From @masak

<masak> someone else get a Parrot backtrace after running
t/spec/S10-packages/basic.rakudo ?
<moritz_> yes
* masak submits rakudobug

On Rakudo ba09b27 on Parrot r39584​:

$ perl6 t/spec/S10-packages/basic.rakudo
1..22
ok 1 - class test
not ok 2 - autovivification(?) for nested packages# TODO ticket uses
role; RT #​62900
not ok 3 - test for working .perl method# TODO RT #​65404
not ok 4 - Non-existant sub through package# TODO RT #​62970
not ok 5 - dot notation with package# TODO RT #​63432
not ok 6 - enum in package# TODO RT #​63826
not ok 7 - access to $?PACKAGE variable# TODO RT #​59484
not ok 8 - some WHO implementation# TODO ticket based only on class; RT #​60446
not ok 9 - WHO implementation with longname# TODO ticket based only on
class; RT #​60446
ok 10 - can refer to role using package
not ok 11 - since role is in package should not need package name#
TODO ticket based on module; RT #​63510
not ok 12 - two different packages should be two different Baz# TODO
ticket based on class(es) not package; RT #​65022
not ok 13 - two different packages should be two different EHomeE#
TODO ticket based on class(es) not package; RT #​65022
ok 14 - # SKIP ticket based on module; RT #​64876
not ok 15 - map in package# TODO RT #​64606
not ok 16 - use outer lexical# TODO ticket used class; RT #​61356
not ok 17 - use outer package var# TODO ticket used class; RT #​63588
not ok 18 - Should be able to update outer package var# TODO ticket
used class; RT #​63588
not ok 19 - simple package case that should not blow platform# TODO RT #​64204
not ok 20 - another simple package case that should not blow platform#
TODO RT #​64204
ok 21 - # SKIP RT #​64606
ok 22 - array initialization works fine in imported subs
# FUDGED!
src/call/pcc.c​:613​: failed assertion 'PObj_is_PMC_TEST(sig_pmc)'
Backtrace - Obtained 26 stack frames (max trace depth is 32).
  Parrot_do_check_events
  Parrot_confess
  Parrot_init_arg_op
  parrot_pass_args
  parrot_pass_args
  Parrot_run_meth_fromc_args
  Parrot_lib_mutablevar_load
  Parrot_lib_mutablevar_load
  Parrot_str_from_int
  enable_event_checking
  gc_register_pmc
  new_runloop_jump_point
  new_runloop_jump_point
  Parrot_runops_fromc_args
  Parrot_ex_throw_from_c
  Parrot_ex_throw_from_op_args
  Parrot_default_ro_get_vtable
  Parrot_str_from_int
  enable_event_checking
  gc_register_pmc
  new_runloop_jump_point
  new_runloop_jump_point
  Parrot_runops_fromc_args
  Parrot_runcode
  start
  start
Abort trap

From @pmichaud

This is a Parrot bug. Recently the Parrot team have been working to
clean up memory leaks in context handling, and as a result we're
starting to see "double free" errors on exit again. Running Rakudo with
Parrot's -G flag avoids the backtrace.

I'll leave this ticket open for now but mark it as "stalled" until we
can find out what's happening in Parrot's context handling.

Pm

The RT System itself - Status changed from 'new' to 'open'

@pmichaud - Status changed from 'open' to 'stalled'

From @kyleha

On Tue Jun 16 05​:23​:25 2009, masak wrote​:

<masak> someone else get a Parrot backtrace after running
t/spec/S10-packages/basic.rakudo ?

I've tried this with Rakudo 54cfe421ee2693a7c9b5a980ef4dc927313cb627,
and it doesn't happen anymore.

The RT System itself - Status changed from 'stalled' to 'open'

From @bbkr

Works also on * release.

$ /usr/bin/perl t/harness --fudge --keep-exit-code --icu=1 --jobs
t/spec/S10-packages/basic.t
t/spec/S10-packages/basic....ok
  6/51 skipped​: various reasons
All tests successful, 6 subtests skipped.
Files=1, Tests=51, 10 wallclock secs ( 7.92 cusr + 0.22 csys = 8.14 CPU)

@bbkr - Status changed from 'open' to 'resolved'