From @timbunce

Created by @timbunce

I'm trying to improve how NYTProf works with Class​::MOP/Moose
and other code that uses #line directives.

Here's a test script​:

------snip------
BEGIN { $^P |= 0x400|0x100|0x10 }
sub a { 0 } 1;
#line 1 "hash-line-first"
sub b { 1 }
#line 1 "hash-line-second"
sub c { 2 }
eval qq{#line 1 "hash-line-eval"
sub d { 3 } 1
} or die;

a(); b(); c(); d();

END {
  for my $file (sort grep { m/_</ } keys %{'main​::'}) {
  print "File​: $file ($$file) Source​:\n";
  my $src = \@​$file;
  print for map { "\t$_​: $src->[$_]" } 1..@​$src;
  print "\n";
  }
  print "Sub​: $_ => $DB​::sub{$_}\n" for sort keys %DB​::sub;
}
------snip------

and the output (using 5.12.2)​:

------snip------
File​: _<(eval 1)[hash-line-second​:2] ((eval 1)[hash-line-second​:2]) Source​:
  1​: #line 1 "hash-line-eval"
  2​: sub d { 3 } 1
  3​:
  4​: ; 5​:
File​: _<hash-line-eval (hash-line-eval) Source​:

File​: _<hash-line-first (hash-line-first) Source​:
  1​: sub b { 1 }
  2​: #line 1 "hash-line-second"
  3​:
File​: _<hash-line-second (hash-line-second) Source​:
  1​: sub c { 2 }
  2​: eval qq{#line 1 "hash-line-eval"
  3​: sub d { 3 } 1
  4​: } or die;
  5​:
  6​: a(); b(); c(); d();
  7​:
  8​: END {
  9​: for my $file (sort grep { m/_</ } keys %{'main​::'}) {
  10​: print "File​: $file ($$file) Source​:\n";
  11​: my $src = \@​$file;
  12​: print for map { "\t$_​: $src->[$_]" } 1..@​$src;
  13​: print "\n";
  14​: }
  15​: print "Sub​: $_ => $DB​::sub{$_}\n" for sort keys %DB​::sub;
  16​: }
  17​:
File​: _<hashline.pl (hashline.pl) Source​:

File​: _<mro.c (mro.c) Source​:

File​: _<perl.c (perl.c) Source​:

File​: _<perlio.c (perlio.c) Source​:

File​: _<perlmain.c (perlmain.c) Source​:

File​: _<universal.c (universal.c) Source​:

Sub​: main​::END => hash-line-second​:8-16
Sub​: main​::a => hashline.pl​:2-2
Sub​: main​::b => hash-line-first​:1-1
Sub​: main​::c => hash-line-second​:1-1
Sub​: main​::d => hash-line-eval​:1-1
------snip------

Things to note​:

a. There's no source code saved for hashline.pl itself because $^P is
  set too late to capture the source. perl -d ... works. Not a problem.

b. #line 1 "hash-line-first" and #line 1 "hash-line-second" work as expected.

c. The subroutines all get recorded in %DB​::sub with the correct filename.

d. The #line inside the string eval does not work as expected.
  The initial *{"_<(eval 1)[...]"} is created and populated, fine.
  Then when it sees the #line it creates a *{"_<hash-line-eval"} glob
  but doesn't store the source code in the array. That's a bug.

So %DB​::sub says subroutine d() was defined in the file "hash-line-eval"
but there's no source code in @​{"_<hash-line-eval"}.

The upshot is that the portion of the source code of a string eval after
a #line directive isn't where it should be, so NYTProf can't find it.
The debugger won't be able to find it either.

Tim.

Flags:
    category=core
    severity=medium

Site configuration information for perl 5.12.2:

Configured by timbo at Sun Oct  3 16:42:33 IST 2010.

Summary of my perl5 (revision 5 version 12 subversion 2) configuration:
  Commit id: 7a3b65c9d99f69553fffe01f73d49fe9abf95120
  Platform:
    osname=darwin, osvers=10.4.0, archname=darwin-thread-multi-2level
    uname='darwin timac.local 10.4.0 darwin kernel version 10.4.0: fri apr 23 18:28:53 pdt 2010; root:xnu-1504.7.4~1release_i386 i386 '
    config_args='-des -Doptimize=-g -DEBUGGING=both -Dusethreads -Dusemultiplicity -Dusesitecustomize -Dusedevel -Uversiononly -Dprefix=/usr/local/perl512-dev'
    hint=recommended, useposix=true, d_sigaction=define
    useithreads=define, usemultiplicity=define
    useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef
    use64bitint=undef, use64bitall=undef, uselongdouble=undef
    usemymalloc=n, bincompat5005=undef
  Compiler:
    cc='cc', ccflags ='-fno-common -DPERL_DARWIN -no-cpp-precomp -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include -I/opt/local/include',
    optimize='-g',
    cppflags='-no-cpp-precomp -fno-common -DPERL_DARWIN -no-cpp-precomp -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include -I/opt/local/include'
    ccversion='', gccversion='4.2.1 (Apple Inc. build 5664)', gccosandvers=''
    intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
    d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=16
    ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t', lseeksize=8
    alignbytes=8, prototype=define
  Linker and Libraries:
    ld='env MACOSX_DEPLOYMENT_TARGET=10.3 cc', ldflags =' -fstack-protector -L/usr/local/lib -L/opt/local/lib'
    libpth=/usr/local/lib /opt/local/lib /usr/lib
    libs=-lgdbm -ldbm -ldl -lm -lutil -lc
    perllibs=-ldl -lm -lutil -lc
    libc=/usr/lib/libc.dylib, so=dylib, useshrplib=false, libperl=libperl.a
    gnulibc_version=''
  Dynamic Linking:
    dlsrc=dl_dlopen.xs, dlext=bundle, d_dlsymun=undef, ccdlflags=' '
    cccdlflags=' ', lddlflags=' -bundle -undefined dynamic_lookup -L/usr/local/lib -L/opt/local/lib -fstack-protector'

Locally applied patches:
    


@INC for perl 5.12.2:
    /usr/local/perl512-dev/lib/site_perl/5.12.2/darwin-thread-multi-2level
    /usr/local/perl512-dev/lib/site_perl/5.12.2
    /usr/local/perl512-dev/lib/5.12.2/darwin-thread-multi-2level
    /usr/local/perl512-dev/lib/5.12.2
    /usr/local/perl512-dev/lib/site_perl/5.12.0/darwin-thread-multi-2level
    /usr/local/perl512-dev/lib/site_perl/5.12.0
    /usr/local/perl512-dev/lib/site_perl
    .


Environment for perl 5.12.2:
    DYLD_LIBRARY_PATH=:/usr/local/pgsql/lib/:/opt/local/lib/mysql5/mysql
    HOME=/Users/timbo
    LANG=en_IE.UTF-8
    LANGUAGE (unset)
    LC_ALL=en_IE.UTF-8
    LD_LIBRARY_PATH (unset)
    LOGDIR (unset)
    PATH=/usr/local/pgsql/bin:/Users/timbo/bin:/usr/local/perl512-dev/bin:/Users/timbo/perl6/rakudo/parrot_install/bin:/usr/local/mysql/bin:/usr/local/bin:/opt/local/bin:/opt/local/sbin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/X11/bin
    PERLCRITIC=/Users/timbo/.setdev/perlcriticrc
    PERLTIDY=/Users/timbo/.setdev/perltidyrc
    PERL_BADLANG (unset)
    SHELL=/bin/bash

From @cpansprout

On Fri Nov 19 04​:11​:52 2010, timbo wrote​:

but there's no source code in @​{"_<hash-line-eval"}.

I can reproduce this with threads, but not without.

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

From @timbunce

On Fri, Nov 19, 2010 at 04​:11​:52AM -0800, Tim Bunce wrote​:

# <URL​: http​://rt.perl.org/rt3/Ticket/Display.html?id=79442 >

Any chance this could get looked into before 5.14?

Tim.

From @cpansprout

On Tue Jan 04 05​:48​:38 2011, timbo wrote​:

On Fri, Nov 19, 2010 at 04​:11​:52AM -0800, Tim Bunce wrote​:

# <URL​: http​://rt.perl.org/rt3/Ticket/Display.html?id=79442 >

Any chance this could get looked into before 5.14?

Yes. :-)
http​://perl5.git.perl.org/perl.git/commitdiff/8818d40

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