[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAAQ0ZWQEhitxNCKBjTNGcv5E+yqwjfbb5GCLoctpJM7u4Zpp=A@mail.gmail.com>
Date: Thu, 10 Dec 2020 22:40:36 +0800
From: Shawn Guo <shawn.guo@...aro.org>
To: Uwe Kleine-König
<u.kleine-koenig@...gutronix.de>
Cc: Bjorn Andersson <bjorn.andersson@...aro.org>,
Andrzej Hajda <a.hajda@...sung.com>,
Neil Armstrong <narmstrong@...libre.com>,
Laurent Pinchart <Laurent.pinchart@...asonboard.com>,
Jonas Karlman <jonas@...boo.se>,
Jernej Skrabec <jernej.skrabec@...l.net>,
David Airlie <airlied@...ux.ie>,
Daniel Vetter <daniel@...ll.ch>,
Thierry Reding <thierry.reding@...il.com>,
Lee Jones <lee.jones@...aro.org>,
Doug Anderson <dianders@...omium.org>,
linux-arm-msm <linux-arm-msm@...r.kernel.org>,
dri-devel <dri-devel@...ts.freedesktop.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
linux-pwm@...r.kernel.org
Subject: Re: [PATCH] drm/bridge: ti-sn65dsi86: Implement the pwm_chip
Hi Uwe,
On Thu, Dec 10, 2020 at 9:05 PM Uwe Kleine-König
<u.kleine-koenig@...gutronix.de> wrote:
> > > @@ -111,6 +118,8 @@
> > >
> > > #define SN_LINK_TRAINING_TRIES 10
> > >
> > > +#define SN_PWM_GPIO 3
> >
> > So this maps to the GPIO4 described in sn65dsi86 datasheet. I'm
> > wondering if it's more readable to define the following SHIFT constants
> > (your code), and use GPIO_MUX_GPIO4_SHIFT >> 2 where you need GPIO
> > offset?
> >
> > #define GPIO_MUX_GPIO1_SHIFT 0
> > #define GPIO_MUX_GPIO2_SHIFT 2
> > #define GPIO_MUX_GPIO3_SHIFT 4
> > #define GPIO_MUX_GPIO4_SHIFT 6
> >
> > If you agree, you may consider to integrate this patch beforehand:
> >
> > https://github.com/shawnguo2/linux/commit/7cde887ffb3b27a36e77a08bee3666d14968b586
>
> My preferred way here would be to add a prefix for the other constants.
> It (IMHO) looks nicer and
>
> GPIO_INPUT_SHIFT
>
> looks like a quite generic name for a hardware specific definition.
While this looks like a reasonable argument, I also like the naming
choice for these constants in the beginning for that distinction
between registers and bits. And changing the names the other way
around means there will be a much bigger diffstat, which I would like
to avoid. I suggest let's just focus on what really matters here -
keep the naming consistent, so that people do not get confused when
they want to add more constants in there.
Shawn
> (Even if up to now there is no other code location using this name.)
Powered by blists - more mailing lists