[U-Boot] [PATCH 04/14] board/cray/L1/flash.c: Fix GCC 4.6 build warning
Marek Vasut
marek.vasut at gmail.com
Thu Dec 1 23:10:50 CET 2011
> Fix:
> flash.c: In function 'flash_erase':
> flash.c:276:24: warning: variable 'l_sect' set but not used
> [-Wunused-but-set-variable]
>
> Signed-off-by: Wolfgang Denk <wd at denx.de>
> ---
> total: 0 errors, 0 warnings, 49 lines checked
> NOTE: Ignored message types: COMPLEX_MACRO CONSIDER_KSTRTO MINMAX
> MULTISTATEMENT_MACRO_USE_DO_WHILE
>
> board/cray/L1/flash.c | 25 +++++++++++--------------
> 1 files changed, 11 insertions(+), 14 deletions(-)
>
> diff --git a/board/cray/L1/flash.c b/board/cray/L1/flash.c
> index a3d893e..77a2100 100644
> --- a/board/cray/L1/flash.c
> +++ b/board/cray/L1/flash.c
> @@ -273,7 +273,7 @@ int flash_erase (flash_info_t *info, int s_first,
int
> s_last) {
> volatile FLASH_WORD_SIZE *addr = (FLASH_WORD_SIZE *)(info->start[0]);
> volatile FLASH_WORD_SIZE *addr2;
> - int flag, prot, sect, l_sect;
> + int flag, prot, sect;
>
> if ((s_first < 0) || (s_first > s_last)) {
> if (info->flash_id == FLASH_UNKNOWN) {
> @@ -303,16 +303,14 @@ int flash_erase (flash_info_t *info, int s_first,
int
> s_last) printf ("\n");
> }
>
> - l_sect = -1;
> -
> /* Disable interrupts which might cause a timeout here */
> flag = disable_interrupts();
>
> /* Start erase on unprotected sectors */
> for (sect = s_first; sect<=s_last; sect++) {
> if (info->protect[sect] == 0) { /* not protected */
> - addr2 = (FLASH_WORD_SIZE *)(info->start[sect]);
> - printf("Erasing sector %p\n", addr2);
> + addr2 = (FLASH_WORD_SIZE *)(info->start[sect]);
> + printf("Erasing sector %p\n", addr2);
>
> addr[ADDR0] = (FLASH_WORD_SIZE)0x00AA00AA;
> addr[ADDR1] = (FLASH_WORD_SIZE)0x00550055;
> @@ -320,15 +318,14 @@ int flash_erase (flash_info_t *info, int s_first,
int
> s_last) addr[ADDR0] = (FLASH_WORD_SIZE)0x00AA00AA;
> addr[ADDR1] = (FLASH_WORD_SIZE)0x00550055;
> addr2[0] = (FLASH_WORD_SIZE)0x00300030; /* sector erase
*/
> - l_sect = sect;
> - /*
> - * Wait for each sector to complete, it's more
> - * reliable. According to AMD Spec, you must
> - * issue all erase commands within a specified
> - * timeout. This has been seen to fail, especially
> - * if printf()s are included (for debug)!!
> - */
> - wait_for_DQ7(info, sect);
> + /*
> + * Wait for each sector to complete, it's more
> + * reliable. According to AMD Spec, you must
> + * issue all erase commands within a specified
> + * timeout. This has been seen to fail, especially
> + * if printf()s are included (for debug)!!
> + */
> + wait_for_DQ7(info, sect);
> }
> }
Logically looks OK
Acked-by: Marek Vasut <marek.vasut at gmail.com>
More information about the U-Boot
mailing list