On Fri, Mar 12, 2021 at 10:12:22AM +0100, Rasmus Villemoes wrote:
> On 12/03/2021 06.28, Yury Norov wrote:
> > On Fri, Feb 19, 2021 at 12:07:27AM +0100, Rasmus Villemoes wrote:
> >> On 18/02/2021 05.05, Yury Norov wrote:
> >>> Many algorithms become simpler if they are passed with relatively small
> >>> input values. One example is bitmap operations when the whole bitmap fits
> >>> into one word. To implement such simplifications, linux/bitmap.h declares
> >>> small_const_nbits() macro.
> >>>
> >>> Other subsystems may also benefit from optimizations of this sort, like
> >>> find_bit API in the following patches. So it looks helpful to generalize
> >>> the macro and extend it's visibility.
> >>
> >> Perhaps, but SMALL_CONST is too generic a name, it needs to keep "bits"
> >> somewhere in there. So why not just keep it at small_const_nbits?
> >>
> >>> Signed-off-by: Yury Norov <yury.norov@xxxxxxxxx>
> >>> ---
> >>> include/asm-generic/bitsperlong.h | 2 ++
> >>> include/linux/bitmap.h | 33 ++++++++++++++-----------------
> >>> 2 files changed, 17 insertions(+), 18 deletions(-)
> >>>
> >>> diff --git a/include/asm-generic/bitsperlong.h b/include/asm-generic/bitsperlong.h
> >>> index 3905c1c93dc2..0eeb77544f1d 100644
> >>> --- a/include/asm-generic/bitsperlong.h
> >>> +++ b/include/asm-generic/bitsperlong.h
> >>> @@ -23,4 +23,6 @@
> >>> #define BITS_PER_LONG_LONG 64
> >>> #endif
> >>>
> >>> +#define SMALL_CONST(n) (__builtin_constant_p(n) && (unsigned long)(n) < BITS_PER_LONG)
> >>> +
> >>> #endif /* __ASM_GENERIC_BITS_PER_LONG */
> >>> diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h
> >>> index adf7bd9f0467..e89f1dace846 100644
> >>> --- a/include/linux/bitmap.h
> >>> +++ b/include/linux/bitmap.h
> >>> @@ -224,9 +224,6 @@ extern int bitmap_print_to_pagebuf(bool list, char *buf,
> >>> * so make such users (should any ever turn up) call the out-of-line
> >>> * versions.
> >>> */
> >>> -#define small_const_nbits(nbits) \
> >>> - (__builtin_constant_p(nbits) && (nbits) <= BITS_PER_LONG && (nbits) > 0)
> >>> -
> >>> static inline void bitmap_zero(unsigned long *dst, unsigned int nbits)
> >>> {
> >>> unsigned int len = BITS_TO_LONGS(nbits) * sizeof(unsigned long);
> >>> @@ -278,7 +275,7 @@ extern void bitmap_to_arr32(u32 *buf, const unsigned long *bitmap,
> >>> static inline int bitmap_and(unsigned long *dst, const unsigned long *src1,
> >>> const unsigned long *src2, unsigned int nbits)
> >>> {
> >>> - if (small_const_nbits(nbits))
> >>> + if (SMALL_CONST(nbits - 1))
> >>
> >> Please don't force most users to be changed to something less readable.
> >> What's wrong with just keeping small_const_nbits() the way it is,
> >> avoiding all this churn and keeping the readability?
> >
> > The wrong thing is that it's defined in include/linux/bitmap.h, and I
> > cannot use it in include/asm-generic/bitops/find.h, so I have to either
> > move it to a separate header, or generalize and share with find.h and
> > other users this way. I prefer the latter option, thougt it's more
> > verbose.
>
> The logical place would be the same place the BITS_PER_LONG macro is
> defined, no?
Yes. This where I placed SMALL_CONST() in current version.
> No need to introduce a new header for that, and all current
> users of small_const_nbits() must already (very possibly indirectly)
> include asm-generic/bitsperlong.h.
>
> I do prefer to keep both the name small_const_nbits() and its current
> semantics, which, although not currently spelled out that way anywhere,
> is "is BITMAP_SIZE(nbits) known at compile time and equal to 1", which
> is precisely what allows the static inlines to unconditionally
> dereference the pointer (that's the "exclude the 0 case") and just deal
> with that one word.
>
> I don't like either SMALL_CONST or small_const_size, because nothing in
> there says it has anything to do with bit ops. As I said, if you have
> some special place that for some reason cannot handle
> nbits==BITS_PER_LONG, then just add that as an additional constraint
> with a comment why.
OK, I'll move small_const_nbits() to the bitsperlong header and
resubmit shortly. My concern still is that nbits is too specific
for bitsperlong.h, but if you're good with it, I'm OK as well.
