[dpdk-dev] app/testpmd: add timer based fwd Rx queue flushing
Commit Message
Testpmd can stuck inside do while loop of the flush_fwd_rx_queues()
function. As non-zero packets are returned always by rte_eth_rx_burst()
function when compiled with no optimizations and if input line rate is
high. "do while" loop must exit at one stage to proceed further to
enable packet forwarding and forward the packets. So timer is set to
exit the do while loop after 1 second.
Fixes: af75078f ("first public release")
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
---
app/test-pmd/testpmd.c | 23 ++++++++++++++++++++++-
1 file changed, 22 insertions(+), 1 deletion(-)
Comments
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Reshma Pattan
> Sent: Friday, July 01, 2016 11:28 AM
> To: dev@dpdk.org
> Cc: Pattan, Reshma
> Subject: [dpdk-dev] [PATCH] app/testpmd: add timer based fwd Rx queue
> flushing
>
> Testpmd can stuck inside do while loop of the flush_fwd_rx_queues()
> function. As non-zero packets are returned always by rte_eth_rx_burst()
> function when compiled with no optimizations and if input line rate is
> high. "do while" loop must exit at one stage to proceed further to
> enable packet forwarding and forward the packets. So timer is set to
> exit the do while loop after 1 second.
>
> Fixes: af75078f ("first public release")
>
> Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
> > Testpmd can stuck inside do while loop of the flush_fwd_rx_queues()
> > function. As non-zero packets are returned always by rte_eth_rx_burst()
> > function when compiled with no optimizations and if input line rate is
> > high. "do while" loop must exit at one stage to proceed further to
> > enable packet forwarding and forward the packets. So timer is set to
> > exit the do while loop after 1 second.
> >
> > Fixes: af75078f ("first public release")
> >
> > Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
>
> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Applied, thanks
@@ -272,6 +272,9 @@ uint32_t bypass_timeout = RTE_BYPASS_TMT_OFF;
#endif
+/* default period is 1 second */
+static uint64_t timer_period = 1;
+
/*
* Ethernet device configuration.
*/
@@ -877,17 +880,35 @@ flush_fwd_rx_queues(void)
uint16_t nb_rx;
uint16_t i;
uint8_t j;
+ uint64_t prev_tsc = 0, diff_tsc, cur_tsc, timer_tsc = 0;
+
+ /* convert to number of cycles */
+ timer_period *= rte_get_timer_hz();
for (j = 0; j < 2; j++) {
for (rxp = 0; rxp < cur_fwd_config.nb_fwd_ports; rxp++) {
for (rxq = 0; rxq < nb_rxq; rxq++) {
port_id = fwd_ports_ids[rxp];
+ /**
+ * testpmd can stuck in the below do while loop
+ * if rte_eth_rx_burst() always returns nonzero
+ * packets. So timer is added to exit this loop
+ * after 1sec timer expiry.
+ */
+ prev_tsc = rte_rdtsc();
do {
nb_rx = rte_eth_rx_burst(port_id, rxq,
pkts_burst, MAX_PKT_BURST);
for (i = 0; i < nb_rx; i++)
rte_pktmbuf_free(pkts_burst[i]);
- } while (nb_rx > 0);
+
+ cur_tsc = rte_rdtsc();
+ diff_tsc = cur_tsc - prev_tsc;
+ timer_tsc += diff_tsc;
+ } while ((nb_rx > 0) &&
+ (timer_tsc < timer_period));
+ prev_tsc = cur_tsc;
+ timer_tsc = 0;
}
}
rte_delay_ms(10); /* wait 10 milli-seconds before retrying */