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 CCS Technical Support
Bugs with Float?

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



Joined: 30 Jan 2005
Posts: 23
Location: Argentina

View user's profile Send private message Send e-mail Yahoo Messenger MSN Messenger

Bugs with Float?
PostPosted: Thu Feb 03, 2005 1:56 pm     Reply with quote
Hello

I have a piece of code, which involves making very complex float maths. I have to calculate the distance from two near points which are given by a GPS. I have the formula to calculate this, but I have found a extrange behaviour when adding two floats values. This is the Piece of Code, and I will tell what the printf shows

#include <16F877.h>
#include <math.h>
#use delay(clock=16000000,RESTART_WDT)
#fuses NOWDT,HS, NOPUT, NOPROTECT, NOBROWNOUT, NOLVP, NOCPD, NOWRT, NODEBUG
#use rs232(baud=57600,parity=N,xmit=PIN_D2,rcv=PIN_D3,bits=8,stream=PC,errors)

float dist1;
float temp1;

void tracking (float lat1,lon1,lat,lon) {

dist1 = lat1;
dist1 = dist1 * 12.44;
dist1 = dist1 + 110574;
temp1 = lat1;
temp1 = temp1 - lat;
}

void main () {

float a1,a2,x1,x2;
a1 = -34.61084;
a2 = -58.43049;
x1 = -34.60900;
x2 = -58.42851;

tracking (a1,a2,x1,2);
fprintf(PC,"dist1: %4.4f temp1: %4.4f",dist1,temp1);
while(1);
}


The printf shows dist1: 110143.4387 temp1: -166.6108

Shouldn't temp1 be: -0.00184?

Thanks
Kimi
C-H Wu
Guest
float bug in function definition ?
PostPosted: Thu Feb 03, 2005 8:41 pm     Reply with quote
Kimi:

I use PCWH 3.187, 3.212, 3.216, 3.218 to compile your code, bug confirmed for both F877 with PCM and 18F4620 with PCH!

void tracking (float lat1,lon1,lat,lon)

Code:
038-03B tracking.lat1  <-- float, 4 bytes
03C     tracking.lon1 <-- become int8, only 1 byte ! bug ?
03D     tracking.lat <-- become int8 ! bug ?
03E     tracking.lon <-- become int8 ! bug ?


workaround :

void tracking (float lat1,float lon1,float lat,float lon) {

Code:

03C-03F tracking.lat1
040-043 tracking.lon1 <-- no more bug
044-047 tracking.lat
048-04B tracking.lon


However, there is no such bug with the following syntax
Code:

void main () {

float a1,a2,x1,x2;


I never notice this CCS feature Twisted Evil , thanks a lot for your information.

You should file a bug report to CCS !

Best wishes

C-H Wu
PCM programmer



Joined: 06 Sep 2003
Posts: 21708

View user's profile Send private message

PostPosted: Thu Feb 03, 2005 9:24 pm     Reply with quote
Quote:

void tracking (float lat1,lon1,lat,lon)

038-03B tracking.lat1 <-- float, 4 bytes
03C tracking.lon1 <-- become int8, only 1 byte ! bug ?
03D tracking.lat <-- become int8 ! bug ?
03E tracking.lon <-- become int8 ! bug ?

It's a bug, but in the sense that the compiler should have given
an error, instead of defaulting to int8.

If I look in the help file for MSVC++ 6.0
and look under "Function Parameters", it says this:

Code:
Each identifier in parameter-type-list must be preceded by its appropriate type specifier, as shown in this example:

void new( double x, double y, double z )
{
    /* Function body here */
}


CCS is based on K&R, and maybe somewhere in K&R it says the
same as above, but I'm not going to spend any more time looking for it.
Mark



Joined: 07 Sep 2003
Posts: 2838
Location: Atlanta, GA

View user's profile Send private message Send e-mail

PostPosted: Thu Feb 03, 2005 10:10 pm     Reply with quote
Quote:

In the ANSI standard:
Definitions must be coded in the same order in which arguments are passed.
Only one identifier for each type specifier can be coded in an ANSI
parameter definition.
The comma ( , ) now separates the definitions between the parentheses.
There is no semicolon to terminate a definition.


The default data type for CCS is int8 thus what you see.
PCM programmer



Joined: 06 Sep 2003
Posts: 21708

View user's profile Send private message

PostPosted: Thu Feb 03, 2005 10:49 pm     Reply with quote
The key part is here in these two sentences:
Quote:

Only one identifier for each type specifier can be coded in an ANSI
parameter definition.
The comma ( , ) now separates the definitions between the parentheses.

That means you can't do this:
Code:
void func(float a, b, c, d);

You must do it this way:
Code:
void func(float a, float b, float c, float d);

Because CCS doesn't give an error for the first declaration above,
it's in violation of the ANSI standard. But CCS doesn't say that
they are in accord with ANSI.

If I try to find some compiler on my system that might be K&R
compatible, I have Quick C for Windows vs. 1.0, copyright 1991.
If I run the following code,
Code:
void my_func(float a, b, c, d)
{

}

then I get this error message:
C:\QCWIN\GENERAL\TEST.C(6) : error C2061: syntax error : identifier 'b'
TEST.EXE - 1 error(s), 0 warning(s)

I have to do this to avoid an error:
Code:
void my_func(float a, float b, float c, float d)
{

}

So CCS is in violation of ANSI and probably K&R as well.
They shouldn't default to int8 for parameters that are missing a
type-definition. They should give an error.
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