src/mesa/main/macros.h

   1 /**
   2  * \file macros.h
   3  * A collection of useful macros.
   4  */
   5
   6 /*
   7  * Mesa 3-D graphics library
   8  *
   9  * Copyright (C) 1999-2006  Brian Paul   All Rights Reserved.
  10  *
  11  * Permission is hereby granted, free of charge, to any person obtaining a
  12  * copy of this software and associated documentation files (the "Software"),
  13  * to deal in the Software without restriction, including without limitation
  14  * the rights to use, copy, modify, merge, publish, distribute, sublicense,
  15  * and/or sell copies of the Software, and to permit persons to whom the
  16  * Software is furnished to do so, subject to the following conditions:
  17  *
  18  * The above copyright notice and this permission notice shall be included
  19  * in all copies or substantial portions of the Software.
  20  *
  21  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
  22  * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  23  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
  24  * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
  25  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  26  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  27  * OTHER DEALINGS IN THE SOFTWARE.
  28  */
  29
  30
  31 #ifndef MACROS_H
  32 #define MACROS_H
  33
  34 #include "util/macros.h"
  35 #include "util/u_math.h"
  36 #include "util/rounding.h"
  37 #include "imports.h"
  38
  39
  40 /**
  41  * \name Integer / float conversion for colors, normals, etc.
  42  */
  43 /*@{*/
  44
  45 /** Convert GLubyte in [0,255] to GLfloat in [0.0,1.0] */
  46 extern GLfloat _mesa_ubyte_to_float_color_tab[256];
  47 #define UBYTE_TO_FLOAT(u) _mesa_ubyte_to_float_color_tab[(unsigned int)(u)]
  48
  49 /** Convert GLfloat in [0.0,1.0] to GLubyte in [0,255] */
  50 #define FLOAT_TO_UBYTE(X)   ((GLubyte) (GLint) ((X) * 255.0F))
  51
  52
  53 /** Convert GLbyte in [-128,127] to GLfloat in [-1.0,1.0] */
  54 #define BYTE_TO_FLOAT(B)    ((2.0F * (B) + 1.0F) * (1.0F/255.0F))
  55
  56 /** Convert GLfloat in [-1.0,1.0] to GLbyte in [-128,127] */
  57 #define FLOAT_TO_BYTE(X)    ( (((GLint) (255.0F * (X))) - 1) / 2 )
  58
  59
  60 /** Convert GLbyte to GLfloat while preserving zero */
  61 #define BYTE_TO_FLOATZ(B)   ((B) == 0 ? 0.0F : BYTE_TO_FLOAT(B))
  62
  63
  64 /** Convert GLbyte in [-128,127] to GLfloat in [-1.0,1.0], texture/fb data */
  65 #define BYTE_TO_FLOAT_TEX(B)    ((B) == -128 ? -1.0F : (B) * (1.0F/127.0F))
  66
  67 /** Convert GLfloat in [-1.0,1.0] to GLbyte in [-128,127], texture/fb data */
  68 #define FLOAT_TO_BYTE_TEX(X)    CLAMP( (GLint) (127.0F * (X)), -128, 127 )
  69
  70 /** Convert GLushort in [0,65535] to GLfloat in [0.0,1.0] */
  71 #define USHORT_TO_FLOAT(S)  ((GLfloat) (S) * (1.0F / 65535.0F))
  72
  73 /** Convert GLfloat in [0.0,1.0] to GLushort in [0, 65535] */
  74 #define FLOAT_TO_USHORT(X)   ((GLuint) ((X) * 65535.0F))
  75
  76
  77 /** Convert GLshort in [-32768,32767] to GLfloat in [-1.0,1.0] */
  78 #define SHORT_TO_FLOAT(S)   ((2.0F * (S) + 1.0F) * (1.0F/65535.0F))
  79
  80 /** Convert GLfloat in [-1.0,1.0] to GLshort in [-32768,32767] */
  81 #define FLOAT_TO_SHORT(X)   ( (((GLint) (65535.0F * (X))) - 1) / 2 )
  82
  83 /** Convert GLshort to GLfloat while preserving zero */
  84 #define SHORT_TO_FLOATZ(S)   ((S) == 0 ? 0.0F : SHORT_TO_FLOAT(S))
  85
  86
  87 /** Convert GLshort in [-32768,32767] to GLfloat in [-1.0,1.0], texture/fb data */
  88 #define SHORT_TO_FLOAT_TEX(S)    ((S) == -32768 ? -1.0F : (S) * (1.0F/32767.0F))
  89
  90 /** Convert GLfloat in [-1.0,1.0] to GLshort in [-32768,32767], texture/fb data */
  91 #define FLOAT_TO_SHORT_TEX(X)    ( (GLint) (32767.0F * (X)) )
  92
  93
  94 /** Convert GLuint in [0,4294967295] to GLfloat in [0.0,1.0] */
  95 #define UINT_TO_FLOAT(U)    ((GLfloat) ((U) * (1.0F / 4294967295.0)))
  96
  97 /** Convert GLfloat in [0.0,1.0] to GLuint in [0,4294967295] */
  98 #define FLOAT_TO_UINT(X)    ((GLuint) ((X) * 4294967295.0))
  99
 100
 101 /** Convert GLint in [-2147483648,2147483647] to GLfloat in [-1.0,1.0] */
 102 #define INT_TO_FLOAT(I)     ((GLfloat) ((2.0F * (I) + 1.0F) * (1.0F/4294967294.0)))
 103
 104 /** Convert GLfloat in [-1.0,1.0] to GLint in [-2147483648,2147483647] */
 105 /* causes overflow:
 106 #define FLOAT_TO_INT(X)     ( (((GLint) (4294967294.0 * (X))) - 1) / 2 )
 107 */
 108 /* a close approximation: */
 109 #define FLOAT_TO_INT(X)     ( (GLint) (2147483647.0 * (X)) )
 110
 111 /** Convert GLfloat in [-1.0,1.0] to GLint64 in [-(1<<63),(1 << 63) -1] */
 112 #define FLOAT_TO_INT64(X)     ( (GLint64) (9223372036854775807.0 * (double)(X)) )
 113
 114
 115 /** Convert GLint in [-2147483648,2147483647] to GLfloat in [-1.0,1.0], texture/fb data */
 116 #define INT_TO_FLOAT_TEX(I)    ((I) == -2147483648 ? -1.0F : (I) * (1.0F/2147483647.0))
 117
 118 /** Convert GLfloat in [-1.0,1.0] to GLint in [-2147483648,2147483647], texture/fb data */
 119 #define FLOAT_TO_INT_TEX(X)    ( (GLint) (2147483647.0 * (X)) )
 120
 121
 122 #define BYTE_TO_UBYTE(b)   ((GLubyte) ((b) < 0 ? 0 : (GLubyte) (b)))
 123 #define SHORT_TO_UBYTE(s)  ((GLubyte) ((s) < 0 ? 0 : (GLubyte) ((s) >> 7)))
 124 #define USHORT_TO_UBYTE(s) ((GLubyte) ((s) >> 8))
 125 #define INT_TO_UBYTE(i)    ((GLubyte) ((i) < 0 ? 0 : (GLubyte) ((i) >> 23)))
 126 #define UINT_TO_UBYTE(i)   ((GLubyte) ((i) >> 24))
 127
 128
 129 #define BYTE_TO_USHORT(b)  ((b) < 0 ? 0 : ((GLushort) (((b) * 65535) / 255)))
 130 #define UBYTE_TO_USHORT(b) (((GLushort) (b) << 8) | (GLushort) (b))
 131 #define SHORT_TO_USHORT(s) ((s) < 0 ? 0 : ((GLushort) (((s) * 65535 / 32767))))
 132 #define INT_TO_USHORT(i)   ((i) < 0 ? 0 : ((GLushort) ((i) >> 15)))
 133 #define UINT_TO_USHORT(i)  ((i) < 0 ? 0 : ((GLushort) ((i) >> 16)))
 134 #define UNCLAMPED_FLOAT_TO_USHORT(us, f)  \
 135         us = ( (GLushort) _mesa_lroundevenf( CLAMP((f), 0.0F, 1.0F) * 65535.0F) )
 136 #define CLAMPED_FLOAT_TO_USHORT(us, f)  \
 137         us = ( (GLushort) _mesa_lroundevenf( (f) * 65535.0F) )
 138
 139 #define UNCLAMPED_FLOAT_TO_SHORT(s, f)  \
 140         s = ( (GLshort) _mesa_lroundevenf( CLAMP((f), -1.0F, 1.0F) * 32767.0F) )
 141
 142 /***
 143  *** UNCLAMPED_FLOAT_TO_UBYTE: clamp float to [0,1] and map to ubyte in [0,255]
 144  *** CLAMPED_FLOAT_TO_UBYTE: map float known to be in [0,1] to ubyte in [0,255]
 145  ***/
 146 #ifndef DEBUG
 147 /* This function/macro is sensitive to precision.  Test very carefully
 148  * if you change it!
 149  */
 150 #define UNCLAMPED_FLOAT_TO_UBYTE(UB, FLT)                               \
 151         do {                                                            \
 152            fi_type __tmp;                                               \
 153            __tmp.f = (FLT);                                             \
 154            if (__tmp.i < 0)                                             \
 155               UB = (GLubyte) 0;                                         \
 156            else if (__tmp.i >= IEEE_ONE)                                \
 157               UB = (GLubyte) 255;                                       \
 158            else {                                                       \
 159               __tmp.f = __tmp.f * (255.0F/256.0F) + 32768.0F;           \
 160               UB = (GLubyte) __tmp.i;                                   \
 161            }                                                            \
 162         } while (0)
 163 #define CLAMPED_FLOAT_TO_UBYTE(UB, FLT)                                 \
 164         do {                                                            \
 165            fi_type __tmp;                                               \
 166            __tmp.f = (FLT) * (255.0F/256.0F) + 32768.0F;                \
 167            UB = (GLubyte) __tmp.i;                                      \
 168         } while (0)
 169 #else
 170 #define UNCLAMPED_FLOAT_TO_UBYTE(ub, f) \
 171         ub = ((GLubyte) _mesa_lroundevenf(CLAMP((f), 0.0F, 1.0F) * 255.0F))
 172 #define CLAMPED_FLOAT_TO_UBYTE(ub, f) \
 173         ub = ((GLubyte) _mesa_lroundevenf((f) * 255.0F))
 174 #endif
 175
 176 static fi_type UINT_AS_UNION(GLuint u)
 177 {
 178    fi_type tmp;
 179    tmp.u = u;
 180    return tmp;
 181 }
 182
 183 static inline fi_type INT_AS_UNION(GLint i)
 184 {
 185    fi_type tmp;
 186    tmp.i = i;
 187    return tmp;
 188 }
 189
 190 static inline fi_type FLOAT_AS_UNION(GLfloat f)
 191 {
 192    fi_type tmp;
 193    tmp.f = f;
 194    return tmp;
 195 }
 196
 197 static inline uint64_t DOUBLE_AS_UINT64(double d)
 198 {
 199    union {
 200       double d;
 201       uint64_t u64;
 202    } tmp;
 203    tmp.d = d;
 204    return tmp.u64;
 205 }
 206
 207 static inline double UINT64_AS_DOUBLE(uint64_t u)
 208 {
 209    union {
 210       double d;
 211       uint64_t u64;
 212    } tmp;
 213    tmp.u64 = u;
 214    return tmp.d;
 215 }
 216
 217 /* First sign-extend x, then return uint32_t. */
 218 #define INT_AS_UINT(x) ((uint32_t)((int32_t)(x)))
 219 #define FLOAT_AS_UINT(x) (FLOAT_AS_UNION(x).u)
 220
 221 /**
 222  * Convert a floating point value to an unsigned fixed point value.
 223  *
 224  * \param frac_bits   The number of bits used to store the fractional part.
 225  */
 226 static inline uint32_t
 227 U_FIXED(float value, uint32_t frac_bits)
 228 {
 229    value *= (1 << frac_bits);
 230    return value < 0.0f ? 0 : (uint32_t) value;
 231 }
 232
 233 /**
 234  * Convert a floating point value to an signed fixed point value.
 235  *
 236  * \param frac_bits   The number of bits used to store the fractional part.
 237  */
 238 static inline int32_t
 239 S_FIXED(float value, uint32_t frac_bits)
 240 {
 241    return (int32_t) (value * (1 << frac_bits));
 242 }
 243 /*@}*/
 244
 245
 246 /** Stepping a GLfloat pointer by a byte stride */
 247 #define STRIDE_F(p, i)  (p = (GLfloat *)((GLubyte *)p + i))
 248 /** Stepping a GLuint pointer by a byte stride */
 249 #define STRIDE_UI(p, i)  (p = (GLuint *)((GLubyte *)p + i))
 250 /** Stepping a GLubyte[4] pointer by a byte stride */
 251 #define STRIDE_4UB(p, i)  (p = (GLubyte (*)[4])((GLubyte *)p + i))
 252 /** Stepping a GLfloat[4] pointer by a byte stride */
 253 #define STRIDE_4F(p, i)  (p = (GLfloat (*)[4])((GLubyte *)p + i))
 254 /** Stepping a \p t pointer by a byte stride */
 255 #define STRIDE_T(p, t, i)  (p = (t)((GLubyte *)p + i))
 256
 257
 258 /**********************************************************************/
 259 /** \name 4-element vector operations */
 260 /*@{*/
 261
 262 /** Zero */
 263 #define ZERO_4V( DST )  (DST)[0] = (DST)[1] = (DST)[2] = (DST)[3] = 0
 264
 265 /** Test for equality */
 266 #define TEST_EQ_4V(a,b)  ((a)[0] == (b)[0] &&   \
 267               (a)[1] == (b)[1] &&   \
 268               (a)[2] == (b)[2] &&   \
 269               (a)[3] == (b)[3])
 270
 271 /** Test for equality (unsigned bytes) */
 272 static inline GLboolean
 273 TEST_EQ_4UBV(const GLubyte a[4], const GLubyte b[4])
 274 {
 275 #if defined(__i386__)
 276    return *((const GLuint *) a) == *((const GLuint *) b);
 277 #else
 278    return TEST_EQ_4V(a, b);
 279 #endif
 280 }
 281
 282
 283 /** Copy a 4-element vector */
 284 #define COPY_4V( DST, SRC )         \
 285 do {                                \
 286    (DST)[0] = (SRC)[0];             \
 287    (DST)[1] = (SRC)[1];             \
 288    (DST)[2] = (SRC)[2];             \
 289    (DST)[3] = (SRC)[3];             \
 290 } while (0)
 291
 292 /** Copy a 4-element unsigned byte vector */
 293 static inline void
 294 COPY_4UBV(GLubyte dst[4], const GLubyte src[4])
 295 {
 296 #if defined(__i386__)
 297    *((GLuint *) dst) = *((GLuint *) src);
 298 #else
 299    /* The GLuint cast might fail if DST or SRC are not dword-aligned (RISC) */
 300    COPY_4V(dst, src);
 301 #endif
 302 }
 303
 304 /** Copy \p SZ elements into a 4-element vector */
 305 #define COPY_SZ_4V(DST, SZ, SRC)  \
 306 do {                              \
 307    switch (SZ) {                  \
 308    case 4: (DST)[3] = (SRC)[3];   \
 309    case 3: (DST)[2] = (SRC)[2];   \
 310    case 2: (DST)[1] = (SRC)[1];   \
 311    case 1: (DST)[0] = (SRC)[0];   \
 312    }                              \
 313 } while(0)
 314
 315 /** Copy \p SZ elements into a homegeneous (4-element) vector, giving
 316  * default values to the remaining */
 317 #define COPY_CLEAN_4V(DST, SZ, SRC)  \
 318 do {                                 \
 319       ASSIGN_4V( DST, 0, 0, 0, 1 );  \
 320       COPY_SZ_4V( DST, SZ, SRC );    \
 321 } while (0)
 322
 323 /** Subtraction */
 324 #define SUB_4V( DST, SRCA, SRCB )           \
 325 do {                                        \
 326       (DST)[0] = (SRCA)[0] - (SRCB)[0];     \
 327       (DST)[1] = (SRCA)[1] - (SRCB)[1];     \
 328       (DST)[2] = (SRCA)[2] - (SRCB)[2];     \
 329       (DST)[3] = (SRCA)[3] - (SRCB)[3];     \
 330 } while (0)
 331
 332 /** Addition */
 333 #define ADD_4V( DST, SRCA, SRCB )           \
 334 do {                                        \
 335       (DST)[0] = (SRCA)[0] + (SRCB)[0];     \
 336       (DST)[1] = (SRCA)[1] + (SRCB)[1];     \
 337       (DST)[2] = (SRCA)[2] + (SRCB)[2];     \
 338       (DST)[3] = (SRCA)[3] + (SRCB)[3];     \
 339 } while (0)
 340
 341 /** Element-wise multiplication */
 342 #define SCALE_4V( DST, SRCA, SRCB )         \
 343 do {                                        \
 344       (DST)[0] = (SRCA)[0] * (SRCB)[0];     \
 345       (DST)[1] = (SRCA)[1] * (SRCB)[1];     \
 346       (DST)[2] = (SRCA)[2] * (SRCB)[2];     \
 347       (DST)[3] = (SRCA)[3] * (SRCB)[3];     \
 348 } while (0)
 349
 350 /** In-place addition */
 351 #define ACC_4V( DST, SRC )          \
 352 do {                                \
 353       (DST)[0] += (SRC)[0];         \
 354       (DST)[1] += (SRC)[1];         \
 355       (DST)[2] += (SRC)[2];         \
 356       (DST)[3] += (SRC)[3];         \
 357 } while (0)
 358
 359 /** Element-wise multiplication and addition */
 360 #define ACC_SCALE_4V( DST, SRCA, SRCB )     \
 361 do {                                        \
 362       (DST)[0] += (SRCA)[0] * (SRCB)[0];    \
 363       (DST)[1] += (SRCA)[1] * (SRCB)[1];    \
 364       (DST)[2] += (SRCA)[2] * (SRCB)[2];    \
 365       (DST)[3] += (SRCA)[3] * (SRCB)[3];    \
 366 } while (0)
 367
 368 /** In-place scalar multiplication and addition */
 369 #define ACC_SCALE_SCALAR_4V( DST, S, SRCB ) \
 370 do {                                        \
 371       (DST)[0] += S * (SRCB)[0];            \
 372       (DST)[1] += S * (SRCB)[1];            \
 373       (DST)[2] += S * (SRCB)[2];            \
 374       (DST)[3] += S * (SRCB)[3];            \
 375 } while (0)
 376
 377 /** Scalar multiplication */
 378 #define SCALE_SCALAR_4V( DST, S, SRCB ) \
 379 do {                                    \
 380       (DST)[0] = S * (SRCB)[0];         \
 381       (DST)[1] = S * (SRCB)[1];         \
 382       (DST)[2] = S * (SRCB)[2];         \
 383       (DST)[3] = S * (SRCB)[3];         \
 384 } while (0)
 385
 386 /** In-place scalar multiplication */
 387 #define SELF_SCALE_SCALAR_4V( DST, S ) \
 388 do {                                   \
 389       (DST)[0] *= S;                   \
 390       (DST)[1] *= S;                   \
 391       (DST)[2] *= S;                   \
 392       (DST)[3] *= S;                   \
 393 } while (0)
 394
 395 /*@}*/
 396
 397
 398 /**********************************************************************/
 399 /** \name 3-element vector operations*/
 400 /*@{*/
 401
 402 /** Zero */
 403 #define ZERO_3V( DST )  (DST)[0] = (DST)[1] = (DST)[2] = 0
 404
 405 /** Test for equality */
 406 #define TEST_EQ_3V(a,b)  \
 407    ((a)[0] == (b)[0] &&  \
 408     (a)[1] == (b)[1] &&  \
 409     (a)[2] == (b)[2])
 410
 411 /** Copy a 3-element vector */
 412 #define COPY_3V( DST, SRC )         \
 413 do {                                \
 414    (DST)[0] = (SRC)[0];             \
 415    (DST)[1] = (SRC)[1];             \
 416    (DST)[2] = (SRC)[2];             \
 417 } while (0)
 418
 419 /** Copy a 3-element vector with cast */
 420 #define COPY_3V_CAST( DST, SRC, CAST )  \
 421 do {                                    \
 422    (DST)[0] = (CAST)(SRC)[0];           \
 423    (DST)[1] = (CAST)(SRC)[1];           \
 424    (DST)[2] = (CAST)(SRC)[2];           \
 425 } while (0)
 426
 427 /** Copy a 3-element float vector */
 428 #define COPY_3FV( DST, SRC )        \
 429 do {                                \
 430    const GLfloat *_tmp = (SRC);     \
 431    (DST)[0] = _tmp[0];              \
 432    (DST)[1] = _tmp[1];              \
 433    (DST)[2] = _tmp[2];              \
 434 } while (0)
 435
 436 /** Subtraction */
 437 #define SUB_3V( DST, SRCA, SRCB )        \
 438 do {                                     \
 439       (DST)[0] = (SRCA)[0] - (SRCB)[0];  \
 440       (DST)[1] = (SRCA)[1] - (SRCB)[1];  \
 441       (DST)[2] = (SRCA)[2] - (SRCB)[2];  \
 442 } while (0)
 443
 444 /** Addition */
 445 #define ADD_3V( DST, SRCA, SRCB )       \
 446 do {                                    \
 447       (DST)[0] = (SRCA)[0] + (SRCB)[0]; \
 448       (DST)[1] = (SRCA)[1] + (SRCB)[1]; \
 449       (DST)[2] = (SRCA)[2] + (SRCB)[2]; \
 450 } while (0)
 451
 452 /** In-place scalar multiplication */
 453 #define SCALE_3V( DST, SRCA, SRCB )     \
 454 do {                                    \
 455       (DST)[0] = (SRCA)[0] * (SRCB)[0]; \
 456       (DST)[1] = (SRCA)[1] * (SRCB)[1]; \
 457       (DST)[2] = (SRCA)[2] * (SRCB)[2]; \
 458 } while (0)
 459
 460 /** In-place element-wise multiplication */
 461 #define SELF_SCALE_3V( DST, SRC )   \
 462 do {                                \
 463       (DST)[0] *= (SRC)[0];         \
 464       (DST)[1] *= (SRC)[1];         \
 465       (DST)[2] *= (SRC)[2];         \
 466 } while (0)
 467
 468 /** In-place addition */
 469 #define ACC_3V( DST, SRC )          \
 470 do {                                \
 471       (DST)[0] += (SRC)[0];         \
 472       (DST)[1] += (SRC)[1];         \
 473       (DST)[2] += (SRC)[2];         \
 474 } while (0)
 475
 476 /** Element-wise multiplication and addition */
 477 #define ACC_SCALE_3V( DST, SRCA, SRCB )     \
 478 do {                                        \
 479       (DST)[0] += (SRCA)[0] * (SRCB)[0];    \
 480       (DST)[1] += (SRCA)[1] * (SRCB)[1];    \
 481       (DST)[2] += (SRCA)[2] * (SRCB)[2];    \
 482 } while (0)
 483
 484 /** Scalar multiplication */
 485 #define SCALE_SCALAR_3V( DST, S, SRCB ) \
 486 do {                                    \
 487       (DST)[0] = S * (SRCB)[0];         \
 488       (DST)[1] = S * (SRCB)[1];         \
 489       (DST)[2] = S * (SRCB)[2];         \
 490 } while (0)
 491
 492 /** In-place scalar multiplication and addition */
 493 #define ACC_SCALE_SCALAR_3V( DST, S, SRCB ) \
 494 do {                                        \
 495       (DST)[0] += S * (SRCB)[0];            \
 496       (DST)[1] += S * (SRCB)[1];            \
 497       (DST)[2] += S * (SRCB)[2];            \
 498 } while (0)
 499
 500 /** In-place scalar multiplication */
 501 #define SELF_SCALE_SCALAR_3V( DST, S ) \
 502 do {                                   \
 503       (DST)[0] *= S;                   \
 504       (DST)[1] *= S;                   \
 505       (DST)[2] *= S;                   \
 506 } while (0)
 507
 508 /** In-place scalar addition */
 509 #define ACC_SCALAR_3V( DST, S )     \
 510 do {                                \
 511       (DST)[0] += S;                \
 512       (DST)[1] += S;                \
 513       (DST)[2] += S;                \
 514 } while (0)
 515
 516 /** Assignment */
 517 #define ASSIGN_3V( V, V0, V1, V2 )  \
 518 do {                                \
 519     V[0] = V0;                      \
 520     V[1] = V1;                      \
 521     V[2] = V2;                      \
 522 } while(0)
 523
 524 /*@}*/
 525
 526
 527 /**********************************************************************/
 528 /** \name 2-element vector operations*/
 529 /*@{*/
 530
 531 /** Zero */
 532 #define ZERO_2V( DST )  (DST)[0] = (DST)[1] = 0
 533
 534 /** Copy a 2-element vector */
 535 #define COPY_2V( DST, SRC )         \
 536 do {                        \
 537    (DST)[0] = (SRC)[0];             \
 538    (DST)[1] = (SRC)[1];             \
 539 } while (0)
 540
 541 /** Copy a 2-element vector with cast */
 542 #define COPY_2V_CAST( DST, SRC, CAST )      \
 543 do {                        \
 544    (DST)[0] = (CAST)(SRC)[0];           \
 545    (DST)[1] = (CAST)(SRC)[1];           \
 546 } while (0)
 547
 548 /** Copy a 2-element float vector */
 549 #define COPY_2FV( DST, SRC )            \
 550 do {                        \
 551    const GLfloat *_tmp = (SRC);         \
 552    (DST)[0] = _tmp[0];              \
 553    (DST)[1] = _tmp[1];              \
 554 } while (0)
 555
 556 /** Subtraction */
 557 #define SUB_2V( DST, SRCA, SRCB )       \
 558 do {                        \
 559       (DST)[0] = (SRCA)[0] - (SRCB)[0];     \
 560       (DST)[1] = (SRCA)[1] - (SRCB)[1];     \
 561 } while (0)
 562
 563 /** Addition */
 564 #define ADD_2V( DST, SRCA, SRCB )       \
 565 do {                        \
 566       (DST)[0] = (SRCA)[0] + (SRCB)[0];     \
 567       (DST)[1] = (SRCA)[1] + (SRCB)[1];     \
 568 } while (0)
 569
 570 /** In-place scalar multiplication */
 571 #define SCALE_2V( DST, SRCA, SRCB )     \
 572 do {                        \
 573       (DST)[0] = (SRCA)[0] * (SRCB)[0];     \
 574       (DST)[1] = (SRCA)[1] * (SRCB)[1];     \
 575 } while (0)
 576
 577 /** In-place addition */
 578 #define ACC_2V( DST, SRC )          \
 579 do {                        \
 580       (DST)[0] += (SRC)[0];         \
 581       (DST)[1] += (SRC)[1];         \
 582 } while (0)
 583
 584 /** Element-wise multiplication and addition */
 585 #define ACC_SCALE_2V( DST, SRCA, SRCB )     \
 586 do {                        \
 587       (DST)[0] += (SRCA)[0] * (SRCB)[0];    \
 588       (DST)[1] += (SRCA)[1] * (SRCB)[1];    \
 589 } while (0)
 590
 591 /** Scalar multiplication */
 592 #define SCALE_SCALAR_2V( DST, S, SRCB )     \
 593 do {                        \
 594       (DST)[0] = S * (SRCB)[0];         \
 595       (DST)[1] = S * (SRCB)[1];         \
 596 } while (0)
 597
 598 /** In-place scalar multiplication and addition */
 599 #define ACC_SCALE_SCALAR_2V( DST, S, SRCB ) \
 600 do {                        \
 601       (DST)[0] += S * (SRCB)[0];        \
 602       (DST)[1] += S * (SRCB)[1];        \
 603 } while (0)
 604
 605 /** In-place scalar multiplication */
 606 #define SELF_SCALE_SCALAR_2V( DST, S )      \
 607 do {                        \
 608       (DST)[0] *= S;                \
 609       (DST)[1] *= S;                \
 610 } while (0)
 611
 612 /** In-place scalar addition */
 613 #define ACC_SCALAR_2V( DST, S )         \
 614 do {                        \
 615       (DST)[0] += S;                \
 616       (DST)[1] += S;                \
 617 } while (0)
 618
 619 /** Assign scalers to short vectors */
 620 #define ASSIGN_2V( V, V0, V1 )  \
 621 do {                            \
 622     V[0] = V0;                  \
 623     V[1] = V1;                  \
 624 } while(0)
 625
 626 /*@}*/
 627
 628 /** Copy \p sz elements into a homegeneous (4-element) vector, giving
 629  * default values to the remaining components.
 630  * The default values are chosen based on \p type.
 631  */
 632 static inline void
 633 COPY_CLEAN_4V_TYPE_AS_UNION(fi_type dst[4], int sz, const fi_type src[4],
 634                             GLenum type)
 635 {
 636    switch (type) {
 637    case GL_FLOAT:
 638       ASSIGN_4V(dst, FLOAT_AS_UNION(0), FLOAT_AS_UNION(0),
 639                 FLOAT_AS_UNION(0), FLOAT_AS_UNION(1));
 640       break;
 641    case GL_INT:
 642       ASSIGN_4V(dst, INT_AS_UNION(0), INT_AS_UNION(0),
 643                 INT_AS_UNION(0), INT_AS_UNION(1));
 644       break;
 645    case GL_UNSIGNED_INT:
 646       ASSIGN_4V(dst, UINT_AS_UNION(0), UINT_AS_UNION(0),
 647                 UINT_AS_UNION(0), UINT_AS_UNION(1));
 648       break;
 649    default:
 650       ASSIGN_4V(dst, FLOAT_AS_UNION(0), FLOAT_AS_UNION(0),
 651                 FLOAT_AS_UNION(0), FLOAT_AS_UNION(1)); /* silence warnings */
 652       assert(!"Unexpected type in COPY_CLEAN_4V_TYPE_AS_UNION macro");
 653    }
 654    COPY_SZ_4V(dst, sz, src);
 655 }
 656
 657 /** \name Linear interpolation functions */
 658 /*@{*/
 659
 660 static inline GLfloat
 661 LINTERP(GLfloat t, GLfloat out, GLfloat in)
 662 {
 663    return out + t * (in - out);
 664 }
 665
 666 static inline void
 667 INTERP_3F(GLfloat t, GLfloat dst[3], const GLfloat out[3], const GLfloat in[3])
 668 {
 669    dst[0] = LINTERP( t, out[0], in[0] );
 670    dst[1] = LINTERP( t, out[1], in[1] );
 671    dst[2] = LINTERP( t, out[2], in[2] );
 672 }
 673
 674 static inline void
 675 INTERP_4F(GLfloat t, GLfloat dst[4], const GLfloat out[4], const GLfloat in[4])
 676 {
 677    dst[0] = LINTERP( t, out[0], in[0] );
 678    dst[1] = LINTERP( t, out[1], in[1] );
 679    dst[2] = LINTERP( t, out[2], in[2] );
 680    dst[3] = LINTERP( t, out[3], in[3] );
 681 }
 682
 683 /*@}*/
 684
 685
 686
 687 static inline unsigned
 688 minify(unsigned value, unsigned levels)
 689 {
 690     return MAX2(1, value >> levels);
 691 }
 692
 693 /**
 694  * Align a value up to an alignment value
 695  *
 696  * If \c value is not already aligned to the requested alignment value, it
 697  * will be rounded up.
 698  *
 699  * \param value  Value to be rounded
 700  * \param alignment  Alignment value to be used.  This must be a power of two.
 701  *
 702  * \sa ROUND_DOWN_TO()
 703  */
 704 static inline uintptr_t
 705 ALIGN(uintptr_t value, int32_t alignment)
 706 {
 707    assert((alignment > 0) && _mesa_is_pow_two(alignment));
 708    return (((value) + (alignment) - 1) & ~((alignment) - 1));
 709 }
 710
 711 /**
 712  * Like ALIGN(), but works with a non-power-of-two alignment.
 713  */
 714 static inline uintptr_t
 715 ALIGN_NPOT(uintptr_t value, int32_t alignment)
 716 {
 717    assert(alignment > 0);
 718    return (value + alignment - 1) / alignment * alignment;
 719 }
 720
 721 /**
 722  * Align a value down to an alignment value
 723  *
 724  * If \c value is not already aligned to the requested alignment value, it
 725  * will be rounded down.
 726  *
 727  * \param value  Value to be rounded
 728  * \param alignment  Alignment value to be used.  This must be a power of two.
 729  *
 730  * \sa ALIGN()
 731  */
 732 static inline uintptr_t
 733 ROUND_DOWN_TO(uintptr_t value, int32_t alignment)
 734 {
 735    assert((alignment > 0) && _mesa_is_pow_two(alignment));
 736    return ((value) & ~(alignment - 1));
 737 }
 738
 739
 740 /** Cross product of two 3-element vectors */
 741 static inline void
 742 CROSS3(GLfloat n[3], const GLfloat u[3], const GLfloat v[3])
 743 {
 744    n[0] = u[1] * v[2] - u[2] * v[1];
 745    n[1] = u[2] * v[0] - u[0] * v[2];
 746    n[2] = u[0] * v[1] - u[1] * v[0];
 747 }
 748
 749
 750 /** Dot product of two 2-element vectors */
 751 static inline GLfloat
 752 DOT2(const GLfloat a[2], const GLfloat b[2])
 753 {
 754    return a[0] * b[0] + a[1] * b[1];
 755 }
 756
 757 static inline GLfloat
 758 DOT3(const GLfloat a[3], const GLfloat b[3])
 759 {
 760    return a[0] * b[0] + a[1] * b[1] + a[2] * b[2];
 761 }
 762
 763 static inline GLfloat
 764 DOT4(const GLfloat a[4], const GLfloat b[4])
 765 {
 766    return a[0] * b[0] + a[1] * b[1] + a[2] * b[2] + a[3] * b[3];
 767 }
 768
 769
 770 static inline GLfloat
 771 LEN_SQUARED_3FV(const GLfloat v[3])
 772 {
 773    return DOT3(v, v);
 774 }
 775
 776 static inline GLfloat
 777 LEN_SQUARED_2FV(const GLfloat v[2])
 778 {
 779    return DOT2(v, v);
 780 }
 781
 782
 783 static inline GLfloat
 784 LEN_3FV(const GLfloat v[3])
 785 {
 786    return sqrtf(LEN_SQUARED_3FV(v));
 787 }
 788
 789 static inline GLfloat
 790 LEN_2FV(const GLfloat v[2])
 791 {
 792    return sqrtf(LEN_SQUARED_2FV(v));
 793 }
 794
 795
 796 /* Normalize a 3-element vector to unit length. */
 797 static inline void
 798 NORMALIZE_3FV(GLfloat v[3])
 799 {
 800    GLfloat len = (GLfloat) LEN_SQUARED_3FV(v);
 801    if (len) {
 802       len = 1.0f / sqrtf(len);
 803       v[0] *= len;
 804       v[1] *= len;
 805       v[2] *= len;
 806    }
 807 }
 808
 809
 810 /** Test two floats have opposite signs */
 811 static inline GLboolean
 812 DIFFERENT_SIGNS(GLfloat x, GLfloat y)
 813 {
 814 #ifdef _MSC_VER
 815 #pragma warning( push )
 816 #pragma warning( disable : 6334 ) /* sizeof operator applied to an expression with an operator may yield unexpected results */
 817 #endif
 818    return signbit(x) != signbit(y);
 819 #ifdef _MSC_VER
 820 #pragma warning( pop )
 821 #endif
 822 }
 823
 824
 825 /** casts to silence warnings with some compilers */
 826 #define ENUM_TO_INT(E)     ((GLint)(E))
 827 #define ENUM_TO_FLOAT(E)   ((GLfloat)(GLint)(E))
 828 #define ENUM_TO_DOUBLE(E)  ((GLdouble)(GLint)(E))
 829 #define ENUM_TO_BOOLEAN(E) ((E) ? GL_TRUE : GL_FALSE)
 830
 831
 832 /* Stringify */
 833 #define STRINGIFY(x) #x
 834
 835 #endif