Any comments on this? We're using this patch in libgcj already. I looked at the reference VMClassLoader, and it seems to have the ProtectionDomain variant of defineClass. Is there a reason to hold off checking this in?
Tom Index: ChangeLog from Tom Tromey <[EMAIL PROTECTED]> * java/lang/reflect/Proxy.java (generate): Pass protection domain to VMClassLoader.defineClass. Index: java/lang/reflect/Proxy.java =================================================================== RCS file: /cvsroot/classpath/classpath/java/lang/reflect/Proxy.java,v retrieving revision 1.10 diff -u -r1.10 Proxy.java --- java/lang/reflect/Proxy.java 8 Oct 2003 18:31:34 -0000 1.10 +++ java/lang/reflect/Proxy.java 8 Oct 2003 18:35:59 -0000 @@ -1326,12 +1326,10 @@ { // XXX Do we require more native support here? - // XXX This is waiting on VM support for protection domains. - Class vmClassLoader = Class.forName("java.lang.VMClassLoader"); Class[] types = {ClassLoader.class, String.class, byte[].class, int.class, int.class, - /* ProtectionDomain.class */ }; + ProtectionDomain.class }; Method m = vmClassLoader.getDeclaredMethod("defineClass", types); // Bypass the security check of setAccessible(true), since this @@ -1340,7 +1338,7 @@ m.flag = true; Object[] args = {loader, qualName, bytecode, new Integer(0), new Integer(bytecode.length), - /* Object.class.getProtectionDomain() */ }; + Object.class.getProtectionDomain() }; Class clazz = (Class) m.invoke(null, args); m.flag = false; _______________________________________________ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath