mtd: r852: make ecc_reg 32-bit in r852_ecc_correct
r852_ecc_correct() reads a 32-bit register into a 16-bit variable,
ecc_reg, but this variable is later used as if it was larger. This is
reported by clang when building the kernel with many warnings:
drivers/mtd/nand/r852.c:512:11: error: shift count >= width of type
[-Werror,-Wshift-count-overflow]
ecc_reg >>= 16;
^ ~~
Fix this by making ecc_reg 32-bit, like the return type of
r852_read_reg_dword().
Signed-off-by: Nicolas Iooss <nicolas.iooss_linux@m4x.org>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
This commit is contained in:
committed by
Brian Norris
parent
a11244c0b2
commit
cef1ed9c6b
@@ -466,7 +466,7 @@ static int r852_ecc_calculate(struct mtd_info *mtd, const uint8_t *dat,
|
|||||||
static int r852_ecc_correct(struct mtd_info *mtd, uint8_t *dat,
|
static int r852_ecc_correct(struct mtd_info *mtd, uint8_t *dat,
|
||||||
uint8_t *read_ecc, uint8_t *calc_ecc)
|
uint8_t *read_ecc, uint8_t *calc_ecc)
|
||||||
{
|
{
|
||||||
uint16_t ecc_reg;
|
uint32_t ecc_reg;
|
||||||
uint8_t ecc_status, err_byte;
|
uint8_t ecc_status, err_byte;
|
||||||
int i, error = 0;
|
int i, error = 0;
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user