[LTP] [PATCH v2] userfaultfd: Add test using UFFDIO_POISON

[Ricardo Branco]

On 2/18/26 11:49 AM, Cyril Hrubis wrote:
> I still do not understand why we need the barrier here. The thread that
> tries to access the memory should block until this thread processes the
> event sicne the usefaultfd range was registered before the memory is
> accessed.
>
> Also the same pattern is present in rest of the usefaultfd tests, so
> either we need this barrier in all of them, or it shouldn't be needed
> here.
>
> And the test seems to work just fine if I remove the barrier code.
I submitted a v3 without the barrier code.  Seems safe.

Best,
R