[dpdk-dev] [PATCH 06/11] ethdev: extend ethdev to support security APIs

Radu Nicolau radu.nicolau at intel.com
Fri Sep 22 13:49:42 CEST 2017


Hi,


On 9/18/2017 8:57 AM, Jerin Jacob wrote:
> -----Original Message-----
>> Date: Thu, 14 Sep 2017 13:56:46 +0530
>> From: Akhil Goyal <akhil.goyal at nxp.com>
>> To: dev at dpdk.org
>> CC: declan.doherty at intel.com, pablo.de.lara.guarch at intel.com,
>>   hemant.agrawal at nxp.com, radu.nicolau at intel.com, borisp at mellanox.com,
>>   aviadye at mellanox.com, thomas at monjalon.net, sandeep.malik at nxp.com,
>>   jerin.jacob at caviumnetworks.com
>> Subject: [PATCH 06/11] ethdev: extend ethdev to support security APIs
>> X-Mailer: git-send-email 2.9.3
>>
>> From: Declan Doherty <declan.doherty at intel.com>
>>   /**
>> @@ -907,6 +912,7 @@ struct rte_eth_conf {
>>   #define DEV_RX_OFFLOAD_QINQ_STRIP  0x00000020
>>   #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
>>   #define DEV_RX_OFFLOAD_MACSEC_STRIP     0x00000080
>> +#define DEV_RX_OFFLOAD_SECURITY         0x00000100
>>   
>>   /**
>>    * TX offload capabilities of a device.
>> @@ -926,6 +932,11 @@ struct rte_eth_conf {
>>   #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO   0x00001000    /**< Used for tunneling packet. */
>>   #define DEV_TX_OFFLOAD_MACSEC_INSERT    0x00002000
>>   #define DEV_TX_OFFLOAD_MT_LOCKFREE      0x00004000
>> +#define DEV_TX_OFFLOAD_SECURITY         0x00008000
> For the documentation side, Sharing the same views as Shahaf's
>
>> +#define DEV_TX_OFFLOAD_SEC_NEED_MDATA   0x00010000
> If i understand correctly, it is more of a fixup if driver needs.I guess
> we can change to negative logic reflect it as offload.
I can make a point for positive logic: if there is no need to update the 
mdata then neither PMD nor app need to do anything. If there is, the PMD 
needs to set the flag and the app needs to check it.

>
>> +#define DEV_TX_OFFLOAD_IPSEC_CRYPTO_HW_TRAILER 0x00020000
>> +#define DEV_TX_OFFLOAD_IPSEC_CRYPTO_TSO        0x00040000
>> +#define DEV_TX_OFFLOAD_IPSEC_CRYPTO_CKSUM      0x00080000
>>   /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
>>    * tx queue without SW lock.
>>    */
> The above text should come after DEV_TX_OFFLOAD_MT_LOCKFREE(Position got
> changed in this patch)
Thanks, will fix.
>
>> @@ -1651,6 +1662,9 @@ struct rte_eth_dev {
>>   	enum rte_eth_dev_state state; /**< Flag indicating the port state */
>>   } __rte_cache_aligned;
>>   



More information about the dev mailing list