[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [aspectj-users] bug 122580

Nishant,

I'd like to enumerate these issues as each one may be a separate bug and should be addressed as such.
1. Bug 122580: You may still be experiencing the same symptoms with the latest build. Please post more details of the failure
2. When applying an "empty" tracing aspect you code fails. You suspect the problem may lie with serializable classes. Weaving can affect the serialVersionUID so you might like to try specifying "-XaddSerialVersionUID" in aop.xml. See https://bugs.eclipse.org/bugs/show_bug.cgi?id=101411
3. You are experiencing problems instrumenting EJBs. This may be related to the loading of XML files.

We also need to establish whether your problems lie with applying aspects or using LTW. Could you try weaving your application before deploying it to WebLogic?

Matthew Webster
AOSD Project
Java Technology Centre, MP146
IBM Hursley Park, Winchester,  SO21 2JN, England
Telephone: +44 196 2816139 (external) 246139 (internal)
Email: Matthew Webster/UK/IBM @ IBMGB, matthew_webster@xxxxxxxxxx
http://w3.hursley.ibm.com/~websterm/


"Nishant Kumar" <knishant@xxxxxxxxx>
Sent by: aspectj-users-bounces@xxxxxxxxxxx

18/09/2006 20:39

Please respond to
aspectj-users@xxxxxxxxxxx
To
aspectj-users@xxxxxxxxxxx
cc
Subject
Re: [aspectj-users] bug 122580





i disabled instrumenting for a package that had ejb code and the application has started working. I will do more testing and let you guys know.

Is there anyway i can instrument ejbs too? Since it is am important layer I would like to see the same in the trace.

thanks and regards,
Nishant
On 9/19/06, Nishant Kumar <knishant@xxxxxxxxx> wrote:
I will also like to add issues regarding classloader. My application is deployed in weblogic 8.1.5 and it is trying to load some xml files from within and ear and it passing when aspectj is not in the picture but failing when aspectj instrumentation is enabled.

Any idea what should i check for this?

Regards,
Nishant


On 9/19/06, Nishant Kumar <knishant@xxxxxxxxx > wrote:
Hi Wes,

i am not running into infinite loop.

I am using something like the following (from the link that you sent)

aspect A {
   pointcut withinTargetClasses() : within(some entirely different package);

   after() returning:  withinTargetClasses() && call(* *(..)) {
       System.out.println("after " + thisJoinPoint);


   }
}

The code is running almost correct. Someway in the middle instead of taking one path it has taken another path. This cause some application exceptions (not system exceptions). I am not sure where this might be happening. That's why I am asking the following general question.

In what kind of application can aspects fail EVEN if the aspects are coded properly? I have pointed out some scenarios in my previous mail just to give an idea what i am asking.

Regards,
Nishant


On 9/19/06, Wes <wes@xxxxxxxxxxxxxx> wrote:
Just to be sure: you're not experiencing the expected infinite recursion from something like this, are you?:
 
  before() : call(* *(..)) { System.out.println("here"); }
 
  http://www.eclipse.org/aspectj/doc/released/faq.html#q:infiniterecursion
 
Wes
 
------------Original Message------------
From: "Nishant Kumar" <knishant@xxxxxxxxx>
To: aspectj-users@xxxxxxxxxxx
Date: Mon, Sep-18-2006 11:26 AM
Subject: Re: [aspectj-users] bug 122580
thanks matthew,

Here are some details about the second point.

I have an aspect that adds tracing calls before and after the methods and constructors, just to trace the flow of events.
It calls a method to do the actual logging. I have commented out all the content from this method to ensure that there is no issue with the content of this method.
So the only issue that i might have is with instrumenting of the code itself.
Even now my application is failing. I exactly do not know where is it failing. I am using aspectj to understand what is happening in the code.

Can a bean which was serializable (with or without serial version) be affect by any before and after advice?
The above should also imply that it would not affect any mbeans that are accessed remotely?
Adding a before or after advice should also not affect any code depending on reflection. Right?

Regards,
Nishant
On 9/18/06, Matthew Webster <matthew_webster@xxxxxxxxxx> wrote:

Nishant,

>yes I was experiencing the same issue. I think I am using the latest download, certainly not as old as june. I am using >weblogic 8.1.5 which comes with jrockit 1.4.2. I will try again downloading the install today.
If it continues to fail could you please run with the "-verbose" option and post the full output along with the stack trace.


>by the way is there any link where i can find about code patterns that might fail when aspects are applied on it. I mean >information about those pieces of code (even not so well designed code) that run fine but when a simple tracing aspect >is applied before and after it, it might fail.
That shouldn't happen. Can you describe the problem in more detail? It sounds like a bug either in your aspect or perhaps in the code generated by the compiler.


Matthew Webster
AOSD Project
Java Technology Centre, MP146
IBM Hursley Park, Winchester,  SO21 2JN, England
Telephone: +44 196 2816139 (external) 246139 (internal)
Email: Matthew Webster/UK/IBM @ IBMGB, matthew_webster@xxxxxxxxxx
http://w3.hursley.ibm.com/~websterm/

"Nishant Kumar" <knishant@xxxxxxxxx>
Sent by: aspectj-users-bounces@xxxxxxxxxxx

18/09/2006 06:10


Please respond to
aspectj-users@xxxxxxxxxxx
To
aspectj-users@xxxxxxxxxxx
cc
Subject
Re: [aspectj-users] bug 122580







yes I was experiencing the same issue. I think I am using the latest download, certainly not as old as june. I am using weblogic 8.1.5 which comes with jrockit 1.4.2. I will try again downloading the install today.

by the way is there any link where i can find about code patterns that might fail when aspects are applied on it. I mean information about those pieces of code (even not so well designed code) that run fine but when a simple tracing aspect is applied before and after it, it might fail.

regards,
Nishant

On 9/15/06, Matthew Webster <matthew_webster@xxxxxxxxxx > wrote:

Nishant,

I don't quite understand your question. Are you experiencing the same problems using JRockit and LTW that are described in the bug report? If so make sure you are using a download later than the date when the bug fix was built i.e. 2006-06-12 10:18  or the latest.

Matthew Webster
AOSD Project
Java Technology Centre, MP146
IBM Hursley Park, Winchester,  SO21 2JN, England
Telephone: +44 196 2816139 (external) 246139 (internal)
Email: Matthew Webster/UK/IBM @ IBMGB, matthew_webster@xxxxxxxxxx
http://w3.hursley.ibm.com/~websterm/
"Nishant Kumar" <knishant@xxxxxxxxx >
Sent by: aspectj-users-bounces@xxxxxxxxxxx

15/09/2006 06:45

Please respond to
aspectj-users@xxxxxxxxxxx
To
aspectj-users@xxxxxxxxxxx
cc
Subject
[aspectj-users] bug 122580






Hi,
  Can someone explain in layman terms what should be done for the
following bug?

https://bugs.eclipse.org/bugs/show_bug.cgi?id=122580

Isn't using the latest download enough?

Regards,
Nishant
_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users


_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users


_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users

_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users

_______________________________________________

aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users

_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users


_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users