I can also observe some failures introduced from the commit. Benchmarks of specJVM fails in server or JET mode, in 32bits/64bits platform. Some GC helper can not be processed (if tunring on -XX:gc.prefetch=true) now. Should we just roll back the commit and commit a more clean one later? Buqi, xiaoming, Alexey, others, do you have opinions? On Mon, Oct 13, 2008 at 3:58 PM, Cheng, BuQi (JIRA) <[EMAIL PROTECTED]> wrote:
> > [ > https://issues.apache.org/jira/browse/HARMONY-5037?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12638963#action_12638963] > > Cheng, BuQi commented on HARMONY-5037: > -------------------------------------- > > Hi Alexey: > > Would you like to provide more information about the patch? The patch is > too bigger to understand clearly. > > Thanks! > > Buqi > > > > > > [drlvm][jit][opt] improper opcode group lookup > > ---------------------------------------------- > > > > Key: HARMONY-5037 > > URL: https://issues.apache.org/jira/browse/HARMONY-5037 > > Project: Harmony > > Issue Type: Bug > > Components: DRLVM > > Reporter: Alexey Varlamov > > Assignee: Alexey Varlamov > > Fix For: 5.0M7 > > > > Attachments: enc64.diff, enc_lookup.patch, > HARMONY-5037-enc_lookup_gt3_May23.patch > > > > > > Encoder of Jitrino.OPT improperly matches OpcodeGroups, allowing to > create invalid instructions. See for example HARMONY-5022 > > -- > This message is automatically generated by JIRA. > - > You can reply to this email to add a comment to the issue online. > >
