scripts on AMD64 systems
running GNU/Linux.
Thanks for reporting this. I have installed the XMKMF patch suggested
by Pavel Roskin in
http://lists.gnu.org/archive/html/autoconf-patches/2005-08/msg00078.html.
Does this fix the problem? (Sorry, I'm not an expert on these systems
yet, so I have
in configure scripts on AMD64 systems
running GNU/Linux.
Thanks for reporting this. I have installed the XMKMF patch suggested
by Pavel Roskin in
http://lists.gnu.org/archive/html/autoconf-patches/2005-08/msg00078.html.
Does this fix the problem? (Sorry, I'm not an expert on these systems
yet, so I
On Fri, 2005-08-26 at 12:42 -0700, Paul Eggert wrote:
Pavel Roskin [EMAIL PROTECTED] writes:
* lib/autoconf/libs.m4 (_AC_PATH_X_XMKMF): Use shell variable
XMKMF to locate xmkmf. Make XMKMF precious. Export CC when
running xmkmf.
Thanks; I installed that. Perhaps it needs
list. It appears that the
traffic is not very heavy. Besides, I have a Spamcop subscription, so I
can reroute spam to my account and report it.
--
Regards,
Pavel Roskin
.
--
Regards,
Pavel Roskin
On Tue, 11 Nov 2003, Alexandre Duret-Lutz wrote:
Paul == Paul Eggert [EMAIL PROTECTED] writes:
Paul Pavel Roskin [EMAIL PROTECTED] writes:
I've seen an announcement in [EMAIL PROTECTED] that Automake
1.7.9 has been released with Autoconf 2.59
OK, I give up. I used
was extremely tolerant to incorrect macro
invocations, but waiting longer endangers the portability of your
package and only delays the conversation to newer Autoconf versions.
You probably mean conversion, not conversation :-)
--
Regards,
Pavel Roskin
, if I understand correctly),
then it's absolutely reasonable to require the developer to spend some
time and make sure that the configure script provides the same level of
compatibility.
--
Regards,
Pavel Roskin
Hi, Dan!
On Thu, 7 Mar 2002, Dan Kegel wrote:
Dumb question: what do I put in configure.in to
make sure nobody accidentally uses autoconf2.13 on it?
I think there's a way to specify minimum version of autoconf
required, but I'm too dumb to find it...
AC_PREREQ(2.50)
--
Regards,
Pavel
macros more that the ones from Autoconf.
Maybe the GNU Coding Standards have changed so that spaces are not allowed
in the project names? But how about non-GNU projects using Autoconf and
Automake?
--
Regards,
Pavel Roskin
(i.e., Autotest
scripts, M4sh-AdHoC scripts and so on).
Fine with me as long as the testsuite doesn't break. I understand that it
doesn't.
--
Regards,
Pavel Roskin
was using some mix of Autoconf and Linux-style
manual configuration. The homepage of GGI is http://www.ggi-project.org/
--
Regards,
Pavel Roskin
-2.52.
--
Regards,
Pavel Roskin
) and
play with it when you want to, not when the testsuite fails.
--
Regards,
Pavel Roskin
$ autoconf
configure.ac:3: warning: syntax
$ autoconf
$ rm -rf autom4te.cache
$ autoconf
configure.ac:3: warning: syntax
$
Removing autom4te.cache between autoconf invocations fixes the test, but I
believe that the actual problem is in autoconf, not in the test.
--
Regards,
Pavel Roskin
That would be a very serious bug in Autoconf, but I don't see
_POSIX_V6_LP64_OFF64 anywhere in the Autoconf sources.
Could you please find out where _POSIX_V6_LP64_OFF64 comes from and how
it is used?
--
Regards,
Pavel Roskin
combination of rules if you
don't want to filter out personal e-mails, only the list,
:0:
* ^Subject: *\(antigen|scanmail)\
* ^List-Post: mailto:[EMAIL PROTECTED]
/dev/null
:0 H B:
* ^Content-Type: application/mixed
* ^List-Post: mailto:[EMAIL PROTECTED]
/dev/null
--
Regards,
Pavel Roskin
.
Try persuading [EMAIL PROTECTED] or [EMAIL PROTECTED] I already
tried it without success.
--
Regards,
Pavel Roskin
to be the reason
why it is used so often.
I'm not aware of any other problems with the CVS version of Autoconf that
could prevent releasing the next stable version.
--
Regards,
Pavel Roskin
I'm missing something obvious or the current scheme with AU_DEFUN
should be reworked. Akim?
--
Regards,
Pavel Roskin
dependencies.
We cannot switch from preprocessing to compilation without seriously
breaking backward compatibility.
--
Regards,
Pavel Roskin
question.
I personally don't want to risk, but if you think that you can make
Automake work with the big macro (maybe by releasing another
Automake-1.4-pX) without spending months on it then I'll support merging
the macros now.
--
Regards,
Pavel Roskin
, or AC_PROG_CPP called after AC_LANG(none).
--
Regards,
Pavel Roskin
it. If you
are committing or adding files from the current directory and from other
directories in the same command always make sure that all the files have
been processed.
--
Regards,
Pavel Roskin
, but a broken
cc is tolerable (eg. to process *.S - *.o)
If a broken cc is tolerable then you should have a mechanism to instruct
configure not to abort. This is quite orthogonal to the preprocessor test.
--
Regards,
Pavel Roskin
earlier or later anyways.
--
Regards,
Pavel Roskin
the remaining bugs by hand?
--
Regards,
Pavel Roskin
something
like set | sort environment.TEST_NUMBER
Maybe the solution is in between: adding a new test which runs all the
macros, and preserve the cache.
Test for working cache is important, but running all macros would be an
overkill in my opinion.
--
Regards,
Pavel Roskin
?) hackerish approach.
Since Autoconf-2.50 has been released, it would be fair to drop support
for Autoconf-2.13 and use new facilities for attaching dependency code to
AC_PROG_CC (provided that they exist and I understand the problem
correctly).
--
Regards,
Pavel Roskin
Regards,
Pavel Roskin
you and everybody on this list!
--
Regards,
Pavel Roskin
PROTECTED]
--
Regards,
Pavel Roskin
, or `m4_ifset', as in
_AC_INIT_CONFIG_LOG
--
Regards,
Pavel Roskin
existed it wouldn't be fixed in 2 years?!
Regards,
Pavel Roskin
ss for you now.
I hope you don't object that I'm moving the discussion to
[EMAIL PROTECTED] This is an important issue and it's not the first time
that we have problems with Autoconf not supplying full path when it's
needed.
Regards,
Pavel Roskin
-stder1
It has been fixed since then, but that was a message on
[EMAIL PROTECTED] that the above warning still appears on Tru64 v5.1.
Please test CVS Autoconf on your system and try to understand what's
happenning if the problem persists.
Regards,
Pavel Roskin
Hello, Akim!
* tests/atgeneral.m4 (AT_INIT): Register at-setup-line and
at-check-line for removal.
Check for the presence of at-check-line only when $at_test was
really a test.
Reported by Pavel.
Thanks!
Go ahead.
Regards,
Pavel Roskin
I've done some search in CVS history. The problem was introduced by the
following change (committed 2000/11/23 09:47:45)
Did you do this by hand?
Shame on me, it wasn't a full-blown script. I used to have an iterator for
CVS, but I lost it since then.
Regards,
Pavel Roskin
ck is useless to find problems in the tests other than the first
one, since a failure to create at-check-line would be masked by the old
at-check-line left from another test.
This test is also useless after banner-* tests. Actually, it's where it's
failing - in banner-1 and banner-2.
Regards,
Pavel Roskin
looking
one) nor the hang with pdksh/zsh are fixed by it.
As a final note, the file BUGS should not say "Do not use." in the
release. Just a reminder, don't take it as offence :-)
Regards,
Pavel Roskin
t.])
fi
I haven't yet figured out why AC_MSG_ERROR doesn't work as expected at
this point.
Important questions:
1) Do we want it to be in AS_SHELL_SANITIZE?
2) Should we check the other bugs, such as string size limit?
Regards,
Pavel Roskin
, ash-0.2, ash-0.3.x, pdksh and
zsh), at least on some *BSD and maybe on Cygwin (but I trashed BIOS on my
Windows machine, so don't expect the later from me).
I'll report the results of my testing tonight.
Regards,
Pavel Roskin
The issues with ash-0.2 can be worked around by creating temporary files,
but I doubt whether we should do it.
Regards,
Pavel Roskin
in Autoconf. User-supplied macros are
not affected either since they don't use conf$$.
Feel free to use conf$$.
Regards,
Pavel Roskin
n:/sbin/cc
+ continue
+ ac_ct_CC=
+ test -n
+ echo configure:1095: result: no
+ echo no
The only change since yesterday was:
2001-02-23 Jim Meyering [EMAIL PROTECTED]
* acgeneral.m4 (_AC_INIT_PREPARE_FS_SEPARATORS): Fix typo: s/;/:/
Regards,
Pavel Roskin
CFLAGS="$CFLAGS -fnord"
LIBS="$LIBS -lol"
$CC $CFLAGS $LIBS -o conf$$ conf$$.c
./conf$$ conf$$.val
rm -f conf$$ conf$$.c
)
ac_result=`cat conf$$.val`
rm -f conf$$.val
Regards,
Pavel Roskin
for "after 2.50".
2001-02-20 Lars J. Aas [EMAIL PROTECTED]
* acgeneral.m4 (_AC_OUTPUT_CONFIG_STATUS): use `AS_EXIT(0)' instead
of `exit 0' when exiting config.status.
Start entries with upper case - Use ...
Everything else is Ok.
Regards,
Pavel Roskin
This is for 2.51 or more. We do have a problem with AC_CHECK_PROG and
relatives, that's for sure. But let's forget about this issue now.
Please, install your proposals etc. in TODO, but focus on releasing
Autoconf.
Ok. I like this approach.
Regards,
Pavel Roskin
to [EMAIL PROTECTED]
It would be great if you could expain what exactly is happening for you.
Regards,
Pavel Roskin
.
If for any reason this can happen without explicit user intervention, it's
a bug in Autoconf (or Automake, if it puts a wrong shell to makefiles).
Regards,
Pavel Roskin
On 16 Feb 2001, Alexandre Oliva wrote:
On Feb 16, 2001, Pavel Roskin [EMAIL PROTECTED] wrote:
I see. But we don't need reentrancy for configure, do we?
Nope. Not even for config.status, AFAIK. We just need concurrency,
not reentrancy. Unless a new requirement has entered the picture
iple macro definitions are
handled.
Regards,
Pavel Roskin
to AT_DATA doesn't end with a newline, then _ATEOF
is glued with the last line.
I believe AT_DATA needs a sanity check.
Regards,
Pavel Roskin
don't want to interrupt my efforts until I succeed or
give up :-)
Regards,
Pavel Roskin
subst([[$1]], [^\[\[\(.*\)\]\]$], [[\1]])])
_
Regards,
Pavel Roskin
it
somewhere. But again, this seems to be an overkill.
Regards,
Pavel Roskin
, as_unset etc
into config.status,
In this case it's easy to support weird things when you stop supporting
other weird things (I mean, reentrancy vs. different shell). But you'll
have to persuade me that your weird thing is more important :-)
Regards,
Pavel Roskin
safe to remove "dnl".
Regards,
Pavel Roskin
on for some projects in a couple of years.
So it's better to be prepared.
Regards,
Pavel Roskin
,
Pavel Roskin
whatever we can.
If your are not sure, don't touch bugs, someone else will take care of
them. And please don't close the bugs that you just don't want to fix.
Either suspend them or leave a comment and move to "analyzed".
Regards,
Pavel Roskin
onftest.sh is clearly better than my conftest.bat, as well as its
contents :-)
Regards,
Pavel Roskin
/m4/whatever).
Since the test is run by bash and not by something else it may be very
tricky to determine the path separator unless you have perl/m4/whatever.
Well, maybe if you try ";" first it should work.
Regards,
Pavel Roskin
ncerned about one project, its development team and its
area of applicability. For example, you can define a variable to "VPATH"
and AC_SUBST it, you can require GNU make etc.
It's all luxury that Autoconf and Automake cannot afford.
Regards,
Pavel Roskin
e Autoconf is our moral obligation before the community -
years of bugfixing should crystallize to a new stable version.
I'm sorry, Akim, that I have to write it, but I feel that we are doing
something wrong.
Regards,
Pavel Roskin
I plan to move the Autoconf package to Automake 1.4a. Any problem
with that? I'll also provide patches to adjust Automake to
configure.ac, and likewise for Libtool.
Why not Automake 1.4b?
ftp://sourceware.cygnus.com/pub/automake/automake-1.4b.tar.gz
Regards,
Pavel Roskin
some new features in
Autoconf. It doesn't mean that every project using the latest Autoconf
(CVS or otherwise) will be required to use an unreleased version of
Automake.
Regards,
Pavel Roskin
.
Regards,
Pavel Roskin
On Sat, 23 Dec 2000, Motoyuki Kasahara wrote:
Hi,
I found a typo in `autoreconf.sh' ($localddir - $localdir).
Fix applied.
Thanks!
Regards,
Pavel Roskin
developed.
Is that - define - going to be dropped too, btw?
Not in 2.50. And only for the sources that indicate (by AC_REQUIRE or
something like that) that they want to work in the new namespace.
Regards,
Pavel Roskin
versions of Autoconf.
Regards,
Pavel Roskin
should be treated in
the same way.
I withdraw my comment that asked for the opposite.
Regards,
Pavel Roskin
"
ITEM NAME="tests/Makefile" REQUIRED="no"
/AC_CONFIG FILES
AC_OUTPUT
Well, enough for today :-)
Regards,
Pavel Roskin
ow to write portable makefiles, so it belongs to
make-stds.texi or maybe also to the Automake manual.
Regards,
Pavel Roskin
PROTECTED]
Regards,
Pavel Roskin
oblems with OpenLDAP without playing with Mac OSX.
Regards,
Pavel Roskin
or should it be fixed?
I guess few simple C++ tests wouldn't hurt, as well as a test for language
selection with AC_LANG.
Regards,
Pavel Roskin
right now by hand.
Regards,
Pavel Roskin
the version number?
I already suggested it, but with your arguments I feel much better about
doing so.
Regards,
Pavel Roskin
be AM_MAINTAINER_MODE in configure.in and @MAINT@
before the explicit rule, but I believe it's an overkill.
Regards,
Pavel Roskin
k. If you insist on having the Systemology section I'll try to put more
relevant content there. I'm updating my .plan
Regards,
Pavel Roskin
autoconf's way to deal with static OS'es. The
autoconf's way is to deal with known problems in possibly unknown
combinations :-)
Regards,
Pavel Roskin
around those issues with QNX 4.25 somehow.
How about this:
ac_output=`command || echo "===ERROR==="`
case $ac_output in
*===ERROR===)
AC_MSG_ERROR([command failed]) ;;
*) ;;
esac
Regards,
Pavel Roskin
Hello, Harlan!
The problem went away after I flushed my build direcory and re-ran
configure.
Congratulations!
It's always a good idea to start from scratch.
I wrongly assumed that you knew it, so it's my fault :-)
Regards,
Pavel Roskin
send us acversion.m4 and acversion.m4.in?
Regards,
Pavel Roskin
tatus, e.g.
ac_file_inputs=`IFS=:
for f in $ac_file_in; do
...
esac
done` || { false; exit; }
I'm afraid that Autoconf will have to reject QNX's shell. It's hard to
work around this kind of shell braindamage :-(
Regards,
Pavel Roskin
in autoupdate, but
now you found a separate problem.
Regards,
Pavel Roskin
+ for loop.
Anybody knows how to rewrite this:
ac_output=`command`; ac_status=$?
in a way that it works even on QNX but doesn't cause significant slowdown?
Regards,
Pavel Roskin
The same. If it's too big send the last 50 lines.
Also please attach config.log
Regards,
Pavel Roskin
that builds all that stuff
automatically, without requiring your intervention when new options are
added.
Your message remains in the archives and will be used some day. Thank you
for your efforts.
Regards,
Pavel Roskin
ffort. You're welcome. - Bruce
I meant you, Bruce. But my point was - we should have templates in CVS and
everything should be regenerated from them automatically whenever anybody
of us changes the template.
Regards,
Pavel Roskin
ntheses. I
don't like either of them.
2) "exit 77" is used in other testsuites, not only in Autoconf. This is
probably not a very strong argument, but why should Autoconf decide what
other testsuites should ignore?
Regards,
Pavel Roskin
ith 0 arguments don't have parentheses, so it
shouldn't be a problem.
I'll have a look whether it's possible to implement AU_ALIAS in such way.
Regards,
Pavel Roskin
an attempt
to accomodate the testsuite at the expence of the users.
Regards,
Pavel Roskin
r package should be the right way. I.e., AC for
autoconf, AM for automake, LT for libtool. It should be easy to rename
A[HSTU]_* to something else before 2.50. It will be harder to do it later.
Regards,
Pavel Roskin
++ background who could rewrite
AC_LANG_SOURCE(C++) and test it on many compilers.
I hope that you will find time to create a well tested patch and submit it
to [EMAIL PROTECTED]
Regards,
Pavel Roskin
appears to be in the definition of AU_DEFUN. It adds
parentheses unconditionally. Maybe we need another AU_DEFUN, without
parentheses?
Regards,
Pavel Roskin
Hello, Ralf!
# autoconf --add-missing
autoconf: option `--add-missing' requires an argument
It's my bug. I'll fix it. Thank you for the report.
Regards,
Pavel Roskin
mv $@.tmp $@
I did use "make dist" to create the tarball - it doesn't help.
Regards,
Pavel Roskin
: undefined macro: lt_AR_FLAGS
configure:5870: error: undefined macro: AR_FLAGS
What should we do?
Return to A[CHMSTU]
Regards,
Pavel Roskin
1 - 100 of 186 matches
Mail list logo
