Pádraig Brady wrote:
I noticed that count_one_bits() branches on the popcount_support
variable on each call, which might negate much of the gain from using this
instruction?
Could be. As far as I know it's never been benchmarked. I stole that code from
Emacs without investigating its perfo
On 05/01/15 21:32, Gisle Vanem wrote:
> Pádraig Brady wrote:
>
>> Does the following address the issue for you?
>> It should drop back to the more generic slow path on 32 bit.
>
> But I do have _BitReverse. That could be an intrinsic.
I was assuming we could do the simpler patch and be
slightly
Pádraig Brady wrote:
Does the following address the issue for you?
It should drop back to the more generic slow path on 32 bit.
But I do have _BitReverse. That could be an intrinsic.
BTW do we have the same issue with __popcnt64 in count-one-bits.h?
Seems so:
__MACHINEX64(unsigned __int6
On 05/01/15 20:05, Gisle Vanem wrote:
> I'm on Win-XP (32bit, MSVC v16) and get a link error on a
> missing '__BitScanReverse64()' in test-count-leading-zeros.c:.
> Here is the dis-asm:
>
> _count_leading_zeros_ll:
> pushebp
> mov ebp,esp
> s
I'm on Win-XP (32bit, MSVC v16) and get a link error on a
missing '__BitScanReverse64()' in test-count-leading-zeros.c:.
Here is the dis-asm:
_count_leading_zeros_ll:
pushebp
mov ebp,esp
sub esp,0x0008
X$72:
mov
