RFR (L): 7153771: array bound check elimination for c1

[Vladimir Kozlov]

I think it is acceptable tradeoff for performance gain we get.

Thanks,
Vladimir

On 2/25/13 7:03 AM, Roland Westrelin wrote:
>> Did C1 compilation time increased? And how much?
>
> Here are increase in compilation time I measure with the reference_server (change in Total compilation time as reported by -XX:+CITime, mean over 5 runs):
>
> jetstream: +2%
> scimark: +10%
> specjbb2000: +4%
> specjbb2005: +4%
> specjvm98: +2%
> volano25 (client): +8%
> volano25 (server): +1%
>
> Roland.
>