Re: [patch] Resubmission in an attempt to get someones attention

[Steve Ellcey <sje at cup dot hp dot com>]

This would have to be a platform specific test, right?  There is no way
to know what bits should wind up where in general though you can check
for a specific pattern on a specific platform.

I am still looking for someone to approve this patch
(http://gcc.gnu.org/ml/gcc-patches/2001-12/msg00717.html), I have a
volunteer to check it in if it is approved but it isn't clear to me if
the patch is not approved because of the lack of a test or if people are
just looking at the test issue and not examining the patch at all.

Steve Ellcey
sje@cup.hp.com

> You don't quite have to go that far.  Easiest is to do stuff like
> 
>   void target(struct badness s) { /* validate s has expected contents */ }
> 
>   int main ()
>   {
>     void (*f) (unsigned long) = (void *) target;
>     (*f) (0x123456789abcdef0);
>   }
> 
> So you're pretty sure that known bits wound up in known registers.