kc3-lang/SDL/src/video/yuv2rgb/yuv_rgb_lsx_func.h
-
Show log
Commit
-
Hash : 50d8642f
Author :
Date : 2022-06-06T20:18:38
Add optimiztion function with LSX in LoongArch 1. yuv420_rgb24_lsx 1. yuv420_rgba_lsx 2. yuv420_bgra_lsx 3. yuv420_argb_lsx 4. yuv420_abgr_lsx
-
src/video/yuv2rgb/yuv_rgb_lsx_func.h
-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372
// Copyright 2016 Adrien Descamps // // Distributed under BSD 3-Clause License #include <lsxintrin.h> #if YUV_FORMAT == YUV_FORMAT_420 #define READ_Y(y_ptr) \ y = __lsx_vld(y_ptr, 0); \ #define READ_UV \ u_temp = __lsx_vld(u_ptr, 0); \ v_temp = __lsx_vld(v_ptr, 0); \ #else #error READ_UV unimplemented #endif #define PACK_RGBA_32(R1, R2, G1, G2, B1, B2, A1, A2, RGB1, RGB2, \ RGB3, RGB4, RGB5, RGB6, RGB7, RGB8) \ { \ __m128i ab_l, ab_h, gr_l, gr_h; \ ab_l = __lsx_vilvl_b(B1, A1); \ ab_h = __lsx_vilvh_b(B1, A1); \ gr_l = __lsx_vilvl_b(R1, G1); \ gr_h = __lsx_vilvh_b(R1, G1); \ RGB1 = __lsx_vilvl_h(gr_l, ab_l); \ RGB2 = __lsx_vilvh_h(gr_l, ab_l); \ RGB3 = __lsx_vilvl_h(gr_h, ab_h); \ RGB4 = __lsx_vilvh_h(gr_h, ab_h); \ ab_l = __lsx_vilvl_b(B2, A2); \ ab_h = __lsx_vilvh_b(B2, A2); \ gr_l = __lsx_vilvl_b(R2, G2); \ gr_h = __lsx_vilvh_b(R2, G2); \ RGB5 = __lsx_vilvl_h(gr_l, ab_l); \ RGB6 = __lsx_vilvh_h(gr_l, ab_l); \ RGB7 = __lsx_vilvl_h(gr_h, ab_h); \ RGB8 = __lsx_vilvh_h(gr_h, ab_h); \ } #define PACK_RGB24_32_STEP(R, G, B, RGB1, RGB2, RGB3) \ RGB1 = __lsx_vilvl_b(G, R); \ RGB1 = __lsx_vshuf_b(B, RGB1, mask1); \ RGB2 = __lsx_vshuf_b(B, G, mask2); \ RGB2 = __lsx_vshuf_b(R, RGB2, mask3); \ RGB3 = __lsx_vshuf_b(R, B, mask4); \ RGB3 = __lsx_vshuf_b(G, RGB3, mask5); \ #define PACK_RGB24_32(R1, R2, G1, G2, B1, B2, RGB1, RGB2, RGB3, RGB4, RGB5, RGB6) \ PACK_RGB24_32_STEP(R1, G1, B1, RGB1, RGB2, RGB3); \ PACK_RGB24_32_STEP(R2, G2, B2, RGB4, RGB5, RGB6); \ #if RGB_FORMAT == RGB_FORMAT_RGB24 #define PACK_PIXEL \ __m128i rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6; \ __m128i rgb_7, rgb_8, rgb_9, rgb_10, rgb_11, rgb_12; \ PACK_RGB24_32(r_8_11, r_8_12, g_8_11, g_8_12, b_8_11, b_8_12, \ rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6) \ PACK_RGB24_32(r_8_21, r_8_22, g_8_21, g_8_22, b_8_21, b_8_22, \ rgb_7, rgb_8, rgb_9, rgb_10, rgb_11, rgb_12) \ #elif RGB_FORMAT == RGB_FORMAT_RGBA #define PACK_PIXEL \ __m128i rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8; \ __m128i rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16; \ __m128i a = __lsx_vldi(0xFF); \ PACK_RGBA_32(r_8_11, r_8_12, g_8_11, g_8_12, b_8_11, b_8_12, a, a, \ rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8) \ PACK_RGBA_32(r_8_21, r_8_22, g_8_21, g_8_22, b_8_21, b_8_22, a, a, \ rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16) \ #elif RGB_FORMAT == RGB_FORMAT_BGRA #define PACK_PIXEL \ __m128i rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8; \ __m128i rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16; \ __m128i a = __lsx_vldi(0xFF); \ PACK_RGBA_32(b_8_11, b_8_12, g_8_11, g_8_12, r_8_11, r_8_12, a, a, \ rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8) \ PACK_RGBA_32(b_8_21, b_8_22, g_8_21, g_8_22, r_8_21, r_8_22, a, a, \ rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16) \ #elif RGB_FORMAT == RGB_FORMAT_ARGB #define PACK_PIXEL \ __m128i rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8; \ __m128i rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16; \ __m128i a = __lsx_vldi(0xFF); \ PACK_RGBA_32(a, a, r_8_11, r_8_12, g_8_11, g_8_12, b_8_11, b_8_12, \ rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8) \ PACK_RGBA_32(a, a, r_8_21, r_8_22, g_8_21, g_8_22, b_8_21, b_8_22, \ rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16) \ #elif RGB_FORMAT == RGB_FORMAT_ABGR #define PACK_PIXEL \ __m128i rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8; \ __m128i rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16; \ __m128i a = __lsx_vldi(0xFF); \ PACK_RGBA_32(a, a, b_8_11, b_8_12, g_8_11, g_8_12, r_8_11, r_8_12, \ rgb_1, rgb_2, rgb_3, rgb_4, rgb_5, rgb_6, rgb_7, rgb_8) \ PACK_RGBA_32(a, a, b_8_21, b_8_22, g_8_21, g_8_22, r_8_21, r_8_22, \ rgb_9, rgb_10, rgb_11, rgb_12, rgb_13, rgb_14, rgb_15, rgb_16) \ #else #error PACK_PIXEL unimplemented #endif #define LSX_ST_UB2(in0, in1, pdst, stride) \ { \ __lsx_vst(in0, pdst, 0); \ __lsx_vst(in1, pdst + stride, 0); \ } #if RGB_FORMAT == RGB_FORMAT_RGB24 \ #define SAVE_LINE1 \ LSX_ST_UB2(rgb_1, rgb_2, rgb_ptr1, 16); \ LSX_ST_UB2(rgb_3, rgb_4, rgb_ptr1 + 32, 16); \ LSX_ST_UB2(rgb_5, rgb_6, rgb_ptr1 + 64, 16); \ #define SAVE_LINE2 \ LSX_ST_UB2(rgb_7, rgb_8, rgb_ptr2, 16); \ LSX_ST_UB2(rgb_9, rgb_10, rgb_ptr2 + 32, 16); \ LSX_ST_UB2(rgb_11, rgb_12, rgb_ptr2 + 64, 16); \ #elif RGB_FORMAT == RGB_FORMAT_RGBA || RGB_FORMAT == RGB_FORMAT_BGRA || \ RGB_FORMAT == RGB_FORMAT_ARGB || RGB_FORMAT == RGB_FORMAT_ABGR \ #define SAVE_LINE1 \ LSX_ST_UB2(rgb_1, rgb_2, rgb_ptr1, 16); \ LSX_ST_UB2(rgb_3, rgb_4, rgb_ptr1 + 32, 16); \ LSX_ST_UB2(rgb_5, rgb_6, rgb_ptr1 + 64, 16); \ LSX_ST_UB2(rgb_7, rgb_8, rgb_ptr1 + 96, 16); \ #define SAVE_LINE2 \ LSX_ST_UB2(rgb_9, rgb_10, rgb_ptr2, 16); \ LSX_ST_UB2(rgb_11, rgb_12, rgb_ptr2 + 32, 16); \ LSX_ST_UB2(rgb_13, rgb_14, rgb_ptr2 + 64, 16); \ LSX_ST_UB2(rgb_15, rgb_16, rgb_ptr2 + 96, 16); \ #else #error SAVE_LINE unimplemented #endif // = u*vr g=u*ug+v*vg b=u*ub #define UV2RGB_16(U, V, R1, G1, B1, R2, G2, B2) \ r_temp = __lsx_vmul_h(V, v2r); \ g_temp = __lsx_vmul_h(U, u2g); \ g_temp = __lsx_vmadd_h(g_temp, V, v2g); \ b_temp = __lsx_vmul_h(U, u2b); \ R1 = __lsx_vilvl_h(r_temp, r_temp); \ G1 = __lsx_vilvl_h(g_temp, g_temp); \ B1 = __lsx_vilvl_h(b_temp, b_temp); \ R2 = __lsx_vilvh_h(r_temp, r_temp); \ G2 = __lsx_vilvh_h(g_temp, g_temp); \ B2 = __lsx_vilvh_h(b_temp, b_temp); \ // Y=(Y-shift)*shift R=(Y+R)>>6,G=(Y+G)>>6,B=(B+Y)>>6 #define ADD_Y2RGB_16(Y1, Y2, R1, G1, B1, R2, G2, B2) \ Y1 = __lsx_vsub_h(Y1, shift); \ Y2 = __lsx_vsub_h(Y2, shift); \ Y1 = __lsx_vmul_h(Y1, yf); \ Y2 = __lsx_vmul_h(Y2, yf); \ R1 = __lsx_vadd_h(R1, Y1); \ G1 = __lsx_vadd_h(G1, Y1); \ B1 = __lsx_vadd_h(B1, Y1); \ R2 = __lsx_vadd_h(R2, Y2); \ G2 = __lsx_vadd_h(G2, Y2); \ B2 = __lsx_vadd_h(B2, Y2); \ R1 = __lsx_vsrai_h(R1, PRECISION); \ G1 = __lsx_vsrai_h(G1, PRECISION); \ B1 = __lsx_vsrai_h(B1, PRECISION); \ R2 = __lsx_vsrai_h(R2, PRECISION); \ G2 = __lsx_vsrai_h(G2, PRECISION); \ B2 = __lsx_vsrai_h(B2, PRECISION); \ #define CLIP(in0, in1, in2, in3, in4, in5) \ { \ in0 = __lsx_vmaxi_h(in0, 0); \ in1 = __lsx_vmaxi_h(in1, 0); \ in2 = __lsx_vmaxi_h(in2, 0); \ in3 = __lsx_vmaxi_h(in3, 0); \ in4 = __lsx_vmaxi_h(in4, 0); \ in5 = __lsx_vmaxi_h(in5, 0); \ in0 = __lsx_vsat_hu(in0, 7); \ in1 = __lsx_vsat_hu(in1, 7); \ in2 = __lsx_vsat_hu(in2, 7); \ in3 = __lsx_vsat_hu(in3, 7); \ in4 = __lsx_vsat_hu(in4, 7); \ in5 = __lsx_vsat_hu(in5, 7); \ } #define YUV2RGB_32 \ __m128i y, u_temp, v_temp; \ __m128i r_8_11, g_8_11, b_8_11, r_8_21, g_8_21, b_8_21; \ __m128i r_8_12, g_8_12, b_8_12, r_8_22, g_8_22, b_8_22; \ __m128i u, v, r_temp, g_temp, b_temp; \ __m128i r_1, g_1, b_1, r_2, g_2, b_2; \ __m128i y_1, y_2; \ __m128i r_uv_1, g_uv_1, b_uv_1, r_uv_2, g_uv_2, b_uv_2; \ \ READ_UV \ \ /* process first 16 pixels of first line */ \ u = __lsx_vilvl_b(zero, u_temp); \ v = __lsx_vilvl_b(zero, v_temp); \ u = __lsx_vsub_h(u, bias); \ v = __lsx_vsub_h(v, bias); \ UV2RGB_16(u, v, r_1, g_1, b_1, r_2, g_2, b_2); \ r_uv_1 = r_1; g_uv_1 = g_1; b_uv_1 = b_1; \ r_uv_2 = r_2; g_uv_2 = g_2; b_uv_2 = b_2; \ READ_Y(y_ptr1) \ y_1 = __lsx_vilvl_b(zero, y); \ y_2 = __lsx_vilvh_b(zero, y); \ ADD_Y2RGB_16(y_1, y_2, r_1, g_1, b_1, r_2, g_2, b_2) \ CLIP(r_1, g_1, b_1, r_2, g_2, b_2); \ r_8_11 = __lsx_vpickev_b(r_2, r_1); \ g_8_11 = __lsx_vpickev_b(g_2, g_1); \ b_8_11 = __lsx_vpickev_b(b_2, b_1); \ \ /* process first 16 pixels of second line */ \ r_1 = r_uv_1; g_1 = g_uv_1; b_1 = b_uv_1; \ r_2 = r_uv_2; g_2 = g_uv_2; b_2 = b_uv_2; \ \ READ_Y(y_ptr2) \ y_1 = __lsx_vilvl_b(zero, y); \ y_2 = __lsx_vilvh_b(zero, y); \ ADD_Y2RGB_16(y_1, y_2, r_1, g_1, b_1, r_2, g_2, b_2) \ CLIP(r_1, g_1, b_1, r_2, g_2, b_2); \ r_8_21 = __lsx_vpickev_b(r_2, r_1); \ g_8_21 = __lsx_vpickev_b(g_2, g_1); \ b_8_21 = __lsx_vpickev_b(b_2, b_1); \ \ /* process last 16 pixels of first line */ \ u = __lsx_vilvh_b(zero, u_temp); \ v = __lsx_vilvh_b(zero, v_temp); \ u = __lsx_vsub_h(u, bias); \ v = __lsx_vsub_h(v, bias); \ UV2RGB_16(u, v, r_1, g_1, b_1, r_2, g_2, b_2); \ r_uv_1 = r_1; g_uv_1 = g_1; b_uv_1 = b_1; \ r_uv_2 = r_2; g_uv_2 = g_2; b_uv_2 = b_2; \ READ_Y(y_ptr1 + 16 * y_pixel_stride) \ y_1 = __lsx_vilvl_b(zero, y); \ y_2 = __lsx_vilvh_b(zero, y); \ ADD_Y2RGB_16(y_1, y_2, r_1, g_1, b_1, r_2, g_2, b_2) \ CLIP(r_1, g_1, b_1, r_2, g_2, b_2); \ r_8_12 = __lsx_vpickev_b(r_2, r_1); \ g_8_12 = __lsx_vpickev_b(g_2, g_1); \ b_8_12 = __lsx_vpickev_b(b_2, b_1); \ \ /* process last 16 pixels of second line */ \ r_1 = r_uv_1; g_1 = g_uv_1; b_1 = b_uv_1; \ r_2 = r_uv_2; g_2 = g_uv_2; b_2 = b_uv_2; \ \ READ_Y(y_ptr2 + 16 * y_pixel_stride) \ y_1 = __lsx_vilvl_b(zero, y); \ y_2 = __lsx_vilvh_b(zero, y); \ ADD_Y2RGB_16(y_1, y_2, r_1, g_1, b_1, r_2, g_2, b_2) \ CLIP(r_1, g_1, b_1, r_2, g_2, b_2); \ r_8_22 = __lsx_vpickev_b(r_2, r_1); \ g_8_22 = __lsx_vpickev_b(g_2, g_1); \ b_8_22 = __lsx_vpickev_b(b_2, b_1); \ \ void LSX_FUNCTION_NAME(uint32_t width, uint32_t height, const uint8_t *Y, const uint8_t *U, const uint8_t *V, uint32_t Y_stride, uint32_t UV_stride, uint8_t *RGB, uint32_t RGB_stride, YCbCrType yuv_type) { const YUV2RGBParam *const param = &(YUV2RGB[yuv_type]); #if YUV_FORMAT == YUV_FORMAT_420 const int y_pixel_stride = 1; const int uv_pixel_stride = 1; const int uv_x_sample_interval = 2; const int uv_y_sample_interval = 2; #endif #if RGB_FORMAT == RGB_FORMAT_RGB565 const int rgb_pixel_stride = 2; #elif RGB_FORMAT == RGB_FORMAT_RGB24 const int rgb_pixel_stride = 3; __m128i mask1 = {0x0504110302100100, 0x0A14090813070612}; __m128i mask2 = {0x1808170716061505, 0x00000000000A1909}; __m128i mask3 = {0x0504170302160100, 0x0A1A090819070618}; __m128i mask4 = {0x1E0D1D0C1C0B1B0A, 0x00000000000F1F0E}; __m128i mask5 = {0x05041C03021B0100, 0x0A1F09081E07061D}; #elif RGB_FORMAT == RGB_FORMAT_RGBA || RGB_FORMAT_BGRA || \ RGB_FORMAT == RGB_FORMAT_ARGB || RGB_FORMAT_ABGR const int rgb_pixel_stride = 4; #else #error Unknown RGB pixel size #endif uint32_t xpos, ypos; __m128i v2r = __lsx_vreplgr2vr_h(param->v_r_factor); __m128i v2g = __lsx_vreplgr2vr_h(param->v_g_factor); __m128i u2g = __lsx_vreplgr2vr_h(param->u_g_factor); __m128i u2b = __lsx_vreplgr2vr_h(param->u_b_factor); __m128i bias = __lsx_vreplgr2vr_h(128); __m128i shift = __lsx_vreplgr2vr_h(param->y_shift); __m128i yf = __lsx_vreplgr2vr_h(param->y_factor); __m128i zero = __lsx_vldi(0); if (width >= 32) { for (ypos = 0; ypos < (height - (uv_y_sample_interval - 1)); ypos += uv_y_sample_interval) { const uint8_t *y_ptr1 = Y + ypos * Y_stride, *y_ptr2 = Y + (ypos + 1) * Y_stride, *u_ptr = U + (ypos/uv_y_sample_interval) * UV_stride, *v_ptr = V + (ypos/uv_y_sample_interval) * UV_stride; uint8_t *rgb_ptr1 = RGB + ypos * RGB_stride, *rgb_ptr2 = RGB + (ypos + 1) * RGB_stride; for (xpos = 0; xpos < (width - 31); xpos += 32){ YUV2RGB_32 { PACK_PIXEL SAVE_LINE1 if (uv_y_sample_interval > 1) { SAVE_LINE2 } } y_ptr1 += 32 * y_pixel_stride; y_ptr2 += 32 * y_pixel_stride; u_ptr += 32 * uv_pixel_stride/uv_x_sample_interval; v_ptr += 32 * uv_pixel_stride/uv_x_sample_interval; rgb_ptr1 += 32 * rgb_pixel_stride; rgb_ptr2 += 32 * rgb_pixel_stride; } } if (uv_y_sample_interval == 2 && ypos == (height - 1)) { const uint8_t *y_ptr = Y + ypos * Y_stride, *u_ptr = U + (ypos/uv_y_sample_interval) * UV_stride, *v_ptr = V + (ypos/uv_y_sample_interval) * UV_stride; uint8_t *rgb_ptr = RGB + ypos * RGB_stride; STD_FUNCTION_NAME(width, 1, y_ptr, u_ptr, v_ptr, Y_stride, UV_stride, rgb_ptr, RGB_stride, yuv_type); } } { int converted = (width & ~31); if (converted != width) { const uint8_t *y_ptr = Y + converted * y_pixel_stride, *u_ptr = U + converted * uv_pixel_stride / uv_x_sample_interval, *v_ptr = V + converted * uv_pixel_stride / uv_x_sample_interval; uint8_t *rgb_ptr = RGB + converted * rgb_pixel_stride; STD_FUNCTION_NAME(width-converted, height, y_ptr, u_ptr, v_ptr, Y_stride, UV_stride, rgb_ptr, RGB_stride, yuv_type); } } } #undef LSX_FUNCTION_NAME #undef STD_FUNCTION_NAME #undef YUV_FORMAT #undef RGB_FORMAT #undef LSX_ALIGNED #undef LSX_ST_UB2 #undef UV2RGB_16 #undef ADD_Y2RGB_16 #undef PACK_RGB24_32_STEP #undef PACK_RGB24_32 #undef PACK_PIXEL #undef PACK_RGBA_32 #undef SAVE_LINE1 #undef SAVE_LINE2 #undef READ_Y #undef READ_UV #undef YUV2RGB_32