diotalevi has asked for the wisdom of the Perl Monks concerning the following question:

I am tracking down a bug in perl, B, B::Utils, or B::Generate. The following code leaves B::Utils::walkallops_simple (and all the other plain-perl subs in B::Utils) with their cop_warnings pointer set to an invalid value. If someone else out there gets a different result from the first print (so something other than 0c), or if the file/line being reported changes, or if it doesn't die - please let me know.

The long string of hex numbers is the hex encoding of the in-memory opcode so that will change since your pointer addresses will be different and such. Don't just compare that long number and report that it is different for you because that is normal and to be expected. If it doesn't die or dies differently or dies elsewhere, or if $w is not 0x0c then let me know. (also try commenting return if $w eq "00"; if it doesn't die. That just keeps a lot of nonsense of stuff that is supposed to work from printing input)

# does *NOT* die when this is not imported - importing
# and using this is what triggers the COP->cop_warnings to be bad.
use B::Utils qw'walkallops_simple'; 
use B::Generate; 

$\ = "\n";
$| = 1;

print "$_ " . $_->VERSION for qw( B B::Utils B::Generate );
B::Utils::walkallops_simple( \ &cb );

sub cb
{
    my $op = shift;
    my $addr = $$op;
    return unless $op=~/COP/;

    my $w = unpack('H*', unpack "P", pack 'L', $addr + 48);
    return if $w eq "00";
    print $w;
    print unpack 'H*', unpack 'P56', pack 'L', $addr;
    print $op->file . ":" . $op->line;

    print $op->warnings; # SEGFAULT HERE
}

[download]
B 1.02
B::Utils 0.04
B::Generate 1.06
0c
28031d0848031d08b0ef0b0800000000ae008411010200000000000054d61508f42016
+086b010000000000001e0100000c00000000000000
/home/josh/perl5.8.3/lib/site_perl/5.8.3/B/Utils.pm:286
Segmentation fault

[download]

Summary of my perl5 (revision 5.0 version 8 subversion 3) configuratio
+n:
  Platform:
    osname=linux, osvers=2.6.3, archname=i686-linux
    uname='linux lik.grenekatz.org 2.6.3 #5 sun feb 29 22:24:31 local 
+time zone must be set--see zic manu i686 amd athlon(tm) xp 2500+ auth
+enticamd gnulinux '
    config_args='-Dmksymlinks -Dprefix=/home/josh/perl5.8.3 -de'
    hint=previous, useposix=true, d_sigaction=define
    usethreads=undef use5005threads=undef useithreads=undef usemultipl
+icity=undef
    useperlio=define d_sfio=undef uselargefiles=define usesocks=undef
    use64bitint=undef use64bitall=undef uselongdouble=undef
    usemymalloc=n, bincompat5005=undef
  Compiler:
    cc='cc', ccflags ='-fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FIL
+E_OFFSET_BITS=64',
    optimize='-O3',
    cppflags='-fno-strict-aliasing -fno-strict-aliasing -D_LARGEFILE_S
+OURCE -D_FILE_OFFSET_BITS=64'
    ccversion='', gccversion='3.3.2 20031218 (Gentoo Linux 3.3.2-r5, p
+ropolice-3.3-7)', gccosandvers=''
    intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
    d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=1
+2
    ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t',
+ lseeksize=8
    alignbytes=4, prototype=define
  Linker and Libraries:
    ld='cc', ldflags =' -L/usr/local/lib'
    libpth=/usr/local/lib /lib /usr/lib
    libs=-lnsl -lndbm -lgdbm -ldb -ldl -lm -lcrypt -lutil -lc
    perllibs=-lnsl -ldl -lm -lcrypt -lutil -lc
    libc=/lib/libc-2.3.2.so, so=so, useshrplib=false, libperl=libperl.
+a
    gnulibc_version='2.3.2'
  Dynamic Linking:
    dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-rdynami
+c'
    cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib'


Characteristics of this binary (from libperl):
  Compile-time options: USE_LARGE_FILES
  Built under linux
  Compiled at Mar  1 2004 14:04:37
  @INC:
    /home/josh/perl5.8.3/lib/5.8.3/i686-linux
    /home/josh/perl5.8.3/lib/5.8.3
    /home/josh/perl5.8.3/lib/site_perl/5.8.3/i686-linux
    /home/josh/perl5.8.3/lib/site_perl/5.8.3
    /home/josh/perl5.8.3/lib/site_perl
    .

[download]

Steve_p - changed pre tags to code tags to get long lines to wrap as suggested
flux8 - same fix but for SOPW view

Replies are listed 'Best First'.
Re: Sounding a segfault in B, B::Utils, or B::Generate - testers requested
by McMahon (Chaplain) on Jun 10, 2004 at 22:22 UTC
    I'll confirm that your bug exists in earlier versions. This output is from FreeBSD 4.9; Perl 5.8.0. 
    B 1.01
B::Utils 0.04
B::Generate 1.06
0c
60571a0880571a0860bd090800000000ae004811010200000000000060f3110888f211
+0856010000000000001e0100000c00000000000000
/usr/local/lib/perl5/site_perl/5.8.0/B/Utils.pm:286
Segmentation fault (core dumped)

    [download]
[reply]
[d/l]
Re: Sounding a segfault in B, B::Utils, or B::Generate - SOLVED
by diotalevi (Canon) on Jun 11, 2004 at 18:47 UTC
[reply]

Back to Seekers of Perl Wisdom