It used to be an ad-hoc hack defined by the x86 version of
<asm/bitops.h> that enabled a couple of library routines to know whether
an integer multiply is faster than repeated shifts and additions.
This just makes it use the real Kconfig system instead, and makes x86
(which was the only architecture that did this) select the option.
NOTE! Even for x86, this really is kind of wrong. If we cared, we would
probably not enable this for builds optimized for netburst (P4), where
shifts-and-adds are generally faster than multiplies. This patch does
*not* change that kind of logic, though, it is purely a syntactic change
with no code changes.
This was triggered by the fact that we have other places that really
want to know "do I want to expand multiples by constants by hand or
not", particularly the hash generation code.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For files only using THIS_MODULE and/or EXPORT_SYMBOL, map
them onto including export.h -- or if the file isn't even
using those, then just delete the include. Fix up any implicit
include dependencies that were being masked by module.h along
the way.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Add support for the hardware version of the Hamming weight function,
popcnt, present in CPUs which advertize it under CPUID, Function
0x0000_0001_ECX[23]. On CPUs which don't support it, we fallback to the
default lib/hweight.c sw versions.
A synthetic benchmark comparing popcnt with __sw_hweight64 showed almost
a 3x speedup on a F10h machine.
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
LKML-Reference: <20100318112015.GC11152@aftab>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Rename the extisting runtime hweight() implementations to
__arch_hweight(), rename the compile-time versions to __const_hweight()
and then have hweight() pick between them.
Suggested-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100318111929.GB11152@aftab>
Acked-by: H. Peter Anvin <hpa@zytor.com>
LKML-Reference: <1265028224.24455.154.camel@laptop>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Optimize hweight32 by using the same technique in hweight64.
The proof of this technique can be found in the commit log for
f9b4192923 ("bitops: hweight()
speedup").
The userspace benchmark on x86_32 showed 20% speedup with
bitmap_weight() which uses hweight32 to count bits for each
unsigned long on 32bit architectures.
int main(void)
{
#define SZ (1024 * 1024 * 512)
static DECLARE_BITMAP(bitmap, SZ) = {
[0 ... 100] = 1,
};
return bitmap_weight(bitmap, SZ);
}
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
LKML-Reference: <1258603932-4590-1-git-send-email-akinobu.mita@gmail.com>
[ only x86 sets ARCH_HAS_FAST_MULTIPLIER so we do this via the x86 tree]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
remove asm/bitops.h includes
including asm/bitops directly may cause compile errors. don't include it
and include linux/bitops instead. next patch will deny including asm header
directly.
Cc: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Based on patch from David Rientjes <rientjes@google.com>, but
changed by AK.
Optimizes the 64-bit hamming weight for x86_64 processors assuming they
have fast multiplication. Uses five fewer bitops than the generic
hweight64. Benchmark on one EMT64 showed ~25% speedup with 2^24
consecutive calls.
Define a new ARCH_HAS_FAST_MULTIPLIER that can be set by other
architectures that can also multiply fast.
Signed-off-by: Andi Kleen <ak@suse.de>
<linux@horizon.com> wrote:
This is an extremely well-known technique. You can see a similar version that
uses a multiply for the last few steps at
http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetParallel whch
refers to "Software Optimization Guide for AMD Athlon 64 and Opteron
Processors"
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/25112.PDF
It's section 8.6, "Efficient Implementation of Population-Count Function in
32-bit Mode", pages 179-180.
It uses the name that I am more familiar with, "popcount" (population count),
although "Hamming weight" also makes sense.
Anyway, the proof of correctness proceeds as follows:
b = a - ((a >> 1) & 0x55555555);
c = (b & 0x33333333) + ((b >> 2) & 0x33333333);
d = (c + (c >> 4)) & 0x0f0f0f0f;
#if SLOW_MULTIPLY
e = d + (d >> 8)
f = e + (e >> 16);
return f & 63;
#else
/* Useful if multiply takes at most 4 cycles */
return (d * 0x01010101) >> 24;
#endif
The input value a can be thought of as 32 1-bit fields each holding their own
hamming weight. Now look at it as 16 2-bit fields. Each 2-bit field a1..a0
has the value 2*a1 + a0. This can be converted into the hamming weight of the
2-bit field a1+a0 by subtracting a1.
That's what the (a >> 1) & mask subtraction does. Since there can be no
borrows, you can just do it all at once.
Enumerating the 4 possible cases:
0b00 = 0 -> 0 - 0 = 0
0b01 = 1 -> 1 - 0 = 1
0b10 = 2 -> 2 - 1 = 1
0b11 = 3 -> 3 - 1 = 2
The next step consists of breaking up b (made of 16 2-bir fields) into
even and odd halves and adding them into 4-bit fields. Since the largest
possible sum is 2+2 = 4, which will not fit into a 4-bit field, the 2-bit
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
"which will not fit into a 2-bit field"
fields have to be masked before they are added.
After this point, the masking can be delayed. Each 4-bit field holds a
population count from 0..4, taking at most 3 bits. These numbers can be added
without overflowing a 4-bit field, so we can compute c + (c >> 4), and only
then mask off the unwanted bits.
This produces d, a number of 4 8-bit fields, each in the range 0..8. From
this point, we can shift and add d multiple times without overflowing an 8-bit
field, and only do a final mask at the end.
The number to mask with has to be at least 63 (so that 32 on't be truncated),
but can also be 128 or 255. The x86 has a special encoding for signed
immediate byte values -128..127, so the value of 255 is slower. On other
processors, a special "sign extend byte" instruction might be faster.
On a processor with fast integer multiplies (Athlon but not P4), you can
reduce the final few serially dependent instructions to a single integer
multiply. Consider d to be 3 8-bit values d3, d2, d1 and d0, each in the
range 0..8. The multiply forms the partial products:
d3 d2 d1 d0
d3 d2 d1 d0
d3 d2 d1 d0
+ d3 d2 d1 d0
----------------------
e3 e2 e1 e0
Where e3 = d3 + d2 + d1 + d0. e2, e1 and e0 obviously cannot generate
any carries.
Signed-off-by: Akinobu Mita <mita@miraclelinux.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch introduces the C-language equivalents of the functions below:
unsigned int hweight32(unsigned int w);
unsigned int hweight16(unsigned int w);
unsigned int hweight8(unsigned int w);
unsigned long hweight64(__u64 w);
In include/asm-generic/bitops/hweight.h
This code largely copied from: include/linux/bitops.h
Signed-off-by: Akinobu Mita <mita@miraclelinux.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
