Carl-Daniel Hailfinger wrote:
On 02.08.2008 16:50, Stefan Reinauer wrote:
Index: src/arch/i386/lib/console_print.c
--- src/arch/i386/lib/console_print.c (revision 3461) +++ src/arch/i386/lib/console_print.c (working copy) @@ -58,58 +58,59 @@ } }
-#define NOINLINE __attribute__((noinline)) -static void print_emerg_char(unsigned char byte) { __console_tx_char(BIOS_EMERG, byte); } -static void print_emerg_hex8(unsigned char value){ __console_tx_hex8(BIOS_EMERG, value); } -static void print_emerg_hex16(unsigned short value){ __console_tx_hex16(BIOS_EMERG, value); } -static void print_emerg_hex32(unsigned int value) { __console_tx_hex32(BIOS_EMERG, value); } -static void print_emerg(const char *str) { __console_tx_string(BIOS_EMERG, str); } +#define STATIC
I don't understand that change. You unconditionally define STATIC as empty token instead of simply removing it.
Yes, absolutely. Someone put it there and was thinking something.
So this allows to put a method in place choosing when it would be required to really set those function prototypes static.
Stefan