Re: [PATCH v2 2/3] lib/find_bit: create find_first_zero_bit_le()

[Yury Norov]

> Just make the macro take that "last word op" as another argument, and
> then you don't need these tricks, and you can do the _le() version in
> lib/find.c file *exactly* like the regular version, using just
> 
>   #ifdef __BIG_ENDIAN
>   unsigned long find_first_zero_bit_le(const unsigned long *addr,
> unsigned long size)
>   { return FIND_FIRST_BIT(~addr[idx], swab(val), size); }
>   #else
>   unsigned long find_first_zero_bit_le(const unsigned long *addr,
> unsigned long size) __alias(find_first_zero_bit);
>   #endif
> 
> or something like that.
> 
> And yes, it means that the "regular" versions need to pass in just
> "val" as a last-word expression, but who cares? It's still cleaner,
> and easier to explain: "The first expression finds the word that
> matches our requirement, the second expression munges the word we
> found into the bit order we use".

OK. I'll do like this:

   #define FIND_FIRST_BIT(FETCH, MUNG, size) ...

   unsigned long find_first_zero_bit(const unsigned long *addr, unsigned long size)
   {
         return FIND_FIRST_BIT(~addr[idx], /* NOP */, size);
   }

   unsigned long find_first_zero_bit_le(const unsigned long *addr, unsigned long size)
   {
         return FIND_FIRST_BIT(~addr[idx], swab, size);
   }

It doesn't actually look that worse.

Thanks,
Yury