This function applied an sf_ahead, bigger than the actual slot
indication message sending (so a message could NEVER arrive on time!).
Reduce to zero, because this is when it should arrive for us.

The previous designs seems to do:
loop {
  poll_ind_queue()
  if (msg)
    scheduler;

  pselect() on messages;
  handle_message {
    if slot_ind
      put_ind_queue()
  }
}

So basically, we artificially put a queue for slot indications in the
middle, but still handle it in the same thread(!). This for some reason
results in a big slow down if the PNF runs faster.

Now, we just do pselect(), waiting for messages. We handle the slot
indication immediately (the schedule should take some microseconds),
then return to pselect().

Assign NFAPI_NR_CONFIG_SLOT_CONFIG_TAG tag in TLVs used for transmitting the TDD Slot configuration through CONFIG.request.

The same function is called in nr_start_resp_cb(), why there and not here?

In the PNF, there is no gNBs section. So we cannot rely on it

Remove useless message bytes printing

Print message if callback missing

less newlines in log

logs

feat(doc): suggested changes: realtime tuned profile, kthread_cpus formulation and further details in PTP pining

feat(doc): update O-RAN 7.2 doc: details in kernel thread isolation, kernel thread isolation for solving PTP clock lag, PCI address of VFs