This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Surprises running gcc 3.2 configure on Cygwin


The first surprise was that everything (configure, make-bootstrap, make-check) worked! That's a surprise because it was consistantly failing only a few days ago and I didn't notice anything changing except the phase of the moon.

The following observations may not be bugs (except the first). They are just things that seem strange to me 'cause I don't know any better. Comments are welcome.
[ BTW, I don't mean to YELL. The caps below are mostly to distinguish my commentary from clippings out of the configure log.

My bootstrap compiler was gcc-3.2 as built by David Billighurst (David.Billinghurst@riotinto.com), with a working Cygwin-Ada.

My configure command was (from bash -x output):

++ /usr/test/gcc-3.2/src/configure --prefix=/usr/local --bindir=/usr/local/bin/i686-pc-cygwin --libdir=/usr/local/lib/gcc3.2/i686-pc-cygwin --infodir=/usr/local/info --mandir=/usr/local/man --datadir=/usr/local/etc/gcc3.2/i686-pc-cygwin --with-system-zlib --enable-nls --without-included-gettext --enable-shared --enable-threads=posix --enable-sjlj-exceptions --with-cpu=i686 --enable-version-specific-runtime-libs --with-newlib --extra-verbose i686-pc-cygwin
Configuring for a i686-pc-cygwin host.
*** This configuration is not supported in the following subdirectories:
target-libffi target-boehm-gc target-zlib target-libjava
(Any other directories should still work fine.)
Created "Makefile" in /usr/test/gcc-3.2/build using "mh-frag" and "mt-frag"
Reading specs from /usr/local/gcc-3_2/bin//../lib/gcc-lib/i686-pc-cygwin/3.2/specs
Configured with: /usr/local/src/gcc-3.2c/configure --prefix=/usr/local/gcc-3_2 --enable-threads=posix --enable-languages=ada --disable-sjlj-exceptions --enable-nls --without-included-gettext --disable-shared
Thread model: posix
gcc version 3.2 20020912 (prerelease)

THE BIGGEST SURPRISE was in the Makefile --- I've done this three or four times so far with similar results

#1
#
# Makefile for directory with subdirs to build.
# Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
# 1999, 2000, 2001 Free Software Foundation
#
# This file is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
#

# Tell GNU make 3.79 not to run the top level in parallel. This
# prevents contention for $builddir/$target/config.cache, as well
# as minimizing scatter in file system caches.
NOTPARALLEL = .NOTPARALLEL
$(NOTPARALLEL):

srcdir = /usr/local/src/gcc-3.2-1

prefix = /usr/local/
exec_prefix = --with-system-zlib
=========^^^^^^^^^^^^^^^^^^
bindir = /usr/local/bin/i686-pc-cygwin


#2: I shuffled configure parameters around and got the same bad selection
# Tell GNU make 3.79 not to run the top level in parallel. This
# prevents contention for $builddir/$target/config.cache, as well
# as minimizing scatter in file system caches.
NOTPARALLEL = .NOTPARALLEL
$(NOTPARALLEL):

srcdir = /usr/local/src/gcc-3.2-1

prefix = /usr/local
exec_prefix = --with-system-zlib
=========^^^^^^^^^^^^^^^^^^^
bindir = /usr/local/bin/i686-pc-cygwin


#3: The run with parameters described at the top, manually corrected:
# as minimizing scatter in file system caches.
NOTPARALLEL = .NOTPARALLEL
$(NOTPARALLEL):

srcdir = /usr/test/gcc-3.2/src

prefix = /usr/local
# exec_prefix = i686-pc-cygwin
==========^^^^^^^^^^^^^^^^^^^
exec_prefix = /usr/local # as corrected!

bindir = /usr/local/bin/i686-pc-cygwin

=========================================================================
IN THE VERY FIRST COMPILER TEST, I get these:

ignoring nonexistent directory "../../include/w32api"
^^^^^^^^^^^^^^^^^^^^^^^
GNU CPP version 3.2 20020912 (prerelease) (cpplib) (80386, BSD syntax)
GNU C version 3.2 20020912 (prerelease) (i686-pc-cygwin)
compiled by GNU C version 3.2 20020912 (prerelease).
ignoring nonexistent directory "/usr/local/gcc-3_2/i686-pc-cygwin/include"
^^^^^^^^^^^^^^^^^^^^^^^^^^
ignoring nonexistent directory "/usr/local/gcc-3_2/i686-pc-cygwin/include"
ignoring duplicate directory "/usr/local/gcc-3_2/lib/gcc-lib/i686-pc-cygwin/3.2/include"
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
#include "..." search starts here:
#include <...> search starts here:
/usr/include/w32api

QUESTION: Are the duplicate / nonexistent directories anything I should care about?
==========================================================================
NEXT SURPRISE:
checking for asprintf... no
checking for atexit... yes
checking for basename... no
===================^^^^^^^^^
but
g:\HOME\Superbiskit>which basename
which basename
/usr/bin/basename

g:\HOME\Superbiskit>
==========================================================================
checking for insque... no
. . . .
checking for mkstemps... no

DOES IT MATTER? Would they be fairly simple?
==========================================================================
checking for vfork.h... no
checking for working vfork... yes

SEEMS STRANGE to have one and not the other. DOES IT MATTER?
==========================================================================
checking for working mmap... no

I THOUGHT WE HAD THAT. Or does it come with a working cygdaemon?
==========================================================================
checking if /usr/local/gcc-3_2/bin//gcc.exe supports -c -o file.o... no
RATHER A SURPRISE?

checking if we can lock with hard links... yes
MOST OF MY FS IS FAT32. Trying to do anything with a hard link is asking for trouble. And making a duplicate then calling it a link is not very good for locking purposes.
==========================================================================
checking if libtool supports shared libraries... yes
checking if package supports dlls... no
checking whether to build shared libraries... no

QUESTION: Does this mean that the gcc package currently isn't set up to use DLL's? WHAT'S THE IMPACT? I know the Cygwin environment is able to build DLL's.

==========================================================================
checking for gnatbind... gnatbind
checking for compiler driver that understands Ada... /usr/local/gcc-3_2/bin//gcc.exe

THAT'S THE GOOD NEWS, Thanks, David!
==========================================================================
checking whether basename is declared... yes
=========================================^^^^ But see above 'checking for basename'

checking whether getopt is declared... no
=======================================^^^^ We do have one, don't we?

checking whether clock is declared... yes
==========================================================================


--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace I am a Christian man, by my actions a great sinner." -- The Way of a Pilgrim; R. M. French, tr.
Life is too short to tolerate crappy software.
.



--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Bug reporting: http://cygwin.com/bugs.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]