[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YAknNNgx66HDtrgQ@google.com>
Date: Wed, 20 Jan 2021 23:03:16 -0800
From: Dmitry Torokhov <dmitry.torokhov@...il.com>
To: Vincent Knecht <vincent.knecht@...loo.org>
Cc: Rob Herring <robh+dt@...nel.org>,
Henrik Rydberg <rydberg@...math.org>,
Michael Srba <Michael.Srba@...nam.cz>,
Linus Walleij <linus.walleij@...aro.org>,
linux-input@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, ~postmarketos/upstreaming@...ts.sr.ht
Subject: Re: [PATCH 2/2] Input: add MStar msg26xx touchscreen driver
Hi Vincent,
On Wed, Jan 20, 2021 at 07:01:08PM +0100, Vincent Knecht wrote:
> +struct packet {
> + u8 y_high : 4;
> + u8 x_high : 4;
This will not work on big endian devices as order of bitfields changes.
I'd recommended treating contact packet as sequence of bytes and parse,
i.e.
x = ((buf[0] & 0x0f) << 8) | buf[1];
x = ((buf[0] & 0xf0) << 4) | buf[2];
...
> + u8 x_low;
> + u8 y_low;
> + u8 pressure;
> +};
> +
> +
> +static void mstar_power_on(struct msg26xx_ts_data *msg26xx)
> +{
> + gpiod_set_value(msg26xx->reset_gpiod, 0);
> + mdelay(10);
> + gpiod_set_value(msg26xx->reset_gpiod, 1);
> + mdelay(FIRMWARE_ON_DELAY);
I am pretty sure this is incorrect. You are saying that you release the
reset line, wait a bit, and then assert it. gpiod is a logical API, with
0 being inactive and 1 being active, and here you want to activate the
reset line, wait appropriate time, release it, and wait for the device
to initialize. What does the datasheet say about reset GPIO polarity?
Thanks.
--
Dmitry
Powered by blists - more mailing lists