1 /* Test program for SSE registers. 2 3 Copyright 2004-2023 Free Software Foundation, Inc. 4 5 This file is part of GDB. 6 7 This program is free software; you can redistribute it and/or modify 8 it under the terms of the GNU General Public License as published by 9 the Free Software Foundation; either version 3 of the License, or 10 (at your option) any later version. 11 12 This program is distributed in the hope that it will be useful, 13 but WITHOUT ANY WARRANTY; without even the implied warranty of 14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 15 GNU General Public License for more details. 16 17 You should have received a copy of the GNU General Public License 18 along with this program. If not, see <http://www.gnu.org/licenses/>. */ 19 20 #include <stdio.h> 21 #include "nat/x86-cpuid.h" 22 23 /* Align sufficient to be able to use movaps. */ 24 #define ALIGN 16 25 26 typedef struct { 27 _Alignas (ALIGN) float f[4]; 28 } v4sf_t; 29 30 31 v4sf_t data_orig[] = 32 { 33 { { 0.0, 0.25, 0.50, 0.75 } }, 34 { { 1.0, 1.25, 1.50, 1.75 } }, 35 { { 2.0, 2.25, 2.50, 2.75 } }, 36 { { 3.0, 3.25, 3.50, 3.75 } }, 37 { { 4.0, 4.25, 4.50, 4.75 } }, 38 { { 5.0, 5.25, 5.50, 5.75 } }, 39 { { 6.0, 6.25, 6.50, 6.75 } }, 40 { { 7.0, 7.25, 7.50, 7.75 } }, 41 #ifdef __x86_64__ 42 { { 8.0, 8.25, 8.50, 8.75 } }, 43 { { 9.0, 9.25, 9.50, 9.75 } }, 44 { { 10.0, 10.25, 10.50, 10.75 } }, 45 { { 11.0, 11.25, 11.50, 11.75 } }, 46 { { 12.0, 12.25, 12.50, 12.75 } }, 47 { { 13.0, 13.25, 13.50, 13.75 } }, 48 { { 14.0, 14.25, 14.50, 14.75 } }, 49 { { 15.0, 15.25, 15.50, 15.75 } }, 50 #endif 51 }; 52 53 54 int 55 have_sse (void) 56 { 57 unsigned int edx; 58 59 if (!x86_cpuid (1, NULL, NULL, NULL, &edx)) 60 return 0; 61 62 if (edx & bit_SSE) 63 return 1; 64 else 65 return 0; 66 } 67 68 #include "../lib/precise-aligned-alloc.c" 69 70 int 71 main (int argc, char **argv) 72 { 73 void *allocated_ptr; 74 v4sf_t *data 75 = precise_aligned_dup (ALIGN, sizeof (data_orig), &allocated_ptr, 76 data_orig); 77 78 if (have_sse ()) 79 { 80 asm ("movaps 0(%0), %%xmm0\n\t" 81 "movaps 16(%0), %%xmm1\n\t" 82 "movaps 32(%0), %%xmm2\n\t" 83 "movaps 48(%0), %%xmm3\n\t" 84 "movaps 64(%0), %%xmm4\n\t" 85 "movaps 80(%0), %%xmm5\n\t" 86 "movaps 96(%0), %%xmm6\n\t" 87 "movaps 112(%0), %%xmm7\n\t" 88 : /* no output operands */ 89 : "r" (data) 90 : "xmm0", "xmm1", "xmm2", "xmm3", "xmm4", "xmm5", "xmm6", "xmm7"); 91 #ifdef __x86_64__ 92 asm ("movaps 128(%0), %%xmm8\n\t" 93 "movaps 144(%0), %%xmm9\n\t" 94 "movaps 160(%0), %%xmm10\n\t" 95 "movaps 176(%0), %%xmm11\n\t" 96 "movaps 192(%0), %%xmm12\n\t" 97 "movaps 208(%0), %%xmm13\n\t" 98 "movaps 224(%0), %%xmm14\n\t" 99 "movaps 240(%0), %%xmm15\n\t" 100 : /* no output operands */ 101 : "r" (data) 102 : "xmm8", "xmm9", "xmm10", "xmm11", "xmm12", "xmm13", "xmm14", "xmm15"); 103 #endif 104 105 asm ("nop"); /* first breakpoint here */ 106 107 asm ( 108 "movaps %%xmm0, 0(%0)\n\t" 109 "movaps %%xmm1, 16(%0)\n\t" 110 "movaps %%xmm2, 32(%0)\n\t" 111 "movaps %%xmm3, 48(%0)\n\t" 112 "movaps %%xmm4, 64(%0)\n\t" 113 "movaps %%xmm5, 80(%0)\n\t" 114 "movaps %%xmm6, 96(%0)\n\t" 115 "movaps %%xmm7, 112(%0)\n\t" 116 : /* no output operands */ 117 : "r" (data) 118 : "xmm0", "xmm1", "xmm2", "xmm3", "xmm4", "xmm5", "xmm6", "xmm7"); 119 #ifdef __x86_64__ 120 asm ( 121 "movaps %%xmm8, 128(%0)\n\t" 122 "movaps %%xmm9, 144(%0)\n\t" 123 "movaps %%xmm10, 160(%0)\n\t" 124 "movaps %%xmm11, 176(%0)\n\t" 125 "movaps %%xmm12, 192(%0)\n\t" 126 "movaps %%xmm13, 208(%0)\n\t" 127 "movaps %%xmm14, 224(%0)\n\t" 128 "movaps %%xmm15, 240(%0)\n\t" 129 : /* no output operands */ 130 : "r" (data) 131 : "xmm8", "xmm9", "xmm10", "xmm11", "xmm12", "xmm13", "xmm14", "xmm15"); 132 #endif 133 134 puts ("Bye!"); /* second breakpoint here */ 135 } 136 137 free (allocated_ptr); 138 139 return 0; 140 } 141