[dpdk-dev] [PATCH] service: add new `rte_service_lcore_count_services` API

Pavan Nikhilesh Bhagavatula pbhagavatula at caviumnetworks.com
Mon Aug 28 13:06:09 CEST 2017


On Mon, Aug 28, 2017 at 10:55:05AM +0000, Van Haaren, Harry wrote:
> > From: Pavan Nikhilesh [mailto:pbhagavatula at caviumnetworks.com]
> > Sent: Wednesday, August 23, 2017 12:17 PM
> > To: Van Haaren, Harry <harry.van.haaren at intel.com>
> > Cc: dev at dpdk.org; Pavan Nikhilesh <pbhagavatula at caviumnetworks.com>
> > Subject: [dpdk-dev] [PATCH] service: add new `rte_service_lcore_count_services` API
> >
> > This new API returns the number of services that are running on a specific
> > service core. It allows an application to decide which service core to run
> > a new service on.
> >
> > Signed-off-by: Pavan Nikhilesh <pbhagavatula at caviumnetworks.com>
>
> Hi,

Hi Harry,

>
> Thanks - yes a useful addition to the API. Comments inline below, apart from that, thanks for the addition!
>
> -H

Thanks for the inputs, will modify the return types appropriately.
>
> > ---
> >  lib/librte_eal/common/include/rte_service.h | 12 ++++++++++++
> >  lib/librte_eal/common/rte_service.c         | 11 +++++++++++
> >  2 files changed, 23 insertions(+)
> >
> > diff --git a/lib/librte_eal/common/include/rte_service.h
> > b/lib/librte_eal/common/include/rte_service.h
> > index 7c6f738..5f0904f 100644
> > --- a/lib/librte_eal/common/include/rte_service.h
> > +++ b/lib/librte_eal/common/include/rte_service.h
> > @@ -374,6 +374,18 @@ int32_t rte_service_lcore_list(uint32_t array[], uint32_t n);
> >   * @warning
> >   * @b EXPERIMENTAL: this API may change without prior notice
> >   *
> > + * Get the numer of services running on the supplied lcore.
> > + *
> > + * @param Lcore Id of the service core.
> > + * @retval >=0 Number of services registered to this core.
> > + * @retval -EINVAL The provided lcore is not a service core.
> > + */
> > +int32_t rte_service_lcore_count_services(uint32_t lcore);
>
> Perhaps -ENOTSUP would be clearer for "not a service core"?
> Then -EINVAL becomes available to indicate an invalid lcore-id pass in?
>
> > +
> > +/**
> > + * @warning
> > + * @b EXPERIMENTAL: this API may change without prior notice
> > + *
> >   * Dumps any information available about the service. If service is NULL,
> >   * dumps info for all services.
> >   */
> > diff --git a/lib/librte_eal/common/rte_service.c b/lib/librte_eal/common/rte_service.c
> > index 7efb76d..ddaed22 100644
> > --- a/lib/librte_eal/common/rte_service.c
> > +++ b/lib/librte_eal/common/rte_service.c
> > @@ -397,6 +397,17 @@ rte_service_lcore_list(uint32_t array[], uint32_t n)
> >  }
> >
> >  int32_t
> > +rte_service_lcore_count_services(uint32_t lcore)
> > +{
>
> /* need to error check lcore here, return -EINVAL if not valid. */
>
> > +	struct core_state *cs = &lcore_states[lcore];
> > +
> > +	if (!cs->is_service_core)
> > +		return -EINVAL;
>
> /* suggesting to use -ENOTSUP instead here */
>
> > +
> > +	return __builtin_popcountll(cs->service_mask);
> > +}
> > +
> > +int32_t
> >  rte_service_start_with_defaults(void)
> >  {
> >  	/* create a default mapping from cores to services, then start the
> > --
> > 2.7.4
>


More information about the dev mailing list