3.4 __farflash Keyword

The IAR __farflash keyword places objects in program memory, and uses 24-bit addresses to read such objects either directly or indirectly via any pointer qualified with this same keyword. Arithmetic on 24-bit wide pointers is only performed on the lower 16 bits, except for comparisons, which are always performed on the entire address.

Suggested Replacement

This keyword can be removed when using a memory-placement feature of MPLAB XC8, or alternatively, there are several MPLAB XC8 qualifiers that performs a similar tasks to this keyword, but there are some differences in their effect.

When enabled (the default state), the MPLAB XC8's const-data-in-program-memory feature places const objects into program memory. In this case, the IAR keyword is redundant and should be removed.

When this feature is disabled, using the -mno-const-data-in-progmem option, use the const qualifier provided by MPLAB XC8 and either the __memx qualifier or one of the__flashn qualifiers, where n is the 64 KB flash segment in which to place and access the object. When using the __memx qualifier, both access and address arithmetic is on a full 24 bit address. With any of the flash qualifiers, pointers and address arithmetic is 16-bit, and the compiler sets RAMPZ to the appropriate 64 KB segment number on each access.

Caveats

The const qualifier must be used with either __memx or any of the flash qualifiers when using MPLAB XC8, whereas this is only preferable when using __farflash with IAR.

If the const-in-program-memory feature is disabled, alternate versions of the string functions normally provided by <string.h> must be used to access string objects located in program memory.

Examples

Consider migrating IAR code such as:

__farflash int x = 200; /* Place and access x somewhere in flash */
volatile int y;
int main() {
    y = x;
}

to MPLAB XC8 code similar to:

const volatile __flash1 int x = 200; /* Place and access x in the 64K-128K segment of flash */
volatile int y;
int main(void) {
    y = x;
}

Further Information

See the Special Type Qualifiers and Options Specific to AVR Devices sections in the MPLAB XC8 C Compiler User's Guide for AVR MCUs for more information on these qualifiers and option.