> -----Original Message----- > From: jdk7-dev-boun...@openjdk.java.net [mailto:jdk7-dev- > boun...@openjdk.java.net] On Behalf Of Paul Hohensee > Sent: Monday, April 11, 2011 2:01 PM > To: jdk7-dev@openjdk.java.net > Subject: Re: New String-related cmdline options > > What 6714404 means is that the +UseStringCache switch only works if you > also specify > +AggressiveOpts. By default, UseStringCache is off, even if > AggressiveOpts is on, because > at the moment it benefits only some applications on some architectures. > At some point, > we'll probably make whether it defaults on or off platform dependent, > but we haven't > completed the necessary analysis yet. > > Recall that AggressiveOpts enables optimizations that we believe will > have general > applicability, but of which we're not yet confident enough to either > make default on > or platform-dependent default-on. We'll almost always do one or the > other in some > future release, so think of AggressiveOpts as a way-station for new > optimizations. > > Paul > > On 4/11/11 9:21 AM, David Dabbs wrote: > > Hi. > > > > I see there are some new String-related cmdline opts. StringTableSize > and > > UseStringCache. Are there docs concerning the new switches and/or > > diagnostics related to tuning? Just for grins, I specified > AggressiveOpts, > > but didn't see UseStringCache set to true, or did I misunderstand > this > > patch? > > > > 6714404: Add UseStringCache switch to enable String caching > under > > AggressiveOpts > > > > > > Thank you, > > > > David > >
Okay. Your description sounds like UnlockExperimentalVMOptions. I was somehow under the impression that specifying AggressiveOpts enabled (as in turned on) new/experimental features as opposed to just allowing them to be enabled. Thank you Paul. David
