Re: [RFT/PATCH] i8042: remove polling timer (v6)

[Dave Neuer]

On 11/3/06, Dmitry Torokhov <dtor@xxxxxxxxxxxxx> wrote:
> On Sunday 29 October 2006 18:20, Dave Neuer wrote:
> >  Or is it also to
> > make the i8042 driver less racy?
>
> I think we agree that i8042_aux_write() is not racy, do you see any other
> races in i8042?

No, the polling seemed like the big race.

> > I ask because I've applied this over
> > (a modified) 2.6.18 on my Compaq Presario X1010us laptop which has
> > been driving me crazy w/ Synaptics problems and keyboard problems
> > (intermittent, but   frequent enough lately that I finally figured I
> > needed to do something about it).
> >
>
> Have you tried limiting Synaptics rate to 40 packets per second (using
> psmouse.rate=40 option)? Some KBD can't handle full Synaptics rate of
> 80 pps; it usually manifests in keyboard troubles.

No, haven't tried that (first I've heard of it, thanks!)

As I said, I have had both keyboard and touchpad problems on this
laptop (formerly more of the former, lately more of the latter);
interestingly, after applying this patch I have had failures much less
frequently (multiple instances of several days w/out failure w/ the
laptop powered on continuously). Also interestingly, and
unfortunately, that may be coincidence; I've discovered a use case
that seems to reliably cause the touchpad to freeze up w/ or w/out the
patch applied (selecting multiple items in modified file list in
EasyTAG). Haven't looked into it yet, on the "hurts when I do this,"
"then don't do that" theory, but when I have time to look at the
EasyTAG code and try to reason about what's happening, I will.

> >
> > I don't really know if or how much the races in this driver are
> > contributing to my problems (keyboard getting stuck repeating last
> > key, or ignoring interrupts, or synaptics touchpad freezing, last of
> > which requires cold boot to fix).
>
> You mean even reloading psmouse module can't revive the touchpad?

Correct. Even rebooting doesn't help. Halt. Start.

> > Maybe more likely an ACPI thing?
>
> Coudl be.

Hmmmm.

Dave
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/