Hi Damyan,
Damyan Ivanov <d...@debian.org> writes:
> Strace log available at
> http://ktnx.net/~dam/hidden/lai0Aede/strace.log.xz
25434 read(5, "=head1 NAME\n\nAnyEvent::Impl::Perl - AnyEvent adaptor for
AnyEvent's pure perl AnyEvent::Loop\n\n=head1 SYNOPSIS\n\n use AnyEvent;\n
use AnyEvent::Loop;\n # this module gets loaded automatically as
required\n\n=head1 DESCRIPTION\n\nThis module provides transparent support for
AnyEvent in case no other\nevent loop could be found or loaded.\n\nIf you want
to use this module instead of autoloading another event loop\nyou can simply
load L<AnyEvent::Loop> before creating the first watcher.\n\nNaturally, it
supports all features of AnyEvent.\n\nSee L<AnyEvent::Loop> for more details on
performance characteristics.\n\n=cut\n\npackage AnyEvent::Impl::Perl;\n\nuse
AnyEvent (); BEGIN { AnyEvent::common_sense }\nuse AnyEvent::Loop;\n\nour
$VERSION = $AnyEvent::VERSION;\n\n# time() is provided via
AnyEvent::Base\n\n*AE::now = \\&AnyEvent::Loop::now;\n*AE::now_update =
\\&AnyEvent::Loop::now_update;\n*AE::io =
\\&AnyEvent::Loop::io;\n*AE::timer = \\&AnyEvent::Loop::timer;\n*AE::idle
= \\&AnyEvent::Loop::idle;\n*_poll =
\\&AnyEvent::Loop::one_event;\n*loop = \\&AnyEvent::Loop::run; #
compatibility with AnyEvent < 6.0\n\nsub now { $AnyEvent::Loop::NOW
}\nsub now_update { AE::now_update }\n\nsub
AnyEvent::CondVar::Base::_wait {\n AnyEvent::Loop::one_event until exists
$_[0]{_ae_sent};\n}\n\nsub io {\n my (undef, %arg) = @_;\n\n
AnyEvent::Loop::io $arg{fh}, $arg{poll} eq \"w\", $arg{cb}\n}\n\nsub timer {\n
my (undef, %arg) = @_;\n\n AnyEvent::Loop::timer $arg{after},
$arg{interval}, $arg{cb}\n}\n\nsub idle {\n my (undef, %arg) = @_;\n\n
AnyEvent::Loop::idle $arg{cb}\n}\n\n=head1 SEE ALSO\n\nL<AnyEvent>.\n\n=head1
AUTHOR\n\n Marc Lehmann <schm...@schmorp.de>\n
http://anyevent.schmorp.de\n\n=cut\n\n1\n\n";, 8192) = 1710
25434 read(5, "", 8192) = 0
25434 close(5) = 0
25434 getuid() = 0
25434 geteuid() = 0
25434 getgid() = 0
25434 getegid() = 0
25434 --- SIGSEGV (Segmentation fault) @ 0 (0) ---
25433 <... select resumed> ) = ? ERESTARTNOHAND (To be restarted)
25433 --- SIGCHLD (Child exited) @ 0 (0) ---
25433 rt_sigreturn(0) = -1 EINTR (Interrupted system call)
25433 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
25433 rt_sigaction(SIGCHLD, NULL, {0x2acdd9b3e700, [], SA_RESTORER,
0x2acdda33c210}, 8) = 0
25433 write(10, "\0", 1) = 1
25433 rt_sigprocmask(SIG_UNBLOCK, [CHLD], NULL, 8) = 0
25433 select(16, [3 4 8], [], NULL, {1, 898767}) = 1 (in [8], left {1, 898765})
25433 read(8, "\0", 9) = 1
25433 wait4(-1, [{WIFSIGNALED(s) && WTERMSIG(s) == SIGSEGV}], WNOHANG, NULL) =
25434
25433 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
25433 rt_sigaction(SIGCHLD, {SIG_DFL, [], SA_RESTORER, 0x2acdda33c210},
{0x2acdd9b3e700, [], SA_RESTORER, 0x2acdda33c210}, 8) = 0
25433 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
25433 write(7, "# exited with exit code 11\n", 27) = 27
25422 <... select resumed> ) = 1 (in [5])
25433 wait4(-1, <unfinished ...>
25422 read(5, <unfinished ...>
25433 <... wait4 resumed> 0x7fff09cebf7c, WNOHANG, NULL) = -1 ECHILD (No child
processes)
25422 <... read resumed> "# exited with exit code 11\n", 65536) = 27
25422 write(2, "# exited with exit code 11\n", 27) = 27
25422 select(8, [3 5], NULL, NULL, NULL <unfinished ...>
25433 write(6, "not ok 1 - plugin messages match expectation\n", 45) = 45
25422 <... select resumed> ) = 1 (in [3])
25422 read(3, "not ok 1 - plugin messages match expectation\n", 65536) = 45
25433 write(7, "\n", 1) = 1
25433 write(7, "# Failed test 'plugin messages match expectation'\n", 52) = 52
25433 write(7, "# at t/004-smtp.t line 80.\n", 29) = 29
25422 write(1, "\rt/004-smtp.t ............ 1/? ", 31) = 31
For some reason, the plugin process segfaults. I don’t know what’s
causing this, but I _really_ doubt that it’s any of kanla’s code. It
rather sounds like your AnyEvent setup is somehow broken. Maybe the bug
should be reassigned.
Anyway, please enable coredumps and provide a backtrace so that we can
figure out what is crashing here.
Also, maybe try installing libev4, which should make AnyEvent use a
different implementation than AnyEvent::Impl::Perl
> Sorry about that. I assumed that you are somehow subscribed to the
> package, either via the PTS, or as a maintainer.
I am, but gmail decided your mail is spam, and I don’t check my spam
folder regularly. Adressing me directly helps with that, I think :).
--
Best regards,
Michael
--
To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
