Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Nalimov: bsf/bsr intrinsics implementation still not optimal

Author: Eugene Nalimov

Date: 10:51:19 09/23/04

Go up one level in this thread


On September 23, 2004 at 13:47:31, Gerd Isenberg wrote:

>On September 23, 2004 at 13:33:44, Eugene Nalimov wrote:
>
>>I worked on IPF/AMD64 implementation. I tried your test case, and here is what I
>>got on AMD64 using latest compiler:
>>
>>PUBLIC	?testbsf@@YAHH@Z				; testbsf
>>; Function compile flags: /Ogtpy
>>_TEXT	SEGMENT
>>mask$ = 8
>>?testbsf@@YAHH@Z PROC					; testbsf
>>; File c:\repro\w.cpp
>>; Line 16
>>	xor	eax, eax
>>; Line 18
>>	test	ecx, ecx
>>	mov	r8d, ecx
>>	je	SHORT $LN7@testbsf
>>	npad	7
>>$LL2@testbsf:
>>; Line 21
>>	bsf	ecx, r8d
>>; Line 22
>>	mov	edx, 1
>>	shl	edx, cl
>>; Line 23
>>	or	eax, edx
>>; Line 24
>>	xor	r8d, edx
>>	jne	SHORT $LL2@testbsf
>>$LN7@testbsf:
>>; Line 28
>>	ret	0
>>?testbsf@@YAHH@Z ENDP					; testbsf
>>
>
>Is there a bsf-intrinsic for unsigned __int64 too?

_BitScanForward64/_BitScanReverse64, but they are supported only on 64-bit
platforms.


Thanks,
Eugene



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.