On Sat, Jun 27, 2020 at 01:55:19AM -0400, Michael Meissner wrote:
> On Thu, Jun 25, 2020 at 12:09:41PM -0500, Segher Boessenkool wrote:
> > On Thu, Jun 04, 2020 at 01:03:51PM -0400, Michael Meissner wrote:
> > > +/* { dg-final { scan-assembler-times {\mpaddi\M|\mpli|\mpla\M} 3 } } */
> >
> > Is there are reason you don't have \M on pli, or is that an oversight?
>
> Just an oversight.
>
> > Okay for trunk with that looked at, and the necessary "no future"
> > changes (and retesting ofc). Thanks!
>
> I simplified the test to use just the instruction (PLI) that is currently
> generated. I was just trying to be general to accomidate possible future code
> generation changes (i.e. instead of loading up the offset with PLI, we could
> potentionally do a PADDI to the base register that was loaded), but we can
> change the test if/when we change the compiler.
That is a good idea, yes. But, please post what you committed, then?
Segher
