Re: withGCC unset

[Ralf Wildenhues]

* Peter O'Gorman wrote on Tue, Jan 15, 2008 at 07:36:48AM CET:
> Ralf Wildenhues wrote:
> > * Peter O'Gorman wrote on Mon, Jan 14, 2008 at 11:08:58PM CET:
> >> This patch should fix issues with withGCC.
> >>
> >> Ok for HEAD?
> > 
> > My only issue with this is: are there situations where we may need to
> > know both whether the current tag compiler is from GCC, and the C
> > compiler is gcc, at the same time?  Or: whether any of the tests that
> > currently use $GCC or $withGCC, actually need to know what the C
> > compiler is.
> 
> I do not think this is the case, however it does lead to my next (rather
> more controversial patch).

Intended before 2.2?

> The only place where this might occur is in _LT_SYS_DYNAMIC_LINKER,
> which we call for every tag, but which sets no tag specific variables.
> 
> I would like to reduce this to just one call to _LT_SYS_DYNAMIC_LINKER
> for the default tag.

Hmm.  That sounds like a quite intrusive patch.  But shouldn't say
anything about stuff I haven't seen yet.

Cheers,
Ralf