Ideally this information should be put as a comment into the code. It is
really not obvious.

On Tuesday 27 March 2018 04:37:19 Masaki Ota wrote:

Hi, 

We can get OTP page 0 value by EA EA E9 commands, but we cannot get it by EA EA EA E9.
As far as I remember, Device initialization finish at EA command, then sends EA EA E9 commands.
In this case we cannot get correct OTP page 0 value.
So I changed this order. (We can get OTP page 1 value by both of EA F0 F0 E9 and F0 F0 E9.)

Best Regards,
Masaki Ota
-----Original Message-----
From: Pali Rohár [mailto:pali.rohar@gmail.com] 
Sent: Monday, March 26, 2018 6:26 AM
To: 太田 真喜 Masaki Ota <redacted>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>; linux-input@vger.kernel.org; linux-kernel@vger.kernel.org; Aaron Ma <redacted>
Subject: Re: [PATCH] Input: ALPS - fix DualPoint flag for 74 03 28 devices

On Tuesday 20 March 2018 11:47:26 Dmitry Torokhov wrote:

quoted

On Mon, Jan 29, 2018 at 2:51 PM, dmitry.torokhov@gmail.com 
[off-list ref] wrote:

quoted

Hi,

On Thu, Nov 16, 2017 at 07:27:02AM +0000, Masaki Ota wrote:

quoted

Hi, Pali, Aaron,

Current code is correct device setting, previous code is wrong.
If the trackstick does not work(DUALPOINT flag disable), Device Firmware setting is wrong.

But recently I received the same report from Thinkpad L570 user, and I checked this device and found this device Firmware setting is wrong. Sorry for our mistake.
Is your laptop L570 ?

I will add code that supports the trackstick for this device.

Sorry for resurrecting this old thread, I am just trying to 
understand what went wrong here. Is the sequence of "f0 f0 e9" and 
"ea ea e9" is important in getting the correct OTP data and we 
originally got this order wrong? It is not clear from the original 
patch and discussion that this change was intentional.

Could I please get an answer to my question?

Thanks!

Masaki, this question is for you ↑↑↑

quoted

quoted

Thanks.

quoted

Best Regards,
Masaki Ota
-----Original Message-----
From: Pali Rohár [mailto:pali.rohar@gmail.com]
Sent: Wednesday, November 15, 2017 5:35 PM
To: 太田 真喜 Masaki Ota <redacted>
Cc: linux-input@vger.kernel.org; linux-kernel@vger.kernel.org; 
dmitry.torokhov@gmail.com; Aaron Ma [off-list ref]
Subject: Re: [PATCH] Input: ALPS - fix DualPoint flag for 74 03 28 
devices

On Wednesday 15 November 2017 14:34:04 Aaron Ma wrote:

quoted

There is a regression of commit 4a646580f793 ("Input: ALPS - fix 
two-finger scroll breakage"), ALPS device fails with log:

psmouse serio1: alps: Rejected trackstick packet from non 
DualPoint device

ALPS device with id "74 03 28" report OTP[0] data 0xCE after 
commit 4a646580f793, after restore the OTP reading order, it 
becomes to 0x10 as before and reports the right flag.

Fixes: 4a646580f793 ("Input: ALPS - fix two-finger scroll 
breakage")
Cc: <redacted>
Signed-off-by: Aaron Ma <redacted>
---
 drivers/input/mouse/alps.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/input/mouse/alps.c 

b/drivers/input/mouse/alps.c index 579b899add26..c59b8f7ca2fc 
100644

--- a/drivers/input/mouse/alps.c
+++ b/drivers/input/mouse/alps.c

@@ -2562,8 +2562,8 @@ static int alps_set_defaults_ss4_v2(struct 

psmouse *psmouse,

    memset(otp, 0, sizeof(otp));

-   if (alps_get_otp_values_ss4_v2(psmouse, 1, &otp[1][0]) ||
-       alps_get_otp_values_ss4_v2(psmouse, 0, &otp[0][0]))
+   if (alps_get_otp_values_ss4_v2(psmouse, 0, &otp[0][0]) ||
+       alps_get_otp_values_ss4_v2(psmouse, 1, &otp[1][0]))
            return -1;

    alps_update_device_area_ss4_v2(otp, priv);

Masaki Ota, please look at this patch as it partially revert your 
commit
4a646580f793 ("Input: ALPS - fix two-finger scroll breakage"). Something smells here.

--
Pali Rohár
pali.rohar@gmail.com

--
Dmitry

--
Pali Rohár
pali.rohar@gmail.com

-- 
Pali Rohár
pali.rohar@gmail.com