Mailing-List: contact commits-help@harmony.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@harmony.apache.org
Message-ID: <29607272.1184674984733.JavaMail.jira@brutus>
Date: Tue, 17 Jul 2007 05:23:04 -0700 (PDT)
From: "Pavel Ozhdikhin (JIRA)" <jira@apache.org>
To: commits@harmony.apache.org
Subject: [jira] Commented: (HARMONY-4476) [drlvm][jit][opt][abcd] Two-state
 Inequality Graph for both Lower and Upper problems, to ensure correctness
 and simplify the code, ability to dump stats
In-Reply-To: <28724278.1184620564998.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HARMONY-4476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12513223 ] 

Pavel Ozhdikhin commented on HARMONY-4476:
------------------------------------------

Egor, I've looked through your patch. Not very deep though, just a generic view.
The code now indeed became more readable! The new graph printing looks nice too.

I have a copule of comments/questions regarding your patch and ABCD implementation in general:

---
classic_abcd.cpp:
Typo in the comment ("constraints"):
    // constrints (edges) rather than old ones;

---
This code:
    static const uint64 min_const_opnd = (min_ssa_opnd) * 3;
    ...
    setID((uint32)min_const_opnd + id);
looks potentially unsafe to me due to type conversion and possible overflow. There should be at least an assert checking that there is no overflow. It's better to be jitrino_assert to fire on release as well.

I've successfully built DRLVM on Win32 with your patch. I also can't run DaCapo on debug build in server_static mode - recent commit for a new BC mapping has broken this. This will be fixed soon.

On release build I ran into verifier issue (this is also a recent change). With verifier disabled I finally managed to run jython, got some info on bounds check but was too lazy to write a script to sum up the results - just made sure something is optimized.

Now I'm running "build test" on win32 and get back with the results later.


> [drlvm][jit][opt][abcd] Two-state Inequality Graph for both Lower and Upper problems, to ensure correctness and simplify the code, ability to dump stats
> --------------------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: HARMONY-4476
>                 URL: https://issues.apache.org/jira/browse/HARMONY-4476
>             Project: Harmony
>          Issue Type: Improvement
>          Components: DRLVM
>         Environment: all
>            Reporter: Egor Pasko
>         Attachments: 0001-Two-state-Inequality-Graph-with-ABCD-Stats-option.txt, igraph.png
>
>
> Creating two Inequality Graphs makes it difficult to make sure the graph looks like the right one. This issue proposes to create two-state InequalityGraph. One state shall represent the upper-bound problem, the other -- lower-bound. Constraints arising from statements like "a=b", "a=b+const" are the same in both states of the graph, whereas constraints arising from statements like "if (a < b)" are different (basically, the edges change direction to the opposite and negate their length)

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.