groovy-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jochen Theodorou <blackd...@gmx.org>
Subject Re: Problem with Groovy 1.8.9 and Java 1.8.0
Date Thu, 23 Apr 2015 11:01:39 GMT
Am 23.04.2015 11:54, schrieb C├ędric Champeau:
> Indy disables primitive optimizations. For pure maths computations, this
> can be an issue. It is also an issue if you use indy in combination with
> @CompileStatic, because the latter depends on the optimized paths.

well, for me that is a tad too general ;)

in my experience indy can be as fast as primopts, but it largely depends 
on the JVM side how well this goes. primopts really shine in a single 
aspect, that is a direct method call under certain circumstances. indy 
can be fast, but it cannot beat a direct method call in Java with a 
mutable callsite in my experience. And of course, primopts can have 
massive slowdowns, if the guarded code blocks are too small. In worst 
case ind can be much faster than primopts. But of course it is true, the 
general trend is towards primopts being faster

As for @CompileStatic.. it gets the information independent of if indy 
is used or not. It is more a question of if dynamic method calls and 
such are done with indy. But since @CompileStatic should not use them, 
there should be no trouble. If there is, it is most likely a bug

bye blackdrag

-- 
Jochen "blackdrag" Theodorou
blog: http://blackdragsview.blogspot.com/


Mime
View raw message