@ sign in C variable declaration

问题:

I found this header file for PIC microcontrollers by the name of pic1250.h and I'm unable to get the hang of some syntax used in it.

The source for the file is:

/*
 *  Header file for the Microchip 
 *  PIC 12c508 chip 
 *  PIC 12c509 chip
 *  Baseline Microcontrollers
 */

static volatile unsigned char   RTCC    @ 0x01;
static volatile unsigned char   TMR0    @ 0x01;
static volatile unsigned char   PCL @ 0x02;
static volatile unsigned char   STATUS  @ 0x03;
static          unsigned char   FSR @ 0x04;
static volatile unsigned char   OSCCAL  @ 0x05;
static volatile unsigned char   GPIO    @ 0x06;

static          unsigned char control   OPTION  @ 0x00;
static volatile unsigned char control   TRIS    @ 0x06;

/*  STATUS bits */
static bit  GPWUF   @ (unsigned)&STATUS*8+7;
static bit  PA0 @ (unsigned)&STATUS*8+5;
static bit  TO  @ (unsigned)&STATUS*8+4;
static bit  PD  @ (unsigned)&STATUS*8+3;
static bit  ZERO    @ (unsigned)&STATUS*8+2;
static bit  DC  @ (unsigned)&STATUS*8+1;
static bit  CARRY   @ (unsigned)&STATUS*8+0;

/*  OPTION bits */
#define     GPWU    (1<<7)
#define     GPPU    (1<<6)
#define     T0CS    (1<<5)
#define     T0SE    (1<<4)
#define     PSA (1<<3)
#define     PS2 (1<<2)
#define     PS1 (1<<1)
#define     PS0 (1<<0)

/*  OSCCAL bits */
static bit  CAL7    @ (unsigned)&OSCCAL*8+7;
static bit  CAL6    @ (unsigned)&OSCCAL*8+6;
static bit  CAL5    @ (unsigned)&OSCCAL*8+5;
static bit  CAL4    @ (unsigned)&OSCCAL*8+4;

/*  GPIO bits   */
static bit  GP5 @ (unsigned)&GPIO*8+5;
static bit  GP4 @ (unsigned)&GPIO*8+4;
static bit  GP3 @ (unsigned)&GPIO*8+3;
static bit  GP2 @ (unsigned)&GPIO*8+2;
static bit  GP1 @ (unsigned)&GPIO*8+1;
static bit  GP0 @ (unsigned)&GPIO*8+0;

#define CONFIG_ADDR 0xFFF
#define FOSC0       0x01
#define FOSC1       0x02
#define WDTE        0x04
#define CP      0x08
#define MCLRE       0x0F

I'm unable to understand the whole modifer-datatype @ declaration-something. Can someone please help me out? I'm just a newbie at this.

回答1:

It's a compiler extension.

From PIC MPLAB XC8 compiler documentation (emphasis mine):

5.5.4 Absolute Variables

Most variables can be located at an absolute address by following its declaration with the construct @ address, where address is the location in memory where the variable is to be positioned. Such a variables is known as an absolute variables.

5.5.4.1 ABSOLUTE VARIABLES IN DATA MEMORY

Absolute variables are primarily intended for equating the address of a C identifier with a special function register, but can be used to place ordinary variables at an absolute address in data memory.

For example:

volatile unsigned char Portvar @ 0x06;

will declare a variable called Portvar located at 06h in the data memory. The compiler will reserve storage for this object (if the address falls into general-purpose RAM) and will equate the variable’s identifier to that address.

Note that MPLAB XC8 is not the only compiler to have the same @ construct to place an object in the specific memory location.

Another well known compiler is Freescale CodeWarrior (at least for HCS08).

Another one is IAR C Compiler (at least for MSP430 and AVR).

回答2:

It's an extension in the PIC compiler, to place a variable at a specific memory position. No other compiler I know have that extension.

回答3:

It's a C language extension supported by the PIC compiler that allows assigning variables to specific RAM addresses.

exemples:

char a @ 0x25;  /* place A at address 0x25 */
bit b @ 0x25.3; /* place B at the third bit of address 0x25 */

There are three uses for this:

• In embedded system you often have very little memory and need to pack several variables in the same byte. This syntax makes it easier, although a standard bit field would work too.
• Another reason is that PIC registers are mapped to RAM, so that you can access them like any other memory location. With this syntax you can define synonyms for the bits you are interested in and use them like normal variables.
• Put you own variable (regardless of size) at a specific location. This is occasionally useful.

Remember that embedded programming is all about total control of the hardware.

回答4:

In addition to what has already been said, please note that the non-standard @ operator is a superfluous feature. You can achieve exactly the same behavior with standard C:

#define RTCC (*(volatile uint8_t*)0x0001u)

Since the variables in this case are hardware registers, you don't need to worry about allocation, they are already present in the hardware. If you want to allocate a variable at a custom address, there should be a linker file of some kind to fix that (since the @ operator only solves specific allocation for variables, not for code).

The main reason why many embedded compilers come up with some non-standard operator like @ is because they can't think outside the box when designing the debugger. They expect some sort of variable to be present in the object file which is fed to the debugger, but if you use #define, no such "debug information object" is allocated.

If the debugger looked at the source code instead, or better yet, had MCU awareness built-in, then non-standard code like this wouldn't be necessary. High quality tools from companies that focus solely on debuggers always come with built-in support for viewing register maps.