Mailing-List: contact dev-help@harmony.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@harmony.apache.org
Received-SPF: pass (herse.apache.org: domain of mike.fursov@gmail.com
 designates 64.233.166.181 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=beta;
        h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references;
        b=jF0V9K+wfaeWfyAkK+vcrucn9LxCdCSURhPkjjBcee+NEWLeG/Sg7fkEpe89wUXm+5EZNMRL7hte9mrLmR4VRFQtFzMn2e1PHpl/AyGMy+f6va4zbvTl6VWuhzh7ngfrMLp4K3JPhQa35gUbF6M/4ap8AmfQFDbgugmemn+lr/0=
Message-ID: <bc79dd600706130134p2653616ch553762e80f95e607@mail.gmail.com>
Date: Wed, 13 Jun 2007 15:34:28 +0700
From: "Mikhail Fursov" <mike.fursov@gmail.com>
To: dev@harmony.apache.org
Subject: Re: [jitrino] synchronization optimization
In-Reply-To: <feb747620706130033g442dee4ai5e2e364bf4ab85c6@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_Part_13826_16075582.1181723668656"
References: <feb747620706102344q62e8069cyd2ba1c7ff9531b9d@mail.gmail.com>
	 <469bff730706122319h7ca3414le6fe1bbe335d32bc@mail.gmail.com>
	 <feb747620706130033g442dee4ai5e2e364bf4ab85c6@mail.gmail.com>

------=_Part_13826_16075582.1181723668656
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Estelle,
We have another low-level optimization for monenter and monexit opcodes:
inlining the implementations of these helpers.
See
vm\thread\javasrc\org\apache\harmony\drlvm\thread\ThreadHelper.java
and
vm\jitrino\src\optimizer\helper_inliner.cpp
for details.

On 6/13/07, Estelle Liu <lyy1105@gmail.com> wrote:
>
> Thanks,Pavel,
> Our group is interested in concurrency analysis,and try to do some
> optimizations on that issue in JIT in the future. We intend to
> get acquainted with the knowledge on the synchronization
> optimizations available in JIT. And we'll appreciate that if you could put
> your improving code in JIRA.
>
>
> 2007/6/13, Pavel Ozhdikhin <pavel.ozhdikhin@gmail.com>:
> >
> > Estelle,
> >
> > I doubt you'll find someone here who knows for sure what it does. :)
> > As you can see 'syncopt' is not used in any of current emconf configs.
> > It does not give any benefit on IA32/EM64T, though it still in the
> > code base to check it on IPF.
> >
> > There is a syncopt2 optimization whose goal is to remove nested
> > synchronization on the same monitor but it's still incomplete. I plan
> > to get back to synchronization optimizations some time and implement
> > removal of nested and repetitive synchronizations - something similar
> > to what is described in [1]. I even implemented a part of this some
> > time ago but then distracted to more urgent JIT tasks.
> >
> > What is your interest in synchronization optimizations? If you would
> > like to work on improving syncopt I can put my code in JIRA.
> >
> > Thanks,
> > Pavel
> >
> > [1] "Automatically Reducing Repetitive Synchronization with a
> > Just-in-Time Compiler for Java", Mark Stoodley, Vijay Sundaresan
> >
> >
> > On 6/11/07, Estelle Liu <lyy1105@gmail.com> wrote:
> > > hi,all,
> > >
> > > I noticed there is a optimization pass "syncopt pass", but I did
> > > master what work does it do. Could anyone give me some information
> about
> > > it?
> > >
> > > --
> > > Estelle
> > >
> >
>
>
>
> --
> Estelle
>


-- 
Mikhail Fursov

------=_Part_13826_16075582.1181723668656--