This is the mail archive of the
java-patches@gcc.gnu.org
mailing list for the Java project.
[RFC] JVMTI CLASS_PREPARE event
- From: Keith Seitz <keiths at redhat dot com>
- To: Java Patch List <java-patches at gcc dot gnu dot org>
- Date: Wed, 24 Jan 2007 13:21:31 -0800
- Subject: [RFC] JVMTI CLASS_PREPARE event
Hi,
If anyone was wondering when the patches were going require serious
review, your wait is over. I've not generated a "proper" diff using "svn
diff" for this "patch" because it is way too verbose with whitespace
changes. Instead, I've simply used "diff -uw" on it. It's much more
readable for purposes of this discussion.
The patch adds the JVMTI event notification for CLASS_PREPARE. This is
now the third or fourth place I've put this notification (at least), and
I know this is still not quite right. It's awfully close, but not perfect.
I welcome suggestions, advice, comments, etc.
Keith
--- .svn/text-base/link.cc.svn-base 2007-01-10 18:35:41.000000000 -0800
+++ link.cc 2007-01-24 13:09:11.000000000 -0800
@@ -33,6 +33,8 @@
#include <limits.h>
#include <java-cpool.h>
#include <execution.h>
+#include <jvmti.h>
+#include "jvmti-int.h"
#include <java/lang/Class.h>
#include <java/lang/String.h>
#include <java/lang/StringBuffer.h>
@@ -1941,6 +1943,7 @@
if (klass->state >= state)
return;
+ {
JvSynchronize sync (klass);
// This is similar to the strategy for class initialization. If we
@@ -2027,3 +2030,15 @@
if (klass->state == JV_STATE_ERROR)
throw new java::lang::LinkageError;
}
+
+ if (klass->state == JV_STATE_LINKED)
+ {
+ if (JVMTI_REQUESTED_EVENT (ClassPrepare))
+ {
+ java::lang::Thread *thread = java::lang::Thread::currentThread ();
+ JNIEnv *jni_env = _Jv_GetCurrentJNIEnv ();
+ _Jv_JVMTI_PostEvent (JVMTI_EVENT_CLASS_PREPARE, thread, jni_env,
+ klass);
+ }
+ }
+}