Community
Participate
Working Groups
I'm using eclipse M4, and this sometimes pops by.. I also use eclipseme, but I'm not aware that it changes any sourcecode.. After cleaning my project / rebuilding / restarting eclipse, the problem still exists.. I also removes the cldc builder (eclipseme) When I open the file where it happens, I get some popups telling me "AST creation failed"
Created attachment 17622 [details] logfile with stacktraces
Could you please provide steps to reproduce and then reopen ? This problem is likely internal to the file where it is reported. Attaching the source of the offending file should be sufficient to get started. When building, there should be a fake problem in problem view recording the compiler failure associated to line 1. This would point at the offending file.
I've created a new project, and removed my sjpp encoding trick. The error happens when I set the jdk compliance to any non-5.0 and uncheck the "Inline finally blocks" (I do this to reduce class file size, as I'm working on midlets.. I've trimmed down the source, and it's very weird... When I come to a certain size, when I remove some a method or field, the error goes away. If I put it back, the error is back..
Created attachment 17624 [details] test project use any non-5.0 jdk, and uncheck the inline finally checkbox
The test project I've created also gives the error on another M4 installation I have (other workstation) Also, I've tried it with the latest integration build. eclipse-SDK-I20050202- 0800-win32. Clean install, no cutomizations.
-SDK-I20050202-0800-win32 .log snippet: !SESSION 2005-02-02 15:00:05.83 ------------------------------------------------ eclipse.buildId=I20050202-0800 java.version=1.5.0 java.vendor=Sun Microsystems Inc. BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=en_US Command-line arguments: -os win32 -ws win32 -arch x86 !ENTRY org.eclipse.core.resources 4 2 2005-02-02 15:00:05.99 !MESSAGE Problems occurred when invoking code from plug- in: "org.eclipse.core.resources". !STACK 0 java.lang.ArrayIndexOutOfBoundsException at java.lang.System.arraycopy(Native Method) at org.eclipse.jdt.internal.compiler.flow.UnconditionalFlowInfo.copy (UnconditionalFlowInfo.java:197) at org.eclipse.jdt.internal.compiler.flow.ExceptionHandlingFlowContext.recordReturn From(ExceptionHandlingFlowContext.java:182) at org.eclipse.jdt.internal.compiler.ast.ReturnStatement.analyseCode (ReturnStatement.java:60) at org.eclipse.jdt.internal.compiler.ast.Block.analyseCode (Block.java:40)
Reproduced. Thanks for testcase.
Problem comes from non redimensionning defNull /defNonNull for extra inits in certain circumstances. Added AssignmentTest#test034. Fixed NOTE: reason for the problem sometimes being revealed or not, is due to the fact we have optimized support for 64 first variables. The bug was visible under some circumstances and only if you had more than 64 variables.
*** Bug 84334 has been marked as a duplicate of this bug. ***
*** Bug 84553 has been marked as a duplicate of this bug. ***
*** Bug 85209 has been marked as a duplicate of this bug. ***
Verified in I20050215-0800
*** Bug 85666 has been marked as a duplicate of this bug. ***
Works for me now using the I20050217-1200 integration build.