Re: ponie's copy of Parrot doesn't build on Windows

Brent Dax wrote:
> 
> A few days ago, a bug was found in one of Parrot's build-time utilities
> that was causing multiple copies of the C compiler to start up on
> Windows.  (Apparently exec() on Windows has some weird asynchronous
> behavior.)  These multiple compilers would then try to open a file, but
> only one would succeed, causing most of the compilations to fail.
> 
> This was fixed in version 1.4 of tools/dev/cc_flags.pl.  I suggest we
> update Ponie's copy of Parrot to include that version.
> 
> If it helps, I've pasted a diff below my sig that will do the same
> thing.  I don't know if applying it is the right thing to do,
> though--there may be some trick with CVS that would be better.
> 
> --Brent Dax <[email protected]>
> Perl and Parrot hacker
> 
> "Yeah, and my underwear is flame-retardant--that doesn't mean I'm gonna
> set myself on fire to prove it."
> 
> --- parrot\tools\dev\cc_flags.pl        (revision 12)
> +++ parrot\tools\dev\cc_flags.pl        (working copy)
> @@ -67,4 +67,4 @@
>  }
> 
>  print "@ARGV\n";
> -exec @ARGV;
> +exit system(@ARGV)/256;

Considering that the parent process *might* want to know if the command
died from signals, as well as maybe from exit()ing, perhaps it should be
something like:

   exec @ARGV unless $^O eq "MSWin32";
   my $n = system(@ARGV);
   die "Error calling system(@ARGV): $!" if $n == -1;
   kill $n & 255, $$ if $n & 255;
   exit $n/256;
   __END__

-- 
$a=24;split//,240513;s/\B/ => /for@@=qw(ac ab bc ba cb ca
);{push(@b,$a),($a-=6)^=1 for 2..$a/6x--$|;print "$@[$a%6
]\n";((6<=($a-=6))?$a+=$_[$a%6]-$a%6:($a=pop @b))&&redo;}