RFR: 8234748: Clean up atomic and orderAccess includes

Thomas Schatzl thomas.schatzl at oracle.com
Wed Nov 27 10:39:58 UTC 2019


Hi,

On 27.11.19 11:34, Stefan Karlsson wrote:
> Hi Thomas,
> 
> On 2019-11-27 11:20, Thomas Schatzl wrote:
>> Hi Stefan,
>>
>>    thanks for tackling this.
>>
>> On 26.11.19 12:11, Stefan Karlsson wrote:
>>> Hi all,
>>>[...]
>> Here are the results:
>>
>> Improvements to orderAccess.hpp includes:
>>
>> - gc/shenandoah/shenandoahVerifier.cpp misses it
>>
>> - the #include from gc/z/zLiveMap.inline.hpp should maybe be moved to 
>> zLiveMap.cpp
>>
>> Improvements to atomic.hpp includes:
>>
>> In the following files the include of atomic.hpp should be removed as 
>> it seems unnecessary:
>> - share/utilities/bitMap.hpp
>> - share/oops/oop.hpp
> 
> These needs atomic.hpp since they use atomic_memory_order.

Okay, did not consider these. Good catch.

> 
>> - share/gc/z/zNMethodTable.cpp
>> - share/gc/shenandoah/shenandoahForwarding.inline.hpp
>> - share/gc/g1/g1CardTable.cpp
>> - os_cpu/solaris_x86/os_solaris_x86.cpp
>>
>> The following need a #include atomic.hpp:
>> - share/oops/methodData.cpp
>> - share/oops/constantPool.cpp
>> - share/memory/metaspace/virtualSpaceNdoe.cpp
>> - os/posix/os_posix.cpp
>>
>> Looks good otherwise.
> 
> Since, I already pushed the original bug, I created a new with your 
> changes:
> https://cr.openjdk.java.net/~stefank/8234897/webrev.01/
> 
> I intend to take this through our build steps, and push it if it's 
> succeeds.
> 

This new change looks good.

Thanks,
   Thomas



More information about the hotspot-dev mailing list