RFR: 8283606: Tests may fail with zh locale on MacOS [v2]

[Vikey Chen]

On Mon, 28 Mar 2022 17:11:54 GMT, Ioi Lam <iklam at openjdk.org> wrote:

>> In the longer term, I think its worth considering implementing LC_ALL as an override. I created an enhancement request for purposes of discussion.
>> [JDK-8283697](https://bugs.openjdk.java.net/browse/JDK-8283697) On macOS, the locale cannot be set using LC_ALL environment value
>> 
>> Jtreg does propagate LC_ALL to the subprocesses it spawns and it is documented.
>> So I'm not sure an immediate fix is necessary if the manual workaround (changing macOs preferences)
>> is suitable for a couple of weeks.
>
>> In the longer term, I think its worth considering implementing LC_ALL as an override. I created an enhancement request for purposes of discussion. [JDK-8283697](https://bugs.openjdk.java.net/browse/JDK-8283697) On macOS, the locale cannot be set using LC_ALL environment value
>> 
>> Jtreg does propagate LC_ALL to the subprocesses it spawns and it is documented. So I'm not sure an immediate fix is necessary if the manual workaround (changing macOs preferences) is suitable for a couple of weeks.
> 
> Roger, do you plan to implement the LC_ALL fix shortly? If so, that may make this PR unnecessary.

@iklam @dholmes-ora Can you please sponsor this PR or review the library version, since 
 https://bugs.openjdk.java.net/browse/JDK-8283697
> the Posix shell variables LANG, and LC_* on macOs do not provide the same information for region and variant that are provided by the Core Foundation locale settings.
And the timing of changes made via the Control Panel is different

-------------

PR: https://git.openjdk.java.net/jdk/pull/7924