Re: EeePC 900 trackpad often not detected at boot in 2.6.30-rc4

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Thu, May 28, 2009 at 05:42:31PM +0100, Sitsofe Wheeler wrote:
> On Thu, May 28, 2009 at 08:54:01AM -0700, Dmitry Torokhov wrote:
> > On Tue, May 26, 2009 at 11:20:49PM +0100, Sitsofe Wheeler wrote:
> > > I only just saw your reply. After about 50 boots the following occurred:
> > > 
> > > [    3.136500] drivers/input/serio/i8042.c: d4 -> i8042 (command) [1964]
> > > [    3.137015] drivers/input/serio/i8042.c: f2 -> i8042 (parameter) [1964]
> > > [    3.140289] ALSA device list:
> > > [    3.140367]   #0: HDA Intel at 0xf7eb8000 irq 16
> > > [    3.140591] TCP cubic registered
> > > [    3.140666] NET: Registered protocol family 17
> > > [    3.140846] Using IPI Shortcut mode
> > > [    3.142878]   Magic number: 9:412:804
> > > [    3.143418] rtc_cmos 00:03: setting system clock to 2009-05-26 21:47:18 UTC (1243374438)
> > > [    3.149589] drivers/input/serio/i8042.c: fa <- i8042 (interrupt, 1, 12) [1977]
> > > [    3.151638] drivers/input/serio/i8042.c: 00 <- i8042 (interrupt, 1, 12) [1979]
> > > [    5.550332] libps2: BINGO!
> > > [    5.550425] drivers/input/serio/i8042.c: d4 -> i8042 (command) [4378]
> > > [    5.551011] drivers/input/serio/i8042.c: f6 -> i8042 (parameter) [4378]
> > > [    5.563688] drivers/input/serio/i8042.c: fa <- i8042 (interrupt, 1, 12) [4391]
> > > [    5.563736] drivers/input/serio/i8042.c: d4 -> i8042 (command) [4391]
> > > [    5.564673] drivers/input/serio/i8042.c: f3 -> i8042 (parameter) [4391]
> > > [    5.572824] drivers/input/serio/i8042.c: fa <- i8042 (interrupt, 1, 12) [4400]
> > > [    5.572863] drivers/input/serio/i8042.c: d4 -> i8042 (command) [4400]
> > > [    5.573811] drivers/input/serio/i8042.c: 0a -> i8042 (parameter) [4400]
> > > [    5.580823] ACPI: Battery Slot [BAT0] (battery present)
> > 
> > Was the mouse in working condition after that? I can't tell from the
> > dmesg since it is trimmed...
> 
> Ah sorry. The mouse/trackpad interrupts (in /proc/interrupts) definitely
> rose (when it goes wrong they usually remain static on some value like
> 15).
> 

OK, great. I also see the mouse is being detected further down in dmesg
now.

> 
> In neither case did I go to X because having the debugging leads to huge
> logs in /var

You can shut off the debugging from userspace by doing:

	echo 0 > /sys/module/i8042/parameters/debug

> but I was rebooting until the following script started
> outputting BINGO!:
> 
> #!/bin/sh
> mount /proc
> grep 12: /proc/interrupts
> dmesg | grep -i bingo
> 
> On a normal boot there will be exactly 115 interrupts if the trackpad
> hasn't been touched but everything is OK just after booting. If there
> are less interrupts you know something has gone wrong...
> 

OK, I guess I will be pushin this to Linus then... Thank you for
testing.

-- 
Dmitry