Loading...

XML

Word

Printable

Issue	Fix Version	Assignee	Priority	Status	Resolution	Resolved In Build
JDK-8194603	11	Vladimir Ivanov	P3	Resolved	Fixed	b01

There's a performance regression in 9 (compared to 8) in Math.pow(x, 2.0D).

Math.pow() was rewritten in 9 (see ~~JDK-8145688~~), but it doesn't have special case for power-of-two in C2 intrinsic implementation.

Original report:
https://www.reddit.com/r/java/comments/7bbh72/i_think_i_found_piece_of_code_which_is_8_times/

backported by

JDK-8194603 C2: missing strength reduction of Math.pow(x, 2.0D) to x*x

duplicates

JDK-8190872 Regression in Math.pow for powers of 2

relates to

JDK-8029302 Performance regression in Math.pow intrinsic

JDK-8145688 update for x86 pow in the math lib

JDK-8209439 C2 library_call can potentially ignore Math.pow intrinsic or use null pointer