wpa_supplicant 2.4/2.5 issue with some Intel chipsets

[Evangelos Foutras]

On 27/12/15 14:29, Peer, Ilan wrote:
> There is no specific description of the issues seen by the users but this is indeed possible. I've just sent 2 additional patches to handle this, can you give them a try?
> 
> http://patchwork.ozlabs.org/patch/561147/
> http://patchwork.ozlabs.org/patch/561146/
> 
> Thanks in advance,
> 
> Ilan.

I built test packages with patch 561146 and asked for user confirmation
that it solves the issue at hand. So far, I have had one confirmation
that it does!

Indeed, there is no single comprehensive report of the issue; my guess
is based on the following clues found in the main task downstream [1]:

- The issue only manifests when using netctl-auto which spawns another
process that attaches to wpa_supplicant; this is also the only scenario
where the -W flag is passed to wpa_supplicant.

- Passing an invalid -m flag to wpa_supplicant works around the problem
(I guess by failing to create a P2P interface?).

Something also clicked when you said that that the "Could not read
interface p2p-dev-wlp2s0 flags: No such device" message was harmless.

Cheers for the fix!

[1] https://bugs.archlinux.org/task/44731