Re: [PATCH -next 01/10] wifi: cfg80211: pre-assign cookie for remain_on_channel, mgmt_tx, probe_peer and tx_control_port

Op 26 juli 2026 22:29:09 schreef Jeff Johnson <[email protected]>:

> On 7/26/2026 12:37 PM, Johannes Berg wrote:
>> On Sun, 2026-07-26 at 21:25 +0200, Arend van Spriel wrote:
>>> Having a single place for cookie assignment and keeping that
>>> responsibility in the cfg80211 subsystem is a logical choice as it
>>> handles the userspace nl80211 API. add_nan_func already does this:
>>> cfg80211 calls cfg80211_assign_cookie() before invoking the driver.
>>> Apply the same pattern to remain_on_channel, mgmt_tx, probe_peer and
>>> tx_control_port by pre-assigning the cookie in the nl80211 command
>>> handlers before the rdev_* call. For tx_control_port the cookie is
>>> only pre-assigned when the caller requests an ack (cookie pointer
>>> non-NULL).
>>>
>>> Drivers may still overwrite the value for now; subsequent patches will
>>> remove per-driver cookie generation. Update the cfg80211_ops kerneldoc
>>> to note that the cookie is pre-assigned and drivers must not modify it.
>>>
>>> Assisted-by: Claude:claude-sonnet-4-6
>>
>> Hah, I _knew_ it, just based on the text style, before even reading
>> beyond the cover letter ... I'm starting to even be able to tell the
>> companies apart (not models yet, i.e. you could've told me this was
>> Opus). E.g. "Patch 1 is the enabling change: [...]", "Non-obvious
>> aspects" or "This means the cookie field is threaded through" etc.
>>
>> Fun. What weird times we live in.
>
> Indeed. And yes, I tend to edit the LLM commit text since the tense is often
> wrong and the vocabulary often isn't want I'd use.

@Johannes: Glad you can see the fun of it ;-)

Agree that LLM text show some typical choice of words and phrasing. I only 
edit (or prompt to correct) the text when I do not understand what is 
written, too much words are used, or missing details I care about.

>
>
>>
>> But you forgot S-o-b.
>
> also wrong branch tag: s/-next/wireless-next/

Patchwork seems to have no problem determining the branch/repo.

Will repost with the S-o-B in place.

Gr. AvS
>