Messages in this thread | | | From | Dmitry Torokhov <> | Subject | Re: serio / atkbd / PS/2 Strangeness with Serial port adapter | Date | Mon, 24 Aug 2009 17:22:16 -0700 |
| |
On Monday 24 August 2009 17:09:20 Matt Causey wrote: > On Mon, Aug 10, 2009 at 1:34 AM, Dmitry > Torokhov<dmitry.torokhov@gmail.com> wrote: > > >> [ 459.004227] drivers/input/serio/i8042.c: fe <- i8042 (interrupt, 1, > >> 12, timeout) [445118] > >> [20171.684870] drivers/input/serio/i8042.c: 0b <- i8042 (interrupt, 0, > >> 1) [20162195] > >> > >> However - the atkbd driver does not pick the device up at all. > >> > > > > Your device is steadely refusing to say that it is a keyboard - and so > > atkbd driver does not bind to it. > > > >> So, I did find a workaround (which is NOT acceptable for this > >> deployment). Remember the PS/2 'wedge' I mentioned? Well, it has an > >> extra PS/2 port for your keyboard. If I connect a PS/2 keyboard to > >> that thing, suddenly my scanner starts working. > >> > >> We see some chatter from the controller driver, then one of these: > >> > >> [23811.805578] input: AT Translated Set 2 keyboard as > >> /devices/platform/i8042/serio0/input/input6 > >> > > > > That happened because apparently the 'wedge' passes the inquirys to > > the real keyboard and it answers propery. > > > > I think the "i8042.dumbkbd" kernel parameter will solve your issue. > > > > -- > > Dmitry > > > > That's great. I think it's the only option that I did not try. ;-) > Works great! > > In case it matters: > > blah blah/ # dmesg | grep i8042 > [ 0.000000] Kernel command line: initrd=ramdisk.gz > video="vesafb:mtrr,ywrap,1024x768-32@85" i8042.dumbkbd > BOOT_IMAGE=bzImage auto > [ 14.046511] serio: i8042 KBD port at 0x60,0x64 irq 1 > [ 14.046578] serio: i8042 AUX port at 0x60,0x64 irq 12 > [ 15.804210] input: AT Translated Set 2 keyboard as > /devices/platform/i8042/serio0/input/input2 > blah blah/ # > > Looks like it just decides that there must be a keyboard there, > whether or not there really is one. > > Are there any downsides to using this kernel parameter? >
The kernel will not control LEDs on the keyboard nor will it be able to control hardware autorepeat parameters (software autorepeat should still work fine).
-- Dmitry
| |