On Thu, Sep 1, 2016 at 10:03 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Michael Paquier <michael.paquier@gmail.com> writes:
>> Also, we could take one extra step forward then, and just introduce
>> ShmemAllocExtended that holds two flags as per the attached:
>> - SHMEM_ALLOC_ZERO that zeros all the fields
>> - SHMEM_ALLOC_NO_OOM that does not fail
>
> Don't see the point really ... it's just more API churn without any
> very compelling reason.
OK, then no objection to your approach. At least I tried.
--
Michael