2013/6/19 Ian Lance Taylor <iant@xxxxxxxxxx>: > And that ought to generate the best results, so it seems that everything might be OK. Yes, that options allowed me to generate the best results. How the gcc classifies the instructions priorities (latency and throughout)? In what source file? Geyslan Gregório Bem hackingbits.com @geyslangb br.linkedin.com/in/geyslan 2013/6/19 Ian Lance Taylor <iant@xxxxxxxxxx>: > On Wed, Jun 19, 2013 at 12:36 PM, Geyslan Gregório Bem > <geyslan@xxxxxxxxx> wrote: >> >> If you still see it as a missed-optimization bug, tell me that I'll report it. > > I may be misreading, but it seems that you got the best results using > -mtune=native -fschedule-insns-2 -O1, or -mtune=native -O3. And that > ought to generate the best results, so it seems that everything might > be OK. > > Ian
