[8u] RFR: Backport of 8067796 (process) Process.waitFor(timeout, unit) doesn't throw NPE if timeout is less than, or equal to zero when unit == null

[Sergey Nazarkin]

I’ve tracked down patchset required for clean backport. Beside 8037866 [2] only 8014066 [1] is required.  I’m a bit worrying about 8014066, the changeset may affect existent applications behavior. 

With 8014066 all subsequent patches applied cleanly except 8067796:
- for non-windows OSes waitFor method is at UNIXProcess.java (not ProcessImpl.java)
- windows version of ProcessImpl.java is placed at different path
- Basic.java: bug list is different

If 8014066 is accepted for backport I’ll create separate RFRs.  In this case patch sequence should be [4] -> [5] -> [6]

[1] https://bugs.openjdk.java.net/browse/JDK-8014066
[2] https://bugs.openjdk.java.net/browse/JDK-8037866
[3] https://bugs.openjdk.java.net/browse/JDK-8067796

[4] http://cr.openjdk.java.net/~snazarki/webrev/8014066/
[5] http://cr.openjdk.java.net/~snazarki/webrev/8037866/
[6] http://cr.openjdk.java.net/~snazarki/webrev/8067796/





Sergey Nazarkin




> On Mar 27, 2020, at 10:26, Andrew Hughes <gnu.andrew at redhat.com> wrote:
> 
> 
> 
> On 26/03/2020 17:39, Sergey Nazarkin wrote:
>> Yes, fixes are jdk8 specific and caused by refactoring introduced at http://hg.openjdk.java.net/jdk9/jdk9/jdk/rev/cb15bc14c26a.  
>> 
>> 
>> Sergey Nazarkin
>> 
>> 
> 
> Ah right, I see the problem; Fun becomes an interface after JDK-8037866,
> allowing the lambdas in the test case to implement it.
> 
> I think the right course here is to backport JDK-8037866. It's a test
> only change and no doubt other test backports will end up hitting this
> problem in the future. Your original webrev should then still be valid.
> 
> Thanks,
> -- 
> Andrew :)
> 
> Senior Free Java Software Engineer
> Red Hat, Inc. (http://www.redhat.com)
> 
> PGP Key: ed25519/0xCFDA0F9B35964222 (hkp://keys.gnupg.net)
> Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222
>