On Thu, Aug 07, 2008 at 12:21:05PM +0200, svn@coreboot.org wrote:
Author: stepan Date: 2008-08-07 12:21:05 +0200 (Thu, 07 Aug 2008) New Revision: 3474
Added: trunk/payloads/libpayload/drivers/options.c Modified: trunk/payloads/libpayload/drivers/Makefile.inc trunk/payloads/libpayload/i386/coreboot.c trunk/payloads/libpayload/include/coreboot_tables.h trunk/payloads/libpayload/include/sysinfo.h Log: add get_option to libpayload, so coreboot cmos options can be queried.
Signed-off-by: Stefan Reinauer stepan@coresystems.de Acked-by: Jordan Crouse jordan.crouse@amd.com Acked-by: Peter Stuge peter@stuge.se
Sorry.
This breaks my build. I should have tested. :\
cc -m32 -Wall -Werror -fno-stack-protector -nostdinc -I./include -I/usr/lib/gcc/i686-pc-linux-gnu/4.2.2/include -c -o drivers/options.o drivers/options.c cc1: warnings being treated as errors drivers/options.c: In function 'get_option': drivers/options.c:93: warning: pointer targets in passing argument 1 of 'memcmp' differ in signedness make: *** [drivers/options.o] Error 1
From the patch:
if (memcmp(cmos_entry->name, name, len))
..
+#define CB_TAG_OPTION 0x00c9 +#define CMOS_MAX_NAME_LENGTH 32 +struct cb_cmos_entries {
- u32 tag;
- u32 size;
- u32 bit;
- u32 length;
- u32 config;
- u32 config_id;
- u8 name[CMOS_MAX_NAME_LENGTH];
+};
memcmp() expects char *
What should change? I'm thinking memcmp() and friends.
Later revisions breaks more stuff. :\
//Peter