CCS C Software and Maintenance Offers
FAQFAQ   FAQForum Help   FAQOfficial CCS Support   SearchSearch  RegisterRegister 

ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

CCS does not monitor this forum on a regular basis.

Please do not post bug reports on this forum. Send them to support@ccsinfo.com

float from two int16's

 
Post new topic   Reply to topic    CCS Forum Index -> General CCS C Discussion
View previous topic :: View next topic  
Author Message
Will Reeve



Joined: 30 Oct 2003
Posts: 209
Location: Norfolk, England

View user's profile Send private message Send e-mail Visit poster's website MSN Messenger

float from two int16's
PostPosted: Sun Jan 30, 2005 10:03 am     Reply with quote

I've got to fix something blind without the hardware (always a pain!). Anyway it comes down to needing a float instead of a int32.

rough code:
int16 a;
int16 b;
float c;

c = a / b;

Code:
....................       R = pforce / sforce;
1DB4:  MOVFF  3D,41B
1DB8:  MOVFF  3C,41A
1DBC:  MOVFF  3F,41D
1DC0:  MOVFF  3E,41C
1DC4:  BRA    192E
1DC6:  MOVFF  02,41B
1DCA:  MOVFF  01,41A
1DCE:  MOVLB  0
1DD0:  BRA    1978
1DD2:  MOVFF  03,419
1DD6:  MOVFF  02,418
1DDA:  MOVFF  01,417
1DDE:  MOVFF  00,416

will this work OK or should i use

c = (float)a / (float)b; for a belts and braces approach? code:

Code:
....................       R = (float)pforce / (float)sforce;
1ECC:  MOVFF  3D,41F
1ED0:  MOVFF  3C,41E
1ED4:  RCALL  192E
1ED6:  MOVFF  00,41A
1EDA:  MOVFF  01,41B
1EDE:  MOVFF  02,41C
1EE2:  MOVFF  03,41D
1EE6:  MOVFF  3F,41F

1EEA:  MOVFF  3E,41E
1EEE:  RCALL  192E
1EF0:  MOVFF  41D,421
1EF4:  MOVFF  41C,420
1EF8:  MOVFF  41B,41F
1EFC:  MOVFF  41A,41E
1F00:  MOVFF  03,425
1F04:  MOVFF  02,424
1F08:  MOVFF  01,423
1F0C:  MOVFF  00,422
1F10:  BRA    1966
1F12:  MOVFF  03,419
1F16:  MOVFF  02,418
1F1A:  MOVFF  01,417
1F1E:  MOVFF  00,416


I would hope both of these would give me exactly the same result, is the compiler doing some good optimisation for me which I am overriding with the cast?

Keep well all,

Will
Charlie U



Joined: 09 Sep 2003
Posts: 183
Location: Somewhere under water in the Great Lakes

View user's profile Send private message

PostPosted: Sun Jan 30, 2005 10:57 am     Reply with quote

Will,

You could try your tests with a quick and dirty program on the simulator in MPLAB to confirm your results.

If you can't do that, I would definitely cast the variables before the divide. Look at one example: if a < b then the divide using ints should yield 0, while the divide with floats should yield a number between 1 and 0 in floating point notation.
Will Reeve



Joined: 30 Oct 2003
Posts: 209
Location: Norfolk, England

View user's profile Send private message Send e-mail Visit poster's website MSN Messenger

PostPosted: Sun Jan 30, 2005 11:31 am     Reply with quote

I wondered if some automatic casting would happen? I know CCS has some strange rules with casts. I haven't used a float for several years!
Charlie U



Joined: 09 Sep 2003
Posts: 183
Location: Somewhere under water in the Great Lakes

View user's profile Send private message

PostPosted: Sun Jan 30, 2005 3:45 pm     Reply with quote

Don't get your hopes up. I have learned the hard way that I should always make an explicit cast when I want thing to work out correctly. Don't rely on automatic/assumed casting.

In the case that you have mentioned, I just ran it on the simulator, and without the cast, the division is performed as int16 first, then the result is cast to a float. So, 2/4 = 0.00. If you cast both a and b to float, then divide, the result is 0.500. I tried this with a few values and the result is the same, int16 division then cast.

So, in your case, since you need to make the change "blind" and need the highest probability of success, cast first, then divide.
Will Reeve



Joined: 30 Oct 2003
Posts: 209
Location: Norfolk, England

View user's profile Send private message Send e-mail Visit poster's website MSN Messenger

PostPosted: Mon Jan 31, 2005 6:25 am     Reply with quote

Thanks. I really should install MPLAB again. When I switched to the CCS IDE I never went back. MPLAB never got installed on my new PC.
SherpaDoug



Joined: 07 Sep 2003
Posts: 1640
Location: Cape Cod Mass USA

View user's profile Send private message

Re: float from two int16's
PostPosted: Mon Jan 31, 2005 8:43 am     Reply with quote

Will Reeve wrote:

int16 a;
int16 b;
float c;

c = a / b;


As I read K&R this should do a 16 bit integer divide producing an integer result which is then cast to a float. I assume that is NOT what you want.
If you want the compiler to do something else you must tell it what you want!
_________________
The search for better is endless. Instead simply find very good and get the job done.
Display posts from previous:   
Post new topic   Reply to topic    CCS Forum Index -> General CCS C Discussion All times are GMT - 6 Hours
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group