |andi@mlm.extern.lrz-muenchen.de Nonsense is better than no sense at all.
> | -NoMeansNo,0-1=2
/Matti Aarnio <mea@nic.funet.fi>