On 4/24/07, Gregory Shimansky <[EMAIL PROTECTED]> wrote:
Xiao-Feng Li wrote:
> I checked this log, and had thought it was a false alarm of failure,
> since I didn't find any failure case at first glance. With more
> careful checking, I found following report:
>
>> ==================================
>> Run JVMTI tests using jitrino.JET
>> ==================================
>>
>> Running Breakpoint1
>> Running PopFrame1
> ---> SIGTRAP in VM code.
> ---> Stack trace:
> ---> <end of stack trace>
>> Running SingleStep1
>> Running VMInit1
>
> I don't know if it is failure report or not. Some JVMTI expert may
> help to explain.
Judging by the log it is a crash because int3 was hit in VM code. It
shouldn't happen because JVMTI should instrument only compiled Java
code. I am running the same VM tests right now and they work just fine,
so I cannot reproduce it. Perhaps analyzing core dump on the testing
machine could help to at least show the stack trace.
Thanks for the explanation. Yes, I had the same feeling about this
strange int3 in VM code, and no, I couldn't reproduce it either. Let's
ignore it for the moment unless it comes again.
Thanks,
xiaofeng
--
Gregory
--
http://xiao-feng.blogspot.com
