harmony-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Pavel Pervov (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HARMONY-4704) [drlvm][jni] JNI improvements
Date Thu, 10 Apr 2008 10:12:04 GMT

    [ https://issues.apache.org/jira/browse/HARMONY-4704?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12587565#action_12587565
] 

Pavel Pervov commented on HARMONY-4704:
---------------------------------------

JNI-complete.patch:
1) Why pop_m2n_noexp in the code, but pop_noexp_m2n in LIL? Is it to make parsing easier?
Is code simplification worth this inconsistency?
2) It would be nice to document added LIL instructions somewhere. I can guess that HSE/HSD
is the pair of hythread_suspend_enable/hythread_suspend_disable.
3) I can also try to guess that 'noexp' means "no exception expected". If that's true - I'd
personally prefer 'noexn' suffix.

> [drlvm][jni] JNI improvements
> -----------------------------
>
>                 Key: HARMONY-4704
>                 URL: https://issues.apache.org/jira/browse/HARMONY-4704
>             Project: Harmony
>          Issue Type: Improvement
>          Components: DRLVM
>            Reporter: Aleksey Shipilev
>         Attachments: JNI-complete-2.patch, JNI-complete.patch
>
>
> The root issue for JNI improvements:
> Clean:
> HARMONY-4705 [drlvm][jni] JNI transition uses ineffective TLS get
> HARMONY-4811 [drlvm][exception] exn_raised() should cache the TLS
> Dirty:
> HARMONY-4714 [drlvm][jni] JNI transition checks exceptions twice
> Proof-of-concept
> HARMONY-4729 [drlvm][jni] JNI transition should inline hythread_suspend_enable() and
hythread_suspend_disable()
> HARMONY-4806 [drlvm][jni] Experiment: rewriting entire JNI transition stub on encoder

-- 
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