RFR (S): 6516521: Doc: should document about the primordial thread attaching to the VM

Daniel D. Daugherty daniel.daugherty at oracle.com
Thu Oct 11 19:23:27 UTC 2018


Thumbs up on the new additional text.

Dan


On 10/11/18 5:45 AM, David Holmes wrote:
> Although the JNI spec itself is not an open document I see no reason 
> not to get feedback on this simple update.
>
> Bug: https://bugs.openjdk.java.net/browse/JDK-6516521
>
> ---
> Chapter 5: The Invocation API
>
> Creating the VM
>
> The JNI_CreateJavaVM() function loads and initializes a Java VM and 
> returns a pointer to the JNI interface pointer. The thread that called 
> JNI_CreateJavaVM() is considered to be the main thread.
> ---
>
> Proposed additional text:
>
> *Note:* Depending on the operating system, the primordial process 
> thread may be subject to special handling that impacts its ability to 
> function properly as a normal Java thread (such as having a limited 
> stack size and being able to throw `StackOverflowError`). It is 
> strongly recommended that the primordial thread is not used to load 
> the Java VM, but that a new thread is created just for that purpose.
>
> Thanks,
> David
>



More information about the hotspot-runtime-dev mailing list