Re: [PATCH V2 net-next 0/6] net: hns3: enhance tc flow offload support

[Simon Horman <horms@kernel.org>]

On Thu, May 28, 2026 at 07:24:28PM +0800, Jijie Shao wrote:
> 
> on 2026/5/28 17:06, Paolo Abeni wrote:
> > On 5/27/26 3:12 PM, Jijie Shao wrote:
> > > on 2026/5/27 17:59, Simon Horman wrote:
> > > > On Sat, May 23, 2026 at 06:54:43PM +0800, Jijie Shao wrote:

...

> > > > Hi Juijie,
> > > > 
> > > > There are AI-generated reviews of this patch-set available at
> > > > https://netdev-ai.bots.linux.dev/ and https://sashiko.dev/
> > > Hi,
> > > 
> > > I'm sorry, I don't understand how to use https://netdev-ai.bots.linux.dev/, it seems to require an API Token?
> > > I find it difficult to locate my own patch-set in Recent Reviews.
> > The nipa sashiko instance has the same web UI as sashiko.dev. The full
> > URL for the former is:
> > 
> > https://netdev-ai.bots.linux.dev/sashiko/
> > 
> > You can lookup your series by the cover letter title or you can use the
> > direct link including the msg id for the cover:
> > 
> > https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260523105449.661848-1-shaojijie%40huawei.com
> 
> Thanks,
> 
> It appears that the two URLs mentioned by Simon Horman are the results of the one AI-generated reviews.

Hi,

Firstly sorry for truncating the netdev-ai URL, that was not my intention.
Secondly, let me explain the relationship between the AI-generated review
on netdev-ai and sashiko.dev.

Both use the same underlying review engine - Sashiko [1].  But they are
using different AI models to run the review.  And I believe that the
prompts also differ a little.  Overall, netdev-ai, has been tuned for the
Netdev review process.  While sashiko.dev is more general.

My observation is that netdev-ai tends to flag things relevant to the
Netdev development process that may be skipped by sashiko.dev. While
sashiko.dev has a tendency to report pre-existing issues that are not
strictly related to the topic of the patch under review.

But, as the underling engine is the same there is also a lot of overlap in
the output produced by these two instances. And in some cases both
instances flag all of the same issues.

[1] https://github.com/sashiko-dev/sashiko