RE: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
Father Chrysostomos wrote on 2013-08-12: George Greer ‘wrote’: ../t/re/pat.t...FAILED 228-231 ../t/op/taint.t.FAILED 592 OK, so which of my commits caused these? They make no sense to me. t/re/pat.t, t/re/pat_thr.t, t/op/taint.t and dist/threads-shared/t/clone.t all fail (in a threaded build) at c9f1f591ba751b24222b22f1f3a6799db0ab0d1b (Read-only COWs), the first of your three merge commits.
Re: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
George Greer âwroteâ: ../t/re/pat.t...FAILED 228-231 ../t/op/taint.t.FAILED 592 OK, so which of my commits caused these? They make no sense to me.
Re: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
I will investigate which commit caused the breakages, but in the meantime here is the op/taint.t dump output, confirming that ${^TAINT} has indeed been written to: C:\Dev\Git\perl\t..\perl harness op\taint.t op/taint.t .. 1/797 SV = PVMG(0x7afa7c) at 0x7d12cc REFCNT = 1 FLAGS = (GMG,SMG,IOK,pIOK) IV = 1 NV = 0 PV = 0 MAGIC = 0x7c275c MG_VIRTUAL = PL_vtbl_sv MG_TYPE = PERL_MAGIC_sv(\0) MG_OBJ = 0x7d12bc MG_LEN = 5 MG_PTR = 0x7d1fcc \24AINT SV = PVMG(0x7afa7c) at 0x7d12cc REFCNT = 1 FLAGS = (GMG,SMG,IOK,pIOK) IV = 0 NV = 0 PV = 0 MAGIC = 0x7c275c MG_VIRTUAL = PL_vtbl_sv MG_TYPE = PERL_MAGIC_sv(\0) MG_OBJ = 0x7d12bc MG_LEN = 5 MG_PTR = 0x7d1fcc \24AINT # Failed test 592 - Assigning to ${^TAINT} fails at op/taint.t line 1648 # got '' # expected /(?^:^Modification of a read-only value attempted)/ op/taint.t .. Failed 1/797 subtests (less 46 skipped subtests: 750 okay) Test Summary Report --- op/taint.t (Wstat: 0 Tests: 797 Failed: 1) Failed test: 592 Files=1, Tests=797, 0 wallclock secs ( 0.08 usr + 0.00 sys = 0.08 CPU) Result: FAIL On 12 August 2013 20:44, Father Chrysostomos spr...@cpan.org wrote: Steve Hay wrote: t/re/pat.t, t/re/pat_thr.t, t/op/taint.t and dist/threads-shared/t/clone.t all fail (in a threaded build) at c9f1f591ba751b24222b22f1f3a6799db0ab0d1b (Read-only COWs), the first of your three merge commits. Thank you. Can you tell which commit in the branch caused it? What do you get if you dump ${^TAINT} before and after the attempt to write to it? diff --git a/t/op/taint.t b/t/op/taint.t index b521408..479db1d 100644 --- a/t/op/taint.t +++ b/t/op/taint.t @@ -1643,8 +1643,10 @@ SKIP: { is(${^TAINT}, 1, '$^TAINT is on'); - +use Devel::Peek; +Dump ${^TAINT}; eval { ${^TAINT} = 0 }; +Dump ${^TAINT}; is(${^TAINT}, 1, '$^TAINT is not assignable'); like($@, qr/^Modification of a read-only value attempted/, 'Assigning to ${^TAINT} fails');
Re: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
The tests fail in commit 3805b5fb04f8819edfb6bd5256d5eefd7b23e730 (and pass in the previous commit, 0ffdaf1a47312f763d4610e9ec65c24561e957a5). On 12 August 2013 22:18, Steve Hay steve.m@googlemail.com wrote: I will investigate which commit caused the breakages, but in the meantime here is the op/taint.t dump output, confirming that ${^TAINT} has indeed been written to: C:\Dev\Git\perl\t..\perl harness op\taint.t op/taint.t .. 1/797 SV = PVMG(0x7afa7c) at 0x7d12cc REFCNT = 1 FLAGS = (GMG,SMG,IOK,pIOK) IV = 1 NV = 0 PV = 0 MAGIC = 0x7c275c MG_VIRTUAL = PL_vtbl_sv MG_TYPE = PERL_MAGIC_sv(\0) MG_OBJ = 0x7d12bc MG_LEN = 5 MG_PTR = 0x7d1fcc \24AINT SV = PVMG(0x7afa7c) at 0x7d12cc REFCNT = 1 FLAGS = (GMG,SMG,IOK,pIOK) IV = 0 NV = 0 PV = 0 MAGIC = 0x7c275c MG_VIRTUAL = PL_vtbl_sv MG_TYPE = PERL_MAGIC_sv(\0) MG_OBJ = 0x7d12bc MG_LEN = 5 MG_PTR = 0x7d1fcc \24AINT # Failed test 592 - Assigning to ${^TAINT} fails at op/taint.t line 1648 # got '' # expected /(?^:^Modification of a read-only value attempted)/ op/taint.t .. Failed 1/797 subtests (less 46 skipped subtests: 750 okay) Test Summary Report --- op/taint.t (Wstat: 0 Tests: 797 Failed: 1) Failed test: 592 Files=1, Tests=797, 0 wallclock secs ( 0.08 usr + 0.00 sys = 0.08 CPU) Result: FAIL On 12 August 2013 20:44, Father Chrysostomos spr...@cpan.org wrote: Steve Hay wrote: t/re/pat.t, t/re/pat_thr.t, t/op/taint.t and dist/threads-shared/t/clone.t all fail (in a threaded build) at c9f1f591ba751b24222b22f1f3a6799db0ab0d1b (Read-only COWs), the first of your three merge commits. Thank you. Can you tell which commit in the branch caused it? What do you get if you dump ${^TAINT} before and after the attempt to write to it? diff --git a/t/op/taint.t b/t/op/taint.t index b521408..479db1d 100644 --- a/t/op/taint.t +++ b/t/op/taint.t @@ -1643,8 +1643,10 @@ SKIP: { is(${^TAINT}, 1, '$^TAINT is on'); - +use Devel::Peek; +Dump ${^TAINT}; eval { ${^TAINT} = 0 }; +Dump ${^TAINT}; is(${^TAINT}, 1, '$^TAINT is not assignable'); like($@, qr/^Modification of a read-only value attempted/, 'Assigning to ${^TAINT} fails');
RE: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
Steve Hay wrote: t/re/pat.t, t/re/pat_thr.t, t/op/taint.t and dist/threads-shared/t/clone.t all fail (in a threaded build) at c9f1f591ba751b24222b22f1f3a6799db0ab0d1b (Read-only COWs), the first of your three merge commits. Thank you. Can you tell which commit in the branch caused it? What do you get if you dump ${^TAINT} before and after the attempt to write to it? diff --git a/t/op/taint.t b/t/op/taint.t index b521408..479db1d 100644 --- a/t/op/taint.t +++ b/t/op/taint.t @@ -1643,8 +1643,10 @@ SKIP: { is(${^TAINT}, 1, '$^TAINT is on'); - +use Devel::Peek; +Dump ${^TAINT}; eval { ${^TAINT} = 0 }; +Dump ${^TAINT}; is(${^TAINT}, 1, '$^TAINT is not assignable'); like($@, qr/^Modification of a read-only value attempted/, 'Assigning to ${^TAINT} fails');
Re: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
Steve Hay wrote: The tests fail in commit 3805b5fb04f8819edfb6bd5256d5eefd7b23e730 Thank you! clang is being âhelpfulâ and preventing me from spotting errors like this. Does the sprout/taint-failure branch make the fail- ures go away?
Re: Smoke [blead] v5.19.2-375-gf41cee3 FAIL(F) MSWin32 Win2000 SP4 (x86/1 cpu) {blead}
On 12 August 2013 23:54, Father Chrysostomos spr...@cpan.org wrote: Steve Hay wrote: The tests fail in commit 3805b5fb04f8819edfb6bd5256d5eefd7b23e730 Thank you! clang is being ‘helpful’ and preventing me from spotting errors like this. Does the sprout/taint-failure branch make the fail- ures go away? Yes, all tests pass on that branch :-) Thanks!