Hi Brian,
On 01/10, Brian Gix wrote:
> This allows one App using the mesh daemon to provision another.
This patch contains also some wording cosmetics and other PB-ADV fixes
that don't seem related to the loopback mechanism.
Could you please split this into a patchset?
> +static void loop_adv(struct pb_adv_session *session, const uint8_t *data,
> + uint16_t len)
> +{
> + struct idle_rx *rx = l_new(struct idle_rx, 1);
> +
> + rx->session = session;
> + rx->len = len;
> + memcpy(rx->data, data, len);
> +
> + l_idle_oneshot(idle_rx_adv, rx, NULL);
> +}
(...)
> + if (session->loop)
> + loop_adv(session->loop, buf, init_size + 10);
> + else
> + mesh_send_pkt(MESH_IO_TX_COUNT_UNLIMITED, 200,
> + buf, init_size + 10);
This condition check is repeated throughout the code. I think it would
be clearer to replace mesh_send_pkt usages with something like:
static void pb_adv_send(struct pb_adv_session *session, uint8_t count,
uint16_t interval, void *data, uint16_t len)
{
if (session->loop) {
struct idle_rx *rx = ...
} else {
mesh_send_pkt(count, interval, data, len);
}
}
--
Michał Lowas-Rzechonek <michal.lowas-rzechonek@xxxxxxxxxxx>
Silvair http://silvair.com
Jasnogórska 44, 31-358 Krakow, POLAND
