I'd probably recommend trying a code style aspect if you can, rather than
annotation style - the woven code will be more optimal. A code style
aspect achieving most of what you want above will not cause creation of the
bytecode for creating the joinpoint:
Object around(int i): execution(* Print.ech*(int)) && args(i) &&
if(Profiler.enabled(
thisJoinPointStaticPart.getSignature().getDeclaringTypeName())) {
...
}
But the proceeding of the joinpoint will have to occur in the advice body
rather than in the Profiler.profile code.
To answer the question. 311749 has not been implemented. If it had been
then the makeJP call would be after the call to the condition check (pc()),
which it isn't right now:
8: invokestatic #32; //Method
org/aspectj/runtime/internal/Conversions.intObject:(I)Ljava/lang/Object;
11: invokestatic #38; //Method
org/aspectj/runtime/reflect/Factory.makeJP:(Lorg/aspectj/lang/JoinPoint$StaticPart;Ljava/lang/Object;Ljava/lang/Object;Ljava/lang/Object;)Lorg/aspectj/lang/JoinPoint;
14: astore 4
16: iload_3
17: getstatic #26; //Field
ajc$tjp_0:Lorg/aspectj/lang/JoinPoint$StaticPart;
20: invokestatic #76; //Method
Azpect.pc:(ILorg/aspectj/lang/JoinPoint$StaticPart;)Z
Code style doesn't need the join point to be passed in for proceeding, so
the cost of building it is completely avoided.
cheers,
Andy
On 26 November 2012 16:30, Hmil P <[email protected]> wrote:
> Hello,
>
> I am using AspectJ to log performance of code and want to minimize the
> impact on code when profiling is disabled. I have decompiled some
> AspectJ-woven code:
> public int echo(int num) {
> int i = num;
> org.aspectj.lang.JoinPoint localJoinPoint =
> org.aspectj.runtime.reflect.Factory
> .makeJP(ajc$tjp_0, this, this,
> Conversions.intObject(i));
> return Conversions.intValue(echo_aroundBody1$advice(this, i,
> localJoinPoint, aj.ScrapAspect.aspectOf(),
> (org.aspectj.lang.ProceedingJoinPoint)
> localJoinPoint, i));
> }
>
>
> Basically, I want to forego the JoinPoint instantiation in the case of
> profiling being disabled for a method by adding an if pointcut calling the
> static Profiler.enabled() function /dependent on the name of the method/.
>
> @Pointcut("(execution(* Print.ech*(int))) && args(i) && if()")
> public static boolean pc(int i, JoinPoint.StaticPart
> thisJoinPointStaticPart) {
> return
>
> Profiler.enabled(thisJoinPointStaticPart.getSignature().getDeclaringTypeName());
> }
> @Around("pc(i)")
> public Object profileMethods(ProceedingJoinPoint thisJoinPoint, int i)
> throws Throwable {
> return Profiler.profile(thisJoinPoint, i);
> }
>
>
> I saw the bug request
> <https://bugs.eclipse.org/bugs/show_bug.cgi?id=311749> and was wondering
> if it has been implemented.
> Also, does my conditional pointcut really stop the join point instantiation
> (because it is only using the static part)?
>
>
>
> --
> View this message in context:
> http://aspectj.2085585.n4.nabble.com/lazyTJP-optimization-for-around-advice-tp4650646.html
> Sent from the AspectJ - users mailing list archive at Nabble.com.
> _______________________________________________
> aspectj-users mailing list
> [email protected]
> https://dev.eclipse.org/mailman/listinfo/aspectj-users
>
_______________________________________________
aspectj-users mailing list
[email protected]
https://dev.eclipse.org/mailman/listinfo/aspectj-users
