[PATCH] Input: psmouse - reset device during probe

[PATCH] Input: psmouse - reset device during probe

[Daniel Drake]

New Acer laptops (Aspire A315-31 and several others) have an EC which
expects the PS/2 reset command to happen during boot.

On Linux, that means that currently the touchpad is unresponsive until
suspend/resume, when psmouse_reconnect() resets the device.

We observe that Windows sends a PS/2 reset command during boot.
Make Linux do the same here, making the touchpad usable on these
new products.

Signed-off-by: Daniel Drake <drake@endlessm.com>
---
 drivers/input/mouse/psmouse-base.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/input/mouse/psmouse-base.c b/drivers/input/mouse/psmouse-base.c
index f73b47b8c578..8f3c903db7b9 100644
--- a/drivers/input/mouse/psmouse-base.c
+++ b/drivers/input/mouse/psmouse-base.c
@@ -1591,6 +1591,12 @@ static int psmouse_connect(struct serio *serio, struct serio_driver *drv)
 	psmouse->resync_time = parent ? 0 : psmouse_resync_time;
 	psmouse->smartscroll = psmouse_smartscroll;
 
+	/* Reset the device, even though we would ordinarily expect it to
+	 * already be in the reset state. This matches Windows behaviour,
+	 * which some ECs require.
+	 */
+	psmouse_reset(psmouse);
+
 	psmouse_switch_protocol(psmouse, NULL);
 
 	if (!psmouse->protocol->smbus_companion) {
-- 
2.11.0

Re: [PATCH] Input: psmouse - reset device during probe

[Pali Rohár]

On Po 5. Jún 2017 22:27 Daniel Drake wrote:
> New Acer laptops (Aspire A315-31 and several others) have an EC which
> expects the PS/2 reset command to happen during boot.
> 
> On Linux, that means that currently the touchpad is unresponsive until
> suspend/resume, when psmouse_reconnect() resets the device.
> 
> We observe that Windows sends a PS/2 reset command during boot.
> Make Linux do the same here, making the touchpad usable on these
> new products.
> 
> Signed-off-by: Daniel Drake <drake@endlessm.com>
> ---
>  drivers/input/mouse/psmouse-base.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/drivers/input/mouse/psmouse-base.c
> b/drivers/input/mouse/psmouse-base.c index f73b47b8c578..8f3c903db7b9
> 100644 --- a/drivers/input/mouse/psmouse-base.c
> +++ b/drivers/input/mouse/psmouse-base.c
> @@ -1591,6 +1591,12 @@ static int psmouse_connect(struct serio *serio,
> struct serio_driver *drv)
>  psmouse->resync_time = parent ? 0 : psmouse_resync_time;
>  psmouse->smartscroll = psmouse_smartscroll;
>  
> +	/* Reset the device, even though we would ordinarily expect it to
> +	 * already be in the reset state. This matches Windows behaviour,
> +	 * which some ECs require.
> +	 */
> +	psmouse_reset(psmouse);
> +
>  psmouse_switch_protocol(psmouse, NULL);
>  
>  if (!psmouse->protocol->smbus_companion) {

This patch will slow down enumeration and initalization of *all* PS/2 mouses.

I guess if there are some problematic machines, it could be hooked...

-- 
Pali Rohár
pali.rohar@gmail.com

Re: [PATCH] Input: psmouse - reset device during probe

[Daniel Drake]

On Thu, Jun 8, 2017 at 1:32 PM, Pali Rohár <pali.rohar@gmail.com> wrote:
> This patch will slow down enumeration and initalization of *all* PS/2 mouses.

I know it's not ideal. but do you think anyone will be able to notice
the difference in initialization time?

Daniel

Re: [PATCH] Input: psmouse - reset device during probe

[Pali Rohár]

[-- Attachment #1: Type: Text/Plain, Size: 728 bytes --]

On Thursday 08 June 2017 22:18:36 Daniel Drake wrote:
> On Thu, Jun 8, 2017 at 1:32 PM, Pali Rohár <pali.rohar@gmail.com>
> wrote:
> > This patch will slow down enumeration and initalization of *all*
> > PS/2 mouses.
> 
> I know it's not ideal. but do you think anyone will be able to notice
> the difference in initialization time?

In past I was working on fixing detection of ALPS trackstick and Dmitry 
refused to add another unconditional reset during probe. That problem 
was fixed by calling reset routine only for ALPS devices and after 
detection that it is ALPS...

So I think Dmitry would complain about this unconditional reset for all 
PS/2 devices again.

-- 
Pali Rohár
pali.rohar@gmail.com

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]