[vox-tech] Shutting off tap-to-click on a Sony Vaio VGN-N270E laptop running Ubuntu Edgy

Issac Trotts issac.trotts at gmail.com
Wed Apr 4 16:59:15 PDT 2007


Ubuntu mostly works great on this laptop, except that wireless access
required writing a script and by default the tap-to-click misfeature is
enabled.  Wireless is working fine now, but I haven't yet figured how to
remove tap-to-click.

The touchpad is an ALPS GlidePoint (http://www.alps.co.jp/index_e.htm), or
at least that's what I gather from inserting some printk's into the mouse
driver and running dmesg.  This is confirmed by cat /proc/bus/input/devices,
which says among other things

I: Bus=3D0011 Vendor=3D0002 Product=3D0001 Version=3D7321
N: Name=3D"PS/2 ALPS GlidePoint"
P: Phys=3Disa0060/serio4/input0
S: Sysfs=3D/class/input/input4
H: Handlers=3Dmouse0 event1 ts0
B: EV=3D7
B: KEY=3D70000 0 0 0 0 0 0 0 0
B: REL=3D3



Here are the things I have tried so far:
- sudo tpconfig -t 0.  This has no effect at all.  I'm using tpconfig
version 3.1.3, which I guess is supposed to work with ALPS.

- Various things people recommended about Section "InputDevice", Identifier
"Synaptics Touchpad" etc. in xorg.conf.  None if it did any good.
Eventually I removed that Section "InputDevice" and the touchpad still
works, since I still have another InputDevice Section with Driver "mouse"
and apparently the touchpad is being detected as a mouse.

- Ripping out all the code for tap-to-click in the synaptics_drv driver and
installing it.  Of course this had no effect because the touchpad is an
ALPS, though I didn't know it yet.

- Modifying /usr/src/linux-source-2.6.17/drivers/input/mouse/alps.c,
specifically changing alps_tap_mode(psmouse, 1) to alps_tap_mode(psmouse, 0)
in the places where it occurs.  I then recompiled the kernel using
make-kpkg, installed it, rebooted, and still had the same problem.

- There is a comment above the function alps_tap_mode() that suggests it
won't work with stickpointer / GlidePoint pads like mine and gives some
advice on how to change it.  My attempt to follow the advice led to the
following new version of that function, which DOES NOT WORK, fyi:

/*
 * Turn touchpad tapping on or off. The sequences are:
 * 0xE9 0xF5 0xF5 0xF3 0x0A to enable,
 * 0xE9 0xF5 0xF5 0xE8 0x00 to disable.
 * My guess that 0xE9 (GetInfo) is here as a sync point.
 * For models that also have stickpointer (DualPoints) its tapping
 * is controlled separately (0xE6 0xE6 0xE6 0xF3 0x14|0x0A) but
 * we don't fiddle with it.
 */
static int alps_tap_mode(struct psmouse *psmouse, int enable)
{
        struct ps2dev *ps2dev =3D &psmouse->ps2dev;
        /* I'm not sure if the comment above is accurate.
         * If it is then we should just use setrate in both cases
         * for stickpointer (whatever that is).
         * I'm not sure how to check for stickpointer in code, but
         * I'm guessing that's what I have, so I'm just hardcoding it for
now.
         * -ijt 2:37am april 4 '07 */
        int cmd =3D enable ? PSMOUSE_CMD_SETRATE : PSMOUSE_CMD_SETRES;
        int stickpoint =3D 1; /* FIXME: set this from a test of some kind */
        unsigned char tap_arg =3D stickpoint
                                    ?(enable ? 0x14 : 0x0A)
                                    :(enable ? 0x0A : 0x00);
        int middle_stuff =3D
            stickpoint ? PSMOUSE_CMD_SETSCALE11: PSMOUSE_CMD_DISABLE;
        unsigned char param[4];

        printk("Setting ALPS tap mode to %i\n",enable);

        if (ps2_command(ps2dev, param, PSMOUSE_CMD_GETINFO) ||
            ps2_command(ps2dev, NULL, middle_stuff) ||
            ps2_command(ps2dev, NULL, middle_stuff) ||
            ps2_command(ps2dev, &tap_arg, cmd))
                return -1;

        if (alps_get_status(psmouse, param))
                return -1;

        return 0;
}

I guess now I'm going to see if there's a new version of the mouse driver in
the latest kernel sources.

Tap-to-click is a cruel joke, so please help if you can.

Thanks,
Issac
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lugod.org/pipermail/vox-tech/attachments/20070404/cf695e7=
b/attachment.htm


More information about the vox-tech mailing list