Hi,
problem was that segments were given different names between kernel 2027
and 2028 and the 32bit mul/shl/shr/div routines ended up being called in
nirvanaland after they were relocated.
This problem can be solved as follows:
--- lib/makefile Tue Mar 4 00:07:29 2003
+++ lib/makefile Tue Mar 4 00:07:56 2003
@@ -11,6 +11,7 @@
libm.lib: $(CLIB)
-$(RM) libm.lib
$(LIBUTIL) $(CLIB) $(MATH_EXTRACT) $(LIBTERM)
+ for %i in (*.obj) do ..\utils\patchobj CODE=LCODE TEXT=_LOWTEXT %i
$(LIBUTIL) libm $(MATH_INSERT) $(LIBTERM)
-$(RM) *.OBJ
and the compiled kernel runs after all (but of course it's less efficient
than the one compiled using Watcom C).
Bart
==^================================================================
This email was sent to: [EMAIL PROTECTED]
EASY UNSUBSCRIBE click here: http://topica.com/u/?b1ddyi.b3hwCs.YXJjaGl2
Or send an email to: [EMAIL PROTECTED]
TOPICA - Start your own email discussion group. FREE!
http://www.topica.com/partner/tag02/create/index2.html
==^================================================================
