On Aug 2 2011, at 15:27 , joe.da...@oracle.com wrote: > Hi Mike. > > Thanks for developing the fix for this. Generally looks good to go back; a > few minor nits. Personally, I would leave the "// Doesn't use > Boolean.getBoolean to avoid class init." note in java.lang.Class unless you > know the comment is not relevant any more.
I actually added the note. I checked with java -verbose:class on the load order and Class is loaded quite a bit before Boolean. It's a note to future maintainers who, like I was, are tempted to convert to Boolean.getBoolean(). > Given the default semantics of jtreg, you could omit the @compile and @run > lines in the test. Will do. > Cheers, > > -Joe > > On 8/2/2011 12:44 PM, Mike Duigou wrote: >> Hello All; >> >> A fairly simple bug to review which snuck through testing. >> >> http://cr.openjdk.java.net/~mduigou/7073296/0/webrev/ >> >> The changes to Class are incidental but so trivial that opted to include >> them. I can remove if anyone feels strongly (or even weakly) that they >> should be excluded. >> >> Mike
