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: <57e9f582340545c79a9ab3433e2a4a7d@realtek.com> <87cztx87tr.fsf@tynnyri.adurom.net>
 <CA+ASDXM9YmhMWmttkrCxB0_tf6Kf6Sz=7Yc26CmjZmMUxyxAfw@mail.gmail.com>
 <87tun7i2rn.fsf@codeaurora.org> <876c29841fcf4d21ab2440cbb5879de2@realtek.com>
In-Reply-To: <876c29841fcf4d21ab2440cbb5879de2@realtek.com>
From:   Brian Norris <briannorris@chromium.org>
Date:   Thu, 13 May 2021 14:08:53 -0700
Message-ID: <CA+ASDXMN__j0x5hN5t98C4k+upOcs+T1=MrivCbxRYa_-ENt_A@mail.gmail.com>
Subject: Re: Request to update 8852a firmware of rtw89 to v0.13.8.0
To:     Pkshih <pkshih@realtek.com>
Cc:     Kalle Valo <kvalo@codeaurora.org>,
        "linux-firmware@kernel.org" <linux-firmware@kernel.org>,
        "linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, May 13, 2021 at 2:38 AM Pkshih <pkshih@realtek.com> wrote:
> > -----Original Message-----
> > From: kvalo=codeaurora.org@mg.codeaurora.org [mailto:kvalo=codeaurora.org@mg.codeaurora.org] On
> > Behalf Of Kalle Valo

> > It can create confusion to the users if during review we make changes
> > how firmware files are handled. Some drivers have windows style .ini
> > files which are not ok in an upstream driver, there could be changes in
> > the file format etc.

Sure, good point. I figured if there's no driver merged anywhere
upstream, people (e.g., me) are taking their own life in their hands
trying to utilize arbitrary versions from the mailing list. But I
could see why you still don't want the potential mismatch.

> > But I have no strong opinion here, my main motivation is just to try to
> > keep things simple for maintaining the "interface" between
> > linux-firmware and kernel wireless drivers. My preference is that the
> > firmware files are ready for submission when a new driver is submitted
> > for review, but the firmware files are submitted only after the driver
> > is accepted.

Ack, that's definitely important to me. The first versions had no
publicly posted firmware, and at least one person (not me at the time)
asked for it. My extension of that is that I don't want to
_discourage_ vendors posting their firmware to public repos :)

> I think I can maintain the firmware in GitHub before driver is accepted,
> and add some information about firmware version and link in the cover of
> patchset. Then, people can take proper firmware.

Sure, that can work.

Thanks,
Brian