Really Fast String Length

strlen.cpp

// This Requires SSE 4.2, Steam hardware survey says this is supported on 97.81% of PCs
// This is REALLY fast, even with -O2 it usually beats the the CRT Implementation 
//  
// In most of my tests this is around 50x faster with small strings with -Od (with -O2 its closer to 10-15x faster) 
//
static size_t 
FastStringLength(char* String)
{
	size_t Result = 0;
	__m128i* Memory = (__m128i *)String;
	const __m128i Zeros = _mm_setzero_si128();
	for (; Memory++; Result += 16)
	{
		const __m128i Data = _mm_loadu_si128(Memory);
		
		if (_mm_cmpistrc(Data, Zeros, _SIDD_CMP_EQUAL_EACH | _SIDD_UBYTE_OPS | _SIDD_LEAST_SIGNIFICANT))
		{
			const int Index = _mm_cmpistri(Data, Zeros, _SIDD_CMP_EQUAL_EACH | _SIDD_UBYTE_OPS | _SIDD_LEAST_SIGNIFICANT);
			return Result + Index;
		}
	}
	
	return(Result);
}