[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: bug in gnu.vm.stack.StackFrame
From: |
Mark Wielaard |
Subject: |
Re: bug in gnu.vm.stack.StackFrame |
Date: |
10 Sep 2002 23:07:34 +0200 |
Hi,
On Fri, 2002-09-06 at 20:59, C. Scott Ananian wrote:
> The gnu/vm/stack/StackFrame class has the following bit of code:
>
> private StackFrame(Object obj, Method method, int lineNum, String filename) {
> this.caller = caller;
> ....
> }
>
> Note that 'caller' is not an argument of the constructor. So the first
> statement of the constructor is effectively a no-op, equivalent to
> this.caller = this.caller
>
> Oops!
Hmmm. There seem to be two implementations of StackFrame.
The one above with bug in vm/reference/gnu/vm/stack/StackFrame.java.
And one in gnu/java/lang/StackFrame.java.
StackFrame, ExecutionStack and StackTrace are a bit old. I don't know if
there are VMs that still use these classes. For Throwable stack traces
we now have java.lang.StackTraceElement and I will soon add my recent
work on VMThrowable that I did for libgcj to Classpath. But for security
managers and for AccessController the ExecutionStack could still be
useful. Do VM implementers actually use/implement support for these
classes?
Cheers,
Mark