Subject:
|
Floating point emulation
|
Newsgroups:
|
lugnet.robotics.rcx.legos
|
Date:
|
Wed, 5 May 1999 20:15:09 GMT
|
Viewed:
|
1514 times
|
| |
|
|
I succeeded in using (emulated) floats in a legOS program. Here are some
instructions in case some of you want/need to try 32 bit FP emulation.
First of all I must point out that I have no merit at all in having got this
result, a part from having stressed a lot the right persons to get help :-)
Thanks to John A. Tamplin who gave me a lot of info and to Marco Beri who
solved the last problems and tested the solution on his legOS installation.
I compile legOS on a Windows PC, so I have cygwin installed. I tried to make
these instruction easy to follow for people not particularly at ease with gnu
tools (I am one of them).
Instructions
------------
1. FP emulation is in the "fp-bit.c" file that compiles into "libgcc.a". Find
the proper "libgcc.a" in your cross compiler egcs path. Mine is in:
..\cygnus\cygwin-b20\H-i586-cygwin32\lib\gcc-lib\h8300-hms\egcs-2.91.57
an its size is 339458.
2. copy it in your legOS\lib directory (in the same place where you find
libc++.a, libc.a, libmint.a).
3. In your Makefile modify the LIBS= line appending "-lgcc" to its end:
LIBS= -lc -lmint -lc++ -lgcc
4. In your Makefile modify the KSOURCES= line adding "memcpy.c":
KSOURCES=kmain.c mm.c systime.c tm.c semaphore.c conio.c lcd.c \
memcpy.c direct-ir.c direct-motor.c direct-sensor.c direct-sound.c
5. Find "memcpy.c" in your legOS\lib\c directory and copy it to your main legOS
directory (the same where you find "direct-motor.c", "direct-ir.c" etc.)
6. Now you can declare and use 32 bit float vars in your source.
Good luck and have fun
Mario
PS: I would be surprised if this solution couldn't be refined by some of you
legOS very experts.
|
|
1 Message in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
