Re: 2.4.23-pre9 ide+XFree+ptrace=Complete hang

From: Maciej Zenczykowski
Date: Wed Nov 12 2003 - 12:18:55 EST


> Maciej Zenczykowski wrote:
>
> > ....
> > sure now that it wouldn't help - the system is just plain totally
> > and completely dead, even the system bios is likely dead - including the
> > System Management Mode code (likely responsible for lighting up the LED
> > on fn key press/release, which no longer works [although not on every
> > crash]).
>
> It is usually the kernel that togles the leds, except when an application (like
> X) requests that the keyboard be put into RAW mode. In RAW mode it is the
> application that is responsible to for updating the leds. If X hangs, Caps-Lock
> and the like will not make the leds toggle anymore.
>
> You can try to issue a SysRq+R to take the keyboard out of RAW mode into XLATE
> so that the kernel translates the keys and you might be able to toggle leds,
> switch consoles, etc.

As I've already written no SysRQ comboes do anything.

I'm speaking about the FN-key led on a laptop which is most definitely not
kernel toggled - there are 3 keyboard LEDS - Caps, Cursor Keys, Numeric
Keypad - the first is toggled by caps by the kernel, the second two have
unusual semantics - the kernel is half responsible and the bios is half
responsible. [There is no scroll lock LED] So the single kernel viewed
numeric lock LED is visible as two LEDS on the laptop - the kernel can
toggle a single bit (numlock: off/on) and the SMM Bios sets up how this is
interpreted. Normally pressing the FN-key causes the SMM Bios to change
this hardware intrepretation of the keyboard LED wiring causing a LED to
light up. This probably wasn't very clear - but it's definetely not
kernel [it works during bios performed suspend to disk, etc.] :)

Cheers,
MaZe.

-
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/