RFR: 8215500: ICRefillVerifierMark does not set the provided verfier as current

Tobias Hartmann tobias.hartmann at oracle.com
Tue Dec 18 10:51:32 UTC 2018


Hi Erik,

looks good to me too.

Best regards,
Tobias

On 18.12.18 10:40, Erik Österlund wrote:
> Hi Per,
> 
> Thanks for the review. Will do that name change before pushing as suggested.
> 
> /Erik
> 
> On 2018-12-17 22:17, Per Liden wrote:
>>
>>
>> On 12/17/2018 08:16 PM, Erik Österlund wrote:
>>> Hi Per,
>>>
>>> On 2018-12-17 19:12, Per Liden wrote:
>>>> On 12/17/2018 06:57 PM, Erik Österlund wrote:
>>>>> Hi,
>>>>>
>>>>> ICRefillVerifierMark::ICRefillVerifierMark sets "this" as the current verifier, as opposed to
>>>>> the verifier passed in. That is wrong.
>>>>>
>>>>> Bug:
>>>>> https://bugs.openjdk.java.net/browse/JDK-8215500
>>>>>
>>>>> Webrev:
>>>>> http://cr.openjdk.java.net/~eosterlund/8215500/webrev.00/
>>>>
>>>> Wow, one of those "how could this have ever worked?" ;) Should we forward declare
>>>> ICRefillVerifier in thread.hpp to get some better type safety here and avoid the
>>>> reinterpret_cast in icBuffer.cpp?
>>>
>>> Sure. New webrev:
>>> http://cr.openjdk.java.net/~eosterlund/8215500/webrev.01/
>>
>> Looks good! However, I couldn't help by noticing that it seem like the
>> "{set_,_,}missed_ic_stub_refill_mark" members in Thread should be called
>> "missed_ic_stub_refill_verifier" now, right? Or maybe just "{set_,_,}ic_refill_verifier" to match
>> the class name. Anyway, you decide, either way I don't need to see a new webrev.
>>
>> cheers,
>> Per
>>
>>>
>>> Thanks,
>>> /Erik
>>>
>>>> cheers,
>>>> Per


More information about the hotspot-compiler-dev mailing list