Information
The topic you requested is included in another documentation set. For convenience, it's displayed below. Choose Switch to see the topic in its original location.
We recommend using Visual Studio 2017

__lzcnt16, __lzcnt, __lzcnt64

 

The new home for Visual Studio documentation is Visual Studio 2017 Documentation on docs.microsoft.com.

The latest version of this topic can be found at __lzcnt16, __lzcnt, __lzcnt64.

Microsoft Specific**

Counts the number of leading zeros in a 16-, 32-, or 64-byte integer.

unsigned short __lzcnt16(  
   unsigned short value  
);  
unsigned int __lzcnt(  
   unsigned int value  
);  
unsigned __int64 __lzcnt64(  
   unsigned __int64 value  
);  

Parameters

[in] value
The 16-, 32-, or 64-bit unsigned integer to scan for leading zeros.

The number of leading zero bits in the value parameter. If value is zero, the return value is the size of the input operand (16, 32, or 64). If the most significant bit of value is one, the return value is zero.

IntrinsicArchitecture
__lzcnt16AMD: Advanced Bit Manipulation (ABM)

Intel: Haswell
__lzcntAMD: Advanced Bit Manipulation (ABM)

Intel: Haswell
__lzcnt64AMD: Advanced Bit Manipulation (ABM) in 64-bit mode.

Intel: Haswell

Header file <intrin.h>

Each of these intrinsics generates the lzcnt instruction. The size of the value that the lzcnt instruction returns is the same as the size of its argument. In 32-bit mode there are no 64-bit general-purpose registers, hence no 64-bit lzcnt.

To determine hardware support for the lzcnt instruction call the __cpuid intrinsic with InfoType=0x80000001 and check bit 5 of CPUInfo[2] (ECX). This bit will be 1 if the instruction is supported, and 0 otherwise. If you run code that uses this intrinsic on hardware that does not support the lzcnt instruction, the results are unpredictable.

On Intel processors that do not support the lzcnt instruction, the instruction byte encoding is executed as bsr (bit scan reverse). If code portability is a concern, consider use of the _BitScanReverse intrinsic instead. For more information, see _BitScanReverse, _BitScanReverse64.

// Compile this test with: /EHsc  
#include <iostream>   
#include <intrin.h>   
using namespace std;   
  
int main()   
{  
  unsigned short us[3] = {0, 0xFF, 0xFFFF};  
  unsigned short usr;  
  unsigned int   ui[4] = {0, 0xFF, 0xFFFF, 0xFFFFFFFF};  
  unsigned int   uir;  
  
  for (int i=0; i<3; i++) {  
    usr = __lzcnt16(us[i]);  
    cout << "__lzcnt16(0x" << hex << us[i] << ") = " << dec << usr << endl;  
  }  
  
  for (int i=0; i<4; i++) {  
    uir = __lzcnt(ui[i]);  
    cout << "__lzcnt(0x" << hex << ui[i] << ") = " << dec << uir << endl;  
  }  
}  
  

__lzcnt16(0x0) = 16  
__lzcnt16(0xff) = 8  
__lzcnt16(0xffff) = 0  
__lzcnt(0x0) = 32  
__lzcnt(0xff) = 24  
__lzcnt(0xffff) = 16  
__lzcnt(0xffffffff) = 0  

Portions of this content are Copyright 2007 by Advanced Micro Devices, Inc. All rights reserved. Reproduced with permission from Advanced Micro Devices, Inc.

Compiler Intrinsics

Show: