Re: phy timestamping "library"

[Richard Cochran <richardcochran@gmail.com>]

On Tue, Nov 26, 2019 at 07:00:02PM +0100, Michael Walle wrote:
> So what I'd like to have in a seperate "library" is the following:
>  - skb queues handling
>  - rx timestamp pool handling
>  - rx timestamp timeout handling
>  - skb delivery to the network stack on either match or timeout

I would resist the temptation to turn this into a "library", unless
there is a clear benefit.

> Now I have the following problem: the original driver attaches more data
> to the timestamp, as is my driver; this data is used by the match
> function.

...

> But how can I give that memory to the lib in phyts_init() and how can
> phyts_init() populate its rx_pool_data. Or maybe I'm getting it all
> wrong and there is a better way ;)

Sounds like the re-factoring of a couple of lists into a "library" is
not worth the effort.  The differences in the matching logic will make
the "library" baroque.

I suggest not over-engineering your new driver.  In any case, it is
hard to say without seeing the patches.

Thanks,
Richard