test/virtual_pmd: Add DEV_RX_OFFLOAD_CRC_STRIP to virtual driver
Checks
Commit Message
Add DEV_RX_OFFLOAD_CRC_STRIP to virtual drivers since they don't
use CRC at all, when an application requires this offload virtual
PMDs should not return error.
Fixes: b219c8c("test: enable HW CRC strip by default")
Signed-off-by: Phil Yang <phil.yang@arm.com>
---
test/test/virtual_pmd.c | 1 +
1 file changed, 1 insertion(+)
Comments
On 7/4/2018 2:31 AM, dev-bounces@dpdk.org wrote:
> Add DEV_RX_OFFLOAD_CRC_STRIP to virtual drivers since they don't
> use CRC at all, when an application requires this offload virtual
> PMDs should not return error.
Agreed.
Can you please check following patch [1] that has been applied to next-net which
adds DEV_RX_OFFLOAD_CRC_STRIP to virtual devices. Is this solves your issue?
[1]
https://patches.dpdk.org/patch/41957/
>
> Fixes: b219c8c("test: enable HW CRC strip by default")
>
> Signed-off-by: Phil Yang <phil.yang@arm.com>
> ---
> test/test/virtual_pmd.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/test/test/virtual_pmd.c b/test/test/virtual_pmd.c
> index f8ddc2d..591b309 100644
> --- a/test/test/virtual_pmd.c
> +++ b/test/test/virtual_pmd.c
> @@ -91,6 +91,7 @@ virtual_ethdev_info_get(struct rte_eth_dev *dev __rte_unused,
> dev_info->max_tx_queues = (uint16_t)512;
>
> dev_info->min_rx_bufsize = 0;
> + dev_info->rx_offload_capa = DEV_RX_OFFLOAD_CRC_STRIP;
> }
>
> static int
>
On 7/4/2018 10:34 AM, Ferruh Yigit wrote:
> On 7/4/2018 2:31 AM, dev-bounces@dpdk.org wrote:
>> Add DEV_RX_OFFLOAD_CRC_STRIP to virtual drivers since they don't
>> use CRC at all, when an application requires this offload virtual
>> PMDs should not return error.
>
> Agreed.
>
> Can you please check following patch [1] that has been applied to next-net which
> adds DEV_RX_OFFLOAD_CRC_STRIP to virtual devices. Is this solves your issue?
My bad, oversight your patch. You already aware of the patch mentioned and this
is to fix a missing part there. Patch looks good, thanks.
>
> [1]
> https://patches.dpdk.org/patch/41957/
>
>>
>> Fixes: b219c8c("test: enable HW CRC strip by default")
>>
>> Signed-off-by: Phil Yang <phil.yang@arm.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
>> ---
>> test/test/virtual_pmd.c | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/test/test/virtual_pmd.c b/test/test/virtual_pmd.c
>> index f8ddc2d..591b309 100644
>> --- a/test/test/virtual_pmd.c
>> +++ b/test/test/virtual_pmd.c
>> @@ -91,6 +91,7 @@ virtual_ethdev_info_get(struct rte_eth_dev *dev __rte_unused,
>> dev_info->max_tx_queues = (uint16_t)512;
>>
>> dev_info->min_rx_bufsize = 0;
>> + dev_info->rx_offload_capa = DEV_RX_OFFLOAD_CRC_STRIP;
>> }
>>
>> static int
>>
>
On 7/4/2018 10:45 AM, Ferruh Yigit wrote:
> On 7/4/2018 10:34 AM, Ferruh Yigit wrote:
>> On 7/4/2018 2:31 AM, dev-bounces@dpdk.org wrote:
>>> Add DEV_RX_OFFLOAD_CRC_STRIP to virtual drivers since they don't
>>> use CRC at all, when an application requires this offload virtual
>>> PMDs should not return error.
<...>
>>>
>>> Fixes: b219c8c("test: enable HW CRC strip by default")
>>>
>>> Signed-off-by: Phil Yang <phil.yang@arm.com>
>
> Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Fixes: 3d12dceed2df ("ethdev: add new offload flag to keep CRC")
Applied to dpdk-next-net/master, thanks.
@@ -91,6 +91,7 @@ virtual_ethdev_info_get(struct rte_eth_dev *dev __rte_unused,
dev_info->max_tx_queues = (uint16_t)512;
dev_info->min_rx_bufsize = 0;
+ dev_info->rx_offload_capa = DEV_RX_OFFLOAD_CRC_STRIP;
}
static int