> > It's strange that 'make test' did not run. (The matter was the missing perl-binary
>in the perl-5.8.0/t - directory ...or what?)
>
> Yes. The odd things is that "make test" should have done that.
> See the target "_test_prep" in Makefile, the "_test_prep" should
> get called (mong other things) with "make test".
Adam: Here is the sclice from my Makefile (static build perl) regarding the test: Is
it ok?
There is no softlink perl-5.8.0/t/perl->../perl after and before I'm making test.
..PHONY: test check test_prep test_prep_nodll test_prep_pre _test_prep \
test_tty test-tty _test_tty test_notty test-notty _test_notty \
utest ucheck test.utf8 check.utf8 test.torture torturetest \
test.third check.third utest.third ucheck.third test_notty.third \
test.deparse test_notty.deparse \
minitest coretest
# Cannot delegate rebuilding of t/perl to make
# to allow interlaced test and minitest
TESTFILE=TEST
_test_prep:
cd t && (rm -f $(PERL)$(EXE_EXT); $(LNS) ../$(PERL)$(EXE_EXT)
$(PERL)$(EXE_EXT))
# Architecture-neutral stuff:
test_prep_pre: preplibrary utilities $(nonxs_ext)
test_prep: test_prep_pre miniperl$(EXE_EXT) perl$(EXE_EXT) $(dynamic_ext)
$(TEST_PERL_DLL)
PERL=./perl $(MAKE) _test_prep
_test_tty:
cd t && $(LDLIBPTH) $(PERL_DEBUG) $(PERL) $(TESTFILE) $(TEST_ARGS) </dev/tty
_test_notty:
cd t && $(LDLIBPTH) $(PERL_DEBUG) PERL_SKIP_TTY_TEST=1 $(PERL) $(TESTFILE)
$(TEST_ARGS)
# The second branch is for testing without a tty or controlling terminal,
# see t/op/stat.t
_test:
if (true </dev/tty) >/dev/null 2>&1; then \
$(MAKE) TEST_ARGS=$(TEST_ARGS) TESTFILE=$(TESTFILE) _test_tty ; \
else \
$(MAKE) TEST_ARGS=$(TEST_ARGS) TESTFILE=$(TESTFILE) _test_notty ; \
fi
@echo "Ran tests" > t/rantests
test check: test_prep
PERL=./perl $(MAKE) _test
test_tty: test_prep
PERL=./perl $(MAKE) _test_tty
test_notty: test_prep
PERL=./perl $(MAKE) _test_notty
utest ucheck test.utf8 check.utf8: test_prep
PERL=./perl TEST_ARGS=-utf8 $(MAKE) _test
coretest: test_prep
PERL=./perl TEST_ARGS=-core $(MAKE) _test
test-prep: test_prep
test-tty: test_tty
test-notty: test_notty
# Torture testing
test.torture torturetest: test_prep
PERL=./perl TEST_ARGS=-torture $(MAKE) _test
# Targets for Third Degree testing.
test_prep.third: test_prep perl.third
PERL=./perl.third $(MAKE) _test_prep
test.third check.third: test_prep.third perl.third
PERL=./perl.third PERL_DEBUG=PERL_3LOG=1 $(MAKE) _test
utest.third ucheck.third: test_prep.third perl.third
PERL=./perl.third PERL_DEBUG=PERL_3LOG=1 TEST_ARGS=-utf8 $(MAKE) _test
test_notty.third: test_prep.third perl.third
PERL=./perl.third $(MAKE) PERL_DEBUG=PERL_3LOG=1 _test_notty
# Targets for Deparse testing.
test.deparse: test_prep
PERL=./perl TEST_ARGS=-deparse $(MAKE) _test
test_notty.deparse: test_prep
PERL=./perl TEST_ARGS=-deparse $(MAKE) _test_notty
# Can't depend on lib/Config.pm because that might be where miniperl
# is crashing.
minitest: miniperl$(EXE_EXT) lib/re.pm
-@test -f lib/lib.pm && test -f lib/Config.pm || \
$(MAKE) lib/Config.pm lib/lib.pm
@echo " "
@echo "You may see some irrelevant test failures if you have been unable"
@echo "to build lib/Config.pm or lib/lib.pm."
@echo " "
- cd t && (rm -f perl$(EXE_EXT); $(LNS) ../miniperl$(EXE_EXT) perl$(EXE_EXT)) \
&& $(LDLIBPTH) ./perl TEST base/*.t comp/*.t cmd/*.t run/*.t io/*.t
op/*.t uni/*.t </dev/tty
# Test via harness
test_harness: test_prep
PERL=./perl $(MAKE) TESTFILE=harness _test
Help!
> > 99.81 is not bad....What are you thinking: Can I work in a productive environment
>with this perl build?
>
> The list of failures you are getting is slightly longer than what I
> got just before 5.8.0 (see pod/perldelta.pod, about line 3655). What
> you have extra are the DProf.t, io_dir.t, lib/h2xs.t, AnyDBM_File.t,
> Dirhandle.t, splice.t. Then again, you are not seeing the op/pat.t,
> op/sprintf.t and op/tr.t failures I did see, which I find kind of odd,
> since I'm pretty certain that op/tr.t should fail a few subtests
> because there's code that needs to be written first before the test
> can ever fully work in EBCDIC... (ditto for pat.t)
>
> But yes, I would say that you should be able to work with this Perl.
> (Hey, there are some Unixes that do not pass 100%...) Also in some
> cases the tests themselves are to blame, some tests still need porting
> to z/OS (the h2xs.t is one example, I think) Just be mindful of the
> errors you got: the list of failures points to the areas (either modules
> or core functionality) where one needs to tread carefully.
>
> > .../ext/Data/Dumper/t/dumper.t 357 8 2.24% 311 314 325 327
> > 331 333 337 339
> > .../ext/Devel/DProf/DProf.t 18 14 77.78% 5-18
> > .../ext/IO/lib/IO/t/io_dir.t 10 2 20.00% 2-3
> > .../ext/IO/lib/IO/t/io_unix.t 5 4 80.00% 2-5
> > .../ext/Storable/t/downgrade.t 10 2560 169 10 5.92% 14-15 46-47 78-79
> > 110-111 150 161
> > .../lib/h2xs.t 6 1536 113 6 5.31% 17 36 55 74 89 108
> > .../lib/AnyDBM_File.t 4 1024 12 4 33.33% 3 5-7
> > .../lib/DirHandle.t 5 2 40.00% 2-3
> > .../lib/ExtUtils/t/Constant.t 51 10 19.61% 32-33 38-42 45-46
> > 49
> > .../lib/ExtUtils/t/Embed.t 9 9 100.00% 1-9
> > .../lib/Tie/Array/splice.t 0 9 ?? ?? % ??
> > uni/fold.t 780 6 0.77% 61 169 196 661
> > 710-711
> > 72 tests and 186 subtests skipped.
> > Failed 12/523 test scripts, 97.71% okay. 75/40471 subtests failed, 99.81% okay.
>
> --
> Jarkko Hietaniemi <[EMAIL PROTECTED]> http://www.iki.fi/jhi/ "There is this special
> biologist word we use for 'stable'. It is 'dead'." -- Jack Cohen
>
_______________________________________________
Adam Fischler
DATEV e.G.
_______________________________________________
