Bug#528154: libpadwalker-perl: using re.pl with SOAP::Lite causes a segfault

Jonathan Yu jonathan.i.yu at gmail.com
Mon May 11 13:36:05 UTC 2009


Hi:

I have been trying to track down this bug, however in the bug tracker
it looks like someone beat me to the punch:

http://rt.cpan.org/Public/Bug/Display.html?id=41710

Also, here is a more detailed backtrace from 'debugperl reproduce.pl'
and 'gdb debugperl core':

(gdb) bt full
#0  Perl_av_fetch (my_perl=0x8da0008, av=0x0, key=0, lval=0) at av.c:200
No locals.
#1  0xb7c56f49 in get_closed_over (cv=0x986d19c, hash=0x9910dec,
indices=0x98c57f4) at PadWalker.xs:362
        i = <value optimized out>
        pad_namelist = (AV *) 0x80c9089
        pad_vallist = (AV *) 0x8da0008
#2  0xb7c57488 in XS_PadWalker_closed_over (my_perl=0x8da0008,
cv=0x92fa6c4) at PadWalker.xs:501
        cv = (CV *) 0x986d19c
        ret = (HV *) 0x9910dec
        targs = (HV *) 0x98c57f4
        sp = (SV **) 0x9034b04
        ax = <value optimized out>
#3  0x080e0bf1 in Perl_pp_entersub (my_perl=0x8da0008) at pp_hot.c:2850
        markix = 0
        sp = (SV **) 0x9034b08
        sv = (SV *) 0x92fa394
        gv = (GV *) 0x935926c
        cv = (CV *) 0x92fa6c4
        cx = <value optimized out>
        gimme = 1
        hasargs = 1 '\001'
#4  0x080a58df in Perl_runops_debug (my_perl=0x8da0008) at dump.c:1931
No locals.
#5  0x080d6d61 in perl_run (my_perl=0x8da0008) at perl.c:2391
        oldscope = 1
        ret = <value optimized out>
        cur_env = {je_prev = 0x8da017c, je_buf = {{__jmpbuf =
{-1208520716, -1074900384, 134626704, -1074900456, 362252199,
        -739819832}, __mask_was_saved = 0, __saved_mask = {__val =
{3085678909, 3220066768, 3086843467, 3086907432, 3086567696, 1,
          1, 0, 134580624, 4, 0, 136880420, 1, 3086468792, 148504584,
3220066912, 148505124, 3220066840, 3086866816, 148522884,
          148504584, 3220066912, 148504584, 3220066840, 135121319,
148504584, 136524852, 5, 10, 0, 0, 134949888}}}}, je_ret = 0,
  je_mustcatch = 0 '\0'}
#6  0x08063f45 in main (argc=2, argv=0xbfee52e4, env=0xbfee52f0) at
perlmain.c:113
        exitstatus = <value optimized out>

Not that the backtrace really helps in this case. It does look like a
bug in PadWalker. Let's try applying that patch and see if it fixes
this script?

Hope this helps,

Jonathan

On Mon, May 11, 2009 at 12:27 AM, Ryan Niebur <ryanryan52 at gmail.com> wrote:
> attached is a smaller test case, which just involves
> Data::Dump::Streamer and SOAP::Lite.
>
> libdata-dump-streamer-perl is version 2.09-1
>
> --
> _________________________
> Ryan Niebur
> ryanryan52 at gmail.com
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.9 (GNU/Linux)
>
> iEYEARECAAYFAkoHqTYACgkQMihv+PacasXmmwCeL0XgkrUEMH6/oL7YlTVetX6X
> UP4An1nn2cVhYISgItf4J7c1sh9bqOCL
> =1LlO
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> pkg-perl-maintainers mailing list
> pkg-perl-maintainers at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/pkg-perl-maintainers
>





More information about the pkg-perl-maintainers mailing list