You can search your Makefile to see if a module is set to compile an
os.c. For example:
[EMAIL PROTECTED] bld]$ egrep 'os' Makefile
# This makefile is suppose to be configured automatically using the
# same unless your are cross-compiling.)
main.lo opcodes.lo os.lo os_unix.lo os_win.lo \
$(TOP)/src/os.c \
$(TOP)/src/os_unix.c \
$(TOP)/src/os_win.c \
$(TOP)/src/os_unix.c \
$(TOP)/src/os_win.c \
$(TOP)/src/os.h \
$(TOP)/src/os_common.h \
# of the most recently modified source code file
os.lo: $(TOP)/src/os.c $(HDR)
$(LTCOMPILE) -c $(TOP)/src/os.c
os_unix.lo: $(TOP)/src/os_unix.c $(HDR)
$(LTCOMPILE) -c $(TOP)/src/os_unix.c
os_win.lo: $(TOP)/src/os_win.c $(HDR)
$(LTCOMPILE) -c $(TOP)/src/os_win.c
#make Borland C++ and/or Microsoft VC import library for the dll
and you should have the expected binary output:
[EMAIL PROTECTED] bld]$ ls -al os*
-rw-rw-r-- 1 rlc rlc 309 Jan 11 15:27 os.lo
-rw-rw-r-- 1 rlc rlc 13640 Jan 11 15:27 os.o
-rw-rw-r-- 1 rlc rlc 324 Jan 11 15:27 os_unix.lo
-rw-rw-r-- 1 rlc rlc 50360 Jan 11 15:27 os_unix.o
-rw-rw-r-- 1 rlc rlc 321 Jan 11 15:27 os_win.lo
-rw-rw-r-- 1 rlc rlc 4248 Jan 11 15:27 os_win.o
and a careful examination of the linkeditor output should show you it
was linked in, too.
So what I'm getting at is that if your attention focuses on a specific
module, ask yourself if it was compiled and linked to begin with. As you
can see, the checking process is simple.
Bob Cochran
Ken & Deb Allen wrote:
Thank you, kind sir. Sometimes it is difficult to see the individual
ferns on the floor of the massive forest!
Oddly enough, the problem seems to have been caused by the lack of the
os.c file being compiled into the project at all? Some files that were
in the 3.2.8 code stream are no longer present and this new os.c file
was added to the 3.3.0 code tree. I had copied to the files from the
3.3.0 code stream into the project subdirectory, but I had missed the
fact that there was a new file, and the compilation did not report the
problem!
I added this file and adjusted the optimization level, and the problem
disappeared -- the code is working as before!
I have adjusted the code so that the debug mode has optimization
turned off, but that it is still enabled in release mode.
Thanks once again.
-ken
On 13-Jan-06, at 8:40 PM, [EMAIL PROTECTED] wrote:
Try recompiling with all optimization turned off. Optimizers
tend to reorder code which makes single-stepping confusing - both
for you and your debugger.