Re: [v6] net/af_xdp: enable uds_path instead of use_cni

[Maryam Tahhan <mtahhan@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1446 bytes --]

On 21/12/2023 11:36, David Marchand wrote:
> Hello,
>
> On Fri, Dec 15, 2023 at 4:40 PM Maryam Tahhan<mtahhan@redhat.com>  wrote:
>> With the original 'use_cni' implementation, (using a
>> hardcoded socket rather than a configurable one),
>> if a DPDK pod is requesting multiple net devices
>> and these devices are from different pools, then
>> the container attempts to mount all the netdev UDSes
>> in the pod as /tmp/afxdp.sock. Which means that at best
>> only 1 netdev will handshake correctly with the AF_XDP
>> DP. This patch addresses this by making the socket
>> parameter configurable using a new vdev param called
>> 'uds_path' and removing the previous 'use_cni' param.
>> This patch also fixes incorrect references to the
>> AF_XDP DP as CNI and updates the documentation with a
>> working example. This change has been tested with the
> If there are fixes mixed in, please separate them in dedicated patches
> so we can backport them to LTS releases.

Thanks for the Feedback David, I don't quite understand the ask here, is 
to to have a separate patch for documentation to fixes? As previous 
feedback in the mailing list was to keep the documentation changes with 
the fix.

There is one fix in this patchset (for for broken socket behaviour). The 
rest of the changes were fixing up all the documentation. So do I need 
to create a patchset with multiple patches, one for the c code and 
separate ones for the documentation?

[-- Attachment #2: Type: text/html, Size: 2111 bytes --]