RFR(S) [8u] 8189170: Add option to disable stack overflow checking in primordial thread for use with JNI_CreateJavaJVM

David Buck david.buck at oracle.com
Fri Jan 26 05:38:54 UTC 2018


Hi!

Thank you Volker and David for caching my mistakes!

I assume additional review is not needed, but just in case I have 
prepared a new webrev:

http://cr.openjdk.java.net/~dbuck/8189170_jdk8_ver01/

This is what I plan to push as soon as CCC approval is obtained.

Cheers,
-Buck

On 2018/01/26 2:42, Volker Simonis wrote:
> Hi David,
> 
> I've built your change on Linux/ppc64 and AIX. On Linux it works fine,
> however on AIX you need the following tiny, additional patch:
> 
> --- a/src/os/aix/vm/os_aix.cpp  Thu Jan 25 17:40:10 2018 +0100
> +++ b/src/os/aix/vm/os_aix.cpp  Thu Jan 25 18:22:55 2018 +0100
> @@ -4646,7 +4646,7 @@
> 
>     // This only works when invoked on a pthread. As we agreed not to use
>     // primordial threads anyway, I assert here
> -  guarantee(!os::Aix::is_primordial_thread(), "not allowed on the
> primordial thread");
> +  guarantee(!os::is_primordial_thread(), "not allowed on the
> primordial thread");
> 
>     // information about this api can be found (a) in the pthread.h header and
>     // (b) in http://publib.boulder.ibm.com/infocenter/pseries/v5r3/index.jsp?topic=/com.ibm.aix.basetechref/doc/basetrf1/pthread_getthrds_np.htm
> 
> because jdk8 apparently had one more call site for
> os::AIX:is_primordial_thread().
> 
> Thanks,
> Volker
> 
> 
> On Thu, Jan 25, 2018 at 5:36 AM, David Holmes <david.holmes at oracle.com> wrote:
>> Sorry David one minor nit.
>>
>> In the <os>_<cpu>.cpp files where you changed is_initial_thread() to
>> is_primordial_thread(), in the following comment it still says
>>
>> // initial thread ...
>>
>> but should now say
>>
>> // primordial thread
>>
>> No need to see updated webrev.
>>
>> Thanks,
>> David
>>
>>
>> On 25/01/2018 1:42 PM, David Holmes wrote:
>>>
>>> Hi David,
>>>
>>> That looks like an accurate backport of the changes. Reviewed.
>>>
>>> Thanks,
>>> David
>>>
>>> On 25/01/2018 12:50 PM, David Buck wrote:
>>>>
>>>> Hi!
>>>>
>>>> Please review my backport of this small change to JDK 8.
>>>>
>>>> bug report:
>>>> https://bugs.openjdk.java.net/browse/JDK-8189170
>>>>
>>>> original JDK changeset:
>>>> http://hg.openjdk.java.net/jdk/jdk/rev/9fd89aabb6cd
>>>>
>>>> code review of original change:
>>>>
>>>> http://mail.openjdk.java.net/pipermail/hotspot-runtime-dev/2017-November/025207.html
>>>>
>>>> JDK 8 webrev for review:
>>>> http://cr.openjdk.java.net/~dbuck/8189170_jdk8_ver00/
>>>>
>>>> All standard JPRT tests run and passed (hotspot testset). Effectiveness
>>>> of change manually confirmed with dummy JNI invocation test case.
>>>>
>>>> Cheers,
>>>> -Buck


More information about the hotspot-runtime-dev mailing list