Request for review: 7034585 Adjust fillinStackTrace filtering to assist 6998871

Mandy Chung mandy.chung at oracle.com
Mon Apr 11 17:03:03 PDT 2011


  On 4/11/11 4:54 PM, David Holmes wrote:
> Tom, Mandy,
>
> Can I get the ok on the revised webrev please. I'd like to push this 
> today.
>

The revised webrev looks good to me.

Mandy

> Thanks,
> David
>
> David Holmes said the following on 04/09/11 15:05:
>> Tom Rodriguez said the following on 04/09/11 10:03:
>>> On Apr 8, 2011, at 5:01 PM, David Holmes wrote:
>>>
>>>> Tom Rodriguez said the following on 04/09/11 09:22:
>>>>> Actually why are you checking for fillInStackTrace* instead of 
>>>>> just fillInStackTrace?
>>>>> if (method->name() == vmSymbols::fillIinStackTrace_name() &&
>>>>>    throwable->is_a(method->method_holder())) {
>>>> Added flexibility. There is still some debate as to the final 
>>>> naming scheme in the JDK changes. At present an overload of 
>>>> fillInStackTrace is used with a dummy int parameter. The normal 
>>>> naming scheme would have the native method be called 
>>>> fillInStackTrace0. That would have required a VM change, hence the 
>>>> overload. But we ended up needing a VM change anyway, hence it is 
>>>> now possible to change the method name. I was trying to accommodate 
>>>> this either way. But perhaps it is simpler to just add the symbol 
>>>> for fillInStackTrace0 and check for either?
>>>
>>> I think that would be more straightforward.
>>
>> Agreed. Updated webrev:
>>
>> http://cr.openjdk.java.net/~dholmes/7034585/webrev.2/
>>
>> Thanks,
>> David
>> ------
>>



More information about the hotspot-runtime-dev mailing list