Re: [PATCH] S/390: shifts and setmems with implicit ANDs

[Andreas Krebbel <krebbel1 at de dot ibm dot com>]

> Sure.  All you need is for the actual argument buried underneath
> the operator to be reloaded.  It doesn't appear in the insn_data
> that reload uses to do its job.
> 
> Frankly, I assume these patters are so rarely used that they just
> don't show up in bootstrap or testsuite in non-trivial ways.  You
> get lucky and local/global handle the pseudo in question, and it
> doesn't require reloading at all.

Thanks I will try to build a testcase for this.
> 
> > > > Almost the same applies to the padding byte operand of mvcle which is
> > > > always truncated to the 8 least significant bits.
> > > 
> > > Here you'd just have to use multiple patterns, iiuc.
> > That would be a pity.
> > Because we have a whole bunch of shift patterns we would need to duplicate
> > a lot of code for a very small performance improvement.
> 
> Huh?  What has setmem got to do with shifts?
Aeh good question. I was obviously still meditating about the shifts when I
wrote this. For the setmem stuff duplicating would be fine there aren't a lot
of these.

Bye,

-Andreas-