Russ Blaine wrote: > A group of us at Sun have been thinking and talking about this for > years. It is certainly an enticing idea - the more data the kernel can > consider when making critical choices, the more efficient it can be > in parceling out resources.
Heh. Once again, I get an idea, something I never remember seeing discussed anywhere. And then, I find some guys at Sun are working on it already ^-^ It's a nice though - suggests my idea was good, and even better, might well get implemented. > Lately, a team here at Sun has been working on doing exactly > that. I'm not sure how much I can say about it publicly, but if more > disclosure is warranted, I will follow-up here. Personally, I only have an abstract interest - it's nice to see better technology. >> PS A big thank-you to whoever added extensive support for CPC in Solaris x86. >> I like using CPC for in-depth benchmark analysis and this helps me greately! > >That was me - thanks for the feedback! It's great to hear things like this. I would do a public article looking at how to use performance counters to better understand benchmarks, but I haven't had time yet, since I'm doing it in my own time (and money) for fun. I know some people who benchmark various systems and write reports, and I've also been trying to get them to use performance counters for extra analysis. Haven't quite succeeded yet... >> PPS Any chance of a an approved (or at least community shared) Java wrapper >> to CPC? I wrote a rather basic one myself, though it's rather flakey. > >What type of wrapper are you talking about? It can be hard to >correlate system activity to JVM behavior down to particular java >code. Or are you talking about measuring performance of the JVM, in >general? We've been thinking about a CPC provider for dtrace for a >while now (and rumor has it that someone is actually prototyping this, >right now). Such a beast would, when used in conjunction with the java >dtrace provider, allow detailed, CPC data-driven analysis of java >apps. The main thing I am/was working on was to analyse a particular Java thread (or set of threads). So I developed a Java/C wrapper where I ask the CPC library to attach to the current thread. That way, you can eliminate what other threads (including GC threads and the Hotspot optimiser) are doing. This message posted from opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
