[dpdk-dev] [PATCH v2] net/pcap: truncate packet if it is too large

[dpdk-dev] [PATCH v2] net/pcap: truncate packet if it is too large

[Zhike Wang]

From: Zhike Wang <wangzhike@jd.com>

Previously large packet would be dropped, instead now it is better to keep it via truncating it.

Signed-off-by: Zhike Wang <wangzk320@163.com>
---
 drivers/net/pcap/rte_eth_pcap.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index aa7ef6f..b4c79d1 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pcap/rte_eth_pcap.c
@@ -313,7 +313,7 @@ struct pmd_devargs_all {
 	struct pcap_pkthdr header;
 	pcap_dumper_t *dumper;
 	unsigned char temp_data[RTE_ETH_PCAP_SNAPLEN];
-	size_t len;
+	size_t len, caplen;
 
 	pp = rte_eth_devices[dumper_q->port_id].process_private;
 	dumper = pp->tx_dumper[dumper_q->queue_id];
@@ -325,28 +325,24 @@ struct pmd_devargs_all {
 	 * dumper */
 	for (i = 0; i < nb_pkts; i++) {
 		mbuf = bufs[i];
-		len = rte_pktmbuf_pkt_len(mbuf);
+		len = caplen = rte_pktmbuf_pkt_len(mbuf);
 		if (unlikely(!rte_pktmbuf_is_contiguous(mbuf) &&
 				len > sizeof(temp_data))) {
-			PMD_LOG(ERR,
-				"Dropping multi segment PCAP packet. Size (%zd) > max size (%zd).",
-				len, sizeof(temp_data));
-			rte_pktmbuf_free(mbuf);
-			continue;
+			caplen = sizeof(temp_data);
 		}
 
 		calculate_timestamp(&header.ts);
 		header.len = len;
-		header.caplen = header.len;
+		header.caplen = caplen;
 		/* rte_pktmbuf_read() returns a pointer to the data directly
 		 * in the mbuf (when the mbuf is contiguous) or, otherwise,
 		 * a pointer to temp_data after copying into it.
 		 */
 		pcap_dump((u_char *)dumper, &header,
-			rte_pktmbuf_read(mbuf, 0, len, temp_data));
+			rte_pktmbuf_read(mbuf, 0, caplen, temp_data));
 
 		num_tx++;
-		tx_bytes += len;
+		tx_bytes += caplen;
 		rte_pktmbuf_free(mbuf);
 	}
 
-- 
1.8.3.1

Re: [dpdk-dev] [PATCH v2] net/pcap: truncate packet if it is too large

[Ferriter, Cian]

Hi Zhike,

I've tested the behavior before and after this patch and can verify that the packets are being correctly truncated.

Code looks good to me too.

Reviewed-by: Cian Ferriter <cian.ferriter@intel.com>

Thanks,
Cian

> -----Original Message-----
> From: Zhike Wang <wangzk320@163.com>
> Sent: Monday 9 December 2019 01:53
> To: dev@dpdk.org
> Cc: wangzhike@jd.com; Yigit, Ferruh <ferruh.yigit@intel.com>; Ferriter, Cian
> <cian.ferriter@intel.com>; Zhike Wang <wangzk320@163.com>
> Subject: [PATCH v2] net/pcap: truncate packet if it is too large
> 
> From: Zhike Wang <wangzhike@jd.com>
> 
> Previously large packet would be dropped, instead now it is better to keep it
> via truncating it.
> 
> Signed-off-by: Zhike Wang <wangzk320@163.com>
> ---
>  drivers/net/pcap/rte_eth_pcap.c | 16 ++++++----------
>  1 file changed, 6 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/net/pcap/rte_eth_pcap.c
> b/drivers/net/pcap/rte_eth_pcap.c index aa7ef6f..b4c79d1 100644
> --- a/drivers/net/pcap/rte_eth_pcap.c
> +++ b/drivers/net/pcap/rte_eth_pcap.c
> @@ -313,7 +313,7 @@ struct pmd_devargs_all {
>  	struct pcap_pkthdr header;
>  	pcap_dumper_t *dumper;
>  	unsigned char temp_data[RTE_ETH_PCAP_SNAPLEN];
> -	size_t len;
> +	size_t len, caplen;
> 
>  	pp = rte_eth_devices[dumper_q->port_id].process_private;
>  	dumper = pp->tx_dumper[dumper_q->queue_id];
> @@ -325,28 +325,24 @@ struct pmd_devargs_all {
>  	 * dumper */
>  	for (i = 0; i < nb_pkts; i++) {
>  		mbuf = bufs[i];
> -		len = rte_pktmbuf_pkt_len(mbuf);
> +		len = caplen = rte_pktmbuf_pkt_len(mbuf);
>  		if (unlikely(!rte_pktmbuf_is_contiguous(mbuf) &&
>  				len > sizeof(temp_data))) {
> -			PMD_LOG(ERR,
> -				"Dropping multi segment PCAP packet. Size
> (%zd) > max size (%zd).",
> -				len, sizeof(temp_data));
> -			rte_pktmbuf_free(mbuf);
> -			continue;
> +			caplen = sizeof(temp_data);
>  		}
> 
>  		calculate_timestamp(&header.ts);
>  		header.len = len;
> -		header.caplen = header.len;
> +		header.caplen = caplen;
>  		/* rte_pktmbuf_read() returns a pointer to the data directly
>  		 * in the mbuf (when the mbuf is contiguous) or, otherwise,
>  		 * a pointer to temp_data after copying into it.
>  		 */
>  		pcap_dump((u_char *)dumper, &header,
> -			rte_pktmbuf_read(mbuf, 0, len, temp_data));
> +			rte_pktmbuf_read(mbuf, 0, caplen, temp_data));
> 
>  		num_tx++;
> -		tx_bytes += len;
> +		tx_bytes += caplen;
>  		rte_pktmbuf_free(mbuf);
>  	}
> 
> --
> 1.8.3.1
> 

Re: [dpdk-dev] [PATCH v2] net/pcap: truncate packet if it is too large

[Ferruh Yigit]

On 12/9/2019 11:26 AM, Ferriter, Cian wrote:

<...>

> 
>> -----Original Message-----
>> From: Zhike Wang <wangzk320@163.com>
>> Sent: Monday 9 December 2019 01:53
>> To: dev@dpdk.org
>> Cc: wangzhike@jd.com; Yigit, Ferruh <ferruh.yigit@intel.com>; Ferriter, Cian
>> <cian.ferriter@intel.com>; Zhike Wang <wangzk320@163.com>
>> Subject: [PATCH v2] net/pcap: truncate packet if it is too large
>>
>> From: Zhike Wang <wangzhike@jd.com>
>>
>> Previously large packet would be dropped, instead now it is better to keep it
>> via truncating it.
>>
>> Signed-off-by: Zhike Wang <wangzk320@163.com>
> 
> Reviewed-by: Cian Ferriter <cian.ferriter@intel.com>

Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Applied to dpdk-next-net/master, thanks.

(Updated signed-off tag to match author identity)