depend: prefer $(...) over `...` in Cygwin-specific recipes
On Cygwin, we can happily assume only POSIX-conforming shells exist.
This commit is useless by itself, but is useful as a preparatory
refactoring in view of a future change.
* lib/am/depend2.am: In recipes for producing '.obj' objects, use
the POSIX form $(...) for command substitution, rather than the
classic Bourne one `...`.
This is just a preparatory change in view of a future refactoring.
No semantic change is intended.
* lib/am/depend2.am: Adjust; for example, using '$@' in all recipes
to indicate the output file (rather than recipe-specific transforms
like '%OBJ%' and '%LTOBJ%'), and not assuming to know the exact
file extension of the output file (e.g., whether it's '.o' or '.lo').
* lib/am/depend2.am: Here. The obsolescent comments removed from this
same file in commit 'v1.13.1-42-g8f06bfb' of 2012-01-09, "depend2.am: fix
comments on verbosity of compilation rules", gave the rationale for why
that code duplication was there in the first place (rationale that, like
those comments, has been obsolete by the silent-rules introduction).
* bin/automake.in: Given the refactoring in 'depend2.am', there is no
longer need to add extra trailing whitespace to the entries of the
'%sourceflags' hash (which are used for the '%SOURCEFLAG%' transform
when 'depend2.am' is processed.
* t/fort2.sh: Adjust to avoid spurious failures.
depend: assume we can always pass '-o' to the C compiler
This is the case after the change in commit v1.13.1-56-g34001a9 of
2013-01-09 (compile: use 'compile' script when "-c -o" is used with
losing compilers).
This is just a simplification; no semantic changes are intended.
* bin/automake.in (handle_languages): Drop transform '%-o%', and
definitions of '$output_flag'.
* lib/am/depend2.am: Adjust to just assume the '?-o?' transform is
true, and the '%-o%' transform expands to "-o".
cleanup: having subdir-objects mandatory allow us some simplifications
Few minor cleanups made possible by earlier changes, plus other minor
cleanups triggered in cascade. No semantic change is intended.
This is a follow-up on previous commit 'v1.13.1d-214-g3ad07d2', and
an adjusted backport of Automake-NG commit 'v1.12.1-315-gc97d41b'
of 2012-06-08 ([ng] cleanup: after enabling of subdir-objects
unconditionally).
* bin/automake.in (LANG_IGNORE, LANG_SUBDIR): Remove.
(handle_languages): Drop the '%DEPBASE%' transform when processing
the '$rule_file'.
(register_language ('name' => 'vala', ...)): Add '.vapi' to the entry
'extensions', and simplify the entry 'output_extensions' to point to
a dummy subroutine (since it wasn't really used anyway).
(handle_single_transform): No longer expect the 'lang_*_rewrite'
subroutines to return a 'LANG_*' constant, but only a transformed
extension, if required. To decide whether further processing of the
source file should be stopped, rely on a new set of 'lang_*_ignore'
subroutines, defaulting to a subroutine that returns false.
Accordingly, don't special case the handling of '.vapi' files anymore,
instead rely on ...
(lang_vala_ignore, lang_header_ignore): ... these new subroutines to
avoid extra processing of C/C++ headers and Vala '.vapi' headers.
(lang_java_rewrite): Remove.
Remove an outdated comment.
* lib/am/depend2.am: Partial rewrite to reduce code duplication and
drop use of the '%DEPBASE%' transform.
* t/compile_f_c_cxx.sh: Adjust.
The fact that Automake-generated Makefiles places compiled object files
in the current directory by default, also when the corresponding source
file is in a subdirectory, is basically an historical accident, due to
the fact that the 'subdir-objects' option had only been introduced in
April 1999, starting with commit 'user-dep-gen-branchpoint-56-g88b5959',
and never made the default, likely to avoid backwards-compatibility
issues.
Since we believe the behaviour enabled by the 'subdir-objects' is the
only natural and most useful one, we make it the only only one available,
simplifying the Automake implementation and APIs a little in the process.
This change is basically an adjusted backport of Automake-NG commit
'v1.12.1-313-g14fe163' of 2012-06-07, "[ng] subdir-objects: enable
unconditionally".
* NEWS: Update.
* doc/automake.texi (Program and Library Variables): The output
object files are no longer placed in the current directory by
default, but rather in the same directory of the source file.
(LIBOBJS): Now the $(LIBOBJS) and $(ALLOCA) variables can also
be used outside of the directory where their sources lie.
(List of Automake options): Report the 'subdir-objects' option
as a no-op, existing only for compatibility with older versions
of Automake.
Other related minor adjustments.
* bin/automake.in (LANG_PROCESS): Remove, it's no longer needed.
(handle_languages): Don't test whether option 'subdir-objects'
is set (just assume it is), and do not use the '%SUBDIROBJ%'
transform when processing '.am' fragments.
(lang_sub_obj): Delete, it would just return 'LANG_SUBDIR'
unconditionally now.
(lang_lex_rewrite): Adjust. Don't test whether the option
'subdir-objects' is set (just assume it is).
(lang_yacc_rewrite): Likewise.
(handle_single_transform): Likewise. Remove an obsolete
comment. Add a proper "FIXME" comments about a fragment
of code that might have become dead code now.
(handle_LIBOBJS_or_ALLOCA): Simplify assuming that the option
'subdir-objects' is always set. Accordingly, there's no need
to warn anymore if '$(LIBOBJS)' or '$(ALLOCA)' are used outside
the '$config_libobj_dir' directory (as specified by autoconf
macro 'AC_CONFIG_LIBOBJ_DIR').
* lib/am/depend2.am: Assume the '?SUBDIROBJ?' Automake time
conditional is always true, and remove its uses accordingly.
* t/compile_f_c_cxx.sh: Adjust.
* t/cscope.tap: Likewise.
* t/depcomp8a.sh: Likewise.
* t/depcomp8b.sh: Likewise.
* t/libtool3.sh: Likewise.
* t/ltlibsrc.sh: Likewise.
* t/pr401.sh: Likewise.
* t/pr401b.sh: Likewise.
* t/pr401c.sh: Likewise.
* t/subobj.sh: Likewise.
* t/lex-line.sh: Likewise.
* t/yacc-line.sh: Likewise.
* t/yacc5.sh: Likewise.
* t/vala-libs.sh: Likewise.
* t/fort4.sh: Likewise, and extend a bit.
* t/fort5.sh: Likewise.
* t/gcj.sh: Likewise.
* t/subpkg.sh: Likewise.
* t/subpkg-yacc.sh: Likewise.
* t/xsource.sh: Likewise.
* t/libobj20a.sh: Remove as obsolete.
* t/libobj20b.sh: Adjust heading comments.
* t/libobj20c.sh: Likewise.
* t/subobj4.sh: Remove as obsolete.
* t/sourcefile-in-subdir.sh: Likewise.
* bin/Makefile.inc (%D%/automake, %D%/aclocal): Make sure that the
directory where the targets scripts are going to be built exists,
before trying to create said scripts.
* maint:
Use AC_DEFUN_ONCE to define AM_PROG_CC_C_O
compile: avoid AC_PROG_CC messy rewrite
options: tiny simplification in dealing with incompatible versions
Nick Bowler [Sat, 11 May 2013 09:45:16 +0000 (11:45 +0200)]
Use AC_DEFUN_ONCE to define AM_PROG_CC_C_O
If AM_PROG_CC_C_O is expanded multiple times, and the compiler does not
support -c and -o together, each expansion of the macro will prepend
the compile script to CC. This can result in the compile script
invoking the compile script, which at best pointless and silly.
Fortunately, there does not appear to be any serious problems as the
first compile invocation strips out -o options, causing subsequent
invocations of the script to merely exec their arguments.
Other than fixing the above, this should not normally cause any changes
to the resulting configure script, except in the (hopefully rare) case
where AM_PROG_CC_C_O is directly expanded (i.e., *not* using AC_REQUIRE)
in the body of a macro defined with AC_DEFUN. In that case, the use of
AC_DEFUN_ONCE may cause the expansion of AM_PROG_CC_C_O to appear
earlier in the configure script.
* m4/minuso.m4: Change the definition of AM_PROG_CC_C_O to use
AC_DEFUN_ONCE, avoiding problems caused by multiple expansions.
Instead, add an hook to AC_OUTPUT to have AM_PROG_CC_C_O invoked
automatically.
See also the long-winded discussion about automake bug#13378.
* m4/minuso.m4 (AM_PROG_CC_C_O): Bring back the old implementation,
from commit v1.13.1-55-g1ab8fb6.
* m4/init.m4 (AC_PROG_CC): Remove this horrible, hacky re-write.
* (AM_INIT_AUTOMAKE): Arrange for AM_PROG_CC_C_O to be called if
necessary.
* t/am-prog-cc-c-o.sh: Adjust to avoid spurious failure.
* t/subobj.sh: Likewise.
Suggested-by: Nick Bowler <nbowler@elliptictech.com> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
options: tiny simplification in dealing with incompatible versions
* lib/Automake/Options.pm (_process_option_list): Here, when an
incompatible version number option is detected, there's no need
to call error() with the "uniq_scope => US_GLOBAL" switch.
In fact, if the same incompatible version number is specified in
AUTOMAKE_OPTIONS in both (say) 'Makefile.am' and 'sub/Makefile.am',
we want each such erroneous usage reported separately, rather than
just the first time it is encountered (as we'd expect to happen
when "uniq_scope => US_GLOBAL" is used).
Ideally, this change should have been folded into the similar
commit 'v1.13.1d-129-gf7ef16f', but we noticed that too late.
Oh well.
options: consistently use return statuses to report errors (more)
This is a follow-up to commit 'v1.13.1d-128-g6e486c5', needed to deal
with errors that are only present in the 'master' branch (while the
reported commit was on the 'maint' branch).
* lib/Automake/Options.pm (_process_option_list): Adjust, where dealing
with the removed options 'dist-shar' and 'dist-tarZ'.
* maint:
options: try to report as much errors as possible
warns: don't tell AM_PROG_MKDIR_P is going to be removed
refactor: fix few "inverted boolean" usages
options: better name for an internal function
options: more consistency in use of return statuses to report errors
options: tiny simplification in dealing with erroneous opts
options: consistently use return statuses to report errors
options: re-enable some sanity checks
THANKS: update Eric Blake's e-mail address
NEWS: typofix
warns: don't tell AM_PROG_MKDIR_P is going to be removed
That is no longer true. For a more extended rationale, see file
'PLANS/obsolete-removed/am-prog-mkdir-p.txt' in the maint branch
(as of commit v1.13.1d-132-g90ec3fe).
* automake.in (scan_autoconf_traces): So adjust the warning message
here.
In some subroutines, we used a return value of 0 to indicate success,
and a return status of 1 to indicate failure. That was not very
consistent with the perl interpretation of 0 as a false value and 1 as
a true value. So we now invert the meaning of the exit statuses.
* lib/Automake/Options.pm (_process_option_list): Here.
(process_global_option_list, process_option_list): And by reflex,
here as well.
* bin/automake.in (handle_options): And here.
(generate_makefile, scan_autoconf_traces): Adjust.
options: tiny simplification in dealing with erroneous opts
* lib/Automake/Options.pm (_process_option_list): Here, when an
invalid option is detected, there's no need to call &error with
the "uniq_scope => US_GLOBAL" switch. In fact, if the same
erroneous option is specified in AUTOMAKE_OPTIONS in both (say)
'Makefile.am' and 'sub/Makefile.am', we want each such erroneous
usage reported separately, rather than just the first time it is
encountered (as happens when "uniq_scope => US_GLOBAL" is used).
See also discussion about automake wishlist bug#13324.
* lib/Automake/Options.pm: Give fatal errors (rather than warnings) if
the 'dist-shar' or 'dist-tarZ' options are used.
* lib/distdir.am: Remove the 'dist-tarZ' and 'dist-shar' targets, and
references to the '.tar.Z' and '.shar' archives.
* bin/automake.in (preprocess_file): Remove 'COMPRESS' and 'SHAR'
transforms.
(handle_dist): Remove lingering references to 'dist-tarZ' and 'dist-shar'
options.
* doc/automake.texi: Adjust, removing references to the removed targets
and distribution formats.
* t/dist-shar.sh: Adjust to expect fatal errors rather than warnings.
* t/dist-tarZ.sh: Likewise.
* maint:
news: document new 'subdir-objects' warning
PLANS: one minor fixlet (mostly cosmetic)
PLANS: we have already dropped support for split info files in master
NEWS: fix a reference to Automake 1.14 where Automake 2.0 was intended
PLANS: fix reference to non-existent 'next' branch
PLANS: fix botched version reference
maintcheck: fix two references to old location of aclocal and automake
dist: deprecated shar and tar+compress formats
* NEWS: Automake 1.14 will warn if a subdir source file is
specified but the 'subdir-objects' option is not given. This
is done to smooth the transition to Automake 2.0, which will
unconditionally assume the behaviour now given only with the
'subdir-objects' option.
* PLANS/rm-f-without-args.txt: Here. The probe checking that "rm -f"
without arguments works will be introduced in Automake 1.14, not in
Automake 1.13.2.
See also discussion about automake wishlist bug#13324.
* lib/Automake/Options.pm: Give proper warnings in the 'obsolete'
category if the 'dist-shar' or 'dist-tarZ' options are used.
* lib/distdir.am: When the 'dist-tarZ' or 'dist-shar' targets are
invoked, make them give a non-fatal warning.
* doc/automake.texi: Report the new deprecations.
* t/dist-shar.sh: New test.
* t/dist-tarZ.sh: Likewise.
* t/lzma.sh: While at it, rename ...
* t/dist-lzma.sh: ... like this, and tweak it to keep more in
sync with the new tests.
* t/dist-formats.tap: Remove references to deprecated formats.
* t/list-of-tests.mk: Adjust.
* maint:
automake: typofix in comments: s/AC_CONFIG_HEADER/AC_CONFIG_HEADERS/
am: prefer a shorter idiom where possible
cosmetics: fix few typos, grammaros and missing whitespace
fixup: remove an obsolete comment
docs: we still don't have the promised better Java interface
build: move automake and aclocal in 'bin' subdir
build: break up monolithic Makefile.am in subdir-specific fragments
+ Extra non-trivial edits:
* m4/Makefile.inc (dist_automake_ac_DATA): Drop lead-dot.m4 and
mkdirp.m4.
* lib/Automake/Makefile.inc (dist_perllib_DATA): Drop Configure_ac.pm.
* branch-1.13.2:
cosmetics: fix few typos, grammaros and missing whitespace
fixup: remove an obsolete comment
docs: we still don't have the promised better Java interface
tests: avoid spurious failure with older flex (2.5.4)
That old version is unfortunately still relevant, being the one
installed on NetBSD 5.1.
* t/lex-multiple.sh: Use the '-o' option rather than the longer
equivalent '--outfile'. The latter is not supported by older
versions of flex (e.g., flex 2.5.4).
* maint.mk (announcement): Here, be prepared to handle the case
in which the first section of the NEWS file is dedicated to report
future backward-incompatibilities and/or other warnings.
* automake.in: Rename ...
* bin/automake.in: ... like this.
* aclocal.in: Rename ...
* bin/aclocal.in: ... like this.
* Makefile.am: Move parts that dealt with the building/distribution
of aclocal and Automake ..
* bin/Makefile.inc): ... in this new included fragment. Adjust as
needed, and make deliberate use of the '%D%' substitution.
* lib/gen-perl-protos: Move ...
* bin/gen-perl-protos: ... here.
* bootstrap.sh, configure.ac, maintainer/rename-tests,
t/wrap/aclocal.in, t/wrap/automake.in, doc/Makefile.inc,
t/ax/tap-setup.sh, .gitignore: Adjust.
* maintainer/syntax-checks.mk: Likewise, and enhance a little.
build: break up monolithic Makefile.am in subdir-specific fragments
This is convenient to do, now that we have improved "relative directory"
support with the '%reladir%' (a.k.a. '%D%') and '%canon_reladir%' (a.k.a.
'%C%') Automake-time substitutions for included makefile fragments.
This move also satisfy our philosophy of using new Automake features in
our own build system, as a way of facilitating early discovery of possible
bugs or interface warts.
* Makefile.am: Break up ...
* doc/Makefile.inc, lib/Automake/Makefile.inc, lib/Makefile.inc,
lib/am/Makefile.inc, m4/Makefile.inc, t/Makefile.inc): ... in this
new included fragments. Adjust as needed, and make deliberate use
of the '%D%' substitution.
* contrib/t/local.am: Rename ...
* contrib/t/Makefile.inc: ... like this.
doc/automake.texi:3885: warning: node next `Public Macros' in menu \
`Private Macros' and in sectioning `Obsolete Macros' differ
doc/automake.texi:4050: warning: node `Private Macros' is next for \
`Obsolete Macros' in sectioning but not in menu
doc/automake.texi:4050: warning: node prev `Obsolete Macros' in menu \
`Private Macros' and in sectioning `Public Macros' differ
doc/automake.texi:4090: warning: node `Obsolete Macros' is next for \
`Private Macros' in menu but not in sectioning
doc/automake.texi:4090: warning: node prev `Private Macros' in menu \
`Public Macros' and in sectioning `Obsolete Macros' differ
* maint:
maint branch: we are going to become Automake 1.14
maint: version bump after beta release 1.13.1d
release: beta release 1.13.1d (will become 1.13.2)
NEWS: document more robust handling/recognition of make options
tests: typofixes in comments in t/preproc-c-compile.sh
tests: remove bashism from a test
* branch-1.13.2:
maint: version bump after beta release 1.13.1d
release: beta release 1.13.1d (will become 1.13.2)
NEWS: document more robust handling/recognition of make options
* maint:
maint: targets and recipes to simplify testing on real-world packages
build: preparatory refactoring
build: tiny reduction in code duplication
make flags analysis: handle more options with args
make flags analysis: use simpler variable names
make flags analysis: whitespace changes
make flags analysis: embed in a subshell
make flags analysis: be more robust
make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
tests: expose weaknesses in make flags analysis
tests: improve debugging output in checks on make flags analysis
make flags analysis: refactor, to reduce code duplication
tests: avoid one tricky use of "make -e"
cosmetics: remove few trailing whitespace occurrences
typofix: in NEWS (extra whitespace)
tests: avoid a spurious error with Solaris make
subdirs: don't return false positives for the '-k' option's presence
header-vars: recognize more make flags ('-k' in particular)
header-vars: simplify how make flags are determined
tests: remove dead code from t/make-dryrun.tap
header-vars: new variable $(am__running_with_option)
tests: expose bug#12554 (false positives for presence of '-k' make option)
* branch-1.13.2:
maint: targets and recipes to simplify testing on real-world packages
build: preparatory refactoring
build: tiny reduction in code duplication
make flags analysis: handle more options with args
make flags analysis: use simpler variable names
make flags analysis: whitespace changes
make flags analysis: embed in a subshell
make flags analysis: be more robust
make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
tests: expose weaknesses in make flags analysis
tests: improve debugging output in checks on make flags analysis
make flags analysis: refactor, to reduce code duplication
tests: avoid one tricky use of "make -e"
tests: avoid a spurious error with Solaris make
subdirs: don't return false positives for the '-k' option's presence
header-vars: recognize more make flags ('-k' in particular)
header-vars: simplify how make flags are determined
tests: remove dead code from t/make-dryrun.tap
header-vars: new variable $(am__running_with_option)
tests: expose bug#12554 (false positives for presence of '-k' make option)
Merge 'better-makeflags-recognition' and 'testing-work' into branch-1.13.2
* better-makeflags-recognition:
make flags analysis: handle more options with args
make flags analysis: use simpler variable names
make flags analysis: whitespace changes
make flags analysis: embed in a subshell
make flags analysis: be more robust
make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
tests: expose weaknesses in make flags analysis
tests: improve debugging output in checks on make flags analysis
make flags analysis: refactor, to reduce code duplication
* testing-work:
maint: targets and recipes to simplify testing on real-world packages
build: preparatory refactoring
build: tiny reduction in code duplication
maint: targets and recipes to simplify testing on real-world packages
We introduce a new section in the maintainer-specific makefile that
contains recipes to test the build system of some well-known GNU
packages with the current development version of Automake. Not the
cleanest way to do so, but good enough for the moment. We'll revisit
the matter after the 1.13.2 release (which we now hope will happen
soon).
* maint.mk (git-sv-host): New.
(SV_GIT_CF, SV_GIT_AC, SV_GIT_GL): Use it to reduce code duplication.
(ALL_PACKAGES, FEW_PACKAGES): New, lists of GNU packages to try out.
(ttp-check, ttp-check-all): New targets, do the checking with said
packages.
(ttp): New, alias for 'ttp-check'.
(ttp-all): New, alias for 'ttp-check-all'.
The code was only duplicated two times, but we are soon going to
need a third occurrence, and that would be one to much.
* Makefile.am (extend_path): New.
(update_mans): Use it instead of copying & pasting its contents.
($(srcdir)/doc/amhello-1.0.tar.gz): Likewise, and minor related
adjustments.
* lib/am/header-vars.am (am__make_running_with_option): Here. Now
that we expect to be run in a subshell, we don't have to worry about
being namespace-safe. And '$foo' is much more pleasant to read than
'$am__foo' -- and pleasant code tends to be more correct.
(am__make_dryrun, am__make_keepgoing): Adjust.
make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
The current development version of GNU make (that is planned to become
GNU make 3.83, sooner or later) has changed the format its $(MFLAGS)
variable slightly, removing the space between an option and its argument:
# With GNU make 3.82, compiled from official tarball:
$ make -f- <<<'all:; @echo "$$MFLAGS"' -I none
-I none
# With development version of GNU make (Git commit b5ea49b):
$ make -f- <<<'all:; @echo "$$MFLAGS"' -I none
-Inone
This was done on purpose, in order to support more easily the new
option '-O', which takes an optional argument; see:
Which was causing a spurious failure on FreeBSD. Not particularly
surprising, given how brittle "make -e" is in general ...
* t/cxx-lt-demo.sh: Instead of forcing $(CC) to be 'false' by
exporting "CC=false" in the environment and then passing the '-e'
option to make, do so by passing "CC=false" on the make command
line, both directly and using AM_MAKEFLAGS.
* fix-pr12554:
tests: avoid a spurious error with Solaris make
subdirs: don't return false positives for the '-k' option's presence
header-vars: recognize more make flags ('-k' in particular)
header-vars: simplify how make flags are determined
tests: remove dead code from t/make-dryrun.tap
header-vars: new variable $(am__running_with_option)
tests: expose bug#12554 (false positives for presence of '-k' make option)
compat: substitute '@mkdir_p@' as '$(MKDIR_P)' unconditionally
We had started to do so only when the obsolete AM_PROG_MKDIR_P variable
was used, but that has recently turned out not to be not good enough,
because Gettext 0.18.2 (the latest one at the moment of writing) is using
the modern macro AC_PROG_MKDIR_P in its '*.m4' files, but at the same time
is still using the obsolete make variable '$(mkdir_p)' in its Makefile
fragments. This has been causing failures in real-world packages
bootstrapped with latest Gettext (0.18.2) and with bleeding-edge automake.
And the worst things is that those failures have been basically silent
and unnoticeable up until the final "make install" --- that is, after
the affected package had been bootstrapped, configured and built. Yikes!
This issue has been reported to the Gettext developers:
But even if they will quickly fix it, Gettext 0.18.2 will remain
affected.
The fix we can implement in Automake is small and unobtrusive enough,
definitely worth having in order to save the users and developers from
having to deal with this hard-to-spot issue.
* m4/init.m4 (AM_INIT_AUTOMAKE): AC_SUBST '@mkdir_p@' to '$(MKDIR_P)'.
* t/am-prog-mkdir-p.sh: Enhance.
* maint:
sync: update files from upstream with "make fetch"
maintcheck: remove outdated whitelisting
tar: format 'ustar' cannot support UID/GID longer than 21 bits
docs: issues with configure substitutions in TESTS
tests: avoid possible autotools caching issues (automake bug#13832)
tests: rename some with more descriptive names
docs: add myself and Ralf Wildenhues as authors
authors: add myself
dry-run: don't get confused by '-I' option
tests: avoid a spurious failure with the Korn Shell
dry-run: with GNU make, prefer $(MFLAGS) over $(MAKEFLAGS)
header vars: can determine whether we are running under GNU make
NEWS: improve wording for automake bug#13514 fix
NEWS: document fix for automake bug#13514
* branch-1.13.2:
sync: update files from upstream with "make fetch"
maintcheck: remove outdated whitelisting
tar: format 'ustar' cannot support UID/GID longer than 21 bits
subdirs: don't return false positives for the '-k' option's presence
This change fixes automake bug#12554.
The old implementation of the code descending into $(SUBDIRS)
entries used the following snippet to decide whether make is running
with the '-k' a.k.a. '--keep-going' option, and thus whether a failure
in a subdirectory should prevent the descent in the following ones:
fail= failcom='exit 1'; \
for f in x $$MAKEFLAGS; do \
case $$f in \
*=* | --[!k]*);; \
*k*) failcom='fail=yes';; \
esac; \
done
It's clear that the second pattern in the 'case' construct could possibly
match false positives, for examples in these two cases:
make check TESTS="x.test k.test"
make -I /usr/local/kool-fragments
which are somewhat unusual, but not invalid. So we need a more resilient
implementation, as we did for the detection of the '-n' flag.
This implementation is now provided by the new private macro
'$(am__make_keepgoing)' (introduced in recent commits); so we can
just us that to fix the bug.
* lib/am/subdirs.am ($(am__recursive_targets)): Use '$(am__make_keepgoing)'
instead of ad-hoc and more brittle checks.
* t/list-of-tests.mk (XFAIL_TESTS): Remove the now-passing test case
't/subdir-keep-going-pr12554.sh'.
Reported-by: Michael Daniels <mdaniels@rim.com> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
header-vars: recognize more make flags ('-k' in particular)
* lib/am/header-vars.am (am__running_with_option): Here.
Few improvements to comments, while at it.
(am__make_keepgoing): New, tell whther make is being runt with
the '-k' option.
* t/make-keepgoing.tap: New test.
* t/list-of-tests.mk: Add it.
* t/make-dryrun.tap: Minor edits to keep it more in sync with
the new test.
* syntax-checks.mk (sc_tests_overriding_macros_on_cmdline): Here.
The test 'make-dryrun.sh' has been since long rewritten as the TAP
test 'make-dryrun.tap', and no longer spuriously triggers this
maintainer check.
header-vars: simplify how make flags are determined
Actually, son far only the '-n' option ("dry mode") was detected,
but this change will allow us to soon detect more options.
* lib/am/header-vars.am (am__running_with_option): Even when $MAKEFLAGS
appears to contain definition of variables with embedded whitespace,
use simple textual pre-processing over $MAKEFLAGS rather than tricky
recursive invocations of make to determine whether the '-n' option was
given. This is enough to correctly handle all the tricky usages covered
in the testsuite.
* t/nodep.sh: Adjust to avoid a spurious failure.
header-vars: new variable $(am__running_with_option)
This is a preparatory refactoring, needed by later patches.
No semantic change is intended.
* lib/am/header-vars.am (am__running_with_option): New, contains
shell code that determines whether the current make instance is
running with a given one-letter option (e.g., -k, -n) that takes
no argument. Actually, the only supported option at the moment
is '-n' (support for '-k' will be added soon).
(am__make_dryrun): Rewrite as a thin wrapper around
'$(am__make_running_with_option)'.
tests: expose bug#12554 (false positives for presence of '-k' make option)
The current implementation of the code descending into $(SUBDIRS)
entries uses the following snippet to decide whether make is running
with the '-k' a.k.a. '--keep-going' option, and thus whether a failure
in a subdirectory should prevent the descent in the following ones:
fail= failcom='exit 1'; \
for f in x $$MAKEFLAGS; do \
case $$f in \
*=* | --[!k]*);; \
*k*) failcom='fail=yes';; \
esac; \
done
It's clear that the second pattern in the 'case' construct can possibly
match false positives, for examples in these two cases:
make check TESTS="x.test k.test"
make -I /usr/local/kool-fragments
which are somewhat unusual, but not invalid. So we need a more resilient
implementation, as we did for the detection of the '-n' flag.
But alas, such an implementation seems quite tricky to obtain in portable
make. So for the moment we content ourselves with exposing the bug, with
the hope of being able to fix soon enough.
* t/subdir-keep-going-pr12554.sh: New test.
* t/list-of-tests.mk (handwritten_TESTS, XFAIL_TESTS): Add it.
* THANKS: Update
Reported-by: Michael Daniels <mdaniels@rim.com> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
tar: format 'ustar' cannot support UID/GID longer than 21 bits
See automake bug#8343 and bug#13588.
POSIX 1988 'ustar' format is defined with *fixed-size* fields. There
is notably a 21 bits limit (2097151) for the UID and the GID.
Tom Rini tom_rini@mentor.com says (in bug#8343):
When the user has a UID or GID that is larger than the ustar format
supports, pax does not error out gracefully in some cases (FC13).
Marc Herbert <marc.herbert@intel.com> adds (in bug#8343):
When "configure" is run by a user with an UID bigger than 21 bits,
BSD pax 3.4 aborts when trying to create the 'conftest.tar' test
archive and leaves an empty or corrupted conftest.tar file behind.
In the next step, pax tries to extract this incomplete or corrupted
archive and this *** hangs the whole ./configure script ***.
Note: GNU cpio 2.9 pretends to pass the test but it is a LIE: it
silently truncates any big UID to its lower 21 bits. I don't know
what can be the consequences of this lie.
I think there is currently a design issue in automake/m4/tar.m4
considering that a ustar archive should should *never* succeed when
./configure is run from a big user ID.
Months later, Petr Hracek <phracek@redhat.com> reports a similar issue
(in bug#13588) for Fedora 17:
I am trying to solve problem in case a user is created with big
UID and during configuration pax hangs with message
ATTENTION! pax archive volume change required.
Ready for archive volume: 1
Input archive name or "." to quit pax.
Archive name >
Time to fix this issue, on the line of a preliminary patch provided by
Petr Hracek in bug#13588. The final patch ended up being remarkably
different from that original proposition, though.
* m4/tar.m4 (_AM_PROG_TAR): If the UID or GID of the current user is
too high (> 2097151), the 'ustar' format cannot work. Adjust checks
accordingly. Some related code reordering and clean-up.
* t/tar-ustar-id-too-high.sh: New test.
* t/list-of-tests.mk: Add it.
* t/tar.sh: While at it, tweak and enhance a little.
* t/tar2.sh: Likewise.
* t/tar3.sh: Likewise.
* t/tar-override.sh: Likewise.
* NEWS: Update.
* THANKS: Likewise.
Helped-by: Pavel Raiskup <praiskup@redhat.com> Helped-by: Petr Hracek <phracek@redhat.com> Helped-by: Marc Herbert <marc.herbert@intel.com> Helped-by: Tom Rini <tom_rini@mentor.com> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
projects / thirdparty / automake.git / log
