Summary: | NullPointerException internal compiler error while reporting deprecation problem | ||
---|---|---|---|
Product: | [Tools] AspectJ | Reporter: | Jean-Pierre Lambert <jean-pierre.lambert> |
Component: | Compiler | Assignee: | Andrew Clement <aclement> |
Status: | RESOLVED FIXED | QA Contact: | |
Severity: | normal | ||
Priority: | P3 | ||
Version: | DEVELOPMENT | ||
Target Milestone: | 1.5.0RC1 | ||
Hardware: | PC | ||
OS: | Windows XP | ||
Whiteboard: |
Description
Jean-Pierre Lambert
2005-11-28 11:32:59 EST
I got exactly the same problem when trying to directly build the big project itself with the AspectJ nature (that is, with the AspectJ compiler instead of the Java compiler). I believe the problem is related to some OutOfMemoryError about the Java heap space. Currently the project compiles without any error if Eclipse is launched with a *really* big Java heap space: 500MB heap space isn't enough, I succeeded to compile the project with more than 700MB heap space. Notice that the error still happen if I try to build the big project as a linked folder; when the error is thrown the Java heap space is used at about 150MB/170MB with still the 700MB maximum heap space. It is crashing trying to report a problem. I don't think it's related to running out of memory in this case. The error is not one I've encountered before: "missingDeprecatedAnnotationForMethod" and we crash here MethodBinding binding = method.binding; presumably as the method passed in is null. The message means a deprecated method has been found without the @Deprecated annotation on it... looking at the code that makes the call to report the deprecation problem, and the code a little further down that method - I wonder if its a compiler bug. The code a little further down allows for a null method but the call to the error reporting code does not. In the latest version of the compiler and AJDT, I have put in a guard that prevents this NPE from occurring. It *may* just cause a different manifestation of the problem though, later in the compilers processing - can you give it a try? You need an AJDT from yesterday afternoon or today (basically the latest) (In reply to comment #3) > In the latest version of the compiler and AJDT, I have put in a guard that > prevents this NPE from occurring. It *may* just cause a different > manifestation of the problem though, later in the compilers processing - can > you give it a try? You need an AJDT from yesterday afternoon or today > (basically the latest) I gave it a try and no problem occured. Furthermore deprecation markers appears and works flawlessly. The bug seems to be fixed, notive however that I hadn't tried extensively. JP ok ... thanks for trying it out so swiftly! I'll close this for now - please reopen if you see a similar manifestation.. |