Re: Recent instabilities?

On Sun, Apr 03, 2022 at 13:03:22 +0200, Havard Eidnes wrote:

> >>> As I tried explaining, "not easily".  This is a Mac Mini G4, with
> >>> a USB keyboard attached.  ddb(4) says
> >>> 
> >>>            macppc    Some modes: <Command>-<Option>-<Power>
> >>
> >> Hmm, the docs need to be updated then, this isn't arch specific but
> >> depends on the keyboard driver - with USB keyboards you can use
> >> ctrl-alt-esc, and I *think* I added that to the adbkbd driver as well.
> >> I'll check.
> >
> > Please do.  I tried Control-Option-ESC on my Apple keyboard model
> > A1048 and it does exactly nothing.

I have just successfully used C-A-Esc to drop into dbb on my mini g4
with a Sun Type 6 keyboard (that's what I have on the table, I will
have to bring the Apple keyboard from another room to specifically
test that).  I did that on a hung currentish (9.99.94) mini nad I
could do ps in ddb to see a bunch of tstiled threads.

As macallan said, it doesn't return back from the polled mode properly
(on a rebooted, unstuck system), but re/plugging it made it work
again.



> "Option" because it also has "Alt" printed on it in smaller
> letters.  Not sure that's correct...  How do I find out?
> "wsconsctl map" gives (among others)
> 
> keycode 226 = Cmd2 Alt_L
> keycode 227 = Cmd Meta_L
[...]
> and "wsconsctl map" shows (among others)
> 
> keycode 41 = Cmd_Debugger Escape
> 
> However, there's no telling which modifiers one needs to use to
> trigger "Cmd_Debugger", and it certainly doesn't follow the
> pattern of the others, e.g.:
> 
> keycode 4 = a A
> 
> etc.  The ESC key certainly doesn't do "Cmd_Debugger" when
> pressed un-shifted(!).  So I have problems translating the code
> and admin command output into practice.  "Help!"

No Cmd1?

If you look a bif earlier in wskbd.c you can see that to activate
Cmd_Foo mapping the driver checks that either Cmd is pressed or Cmd1
and Cmd2 are pressed.  So I guess Command-Esc (i.e. Cmd +
Cmd_Debugger) should do the trick for you.


-uwe