PING Re: RFR: 8163589: Add back class id intrinsic method for event based tracing
Robbin Ehn
robbin.ehn at oracle.com
Fri Aug 26 07:36:20 UTC 2016
Hi all,
Please have a look, thanks!
/Robbin
(updated issue heading also)
On 08/23/2016 06:06 PM, Robbin Ehn wrote:
> Hi all,
>
> C1 in 32-bit actually don't support shift operation which was added in v3 of this series.
>
> We hit this assert in 32-bit:
> # Internal Error (/home/rehn/source/jdk/8163589/hotspot/src/cpu/x86/vm/c1_LIRAssembler_x86.cpp:2917), pid=22445, tid=22493
> # Error: Unimplemented()
>
> So we will just live without this intrinsic in C1 32-bit.
>
> Also added specific test for intrinsic and 'event based testing', see bug (https://bugs.openjdk.java.net/browse/JDK-8163589).
>
> I really hope I haven't missed anything else!
>
> Incr:
> http://cr.openjdk.java.net/~rehn/8163589/hotspot.03-04/webrev/
>
> Full:
> http://cr.openjdk.java.net/~rehn/8163589/hotspot.04/webrev/
>
> Thanks!
>
> /Robbin
>
> On 08/10/2016 04:05 PM, Robbin Ehn wrote:
>> Hi all,
>>
>> In 8 we had this intrinsic, I'm adding it back
>> Used for event based tracing.
>>
>> Webrev: http://cr.openjdk.java.net/~rehn/8163589/hotspot.01/webrev/
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8163589
>>
>> Thanks!
>>
>> /Robbin
More information about the hotspot-compiler-dev
mailing list