Review Request for 6751338: ZIP inflater/deflater performance

Dave Bristor bristor at
Tue Apr 5 15:41:00 UTC 2011

@Alan: Great idea about looking at runs with +PrintGCDetails and +PrintJNIGCStalls!  Wish I had thought of that.  About @ignore: it seems fine remove that, though if there are performance degradations the test won't catch them of course.  Renaming the test seems fine.

@Sherman: updating this_off is fine as "documentation".


--- On Mon, 4/4/11, Alan Bateman <Alan.Bateman at> wrote:

> From: Alan Bateman <Alan.Bateman at>
> Subject: Re: Review Request for 6751338: ZIP inflater/deflater performance
> To: "Xueming Shen" <xueming.shen at>
> Cc: "Dave Bristor" <bristor at>, "core-libs-dev" <core-libs-dev at>
> Date: Monday, April 4, 2011, 2:09 AM
> Xueming Shen wrote:
> > Dave, Alan,
> > 
> > Here is the final webrev based on Dave's patch and the
> jdk1.5 code that does not
> > have the change for 6206933. JPRT job result suggests
> no new testing failure and
> > my "non-scientific" benchmark test (to use
> GZIPOu/InputStream to compress/
> > decompress the rt.jar) does show relative performance
> gain. Will try to run more
> > tests the weekend, but here is the webrev.
> > 
> >
> I went through the webrev and also checked the old
> (pre-OpenJDK) code from before the changes for 6206933. The
> changes look okay to me. When testing with HotSpot then
> running with +PrintGCDetails and +PrintJNIGCStalls may be
> useful.
> One comment on the test is that it
> might be better to push this without the jtreg tags as the
> @ignore will cause it to be reported by jtreg as an "error".
> On the naming then maybe might be
> better.
> -Alan

More information about the core-libs-dev mailing list