[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: some more 110h4 clues...



>  thanx!  i'm sorry for all the trouble, i still have no '030 to test MMU
> stuff myself... :/  i hope i this is better.

 It is *better*, but still not *perfect*. :-(

 All cases of directly susequent p_fork/p_vfork and p_exec seem to work now,
every shell can start every program, either in background or in foreground
without problems, but getty still crashes with something like:

type = free  AA=11BE024  BP=11BE000

 when overlaying itself with login. Looks like the process doesn't own its
own basepage. To make it even more complicate: This all only happens if
there's actually more than one getty running, say, the cain:

PID 001 (init): p_vfork()
PID 002 (init): p_exec(200, "getty")
PID 002 (getty): p_exec(200, "login")

 works perfectly well, but as soon as you start several getties, so that it
looks like:

PID 001 (init): p_vfork()
PID 002 (init): p_exec(200, "getty")
PID 001 (init): p_vfork()
PID 003 (init): p_exec(200, "getty")
PID 002 (getty): p_exec(200, "login")

 you'll run into problems. Note that it's the first getty started (PID 2) that
crashes! If you log onto another tty so that the *last* started getty (in this
case PID 3) will overlay itself, everything is back ok again...

ciao,
TeSche

PS: Is it ok that you used a different adress for this message? I replied
to both the new and the known one, to be sure...
-- 
Torsten Scherer (TeSche, Schiller...), itschere@techfak.uni-bielefeld.de
Faculty of Technology, University of Bielefeld, Germany, Europe, Earth...
| Use any of "finger itschere@129.70.131.2-15" for adresses and more.|
| Last updated: 14. April 1994.|