mutter/cogl/cogl-util.h

210 lines
6.5 KiB
C
Raw Normal View History

/*
* Cogl
*
* An object oriented GL/GLES Abstraction/Utility Layer
*
* Copyright (C) 2007,2008,2009,2010 Intel Corporation.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library. If not, see <http://www.gnu.org/licenses/>.
*
*
*/
#ifndef __COGL_UTIL_H
#define __COGL_UTIL_H
#include <glib.h>
#include <math.h>
#include <cogl/cogl-defines.h>
#include "cogl-types.h"
#ifndef COGL_HAS_GLIB_SUPPORT
#include <stdio.h>
#endif
/* When compiling with Visual Studio, symbols that represent data that
are exported out of the DLL need to be marked with the dllexport
attribute. */
#ifdef _MSC_VER
#ifdef COGL_BUILD_EXP
#define COGL_EXPORT __declspec(dllexport)
#else
#define COGL_EXPORT __declspec(dllimport)
#endif
#else
#define COGL_EXPORT
#endif
int
_cogl_util_next_p2 (int a);
/* The signbit macro is defined by ISO C99 so it should be available,
however if it's not we can fallback to an evil hack */
#ifdef signbit
#define cogl_util_float_signbit(x) signbit(x)
#else
/* This trick was stolen from here:
http://lists.boost.org/Archives/boost/2006/08/108731.php
It xors the integer reinterpretations of -1.0f and 1.0f. In theory
they should only differ by the signbit so that gives a mask for the
sign which we can just test against the value */
static inline gboolean
cogl_util_float_signbit (float x)
{
static const union { float f; guint32 i; } negative_one = { -1.0f };
static const union { float f; guint32 i; } positive_one = { +1.0f };
union { float f; guint32 i; } value = { x };
return !!((negative_one.i ^ positive_one.i) & value.i);
}
#endif
/* This is a replacement for the nearbyint function which always
rounds to the nearest integer. nearbyint is apparently a C99
function so it might not always be available but also it seems in
glibc it is defined as a function call so this macro could end up
faster anyway. We can't just add 0.5f because it will break for
negative numbers. */
#define COGL_UTIL_NEARBYINT(x) ((int) ((x) < 0.0f ? (x) - 0.5f : (x) + 0.5f))
/* Returns whether the given integer is a power of two */
static inline gboolean
_cogl_util_is_pot (unsigned int num)
{
/* Make sure there is only one bit set */
return (num & (num - 1)) == 0;
}
/* Split Bob Jenkins' One-at-a-Time hash
*
* This uses the One-at-a-Time hash algorithm designed by Bob Jenkins
* but the mixing step is split out so the function can be used in a
* more incremental fashion.
*/
static inline unsigned int
_cogl_util_one_at_a_time_hash (unsigned int hash,
Use GL_ARB_sampler_objects GL_ARB_sampler_objects provides a GL object which overrides the sampler state part of a texture object with different values. The sampler state that Cogl currently exposes is the wrap modes and filters. Cogl exposes the state as part of the pipeline layer state but without this extension GL only exposes it as part of the texture object state. This means that it won't work to use a single texture multiple times in one primitive with different sampler states. It also makes switching between different sampler states with a single texture not terribly efficient because it has to change the texture object state every time. This patch adds a cache for sampler states in a shared hash table attached to the CoglContext. The entire set of parameters for the sampler state is used as the key for the hash table. When a unique state is encountered the sampler cache will create a new entry, otherwise it will return a const pointer to an existing entry. That means we can have a single pointer to represent any combination of sampler state. Pipeline layers now just store this single pointer rather than storing all of the sampler state. The two separate state flags for wrap modes and filters have now been combined into one. It should be faster to compare the sampler state now because instead of comparing each value it can just compare the pointers to the cached sampler entries. The hash table of cached sampler states should only need to perform its more expensive hash on the state when a property is changed on a pipeline, not every time it is flushed. When the sampler objects extension is available each cached sampler state will also get a sampler object to represent it. The common code to flush the GL state will now simply bind this object to a unit instead of flushing the state though the CoglTexture when possible. Reviewed-by: Robert Bragg <robert@linux.intel.com>
2012-04-04 21:20:04 +00:00
const void *key,
size_t bytes)
{
Use GL_ARB_sampler_objects GL_ARB_sampler_objects provides a GL object which overrides the sampler state part of a texture object with different values. The sampler state that Cogl currently exposes is the wrap modes and filters. Cogl exposes the state as part of the pipeline layer state but without this extension GL only exposes it as part of the texture object state. This means that it won't work to use a single texture multiple times in one primitive with different sampler states. It also makes switching between different sampler states with a single texture not terribly efficient because it has to change the texture object state every time. This patch adds a cache for sampler states in a shared hash table attached to the CoglContext. The entire set of parameters for the sampler state is used as the key for the hash table. When a unique state is encountered the sampler cache will create a new entry, otherwise it will return a const pointer to an existing entry. That means we can have a single pointer to represent any combination of sampler state. Pipeline layers now just store this single pointer rather than storing all of the sampler state. The two separate state flags for wrap modes and filters have now been combined into one. It should be faster to compare the sampler state now because instead of comparing each value it can just compare the pointers to the cached sampler entries. The hash table of cached sampler states should only need to perform its more expensive hash on the state when a property is changed on a pipeline, not every time it is flushed. When the sampler objects extension is available each cached sampler state will also get a sampler object to represent it. The common code to flush the GL state will now simply bind this object to a unit instead of flushing the state though the CoglTexture when possible. Reviewed-by: Robert Bragg <robert@linux.intel.com>
2012-04-04 21:20:04 +00:00
const unsigned char *p = key;
int i;
for (i = 0; i < bytes; i++)
{
hash += p[i];
hash += (hash << 10);
hash ^= (hash >> 6);
}
return hash;
}
unsigned int
_cogl_util_one_at_a_time_mix (unsigned int hash);
/* These two builtins are available since GCC 3.4 */
#if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4)
#define COGL_UTIL_HAVE_BUILTIN_FFSL
#define COGL_UTIL_HAVE_BUILTIN_POPCOUNTL
#endif
/* The 'ffs' function is part of C99 so it isn't always available */
#ifdef HAVE_FFS
#define _cogl_util_ffs ffs
#else
int
_cogl_util_ffs (int num);
#endif
/* The 'ffsl' function is non-standard but GCC has a builtin for it
since 3.4 which we can use */
#ifdef COGL_UTIL_HAVE_BUILTIN_FFSL
#define _cogl_util_ffsl __builtin_ffsl
#else
/* If ints and longs are the same size we can just use ffs. Hopefully
the compiler will optimise away this conditional */
#define _cogl_util_ffsl(x) \
(sizeof (long int) == sizeof (int) ? _cogl_util_ffs ((int) x) : \
_cogl_util_ffsl_wrapper (x))
int
_cogl_util_ffsl_wrapper (long int num);
#endif /* COGL_UTIL_HAVE_BUILTIN_FFSL */
#ifdef COGL_UTIL_HAVE_BUILTIN_POPCOUNTL
#define _cogl_util_popcountl __builtin_popcountl
#else
extern const unsigned char _cogl_util_popcount_table[256];
/* There are many ways of doing popcount but doing a table lookup
seems to be the most robust against different sizes for long. Some
pages seem to claim it's the fastest method anyway. */
static inline int
_cogl_util_popcountl (unsigned long num)
{
int i;
int sum = 0;
/* Let's hope GCC will unroll this loop.. */
for (i = 0; i < sizeof (num); i++)
sum += _cogl_util_popcount_table[(num >> (i * 8)) & 0xff];
return sum;
}
#endif /* COGL_UTIL_HAVE_BUILTIN_POPCOUNTL */
#ifdef COGL_HAS_GLIB_SUPPORT
#define _COGL_RETURN_IF_FAIL(EXPR) g_return_if_fail(EXPR)
#define _COGL_RETURN_VAL_IF_FAIL(EXPR, VAL) g_return_val_if_fail(EXPR, VAL)
#else
#define _COGL_RETURN_IF_FAIL(EXPR) do { \
if (!(EXPR)) \
{ \
fprintf (stderr, "file %s: line %d: assertion `%s' failed", \
__FILE__, \
__LINE__, \
#EXPR); \
return; \
}; \
} while(0)
#define _COGL_RETURN_VAL_IF_FAIL(EXPR, VAL) do { \
if (!(EXPR)) \
{ \
fprintf (stderr, "file %s: line %d: assertion `%s' failed", \
__FILE__, \
__LINE__, \
#EXPR); \
return (VAL); \
}; \
} while(0)
#endif /* COGL_HAS_GLIB_SUPPORT */
/* Match a CoglPixelFormat according to channel masks, color depth,
* bits per pixel and byte order. These information are provided by
* the Visual and XImage structures.
*
* If no specific pixel format could be found, COGL_PIXEL_FORMAT_ANY
* is returned.
*/
CoglPixelFormat
_cogl_util_pixel_format_from_masks (unsigned long r_mask,
unsigned long g_mask,
unsigned long b_mask,
int depth, int bpp,
int byte_order);
#endif /* __COGL_UTIL_H */