| Message ID | 1405923752-3412-1-git-send-email-venkatesh.vivekanandan@linaro.org |
|---|---|
| State | New |
| Headers | show |
I think this patch should be named differently. Maxim. On 07/21/2014 10:22 AM, venkatesh.vivekanandan@linaro.org wrote: > From: Venkatesh Vivekanandan <venkatesh.vivekanandan@linaro.org> > > - Fix to give transmitted buffers back to buffer pool. > - If there are issues seen while sending traffic, ONLY then this patch > has to be applied. Intention of this patch is to fix the traffic > issue. > > Signed-off-by: Venkatesh Vivekanandan <venkatesh.vivekanandan@linaro.org> > --- > platform/linux-dpdk/odp_packet_io.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/platform/linux-dpdk/odp_packet_io.c b/platform/linux-dpdk/odp_packet_io.c > index d8d127f..980a15c 100644 > --- a/platform/linux-dpdk/odp_packet_io.c > +++ b/platform/linux-dpdk/odp_packet_io.c > @@ -230,6 +230,9 @@ int odp_pktio_recv(odp_pktio_t id, odp_packet_t pkt_table[], unsigned len) > if (pktio_entry == NULL) > return -1; > > + if (pktio_entry->s.params.type == ODP_PKTIO_TYPE_DPDK) > + odp_pktio_send(id, pkt_table, 0); > + > lock_entry(pktio_entry); > pkts = recv_pkt_dpdk(&pktio_entry->s.pkt_dpdk, pkt_table, len); > unlock_entry(pktio_entry);
As this is just a temporary fix, it is ok to leave it this way. Please let me know if there is any specific information presence in the subject. Venky On 21 July 2014 13:46, Maxim Uvarov <maxim.uvarov@linaro.org> wrote: > I think this patch should be named differently. > > Maxim. > > > On 07/21/2014 10:22 AM, venkatesh.vivekanandan@linaro.org wrote: > >> From: Venkatesh Vivekanandan <venkatesh.vivekanandan@linaro.org> >> >> - Fix to give transmitted buffers back to buffer pool. >> - If there are issues seen while sending traffic, ONLY then this patch >> has to be applied. Intention of this patch is to fix the traffic >> issue. >> >> Signed-off-by: Venkatesh Vivekanandan <venkatesh.vivekanandan@linaro.org> >> --- >> platform/linux-dpdk/odp_packet_io.c | 3 +++ >> 1 file changed, 3 insertions(+) >> >> diff --git a/platform/linux-dpdk/odp_packet_io.c >> b/platform/linux-dpdk/odp_packet_io.c >> index d8d127f..980a15c 100644 >> --- a/platform/linux-dpdk/odp_packet_io.c >> +++ b/platform/linux-dpdk/odp_packet_io.c >> @@ -230,6 +230,9 @@ int odp_pktio_recv(odp_pktio_t id, odp_packet_t >> pkt_table[], unsigned len) >> if (pktio_entry == NULL) >> return -1; >> + if (pktio_entry->s.params.type == ODP_PKTIO_TYPE_DPDK) >> + odp_pktio_send(id, pkt_table, 0); >> + >> lock_entry(pktio_entry); >> pkts = recv_pkt_dpdk(&pktio_entry->s.pkt_dpdk, pkt_table, len); >> unlock_entry(pktio_entry); >> > > > _______________________________________________ > lng-odp mailing list > lng-odp@lists.linaro.org > http://lists.linaro.org/mailman/listinfo/lng-odp >
Looks like dpdk does the same hack:
examples/multi_process/client_server_mp/mp_client/client.c
for (;;) {
uint16_t i, rx_pkts = PKT_READ_SIZE;
uint8_t port;
/* try dequeuing max possible packets first, if that
fails, get the
* most we can. Loop body should only execute once,
maximum */
while (rx_pkts > 0 &&
unlikely(rte_ring_dequeue_bulk(rx_ring, pkts, rx_pkts) != 0))
rx_pkts =
(uint16_t)RTE_MIN(rte_ring_count(rx_ring), PKT_READ_SIZE);
if (unlikely(rx_pkts == 0)){
if (need_flush)
for (port = 0; port < ports->num_ports;
port++)
send_packets(ports->id[port]);
need_flush = 0;
continue;
}
for (i = 0; i < rx_pkts; i++)
handle_packet(pkts[i]);
On 07/23/2014 03:31 PM, Venkatesh Vivekanandan wrote:
> As this is just a temporary fix, it is ok to leave it this way. Please
> let me know if there is any specific information presence in the subject.
>
> Venky
>
>
> On 21 July 2014 13:46, Maxim Uvarov <maxim.uvarov@linaro.org
> <mailto:maxim.uvarov@linaro.org>> wrote:
>
> I think this patch should be named differently.
>
> Maxim.
>
>
> On 07/21/2014 10:22 AM, venkatesh.vivekanandan@linaro.org
> <mailto:venkatesh.vivekanandan@linaro.org> wrote:
>
> From: Venkatesh Vivekanandan
> <venkatesh.vivekanandan@linaro.org
> <mailto:venkatesh.vivekanandan@linaro.org>>
>
> - Fix to give transmitted buffers back to buffer pool.
> - If there are issues seen while sending traffic, ONLY then
> this patch
> has to be applied. Intention of this patch is to fix the
> traffic
> issue.
>
> Signed-off-by: Venkatesh Vivekanandan
> <venkatesh.vivekanandan@linaro.org
> <mailto:venkatesh.vivekanandan@linaro.org>>
> ---
> platform/linux-dpdk/odp_packet_io.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/platform/linux-dpdk/odp_packet_io.c
> b/platform/linux-dpdk/odp_packet_io.c
> index d8d127f..980a15c 100644
> --- a/platform/linux-dpdk/odp_packet_io.c
> +++ b/platform/linux-dpdk/odp_packet_io.c
> @@ -230,6 +230,9 @@ int odp_pktio_recv(odp_pktio_t id,
> odp_packet_t pkt_table[], unsigned len)
> if (pktio_entry == NULL)
> return -1;
> + if (pktio_entry->s.params.type == ODP_PKTIO_TYPE_DPDK)
> + odp_pktio_send(id, pkt_table, 0);
> +
> lock_entry(pktio_entry);
> pkts = recv_pkt_dpdk(&pktio_entry->s.pkt_dpdk,
> pkt_table, len);
> unlock_entry(pktio_entry);
>
>
>
> _______________________________________________
> lng-odp mailing list
> lng-odp@lists.linaro.org <mailto:lng-odp@lists.linaro.org>
> http://lists.linaro.org/mailman/listinfo/lng-odp
>
>
diff --git a/platform/linux-dpdk/odp_packet_io.c b/platform/linux-dpdk/odp_packet_io.c index d8d127f..980a15c 100644 --- a/platform/linux-dpdk/odp_packet_io.c +++ b/platform/linux-dpdk/odp_packet_io.c @@ -230,6 +230,9 @@ int odp_pktio_recv(odp_pktio_t id, odp_packet_t pkt_table[], unsigned len) if (pktio_entry == NULL) return -1; + if (pktio_entry->s.params.type == ODP_PKTIO_TYPE_DPDK) + odp_pktio_send(id, pkt_table, 0); + lock_entry(pktio_entry); pkts = recv_pkt_dpdk(&pktio_entry->s.pkt_dpdk, pkt_table, len); unlock_entry(pktio_entry);