[lkml]   [2003]   [Jun]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH] Synaptics TouchPad driver for 2.5.70
Vojtech Pavlik <> writes:

> On Sun, Jun 15, 2003 at 02:18:57PM +0200, Peter Osterlund wrote:
> >
> > - if (hw.w != priv->old_w) {
> > - input_event(dev, EV_MSC, MSC_GESTURE, hw.w);
> > - priv->old_w = hw.w;
> > - }
> > + /*
> > + * This will generate an event even if w is unchanged, but that is
> > + * exactly what we want, because user space drivers may depend on
> > + * this for gesture decoding.
> > + */
> > + input_event(dev, EV_MSC, MSC_GESTURE, hw.w);
> This assumption is not nice. It should instead rely on input_sync() /
> EV_SYN, SYN_REPORT events for complete packet decoding. Can you do
> something about that?

The X driver already relies on EV_SYN to decide when it should act on
the data from the kernel. The problem is that the packet stream is
used as a time base for gesture decoding, because the touchpad was
designed like that to make driver implementation simpler. From the
Synaptics manual:

(Specifically, the TouchPad begins sending packets when Z is 8
or more.) The TouchPad also begins sending packets whenever
any button is pressed or released. Once the TouchPad begins
transmitting, it continues to send packets for one second
after Z falls below 8 and the buttons stop changing. The
TouchPad does this partly to allow host software to use the
packet stream as a time base for gesture decoding, and also to
minimize the impact if the system occasionally drops a packet.

For example, if I press the left button, the X driver can not
immediately generate a left button down event, because maybe I will
press the right button real soon, in which case the middle mouse
button emulation will be activated and generate a middle button down
event. This and similar things are easy to implement by just counting

I guess it would be possible to rewrite the driver so that it doesn't
rely on the packet stream for timing, but it would make the driver
more complicated.

If I could generate only EV_SYN events from the kernel without the
EV_MSC events, that would of course be OK too, but I don't know if
that is possible.

The event parsing code int the X driver currently looks like this:

static Bool
SynapticsParseEventData(LocalDevicePtr local, SynapticsPrivatePtr priv,
struct SynapticsHwState *hw)
struct input_event ev;

while (SynapticsReadEvent(priv, &ev) == Success) {
switch (ev.type) {
case 0x00: /* SYN */
*hw = priv->hwState;
return Success;
case 0x01: /* KEY */
switch (ev.code) {
case 0x110: /* BTN_LEFT */
priv->hwState.left = (ev.value ? TRUE : FALSE);
case 0x111: /* BTN_RIGHT */
priv->hwState.right = (ev.value ? TRUE : FALSE);
case 0x115: /* BTN_FORWARD */
priv->hwState.up = (ev.value ? TRUE : FALSE);
case 0x116: /* BTN_BACK */
priv->hwState.down = (ev.value ? TRUE : FALSE);
case 0x03: /* ABS */
switch (ev.code) {
case 0x00: /* ABS_X */
priv->hwState.x = ev.value;
case 0x01: /* ABS_Y */
priv->hwState.y = ev.value;
case 0x18: /* ABS_PRESSURE */
priv->hwState.z = ev.value;
case 0x04: /* MSC */
switch (ev.code) {
case 0x02: /* MSC_GESTURE */
priv->hwState.w = ev.value;
return !Success;

static Bool
SynapticsReadEvent(SynapticsPrivatePtr priv, struct input_event *ev)
int i, c;
unsigned char *pBuf, u;

for (i = 0; i < sizeof(struct input_event); i++) {
if ((c = XisbRead(priv->buffer)) < 0)
return !Success;
u = (unsigned char)c;
pBuf = (unsigned char *)ev;
pBuf[i] = u;
return Success;

Peter Osterlund -
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:36    [W:0.042 / U:1.412 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site