RFR (XXS): 8135327: C1 for arm* should support conditional card marking

[Bertrand Delsart]

Hi Thomas,

Avoiding the new set_card was indeed a good idea since it was not called 
by open code.

However, there is no longer any need to change open code to fix 8135327 
and the proposed change is IMHO counter productive: since 
CardTableModRef_post_barrier_helper is called from open code (when 
CARDTABLEMODREF_POST_BARRIER_HELPER is defined), its prototype should 
IMHO be declared in open code.

Regards,

Bertrand.

On 23/09/2015 21:42, Thomas Schatzl wrote:
> Hi all,
>
>    after some talk with Bengt I found a solution that removes the ugly
> #define from the code completely. I prefer this solution.
>
> New webrev:
> http://cr.openjdk.java.net/~tschatzl/8135327/webrev.1/
> Testing:
> jprt
>
> Thanks,
>    Thomas
>
> On Wed, 2015-09-23 at 09:20 +0200, Thomas Schatzl wrote:
>> Hi,
>>
>> On Tue, 2015-09-22 at 09:19 -0400, Eric Caspole wrote:
>>> Looks good.
>>> Eric
>>>
>>
>>    thanks for the review.
>>
>> Thomas
>>
>>
>
>


-- 
Bertrand Delsart,                     Grenoble Engineering Center
Oracle,         180 av. de l'Europe,          ZIRST de Montbonnot
38330 Montbonnot Saint Martin,                             FRANCE
bertrand.delsart at oracle.com             Phone : +33 4 76 18 81 23

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE: This email message is for the sole use of the intended
recipient(s) and may contain confidential and privileged
information. Any unauthorized review, use, disclosure or
distribution is prohibited. If you are not the intended recipient,
please contact the sender by reply email and destroy all copies of
the original message.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~