| 1 | /* |
| 2 | * Copyright (C) 2010 Mans Rullgard <mans@mansr.com> |
| 3 | * |
| 4 | * This file is part of FFmpeg. |
| 5 | * |
| 6 | * FFmpeg is free software; you can redistribute it and/or |
| 7 | * modify it under the terms of the GNU Lesser General Public |
| 8 | * License as published by the Free Software Foundation; either |
| 9 | * version 2.1 of the License, or (at your option) any later version. |
| 10 | * |
| 11 | * FFmpeg is distributed in the hope that it will be useful, |
| 12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 14 | * Lesser General Public License for more details. |
| 15 | * |
| 16 | * You should have received a copy of the GNU Lesser General Public |
| 17 | * License along with FFmpeg; if not, write to the Free Software |
| 18 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 19 | */ |
| 20 | |
| 21 | #ifndef AVCODEC_ARM_VP56_ARITH_H |
| 22 | #define AVCODEC_ARM_VP56_ARITH_H |
| 23 | |
| 24 | #if CONFIG_THUMB |
| 25 | # define A(x) |
| 26 | # define T(x) x |
| 27 | #else |
| 28 | # define A(x) x |
| 29 | # define T(x) |
| 30 | #endif |
| 31 | |
| 32 | #if CONFIG_THUMB || defined __clang__ |
| 33 | # define L(x) |
| 34 | # define U(x) x |
| 35 | #else |
| 36 | # define L(x) x |
| 37 | # define U(x) |
| 38 | #endif |
| 39 | |
| 40 | #if HAVE_ARMV6_INLINE |
| 41 | |
| 42 | #define vp56_rac_get_prob vp56_rac_get_prob_armv6 |
| 43 | static inline int vp56_rac_get_prob_armv6(VP56RangeCoder *c, int pr) |
| 44 | { |
| 45 | unsigned shift = ff_vp56_norm_shift[c->high]; |
| 46 | unsigned code_word = c->code_word << shift; |
| 47 | unsigned high = c->high << shift; |
| 48 | unsigned bit; |
| 49 | |
| 50 | __asm__ ("adds %3, %3, %0 \n" |
| 51 | "itt cs \n" |
| 52 | "cmpcs %7, %4 \n" |
| 53 | L("ldrcsh %2, [%4], #2 \n") |
| 54 | U("ldrhcs %2, [%4], #2 \n") |
| 55 | "rsb %0, %6, #256 \n" |
| 56 | "smlabb %0, %5, %6, %0 \n" |
| 57 | T("itttt cs \n") |
| 58 | "rev16cs %2, %2 \n" |
| 59 | T("lslcs %2, %2, %3 \n") |
| 60 | T("orrcs %1, %1, %2 \n") |
| 61 | A("orrcs %1, %1, %2, lsl %3 \n") |
| 62 | "subcs %3, %3, #16 \n" |
| 63 | "lsr %0, %0, #8 \n" |
| 64 | "cmp %1, %0, lsl #16 \n" |
| 65 | "ittte ge \n" |
| 66 | "subge %1, %1, %0, lsl #16 \n" |
| 67 | "subge %0, %5, %0 \n" |
| 68 | "movge %2, #1 \n" |
| 69 | "movlt %2, #0 \n" |
| 70 | : "=&r"(c->high), "=&r"(c->code_word), "=&r"(bit), |
| 71 | "+&r"(c->bits), "+&r"(c->buffer) |
| 72 | : "r"(high), "r"(pr), "r"(c->end - 1), |
| 73 | "0"(shift), "1"(code_word) |
| 74 | : "cc"); |
| 75 | |
| 76 | return bit; |
| 77 | } |
| 78 | |
| 79 | #define vp56_rac_get_prob_branchy vp56_rac_get_prob_branchy_armv6 |
| 80 | static inline int vp56_rac_get_prob_branchy_armv6(VP56RangeCoder *c, int pr) |
| 81 | { |
| 82 | unsigned shift = ff_vp56_norm_shift[c->high]; |
| 83 | unsigned code_word = c->code_word << shift; |
| 84 | unsigned high = c->high << shift; |
| 85 | unsigned low; |
| 86 | unsigned tmp; |
| 87 | |
| 88 | __asm__ ("adds %3, %3, %0 \n" |
| 89 | "itt cs \n" |
| 90 | "cmpcs %7, %4 \n" |
| 91 | L("ldrcsh %2, [%4], #2 \n") |
| 92 | U("ldrhcs %2, [%4], #2 \n") |
| 93 | "rsb %0, %6, #256 \n" |
| 94 | "smlabb %0, %5, %6, %0 \n" |
| 95 | T("itttt cs \n") |
| 96 | "rev16cs %2, %2 \n" |
| 97 | T("lslcs %2, %2, %3 \n") |
| 98 | T("orrcs %1, %1, %2 \n") |
| 99 | A("orrcs %1, %1, %2, lsl %3 \n") |
| 100 | "subcs %3, %3, #16 \n" |
| 101 | "lsr %0, %0, #8 \n" |
| 102 | "lsl %2, %0, #16 \n" |
| 103 | : "=&r"(low), "+&r"(code_word), "=&r"(tmp), |
| 104 | "+&r"(c->bits), "+&r"(c->buffer) |
| 105 | : "r"(high), "r"(pr), "r"(c->end - 1), "0"(shift) |
| 106 | : "cc"); |
| 107 | |
| 108 | if (code_word >= tmp) { |
| 109 | c->high = high - low; |
| 110 | c->code_word = code_word - tmp; |
| 111 | return 1; |
| 112 | } |
| 113 | |
| 114 | c->high = low; |
| 115 | c->code_word = code_word; |
| 116 | return 0; |
| 117 | } |
| 118 | |
| 119 | #endif |
| 120 | |
| 121 | #endif /* AVCODEC_ARM_VP56_ARITH_H */ |