RFR: Shenandoah critical native support

[Zhengyu Gu]

On 03/29/2018 03:58 AM, Aleksey Shipilev wrote:
> On 03/29/2018 03:44 AM, Zhengyu Gu wrote:
>> On 03/28/2018 07:02 PM, Aleksey Shipilev wrote:
>>> On 03/28/2018 07:54 PM, Zhengyu Gu wrote:
>>>> Webrev: http://cr.openjdk.java.net/~zgu/shenandoah/sh-critical-natives/webrev.00/index.html
>>>
>>> This change has a few problems:
>>>
>>>    *) Breaks our nightlies with:
>>>
>>> * For target support_test_hotspot_jtreg_native_support_libCriticalNative_libCriticalNative.o:
>>> /home/jenkins/workspace/nightly/shenandoah-jdk/test/hotspot/jtreg/gc/shenandoah/libCriticalNative.c:
>>> In function ‘JavaCritical_CriticalNativeStress_sum1’:
>>> /home/jenkins/workspace/nightly/shenandoah-jdk/test/hotspot/jtreg/gc/shenandoah/libCriticalNative.c:29:3:
>>>
>>> error: ‘for’ loop initial declarations are only allowed in C99 mode
>>>      for (jint index = 0; index < length; index ++) {
>>>      ^
>>>
>>>    *) Typos: "ciritical",
>>>
>>> +JRT_LEAF(oopDesc*, SharedRuntime::pin_ciritcal_native_array(oopDesc* obj))
>>>
>>> ...
>>>
>>> +JRT_LEAF(void, SharedRuntime::unpin_ciritcal_native_array(oopDesc* obj))
>>
>> Oops!
>>
>> Fixed here: http://cr.openjdk.java.net/~zgu/shenandoah/sh-critical-natives/webrev.01/
> 
> I don't think this is correct:
> 
>    29   jint index;
>    30   for (jint index = 0; index < length; index ++) {
> 
> Need to drop declaration at L30. After that, the patched version passes on CI host.

Sorry, never a good ideal to send out a patch before went to bed :-(

http://cr.openjdk.java.net/~zgu/shenandoah/sh-critical-natives/webrev.02/

Thanks,

-Zhengyu


> 
> Thanks,
> -Aleksey
> 
>