Author: Leen Ammeraal
Date: 09:22:15 01/22/02
Go up one level in this thread
Hi Miguel,
I tested Dann's version, LOWBIT, not your routine.
I forgot to write that I admire
this almost magical algorithm very much and that I would
be very pleased if someone could explain it. If your
experiments show that your version is better, please let
me (or rather, us) know. I remember that this assembly
routine was faster than the C(++) routine that I replaced
(a long time ago).
By the way, it is a long time since I wrote assembly
language myself, and I don't understand this routine
either. In particular, I find it miraculous that it
contains neither a loop nor a table-lookup. Any explanation
would be very welcome, preferably in a C equivalent,
if possible, just to show how it works.
Leen
On January 22, 2002 at 12:01:07, Miguel A. Ballicora wrote:
>On January 22, 2002 at 11:23:36, Leen Ammeraal wrote:
>
>>I did an experiment with your function lowbit, adapting
>>it (by using a union of both an __int64 and a
>>struct of two ints, in view of 64 bit bitboards)
>>but it was slightly slower (with very little difference)
>>than the following assembly routine which you (Dann) also
>>provided:
>>
>>__forceinline
>>int FirstPiece(BITBOARD bits)
>>{
>>// Given is: bits != 0
>>__asm {
>>; 64-bit number to move into ECX:EBX, will never be zero!
>> mov ecx,dword ptr [bits+4]
>> mov ebx,dword ptr [bits]
>> bsf eax,ecx
>> add eax,32
>> bsf eax,ebx
>> }
>>}
>>
>>So I think I will stick to this assembly function.
>>Leen
>
>Hi Leen,
>
>I am a bit confused. You answer to Dann but you talk about the function lowbit()
>I posted a function lowbit and Dann posted a macro LOWBIT (by L.Kirby).
>Which one you say you tested and is slightly slower than the assembler version?
>
>If the difference is not high, maybe the C version of (LOWBIT or lowbit or
>whatever) could be optimized and or coded in assembler.
>
>Regards,
>Miguel
>PS: did you test lowbit2?
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.