[PATCH] wireless-regdb: add regulatory rules for Iraq (IQ)

[Chen-Yu Tsai]

On Tue, May 19, 2026 at 4:55 PM Johannes Berg <johannes at sipsolutions.net> wrote:
>
> Hi,
>
> > As for 10 MHz, I don't think it is supported, so we don't really consider
> > that case. I could be wrong though.
> >
> > Johannes, could you shed some light on 10 MHz channel width support?
>
> It's ... complicated?
>
> We support a _very_ limited set of 5/10 MHz operation, but I think most
> of it is fairly much unreachable from userspace. I asked about removing
> it entirely a few years ago, and some people were opposed, but said
> people also haven't actually helped take care of it or anything ... I
> was just tempted again a little while back to remove it due to the rates
> issues.
>
> Ever since my commit 5add321c329b ("wifi: cfg80211: remove scan_width
> support") I believe it has been unreachable on the *client* side, but
> given that we still have some support in _nl80211_parse_chandef() and
> chandef functions, I expect that it would be possible to still configure
> an AP or monitor interface with 5/10 MHz, though only with drivers that
> have WIPHY_FLAG_SUPPORTS_5_10_MHZ, i.e. ath5k, ath9k and hwsim. I
> wouldn't mind removing it all though.
>
>
> Is this a concern from a regulatory POV right now, due to say power
> density by channel width? This feels somewhat familiar even with higher
> channel widths - maybe it's time to add such an attribute to the regdb?

There are some regulations that say X dBm for 10 MHz or X*2 dBm for 20 MHz.
I've mostly been ignoring 10 MHz since that doesn't seem to be a thing
in modern WiFi. This is slightly different from the newer PSD rules we
are seeing with 6 GHz, since it only gives two points instead of an
actual density value.


ChenYu