icc

Are there any C compilers that have extensions to store an array in column-major order instead of the standard row-major order? Short answer is "No". Long answer is that storing an array in column-major order would break the one-to-one correspondence between array index operations and pointer arithmetics, and the way an N-dimension array is sliced into N-1 dimension arrays. Consider a 10x20 array stored in column-major order. This means that cells in adjacent columns would be next to each other in memory. On the other hand, converting a pointer to array element at i, j to an element pointer

I've noticed that accessing __m128 fields by index is possible in gcc , without using the union trick. __m128 t; float r(t[0] + t[1] + t[2] + t[3]); I can also load a __m128 just like an array: __m128 t{1.f, 2.f, 3.f, 4.f}; This is all in line with gcc 's vector extensions. These, however, may not be available elsewhere. Are the loading and accessing features supported by the intel compiler and msvc? To load a __m128 , you can write _mm_setr_ps(1.f, 2.f, 3.f, 4.f) , which is supported by GCC, ICC, MSVC and clang. So far as I know, clang and recent versions of GCC support accessing __m128

I am using icpc (non optional) and I am compiling with -std=c++0x so I can use lambas. However when I do so it creates havok with gcc stdlib with features that one supports that the other doesn't. I have tried defining __GXX_EXPERIMENTAL_CXX0X__ but that didn't help. So ideally what I am asking for is the ability to use the c++0x language features with the C++03 stdlib. gcc 4.6 icc 12.1] EDIT example of error: /usr/include/c++/4.6.2/type_traits(74): error: identifier "constexpr" is undefined static constexpr _Tp value = __v; ^ /usr/include/c++/4.6.2/type_traits(74): error: expected a ";"

This is a followup on this question . The code below for a 4x4 matrix multiplication C = AB compiles fine on ICC on all optimization settings. It executes correctly on -O1 and -O2, but gives an incorrect result on -O3. The problem seems to come from the _mm256_storeu_pd operation, as substituting it (and only it) with the asm statement below gives correct results after execution. Any ideas? inline void RunIntrinsics_FMA_UnalignedCopy_MultiplyMatrixByMatrix(double *A, double *B, double *C) { size_t i; /* the registers you use */ __m256d a0, a1, a2, a3, b0, b1, b2, b3, sum; // __m256d *C256 = (_

I am trying to compile a program on a 64-bit machine with icpc. Unfortunately, I get an error message of catastrophic error: cannot open source file "bits/c++config.h" . I have adopded some advice from here but with no success. Also, I do not forget to run source /opt/intel/bin/compilervars.sh intel64 before hand and I'm on Ubuntu 13.10 in case this is important. Sven First, find the missing file: find /usr -name c++config.h (Headers are in /usr , most of the time.) Then, add its top dir to the include path of your compilation command, so the compiler will find "bits/c++config.h", using the -I

GCC and Clang have the __int128_t and __uint128_t extensions for 128-bit integer arithmetic. I was hopeful that __m128i would give something similar for the Intel C Compiler, but (if it's even possible) it looks to me like I'd have to write explicit SSE2 function calls in order to use __m128i , instead of using "built-in" operators like + , - , * , / , and % . I was hoping to do something like this (this doesn't work): #if defined(__INTEL_COMPILER) && defined(__SSE2__) #include "xmmintrin.h" typedef __u128 uint128_t; #elif defined (__GNUC__) typedef __uint128_t uint128_t; #else #error For 128

simd pragma can be used with icc compiler to perform a reduction operator: #pragma simd #pragma simd reduction(+:acc) #pragma ivdep for(int i( 0 ); i < N; ++i ) { acc += x[i]; } Is there any equivalent solution in msvc or/and gcc? Ref(p28): http://d3f8ykwhia686p.cloudfront.net/1live/intel/CompilerAutovectorizationGuide.pdf GCC definitely can vectorize. Suppose you have file reduc.c with contents: int foo(int *x, int N) { int acc, i; for( i = 0; i < N; ++i ) { acc += x[i]; } return acc; } Compile it (I used gcc 4.7.2) with command line: $ gcc -O3 -S reduc.c -ftree-vectorize -msse2 Now you can