I don't know if this is the way to go on principal, and I definitely feel
that passing flags around in environment variables and unadorned on command
lines is a messy hack, but I wanted to run the test suite with the JIT.
So I did this:
--- Makefile~ Tue Jan 1 20:10:58 2002
+++ Makefile Tue Jan 1 22:19:10 2002
@@ -331,10 +331,22 @@
#
###############################################################################
-test: $(TEST_PROG) assemble.pl .test_dummy
+test_common: $(TEST_PROG) assemble.pl
+
+test: test_common .test_dummy
.test_dummy:
$(PERL) t/harness
+
+test_jit: test_common .test_jit_dummy
+
+.test_jit_dummy:
+ $(PERL) t/harness -j
+
+test_prederef: test_common .test_prederef_dummy
+
+.test_prederef_dummy:
+ $(PERL) t/harness -P
mopstest: $(TEST_PROG) examples/assembly/mops.pbc
$(TEST_PROG) examples/assembly/mops.pbc
--- t/harness~ Sat Dec 15 14:21:46 2001
+++ t/harness Tue Jan 1 20:45:27 2002
@@ -3,5 +3,9 @@
use strict;
use Test::Harness qw(runtests);
+if (@ARGV) {
+ $ENV{PARROT_TEST_OPTS} = shift;
+}
+
my @tests = map { glob( "t/$_/*.t" ) } ( qw(op) );
runtests( @tests );
--- Parrot/Test.pm~ Sat Dec 15 14:19:56 2001
+++ Parrot/Test.pm Tue Jan 1 20:47:06 2002
@@ -12,6 +12,9 @@
@EXPORT = ( qw(output_is output_like output_isnt), @Test::More::EXPORT );
@ISA = qw(Exporter Test::More);
+my $test_opts = exists $ENV{PARROT_TEST_OPTS} ? $ENV{PARROT_TEST_OPTS} : '';
+
+
sub import {
my( $class, $plan, @args ) = @_;
@@ -60,7 +63,7 @@
close ASSEMBLY;
_run_command( "$PConfig{perl} assemble.pl $as_f --output $by_f" );
- _run_command( "./$PConfig{test_prog} $by_f", 'STDOUT' => $out_f, 'STDERR' =>
$out_f);
+ _run_command( "./$PConfig{test_prog} $test_opts $by_f", 'STDOUT' => $out_f,
+'STDERR' => $out_f);
my $prog_output;
open OUTPUT, "< $out_f";
after make shared I can do this like normal:
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib make test
/usr/bin/perl t/harness
t/op/basic..........ok
t/op/bitwise........ok
t/op/debuginfo......ok
t/op/hacks..........ok
t/op/integer........ok
t/op/interp.........ok
t/op/macro..........ok
t/op/number.........ok
t/op/pmc............ok, 1/47 subtest skipped
t/op/pmc_perlstring.ok
t/op/stacks.........ok, 1/19 subtest skipped
t/op/string.........ok
t/op/time...........ok
t/op/trans..........ok
All tests successful, 2 subtests skipped.
Files=14, Tests=249, 710 wallclock secs (577.37 cusr + 49.16 csys = 626.53 CPU)
and I can test the prederef like this:
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib make test_prederef
/usr/bin/perl t/harness -P
t/op/basic..........# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: 'start
lab 1
done
'
# Looks like you failed 1 tests of 8.
dubious
Test returned status 1 (wstat 256, 0x100)
DIED. FAILED test 8
Failed 1/8 tests, 87.50% okay
t/op/bitwise........ok
t/op/debuginfo......ok
t/op/hacks..........ok
t/op/integer........# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: 'the word
done 1
done 2
done 3
'
# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: 'start
done 1
done 2
done 3
'
# Looks like you failed 2 tests of 30.
dubious
Test returned status 2 (wstat 512, 0x200)
DIED. FAILED tests 12, 15
Failed 2/30 tests, 93.33% okay
t/op/interp.........ok
t/op/macro..........ok
t/op/number.........ok
t/op/pmc............ok, 1/47 subtest skipped
t/op/pmc_perlstring.# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: 'x
12
xxxxxxxxxxxx
'
# Looks like you failed 1 tests of 4.
dubious
Test returned status 1 (wstat 256, 0x100)
DIED. FAILED test 4
Failed 1/4 tests, 75.00% okay
t/op/stacks.........ok, 1/19 subtest skipped
t/op/string.........# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: 'ok 1
ok 2
ok 3
ok 4
'
# Looks like you failed 1 tests of 63.
dubious
Test returned status 1 (wstat 256, 0x100)
DIED. FAILED test 25
Failed 1/63 tests, 98.41% okay
t/op/time...........ok
t/op/trans..........ok
Failed Test Status Wstat Total Fail Failed List of failed
-------------------------------------------------------------------------------
t/op/basic.t 1 256 8 1 12.50% 8
t/op/integer.t 2 512 30 2 6.67% 12, 15
t/op/pmc_perlst 1 256 4 1 25.00% 4
t/op/string.t 1 256 63 1 1.59% 25
2 subtests skipped.
Failed 4/14 test scripts, 71.43% okay. 5/249 subtests failed, 97.99% okay.
*** Error code 2
and the JIT like this (and I've no idea why '1' doesn't match '1')
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib make test_jit
/usr/bin/perl t/harness -j
t/op/basic..........# Failed test (Parrot/Test.pm at line 79)
# got: '1'
# expected: '1'
# Failed test (Parrot/Test.pm at line 79)
# got: '42'
# expected: '42'
# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: '42
1234
'
# Failed test (Parrot/Test.pm at line 79)
# got: 'start
done
'
# expected: 'start
lab 1
done
'
# Looks like you failed 4 tests of 8.
dubious
Test returned status 4 (wstat 1024, 0x400)
DIED. FAILED tests 2, 6-8
Failed 4/8 tests, 50.00% okay
t/op/bitwise........# Failed test (Parrot/Test.pm at line 79)
# got: ''
# expected: '6
6
5
12
'
^C
So, what's going wrong, and why all the segvs?
I figured it was better to start by wondering about the prederef code.
But I might just go to bed instead. Or fix compiler warnings. or perl5
The prederef does work, as does the JIT:
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib ./test_parrot
examples/assembly/mops.pbc
Iterations: 100000000
Estimated ops: 200000000
Elapsed time: 91.229787
M op/s: 2.192266
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib ./test_parrot -P
examples/assembly/mops.pbc
Iterations: 100000000
Estimated ops: 200000000
Elapsed time: 60.178785
M op/s: 3.323430
nick@thinking-cap [parrot]$ LD_LIBRARY_PATH=blib/lib ./test_parrot -j
examples/assembly/mops.pbc
Iterations: 100000000
Estimated ops: 200000000
Elapsed time: 4.572718
M op/s: 43.737663
For information, if this is relevant:
FreeBSD thinking-cap.moo 4.0-RELEASE FreeBSD 4.0-RELEASE #0: Mon Mar 20 22:50:22 GMT
2000 [EMAIL PROTECTED]:/usr/src/sys/compile/GENERIC i386
it's an AMD K5 133 with 16M RAM and a couple of very slow, small hard disks.
Nicholas Clark
