[m-dev.] diff: work around gcc global register problem

David Glen JEFFERY dgj at cs.mu.OZ.AU
Wed Nov 15 16:25:18 AEDT 2000


On 15-Nov-2000, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:
> > FYI, this bug also occurs in the random module in the library in one of
> > the HAL grades (asm_fast.gc.tr.rt.se.init, IIRC).
> > 
> > For now, I have just disabled the random library on the installation I do
> > at Monash. Do you want a bug report? It isn't affecting us at all, but maybe
> > the fix is as simple as the one in string.m, so you may want to fix it.
> 
> The change to string.m is a work-around, not a fix.
> If you already have a work-around that works for you, then be happy ;-)
> I'd rather spend time working on a fix than on another work-around.

Yes, I figured that was the case. Disabling `random' is causing no pain at
the moment, and until it does I won't bother trying to find another workaround.

> There's two possible fixes in this case: (1) fix gcc itself (the gcc
> maintainers have shown no interest in this) or (2) avoid using gcc
> global registers, e.g. by switching to the MLDS back-end.  Recently
> I've been working on improving the MLDS back-end, to make (2) more feasible.

At some stage, I suspect we will do (2). That may require a fair bit of
work, though, as the HAL implementation does a bit of low level trickery.


dgj
-- 
David Jeffery (dgj at cs.mu.oz.au) | If your thesis is utterly vacuous
PhD student,                    | Use first-order predicate calculus.
Dept. of Comp. Sci. & Soft. Eng.|     With sufficient formality
The University of Melbourne     |     The sheerist banality
Australia                       | Will be hailed by the critics: "Miraculous!"
                                |     -- Anon.
--------------------------------------------------------------------------
mercury-developers mailing list
Post messages to:       mercury-developers at cs.mu.oz.au
Administrative Queries: owner-mercury-developers at cs.mu.oz.au
Subscriptions:          mercury-developers-request at cs.mu.oz.au
--------------------------------------------------------------------------



More information about the developers mailing list