RFR: JDK-8212828 Allow POSIX_SPAWN to be used for ProcessImpl on Linux

[Roger Riggs]

Hi Thomas,

The webrev looks fine.

Please remove the @author tag in the Linux (2nd) test block in Basic.java.
Author tags are losing favor and there's no need to repeat it.

I ran the change through our tests without errors.

I'd give it another 24hours before pushing in case anyone else wants to 
review it.

Thanks, Roger

p.s. The issue to change the default is: 
https://bugs.openjdk.java.net/browse/JDK-8213192


On 10/30/2018 11:41 AM, Thomas Stüfe wrote:
> Hi Roger,
>
>
> On Tue, Oct 30, 2018 at 3:46 PM Roger Riggs <Roger.Riggs at oracle.com> wrote:
>> Hi Thomas,
>>
>> On 10/29/2018 12:04 PM, Thomas Stüfe wrote:
>>
>> Hi Roger,
>>
>> On Thu, Oct 25, 2018 at 10:45 PM Roger Riggs <Roger.Riggs at oracle.com> wrote:
>>
>> Hi Thomas,
>>
>> In an abundance of caution, I was thinking that it would be a change right
>> at the beginning of a new release so it gets the most exercise and
>> users in early access, etc.
>>
>> Okay, I understand that.
>>
>> Over the next days I will run tests with posix_spawn enabled by
>> default on our landscape. We have many different Linuxes on different
>> architectures and different levels of glibc, so this is a reasonable
>> test.
>>
>> If I do not encounter red flags, I would consider the posix_spawn path
>> tested well enough to ship it at least as a non-default, experimental
>> option. Like David originally intended. Then, start of next release,
>> we can make it default and see how that goes.
>>
>> Does that sound ok to you?
>>
>>
>> That's fine, until it becomes the default, it will be opt in.
>>
>> Is there an updated webrev with the corrected test executions?
>>
> Here you go.
>
> http://cr.openjdk.java.net/~stuefe/webrevs/JDK-8212828-posix_spawn-on-linux/webrev_0/webrev/
>
> Tested locally, seems to work fine. I am re-running the submit tests too.
>
>