ProcessImpl: Can we obsolete the vfork mode?

[Thomas Stüfe]

Side note, I did not find any tests for the VFORK mode. Even the FORK mode
seems not that well covered ( we only seem to run
java/lang/ProcessBuilder/Basic.java in FORK mode)

On Fri, May 16, 2025 at 6:59 PM Thomas Stüfe <thomas.stuefe at gmail.com>
wrote:

> I am fine with waiting past 25, though the continued existence of this
> mode makes me apprehensive. Back at SAP I have seen real customer issues
> from these kind of problems, and we only catched those out of accident.
>
> A separate deprecation for JDK 25 may be a good first step (a warning, but
> continued acceptance, of the VFORK mode)
>
> On Fri, May 16, 2025 at 6:20 PM Joseph D. Darcy <joe.darcy at oracle.com>
> wrote:
>
>> I concur that this work is better down early in JDK 26.
>>
>> Thanks,
>>
>> -Joe
>>
>> On 5/16/2025 12:26 AM, Thomas Stüfe wrote:
>>
>> Okay, moved the release version to 26.
>>
>> Could one of you (you or @Roger Riggs <roger.riggs at oracle.com> ) review
>> the CSR if you have the cycles to spare?
>>
>> https://bugs.openjdk.org/browse/JDK-8357090
>>
>> Thanks!
>>
>> On Fri, May 16, 2025 at 7:56 AM Alan Bateman <alan.bateman at oracle.com>
>> wrote:
>>
>>> On 16/05/2025 05:10, Thomas Stüfe wrote:
>>>
>>>
>>>
>>>>
>>> Thank you, Alan. I will prepare a patch and a CSR.
>>>
>>> Okay, and maybe something for early in JDK 26 to avoid touching this
>>> area close to the fork for JDK 25 RDP1.
>>>
>>> -Alan
>>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/core-libs-dev/attachments/20250517/465957da/attachment.htm>