RFR(S): JDK-8043780 Use open(O_CLOEXEC) instead of fcntl(FD_CLOEXEC)

[Martin Buchholz]

Comments reworked as suggested and webrev refreshed.
http://cr.openjdk.java.net/~martin/webrevs/openjdk9/O_CLOEXEC/

On Wed, Oct 22, 2014 at 7:48 PM, David Holmes <david.holmes at oracle.com>
wrote:

> Hi Martin,
>
> On 23/10/2014 7:34 AM, Martin Buchholz wrote:
>
>> Re-ping.
>>
>> I have a simple change, reviewed by hotspot engineer, and I just need a
>> jprt run.
>>
>> Optionally, we could also use a simple port to other platforms that might
>> support O_CLOEXEC, but that can be done later.
>>
>
> Functionally this seems okay to me. The only suggestion I have is to move
> the large comment block about close-on-exec to the point where you set
> O_CLOEXEC and explain the attempt and fallback.
>
> If you prepare the final changeset I will sponsor it for you.
>
> Cheers,
> David
>
>
>  On Tue, Jul 1, 2014 at 2:49 PM, Martin Buchholz <martinrb at google.com>
>> wrote:
>>
>>  Ping. Please review this change.
>>>
>>> It is in production use locally on Linux.  Please either accept this
>>> change as is (Linux only), or work with this patch to create a more
>>> ambitious patch that addresses other Unix variants.
>>>
>>> Adding porters-dev.
>>>
>>>
>>> On Thu, May 22, 2014 at 10:14 AM, Martin Buchholz <martinrb at google.com>
>>> wrote:
>>>
>>>  Hi,
>>>>
>>>> I'd like to contribute a patch.
>>>>
>>>> Webrev:
>>>>    http://cr.openjdk.java.net/~martin/webrevs/openjdk9/O_CLOEXEC/
>>>>
>>>> Bug:
>>>>    https://bugs.openjdk.java.net/browse/JDK-8043780
>>>>
>>>> If accepted, corresponding changes should be made to the other Unix
>>>> variants, but I can't do any testing there.
>>>>
>>>>
>>>
>>>