odp_init_global -> odp_term_global This pair work for me, I assume that odp_term_global takes no arguments, any things it needs to know are already known as global data in the implementation. It should just return success or failure - possibly setting errno ?
Mike On 14 October 2014 07:59, Mike Holmes <[email protected]> wrote: > > > On 14 October 2014 03:53, Taras Kondratiuk <[email protected]> > wrote: > >> On 10/13/2014 09:56 PM, Mike Holmes wrote: >> >>> Unit testing replicates a situation similar to an NFV case where the >>> workload >>> is varied without rebooting the HW. >>> In this use case there have to be hooks for the application to inform the >>> implementation that HW needs to be released. >>> >> >> odp_init_global -> odp_term_global ? >> What about local termination? > > > I hoped to tackle them one at a time to define how the API signature > should look, but yes, we need that too. > > >> >> >> >>> Signed-off-by: Mike Holmes <[email protected]> >>> --- >>> >>> This patch is intended to specify the API, the implementation for it will >>> require a number of other APIs to be created before it is complete. >>> If accepted the follow on work would require APIs and implementations for >>> >>> odp_shm_terminate_global >>> odp_thread_terminate_global >>> odp_buffer_pool_terminate_global >>> odp_queue_terminate_global >>> odp_schedule_terminate_global >>> odp_pktio_terminate_global >>> odp_timer_terminate_global >>> odp_crypto_terminate_global >>> >> >> IMO it is an implementation details. Do you want to make them public? >> > > No, but presumably to actually implement the termination these areas will > need addressing so I wanted to make the scope clear for what had to happen > under the covers. > > > -- > *Mike Holmes* > Linaro Sr Technical Manager > LNG - ODP > -- *Mike Holmes* Linaro Sr Technical Manager LNG - ODP
_______________________________________________ lng-odp mailing list [email protected] http://lists.linaro.org/mailman/listinfo/lng-odp
