The status bits included in the report-id field of periodic OA reports aren't documented for Haswell, so we could potentially reserve this field for later use by explicitly zeroing it before copying it.
That said; while writting i-g-t tests I inferred that bit 10 of this field does seem to reliably indiciate that a report was timer triggered which is useful when trying to test the periodic sampling exponent.
If we don't reserve/zero the field then i-g-t can also verify that this field != 0 when testing if a report has landed somewhere (since zero is a valid value for all other fields).