linux-kernel - Re: [PATCH] wifi: mwifiex: enable host mlme on sdio W8997 chipsets

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <aDdCP_lvlyvk9J3n@google.com>
Date: Wed, 28 May 2025 10:05:03 -0700
From: Brian Norris <briannorris@...omium.org>
To: Francesco Dolcini <francesco@...cini.it>
Cc: rafael@...ms.me, Rafael Beims <rafael.beims@...adex.com>,
	linux-wireless@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] wifi: mwifiex: enable host mlme on sdio W8997 chipsets

Hi Francesco,

On Fri, May 23, 2025 at 09:26:04AM +0200, Francesco Dolcini wrote:
> On Thu, May 22, 2025 at 10:24:14AM -0700, Brian Norris wrote:
> > On Wed, May 21, 2025 at 07:19:34AM -0300, rafael@...ms.me wrote:
> > > From: Rafael Beims <rafael.beims@...adex.com>
> > > 
> > > Enable the host mlme flag for W8997 chipsets so WPA3 can be used.
> > > This feature depends on firmware support (V2 API key), which may not be
> > > available in all available firmwares.
> > 
> > Is it available in *any* W8997 firmware? Or particularly, is it
> > available in the firmware in linux-firmware.git? Judging by its git
> > history, the answer is "no", in which case this is definitely NAK'd.
> 
> mrvl/sdsd8997_combo_v4.bin, from linux-firmware GIT, 16.92.21.p137
> version. From an off-list chat with Rafael he confirmed me that this is
> what he used for testing.

linux-firmware.git claims to hold W16.68.1.p197.1. Either WHENCE is
wrong, or that sounds like a completely different branch. Are you sure
about that?

> > users on the old FW version. So, we'd need to teach the driver to know
> > the difference between v1 and v2 API here, and choose accordingly.
> 
> This is already implemented. From mwifiex_ret_get_hw_spec()
> 
> ```
>         if (adapter->key_api_major_ver != KEY_API_VER_MAJOR_V2)
>                 adapter->host_mlme_enabled = false;
> ```


Huh, I forgot about that. I also assumed "v2 API" wasn't something
available on most older chips, but I guess I might have been wrong.

> To me the patch is ok.

The firmware versions above don't match up to me. But if we confirm
that:
(a) this API is available in a linux-firmware.git firmware for this chip
    and
(b) someone includes a valid linux-firmware.git version that they tested
    in the changelog

then maybe this is OK. As it stands now though, no.

Brian