Hello,

I initially proposed this attached patch 3 years ago [1]
I was wondering if this patch could be applied against the version 8. I
tested it against the JDK 7 and 8 (the patch applies without any problem
to both forest).

This would save a lot of time for people who are getting this bug.
Without it, it is hard to find out what is wrong in the JNI code.

Thanks,
Sylvestre
[1]
http://mail.openjdk.java.net/pipermail/hotspot-dev/2010-September/003340.html
PS: please C/C me


JNI_EEXIST-usage.diff

# HG changeset patch
# User Sylvestre Ledru <sylvestre@debian.org>
# Date 1372757479 -7200
# Node ID d02f633c4133c3e66d3a81d39923fff0c59652cb
# Parent b6d1e42655cdf72f6747224366cd1b78b4ec9548
Return the appropriate JNI error message (instead of the generic one) when the JVM is already started

diff --git a/src/share/vm/prims/jni.cpp b/src/share/vm/prims/jni.cpp
--- a/src/share/vm/prims/jni.cpp
+++ b/src/share/vm/prims/jni.cpp
@@ -5097,7 +5097,7 @@
   // function used to determine this will always return false. Atomic::xchg
   // does not have this problem.
   if (Atomic::xchg(1, &vm_created) == 1) {
- return JNI_ERR; // already created, or create attempt in progress
+ return JNI_EEXIST; // already created, or create attempt in progress
   }
   if (Atomic::xchg(0, &safe_to_recreate_vm) == 0) {
     return JNI_ERR; // someone tried and failed and retry not allowed.