Re: [PATCH] ath10k: remove not needed warning when peer unmap

[Kalle Valo <kvalo@qca.qualcomm.com>]

Janusz Dziedzic <janusz.dziedzic@tieto.com> writes:

> Remove not needed warning when get peer unmap
> event from the firmware. This is not critical
> message. Instead print this as a debug message.

I don't agree with that statement. If that would be true, we could
remove a lot of warnings from ath10k. We have these warnings to catch
problems early, which again improves the quality of the driver.

Your commit log was again missing the "why?" part. I assume the reason
for this patch is the problem of seeing the warning "unknown peer id 2"
when putting the interface is down, which again is a spurious event from
the firmware? You should document that in the commit log as well as add
a short comment to the code explaining why we only print a debug message
when that happens.

Other idea I had would be to keep the warning message but add a new test
to detect this problematic case, but I guess for that we would need to
add a new state "stopping" to catch that? For example, something like
this:

if state == stopping and event->id == 2
        dbg("foo")
else
        warn("bar")

-- 
Kalle Valo