«

Am writing a JNI wrapper class.
Are there any performance advantages of the JNI C routine calling a Java
method to set some variables all at once, or using SetIntField,
SetByteField, etc, to set the variables?

Which is normally recommended? Much less code to just call the method.

Thanks,

On Sun, 11 Sep 2005 18:06:41 -0400, "Jim"
<> wrote or quoted :

>Which is normally recommended? Much less code to just call the method.

the fewer the number of JNI calls the faster the code will be.
--
Canadian Mind Products, Roedy Green.
http://mindprod.com Again taking new Java programming contracts.

On Mon, 12 Sep 2005 03:38:29 +0000, Roedy Green wrote:

> On Sun, 11 Sep 2005 18:06:41 -0400, "Jim"
> <> wrote or quoted :
>
>>Which is normally recommended? Much less code to just call the method.
>
> the fewer the number of JNI calls the faster the code will be.

That is right. However, the more atomic the native call, the less time is
spend in the most tedious and complicated debugging process i know
(ideally, none).

I am all for speed, but when it comes to JNI, i usually wrap as tightly as
i can, often creating a (private, static) method per native function.

--
You can't run away forever,
But there's nothing wrong with getting a good head start.
--- Jim Steinman, "Rock and Roll Dreams Come Through"