harmony-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexei Fedotov (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HARMONY-5679) [classlib][archive] UTF-8 decoder & Attributes.Name.hashCode optimizations
Date Thu, 03 Apr 2008 03:00:24 GMT

     [ https://issues.apache.org/jira/browse/HARMONY-5679?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Alexei Fedotov updated HARMONY-5679:
------------------------------------

    Attachment: utf8_decode_test.patch
                utf8_decode_pos.patch

As a part of a careful check I have cloned decoder tests to accept a read only buffer as an
input buffer. This invalidates hasArray() check for the given buffer, hence another branch
of a decoder code is chosen, where pos++ was placed incorrectly. ~80 new tests do nod add
new failures for RI. 

There is a small change that should be done to main sources to make the regression test pass.

> [classlib][archive] UTF-8 decoder & Attributes.Name.hashCode optimizations
> --------------------------------------------------------------------------
>
>                 Key: HARMONY-5679
>                 URL: https://issues.apache.org/jira/browse/HARMONY-5679
>             Project: Harmony
>          Issue Type: Sub-task
>          Components: Classlib
>            Reporter: Alexei Fedotov
>            Assignee: Tim Ellison
>         Attachments: attr_name_hash.patch, utf8_decode.patch, utf8_decode_pos.patch,
utf8_decode_review.patch, utf8_decode_test.patch
>
>
> Found a bug in UTF-8 patch. pos++ is not incremented on fast path. Seems we don't have
a test for this.

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


Mime
View raw message