Re: [Gnu-arch-users] What's the rationale behind the cmd-build-config.c restrictions?

[Andrew Suffield]

On Fri, Jan 28, 2005 at 09:10:52AM -0600, Rob Browning wrote:
> Andrew Suffield <address@hidden> writes:
> 
> > On Thu, Jan 27, 2005 at 09:46:41AM -0600, Rob Browning wrote:
> >> 
> >> In cmd-build-config.c we have:
> >> 
> >>     if (!arch_valid_config_name (config))
> >>       {
> >>         safe_printfmt (2, "%s: invalid config name (%s)\n",
> >>                        argv[0], config);
> >>         exit (2);
> >>       }
> >> 
> >>     tree_root = arch_tree_root (0, dir, 0);
> >> 
> >>     if (!tree_root)
> >>       {
> >>         safe_printfmt (2, "%s: not in project tree (%s)\n",
> >>                        argv[0], dir);
> >>         exit (1);
> >>       }
> >> 
> >> What's the rationale behind these restrictions?
> >
> > Nothing relevant, it's a bug. It's a very old bug that should have
> > been fixed a long time ago.
> 
> Do you have any idea if the fix is likely to be as simple as removing
> that bit, or are there probably other changes that would need to be
> made?

If you knock out that one then an invariant() call deeper in will trap
you. If you remove them both then it'll work. There's also a similar,
even more broken test in catcfg.

-- 
  .''`.  ** Debian GNU/Linux ** | Andrew Suffield
 : :' :  http://www.debian.org/ |
 `. `'                          |
   `-             -><-          |

signature.asc
Description: Digital signature