harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gregory Shimansky <gshiman...@gmail.com>
Subject Re: [drlvm][regression] the test org.apache.harmony.logging.tests.java.util.logging.SocketHandlerTest is hang up today
Date Wed, 22 Nov 2006 12:26:28 GMT
Alexei Fedotov wrote:
> Gregory,
> 
> Thanks for your interest. The bug is reproducible on an interpreter.
> It seems like a hardware exception is fired from internals of Winsock
> listen() function. Could it be true?

Probably yes, this is something I've encountered in another place (see 
the test crashes on win32 thread). What is the exception code? If it is 
0xC0000008 or 0xC0000024, it may be the problem with a bad handle.

> I presume that this can be caused by some linking problem. I'm
> rebuilding a workspace now.
> 
> With best regards,
> Alexei
> 
> 
> 
> 
> 
> 
> On 11/21/06, Gregory Shimansky <gshimansky@gmail.com> wrote:
>> Fedotov, Alexei A wrote:
>> > Vladimir,
>> >
>> > Yes, I can reproduce the issue with
>> > org.apache.harmony.logging.tests.java.util.logging.SocketHandlerTest on
>> > Windows 2003 server.
>> >
>> > Unfortunately, all frames of the stack trace above the exception 
>> handler
>> > are unrecognizable (JITted?)
>>
>> You can try to evaluate st_print() in watch window. If it succeeds (it
>> may fail on assertion about suspend enabled/disabled if called from a
>> wrong part of code), it will print stack dump to console.
>>
>> I think it is mentioned in VM debugging tips.
>>
>> > I'm also able to reproduce few more failures, including a failure in
>> > archive module, will write back soon.
>>
>> -- 
>> Gregory
>>
>>
> 
> 


-- 
Gregory


Mime
View raw message