[GitHub] [mynewt-nimble] wes3 commented on a change in pull request #399: host: Always wake tx queues on rx of num-cmp-pkts

[GitBox <gi...@apache.org>]

wes3 commented on a change in pull request #399: host: Always wake tx queues on rx of num-cmp-pkts
URL: https://github.com/apache/mynewt-nimble/pull/399#discussion_r270943312
 
 

 ##########
 File path: nimble/host/src/ble_hs_hci_evt.c
 ##########
 @@ -264,19 +256,14 @@ ble_hs_hci_evt_num_completed_pkts(uint8_t event_code, uint8_t *data, int len)
                     conn->bhc_outstanding_pkts -= num_pkts;
                 }
 
-                if (ble_hs_hci_avail_pkts == 0) {
 
 Review comment:
   Here is the section from the spec:
   While the Controller has HCI data packets in its buffer, it must keep sending the Number Of Completed Packets event to the Host at least periodically, until it finally reports that all the pending ACL Data Packets have been transmitted or flushed. The rate with which this event is sent is manufacturer specific.
   
   As Chris says, the controller will not send 0 in its num completed packet event unless there is data enqueued. The idea behind not doing this periodically is that we did not want to allocate an event buffer for it if not needed.
   
   I do not see why the controller would be prohibited from doing this but why should it? It is just extra work that does not need to be done (imo).

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services