[RFA] JDWP implment getFrameCount

[Kyle Galloway <kgallowa at redhat dot com>]

Keith Seitz wrote:
> Kyle Galloway wrote:
>
> > Comments, questions?
>
> I'm not a maintainer, but since I am quite involved in this code, I 
> would really, really appreciate it if you could break this into 
> smaller patches. It just makes it easier (and therefore faster) to 
> parse. [Hey, I'm getting old! :-)]
>
> [And you need to submit the classpath changes upstream to 
> classpath-patches@gnu.org first.]
As requested, I have submitted the classpath changes, and once they are 
approved will submit the getFrame and getFrames patches in separate 
chunks.  For now, getFrameCount involves no classpath changes so I have 
started with this.

This patch implements the JDWP method getFrameCount from 
natVMVirtualMachine.cc.  It calls jvmti->GetFrameCount then returns the 
result.

Changelog

2007-02-01  Kyle Galloway  <kgallowa@redhat.com>

   * gnu/classpath/jdwp/natVMVirtualMachine.cc (getFrameCount): Implment.

Comments?

- Kyle

Index: /notnfs/kgallowa/gcc/libjava/gnu/classpath/jdwp/natVMVirtualMachine.cc
===================================================================
--- /notnfs/kgallowa/gcc/libjava/gnu/classpath/jdwp/natVMVirtualMachine.cc	(revision 121464)
+++ /notnfs/kgallowa/gcc/libjava/gnu/classpath/jdwp/natVMVirtualMachine.cc	(working copy)
@@ -369,7 +369,15 @@
 gnu::classpath::jdwp::VMVirtualMachine::
 getFrameCount (MAYBE_UNUSED Thread *thread)
 {
-  return 0;
+  jint frame_count;
+  jthread thr = reinterpret_cast<jthread> (thread);
+  
+  jvmtiError jerr = _jdwp_jvmtiEnv->GetFrameCount (thr, &frame_count);
+  
+  if (jerr != JVMTI_ERROR_NONE)
+    throw_jvmti_error (jerr);
+  
+  return frame_count;
 }
 
 jint