Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH v1 2/2] ASoC: tegra: Unify ASoC machine drivers
To:     Mark Brown <broonie@kernel.org>, Rob Herring <robh@kernel.org>
Cc:     Thierry Reding <thierry.reding@gmail.com>,
        Jonathan Hunter <jonathanh@nvidia.com>,
        Takashi Iwai <tiwai@suse.com>,
        Jaroslav Kysela <perex@perex.cz>,
        Ion Agorria <ion@agorria.com>,
        Svyatoslav Ryhel <clamor95@gmail.com>,
        Frank Rowand <frowand.list@gmail.com>,
        Liam Girdwood <lgirdwood@gmail.com>,
        linux-tegra@vger.kernel.org, devicetree@vger.kernel.org,
        alsa-devel@alsa-project.org, linux-kernel@vger.kernel.org
References: <20210518001356.19227-1-digetx@gmail.com>
 <20210518001356.19227-3-digetx@gmail.com>
 <20210518180949.GA949047@robh.at.kernel.org>
 <20210518183455.GE4358@sirena.org.uk>
From:   Dmitry Osipenko <digetx@gmail.com>
Message-ID: <b924f76a-56dd-5ca3-229d-293f7ff1e81a@gmail.com>
Date:   Tue, 18 May 2021 23:16:48 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.8.1
MIME-Version: 1.0
In-Reply-To: <20210518183455.GE4358@sirena.org.uk>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

18.05.2021 21:34, Mark Brown пишет:
> On Tue, May 18, 2021 at 01:09:49PM -0500, Rob Herring wrote:
>> On Tue, May 18, 2021 at 03:13:56AM +0300, Dmitry Osipenko wrote:
> 
>>> Squash all machine drivers into a single-universal one. This reduces
>>> code duplication, eases addition of a new drivers and upgrades older
>>> code to a modern Linux kernel APIs.
> 
>> Nice, I never understood why each codec needed it's own machine driver 
>> (and typically in turn compatible string).
> 
> It's generally the clocking configuration, things with fancy clock trees
> don't play so nicely with generic systems where we want to configure
> them at runtime, or things with complex routing in the CPU where we
> only recently started getting bindings that were functional enough to
> use.  Tegra is fairly complex internally.
> 

This is true, but still it's more optimal to use flags in a single
common driver for the minor differences rather than to duplicate 90% of
the code. In practice majority of tegra boards are derived from the same
reference design, thus the h/w configurations are mostly the same.