> Am 06.04.2021 um 13:24 schrieb Huacai Chen <chenhuacai@xxxxxxxxxx>:
>
> Only 32bit kernel need __div64_32(), but commit c21004cd5b4cb7d479514d4
> ("MIPS: Rewrite <asm/div64.h> to work with gcc 4.4.0.") makes it depend
> on 64bit kernel by mistake. This patch fix this longstanding error.
>
> Fixes: c21004cd5b4cb7d479514d4 ("MIPS: Rewrite <asm/div64.h> to work with gcc 4.4.0.")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Huacai Chen <chenhuacai@xxxxxxxxxxx>
> ---
> arch/mips/include/asm/div64.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/mips/include/asm/div64.h b/arch/mips/include/asm/div64.h
> index dc5ea5736440..d199fe36eb46 100644
> --- a/arch/mips/include/asm/div64.h
> +++ b/arch/mips/include/asm/div64.h
> @@ -11,7 +11,7 @@
>
> #include <asm-generic/div64.h>
>
> -#if BITS_PER_LONG == 64
> +#if BITS_PER_LONG == 32
>
> #include <linux/types.h>
Hm.
Ingenic jz4780/CI20 build attempt on top ov v5.12-rc6:
In file included from ./include/linux/math64.h:7:0,
from ./include/linux/time.h:6,
from ./include/linux/compat.h:10,
from arch/mips/kernel/asm-offsets.c:12:
./include/linux/math64.h: In function 'div_u64_rem':
./arch/mips/include/asm/div64.h:29:11: error: invalid type argument of unary '*' (have 'long long unsigned int')
__high = *__n >> 32; \
^
./include/asm-generic/div64.h:243:11: note: in expansion of macro '__div64_32'
__rem = __div64_32(&(n), __base); \
^
./include/linux/math64.h:91:15: note: in expansion of macro 'do_div'
*remainder = do_div(dividend, divisor);
^
Or does it just reveal an unknown bug in my compiler?
