Re: [dpdk-dev] [PATCH v1 00/20] net/txgbe: add VF driver support

[Ferruh Yigit <ferruh.yigit@intel.com>]

On 2/3/2021 7:48 AM, Jiawen Wu wrote:
> On February 3, 2021 2:08 AM, Ferruh Yigit wrote:
>> On 1/22/2021 9:47 AM, Jiawen Wu wrote:
>>> This patch adds txgbevf PMD as a part of txgbe PMD.
>>>
>>> Jiawen Wu (20):
>>>     net/txgbe: add ethdev probe and remove for VF device
>>>     net/txgbe: add base code for VF driver
>>>     net/txgbe: support add and remove VF device MAC address
>>>     net/txgbe: get VF device information
>>>     net/txgbe: add interrupt operation for VF device
>>>     net/txgbe: get link status of VF device
>>>     net/txgbe: add Rx and Tx unit init for VF device
>>>     net/txgbe: add VF device stats and xstats get operation
>>>     net/txgbe: add VLAN handle support to VF driver
>>>     net/txgbe: add RSS support for VF device
>>>     net/txgbe: add VF device promiscuous and allmulticast mode
>>>     net/txgbe: support multicast MAC filter for VF driver
>>>     net/txgbe: support to update MTU on VF device
>>>     net/txgbe: support register dump on VF device
>>>     net/txgbe: start and stop VF device
>>>     net/txgbe: add some supports as PF driver implemented
>>>     net/txgbe: support VF representor
>>>     net/txgbe: hardware support for VF representor
>>>     net/txgbe: support VLAN filter for VF representor
>>>     doc: update release note for txgbe
>>>
>>
>> Hi Jiawen,
>>
>> I put some comments on some patches, can you please send a new version
>> with updates?
>>
>> And I want to note that it is specially hard to review this PMD because it is
>> clone of complete and complex ixgbe driver.
>> So most of the code is already ready, we are tying to structure it logically for
>> upstreaming it but it is not possible to review the details of this much feature in
>> this short time.
>>
>> For those feature we are relying on the Wangxun to test all these features, and
>> to confirm this, can it be possible for Wangxun to send some test results for
>> this release?
>>
> 
> Hi Ferruh,
> 
> Should I provide test results of PF part applied to the main repo, and VF part with
> some modifications? I mean, the latest version that I format the patches.
> 

Hi Jiawen,

Sorry for the delay, I wasn't sure how to put this, there is no official defined 
process around it for now.

Testing reports can help on both:
1) A release with a new PMD can put some proof that new PMD is working as expected.
2) Gives some confidence for new patches to be merged.


>> Similar to those reports:
>> https://patches.dpdk.org/patch/84436/
>> https://patches.dpdk.org/patch/84500/
>> https://patches.dpdk.org/patch/84517/
>>
> 
> Those reports show the NICs supported platforms. All features in dts report don't
> to be needed, do they?
> 

This is to cover the case 1) above, as you said this is to document the 
platform/NIC tested for that release.

>>
>> Also quick test results for the -rc2 & -rc3 can be very helpful, if you can provide,
>> like:
>> http://inbox.dpdk.org/dev/c2bd98a6da8d4955b19b8838c9bf60b7@intel.com/#
>> t
>> http://inbox.dpdk.org/dev/92ce632a-dc21-f50c-d4d6-85e7961b6e3b@linux.vne
>> t.ibm.com/#t
>>

And this is for 2) above, it can have more details on the features tested, so it 
gives confidence and gives chance to detect and fix defect for the release.