math.h: Add unsigned 8 bits fractional numbers type

Some users may be requiring only rather small numbers as both
numerator and denominator: add signed and unsigned 8 bits
structs {s8,u8}_fract.

Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Link: https://patch.msgid.link/20240604123008.327424-4-angelogioacchino.delregno@collabora.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>

Author: AngeloGioacchino Del Regno

include/linux/math.h

@@ -112,6 +112,8 @@ struct type##_fract {					\
 	__##type numerator;				\
 	__##type denominator;				\
 };
+__STRUCT_FRACT(s8)
+__STRUCT_FRACT(u8)
 __STRUCT_FRACT(s16)
 __STRUCT_FRACT(u16)
 __STRUCT_FRACT(s32)