Re: [gomp] New tests

[Jakub Jelinek <jakub at redhat dot com>]

On Mon, Oct 17, 2005 at 12:44:14PM -0400, Diego Novillo wrote:
> > So, IMHO either we should translate it into GOMP_flush function call and
> > do the __sync_synchronize (); in libgomp, or use something equally
> > strong to a non-pure function call to prevent optimizations across it.
> >
> Yeah, that'd be the easiest approach.

Maybe for flush where just a few variables are listed simple
__asm __volatile ("" : "=m" (foo) : "m" (foo), "g" (&foo));
__sync_synchronize ();
would be enough too (all variables for which it makes sense to flush them
must live in memory, otherwise other threads couldn't get at them).

Another thing, I'm getting tcl tracebacks when running make
check-target-libgomp, as one of {appendix-a.exp,dg.exp} is trying to
unset lang_* variables that were already unset by the other *.exp script.

I'm attaching 2 patches to solve this, the one that removes the special
appendix-a/*.exp seems preferrable to me (as it really doesn't need anything
different from the dg.exp).  Ok?

	Jakub

Attachment: X1
Description: Text document

Attachment: X2
Description: Text document