Christophe Lyon | 073831a | 2011-01-24 17:37:40 +0100 | [diff] [blame^] | 1 | /* |
| 2 | |
| 3 | Copyright (c) 2009, 2010, 2011 STMicroelectronics |
| 4 | Written by Christophe Lyon |
| 5 | |
| 6 | Permission is hereby granted, free of charge, to any person obtaining a copy |
| 7 | of this software and associated documentation files (the "Software"), to deal |
| 8 | in the Software without restriction, including without limitation the rights |
| 9 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 10 | copies of the Software, and to permit persons to whom the Software is |
| 11 | furnished to do so, subject to the following conditions: |
| 12 | |
| 13 | The above copyright notice and this permission notice shall be included in |
| 14 | all copies or substantial portions of the Software. |
| 15 | |
| 16 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 17 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 18 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 19 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 20 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 21 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN |
| 22 | THE SOFTWARE. |
| 23 | |
| 24 | */ |
| 25 | |
| 26 | #ifdef __arm__ |
| 27 | #include <arm_neon.h> |
| 28 | #else |
| 29 | #error Target not supported |
| 30 | #endif |
| 31 | |
| 32 | #include "stm-arm-neon-ref.h" |
| 33 | |
| 34 | #define INSN vqdmull |
| 35 | #define TEST_MSG "VQDMULL_LANE" |
| 36 | #define FNNAME1(NAME) void exec_ ## NAME ## _lane (void) |
| 37 | #define FNNAME(NAME) FNNAME1(NAME) |
| 38 | |
| 39 | FNNAME (INSN) |
| 40 | { |
| 41 | int i; |
| 42 | |
| 43 | /* vector_res = vqdmull_lane(vector,vector2,lane), then store the result. */ |
| 44 | #define TEST_VQDMULL_LANE2(INSN, T1, T2, W, W2, N, L) \ |
| 45 | Neon_Overflow = 0; \ |
| 46 | VECT_VAR(vector_res, T1, W2, N) = \ |
| 47 | INSN##_lane_##T2##W(VECT_VAR(vector, T1, W, N), \ |
| 48 | VECT_VAR(vector2, T1, W, N), \ |
| 49 | L); \ |
| 50 | vst1q_##T2##W2(VECT_VAR(result, T1, W2, N), \ |
| 51 | VECT_VAR(vector_res, T1, W2, N)); \ |
| 52 | dump_neon_overflow(TEST_MSG, xSTR(INSN##_lane_##T2##W)) |
| 53 | |
| 54 | /* Two auxliary macros are necessary to expand INSN */ |
| 55 | #define TEST_VQDMULL_LANE1(INSN, T1, T2, W, W2, N, L) \ |
| 56 | TEST_VQDMULL_LANE2(INSN, T1, T2, W, W2, N, L) |
| 57 | |
| 58 | #define TEST_VQDMULL_LANE(T1, T2, W, W2, N, L) \ |
| 59 | TEST_VQDMULL_LANE1(INSN, T1, T2, W, W2, N, L) |
| 60 | |
| 61 | |
| 62 | /* With ARM RVCT, we need to declare variables before any executable |
| 63 | statement */ |
| 64 | DECL_VARIABLE(vector, int, 16, 4); |
| 65 | DECL_VARIABLE(vector, int, 32, 2); |
| 66 | DECL_VARIABLE(vector2, int, 16, 4); |
| 67 | DECL_VARIABLE(vector2, int, 32, 2); |
| 68 | |
| 69 | DECL_VARIABLE(vector_res, int, 32, 4); |
| 70 | DECL_VARIABLE(vector_res, int, 64, 2); |
| 71 | |
| 72 | clean_results (); |
| 73 | |
| 74 | /* Initialize vector */ |
| 75 | TEST_VDUP(vector, , int, s, 16, 4, 0x1000); |
| 76 | TEST_VDUP(vector, , int, s, 32, 2, 0x1000); |
| 77 | |
| 78 | /* Initialize vector2 */ |
| 79 | TEST_VDUP(vector2, , int, s, 16, 4, 0x4); |
| 80 | TEST_VDUP(vector2, , int, s, 32, 2, 0x2); |
| 81 | |
| 82 | /* Choose lane arbitrarily */ |
| 83 | fprintf(ref_file, "\n%s overflow output:\n", TEST_MSG); |
| 84 | TEST_VQDMULL_LANE(int, s, 16, 32, 4, 2); |
| 85 | TEST_VQDMULL_LANE(int, s, 32, 64, 2, 1); |
| 86 | |
| 87 | |
| 88 | fprintf(ref_file, "\n%s output:\n", TEST_MSG); |
| 89 | DUMP(TEST_MSG, int, 32, 4, PRIx32); |
| 90 | DUMP(TEST_MSG, int, 64, 2, PRIx64); |
| 91 | |
| 92 | |
| 93 | |
| 94 | TEST_VDUP(vector, , int, s, 16, 4, 0x8000); |
| 95 | TEST_VDUP(vector2, , int, s, 16, 4, 0x8000); |
| 96 | TEST_VDUP(vector, , int, s, 32, 2, 0x80000000); |
| 97 | TEST_VDUP(vector2, , int, s, 32, 2, 0x80000000); |
| 98 | fprintf(ref_file, "\n%s overflow output:\n", TEST_MSG " (check mul overflow)"); |
| 99 | TEST_VQDMULL_LANE(int, s, 16, 32, 4, 2); |
| 100 | TEST_VQDMULL_LANE(int, s, 32, 64, 2, 1); |
| 101 | |
| 102 | fprintf (ref_file, "\n%s output:\n", TEST_MSG " (check mul overflow)"); |
| 103 | DUMP(TEST_MSG, int, 32, 4, PRIx32); |
| 104 | DUMP(TEST_MSG, int, 64, 2, PRIx64); |
| 105 | } |