| OLD | NEW |
| 1 /* | 1 /* |
| 2 * Copyright (c) 2013 The WebM project authors. All Rights Reserved. | 2 * Copyright (c) 2013 The WebM project authors. All Rights Reserved. |
| 3 * | 3 * |
| 4 * Use of this source code is governed by a BSD-style license | 4 * Use of this source code is governed by a BSD-style license |
| 5 * that can be found in the LICENSE file in the root of the source | 5 * that can be found in the LICENSE file in the root of the source |
| 6 * tree. An additional intellectual property rights grant can be found | 6 * tree. An additional intellectual property rights grant can be found |
| 7 * in the file PATENTS. All contributing project authors may | 7 * in the file PATENTS. All contributing project authors may |
| 8 * be found in the AUTHORS file in the root of the source tree. | 8 * be found in the AUTHORS file in the root of the source tree. |
| 9 */ | 9 */ |
| 10 | 10 |
| 11 #include <assert.h> | 11 #include <assert.h> |
| 12 #include <stdio.h> | 12 #include <stdio.h> |
| 13 | 13 |
| 14 #include "./vpx_config.h" | 14 #include "./vpx_config.h" |
| 15 #include "./vp9_rtcd.h" | 15 #include "./vp9_rtcd.h" |
| 16 #include "vp9/common/vp9_common.h" | 16 #include "vp9/common/vp9_common.h" |
| 17 #include "vp9/common/vp9_blockd.h" | 17 #include "vp9/common/vp9_blockd.h" |
| 18 #include "vp9/common/vp9_idct.h" | 18 #include "vp9/common/vp9_idct.h" |
| 19 #include "vp9/common/mips/dspr2/vp9_common_dspr2.h" | 19 #include "vp9/common/mips/dspr2/vp9_common_dspr2.h" |
| 20 | 20 |
| 21 #if HAVE_DSPR2 | 21 #if HAVE_DSPR2 |
| 22 static void idct32_1d_rows_dspr2(const int16_t *input, int16_t *output, | 22 static void idct32_rows_dspr2(const int16_t *input, int16_t *output, |
| 23 uint32_t no_rows) { | 23 uint32_t no_rows) { |
| 24 int16_t step1_0, step1_1, step1_2, step1_3, step1_4, step1_5, step1_6; | 24 int16_t step1_0, step1_1, step1_2, step1_3, step1_4, step1_5, step1_6; |
| 25 int16_t step1_7, step1_8, step1_9, step1_10, step1_11, step1_12, step1_13; | 25 int16_t step1_7, step1_8, step1_9, step1_10, step1_11, step1_12, step1_13; |
| 26 int16_t step1_14, step1_15, step1_16, step1_17, step1_18, step1_19, step1_20; | 26 int16_t step1_14, step1_15, step1_16, step1_17, step1_18, step1_19, step1_20; |
| 27 int16_t step1_21, step1_22, step1_23, step1_24, step1_25, step1_26, step1_27; | 27 int16_t step1_21, step1_22, step1_23, step1_24, step1_25, step1_26, step1_27; |
| 28 int16_t step1_28, step1_29, step1_30, step1_31; | 28 int16_t step1_28, step1_29, step1_30, step1_31; |
| 29 int16_t step2_0, step2_1, step2_2, step2_3, step2_4, step2_5, step2_6; | 29 int16_t step2_0, step2_1, step2_2, step2_3, step2_4, step2_5, step2_6; |
| 30 int16_t step2_7, step2_8, step2_9, step2_10, step2_11, step2_12, step2_13; | 30 int16_t step2_7, step2_8, step2_9, step2_10, step2_11, step2_12, step2_13; |
| 31 int16_t step2_14, step2_15, step2_16, step2_17, step2_18, step2_19, step2_20; | 31 int16_t step2_14, step2_15, step2_16, step2_17, step2_18, step2_19, step2_20; |
| 32 int16_t step2_21, step2_22, step2_23, step2_24, step2_25, step2_26, step2_27; | 32 int16_t step2_21, step2_22, step2_23, step2_24, step2_25, step2_26, step2_27; |
| 33 int16_t step2_28, step2_29, step2_30, step2_31; | 33 int16_t step2_28, step2_29, step2_30, step2_31; |
| (...skipping 841 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
| 875 uint32_t pos = 45; | 875 uint32_t pos = 45; |
| 876 | 876 |
| 877 /* bit positon for extract from acc */ | 877 /* bit positon for extract from acc */ |
| 878 __asm__ __volatile__ ( | 878 __asm__ __volatile__ ( |
| 879 "wrdsp %[pos], 1 \n\t" | 879 "wrdsp %[pos], 1 \n\t" |
| 880 : | 880 : |
| 881 : [pos] "r" (pos) | 881 : [pos] "r" (pos) |
| 882 ); | 882 ); |
| 883 | 883 |
| 884 // Rows | 884 // Rows |
| 885 idct32_1d_rows_dspr2(input, outptr, 32); | 885 idct32_rows_dspr2(input, outptr, 32); |
| 886 | 886 |
| 887 // Columns | 887 // Columns |
| 888 vp9_idct32_1d_cols_add_blk_dspr2(out, dest, dest_stride); | 888 vp9_idct32_cols_add_blk_dspr2(out, dest, dest_stride); |
| 889 } | 889 } |
| 890 | 890 |
| 891 void vp9_idct32x32_34_add_dspr2(const int16_t *input, uint8_t *dest, | 891 void vp9_idct32x32_34_add_dspr2(const int16_t *input, uint8_t *dest, |
| 892 int stride) { | 892 int stride) { |
| 893 DECLARE_ALIGNED(32, int16_t, out[32 * 32]); | 893 DECLARE_ALIGNED(32, int16_t, out[32 * 32]); |
| 894 int16_t *outptr = out; | 894 int16_t *outptr = out; |
| 895 uint32_t i; | 895 uint32_t i; |
| 896 uint32_t pos = 45; | 896 uint32_t pos = 45; |
| 897 | 897 |
| 898 /* bit positon for extract from acc */ | 898 /* bit positon for extract from acc */ |
| 899 __asm__ __volatile__ ( | 899 __asm__ __volatile__ ( |
| 900 "wrdsp %[pos], 1 \n\t" | 900 "wrdsp %[pos], 1 \n\t" |
| 901 : | 901 : |
| 902 : [pos] "r" (pos) | 902 : [pos] "r" (pos) |
| 903 ); | 903 ); |
| 904 | 904 |
| 905 // Rows | 905 // Rows |
| 906 idct32_1d_rows_dspr2(input, outptr, 8); | 906 idct32_rows_dspr2(input, outptr, 8); |
| 907 | 907 |
| 908 outptr += 8; | 908 outptr += 8; |
| 909 __asm__ __volatile__ ( | 909 __asm__ __volatile__ ( |
| 910 "sw $zero, 0(%[outptr]) \n\t" | 910 "sw $zero, 0(%[outptr]) \n\t" |
| 911 "sw $zero, 4(%[outptr]) \n\t" | 911 "sw $zero, 4(%[outptr]) \n\t" |
| 912 "sw $zero, 8(%[outptr]) \n\t" | 912 "sw $zero, 8(%[outptr]) \n\t" |
| 913 "sw $zero, 12(%[outptr]) \n\t" | 913 "sw $zero, 12(%[outptr]) \n\t" |
| 914 "sw $zero, 16(%[outptr]) \n\t" | 914 "sw $zero, 16(%[outptr]) \n\t" |
| 915 "sw $zero, 20(%[outptr]) \n\t" | 915 "sw $zero, 20(%[outptr]) \n\t" |
| 916 "sw $zero, 24(%[outptr]) \n\t" | 916 "sw $zero, 24(%[outptr]) \n\t" |
| (...skipping 23 matching lines...) Expand all Loading... |
| 940 "sw $zero, 36(%[outptr]) \n\t" | 940 "sw $zero, 36(%[outptr]) \n\t" |
| 941 "sw $zero, 40(%[outptr]) \n\t" | 941 "sw $zero, 40(%[outptr]) \n\t" |
| 942 "sw $zero, 44(%[outptr]) \n\t" | 942 "sw $zero, 44(%[outptr]) \n\t" |
| 943 | 943 |
| 944 : | 944 : |
| 945 : [outptr] "r" (outptr) | 945 : [outptr] "r" (outptr) |
| 946 ); | 946 ); |
| 947 } | 947 } |
| 948 | 948 |
| 949 // Columns | 949 // Columns |
| 950 vp9_idct32_1d_cols_add_blk_dspr2(out, dest, stride); | 950 vp9_idct32_cols_add_blk_dspr2(out, dest, stride); |
| 951 } | 951 } |
| 952 | 952 |
| 953 void vp9_idct32x32_1_add_dspr2(const int16_t *input, uint8_t *dest, | 953 void vp9_idct32x32_1_add_dspr2(const int16_t *input, uint8_t *dest, |
| 954 int stride) { | 954 int stride) { |
| 955 int r, out; | 955 int r, out; |
| 956 int32_t a1, absa1; | 956 int32_t a1, absa1; |
| 957 int32_t vector_a1; | 957 int32_t vector_a1; |
| 958 int32_t t1, t2, t3, t4; | 958 int32_t t1, t2, t3, t4; |
| 959 int32_t vector_1, vector_2, vector_3, vector_4; | 959 int32_t vector_1, vector_2, vector_3, vector_4; |
| 960 uint32_t pos = 45; | 960 uint32_t pos = 45; |
| (...skipping 106 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
| 1067 : [t1] "=&r" (t1), [t2] "=&r" (t2), [t3] "=&r" (t3), [t4] "=&r" (t4), | 1067 : [t1] "=&r" (t1), [t2] "=&r" (t2), [t3] "=&r" (t3), [t4] "=&r" (t4), |
| 1068 [vector_1] "=&r" (vector_1), [vector_2] "=&r" (vector_2), | 1068 [vector_1] "=&r" (vector_1), [vector_2] "=&r" (vector_2), |
| 1069 [vector_3] "=&r" (vector_3), [vector_4] "=&r" (vector_4), | 1069 [vector_3] "=&r" (vector_3), [vector_4] "=&r" (vector_4), |
| 1070 [dest] "+&r" (dest) | 1070 [dest] "+&r" (dest) |
| 1071 : [stride] "r" (stride), [vector_a1] "r" (vector_a1) | 1071 : [stride] "r" (stride), [vector_a1] "r" (vector_a1) |
| 1072 ); | 1072 ); |
| 1073 } | 1073 } |
| 1074 } | 1074 } |
| 1075 } | 1075 } |
| 1076 #endif // #if HAVE_DSPR2 | 1076 #endif // #if HAVE_DSPR2 |
| OLD | NEW |