On 2021-02-19 21:28, Xie He wrote:
On Fri, Feb 19, 2021 at 10:39 AM Jakub Kicinski <kuba@xxxxxxxxxx>
wrote:
Not entirely sure what the argument is about but adding constants
would
certainly help.
Leon wants me to replace this:
dev->needed_headroom = 3 - 1;
with this:
/* 2 is the result of 3 - 1 */
dev->needed_headroom = 2;
But I don't feel his way is better than my way.
More fundamentally IDK if we can make such a fundamental change here.
When users upgrade from older kernel are all their scripts going to
work the same? Won't they have to bring the new netdev up?
Yes, this patch will break backward compatibility. Users with old
scripts will find them no longer working.
However, it's hard for me to find a better way to solve the problem
described in the commit message.
So I sent this as an RFC to see what people think about this. (Martin
Schiller seems to be OK with this.)
Well, I said I would like to test it.
I'm not really happy with this change because it breaks compatibility.
We then suddenly have 2 interfaces; the X.25 routings are to be set via
the "new" hdlc<x>_x25 interfaces instead of the hdlc<x> interfaces.
I currently just don't have a nicer solution to fix this queueing
problem either. On the other hand, since the many years we have been
using the current state, I have never noticed any problems with
discarded frames. So it might be more a theoretical problem than a
practical one.
I think users who don't use scripts can adapt quickly and users who
use scripts can also trivally fix their scripts.
Actually many existing commits in the kernel also (more or less) cause
some user-visible changes. But I admit this patch is a really big
change.
