Ada bootstrap failures on various platforms

[Laurent GUERBY]

On Fri, 2004-04-16 at 22:28, Rainer Orth wrote:
> I've just started bootstraps of current mainline on several platforms and
> ran into Ada bootstrap failures almost everywhere (only alpha-dec-osf* is
> ok so far):

This reminds me of one bug fix I commited a while ago,
it was coming from:

x = f(f1(a), f2(b))

where stage1 was calling f1 before f2 and stage2 f2
before f1 and this was causing debuging info to be generated differently
(f1 and f2 where building standard types). The fix was to use
temporaries to fix the order of calls. May be some recent commited code
could have such a construct.

Laurent

2001-10-27  Laurent Guerby <guerby@acm.org>

	* trans.c (gigi): Fix non determinism leading to bootstrap
	comparison failures for debugging information.