[PATCH PR90240][RFC]Avoid scaling cost overflow by introducing scaling bound

Martin Liška mliska@suse.cz
Mon Apr 29 08:00:00 GMT 2019

On 4/27/19 6:12 AM, bin.cheng wrote:
> HI Martin, I remember you introduced comp_cost/cost_scaling to improve one case
> in spec2017.  Unfortunately I don't have access to the benchmark now, could you help
> verify that if this patch has performance issue on it please?  Thanks

Yes, it's 548.exchange2_r test from SPEC2017 and I can confirm it's fine after your patch.
I see the same speed with -Ofast -march=native.


More information about the Gcc-patches mailing list