[regression][shutdown] JVMTI test fail

[Alexei Fedotov <al...@gmail.com>]

Hello gurus,

Mikhail wrote,

>          HyNLSHashEntry *next = entry->next;

Smoke JVMTI tests fail for me at the same line. Its seems that
nls->hash_buckets[i]-> next contain 0xeeffeeff which resembles a
DEADBEAF for me. Do we have a mind here that can handle the complexity
of hashmap shutdown problems?

BTW, I can see that from recent commits Tim's one adds some
http://svn.apache.org/viewvc?view=rev&revision=547205 hash table
operation.

    [junit] DEBUG: Breakpoint on line 47
    [junit] DEBUG: Disabling Breakpoint event...
    [junit] DEBUG: done
    [junit] DEBUG: Setting success == true...
    [junit] DEBUG: done
    [junit] DEBUG: Deallocating line number table...
    [junit] DEBUG: done
    [echo] Running VMInit1
    <crashed here>

On 3/20/07, Mikhail Fursov (JIRA) <ji...@apache.org> wrote:
>
>    [ https://issues.apache.org/jira/browse/HARMONY-3428?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12482357 ]
>
> Mikhail Fursov commented on HARMONY-3428:
> -----------------------------------------
>
> After applying  "tmp_workaround-nef.patch"  HWA craches in JET mode in "modules\portlib\src\main\native\port\shared\hynls.c"
>
> for (i = 0; i < HYNLS_NUM_HASH_BUCKETS; i++)
>    {
>      entry = nls->hash_buckets[i];
>      while (entry)
>        {
>          HyNLSHashEntry *next = entry->next;
>
> with entry=0x00000001
>
> > [drlvm][winx64] HWA fails with new VEH in Interpreter/debug mode
> > ----------------------------------------------------------------
> >
> >                 Key: HARMONY-3428
> >                 URL: https://issues.apache.org/jira/browse/HARMONY-3428
> >             Project: Harmony
> >          Issue Type: Bug
> >          Components: DRLVM
> >         Environment: windows/em64t
> >            Reporter: Vladimir Beliaev
> >            Priority: Critical
> >         Attachments: tmp_workaround-nef.patch
> >
> >
> > With recently committed Vectored Exception Handler update the HWA (Hello World Application) does not work anymore even with Interpterer in debug mode on windows/em64t.
> > It fails with assert (p_TLS_vmthread).
> > Thanks
> > Vladimir Beliaev
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>


-- 
With best regards,
Alexei,
ESSD, Intel