What’s the Hidden Trick Behind This Unusual C Macro in Kernel Code?

While browsing kernel source late at night, the author encountered a macro definition that differs from the usual pattern and decided to share it for readers who might not have seen it before.

Typical macro usage in C/embedded code includes a parameter list, for example:

#include "stdio.h"
#define UART_RBR(base)         (unsigned long)(base+0x01) /* Read only */

unsigned int mtk_uart_read_byte()
{
    unsigned long base = 0x1231231;
    return UART_RBR(base);
}

int main()
{
    printf("%lx
", mtk_uart_read_byte());
    getchar();
    return 0;
}

The author then shows the same code where the macro is defined without the parameter list:

#include "stdio.h"
#define UART_RBR         (unsigned long)(base+0x01) /* Read only */

unsigned int mtk_uart_read_byte()
{
    unsigned long base = 0x1231231;
    return UART_RBR;
}

int main()
{
    printf("%lx
", mtk_uart_read_byte());
    getchar();
    return 0;
}

Even though the macro lacks an argument list, it still expands correctly because the identifier UART_RBR is replaced by the expression that uses the variable base defined in the surrounding function scope. This demonstrates a useful macro trick for low‑level C and kernel development.