[comp.sources.games.bugs] nethack 2.3 on SCO Xenix

jl42+@andrew.cmu.edu (Jay Mathew Libove) (04/22/88)

I patched my 2.2 nethack sources to 2.3 with the recently posted kit
and built the system, entirely without complaints from the machine (a
PCs Limited AT running SCO Xenix SysV/286 release 2.2.1, with
development system 2.1.4d).

However, when I play the game, I eventually end up hanging in the
termcap.c module, delay_output() routine. I played with it by
#ifdef'ing M_XENIX (automatically defined by my compiler) and if so
using "nap(50)" rather than the rest of the delay_output() routine.
Since the hanging continued, I put a "fprintf(stderr,"something")"
before and after the nap() call, and when the game hangs, I see the
before note, and nothing else. I checked the running time of the game
and found that as of the hang it uses no CPU, and it does still
respond to the kill signal.

Has anyone else hit this, and maybe can shed some light on the
situation?

---------------------------------------------------

I just ran nethack under adb on my SCO Xenix system (this is a
continuation of my most recent post, which describes my problem with
nethack 2.3 hanging on my SCO Xenix system) and came up with:

nethack 2.3 on my SCO Xenix system (SysV/286 v2.2.1, development
system 2.1.4d) hangs in the termcap.c module, delay_output() call (in
which I have #ifdef'd in M_XENIX, automatically defined by my
compiler, to replace the whole routine with "nap(50)") during throws,
e.g. throwing shurikens at monsters. It does not always hang when I
throw, but when it does, it hangs with the function call backtrace:

_delay_output() from _tmp_at+81.
_tmp_at(26., 16.) from _bhit+470.
_bhit(1., 0., 6., 41., 0., 0., 0., 0., 1994., 127.) from _throwit+419.
_throwit(1994., 127.) from _dothrow+305.
????????() from _rhack+603.
_rhack(29752., 103.) from _main+3258.
_main(2., -30814., 103., -30802., 103.) from __start+94.
__start() from start0+7.

[ The above is from one particular crack up; it may be that that is
not exactly what happens every time, but the functions are the same
].

Again, if anyone can help me get this running on SCO, I'd be really
happy. Thanks!

Jay Libove
Arpa:   Jay.Libove@andrew.cmu.edu	Bitnet: Jay.Libove@drycas.bitnet
UUCP:   ...!{uunet, ucbvax, harvard}!andrew.cmu.edu!Jay.Libove
UUCP:   ...!{pitt | bellcore} !darth!libove!libove