Paolo Bonzini <bonzini <at> gnu.org> writes:

> 
> Subconfigures are invoked with $SHELL, not with $CONFIG_SHELL.  And
> config.status is also created with
> 
> #! $SHELL
> 
> which will not pick the CONFIG_SHELL if it was given by the user on the
> command line.  So is this patch correct?

Yes, the concept looks correct.  However, I think you are missing 'export 
SHELL', which could lead to bugs on the shells where variables imported from 
the environment but subsequently modified are not passed on to child processes 
unless re-exported.  Please apply, once you've fixed that.

-- 
Eric Blake





Reply via email to