task manager test segfaults with DBI >= 1.619

[dod]

Hello

With new DBI 1.619 (or 1.620), padre task_manager test segfaults. Here's the output of the test for padre 0.94:

$ perl -Ilib t/28_task_manager.t 
1..16
ok 1 - use Test::NoWarnings;
ok 2 - The object isa Padre::Wx::App
ok 3 - The object isa t::lib::Padre::NullWindow
ok 4 - The object isa Padre::TaskManager
ok 5 - No threads
ok 6 - ->start ok
ok 7 - Three threads exists
ok 8 - The object isa Padre::Task::Addition
ok 9 - ->schedule ok
ok 10 - ->{prepare} is false
ok 11 - ->{run}     is false
ok 12 - ->{finish}  is false
Segmentation fault

stack trace:

$ gdb /usr/bin/perl core
GNU gdb (GDB) 7.4.1-debian
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/bin/perl...(no debugging symbols found)...done.
[New LWP 27589]
[New LWP 27586]

warning: Can't read pathname for load map: Input/output error.
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `perl -Ilib t/38_task_manager.t'.
Program terminated with signal 11, Segmentation fault.
#0  0x00007f7623f865ab in ?? () from /usr/lib/perl5/auto/DBI/DBI.so
(gdb) bt
#0  0x00007f7623f865ab in ?? () from /usr/lib/perl5/auto/DBI/DBI.so
#1  0x00007f762f91ef86 in Perl_mg_dup () from /usr/lib/libperl.so.5.14
#2  0x00007f762f91e4d4 in ?? () from /usr/lib/libperl.so.5.14
#3  0x00007f762f91e940 in Perl_sv_dup_inc () from /usr/lib/libperl.so.5.14
#4  0x00007f762f91eae7 in ?? () from /usr/lib/libperl.so.5.14
#5  0x00007f762f91f060 in Perl_mg_dup () from /usr/lib/libperl.so.5.14
#6  0x00007f762f91e4d4 in ?? () from /usr/lib/libperl.so.5.14
#7  0x00007f762f91e940 in Perl_sv_dup_inc () from /usr/lib/libperl.so.5.14
#8  0x00007f762f91f162 in Perl_gp_dup () from /usr/lib/libperl.so.5.14
#9  0x00007f762f91e68e in ?? () from /usr/lib/libperl.so.5.14
#10 0x00007f762f91e940 in Perl_sv_dup_inc () from /usr/lib/libperl.so.5.14
#11 0x00007f762f9066aa in Perl_he_dup () from /usr/lib/libperl.so.5.14
#12 0x00007f762f91e040 in ?? () from /usr/lib/libperl.so.5.14
#13 0x00007f762f91e940 in Perl_sv_dup_inc () from /usr/lib/libperl.so.5.14
#14 0x00007f762f91f162 in Perl_gp_dup () from /usr/lib/libperl.so.5.14
#15 0x00007f762f91e68e in ?? () from /usr/lib/libperl.so.5.14
#16 0x00007f762f91e940 in Perl_sv_dup_inc () from /usr/lib/libperl.so.5.14
#17 0x00007f762f9066aa in Perl_he_dup () from /usr/lib/libperl.so.5.14
#18 0x00007f762f91e040 in ?? () from /usr/lib/libperl.so.5.14
#19 0x00007f762f91e8e0 in Perl_sv_dup () from /usr/lib/libperl.so.5.14
#20 0x00007f762f91e661 in ?? () from /usr/lib/libperl.so.5.14
#21 0x00007f762f91e8e0 in Perl_sv_dup () from /usr/lib/libperl.so.5.14
#22 0x00007f762f9238bc in perl_clone () from /usr/lib/libperl.so.5.14
#23 0x00007f762e1d009f in XS_threads_create () from /usr/lib/perl/5.14/auto/threads/threads.so
#24 0x00007f762f91472f in Perl_pp_entersub () from /usr/lib/libperl.so.5.14
#25 0x00007f762f90bbc6 in Perl_runops_standard () from /usr/lib/libperl.so.5.14
#26 0x00007f762f8a74ce in Perl_call_sv () from /usr/lib/libperl.so.5.14
#27 0x00007f762e1ce681 in ?? () from /usr/lib/perl/5.14/auto/threads/threads.so
#28 0x00007f762f1c0b50 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0
#29 0x00007f762ef0b90d in clone () from /lib/x86_64-linux-gnu/libc.so.6
#30 0x0000000000000000 in ?? ()
(gdb) quit

I have verified that the test works fine with DBI 1.618. Padre 0.94 and 0.96 give the same result, on i386 or amd64

[bowtie]

Test against trunk
with DBI => 1.620

                   Padre 0.97
Core...
                  osname linux
                archname i686-linux-thread-multi
            Distribution Ubuntu 11.10 
                  Kernel 3.0.0-17-generic
                    Perl 5.14.2
                 Threads 3
                     RAM (unsupported)
Wx...
                      Wx 0.9906
               WxWidgets 2.8.11
                 unicode 1
        Alien::wxWidgets 0.59
 Wx::Perl::ProcessStream 0.32
           Wx::Scintilla 0.3801
Other...
                     PPI 1.215
           Debug::Client 0.21_01
       PPIx::EditorTools 0.15_02
                  Config /usr/src/Padre/Padre/.padre

$ prove -lv t/38_task_manager.t 
t/38_task_manager.t .. 
1..16
ok 1 - use Test::NoWarnings;
ok 2 - The object isa Padre::Wx::App
ok 3 - The object isa t::lib::Padre::NullWindow
ok 4 - The object isa Padre::TaskManager
ok 5 - No threads
ok 6 - ->start ok
ok 7 - Three threads exists
ok 8 - The object isa Padre::Task::Addition
ok 9 - ->schedule ok
ok 10 - ->{prepare} is false
ok 11 - ->{run}     is false
ok 12 - ->{finish}  is false
Failed 4/16 subtests 

Test Summary Report
-------------------
t/38_task_manager.t (Wstat: 11 Tests: 12 Failed: 0)
  Non-zero wait status: 11
  Parse errors: Bad plan.  You planned 16 tests but ran 12.
Files=1, Tests=12,  3 wallclock secs ( 0.04 usr  0.01 sys +  0.94 cusr  0.08 csys =  1.07 CPU)
Result: FAIL

[bowtie]

see -> ​http://irclog.perlgeek.de/padre/2012-04-12#i_5431780

[bowtie]

Test against padre 0.96
with DBI => 1.620

                   Padre 0.96
Core...
                  osname linux
                archname i686-linux-thread-multi
            Distribution Ubuntu 11.10 
                  Kernel 3.0.0-17-generic
                    Perl 5.14.2
                 Threads 3
                     RAM 212.0MB
Wx...
                      Wx 0.9906
               WxWidgets 2.8.11
                 unicode 1
        Alien::wxWidgets 0.59
 Wx::Perl::ProcessStream 0.32
           Wx::Scintilla 0.3801
Other...
                     PPI 1.215
           Debug::Client 0.21_01
                  Config /home/kevin/.local/share/.padre

$ prove -lv t/38_task_manager.t 
t/38_task_manager.t .. 
1..16
ok 1 - use Test::NoWarnings;
ok 2 - The object isa Padre::Wx::App
ok 3 - The object isa t::lib::Padre::NullWindow
ok 4 - The object isa Padre::TaskManager
ok 5 - No threads
ok 6 - ->start ok
ok 7 - Three threads exists
ok 8 - The object isa Padre::Task::Addition
ok 9 - ->schedule ok
ok 10 - ->{prepare} is false
ok 11 - ->{run}     is false
ok 12 - ->{finish}  is false
Failed 4/16 subtests 

Test Summary Report
-------------------
t/38_task_manager.t (Wstat: 11 Tests: 12 Failed: 0)
  Non-zero wait status: 11
  Parse errors: Bad plan.  You planned 16 tests but ran 12.
Files=1, Tests=12,  2 wallclock secs ( 0.04 usr  0.00 sys +  0.78 cusr  0.14 csys =  0.96 CPU)
Result: FAIL

[bowtie]

I added more text to tests see r19061
from 38_task_manager.t

# Schedule the startup timer
Wx::Event::EVT_TIMER( $wxapp, TIMER_POSTINIT, \&startup );
my $timer1 = Wx::Timer->new( $wxapp, TIMER_POSTINIT );

# Schedule the failure timeout
Wx::Event::EVT_TIMER( $wxapp, TIMER_LASTRESORT, \&timeout );
my $timer2 = Wx::Timer->new( $wxapp, TIMER_LASTRESORT );

if you comment out each of above in turn,
it turns out that the error is generated by \&startup

sub startup {

    # Run the startup process
    ok( $manager->start, '->start, startup process ok' );
    Time::HiRes::sleep(1);
    is( scalar( threads->list ), 1, 'Three threads exists' );

    # Create the sample task
    my $addition = Padre::Task::Addition->new(
        x => 2,
        y => 3,
    );
    isa_ok( $addition, 'Padre::Task::Addition' );

    # Schedule the task (which should trigger it's execution)
    ok( $manager->schedule($addition), '->schedule startup ok' );
    is( $addition->{prepare}, 1, '->{prepare} startup is false' ); #should this be true as 1
    is( $addition->{run},     0, '->{run}     startup is false' );
    is( $addition->{finish},  0, '->{finish}  startup is false' );
}

the \&timeout works with out error

[dod]

Looks like this bug was fixed with DBI 1.621:

Fixed segmentation fault when a thread is created from

within another thread RT#77137, thanks to Dave Mitchell.

The test now pass on Debian amd64.

All the best

[bowtie]

I concur
hence closing