RFR JDK-8059510 Compact symbol table layout inside shared archive

Jiangli Zhou jiangli.zhou at oracle.com
Thu Oct 23 16:45:23 UTC 2014


Hi Staffan,

No problem, the review request is still open. :) Thanks for the 
suggestion. I'll try to incorporate it.

Thanks!
Jiangli

On 10/22/2014 11:15 PM, Staffan Larsen wrote:
> Jiangli,
>
> Sorry for being late to the review. I would suggest moving the diagnostic command implementation from diagnosticCommand.cpp/hpp to symbolTable.cpp/hpp. If all dcmds are implemented in diagnosticCommand.cpp, then that file will grow to become quite large, so I would prefer having the commands close to the code they interact with (there are probably a couple more that should be moved). The only problem you may run into is to find a good place to call DCmdFactory::register_DCmdFactory(), but that can be done anywhere during the initialization of the VM.
>
> Thanks,
> /Staffan
>
>
> On 22 okt 2014, at 02:09, Jiangli Zhou <jiangli.zhou at oracle.com> wrote:
>
>> Hi Gerard,
>>
>> Please see the updated webrev: http://cr.openjdk.java.net/~jiangli/8059510/webrev.03/.
>>
>> Thanks,
>> Jiangli
>>
>> On 10/21/2014 02:57 PM, Jiangli Zhou wrote:
>>> Hi Gerard,
>>>
>>> On 10/21/2014 02:32 PM, Gerard Ziemski wrote:
>>>> On 10/21/2014 1:45 PM, Jiangli Zhou wrote:
>>>>> There are pros and cons for both dynamically and statically allocating the _shared_table. Dynamically allocating the _shared_table as you suggested adds overhead of the memory allocation and the overhead of the extra NULL check when the _shared_table is in use. When doing optimization, we try to have the least overhead for the fast case, so we could make the fast case faster (that's when CDS is used).
>>>>>
>>>>> Maybe we could make _shared_table.lookup() inlined and reduce the overhead of the function call when CDS is not used.
>>>> I think that would be OK, though it will require a reader of the code to follow the call through to find out it's cheap NOP when CDS is off, unless we add a nice comment here?
>>> I'll add some comments. Thanks for the suggestion.
>>>
>>> Thanks,
>>> Jiangli
>>>
>>>>
>>>> cheers
>>>>



More information about the hotspot-runtime-dev mailing list