We investigating in multithreading possibilities. In connection with the
GC I try to organize the mark phase into a parallel thread.
The idea is to have 2 heap, serve allocations from the active one, while
marking the other. When the active is full, swap them.
Currently, it is in a non stable and non complete state, and it is not
100% that it will be ever complete.

On 04/19/2010 11:08 PM, Geoffrey Garen wrote:
> Hi Balazs.
>
> I just checked in those two tests to JavaScriptCore/tests/perf.
>
> What kind of GC hacking are you doing?
>
> Geoff
>
> On Apr 19, 2010, at 5:44 AM, Balazs Kelemen wrote:
>
>   
>>  Hi folks!
>>
>> I am doing some hacking around the GC, and want to test and benchmark it.
>> In GC related changes I see results of these two benchmarks, but I do
>> not find them in the tree.
>> Are those living in the tree? If not, I think it would be useful to
>> check-in them.
>>
>> Balazs Kelemen
>>
>> _______________________________________________
>> webkit-dev mailing list
>> [email protected]
>> http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
>>     
>   

_______________________________________________
webkit-dev mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev

Reply via email to