Using fast Intel random generator(SSE2) fails with stack around ... is corrupted

Question

I need very fast(fastest) random generator. I found this one from Intel: Fast Intel Random Number Generator

Looks good. So i created project at MS Visual Studio 2013:

//FastRandom.h:
#pragma once
#include "emmintrin.h"
#include <time.h>
//define this if you wish to return values similar to the standard rand();
#define COMPATABILITY

namespace Brans
{
        __declspec(align(16)) static __m128i cur_seed;

        // uncoment this if you are using intel compiler
        // for MS CL the vectorizer is on by default and jumps in if you
        // compile with /O2 ...
        //#pragma intel optimization_parameter target_arch=avx
        //__declspec(cpu_dispatch(core_2nd_gen_avx, core_i7_sse4_2, core_2_duo_ssse3, generic )
        inline void rand_sse(unsigned int* result)
        {
            __declspec(align(16)) __m128i cur_seed_split;

            __declspec(align(16)) __m128i multiplier;

            __declspec(align(16)) __m128i adder;

            __declspec(align(16)) __m128i mod_mask;

            __declspec(align(16)) __m128i sra_mask;

            __declspec(align(16)) __m128i sseresult;

            __declspec(align(16)) static const unsigned int mult[4] =

            { 214013, 17405, 214013, 69069 };

            __declspec(align(16)) static const unsigned int gadd[4] =

            { 2531011, 10395331, 13737667, 1 };

            __declspec(align(16)) static const unsigned int mask[4] =

            { 0xFFFFFFFF, 0, 0xFFFFFFFF, 0 };

            __declspec(align(16)) static const unsigned int masklo[4] =

            { 0x00007FFF, 0x00007FFF, 0x00007FFF, 0x00007FFF };



            adder = _mm_load_si128((__m128i*) gadd);

            multiplier = _mm_load_si128((__m128i*) mult);

            mod_mask = _mm_load_si128((__m128i*) mask);

            sra_mask = _mm_load_si128((__m128i*) masklo);

            cur_seed_split = _mm_shuffle_epi32(cur_seed, _MM_SHUFFLE(2, 3, 0, 1));



            cur_seed = _mm_mul_epu32(cur_seed, multiplier);

            multiplier = _mm_shuffle_epi32(multiplier, _MM_SHUFFLE(2, 3, 0, 1));

            cur_seed_split = _mm_mul_epu32(cur_seed_split, multiplier);


            cur_seed = _mm_and_si128(cur_seed, mod_mask);

            cur_seed_split = _mm_and_si128(cur_seed_split, mod_mask);

            cur_seed_split = _mm_shuffle_epi32(cur_seed_split, _MM_SHUFFLE(2, 3, 0, 1));

            cur_seed = _mm_or_si128(cur_seed, cur_seed_split);

            cur_seed = _mm_add_epi32(cur_seed, adder);


#ifdef COMPATABILITY



            // Add the lines below if you wish to reduce your results to 16-bit vals...

            sseresult = _mm_srai_epi32(cur_seed, 16);

            sseresult = _mm_and_si128(sseresult, sra_mask);

            _mm_storeu_si128((__m128i*) result, sseresult);

            return;

#endif


            _mm_storeu_si128((__m128i*) result, cur_seed);

            return;
        }

        inline void srand_sse(unsigned int seed)
        {
            cur_seed = _mm_set_epi32(seed, seed + 1, seed, seed + 1);
        }

        inline void srand_sse()
        {
            unsigned int seed = (unsigned int)time(0);
            cur_seed = _mm_set_epi32(seed, seed + 1, seed, seed + 1);
        }

        inline unsigned int GetRandom(unsigned int low, unsigned int high)
        {
            unsigned int ret = 0;
            rand_sse(&ret);
            return ret % (high - low + 1) + low;
        }

    };


// Test.cpp : Defines the entry point for the console application.
//

#include "stdafx.h"
#include "FastRandom.h"
#include <iostream>

using namespace Brans;

int _tmain(int argc, _TCHAR* argv[])
{
    srand_sse();
    unsigned int result = 0;
    for (size_t i = 0; i < 10000; i++)
    {
        result += GetRandom(1, 50);
        result -= GetRandom(1, 50);
    }

    std::cout << result << std::endl;
    return 0;
}

I expect 0 result +- 50. But when i run program in Debug, i got: Run-Time Check Failure #2 - Stack around the variable 'ret' was corrupted. at GetRandom(...). When i run it in release i got undefined result, up to max unsigned int. (I am using intel i5 processor).

What is wrong?

========= Add to accepted answer, also i have mistake that i should use long instead of unsigned int because negative result became large positive for unsigned.

Answer 1

From the docs of Intel Fast Random Generator:

The rand_sse() function implements a vectorized version of this fast_rand() function, where the integer math operations are done in fours, using the SIMD architecture.

It means rand_sse generates you 4 random numbers at once using sse2.

So you need to give it array of unsigned int's:

unsigned int result[4];
rand_sse( result );