On Mon, Nov 21, 2016 at 01:27:59PM -0500, Michael Meissner wrote: > On Fri, Nov 18, 2016 at 05:07:21PM -0600, Segher Boessenkool wrote: > > On Fri, Nov 18, 2016 at 05:52:12PM -0500, Michael Meissner wrote: > > > On Fri, Nov 18, 2016 at 04:43:40PM -0600, Segher Boessenkool wrote: > > > > Could you also test with reload please? Just LE is enough I guess. > > > > We'd like to keep reload working for GCC 7 at least, and these cost > > > > prefixes tend to break mov patterns :-/ > > > > > > Argh, I guess you are right, but then if reload doesn't work, I will > > > likely > > > submit the patch where there are three different movdi's (one for 32-bit > > > without the change, one for 64-bit with reload, and one for 64-bit with > > > lra). > > > I would prefer not to do that. > > > > Let's hope it just works :-) > > I did test it over the weekend. > > 29 of the 30 spec 2006 benchmarks currently build with reload (gamess fails). > The same 29 build and run with the new patch. Like the patch under LRA, there > are no regressions in performance, and one FP benchmark is faster. > > Under LRA, sphinx3 is 2.5% faster (compared to LRA without the patch). > > Under reload, sphinx3 is roughly the same performance, but calculix is 3.8% > faster.
Great, thanks for testing. > Can I apply the patch? Okay, if you change the changelog to say what the patch actually does ;-) And please watch for fallout. Segher
