Re: Broken code in checks/binaries
On Tue, May 20, 2008 at 03:57:09AM +0200, Frank Lichtenheld wrote:
> On Mon, May 19, 2008 at 06:45:52PM -0700, Russ Allbery wrote:
> > Frank Lichtenheld <djpig@debian.org> writes:
> > > checks/binaries contains the following piece of code:
> > > 127 sub lib_soname_path {
> > > 128 my (@paths) = @_;
> > > 129 foreach my $path (@paths) {
> > > 130 return 1 if $path =~ m%^(\.?/)?lib/[^/]+$%;
> > > 131 return 1 if $path =~ m%^(\.?/)?usr/lib/[^/]+$%;
> > > 132 return 1 if $path =~ m%^(\.?/)?usr/X11R6/lib/[^/]+$%;
> > > 133 return 1 if $path =~ m%^(\.?/)?lib/libnss_[^.]+\.so(\.[0-9]+)$%;
> > > 134 }
> > > 135 return 0;
> > > 136 }
> > >
> > > Obviously line 133 is never reached because 131 already matches such
> > > paths. And return 1 seems to be wrong thing to do for line 133
> > > anyway. What is line 133 supposed to do?
> >
> > return 0 and be above all of the rest, I think. Good catch. I wonder
>
> Not "next" instead of "return"?
>
> > what I was thinking there. The goal is to not require NSS modules to have
> > SONAMEs or shlibs, since they're weird special cases.
Russ, I was waiting for you to fix this. Did you do the same in reverse or
did you simply forget?
Gruesse,
--
Frank Lichtenheld <djpig@debian.org>
www: http://www.djpig.de/
Reply to: