BoostC strangeness

[futz]

Hi guys. After a long break from it I'm back tinkering with MCUs.

I'm trying to compile a program with BoostC and getting bogus error messages.

The errors follow that pixel() function when I move it elsewhere in the source file, so it isn't being caused by something above it.

Here's the prototype line from the header file:

void pixel(unsigned char, unsigned char, unsigned char);

I seem to have some vague memory of having this problem back when I first started using BoostC, and solving it. But it's gone now.

This code compiles correctly on another compiler (for another MCU).

 

[AtomSoft]

check before that prototype heh

You should know this futz its hard to help without full code heh

 

[AtomSoft]

Oh try this:
(as the prototype declaration at the top or in header)

void pixel(unsigned char x, unsigned char y, unsigned char color);

 

[futz]

You should know this futz its hard to help without full code heh

Heh I shoulda known that was coming. But you've seen this code - I think I ported it from source I got from you.

 

[AtomSoft]

heh check before your post. try naming the variables in the declaration then get back to me heh also.... WELCOME BACK!

 

[futz]

Oh try this:
(as the prototype declaration at the top or in header)

void pixel(unsigned char x, unsigned char y, unsigned char color);

Variable names in prototypes are always ignored. Unnecessary. I tried it anyway - no change.

 

[AtomSoft]

ah... was worth a try heh

did you check the prototypes before that one ?

 

[futz]

Here's the entire header file:

#include <system.h>
#pragma	CLOCK_FREQ	8000000
#pragma DATA    _CONFIG1H, _INTIO2_OSC_1H
#pragma DATA    _CONFIG2H, _WDT_OFF_2H
#pragma DATA    _CONFIG3H, _MCLRE_ON_3H
#pragma DATA    _CONFIG4L, _LVP_OFF_4L

#define CS      portb.0
#define DATACHAR    portb.1
#define SCLK    portb.2
#define SDATA   portb.3
#define RST     portb.4

void update(void);
void line(int,int,int,int,unsigned char);
void cls(void);
void lcd_init(void);
void lcd_reset(void);
void sendbyte(unsigned char);
void lcd_send(unsigned char, unsigned char);
void pixel(unsigned char, unsigned char, unsigned char);

 

[futz]

And here is the entire source file. I know that 864 byte frame buffer won't work on a 18F1320 - I just wanted to do a test compile before I figure out what to do about that (might end up moving to a bigger PIC):

#include "7110_lcd.h"

int start,count,result,done = 0;
static unsigned int buffer = 0;
char number[] = "                ";
unsigned char framebuff[864];

void main()
{
    unsigned char x,y;
    osccon = 0x72;          //set int osc to 8MHz
    trisb = 0;
    adcon1 = 0b01111111;    //all digital
    while(1){
        for(x=10;x<36;x++){
            for(y=10;y<36;y++){
                pixel(x,y,1);
            }    
        }
        delay_s(1);
        cls();
    }    
}

void update(void){      //display frame buffer
    unsigned char i,x,outbyte,page;
    int count = 0;
    page = 0xb0;            //page address variable
    for(x=0;x<8;x++){
        lcd_send(page,0);       //set page address
        lcd_send(0x11,0);       //set column address
        lcd_send(0x02,0);
        for(i=0;i<0x60;i++){    //display a row
            outbyte = framebuff[count++];
            lcd_send(outbyte,1);
        }
        page++;
    }
}           

void lcd_init(void){
    DATACHAR = 1;
    delay_ms(2);
    lcd_reset();
    lcd_send(0xa6,0);   //Display: Normal   
    lcd_send(0xA3,0);   //LCD Bias Settings: 1/7
    lcd_send(0xA1,0);   //ADC Selection: Reverse
    lcd_send(0xC0,0); //Common Output: Normal Direction
//    lcd_send(0xC8,0); //Common Output: Upside Down
    lcd_send(0x22,0);   //Set the V5 output Voltage
    lcd_send(0x81,0);   //set Electronic Volume - brightness
    lcd_send(0x2f,0);   
    lcd_send(0x2E,0);   //Power Controller Set: Booster circuit: ON/Voltage regulator circuit: ON/Voltage follower circuit: OFF
    lcd_send(0x2F,0);   //Power Controller Set: Voltage follower circuit: ON
//    lcd_send(0xE3,0);   //Non-OPeration Command
//    lcd_send(0x40,0);   //Set the start line
    lcd_send(0xAF,0);   //LCD On
    lcd_send(0xA4,0);   //Display All Points: NORMAL
    cls();
}

void cls(void){
    unsigned char i,x,line;
    int j;
    line = 0xb0;                //page address variable
    for(x=0;x<9;x++){
        lcd_send(line,0);       //set page address
        lcd_send(0x11,0);       //set column address
        lcd_send(0x02,0);
        for(i=0;i<0x60;i++){    //write zeros to display RAM
            lcd_send(0x00,1);
        }
        line++;
    }
    for(j=0;j<865;j++)
        framebuff[j] = 0;
}

void lcd_send(unsigned char cmd, unsigned char type){
    if(type)
        DATACHAR = 1;
    else
        DATACHAR = 0;
    CS = 0;
    sendbyte(cmd);
    CS = 1;
}    

void sendbyte(unsigned char data){
    char x;
    for(x=0;x<9;x++){
        SCLK = 0;
        SDATA = 0;
        if(data & 0x80)
            SDATA = 1;
        SCLK = 1;
        data <<= 1;
    }
}
        
void lcd_reset(void){
    CS = 1;
    delay_ms(2);
    RST = 1;
    delay_ms(2);
}

void pixel(unsigned char x, unsigned char y, unsigned char color)
{
    unsigned char temp,readtemp;
    char bit,div;
    int offset;
    div = y >> 3;
    offset = div * 96 + x;
    bit = (y - (div << 3));
    temp = 1 << bit;
    readtemp = framebuff[offset];
    if(!color)
        y = readtemp & ~temp;
    else
        y = readtemp | temp;
    framebuff[offset] = y;
}

 

[AtomSoft]

Fixed.. you cant use the name "bit"

try this

void pixel(unsigned char x, unsigned char y, unsigned char color)
{
    unsigned char temp,readtemp;
    char [B]mbit[/B], div;
    int offset;
    div = y >> 3;
    offset = div * 96 + x;
    [B]mbit [/B]= (y - (div << 3));
    temp = 1 << [B]mbit[/B];
    readtemp = framebuff[offset];
    if(!color)
        y = readtemp & ~temp;
    else
        y = readtemp | temp;
    framebuff[offset] = y;
}

I guess in BoostC "bit" is a type

You see posting full code helped a great deal

 

[futz]

D'oh! I should have spotted that highlighted bit there! That was it. Thank you.

What a strangely misleading error message.

 

[AtomSoft]

yeah.. heh . no problem im glad i could finally help you you always help me so... have fun and stay in the mcu department heh