Message-ID: <7f996d2efd23cefd17074edaeed0a6bbbd9f1a99.camel@sipsolutions.net>
Subject: Re: [PATCH 1/3] wifi: ipw2x00: convert ipw_fw_error->elem to
 flexible array[]
From:   Johannes Berg <johannes@sipsolutions.net>
To:     Jacob Keller <jacob.e.keller@intel.com>
Cc:     Kalle Valo <kvalo@kernel.org>, linux-wireless@vger.kernel.org,
        Stanislav Yakovlev <stas.yakovlev@gmail.com>
Date:   Tue, 28 Feb 2023 18:46:36 +0100
In-Reply-To: <d393ba90-ecdd-ffea-540b-d6db15571d5b@intel.com>
References: <20230228162827.3876606-1-jacob.e.keller@intel.com>
         <a8798dce4ae87aee64dfd56721b1668f8c969951.camel@sipsolutions.net>
         <d393ba90-ecdd-ffea-540b-d6db15571d5b@intel.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.3 (3.46.3-1.fc37) 
MIME-Version: 1.0
Precedence: bulk

On Tue, 2023-02-28 at 09:44 -0800, Jacob Keller wrote:
>=20
> Previous to this change, error struct has two pointers to sections of
> memory allocated at the end of the buffer.
>=20
> The code used to be:
>=20
> -	error =3D kmalloc(sizeof(*error) +
> -			sizeof(*error->elem) * elem_len +
> -			sizeof(*error->log) * log_len, GFP_ATOMIC);
>=20
> i.e. the elem_len is multiplying sizeof(*error->elem).
>=20
> The code is essentially trying to get two flexible arrays in the same
> allocation, and its a bit messy to do that. I don't see how elem_len
> could be anything other than "number of elems" given this code I removed.

Yeah, you're right. I was thinking of more modern HW/FW too much I
guess, I see now even in the driver we have an array walk here (and it
trusts the elem_len from firmware... ahrg!)

> I posted these mainly because I was trying to resolve all of the hits
> that were found by the coccinelle patch I made, posted at [1]. I wanted
> to get it to run clean so that we had no more struct_size hits.
>=20
> Dropping this would just make that patch have some hits until the driver
> is removed, eventually...
>=20
> Not really a big deal to me, I just didn't want to post a coccinelle
> patch without also trying to fix the handful of problems it reported,
> since the total number of reports was small.
>=20

Makes sense. I don't think we'll drop the driver at any point soon, but
I also don't see it being changed much :)

johannes