[PATCH 1/2] app/testpmd: fix stats get when display fwd stats

[Min Hu (Connor)]

Hi, Singh,

在 2022/4/27 20:15, Singh, Aman Deep 写道:
> 
> 
> On 4/6/2022 2:15 PM, Min Hu (Connor) wrote:
>> In function 'fwd_stats_display', if function 'rte_eth_stats_get' fails,
>> 'stats' is uncertainty value. The display result will be abnormal.
>>
>> This patch check the return value of 'rte_eth_stats_get' to avoid
>> display abnormal stats.
>>
>> Fixes: 53324971a14e ("app/testpmd: display/clear forwarding stats on 
>> demand")
>> Cc: stable at dpdk.org
>>
>> Signed-off-by: Min Hu (Connor) <humin29 at huawei.com>
>> ---
>>   app/test-pmd/config.c  | 10 ++++++++--
>>   app/test-pmd/testpmd.c | 16 ++++++++++++++--
>>   2 files changed, 22 insertions(+), 4 deletions(-)
>>
>> diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
>> index cc8e7aa138..bd689f9f86 100644
>> --- a/app/test-pmd/config.c
>> +++ b/app/test-pmd/config.c
>> @@ -249,14 +249,20 @@ nic_stats_display(portid_t port_id)
>>                                   diff_ns;
>>       uint64_t mpps_rx, mpps_tx, mbps_rx, mbps_tx;
>>       struct rte_eth_stats stats;
>> -
>>       static const char *nic_stats_border = "########################";
>> +    int ret;
>>       if (port_id_is_invalid(port_id, ENABLED_WARN)) {
>>           print_valid_ports();
>>           return;
>>       }
>> -    rte_eth_stats_get(port_id, &stats);
>> +    ret = rte_eth_stats_get(port_id, &stats);
>> +    if (ret != 0) {
>> +        fprintf(stderr,
>> +            "%s: Error: failed to get stats (port %u): %d",
>> +            __func__, port_id, ret);
>> +        return;
>> +    }
>>       printf("\n  %s NIC statistics for port %-2d %s\n",
>>              nic_stats_border, port_id, nic_stats_border);
>> diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
>> index fe2ce19f99..79bb23264b 100644
>> --- a/app/test-pmd/testpmd.c
>> +++ b/app/test-pmd/testpmd.c
>> @@ -1982,6 +1982,7 @@ fwd_stats_display(void)
>>       struct rte_port *port;
>>       streamid_t sm_id;
>>       portid_t pt_id;
>> +    int ret;
>>       int i;
>>       memset(ports_stats, 0, sizeof(ports_stats));
>> @@ -2013,7 +2014,13 @@ fwd_stats_display(void)
>>           pt_id = fwd_ports_ids[i];
>>           port = &ports[pt_id];
>> -        rte_eth_stats_get(pt_id, &stats);
>> +        ret = rte_eth_stats_get(pt_id, &stats);
>> +        if (ret != 0) {
>> +            fprintf(stderr,
>> +                "%s: Error: failed to get stats (port %u): %d",
>> +                __func__, pt_id, ret);
>> +            continue;
>> +        }
>>           stats.ipackets -= port->stats.ipackets;
>>           stats.opackets -= port->stats.opackets;
>>           stats.ibytes -= port->stats.ibytes;
>> @@ -2108,11 +2115,16 @@ fwd_stats_reset(void)
>>   {
>>       streamid_t sm_id;
>>       portid_t pt_id;
>> +    int ret;
>>       int i;
>>       for (i = 0; i < cur_fwd_config.nb_fwd_ports; i++) {
>>           pt_id = fwd_ports_ids[i];
>> -        rte_eth_stats_get(pt_id, &ports[pt_id].stats);
>> +        ret = rte_eth_stats_get(pt_id, &ports[pt_id].stats);
>> +        if (ret != 0)
>> +            fprintf(stderr,
>> +                "%s: Error: failed to clear stats (port %u):%d",
>> +                __func__, pt_id, ret);
> Should we clear "ports[pt_id].stats" in this condition.
when done 'rte_eth_stats_get',
if success, 'ports[pt_id].stats' will be cleared in this function.
if failed, no need to clear 'ports[pt_id].stats' as the log said.

>>       }
>>       for (sm_id = 0; sm_id < cur_fwd_config.nb_fwd_streams; sm_id++) {
>>           struct fwd_stream *fs = fwd_streams[sm_id];
> As such LGTM
Thanks for your comment.
> Acked-by: Aman Singh <aman.deep.singh at intel.com>
> .