Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20201112200906.991086-1-kuabhs@chromium.org> <20201112200856.v2.1.Ia526132a366886e3b5cf72433d0d58bb7bb1be0f@changeid>
 <CAD=FV=XKCLgL6Bt+3KfqKByyP5fpwXOh6TNHXAoXkaQJRzjKjQ@mail.gmail.com>
 <002401d6c242$d78f2140$86ad63c0$@codeaurora.org> <CAD=FV=UnecON-M9eZVQePuNpdygN_E9OtLN495Xe1GL_PA94DQ@mail.gmail.com>
 <002d01d6c2dd$4386d880$ca948980$@codeaurora.org> <CAD=FV=WQPMnor3oTefDHd6JP6UmpyBo7UsOJ1Sg4Ly1otxr6hw@mail.gmail.com>
 <004301d6c968$12ef1b10$38cd5130$@codeaurora.org>
In-Reply-To: <004301d6c968$12ef1b10$38cd5130$@codeaurora.org>
From:   Doug Anderson <dianders@chromium.org>
Date:   Thu, 3 Dec 2020 07:33:52 -0800
Message-ID: <CAD=FV=VCbjRUxUsmyk=64FLDGU=W41EXh5tdfQr1Lg83T8jiEA@mail.gmail.com>
Subject: Re: [PATCH v2 1/1] ath10k: add option for chip-id based BDF selection
To:     Rakesh Pillai <pillair@codeaurora.org>
Cc:     Abhishek Kumar <kuabhs@chromium.org>,
        Kalle Valo <kvalo@codeaurora.org>,
        LKML <linux-kernel@vger.kernel.org>,
        ath10k <ath10k@lists.infradead.org>,
        Brian Norris <briannorris@chromium.org>,
        linux-wireless <linux-wireless@vger.kernel.org>,
        "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        netdev <netdev@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi,

On Thu, Dec 3, 2020 at 3:33 AM Rakesh Pillai <pillair@codeaurora.org> wrote:
>
> > What I'm trying to say is this.  Imagine that:
> >
> > a) the device tree has the "variant" property.
> >
> > b) the BRD file has two entries, one for "board-id" (1) and one for
> > "board-id + chip-id" (2).  It doesn't have one for "board-id + chip-id
> > + variant" (3).
> >
> > With your suggestion we'll see the "variant" property in the device
> > tree.  That means we'll search for (1) and (3).  (3) isn't there, so
> > we'll pick (1).  ...but we really should have picked (2), right?
>
> Do we expect board-2.bin to not be populated with the bdf with variant field (if its necessary ?)

The whole fact that there is a fallback to begin with implies that
there can be a mismatch between the board-2.bin and the device tree
file.  Once we accept that there can be a mismatch, it seems good to
try all 3 fallbacks in order.


> Seems fine for me, if we have 2 fallback names if that is needed.

OK, sounds good.  So hopefully Abhishek can post a v3 based on what's
in <https://crrev.com/c/2556437> and you can confirm you're good with
it there?

-Doug