On Sun, Nov 12, 2023 at 9:39 PM Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> On Sun, 12 Nov 2023 at 10:54, Kalle Valo <kvalo@xxxxxxxxxx> wrote:
> >
> > Linus, I suspect it will take a while before we make a new wireless pull
> > request (there's currently just one commit in the wireless tree). So if
> > you want to take this directly to your tree please go ahead, otherwise
> > you get it via the normal route in a week or two. Just let us know what
> > you prefer.
> >
> > Acked-by: Kalle Valo <kvalo@xxxxxxxxxx>
>
> Ok, I just tested it in my private tree, and it fixes the WARN_ON() as
> expected, so I'll apply it for real.
Good to know - thanks.
>
> However, now that I don't have that big warning in there, I do note
> another iwlwifi issue that is new to this meger window:
>
> debugfs: Directory 'iwlmvm' with parent 'netdev:wlo2' already present!
> iwlwifi 0000:45:00.0: Failed to create debugfs directory under netdev:wlo2
> debugfs: Directory 'iwlmvm' with parent 'netdev:wlo2' already present!
> iwlwifi 0000:45:00.0: Failed to create debugfs directory under netdev:wlo2
>
> and looking at my system logs, this is new.
>
> It looks like it is probably due to
>
> c36235acb34f ("wifi: iwlwifi: mvm: rework debugfs handling")
> e9dd25550770 ("wifi: iwlwifi: mvm: add a per-link debugfs")
>
> but that's just from looking at the patches (ie no bisection or any
> real effort).
Don't bother, we'll take a look.
>
> This is after the system has come up, so I assume it's when
> networkmanager or whatever actually is setting that wireless thing up.
Most likely the network-manager brings the device up several times or something
like that, but let's not blame them before we check what happens internally.
>
> Again: I actually don't *use* the WiFi on this machine - but from a
> quick check it does seem to work. So this is just an annoyance and a
> sign that somebody didn't do all the details right (possibly triggered
> by odd user land behavior, of course, but still...)
>
Oh, we know the rules :)
> Linus
