* Makefile.am (sc_diff_automake_in_automake): Update, as we
now expect 9 lines, not 8, to be changed from `automake.in'
to `automake'.
(sc_diff_aclocal_in_aclocal): New maintainer check, similar to
the above, and checking that only 10 lines are changed from
`aclocal.in' to `aclocal'.
(syntax_check_rules): Update.
(sc_tests_Exit_not_exit): Exempt self tests `self-check-*.test'
from this check, as they can legitimately use the bare `exit'
builtin in various places.
* doc/automake.texi (Python): Remove stray `@' from the end of
a line. Typo introduced in commit `v1.11-312-g5bf7af6'.
* tests/depcomp8a.test: Pass DISTCHECK_CONFIGURE_FLAGS to make
from the environment rather than from the command line, to
pacify the `sc_tests_overriding_macros_on_cmdline' maintainer
check.
* tests/depcomp8b.test: Likewise.
tests: don't require GNU make where it's not strictly needed
* tests/autohdr3.test: Rewrite to be stricter when make is GNU
make. Drop the requirement of GNU make.
* tests/aclocal5.test: Drop GNU make requirement, it's not truly
needed. Add reference to similar tests `remake-subdir*.test'.
* remake-subdir-gnu.test, remake-subdir-from-subdir.test,
remake-subdir.test, remake-subdir2.test: Add reference to
each other, and to related test `aclocal5.test'.
* tests/aclocal6.test: Drop GNU make requirement, it's not
truly needed.
* tests/confh6.test: Likewise.
* tests/lex3.test: Likewise.
* tests/remake11.test: Likewise.
* tests/subdir5.test: Likewise.
* tests/subdir8.test: Likewise.
* tests/werror2.test: Likewise.
* tests/conff.test: Likewise, and ensure verbose printing of
captured make output.
* tests/lex5.test: Tweak so that GNU make is no more required.
* tests/version7.test: Likewise.
* tests/maken2.test: Add explicative comment for why this test
required GNU make.
* tests/maken4.test: Let it run also with BSD makes supporting
the `.MAKE' special target.
* tests/output6.test: Use proper m4 quoting in configure.in.
Expand make macros with one-character name using `$(x)', not
`$x', for portability. Move checks in the makefiles, rather
than relying on grepping the output from make. Drop the now
unneeded GNU make requirement.
lex tests: do not force the use of flex unconditionally
* tests/defs (lex): Act more similarly to what the `yacc'
requirement does, i.e., only force the use of flex if the
$LEX variable is left unset by the user.
(flex): Use `skip_' to skip the test if flex is not found.
* tests/lex3.test (foo.l:yywrap): Return 1, not 0, to avoid hangs.
Bug introduced in commit 'v1.11-871-geb147a1'.
(Makefile.am): Do not add `@LEXLIB@' to `$(LDADD)', as we define
our own `yywrap' function.
* tests/lex.test (tscan.l): In `yywrap', return 1, not 0, for
consistency with the default flex implementation.
* tests/lex-libobj.test (yywrap.c): Likewise.
* tests/lex-subobj-nodep.test (s1.l): Likewise.
* tests/lexvpath.test (foo.c): Likewise.
* tests/silent-lex-gcc (foo.l): Likewise.
* tests/silent-lex-generic (foo.l): Likewise.
* tests/silent-many-gcc (foo5.l): Likewise.
* tests/silent-many-generic (foo5.l): Likewise.
* tests/lex-lib.test (mu.c): Likewise.
Update heading comments, to refer to ...
* tests/lex-lib-external.test: ... this new test, which checks
that we can get use the `yywrap' function from a system-wide
library, if that's available.
* tests/self-check-cleanup.test: When the test must be skipped,
use `skip_ REASON' instead of `echo REASON; Exit 77'. Also,
make the skip message shorter and clearer.
Our ad-hoc usage of `tests/defs' in the testsuite's self tests
stopped working properly when we made the test scripts re-execute
themselves with the configure-time $SHELL. Fix this.
* tests/defs: Only check that we can find the client test script
when we must re-execute it.
* tests/self-check-cleanup.test: Export `AM_TESTS_REEXEC' to "no"
before running the self tests.
* tests/self-check-dir.test: Likewise.
* tests/self-check-explicit-skips.test: Likewise.
* tests/self-check-me.test: Likewise.
* tests/self-check-sanity.test: Likewise.
* tests/self-check-reexec.test: New test.
* tests/Makefile.am (TESTS): Update.
tests: prefer ulimit over timeout in test on a memory-hogging bug
* tests/cond29.test: Use 'ulimit' instead of 'timeout' to ensure
that automake does not uses up too much resources. This is really
and improvement because the bug tested by this script wasn't just
a "it hangs" or "it runs too slow" bug, but rather a memory-hogging
bug (due to combinatorial explosion when many Automake conditionals
had to be handled) which could easily crash the whole system, which
is unacceptable. The requirement of a proper and working 'ulimit'
builtin might cause the test to be skipped on more systems, but
that shouldn't be a problem since the bug isn't about a portability
issue, but is rather an automake internal implementation problem.
I've verified that the test as updated by this patch still passes
with automake 1.8.5, automake 1.10.2, and obviously the development
version of automake, and that it fails with automake 1.7.9.
testsuite: avoid generating `*-p.test' tests, use a wrapper script
The generated `*-p.test' tests had already become just thin layers
around the corresponding test scripts. This change makes the final
step, converting to the use of a generic wrapper script and thus
avoiding the extra test generation (similarly to what is done for
the `*.instspc' and `*.depmod' tests).
* tests/parallel-tests.sh: New file, driver script to run checks
the on the `parallel-tests' semantics by wrapping tests that use
the generic "Simple Tests" driver.
* tests/gen-parallel-tests: Update, mostly to reflect the new
`.ptest' extensions used for tests in $(parallel_tests).
* Makefile.am (TESTS_EXTENSIONS): Add `.ptest'.
(PTEST_LOG_COMPILER): Define, it calls `parallel-tests.sh'.
($(parallel_tests)): Do not really generate `*-p.test' tests
anymore; this is now just a dummy dependency declaration required
in order to have make actually produce expected log files from
the `.ptest.log' suffix rule.
(EXTRA_DIST): Distribute `parallel-tests.sh'.
(MAINTAINERCLEANFILES): Don't remove the `$(parallel_tests)', it
is not necessary anymore.
(generated_tests): Variable definition removed.
(TESTS): Update, by listing `$(parallel_tests)' directly instead
of `$(generated_tests)'
(expected_list_of_tests): Remove `$(generated_tests)'.
(maintainer-check-list-of-tests): No need to explicitly depend on
`$(expected_list_of_tests)' anymore.
* master:
automake, aclocal: honour configure-time AUTOCONF and AUTOM4TE
build: the user can override AUTOM4TE, AUTORECONF and AUTOUPDATE too
tests/README: fix example about `make -e' usage
* maint:
automake, aclocal: honour configure-time AUTOCONF and AUTOM4TE
build: the user can override AUTOM4TE, AUTORECONF and AUTOUPDATE too
tests/README: fix example about `make -e' usage
automake, aclocal: honour configure-time AUTOCONF and AUTOM4TE
Currently, the Automake's own configure script allow definition
of AUTOCONF and AUTOM4TE, expected to point respectively to an
autoconf and autom4te programs. But while these definitions are
honoured in the Automake's build systems and test suite, they
were *not* honoured in the generated `automake' and `aclocal'
scripts. This behaviour, apart from being wrong in that it does
not allow the user enough freedom in choosing his tools, also
caused inconsistencies in the test suite, brining to spurious
failures.
Problem reported by Graham Reitz on the automake list; see thread:
<http://lists.gnu.org/archive/html/automake/2011-05/msg00022.html>
* automake.in ($traces): Use `@am_AUTOCONF', not simply `autoconf'.
* aclocal.in ($traces): Use `@am_AUTOM4TE', not simply `autom4te'.
* Makefile.am (do_subst): Substitute also `@am_AUTOCONF' and
`@am_AUTOM4TE'.
* NEWS: Update.
* THANKS: Update.
build: the user can override AUTOM4TE, AUTORECONF and AUTOUPDATE too
Our build system allows the user to override AUTOCONF and AUTOHEADER
at configure time, and honours these overrides in our testsuite.
But it didn't do the same with AUTOM4TE, AUTORECONF and AUTOUPDATE.
This change fixes that inconsistency.
* configure.ac (am_AUTOM4TE, am_AUTOUPDATE, am_AUTORECONF): New
AC_SUBSTitutions. Update comments.
* tests/defs.in ($AUTOUPDATE): Default to `@am_AUTOUPDATE@' now.
($AUTOM4TE): New variable, defaulting to `@am_AUTOM4TE@'.
($AUTORECONF): New variable, defaulting to `@am_AUTORECONF@'.
* doc/Makefile.am ($(srcdir)/amhello-1.0.tar.gz): Update.
remake: behave better with non-GNU make in subdirectories
Currently, with every decent make program, it is possible to
rebuild out-of-date autotools-generated files with a simple
"make Makefile" -- but for this to work reliably with non-GNU
make implementations, the command must be issued from the
top-level directory. This patch removes such limitation.
* lib/am/configure.am (am--refresh): Depend on `%MAKEFILE%'.
* tests/defs.in (using_gmake): New function, backported from the
`master' branch (and simplified).
* tests/remake-subdir.test: New test.
* tests/remake-subdir2.test: Likewise.
* tests/remake-subdir-gnu.test: Likewise.
* tests/remake-subdir-from-subdir.test: Likewise.
* tests/Makefile.am (TESTS): Update.
tests: don't require gfortran if any fortran compiler is enough
* tests/silentf90.test ($required): Require `fortran', not
`gfortran'.
* tests/silentf77.test ($required): Require `fortran77', not
`gfortran'.
* tests/silent-many-generic.test ($required): Require `fortran'
and `fortran77' rather than `gfortran'.
tests: improve `ccnoco*.test', better cross-compiling support
* tests/ccnoco3.test (Mycomp): Use the `$CC' chosen by `tests/defs'
instead of forcing `gcc' unconditionally. This ensures better
coverage in case of cross-compiling, when GCC can be named e.g.,
`i586-mingw32msvc-gcc', instead of simply `gcc'.
* tests/ccnoco.test: Likewise. Remove redundant checks. Modernize
the created `configure.in'. Run tests both in-tree and in VPATH.
Export `CC' to the overridden value only once.
* tests/ccnoco2.test: Slightly stricter grepping of automake
stderr. Add trailing `:' command.
testsuite: use $SHELL to run tests which are shell scripts
* tests/Makefile.am (TEST_LOG_COMPILER): Define so that the
configure-time $SHELL is used to run the tests.
* tests/defs: Add code to re-execute by default the running
test script with configure-time $SHELL. Updated comments.
* configure.ac: Check that `set -e' is working for $SHELL,
not for /bin/sh.
* tests/defs-static.in: Update comments.
* tests/README (Supported shells): Updated.
(Getting details from failures): Don't tell that tests are
run by /bin/sh by default.
tests: split 'subst2.test' to improve modularity and coverage
* tests/subst2.test: Test removed, split into ...
* tests/programs-primary-rewritten.test: ... this test ...
* tests/subst-no-trailing-empty-line.test: ... and this one ...
* tests/extra-programs-empty.test: ... and this one.
* tests/Makefile.am (TESTS): Update.
tests: tweak and improve tests on Automake conditionals
The "avoid the requirement of a working compiler" we refer about
below is obtained by disabling automatic dependency tracking and
defining `CC' (and `EXEEXT', `OBJEXT', `CXX', etc., if needed) to
dummy values, either directly in the `Makefile.am', or by AC_SUBST
in `configure.in'.
The "cosmetic changes" we refer about below consists in, e.g.,
adding a trailing `:' command to a script, using proper m4 quoting
in `configure.in' files, adding commands that offer more debugging
output, explicitly declaring phony targets as such in Makefiles,
making use of idioms more consistent with those employed in other
tests, following the GNU formatting standards more closely, etc.
* tests/cond.test: Cosmetic changes.
* tests/cond3.test: Likewise.
* tests/cond10.test: Likewise.
* tests/cond14.test: Likewise.
* tests/cond15.test: Likewise.
* tests/cond17.test: Likewise.
* tests/cond40.test: Likewise.
* tests/cond13.test: Likewise, and make grepping of the generated
`Makefile.in' slightly stricter.
* tests/cond6.test: Extend by also running ./configure, make and
"make install". Few cosmetic changes.
* tests/cond8.test: Likewise.
* tests/cond2.test: Do few cosmetic changes. Make grepping of
automake stderr stricter.
* tests/cond20.test: Likewise.
* tests/cond23.test: Likewise.
* tests/cond24.test: Likewise.
* tests/cond42.test: Likewise.
* tests/cond46.test: Likewise.
* tests/cond9.test: Move more checks in the `Makefile.am' instead
of grepping make's output.
* tests/cond38.test: Likewise.
* tests/cond11.test: Likewise. Avoid the requirement of a working
C compiler.
* tests/cond16.test: Likewise.
* tests/cond22.test: Likewise.
* tests/cond30.test: Likewise.
* tests/cond31.test: Likewise.
* tests/cond4.test: Likewise, and drop the now-useless GNU make
requirement as well.
* tests/cond18.test: Likewise.
* tests/cond19.test: Likewise.
* tests/cond22.test: Likewise.
* tests/cond25.test: Add trailing `:' command.
* tests/cond26.test: Likewise.
* tests/cond27.test: Likewise.
* tests/cond28.test: Likewise.
* tests/cond29.test: Use `unindent' function for better formatting.
Other minor cosmetic changes. Use a `timeout' program (if it's
available) to determine whether the script takes too long, instead
of just hoping that the user will notice an abnormally long test
execution time.
* tests/cond33.test: Extend a bit. Fix heading comments. Few more
cosmetic changes.
* tests/cond37.test: Ensure verbose printing of captured make
output. Minor cosmetic changes.
* tests/cond34.test: Likewise. Avoid the requirement of a working
C compiler. Make grepping of make output slightly stricter.
* tests/cond35.test: Quote literal dots in grep regexp. Add extra
debugging output. Minor cosmetic changes.
* tests/cond36.test: Likewise.
* tests/cond41.test: Do not simply check that aclocal fails, but
also grep its stderr for the expected error message.
* tests/cond43.test: Likewise (but the program being automake).
Few cosmetic changes.
* tests/condd.test: Avoid the requirement of a working C compiler.
* tests/condman3.test: Extend by using more man pages and more
manpage sections.
* tests/condman.test: Removed, it's completely superseded by
`condman3.test'.
* tests/Makefile.am (TESTS): Update.
lex tests: avoid spurious failures when LEXLIB isn't found
The AC_PROG_LEX Autoconf macro does not diagnose a failure to find
the "lex library" expected to provide a `yywrap' function (function
which is required to link most lex-generated programs). On the
contrary, when all the link attempts (i.e., with `-ll' and `-lfl')
fail, configure declares that no lex library is needed, and simply
proceeds with the configuration process -- only for the build to
possibly fail later, at make time.
This behaviour might be (partly) intended; the Autoconf manual
reads:
``You are encouraged to use Flex in your sources, since it is
both more pleasant to use than plain Lex and the C source
it produces is portable. In order to ensure portability,
however, you must either provide a function `yywrap' or, if
you don't use it (e.g., your scanner has no `#include'-like
feature), simply include a `%noyywrap' statement in the
scanner's source.''
This AC_PROG_LEX behaviour is causing some spurious failures of the
Automake testsuite in environments which lack a proper library
providing `yywrap' (this happens for example in Linux->MinGW cross
compilations). But at this point is clear that a proper workaround
is to simply provide a fall-back implementation of `yywrap' in our
lexers.
* tests/cond35.test: Provide a dummy `yywrap' function.
* tests/lex3.test: Likewise.
* tests/lexvpath.test: Likewise.
* tests/silent-many-gcc.test: Likewise.
* tests/silent-many-generic.test: Likewise.
* tests/silent-lex-gcc.test: Likewise, and a dummy `main' too.
* tests/silent-lex-generic.test: Likewise.
* tests/lex-lib.test: New test.
* tests/lex-libobj.test: New test.
* tests/lex-nowrap.test: New test.
* tests/Makefile.am (TESTS): Update.
* THANKS: Update.
testsuite: require C++ compiler explicitly in tests needing it
The list of the test scripts which needed a C++ compiler but
failed to require it explicitly has been found by running:
$ tests=`grep -lE 'CXX|\.c(c|\+\+|xx|pp)' *.test | tr '\n' ' '`
$ am_explicit_skips=yes CXX=false make check TESTS="$tests"
and looking for tests that reported FAIL instead of SKIP.
After the present change, all those previously failing tests
either pass or get correctly skipped.
* tests/subobj9.test: Require `c++', instead of explicitly
skipping on configure failure. Add excerpts and/or details from
the original bug report that prompted this tests to be written.
* tests/silentcxx-gcc.test: Require g++.
* tests/silentcxx.test: Require c++.
* tests/suffix3.test: Likewise.
testsuite: require C compiler explicitly in tests needing it
The list of the test scripts which needed a C compiler but
failed to require it explicitly has been found by running:
$ am_explicit_skips=yes CC=false make check
and looking for tests that reported FAIL instead of SKIP.
After the present change, all those previously failing tests
either pass or get correctly skipped.
testsuite: allow user to ask for stricter skip semantics
Given how the Automake testsuite is currently structured, if a
command in a test script fails with status `77', the global test
outcome is considered a SKIP, because the value of `77' for `$?'
is passed to the exit trap. Indeed, this happens in practice, as
an autoconf-generated configure script can exit with status `77'
if it fails to find, e.g., a required compiler.
While this behaviour is quite useful for avoiding spurious test
failures in the wild, it can also sometimes prevent the Automake
developers to easily see and declare the requirements of their
tests.
This change introduces a new variable `am_explicit_skips', meant to
be user-overridable, and which, when set to a "true" value (i.e.,
`yes' or `1'), require a direct call to `Exit 77' in order to make
the test outcome be considered a SKIP.
* tests/defs.in ($am__test_skipped): New variable, initialized
to `no'.
(Exit): Set `$am__test_skipped' to `yes' if passed an exit status
of 77.
(trap '...' 0): When `$am_explicit_skips' is set to a "true" value,
reset an exit status of `77' to `78' if $am__test_skipped is not
set to `yes'.
* tests/self-check-exit.test: Adjust: unset `am_explicit_skips'.
* tests/self-check-explicit-skips.test: New test.
* tests/Makefile.am (TESTS): Update.
testsuite: user can force skipping of compiler-requiring tests
* tests/defs (cc, c++, fortran, fortran77): Skip the test if
the relevant compiler is disabled by having the corresponding
variable (CC, CXX, FC and F77, respectively) set to "false".
(yacc): For consistency, skip the test when the YACC variable
is set to "false", not when it's set to "no". Since we are at
it, fix the skip message to be shorter and more consistent.
tests: tweak and improve tests on "Simple Tests" driver
* tests/check.test: Make grepping of generated Makefile.in
slightly stricter. Add trailing `:' command.
* tests/check2.test: Ensure verbose printing of captured make
output. Make grepping of generated Makefile.in and of make
output stricter. Add trailing `:' command.
* tests/check3.test: Likewise. Also, fix minor error in a
comment.
* tests/check4.test: Stricter grepping of make output. Use
`using_gmake' subroutine instead of copy&paste from defs.in
to determine whether $MAKE is GNU make. Add a trailing `:'
command.
* tests/check6.test: Add trailing `:' command.
* tests/check10.test: Prefer redirection `>' over `>>' when
creating new files. Ensure verbose printing of captured make
output.
* tests/check11.test: Likewise.
* tests/check7.test: Likewise. Make grepping of make output
stricter. Prefer "make distcheck" over a simple "make check"
in a VPATH build. Add trailing `:' command.
(a.c): Format following GNU coding standards.
(Makefile.am): Explicitly declare target `print-xfail-tests'
as PHONY. Distribute $(check_SCRIPTS), for distcheck.
* tests/parallel-tests.test: Prefer redirection `>' over `>>'
when creating new files.
* tests/parallel-tests2.test: Likewise.
* tests/parallel-tests9.test: Likewise.
* tests/parallel-tests8.test: Likewise. Make grepping of
automake stderr stricter.
* tests/check8.test: Likewise. Make formatting of created
*.c files more consistent with GNU Coding Standards.
* tests/check-subst-prog.test (configure.in): Break overly
long lines.
(pass-prog.c, xfail-prog.c): Format them more consistently
with GNU Coding Standards.
* tests/parallel-tests3.test: Use a shorter and clearer skip
message. Use the `unindent' subroutine to improve readability
of indented loop bodies. Simplify the hack used to please
maintainer-check.
* tests/parallel-tests5.test: Make formatting of created *.c
files more consistent with GNU Coding Standards. Use the
`unindent' subroutine to improve readability of indented loop
bodies.
* tests/parallel-tests10.test: Add blank line, for clarity.
Prefer trailing `:' over trailing `Exit 0', for consistency
with other tests.
Ralf Wildenhues [Sun, 22 May 2011 17:02:27 +0000 (19:02 +0200)]
tests/README: fix example about `make -e' usage
* tests/README (Section "Writing test cases" subsection "Do"): When
some variable is never initialized in the Makefile, `-e' is not
necessary in order to override it. DESTDIR is such a variable: we
ensure that we do not ever initialize it. And as such, it is quite
portable to use:
$ make DESTDIR=/foo/bar install
and in fact, quite widely used.
So our example about when `make -e' is required, which references
the `DESTDIR' variable, is poorly chosen, if not downright wrong.
Rewrite it to use `prefix' as the overridden variable instead.
parallel-tests: stricter checks on DISABLE_HARD_ERRORS support
* tests/parallel-tests-harderror.test: New test, doing more
in-depth checks on DISABLE_HARD_ERRORS.
* tests/parallel-tests.test: Remove tests on DISABLE_HARD_ERRORS,
now redundant.
* tests/Makefile.am (TESTS): Update.
* tests/README (Section "User interface" subsection "Running the
tests"): In examples, be more consistent w.r.t. the differences
between GNU and non-GNU make.
tests/README: suggest how to run tests in cross-compile mode
* tests/README (Section "User interface" subsection "Running the
tests"): Briefly explain how to override 'host_alias' at runtime
to force the use of cross-compilers by the testsuite. Give an
example.
* configure.ac (AC_CANONICAL_HOST): New, probably not strictly
necessary, but useful to complements AC_CANONICAL_BUILD.
* tests/defs-static.in ($host_alias, $build_alias): New variables,
exported, user-overridable, defaults AC_SUBST'd by configure.
* tests/defs (cross_compiling): Consider cross-compilation mode
active only if `$host_alias' is not empty *and* it differs from
`$build_alias'. This is more faithful to the way configure
determines whether it is cross-compiling, but still allows the
user to easily force cross-compilation by overriding `host_alias'
and `build_alias' at runtime.
* maint:
testsuite: avoid re-running few tests with 'parallel-tests' option
testsuite: allow tests to avoid the use of 'parallel-tests' option
tests/README: update obsoleted advice
test defs: rename requirement 'non-cross' -> 'native'
testsuite: be more cross-compile friendly
* master:
testsuite: avoid re-running few tests with 'parallel-tests' option
testsuite: allow tests to avoid the use of 'parallel-tests' option
tests/README: update obsoleted advice
* maint:
testsuite: avoid re-running few tests with 'parallel-tests' option
testsuite: allow tests to avoid the use of 'parallel-tests' option
tests/README: update obsoleted advice
* maint:
testsuite: avoid re-running few tests with 'parallel-tests' option
testsuite: allow tests to avoid the use of 'parallel-tests' option
tests/README: update obsoleted advice
With extra edits to these files:
- tests/java-compile-run-flat.test
- tests/java-compile-run-nested.test
- ChangeLog
- tests/Makefile.in (autogenerated)
* maint:
testsuite: avoid re-running few tests with 'parallel-tests' option
testsuite: allow tests to avoid the use of 'parallel-tests' option
tests/README: update obsoleted advice
testsuite: avoid re-running few tests with 'parallel-tests' option
Some tests in our testsuite use the 'simple-tests' driver only
marginally, or simply as a mean to conveniently check unrelated
invariants. It makes little sense to force these tests to also
run with the 'parallel-tests' Automake option active, as doing so
offers no real gain in coverage, while often causing a measurable
overhead in execution time (for an already too-slow testsuite).
* tests/pr401.test (parallel_tests): Define to "no", to prevent
the generation of a sibling test script using the 'parallel-tests'
driver.
* tests/pr401b.test: Likewise.
* tests/pr401c.test: Likewise.
testsuite: allow tests to avoid the use of 'parallel-tests' option
* tests/gen-parallel-tests: Do not generate "siblings" for tests
that explicitly define the `parallel_tests', whether to "yes" or
to any other value. Extend heading comments to give a rationale
for this behaviour.
* tests/README: Update.
* tests/README (Section "Writing test cases" subsection "Do"):
Do not suggest to use the `*-p.test' pattern for the names of
hand-written tests which use the `parallel-tests' Automake option.
Not only is this not respected by the existing tests, but it is
more likely to cause conflicts with auto-generated tests.
So, suggest to *avoid* using the `*-p.test' pattern in names
of hand-written tests instead.
(Section "Writing test cases" subsection "Do not"): When
suggesting not to override Makefile variables using command
line arguments, do not use the badly outdated variables `U'
and 'ANSI2KNR' in the example; instead, use the more common
and typical `DESTDIR'.
tests: avoid spurious failures in cross-compile mode
* tests/depcomp2.test: Ensure verbose printing of captured stderr
from configure.
* tests/ansi3.test ($required): Add 'native', as the ansi2knr
functionality is not meant to work with a cross-compiler.
* tests/ansi3b.test: Likewise.
* tests/ansi5.test: Likewise.
* tests/ansi6.test: Likewise.
* tests/ansi7.test: Likewise.
* tests/ansi8.test: Likewise.
* tests/ansi10.test: Likewise.
* tests/subobj3.test: Likewise.
* tests/check8.test ($required): Add 'native', as this test is
not meant to work with a cross-compiler.
* tests/gnits2.test: Likewise.
* tests/ltconv.test: Likewise.
* tests/ltcond2.test: Likewise.
* tests/parallel-tests5.test: Likewise.
* tests/specflg7.test: Likewise.
* tests/specflg8.test: Likewise.
* tests/check5.test: Likewise. Also, ensure verbose printing of
captured make output, and thrown in few cosmetic and consistency
improvements.
* tests/ltinit.test: Be laxer in grepping configure output, to
avoid spurious failures on systems which lack POSIX dynamic
linking (e.g., MinGW), or when cross-compiling for such systems.
* tests/ansi9.test: Skip parts of the test that would require a
native compiler.
* tests/backcompat6.test: Likewise.
* tests/depcomp8a.test: Likewise.
* tests/pr401.test: Likewise.
* tests/pr401b.test: Likewise.
* tests/pr401c.test: Likewise.
* tests/target-cflags.test: Likewise.
* tests/lex3.test: Likewise.
* tests/yacc-basic.test: Likewise.
* tests/autohdr4.test: Likewise. Also, extend the test to ensure
more coverage in cross-compile mode.
* tests/canon7.test: Likewise.
* tests/check9.test: Removed, split into two new tests ...
* tests/check-subst-prog.test: ... this one, which requires a
native C compiler ...
* tests/check-subst.test: ... and this one, which doesn't require
any compiler.
* tests/parallel-tests4.test: Removed, split into two new tests ...
* tests/parallel-tests-suffix-prog.test: ... this one, which
requires a native C compiler ...
* tests/parallel-tests-suffix.test: ... and this one, which
doesn't require any compiler.
* tests/parallel-tests7.test: Removed, split into two new tests ...
* tests/parallel-tests-ext-driver-prog.test: ... this one, which
requires a native C compiler ...
* tests/parallel-tests-ext-driver.test: ... and this one, which
doesn't require any compiler.
* tests/Makefile.am (TESTS): Update.
* maint:
test defs: rename requirement 'non-cross' -> 'native'
testsuite: be more cross-compile friendly
tests: fix spurious failure of txinfo21.test on FreeBSD
* maint:
test defs: rename requirement 'non-cross' -> 'native'
testsuite: be more cross-compile friendly
tests: fix spurious failure of txinfo21.test on FreeBSD
* tests/defs.in (cross_compiling): New subroutine.
(am__tool_prefix): New internal variable.
(gcc, g++, gcj): Force the use of the correct "tool prefix"
when cross compiling.
(gfortran, g77, non-cross): New requirements.
tests: fix portability issues in 'repeated-options.test'
* tests/repeated-options.test: Do not assume that object files
have `.o' suffix and executables have no default suffix; instead,
use `$(OBJEXT)' and `$(EXEEXT)'.
lex: "make clean" removes .c files from non-distributed .l
Previously, while automake did *not* distribute C source and header
files derived from non-distributed Lex sources, it still caused
them to be removed only by "make maintainer-clean" only, and not by
simply "make clean" or "make distclean".
This caused "make distcheck" to fail, unless the developer put
those generated .c files in CLEANFILES or in DISTCLEANFILES by
hand.
This change fixes this issue, by making non-distributed `.c' files
generated by non-distributed Lex sources cleaned by "make clean".
A similar problem for Yacc support had been fixed with the commit v1.11-263-ged2c8bc.
* tests/automake.in (lang_lex_target_hook): Make C source files
derived from non-distributed Lex files cleaned by "make clean",
not only by "make maintainer-clean".
* tests/lex-clean.test: New test.
* tests/lex-clean-cxx.test: Likewise.
* tests/lex-nodist.test: Likewise.
* tests/lex-pr204.test: Likewise.
* tests/pr204.test: For consistency, renamed ...
* tests/yacc-pr204.test: ... to this, and updated to keep it
more in sync with 'lex-pr204.test'.
* tests/yacc-nodist.test: Updated to keep it more in sync with
'lex-nodist.test'.
* tests/Makefile.am (TESTS): Update.
* NEWS: Update.
lex tests: remove erroneous check about ylwrap distribution
* tests/lex5.test: Do not check that the ylwrap script is *not*
distributed when there is only one lexer, as ylwrap is in fact
expected to distributed unconditionally (at least starting from
commit `Release-1-9-323-gc5881cc' of 19-08-2006). Note that this
bogus check wasn't triggering any failure because it was done at
the wrong time (i.e. from the VPATH build directory), so that it
could never fail anyway.
Instead, check that ylwrap is distributed even when there is only
one lexer.
Since we are at it, add a couple of `ls -l' calls, to get better
debugging info.
yacc tests: "make clean" removes C++ files from non-distributed .y
* tests/yacc-clean-cxx.test: New test, sister of `yacc-clean.test'.
* tests/yacc-clean.test: Update heading comment to reference the
new sister test.
* tests/Makefile.am (TESTS): Update.
The parallel-tests driver has now been used quite extensively
by a fair number of real-world applications (e.g., GNU coreutils,
GNU libtool, GNU grep, and various packages using Gnulib), and
thus exposed to adequate on-field testing. So there's no point
in declaring it experimental anymore (which would risk to make
potential users shy away from it).
* doc/automake.texi (Simple Tests using parallel-tests): Do not
declare the parallel-tests driver as "experimental" anymore.
check: document and test $(TEST_SUITE_LOG) overriding
* doc/automake.texi (Simple Tests using parallel-tests): Explain
how and why TEST_SUITE_LOG can be overridden at runtime.
* tests/parallel-tests-log-override-1.test: New test, check that
the newly documented idiom and similar usages are truly supported.
* tests/parallel-tests-log-override-2.test: Likewise.
* tests/parallel-tests-log-override-recheck.test: Likewise.
* tests/Makefile.am (TESTS): Update.
tests: fix spurious failure of extradep.test on FreeBSD
* tests/extradep.test: When using `ls -t' to determine whether a
file has been updated, make sure to use as reference a file whose
timestamp is expected to be *strictly* older that that of the file
being checked. This is required because at least FreeBSD `ls' do
not sort files with the same timestamp in alphabetical order when
using the `-t' option.
* tests/extradep2.test: Likewise.
tests: fix spurious failure of txinfo21.test on FreeBSD
* tests/txinfo21.test: Use the `is_newest' subroutine instead of
the `ls -t' hack to to determine whether a file has been updated.
This is required because at least FreeBSD `ls' do not sort files
with the same timestamp in alphabetical order when using the `-t'
option.
* tests/depmod-tests.sh (get_depmodes): Use `echo $all_depmodes'
instead of `... | tr "$nl" "$sp"' to strip extra whitespaces and
newlines from `$all_depmodes'. The latter idiom with `tr' is
wrong since it gets run before `./defs' is sourced, and thus `$nl'
and `$sp' are undefined.
Bug revealad by FreeBSD tr(1) implementation, which doesn't accept
empty strings as arguments.
* tests/yacc-d-cxx.test (Makefile.am): Add temporary workaround
for automake bug#8485. The bug is still exposed by the testcase
'yacc-dist-nobuild-subdir.test'.
tests: add excerpts from few original bug reports, for clarity.
* tests/comment6.test: Add excerpts and/or details from the
original bug report that prompted this tests to be added and/or
extended.
* tests/dejagnu4.test: Likewise.
* tests/installdir.test: Likewise.
With some edits to `tests/defs' and `ChangeLog', to reflect the
fact that in master we already had a `yacc' requirement, and no
more had an explicitly-recognized `bison' requirement.
* tests/lex-subobj-nodep.test ($required): Require `lex',
not `flex', as this test is expected to work with any lex
implementation.
* tests/silent-lex.test: Likewise.
tests defs: allow requirements for compilers (mostly dummy)
Most of the new requirements that are now accepted in `$required'
as consequence of this patch are still dummy. They are planned
to be implemented only in master (or in some derived branch), but
having them here (even just as no-op) will allow for an easier
integration/backporting of potential new testcases.
* tests/defs.in (cc, c++, fortran, fortran77): New requirements,
still dummy.
(flex): New requirement, picking LEX for configure.
(lex): New requirement, alias for `flex'. A more appropriate
implementation, looking for a generic `lex' program, will follow
in the future.
(yacc): New requirement, alias for `bison'. A more appropriate
implementation, looking for a generic `yacc' program, will follow
in the future.
tests: 'silent-many-gcc.test' improved and made more robust
* tests/silent-many-gcc.test: Also force "fast" gcc depmode
for C++ compilations. Add sanity checks verifying that the
cache variables we force are really used by configure. Fix
typo in comments.
* tests/lflags2.test ($required): Remove 'g++', as any working
C++ compiler should be enough, and ./configure will cause the
test to skip if no working C++ compiler is found.
* tests/yflags2.test: Likewise.
* tests/subobj9.test: Likewise.
* tests/silentcxx.test: Likewise. Also, do not force depmodes
that could cause non-GNU C++ compilers to fail.
* tests/silentcxx-gcc.test: New test, like `silentcxx.test',
but forcing "fast" gcc depmode (and thus requiring the GNU C++
compiler).
* tests/specflg10.test: Add proper "fixme" comment telling that
we should make this test work with a generic C++ compiler.
* tests/Makefile.am (TESTS): Update.
tests defs: one more environment sanitization (corner case)
* tests/defs (am__using_gmake): Initialize. The `using_gmake'
subroutine was using this variable for caching, but wasn't
initializing it, which could cause problems in the (admittedly
very unlikely) case in which it was pre-existent in the
environment.
* tests/Makefile.am (TESTS): Added test scripts man[678].test,
which have been present on the filesystem and in the repository,
but which had erroneously been left out from $(TESTS).
* CheckListOfTests.am: New file.
(maintainer-check-list-of-tests): New target,
check for consistency between list of tests defined in the
including Makefile and list of tests on the filesystem.
(clean-maintcheck-testslist-tmp): New rule, to clean up the
temporary files that might be left around by the rules associated
with the previous target.
(clean-local): Depend on it.
* lib/Automake/tests/Makefile.am: Include `CheckListOfTests.am'.
(maintainer-check): Added dependency from the new target
`maintainer-check-list-of-tests'.
* tests/Makefile.am: Likewise.
* Makefile.am (maintainer-check-list-of-test): New target,
calling recursively into `tests/' and `lib/Automake/tests/',
using ...
(TEST_SUBDIRS): ... this new variable.
projects / thirdparty / automake.git / log
