gistfile1.txt

~ gcc --version
gcc (GCC) 6.3.1 20170306
Copyright (C) 2016 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

~ gcc -march=native -Wall -Wextra -O3 test.c -o test
test.c: In function ‘main’:
test.c:5:14: warning: unused parameter ‘argc’ [-Wunused-parameter]
 int main(int argc, char** argv) {
              ^~~~
test.c:5:27: warning: unused parameter ‘argv’ [-Wunused-parameter]
 int main(int argc, char** argv) {
                           ^~~~
~ ./test
15
15
0
~ valgrind --tool=callgrind ./test 
==19984== Callgrind, a call-graph generating cache profiler
==19984== Copyright (C) 2002-2015, and GNU GPL'd, by Josef Weidendorfer et al.
==19984== Using Valgrind-3.12.0 and LibVEX; rerun with -h for copyright info
==19984== Command: ./test
==19984== 
==19984== For interactive control, run 'callgrind_control -h'.
15
0
0
==19984== 
==19984== Events    : Ir
==19984== Collected : 158231
==19984== 
==19984== I   refs:      158,231
~ valgrind --tool=memcheck ./test 
==19988== Memcheck, a memory error detector
==19988== Copyright (C) 2002-2015, and GNU GPL'd, by Julian Seward et al.
==19988== Using Valgrind-3.12.0 and LibVEX; rerun with -h for copyright info
==19988== Command: ./test
==19988== 
15
15
0
==19988== 
==19988== HEAP SUMMARY:
==19988==     in use at exit: 0 bytes in 0 blocks
==19988==   total heap usage: 1 allocs, 1 frees, 1,024 bytes allocated
==19988== 
==19988== All heap blocks were freed -- no leaks are possible
==19988== 
==19988== For counts of detected and suppressed errors, rerun with: -v
==19988== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

test.c

#include <stdio.h>
#include <immintrin.h>


int main(int argc, char** argv) {
    const __m256i one = _mm256_set_epi64x(0xffffffffffffffff, 0xffffffffffffffff, 0xffffffffffffffff, 0xffffffffffffffff);
    long long unsigned p[3] = {0};
    long long unsigned tmp[4];
    
    __m256i B1 = _mm256_set_epi64x(0, 0xf, 0, 0xf);
    __m256i B2 = _mm256_set_epi64x(0, 0, 0, 0);

    _mm256_maskstore_epi64((long long int*)tmp, one, B1);
    p[0] = tmp[2];
    p[1] = tmp[0];
    _mm256_maskstore_epi64((long long int*)tmp, one, B2);

    for(int i = 0 ; i <3; ++i) { //Somehow it is important that this loop is not unrolled
        printf("%d\n", (int)p[i]);
    }
}

test.s

	.file	"test.c"
	.section	.rodata.str1.1,"aMS",@progbits,1
.LC1:
	.string	"%d\n"
	.section	.text.startup,"ax",@progbits
	.p2align 4,,15
	.globl	main
	.type	main, @function
main:
.LFB4605:
	.cfi_startproc
	pushq	%rbp
	.cfi_def_cfa_offset 16
	.cfi_offset 6, -16
	vpcmpeqd	%ymm0, %ymm0, %ymm0
	movq	%rsp, %rbp
	.cfi_def_cfa_register 6
	pushq	%r12
	pushq	%r10
	pushq	%rbx
	.cfi_offset 12, -24
	.cfi_offset 10, -32
	.cfi_offset 3, -40
	leaq	-96(%rbp), %rbx
	leaq	24(%rbx), %r12
	subq	$72, %rsp
	vmovdqa	.LC0(%rip), %ymm1
	movq	$0, -80(%rbp)
	vpmaskmovq	%ymm1, %ymm0, -64(%rbp)
	movq	-48(%rbp), %rax
	vpxor	%xmm1, %xmm1, %xmm1
	movq	%rax, -96(%rbp)
	movq	-64(%rbp), %rax
	vpmaskmovq	%ymm1, %ymm0, -64(%rbp)
	movq	%rax, -88(%rbp)
	vzeroupper
.L2:
	movl	(%rbx), %esi
	movl	$.LC1, %edi
	xorl	%eax, %eax
	addq	$8, %rbx
	call	printf
	cmpq	%r12, %rbx
	jne	.L2
	addq	$72, %rsp
	xorl	%eax, %eax
	popq	%rbx
	popq	%r10
	popq	%r12
	popq	%rbp
	.cfi_def_cfa 7, 8
	ret
	.cfi_endproc
.LFE4605:
	.size	main, .-main
	.section	.rodata.cst32,"aM",@progbits,32
	.align 32
.LC0:
	.quad	15
	.quad	0
	.quad	15
	.quad	0
	.ident	"GCC: (GNU) 6.3.1 20170306"
	.section	.note.GNU-stack,"",@progbits

test2.s

	.file	"test.c"
	.section	.rodata.str1.1,"aMS",@progbits,1
.LC1:
	.string	"%d\n"
	.section	.text.startup,"ax",@progbits
	.p2align 4,,15
	.globl	main
	.type	main, @function
main:
.LFB4605:
	.cfi_startproc
	pushq	%rbp
	.cfi_def_cfa_offset 16
	.cfi_offset 6, -16
	vpcmpeqd	%ymm0, %ymm0, %ymm0
	movl	$.LC1, %edi
	xorl	%eax, %eax
	movq	%rsp, %rbp
	.cfi_def_cfa_register 6
	pushq	%r10
	subq	$40, %rsp
	.cfi_offset 10, -24
	vmovdqa	.LC0(%rip), %ymm1
	vpmaskmovq	%ymm1, %ymm0, -48(%rbp)
	movq	-48(%rbp), %rsi
	vpxor	%xmm1, %xmm1, %xmm1
	vpmaskmovq	%ymm1, %ymm0, -48(%rbp)
	vzeroupper
	call	printf
	xorl	%esi, %esi
	movl	$.LC1, %edi
	xorl	%eax, %eax
	call	printf
	addq	$40, %rsp
	xorl	%eax, %eax
	popq	%r10
	popq	%rbp
	.cfi_def_cfa 7, 8
	ret
	.cfi_endproc
.LFE4605:
	.size	main, .-main
	.section	.rodata.cst32,"aM",@progbits,32
	.align 32
.LC0:
	.quad	15
	.quad	0
	.quad	15
	.quad	0
	.ident	"GCC: (GNU) 6.3.1 20170306"
	.section	.note.GNU-stack,"",@progbits