Fix maintainer-check failure

Applied to master and branch-1-10. 2008-01-30 Ralf Wildenhues <[EMAIL PROTECTED]> * tests/man3.test: Avoid `make VAR=value'. diff --git a/tests/man3.test b/tests/man3.test index 2527e9c..fe9394d 100755 --- a/tests/man3.test +++ b/tests/man3.test @@ -41,4 +41,4 @@ $ACLOCAL

Re: Logic-o in automake manual

Hello Reuben, * Reuben Thomas wrote on Mon, Feb 04, 2008 at 12:50:45PM CET: > You mean "appear newer", not "appear older", as "configure.ac" is > lexically after "configure". If configure.ac "appeared older" than > configure, then configure would not be rebuilt, spuriously or otherwise. > > By t

Re: Config.guess support for i586-pc-winnt

Hello Markus, * Duft Markus wrote on Mon, Feb 11, 2008 at 04:06:47PM CET: > (where does config.guess really come from? I know there is a gnuconfig > package at gentoo). [...] > Sorry if this is not the correct place for this, maybe you can point me > there if it's not... That's really weird that

Re: 'nonemtpy' typo in lib/am/tags.am

Hi Colin, * Colin Watson wrote on Sun, Feb 17, 2008 at 05:58:58PM CET: > The attached patch fixes a typo I happened to notice while upgrading > man-db to Automake 1.10.1. Thanks! Applied like this, master and branch-1-10. Cheers, Ralf 2008-02-17 Colin Watson <[EMAIL PROTECTED]> (tiny change

Re: AM_INIT_AUTOMAKE uses a deprecated macro

test and greatest" tools. Has this been updated? If so, how/where > do I get the update? Fixed thusly, master and branch-1-10. Thanks for the report. Cheers, Ralf 2008-02-19 Ralf Wildenhues <[EMAIL PROTECTED]> PR automake/498 * m4/options.m4 (_AM_SET_OPTION

Re: XFAIL and make

27;t with this shell. Please try the patch (against master) below. You need to re-bootstrap to propagate the change into tests/Makefile.in. 3) Is this failure exposed by the check[67].test tests? If not, then we need to add a testcase. Thanks for the report. Ralf 2008-02-22 Ralf Wildenhu

Re: XFAIL and make

* Ralf Wildenhues wrote on Fri, Feb 22, 2008 at 06:47:54PM CET: > > 2) NetBSD ksh has a bug. At least I *think* that is a bug, I haven't > checked SUSv3 yet. This: > foo=foo; t='[ ]'; case " foo bar " in *$t$foo$t*) echo yes;; esac > > should

Re: Tidy-up of node "Why doesn't automake support wildcards?"

Hi Reuben, * Reuben Thomas wrote on Wed, Feb 27, 2008 at 12:49:24AM CET: > 2008-02-23 Reuben Thomas <[EMAIL PROTECTED]> > > * doc/automake.texi: Improve "Why doesn't Automake support > wildcards" node's English and sense. Applied. Thanks! Cheers, Ralf

Re: XFAIL and make

Hi Patrick, * Patrick Welche wrote on Tue, Feb 26, 2008 at 02:05:59PM CET: > On Sat, Feb 23, 2008 at 11:06:52AM +0100, Ralf Wildenhues wrote: > > It seems NetBSD ksh ( == @(#)PD KSH v5.2.14 99/07/13.2) has lots of bugs. > (I should have remembered make uses sh which works, but gma

instsh2.test failure on Solaris (was: Please report to [EMAIL PROTECTED])

Hello Younes, * Younes Younes wrote on Mon, Mar 03, 2008 at 11:21:29AM CET: > > the test file is instsh2.test > > hier is the report Thanks. [...] > /home/younes/built/automake-1.10.1/tests:/usr/local/bin:/usr/ccs/bin:/usr/ucb:/usr/bin:/usr/sfw/bin:/opt/csw/bin:/usr/openwin/bin:/usr/lang/bin:/op

Re: Implementing notrans_man_MANS

Hello Peter, * Peter Breitenlohner wrote on Wed, Feb 20, 2008 at 02:20:04PM CET: > > attached are five patches (the diffs are actually against the result of > applying your patch for PR 516 from 2008-01-22): Thanks for your work in this, and sorry for the delay. I saw that your paperwork is now

Re: Implementing notrans_man_MANS

Hi Peter, * Peter Breitenlohner wrote on Thu, Mar 06, 2008 at 10:43:20AM CET: > On Wed, 5 Mar 2008, Ralf Wildenhues wrote: > >> This triple loop is already at 72 iterations. Not good for performance. >> If this grows further, we may have to access variables in a different >

Re: Implementing notrans_man_MANS

* Peter Breitenlohner wrote on Thu, Mar 06, 2008 at 01:27:05PM CET: > On Thu, 6 Mar 2008, Ralf Wildenhues wrote: > >>>> Also, the whole thing would probably be a lot more readable if the >>>> ?NOTR?/?DOTR? prefixes vanished and were replaced by full rules, wrap

[PATCH] Add support for newer Fortran extensions '.[fF]0[38]'.

* automake.in (scan_autoconf_traces): Accept .f03, .f08 for Fortran, .F03 and .F08 for preprocessed Fortran files. Use $(FCFLAGS_f[09]x) accordingly. * doc/automake.texi (Compiling Fortran 9x Files): Update. * tests/fort2.test: Add more tests. * NEWS: Update. --- I noticed that GCC now also suppor

drop usage of obsolete AC_OBJEXT and AC_EXEEXT

Applied to master and branch-1-10. Cheers, Ralf 2008-03-08 Ralf Wildenhues <[EMAIL PROTECTED]> Drop usage of obsolete macros AC_EXEEXT and AC_OBJEXT. * tests/ansi4.test: Drop AC_OBJEXT and AC_EXEEXT. * tests/ansi6.test: Likewise. * tests/ansi7.test: Li

FYI: master now depends on unreleased Autoconf

early and provide a sensible error message. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> diff --git a/configure.ac b/configure.ac index 11e989e..5198d9e 100644 --- a/configure.ac +++ b/configure.ac @@ -1,7 +1,7 @@ # Process this file with autoconf to produce a configure script

Re: fast make install

* Ralf Wildenhues wrote on Thu, Jan 24, 2008 at 12:13:13AM CET: > Here's the proposed patch (generated files not shown), still missing > Autoconf version requirement. It does the fast install for HEADERS, > DATA, PYTHON, LIBRARIES, and TEXINFOS outputs. > > Comments, re

Re: Implementing notrans_man_MANS

Hello Peter, * Peter Breitenlohner wrote on Mon, Mar 10, 2008 at 01:26:08PM CET: > > [...] I can only use if/endif if/endif and not if/else/endif > because a section may have trans_ as well as notrans_ man pages. Ah yes, sure. > Attached are revised versions of the three patches. I hope they tak

Re: fast make install

Hello Markus, First, please don't top-post, thank you. * Duft Markus wrote on Mon, Mar 10, 2008 at 08:59:52AM CET: > Ralf Wildenhues <> wrote: > > Applied as shown in > > > <http://thread.gmane.org/gmane.comp.sysutils.automake.patches/2945/focus=3088>, >

Re: Autoconf's copy of ChannelDefs.pm

Hi Stepan, * Stepan Kasal wrote on Fri, Mar 14, 2008 at 03:20:53PM CET: > several of the lib/Automake/*.pm files are used by Autoconf as > lib/Autom4te/*.pm. Why isn't ChannelDefs.pm among them. No idea. I can only guess that it's omission or due to -Wcross. > The only question is about -Wcr

fix AC_CONFIG_LINKS([X:X]) (was: New module gnu-make)

hmm, this was long ago: * Ralf Wildenhues wrote on Fri, Nov 16, 2007 at 05:52:54PM CET: > > Current Autoconf allows you to do this: > AC_CONFIG_LINKS([GNUmakefile:GNUmakefile]) > > the only bit that's missing is that automake shouldn't remove > the original

conditional AC_CONFIG_FILES: AM_COND_IF.

Proposed patch for conditional AC_CONFIG_FILES. This enables to fix one half of (the other half is the FIXME in the test below.) Comments, reviews appreciated. autoconf/lib/autom4te.cfg should preselect _AM_COND_{IF,ELSE,ENDIF} if this go

Re: conditional AC_CONFIG_FILES: AM_COND_IF.

* Ralf Wildenhues wrote on Sun, Mar 16, 2008 at 10:26:54PM CET: > Proposed patch for conditional AC_CONFIG_FILES. This enables to fix one > half of <http://gcc.gnu.org/ml/gcc-patches/2008-03/msg00941.html> > (the other half is the FIXME in the test below.) > > Comments

Re: conditional AC_CONFIG_FILES:AM_COND_IF.

Hi Eric, On Fri, March 21, 2008 14:02, Eric Blake wrote: > Ralf Wildenhues gmx.de> writes: > >> Comments, reviews appreciated. > > Looks like a good idea. Sorry for the delayed review. Hey, it's great to have review at all, no matter how delayed. &g

Re: conditional AC_CONFIG_FILES:AM_COND_IF.

> On Fri, March 21, 2008 14:02, Eric Blake wrote: > > Ralf Wildenhues gmx.de> writes: > >> + _AM_COND_IF => 1, > >> + _AM_COND_ELSE => 1, > >> + _AM_COND_ENDIF => 1, > > > > Based on your question about tr

Re: conditional AC_CONFIG_FILES:AM_COND_IF.

* Eric Blake wrote on Wed, Mar 26, 2008 at 02:30:14AM CET: > > Looks okay from my perspective. Thanks, applied. > | [EMAIL PROTECTED] macros may be nested. > > Maybe add a line here about needing proper quoting to do nesting? Done like this: @code{AM_COND_IF} macros may be nested when m4 quotat

Re: gnupload tweak

Hi Eric, * Eric Blake wrote on Wed, Apr 02, 2008 at 09:47:12PM CEST: > Is this okay to apply, so that my VPATH invocation will actually work > with the ftp-upload.gnu.org directive files? Sure, thanks! Cheers, Ralf > * lib/gnupload: Allow uploads from separate directory.

Re: AC_AUTOCONF_VERSION quoting

Hi Eric, * Eric Blake wrote on Thu, Apr 10, 2008 at 05:59:13PM CEST: > Automake should set a good example of proper macro quoting. Expanding > AC_AUTOCONF_VERSION is dangerous, since snapshot versions of autoconf include > letters in the version string, which in turn could expand as inadvertent

Re: support of the cegcc compiler

Hello Vincent, * Vincent Torri wrote on Sat, Apr 12, 2008 at 10:22:22AM CEST: > > right now, automake supports only the mingw32ce compiler that the cegcc > project provides. That project also provides the cegcc compiler, which is > not supported right now. > > I've attached a patch that add ceg

Re: [PATCH] eliminate some installation overhead

Hi Jim, * Jim Meyering wrote on Mon, May 26, 2008 at 05:05:01PM CEST: > > Since part of coreutils' install program (currently ifdef'd out) incurs > a performance hit due to an SELinux implementation problem (I hear that > a solution is in the works), I'm motivated to continue Ralf's work of > min

Re: fix AC_CONFIG_LINKS([X:X]) (was: New module gnu-make)

<http://lists.gnu.org/archive/html/automake-patches/2008-03/msg00018.html> * Ralf Wildenhues wrote on Sat, Mar 15, 2008 at 06:28:40PM CET: > > > > Current Autoconf allows you to do this: > > AC_CONFIG_LINKS([GNUmakefile:GNUmakefile]) > > > > the only

Re: conditional AC_CONFIG_FILES:AM_COND_IF.

* Ralf Wildenhues wrote on Wed, Mar 26, 2008 at 12:14:00AM CET: > > I think I have a followup patch to address the FIXME. This is to fix the other half of <http://gcc.gnu.org/ml/gcc-patches/2008-03/msg00941.html> <http://thread.gmane.org/gmane.comp.sysutils.automake.patches/3129&

am__install_max

Let's factorize the arbitrary constant. Applied to master. Cheers, Ralf * lib/am/inst-vars.am (am__install_max): New variable. (am__nobase_list): Use it. diff --git a/lib/am/inst-vars.am b/lib/am/inst-vars.am index acdfc9a..8f89009 100644 --- a/lib/am/inst-vars.am +++ b/lib/am/in

Re: [trunk][patch] fix --enable-shared=some_libraries

Hello, * Rafael Espindola wrote on Wed, May 14, 2008 at 07:18:52PM CEST: > 2008/5/14 Paolo Bonzini <[EMAIL PROTECTED]>: > > Ok for trunk and 4.3, but please also handle enable_static in the > > same way. The patch is preapproved with this change, but please repost > > it so that it can also go in

facilitate 'libtool --dry-run': new macro $(am__mv)

Before depend2.am was optimized to use several command lines (in some cases) for compilation (d505ab9091905a56e01c5e13d9e2d55b50420691), make LIBTOOLFLAGS=--dry-run used to work reasonably well, actually because the exit status of the mv command was ignored with most make implementations. Now t

Re: Python 3.0 support

Hi Johan, * Johan Dahlin wrote on Sun, Jul 27, 2008 at 03:53:58PM CEST: > Hi, here are 4 patches adding Python 3.0 support for Automake. Thank you for working on these patches. > Python 3.0 changes has major backwards incompatible changes, we have to > be rather careful to use a subset which wo

Fix maintainer-check failures.

Applied to master and (the tests part) branch-1-10. Cheers, Ralf Fix maintainer-check failures. * doc/automake.texi (Conditionals): Drop TAB. * tests/missing6.test: Use 'rm -f'. diff --git a/doc/automake.texi b/doc/automake.texi index 5cb5642..eccbc5d 100644 --- a/doc/aut

Use multiple files in notrans.test.

Applied to master. This helps finding bugs in multi-file install. Cheers, Ralf 2008-08-19 Ralf Wildenhues <[EMAIL PROTECTED]> * tests/notrans.test: Use multiple man files of each type. diff --git a/tests/notrans.test b/tests/notrans.test index 6d73547..bf7ad7f 100755 --- a

Re: patch: Makefile.in in subdir not generated

he common cases (updated Makefile.am, or removed Makefile.in), and that automake is never called more than once for each case. The second case ensures that 'make' fails when any rebuilding rule would fail. I'm still pondering whether to put this in branch-1-10 as well. Cheers, and

Re: vanity patch

Hi Ilya, * Ilya N. Golubev wrote on Thu, Aug 28, 2008 at 05:15:04PM CEST: > > put you in THANKS > > Looks like that file is in `utf-8', not restricted to ascii or iso > 8859-1, unlike name that is normally put in `From:' field of my > messages (including this one) to english speaking lists, fora.

Re: duplicate automake runs? [patch: Makefile.in in subdir not generated]

Hello Ilya, * Ilya N. Golubev wrote on Thu, Aug 28, 2008 at 05:40:03PM CEST: > > causes duplicate automake reruns in the common case > > where toplevel Makefile.am is updated. > > Explicitly tried avoid that, and stated that in change log entry: > > > Added elimination of > > duplicate `$(AUTOMA

Re: patch: Makefile.in in subdir not generated

* Ralf Wildenhues wrote on Fri, Aug 22, 2008 at 02:57:50PM CEST: > I'm still pondering whether to put this in branch-1-10 as well. I think this is safe enough, I've pulled that to branch-1-10 (including the THANKS name fix, of course). Cheers, Ralf > 2008-08-22 Ralf Wil

Avoid getting bitten by CDPATH

m: Likewise. * lib/am/subdirs.am: Likewise. * lib/am/tags.am: Likewise. * lib/am/texibuild.am: Likewise. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> diff --git a/HACKING b/HACKING index 3be23bc..5e7191f 100644 --- a/HACKING +++ b/HACKING @@ -64,6 +64,9 @

[PATCH] Implement AM_MAINTAINER_MODE([DEFAULT-MODE]).

e. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> --- [ moving from bug-automake ] * Vincent Lefevre wrote on Sun, Aug 31, 2008 at 01:54:23PM CEST: > On 2008-08-31 12:19:54 +0200, Ralf Wildenhues wrote: > > * Vincent Lefevre wrote on Sun, Aug 31, 2008 at 12:11:04PM CEST: >

Re: [PATCH] Implement AM_MAINTAINER_MODE([DEFAULT-MODE]).

Hi Eric, * Eric Blake wrote on Mon, Sep 01, 2008 at 04:43:08AM CEST: > Ralf Wildenhues gmx.de> writes: > > > Somebody please check what I'm currently not seeing in the m4 fu I > > tried (the test fails, because there is something wrong in > > m4/maintainer.m4).

Re: [PATCH] Implement AM_MAINTAINER_MODE([DEFAULT-MODE]).

* Vincent Lefevre wrote on Thu, Sep 04, 2008 at 03:23:37AM CEST: > Thanks, it seems to work well. (Now I have another problem because > awk isn't available, while it wasn't required in the past; I'll try > to build it if its build doesn't depend on itself.) Yes, awk is needed more and more in Auto

Re: [PATCH] Implement AM_MAINTAINER_MODE([DEFAULT-MODE]).

* Vincent Lefevre wrote on Thu, Sep 04, 2008 at 02:17:06PM CEST: > On 2008-09-04 08:24:57 +0200, Ralf Wildenhues wrote: > > GNU awk should be cross compilable, but also it has not, as far as I > > know, yet updated to autotools which need an awk to be present. Since > > this

[PATCH] Fix nobase-python.test for BSD make.

* tests/nobase-python.test: Listing a dependency to be found via VPATH search explicitly, without the path, as a target, will turn off VPATH search for FreeBSD make. So move the creation of the files outside of the Makefile. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> --- Applyin

Re: Avoid getting bitten by CDPATH

* Ralf Wildenhues wrote on Sun, Aug 31, 2008 at 09:56:33AM CEST: > I'm not sure whether this patch is (at least partly) overkill, > so I'd be glad about some feedback on this; thanks. FYI, pushed. > For all possibly-relative subdirs, use $(am__cd). > > Up t

[PATCH] Avoid Heisenbug with verbose testing on HP-UX.

* tests/output-order.test: Remove spurious redirection line from output for comparison. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> --- This shell issue is described in info Autoconf "File Descriptors" Applied to HEAD and branch-1-10. Cheers, Ralf ChangeLog

Use Exit, not exit, in test suite

he tests do not use `exit' with an argument, but use `Exit' instead. * tests/defs.in (Exit): New function. Use it throughout, starting with the introduction of the exit trap. * tests/*.test: Use `Exit $arg' instead of `exit $arg' throughout, except insi

Re: Updated Vala support for automake

Hello Mathias, all, let me please say that I'm a bit ashamed that I put this off for so long. Please accept my apologies. I promise to push this through quickly now. But first of all, thanks for your contributions to Automake! So here we go with a review of

patch series: fast install and uninstall

Hello, Continuing the patch series against git master Automake, to fix and improve 'make install' and 'make uninstall' (and 'make clean' a bit). There's about a dozen patches coming up. Jim and Akim pushed towards multi-file 'install' because the GNU coreutils install program is rather heavy, esp

Fix multi-file installation for files in srcdir.

This patch lowers the limit to 40 files installed. I figured 80 bytes for `$(srcdir)/' plus relative file name were reasonable, plus maybe a few hundred for `$(DESTDIR)$(foodir)', which brings us to roughly 3500. Leaves 600 for the environment, to stay below the POSIX minimum limit of 4096. Not m

Re: Fix multi-file installation for files in srcdir.

* Ralf Wildenhues wrote on Sun, Sep 07, 2008 at 12:20:06AM CEST: > This patch lowers the limit to 40 files installed. I figured 80 bytes > for `$(srcdir)/' plus relative file name were reasonable, plus maybe a > few hundred for `$(DESTDIR)$(foodir)', which brings us to roughly

Exit upon `install' failure, part one.

that SCRIPTS aren't installed efficiently yet, it merely slow down the test execution. Cheers, Ralf 2008-09-07 Ralf Wildenhues <[EMAIL PROTECTED]> Exit upon `install' failure, part one. * lib/am/data.am (install-%DIR%%PRIMARY%): Exit if any install

Fail upon install or ranlib failure, part two.

This is the second part of the install failure fixes. Nothing surprising really, but it differs from the first part in a couple of details: * The tests do not try to deal with a huge number of files. Not sure if I want to change this later when all files are installed efficiently, but simply build

Re: automake (GNU automake) 1.10.1 9 of 538 tests failed

Hello John, * Trudeau, John F wrote on Fri, Aug 15, 2008 at 01:23:53AM CEST: > The bash version that was used in the original report was: > > /home/trudeajo$ /bin/bash --version > GNU bash, version 2.05b.0(1)-release (powerpc-ibm-aix5.1) > Copyright (C) 2002 Free Software Foundation, Inc. > Afte

Faster `make uninstall'; fast install for JAVA.

The next in the series. Since uninstall is generally easier to do efficiently than install, let's fix that first. Note that, simply by entering the destination directory, we can avoid all command line length issues. I chose to echo the commands issued in subshells, so that cut and pasting multip

Reliable multi-file install for man pages.

bits. And considerably shorter than its previous version. :-) Cheers, Ralf 2008-09-07 Ralf Wildenhues <[EMAIL PROTECTED]> Jim Meyering <[EMAIL PROTECTED]> Reliable multi-file install for man pages. * lib/am/mans.am: Include inst-vars.am.

Multi-file install for PROGRAMS.

ol install mode. (I guess that's what I'm looking at next after the next Automake release.) Cheers, Ralf 2008-09-07 Ralf Wildenhues <[EMAIL PROTECTED]> Multi-file install for PROGRAMS. * lib/am/progs.am (install-%DIR%PROGRAMS): Allow to install several

Re: automake generates delete targets twice

Hello Toralf, * Ralf Wildenhues wrote on Wed, Jul 30, 2008 at 09:02:01PM CEST: > * Toralf Förster wrote on Wed, Jul 30, 2008 at 09:26:10AM CEST: > > > > regarding to https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=2749 I'm > > wondering whether at UNIX systems aut

New doc section about command line length limits.

Review and comments of whatever sort appreciated. I haven't pushed this yet. Thanks, Ralf New doc section about command line length limits. * doc/automake.texi (Length limitations): New node. (Install): Mention multi-file install. * NEWS: Expand on the issue, list some more user-

Fast install and uninstall for SCRIPTS.

Once DATA and PROGRAMS are done, SCRIPTS follow naturally by mixing the features appropriately: scripts are programs without the libtool hassle, the .exe suffix, and without being present only optionally, but they can live in the source tree, too, so need line length check. Viewed the other way ro

Remove unneeded per-rule *INSTALL variables.

For now, this is the last (proposed) patch of the series. It cleans up the per-rule *INSTALL variables, which are in most cases not needed any more because the rules now ensure that the respective directories exist. These names have never been documented. However, some packages, notably coreutil

Faster install for libtool outputs.

Hmm, apparently I failed to notice that `libtool --mode=install' doesn't mind being invoked for multiple files at once, even with libtool 1.5.x. This makes the rules more consistent with the rest (never mind the apparent increased duplication in ltlib.am). Pushed. Cheers, Ralf Faster install

Fix installation of scripts and programs with destructive $(transform)

Jim notified me about a distcheck failure of coreutils with recent git Automake. It deals with a destructive $(transform) like this: program_transform_name='s/.*/foo/' which happens to let multiple files end up with identical installed names. The failure happens with PROGRAMS and SCRIPTS. The

Re: Faster install for libtool outputs.

* Ralf Wildenhues wrote on Sat, Sep 13, 2008 at 09:46:47AM CEST: > Hmm, apparently I failed to notice that `libtool --mode=install' doesn't > mind being invoked for multiple files at once, even with libtool 1.5.x. > This makes the rules more consistent with the rest (never

Man pages for automake and aclocal.

`.so' man page links to the versioned pages. * NEWS: Update. Signed-off-by: Ralf Wildenhues <[EMAIL PROTECTED]> diff --git a/NEWS b/NEWS index a2b336e..dac6a9e 100644 --- a/NEWS +++ b/NEWS @@ -43,6 +43,8 @@ New in 1.10a: Note that Automake will never overwrite an exist

Re: Faster install for libtool outputs.

Let's start fixing the issues I noted last weekend: * Ralf Wildenhues wrote on Sun, Sep 14, 2008 at 03:23:45PM CEST: > While writing a test case for this, I noticed that nobase_ is broken > for LTLIBRARIES in more ways, and before the install changes: > > * -rpath is not adjuste

Re: Faster install for libtool outputs.

* Ralf Wildenhues wrote on Sun, Sep 14, 2008 at 03:23:45PM CEST: > > Sigh, this was overly optimistic. With libtool libraries, the > eventual need to relink at install time requires that the ordering > is exactly the same as given in the Makefile.am lines. > > This is becau

Re: Faster install for libtool outputs.

* Ralf Wildenhues wrote on Sun, Sep 14, 2008 at 03:23:45PM CEST: > * unlike lib_LTLIBRARIES, nobase_lib_LTLIBRARIES aren't consistently > installed before bin_PROGRAMS. :-/ Yes, this fix is more of an ugly ugly hack, just like the hack done for lib_LTLIBRARIES. I figured two consi

Re: Faster install for libtool outputs.

Hi Jim, * Jim Meyering wrote on Sat, Sep 20, 2008 at 02:02:36PM CEST: > Ralf Wildenhues <[EMAIL PROTECTED]> wrote: > > > > One downside of the large test suite is the time it takes to complete. > > Since there are lots of instances of $sleep in it, I've foun

Re: New doc section about command line length limits.

* Ralf Wildenhues wrote on Sun, Sep 07, 2008 at 01:16:46PM CEST: > Review and comments of whatever sort appreciated. > I haven't pushed this yet. Pushed as below, slightly amended to cater for - libtool multi-file install, - nobase is not equivalent to several vars for LTLIBRARIES, as

Compactify program rules using ternary operators.

I keep forgetting about the ternary operators that Akim introduced a few months ago. Pushed. Cheers, Ralf Compactify program rules using ternary operators. * lib/am/progs.am (install-%DIR%PROGRAMS): Uglify^WCompactify using ternary operators `%KEY?TRUE:FALSE%'. (uninstall-%D

Re: Remove unneeded per-rule *INSTALL variables.

* Ralf Wildenhues wrote on Sun, Sep 07, 2008 at 01:46:15PM CEST: > For now, this is the last (proposed) patch of the series. It cleans up > the per-rule *INSTALL variables, which are in most cases not needed any > more because the rules now ensure that the respective directories exist. &

Re: Man pages for automake and aclocal.

Hi Eric, * Eric Blake wrote on Sat, Sep 20, 2008 at 11:28:47PM CEST: > Ralf Wildenhues gmx.de> writes: > > > > While adding man pages to Libtool, let's do it for Automake, too. > I'm now seeing this trying to bootstrap automake.git on a Solaris 10 box: > >

[PATCH] Fix some comment typos.

Pushed master and branch-1-10. Cheers, Ralf Fix some comment typos. * automake.in: Fix some comment typos. * lib/Automake/Condition.pm: Likewise. * lib/Automake/DisjConditions.pm: Likewise. * lib/Automake/Variable.pm: Likewise. * lib/Automake/tests/DisjConditions.pl:

Re: failure in make check

[ moving from bug-automake ] * Angeline Gail Burrell wrote on Mon, Sep 15, 2008 at 07:55:19AM CEST: > > The attachment make_auto.am has the > stdout from $make check, while make_auto.tests has the stdout from $make > check for tests ausdir2, cond17, lisp3, lisp6, lisp8, and txinfo5 in > verbose

Re: automake (GNU automake) 1.10.1 9 of 538 tests failed

[ moving from bug-automake ] * Ralf Wildenhues wrote on Mon, Sep 29, 2008 at 09:30:28PM CEST: > > The depcomp6.test one looks like bad interaction between AC_PROG_CC_C_O > and AM_PROG_CC_C_O, or a shell bug: although the compiler is deemed > potent by the former, the latter still rew

Re: automake (GNU automake) 1.10.1 9 of 538 tests failed

* Ralf Wildenhues wrote on Sun, Oct 05, 2008 at 09:54:48PM CEST: > > I'm still stumped by the fact that sh on AIX evaluates these lines > wrongly, and haven't gotten down to the bottom of this shell bug. > > Reproduced it though, thus pushed the patch below, including

Re: automake (GNU automake) 1.10.1 9 of 538 tests failed

* Ralf Wildenhues wrote on Sun, Oct 05, 2008 at 09:58:15PM CEST: > - I don't understand how this patch should have any impact on the other > two test failures you have been seeing. I still see them failing. Never mind. They are SKIPped in your rerun; presumably because either li

[PATCH] Improve test coverage of current TESTS semantics.

I noted that the coreutils check.mk script, when used for automake/tests/, would cause no tests failures, which is great. But there is a bug in check.mk wrt. matching of X{FAIL,PASS} and tests in subdirs. (And it outputs the names of the logs instead of the tests.) Also, I noted that @substitute

Fix bootstrap to remove read-only directories right.

When an earlier in-tree 'make dist' was interrupted, it may leave behind a directory automake-$APIVERSION with read-only subdirs, breaking a subsequent bootstrap. This trivial patch fixes that. Pushed to master and branch-1-10. Cheers, Ralf Fix bootstrap to remove read-only directories righ

Re: Updated Vala support for automake

Hello people interested in Automake support for Vala, (please indicate if you don't want to be Cc:ed.) * Ralf Wildenhues wrote on Sat, Sep 06, 2008 at 08:53:06PM CEST: > <http://thread.gmane.org/gmane.comp.sysutils.automake.patches/2860/focus=513>. I've udpated Mathias'

Re: Fixing plural case in Banner

Hello William, thanks for the patches. Here's a pretty nitty review, I hope you don't mind. * William Pursell wrote on Sat, Oct 11, 2008 at 05:27:25AM CEST: > Ralf Wildenhues wrote: >>> + if test "$$all" -eq 1; then \ >>> + tests="

Re: typo in comment in tests/defs.in

Hi William, * William Pursell wrote on Sat, Oct 11, 2008 at 06:58:47AM CEST: > --- a/tests/defs.in > +++ b/tests/defs.in > @@ -152,7 +152,7 @@ do >CC=icc >export CC ># There is no way to ask *only* the compiler's version. > - # This tool always want to do something (by

[PATCH] Nicer 'make dist' output.

The output of 'make dist' has become more and more cluttered. I am thinking of two choices to clean it up: - either post all commands, i.e., each cp and all, - or just post basically one line per directory, - or post nothing. The patch below goes the second way. I've pushed it now, but I'd apprec

Parallel tests execution [0/4]

eated if all tests passed? - should skipped tests show up in test-suite.log? - is everybody ok with the following authorship for patch 1/4? 2008-10-XX Akim Demaille <[EMAIL PROTECTED]> Jim Meyering <[EMAIL PROTECTED]> Benoit Sigoure <[EMAIL PROTECTED]&g

Parallel test execution: new option `parallel-tests': [1/4]

Parallel test execution: new option `parallel-tests'. 2008-07-05 Akim Demaille <[EMAIL PROTECTED]> Jim Meyering <[EMAIL PROTECTED]> Benoit Sigoure <[EMAIL PROTECTED]> Ralf Wildenhues <[EMAIL PROTECTED]> * automake.in (handl

parallel-tests: Ensure backward-compatible semantics. [2/4]

parallel-tests: Ensure backward-compatible semantics. For each test in Automake's test suite that uses TESTS, generate an identical one that uses the `parallel-tests' option, for coverage of backward-compatible functionality. * tests/gen-parallel-tests: New file, gener

New tests for `parallel-tests'. [3/4]

New tests for `parallel-tests'. * tests/parallel-tests.test: New file, to expose a bunch of features of `parallel-tests': VERBOSE, clean, TEST_SUITE_LOG, test dependencies, DISABLE_HARD_ERRORS, LAZY_TEST_SUITE. * tests/parallel-tests2.test: New file, test check-html. *

Documentation for the parallel-tests driver. [4/4]

Documentation for the parallel-tests driver. * doc/automake.texi (Tests, Options): Document the `parallel-tests' option, including new features of the test driver. * NEWS: Update. diff --git a/NEWS b/NEWS index ed71e9e..6171e2e 100644 --- a/NEWS +++ b/NEWS @@ -101,6 +101,11 @@

Re: magic variables for included fragments (was: Feature request)

[ moving to automake-patches; this is <http://thread.gmane.org/gmane.comp.sysutils.automake.general/9824> ] * Ralf Wildenhues wrote on Sun, Oct 12, 2008 at 10:46:06PM CEST: > > I'll follow up on automake-patches with a patch to test. Here we go. WDYT? Cheers, Ralf Recurs

Re: [PATCH] Nicer 'make dist' output.

Hi Eric, thanks for the feedback. * Eric Blake wrote on Mon, Oct 13, 2008 at 04:18:03PM CEST: > > I like the shorter output. However, ... > > > new_top_distdir=$$reldir; \ > > + echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) > > top_distdir="$$new_top_distdir" distdir="$$new_dis

Re: Parallel tests execution [0/4]

Hi Jim, * Jim Meyering wrote on Wed, Oct 15, 2008 at 09:13:54PM CEST: > Ralf Wildenhues <[EMAIL PROTECTED]> wrote: > > 12) allow for additional output on stdout/stderr? > > example: test was skipped because of $reason. > > This would be nice. > It

Re: Parallel tests execution [0/4]

Hi Akim, thanks for the feedback! * Akim Demaille wrote on Thu, Oct 16, 2008 at 04:05:47PM CEST: > >>> "RW" == Ralf Wildenhues <[EMAIL PROTECTED]> writes: > > > These four patches implement parallel execution of TESTS in Automake, > > adapted

Re: Documentation for the parallel-tests driver. [4/4]

* Akim Demaille wrote on Thu, Oct 16, 2008 at 05:16:20PM CEST: > >>> "RW" == Ralf Wildenhues <[EMAIL PROTECTED]> writes: > > > [EMAIL PROTECTED] Simple tests using @samp{parallel-tests} > > [EMAIL PROTECTED] @option{parallel-tests}, Using > > +T

Re: Parallel tests execution [0/4]

* Ralf Wildenhues wrote on Wed, Oct 15, 2008 at 11:35:07PM CEST: > * Jim Meyering wrote on Wed, Oct 15, 2008 at 09:13:54PM CEST: > > Ralf Wildenhues <[EMAIL PROTECTED]> wrote: > > > 12) allow for additional output on stdout/stderr? > > > example: te

AM_DEFAULT_SOURCE_EXT (was: Parallel tests execution [0/4])

* Akim Demaille wrote on Thu, Oct 16, 2008 at 04:11:22PM CEST: > > I wish I could define > > AM_DEFAULT_SOURCE_EXT = .cc > > or whatever, so that I wouldn't have to define all the foo_SOURCES. I agree "extension" is better than "suffix". (Even if Automake itself is inconsistent here and uses

<    2   3   4   5   6   7   8   9   10   11   >