[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20171108054759.GN3187@localhost>
Date: Wed, 8 Nov 2017 11:17:59 +0530
From: Vinod Koul <vinod.koul@...el.com>
To: Sinan Kaya <okaya@...eaurora.org>
Cc: Robin Murphy <robin.murphy@....com>, dmaengine@...r.kernel.org,
timur@...eaurora.org, linux-arm-msm@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH 2/3] dmaengine: qcom_hidma: add support for the new
revision
On Mon, Nov 06, 2017 at 05:09:29PM -0500, Sinan Kaya wrote:
> On 11/6/2017 1:15 PM, Sinan Kaya wrote:
> > On 11/6/2017 1:03 PM, Robin Murphy wrote:
> >>> #ifdef CONFIG_ACPI
> >>> ret = strcmp(acpi_device_hid(adev), "QCOM8062");
> >>> + if (ret)
> >>> + ret = strcmp(acpi_device_hid(adev), "QCOM8063");
> >> This string-juggling looks to have already hit the point at which it
> >> doesn't scale well - it would be a lot nicer to make use of
> >> of_device_get_match_data() and the ACPI equivalent to abstract the
> >> version-specific data appropriately.
> >
> > Sure, let me do some research.
> >
>
> I just wanted to double check here.
>
> This is what I can do:
> 1. Maintain different match tables for different driver capabilities.
> 2. Instead of doing open-coded strcmp, I can do match against different tables
> using acpi_match_device/of_match_device.
>
> Hope this works for you.
Sounds better to me. You should have common match where capabilities are
same.
--
~Vinod
Powered by blists - more mailing lists