/*
* Mesa 3-D graphics library
- * Version: 7.1
*
* Copyright (C) 1999-2007 Brian Paul All Rights Reserved.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
- * BRIAN PAUL BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
- * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
- * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
*/
-
-
+#include <stdio.h>
+#include <stdarg.h>
+#include "c99_math.h"
+#include "util/rounding.h" /* for _mesa_roundeven */
#include "imports.h"
#include "context.h"
#include "mtypes.h"
#endif
-#ifdef WIN32
+#ifdef _WIN32
#define vsnprintf _vsnprintf
-#elif defined(__IBMC__) || defined(__IBMCPP__) || ( defined(__VMS) && __CRTL_VER < 70312000 )
+#elif defined(__IBMC__) || defined(__IBMCPP__)
extern int vsnprintf(char *str, size_t count, const char *fmt, va_list arg);
-#ifdef __VMS
-#include "vsnprintf.c"
-#endif
#endif
/**********************************************************************/
if (err)
return NULL;
return mem;
-#elif defined(_WIN32) && defined(_MSC_VER)
+#elif defined(_WIN32)
return _aligned_malloc(bytes, alignment);
#else
uintptr_t ptr, buf;
- ASSERT( alignment > 0 );
+ assert( alignment > 0 );
- ptr = (uintptr_t) malloc(bytes + alignment + sizeof(void *));
+ ptr = (uintptr_t)malloc(bytes + alignment + sizeof(void *));
if (!ptr)
return NULL;
}
return mem;
-#elif defined(_WIN32) && defined(_MSC_VER)
+#elif defined(_WIN32)
void *mem;
mem = _aligned_malloc(bytes, alignment);
#else
uintptr_t ptr, buf;
- ASSERT( alignment > 0 );
+ assert( alignment > 0 );
- ptr = (uintptr_t) calloc(1, bytes + alignment + sizeof(void *));
+ ptr = (uintptr_t)calloc(1, bytes + alignment + sizeof(void *));
if (!ptr)
return NULL;
* \param ptr pointer to the memory to be freed.
* The actual address to free is stored in the word immediately before the
* address the client sees.
+ * Note that it is legal to pass NULL pointer to this function and will be
+ * handled accordingly.
*/
void
_mesa_align_free(void *ptr)
{
#if defined(HAVE_POSIX_MEMALIGN)
free(ptr);
-#elif defined(_WIN32) && defined(_MSC_VER)
+#elif defined(_WIN32)
_aligned_free(ptr);
#else
- void **cubbyHole = (void **) ((char *) ptr - sizeof(void *));
- void *realAddr = *cubbyHole;
- free(realAddr);
+ if (ptr) {
+ void **cubbyHole = (void **) ((char *) ptr - sizeof(void *));
+ void *realAddr = *cubbyHole;
+ free(realAddr);
+ }
#endif /* defined(HAVE_POSIX_MEMALIGN) */
}
_mesa_align_realloc(void *oldBuffer, size_t oldSize, size_t newSize,
unsigned long alignment)
{
-#if defined(_WIN32) && defined(_MSC_VER)
+#if defined(_WIN32)
(void) oldSize;
return _aligned_realloc(oldBuffer, newSize, alignment);
#else
if (newBuf && oldBuffer && copySize > 0) {
memcpy(newBuf, oldBuffer, copySize);
}
- if (oldBuffer)
- _mesa_align_free(oldBuffer);
+
+ _mesa_align_free(oldBuffer);
return newBuf;
#endif
}
-
-
-/** Reallocate memory */
-void *
-_mesa_realloc(void *oldBuffer, size_t oldSize, size_t newSize)
-{
- const size_t copySize = (oldSize < newSize) ? oldSize : newSize;
- void *newBuffer = malloc(newSize);
- if (newBuffer && oldBuffer && copySize > 0)
- memcpy(newBuffer, oldBuffer, copySize);
- if (oldBuffer)
- free(oldBuffer);
- return newBuffer;
-}
-
-/**
- * Fill memory with a constant 16bit word.
- * \param dst destination pointer.
- * \param val value.
- * \param n number of words.
- */
-void
-_mesa_memset16( unsigned short *dst, unsigned short val, size_t n )
-{
- while (n-- > 0)
- *dst++ = val;
-}
-
/*@}*/
/*@{*/
-/**
- inv_sqrt - A single precision 1/sqrt routine for IEEE format floats.
- written by Josh Vanderhoof, based on newsgroup posts by James Van Buskirk
- and Vesa Karvonen.
-*/
-float
-_mesa_inv_sqrtf(float n)
-{
-#if defined(USE_IEEE) && !defined(DEBUG)
- float r0, x0, y0;
- float r1, x1, y1;
- float r2, x2, y2;
-#if 0 /* not used, see below -BP */
- float r3, x3, y3;
-#endif
- fi_type u;
- unsigned int magic;
-
- /*
- Exponent part of the magic number -
-
- We want to:
- 1. subtract the bias from the exponent,
- 2. negate it
- 3. divide by two (rounding towards -inf)
- 4. add the bias back
-
- Which is the same as subtracting the exponent from 381 and dividing
- by 2.
-
- floor(-(x - 127) / 2) + 127 = floor((381 - x) / 2)
- */
-
- magic = 381 << 23;
-
- /*
- Significand part of magic number -
-
- With the current magic number, "(magic - u.i) >> 1" will give you:
-
- for 1 <= u.f <= 2: 1.25 - u.f / 4
- for 2 <= u.f <= 4: 1.00 - u.f / 8
-
- This isn't a bad approximation of 1/sqrt. The maximum difference from
- 1/sqrt will be around .06. After three Newton-Raphson iterations, the
- maximum difference is less than 4.5e-8. (Which is actually close
- enough to make the following bias academic...)
-
- To get a better approximation you can add a bias to the magic
- number. For example, if you subtract 1/2 of the maximum difference in
- the first approximation (.03), you will get the following function:
-
- for 1 <= u.f <= 2: 1.22 - u.f / 4
- for 2 <= u.f <= 3.76: 0.97 - u.f / 8
- for 3.76 <= u.f <= 4: 0.72 - u.f / 16
- (The 3.76 to 4 range is where the result is < .5.)
-
- This is the closest possible initial approximation, but with a maximum
- error of 8e-11 after three NR iterations, it is still not perfect. If
- you subtract 0.0332281 instead of .03, the maximum error will be
- 2.5e-11 after three NR iterations, which should be about as close as
- is possible.
-
- for 1 <= u.f <= 2: 1.2167719 - u.f / 4
- for 2 <= u.f <= 3.73: 0.9667719 - u.f / 8
- for 3.73 <= u.f <= 4: 0.7167719 - u.f / 16
-
- */
-
- magic -= (int)(0.0332281 * (1 << 25));
-
- u.f = n;
- u.i = (magic - u.i) >> 1;
-
- /*
- Instead of Newton-Raphson, we use Goldschmidt's algorithm, which
- allows more parallelism. From what I understand, the parallelism
- comes at the cost of less precision, because it lets error
- accumulate across iterations.
- */
- x0 = 1.0f;
- y0 = 0.5f * n;
- r0 = u.f;
-
- x1 = x0 * r0;
- y1 = y0 * r0 * r0;
- r1 = 1.5f - y1;
-
- x2 = x1 * r1;
- y2 = y1 * r1 * r1;
- r2 = 1.5f - y2;
-
-#if 1
- return x2 * r2; /* we can stop here, and be conformant -BP */
-#else
- x3 = x2 * r2;
- y3 = y2 * r2 * r2;
- r3 = 1.5f - y3;
-
- return x3 * r3;
-#endif
-#else
- return (float) (1.0 / sqrt(n));
-#endif
-}
-
-#ifndef __GNUC__
-/**
- * Find the first bit set in a word.
- */
-int
-ffs(int i)
-{
- register int bit = 0;
- if (i != 0) {
- if ((i & 0xffff) == 0) {
- bit += 16;
- i >>= 16;
- }
- if ((i & 0xff) == 0) {
- bit += 8;
- i >>= 8;
- }
- if ((i & 0xf) == 0) {
- bit += 4;
- i >>= 4;
- }
- while ((i & 1) == 0) {
- bit++;
- i >>= 1;
- }
- bit++;
- }
- return bit;
-}
-
-
-/**
- * Find position of first bit set in given value.
- * XXX Warning: this function can only be used on 64-bit systems!
- * \return position of least-significant bit set, starting at 1, return zero
- * if no bits set.
- */
-int
-ffsll(long long int val)
-{
- int bit;
-
- assert(sizeof(val) == 8);
-
- bit = ffs((int) val);
- if (bit != 0)
- return bit;
-
- bit = ffs((int) (val >> 32));
- if (bit != 0)
- return 32 + bit;
-
- return 0;
-}
-#endif /* __GNUC__ */
-
-
-#if !defined(__GNUC__) ||\
- ((__GNUC__ * 100 + __GNUC_MINOR__) < 304) /* Not gcc 3.4 or later */
+#ifndef HAVE___BUILTIN_POPCOUNT
/**
* Return number of bits set in given GLuint.
*/
}
return bits;
}
+#endif
+#ifndef HAVE___BUILTIN_POPCOUNTLL
/**
* Return number of bits set in given 64-bit uint.
*/
}
#endif
-
-/**
- * Convert a 4-byte float to a 2-byte half float.
- * Based on code from:
- * http://www.opengl.org/discussion_boards/ubb/Forum3/HTML/008786.html
- */
-GLhalfARB
-_mesa_float_to_half(float val)
-{
- const fi_type fi = {val};
- const int flt_m = fi.i & 0x7fffff;
- const int flt_e = (fi.i >> 23) & 0xff;
- const int flt_s = (fi.i >> 31) & 0x1;
- int s, e, m = 0;
- GLhalfARB result;
-
- /* sign bit */
- s = flt_s;
-
- /* handle special cases */
- if ((flt_e == 0) && (flt_m == 0)) {
- /* zero */
- /* m = 0; - already set */
- e = 0;
- }
- else if ((flt_e == 0) && (flt_m != 0)) {
- /* denorm -- denorm float maps to 0 half */
- /* m = 0; - already set */
- e = 0;
- }
- else if ((flt_e == 0xff) && (flt_m == 0)) {
- /* infinity */
- /* m = 0; - already set */
- e = 31;
- }
- else if ((flt_e == 0xff) && (flt_m != 0)) {
- /* NaN */
- m = 1;
- e = 31;
- }
- else {
- /* regular number */
- const int new_exp = flt_e - 127;
- if (new_exp < -24) {
- /* this maps to 0 */
- /* m = 0; - already set */
- e = 0;
- }
- else if (new_exp < -14) {
- /* this maps to a denorm */
- unsigned int exp_val = (unsigned int) (-14 - new_exp); /* 2^-exp_val*/
- e = 0;
- switch (exp_val) {
- case 0:
- _mesa_warning(NULL,
- "float_to_half: logical error in denorm creation!\n");
- /* m = 0; - already set */
- break;
- case 1: m = 512 + (flt_m >> 14); break;
- case 2: m = 256 + (flt_m >> 15); break;
- case 3: m = 128 + (flt_m >> 16); break;
- case 4: m = 64 + (flt_m >> 17); break;
- case 5: m = 32 + (flt_m >> 18); break;
- case 6: m = 16 + (flt_m >> 19); break;
- case 7: m = 8 + (flt_m >> 20); break;
- case 8: m = 4 + (flt_m >> 21); break;
- case 9: m = 2 + (flt_m >> 22); break;
- case 10: m = 1; break;
- }
- }
- else if (new_exp > 15) {
- /* map this value to infinity */
- /* m = 0; - already set */
- e = 31;
- }
- else {
- /* regular */
- e = new_exp + 15;
- m = flt_m >> 13;
- }
- }
-
- result = (s << 15) | (e << 10) | m;
- return result;
-}
-
-
-/**
- * Convert a 2-byte half float to a 4-byte float.
- * Based on code from:
- * http://www.opengl.org/discussion_boards/ubb/Forum3/HTML/008786.html
- */
-float
-_mesa_half_to_float(GLhalfARB val)
-{
- /* XXX could also use a 64K-entry lookup table */
- const int m = val & 0x3ff;
- const int e = (val >> 10) & 0x1f;
- const int s = (val >> 15) & 0x1;
- int flt_m, flt_e, flt_s;
- fi_type fi;
- float result;
-
- /* sign bit */
- flt_s = s;
-
- /* handle special cases */
- if ((e == 0) && (m == 0)) {
- /* zero */
- flt_m = 0;
- flt_e = 0;
- }
- else if ((e == 0) && (m != 0)) {
- /* denorm -- denorm half will fit in non-denorm single */
- const float half_denorm = 1.0f / 16384.0f; /* 2^-14 */
- float mantissa = ((float) (m)) / 1024.0f;
- float sign = s ? -1.0f : 1.0f;
- return sign * mantissa * half_denorm;
- }
- else if ((e == 31) && (m == 0)) {
- /* infinity */
- flt_e = 0xff;
- flt_m = 0;
- }
- else if ((e == 31) && (m != 0)) {
- /* NaN */
- flt_e = 0xff;
- flt_m = 1;
- }
- else {
- /* regular */
- flt_e = e + 112;
- flt_m = m << 13;
- }
-
- fi.i = (flt_s << 31) | (flt_e << 23) | flt_m;
- result = fi.f;
- return result;
-}
-
-/*@}*/
-
-
-/**********************************************************************/
-/** \name Sort & Search */
-/*@{*/
-
-/**
- * Wrapper for bsearch().
- */
-void *
-_mesa_bsearch( const void *key, const void *base, size_t nmemb, size_t size,
- int (*compar)(const void *, const void *) )
-{
-#if defined(_WIN32_WCE)
- void *mid;
- int cmp;
- while (nmemb) {
- nmemb >>= 1;
- mid = (char *)base + nmemb * size;
- cmp = (*compar)(key, mid);
- if (cmp == 0)
- return mid;
- if (cmp > 0) {
- base = (char *)mid + size;
- --nmemb;
- }
- }
- return NULL;
-#else
- return bsearch(key, base, nmemb, size, compar);
-#endif
-}
-
-/*@}*/
-
-
-/**********************************************************************/
-/** \name Environment vars */
-/*@{*/
-
-/**
- * Wrapper for getenv().
- */
-char *
-_mesa_getenv( const char *var )
-{
-#if defined(_XBOX) || defined(_WIN32_WCE)
- return NULL;
-#else
- return getenv(var);
-#endif
-}
-
/*@}*/
/** \name String */
/*@{*/
-/**
- * Implemented using malloc() and strcpy.
- * Note that NULL is handled accordingly.
- */
-char *
-_mesa_strdup( const char *s )
-{
- if (s) {
- size_t l = strlen(s);
- char *s2 = (char *) malloc(l + 1);
- if (s2)
- strcpy(s2, s);
- return s2;
- }
- else {
- return NULL;
- }
-}
-
-/** Wrapper around strtof() */
-float
-_mesa_strtof( const char *s, char **end )
-{
-#if defined(_GNU_SOURCE) && !defined(__CYGWIN__) && !defined(__FreeBSD__) && \
- !defined(ANDROID) && !defined(__HAIKU__)
- static locale_t loc = NULL;
- if (!loc) {
- loc = newlocale(LC_CTYPE_MASK, "C", NULL);
- }
- return strtof_l(s, end, loc);
-#elif defined(_ISOC99_SOURCE) || (defined(_XOPEN_SOURCE) && _XOPEN_SOURCE >= 600)
- return strtof(s, end);
-#else
- return (float)strtod(s, end);
-#endif
-}
/** Compute simple checksum/hash for a string */
unsigned int