Re: address@hidden: 21.2.90: configure problems -- using gcc 3.1 and --x-includes option]

[Akim Demaille]

|     Are you sure this is 2.53?  This is the kind of problems that should
|     happen with 2.13 only.  Can we see config.log?  Thanks!
| 
| The Emacs bug report came from puneet, not from me.  He's the one
| you'd need to ask to get config.log.

Good.  Please, do send it to address@hidden

| It appears that we used Autoconf 2.52 to generate the Emacs configure
| file.  (Old versions won't work at all.)  Should we try it using
| Autoconf 2.53?

It shouldn't make any difference, the added robustness to warnings was
implement in the early times of 2.50 IIRC (but my memory fades).  We
need config.log to understand what happened anyway.

My opinion is just the same as my partners': we spent months to have a
correct handling of warnings vs errors.  Just to make one doubious
situation easier to handle, we might jeopardize much of what we have.

And then again, Autoconf *does* try to distinguish warnings from
errors (this is CVS Autoconf) because some compilers don't set the
exit status properly.

/tmp % cat configure.ac                                           nostromo 9:01
AC_INIT
AC_CHECK_HEADERS(stdio.h)
/tmp % cat gcc                                                    nostromo 9:01
#! /bin/sh
echo "$0: warning: defcon 1" >&2
exec /usr/bin/gcc "$@"
/tmp % ./configure CC=./gcc CPP='./gcc -E'                        nostromo 9:01
checking for gcc... ./gcc
checking for C compiler default output... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether ./gcc accepts -g... yes
checking for ./gcc option to accept ANSI C... none needed
checking how to run the C preprocessor... ./gcc -E
checking for ANSI C header files... yes
checking for sys/types.h... yes
checking for sys/stat.h... yes
checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking stdio.h usability... yes
checking stdio.h presence... yes
checking for stdio.h... yes


configure has detected it can ignore the warnings.  In the following
case it still can use stderr:

/tmp % cat cc                                                    nostromo Err 1
#! /bin/sh
/usr/bin/gcc "$@"
exit 0
/tmp % ./configure CC=./cc CPP='./cc -E'                          nostromo 9:03
checking for gcc... ./cc
checking for C compiler default output... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether ./cc accepts -g... yes
checking for ./cc option to accept ANSI C... none needed
checking how to run the C preprocessor... ./cc -E
checking for ANSI C header files... yes
checking for sys/types.h... yes
checking for sys/stat.h... yes
checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking stdio.h usability... yes
checking stdio.h presence... yes
checking for stdio.h... yes

In the following case (meaningless $? _and_ stderr) it cannot:

/tmp % cat deadcc                                                 nostromo 9:03
#! /bin/sh
echo "$0: warning: defcon 1" >&2
/usr/bin/gcc "$@"
exit 0
/tmp % ./configure CC=./deadcc CPP='./deadcc -E'                  nostromo 9:03
checking for gcc... ./deadcc
checking for C compiler default output... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether ./deadcc accepts -g... yes
checking for ./deadcc option to accept ANSI C... none needed
checking how to run the C preprocessor... ./deadcc -E
configure: error: C preprocessor "./deadcc -E" fails sanity check




So really, I believe much is already done to cope with warnings, when
we can.