RFR: 8260930: AARCH64: Invalid value passed to critical JNI function

[Andrew Haley]

On 2/4/21 3:30 AM, Yangfei (Felix) wrote:
> Here, the aarch64 VM crashes when executing an existing regression test.

>  And backporting the fix should be low in risk given that that this feature is rarely used so far. 
> I am also worried that people with 8u may create new code or modify existing code making use of the critical JNI feature for performance reasons. 

I guess that's possible.

> Instead of crashing and possibly giving people the impression that the aarch64 VM is not quite stable, I think it will be better to tell them explicitly that this feature is not supported on this platform. 
> I will let you 8u maintainers to decide. 
> 
> [1] https://wiki.openjdk.java.net/display/jdk8u/Guidelines+for+working+on+jdk8u

OK, if you like, but bear in mind that it's significant fixes only. 8u is supposed
to be the long-term-stable release.

-- 
Andrew Haley  (he/him)
Java Platform Lead Engineer
Red Hat UK Ltd. <https://www.redhat.com>
https://keybase.io/andrewhaley
EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671