Mailing-List: contact commits-help@harmony.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@harmony.apache.org
Message-ID: <18294936.1181137586089.JavaMail.jira@brutus>
Date: Wed, 6 Jun 2007 06:46:26 -0700 (PDT)
From: "Mikhail Fursov (JIRA)" <jira@apache.org>
To: commits@harmony.apache.org
Subject: [jira] Updated: (HARMONY-4073) [drlvm][jit] RCE must not substitute
 EFLAGS values produced by CMP with ones produced by SBB
In-Reply-To: <23461316.1181135546118.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-4073?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mikhail Fursov updated HARMONY-4073:
------------------------------------

    Attachment: i8.fix

i8.fix - the fix and regression test

> [drlvm][jit] RCE must not substitute EFLAGS values produced by CMP with ones produced by SBB
> --------------------------------------------------------------------------------------------
>
>                 Key: HARMONY-4073
>                 URL: https://issues.apache.org/jira/browse/HARMONY-4073
>             Project: Harmony
>          Issue Type: Bug
>          Components: DRLVM
>            Reporter: Mikhail Fursov
>         Attachments: i8.fix
>
>
> I found that LongTest from luni module fails when run with OPT.
> The minimized testcase is:
>     System.out.println(foo(Long.MIN_VALUE));
>     static boolean foo(long v) {
>         v = -v;
>         return v < 0;    
>     }
> RI prints 'true' but DRLVM in -Xem:opt mode prints 'false' here.
> The reason of failure is illegal RCE behaviour: it must not substitute EFLAGS values produced by CMP with ones produced by SBB

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