mutter/cogl/cogl-texture.c

1022 lines
30 KiB
C
Raw Normal View History

/*
* Cogl
*
* An object oriented GL/GLES Abstraction/Utility Layer
*
* Copyright (C) 2007,2008,2009 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, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 02111-1307, USA.
*
* Authors:
* Matthew Allum <mallum@openedhand.com>
* Neil Roberts <neil@linux.intel.com>
* Robert Bragg <robert@linux.intel.com>
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "cogl.h"
#include "cogl-internal.h"
#include "cogl-util.h"
#include "cogl-bitmap.h"
#include "cogl-bitmap-private.h"
#include "cogl-buffer-private.h"
#include "cogl-pixel-buffer-private.h"
#include "cogl-texture-private.h"
#include "cogl-texture-driver.h"
#include "cogl-texture-2d-sliced-private.h"
#include "cogl-texture-2d-private.h"
#include "cogl-sub-texture-private.h"
#include "cogl-atlas-texture-private.h"
#include "cogl-material.h"
#include "cogl-context.h"
Removed COGLhandle and changed shader and program functions to be wrapped in reference-counted CoglHandles instead. * clutter/cogl/gl/cogl-shader.c: * clutter/cogl/gl/cogl-shader.h: * clutter/cogl/gl/cogl-program.c: * clutter/cogl/gl/cogl-program.h: New files to hold the shader and program functions. * clutter/cogl/gl/cogl.c: Removed shader and program functions. * clutter/cogl/common/cogl-handle.h: New header to define COGL_HANDLE_DEFINE which helps build functions to create reference-counted handles. This reduces the amount of duplicated code. * clutter/cogl/gl/cogl-texture.c: * clutter/cogl/gles/cogl-texture.c: * clutter/cogl/gl/cogl-fbo.c: Converted to use COGL_HANDLE_DEFINE from cogl-handle.h to avoid duplicating some of the common code. * clutter/cogl/gles/cogl-defines.h.in: * clutter/cogl/gl/cogl-defines.h.in: Removed COGLhandle * clutter/cogl/gl/cogl-context.h: Added handle arrays for programs and shaders. * clutter/cogl/gl/cogl-context.c (cogl_create_context): Added initialisers for shader_handles and program_handles. (cogl_destroy_context): Added calls to g_array_free for all handle arrays. * clutter/cogl/gl/Makefile.am (libclutter_cogl_la_SOURCES): Added cogl-{program,shader}.{c,h} * clutter/cogl/common/Makefile.am (libclutter_cogl_common_la_SOURCES): Added cogl-handle.h * clutter/cogl/gles/cogl.c: * clutter/cogl/cogl.h.in: Programs and shaders are now wrapped in CoglHandles instead of COGLhandles. cogl_program_destroy and cogl_shader_destroy is now replaced with cogl_program_unref and cogl_shader_unref. cogl_program_ref and cogl_shader_ref are also added. * clutter/clutter-shader.c: Converted to use CoglHandles for the programs and shaders instead of COGLhandles. * cogl/cogl-sections.txt: Added cogl_shader_ref, cogl_shader_unref, cogl_is_shader, cogl_program_ref, cogl_program_unref, cogl_is_program and cogl_is_offscreen.
2008-04-29 12:10:37 -04:00
#include "cogl-handle.h"
[cogl] Improving Cogl journal to minimize driver overheads + GPU state changes Previously the journal was always flushed at the end of _cogl_rectangles_with_multitexture_coords, (i.e. the end of any cogl_rectangle* calls) but now we have broadened the potential for batching geometry. In ideal circumstances we will only flush once per scene. In summary the journal works like this: When you use any of the cogl_rectangle* APIs then nothing is emitted to the GPU at this point, we just log one or more quads into the journal. A journal entry consists of the quad coordinates, an associated material reference, and a modelview matrix. Ideally the journal only gets flushed once at the end of a scene, but in fact there are things to consider that may cause unwanted flushing, including: - modifying materials mid-scene This is because each quad in the journal has an associated material reference (i.e. not copy), so if you try and modify a material that is already referenced in the journal we force a flush first) NOTE: For now this means you should avoid using cogl_set_source_color() since that currently uses a single shared material. Later we should change it to use a pool of materials that is recycled when the journal is flushed. - modifying any state that isn't currently logged, such as depth, fog and backface culling enables. The first thing that happens when flushing, is to upload all the vertex data associated with the journal into a single VBO. We then go through a process of splitting up the journal into batches that have compatible state so they can be emitted to the GPU together. This is currently broken up into 3 levels so we can stagger the state changes: 1) we break the journal up according to changes in the number of material layers associated with logged quads. The number of layers in a material determines the stride of the associated vertices, so we have to update our vertex array offsets at this level. (i.e. calling gl{Vertex,Color},Pointer etc) 2) we further split batches up according to material compatability. (e.g. materials with different textures) We flush material state at this level. 3) Finally we split batches up according to modelview changes. At this level we update the modelview matrix and actually emit the actual draw command. This commit is largely about putting the initial design in-place; this will be followed by other changes that take advantage of the extended batching.
2009-06-17 13:46:42 -04:00
#include "cogl-primitives.h"
#include "cogl-framebuffer-private.h"
#include <string.h>
#include <stdlib.h>
#include <math.h>
/* XXX:
* The CoglHandle macros don't support any form of inheritance, so for
* now we implement the CoglHandle support for the CoglTexture
* abstract class manually.
*/
gboolean
cogl_is_texture (CoglHandle handle)
{
CoglHandleObject *obj = (CoglHandleObject *)handle;
if (handle == COGL_INVALID_HANDLE)
return FALSE;
return (obj->klass->type == _cogl_handle_texture_2d_get_type () ||
obj->klass->type == _cogl_handle_atlas_texture_get_type () ||
obj->klass->type == _cogl_handle_texture_2d_sliced_get_type () ||
obj->klass->type == _cogl_handle_sub_texture_get_type ());
}
CoglHandle
cogl_texture_ref (CoglHandle handle)
{
if (!cogl_is_texture (handle))
return COGL_INVALID_HANDLE;
_COGL_HANDLE_DEBUG_REF (CoglTexture, handle);
cogl_handle_ref (handle);
return handle;
}
void
cogl_texture_unref (CoglHandle handle)
{
if (!cogl_is_texture (handle))
{
g_warning (G_STRINGIFY (cogl_texture_unref)
": Ignoring unref of Cogl handle "
"due to type mismatch");
return;
}
_COGL_HANDLE_DEBUG_UNREF (CoglTexture, handle);
cogl_handle_unref (handle);
}
static gboolean
_cogl_texture_needs_premult_conversion (CoglPixelFormat src_format,
CoglPixelFormat dst_format)
{
return ((src_format & COGL_A_BIT) &&
src_format != COGL_PIXEL_FORMAT_A_8 &&
(src_format & COGL_PREMULT_BIT) !=
(dst_format & COGL_PREMULT_BIT));
}
CoglPixelFormat
_cogl_texture_determine_internal_format (CoglPixelFormat src_format,
CoglPixelFormat dst_format)
{
/* If the application hasn't specified a specific format then we'll
* pick the most appropriate. By default Cogl will use a
* premultiplied internal format. Later we will add control over
* this. */
if (dst_format == COGL_PIXEL_FORMAT_ANY)
{
if ((src_format & COGL_A_BIT) &&
src_format != COGL_PIXEL_FORMAT_A_8)
return src_format | COGL_PREMULT_BIT;
else
return src_format;
}
else
return dst_format;
}
gboolean
_cogl_texture_prepare_for_upload (CoglBitmap *src_bmp,
CoglPixelFormat dst_format,
CoglPixelFormat *dst_format_out,
CoglBitmap *dst_bmp,
gboolean *copied_bitmap,
GLenum *out_glintformat,
GLenum *out_glformat,
GLenum *out_gltype)
{
dst_format = _cogl_texture_determine_internal_format (src_bmp->format,
dst_format);
*copied_bitmap = FALSE;
*dst_bmp = *src_bmp;
/* If the source format does not have the same premult flag as the
dst format then we need to copy and convert it */
if (_cogl_texture_needs_premult_conversion (src_bmp->format,
dst_format))
{
dst_bmp->data = g_memdup (dst_bmp->data,
dst_bmp->height * dst_bmp->rowstride);
*copied_bitmap = TRUE;
if (!_cogl_bitmap_convert_premult_status (dst_bmp,
src_bmp->format ^
COGL_PREMULT_BIT))
{
g_free (dst_bmp->data);
return FALSE;
}
}
/* Use the source format from the src bitmap type and the internal
format from the dst format type so that GL can do the
conversion */
_cogl_pixel_format_to_gl (src_bmp->format,
NULL, /* internal format */
out_glformat,
out_gltype);
_cogl_pixel_format_to_gl (dst_format,
out_glintformat,
NULL,
NULL);
if (dst_format_out)
*dst_format_out = dst_format;
return TRUE;
}
void
_cogl_texture_prep_gl_alignment_for_pixels_upload (int pixels_rowstride)
{
if (!(pixels_rowstride & 0x7))
GE( glPixelStorei (GL_UNPACK_ALIGNMENT, 8) );
else if (!(pixels_rowstride & 0x3))
GE( glPixelStorei (GL_UNPACK_ALIGNMENT, 4) );
else if (!(pixels_rowstride & 0x1))
GE( glPixelStorei (GL_UNPACK_ALIGNMENT, 2) );
else
GE( glPixelStorei (GL_UNPACK_ALIGNMENT, 1) );
}
void
_cogl_texture_prep_gl_alignment_for_pixels_download (int pixels_rowstride)
{
if (!(pixels_rowstride & 0x7))
GE( glPixelStorei (GL_PACK_ALIGNMENT, 8) );
else if (!(pixels_rowstride & 0x3))
GE( glPixelStorei (GL_PACK_ALIGNMENT, 4) );
else if (!(pixels_rowstride & 0x1))
GE( glPixelStorei (GL_PACK_ALIGNMENT, 2) );
else
GE( glPixelStorei (GL_PACK_ALIGNMENT, 1) );
}
/* FIXME: wrap modes should be set on materials not textures */
void
_cogl_texture_set_wrap_mode_parameter (CoglHandle handle,
GLenum wrap_mode)
{
CoglTexture *tex = COGL_TEXTURE (handle);
tex->vtable->set_wrap_mode_parameter (tex, wrap_mode);
}
/* This is like CoglSpanIter except it deals with floats and it
effectively assumes there is only one span from 0.0 to 1.0 */
typedef struct _CoglTextureIter
{
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
float pos, end, next_pos;
gboolean flipped;
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
float t_1, t_2;
} CoglTextureIter;
static void
_cogl_texture_iter_update (CoglTextureIter *iter)
{
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
float t_2;
float frac_part;
frac_part = modff (iter->pos, &iter->next_pos);
/* modff rounds the int part towards zero so we need to add one if
we're meant to be heading away from zero */
if (iter->pos >= 0.0f || frac_part == 0.0f)
iter->next_pos += 1.0f;
if (iter->next_pos > iter->end)
t_2 = iter->end;
else
t_2 = iter->next_pos;
if (iter->flipped)
{
iter->t_1 = t_2;
iter->t_2 = iter->pos;
}
else
{
iter->t_1 = iter->pos;
iter->t_2 = t_2;
}
}
static void
_cogl_texture_iter_begin (CoglTextureIter *iter,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
float t_1, float t_2)
{
if (t_1 <= t_2)
{
iter->pos = t_1;
iter->end = t_2;
iter->flipped = FALSE;
}
else
{
iter->pos = t_2;
iter->end = t_1;
iter->flipped = TRUE;
}
_cogl_texture_iter_update (iter);
}
static void
_cogl_texture_iter_next (CoglTextureIter *iter)
{
iter->pos = iter->next_pos;
_cogl_texture_iter_update (iter);
}
static gboolean
_cogl_texture_iter_end (CoglTextureIter *iter)
{
return iter->pos >= iter->end;
}
/* This invokes the callback with enough quads to cover the manually
repeated range specified by the virtual texture coordinates without
emitting coordinates outside the range [0,1] */
void
_cogl_texture_iterate_manual_repeats (CoglTextureManualRepeatCallback callback,
float tx_1, float ty_1,
float tx_2, float ty_2,
void *user_data)
{
CoglTextureIter x_iter, y_iter;
for (_cogl_texture_iter_begin (&y_iter, ty_1, ty_2);
!_cogl_texture_iter_end (&y_iter);
_cogl_texture_iter_next (&y_iter))
for (_cogl_texture_iter_begin (&x_iter, tx_1, tx_2);
!_cogl_texture_iter_end (&x_iter);
_cogl_texture_iter_next (&x_iter))
{
float coords[4] = { x_iter.t_1, y_iter.t_1, x_iter.t_2, y_iter.t_2 };
callback (coords, user_data);
}
}
CoglHandle
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
cogl_texture_new_with_size (unsigned int width,
unsigned int height,
[cogl] Remove max_waste argument from Texture ctors The CoglTexture constructors expose the "max-waste" argument for controlling the maximum amount of wasted areas for slicing or, if set to -1, disables slicing. Slicing is really relevant only for large images that are never repeated, so it's a useful feature only in controlled use cases. Specifying the amount of wasted area is, on the other hand, just a way to mess up this feature; 99% the times, you either pull this number out of thin air, hoping it's right, or you try to do the right thing and you choose the wrong number anyway. Instead, we can use the CoglTextureFlags to control whether the texture should not be sliced (useful for Clutter-GST and for the texture-from-pixmap actors) and provide a reasonable value for enabling the slicing ourself. At some point, we might even provide a way to change the default at compile time or at run time, for particular platforms. Since max_waste is gone, the :tile-waste property of ClutterTexture becomes read-only, and it proxies the cogl_texture_get_max_waste() function. Inside Clutter, the only cases where the max_waste argument was not set to -1 are in the Pango glyph cache (which is a POT texture anyway) and inside the test cases where we want to force slicing; for the latter we can create larger textures that will be bigger than the threshold we set. Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com> Signed-off-by: Robert Bragg <robert@linux.intel.com> Signed-off-by: Neil Roberts <neil@linux.intel.com>
2009-05-23 14:18:18 -04:00
CoglTextureFlags flags,
CoglPixelFormat internal_format)
{
CoglHandle tex;
/* First try creating a fast-path non-sliced texture */
tex = _cogl_texture_2d_new_with_size (width, height, flags, internal_format);
/* If it fails resort to sliced textures */
if (tex == COGL_INVALID_HANDLE)
tex = _cogl_texture_2d_sliced_new_with_size (width,
height,
flags,
internal_format);
return tex;
}
CoglHandle
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
cogl_texture_new_from_data (unsigned int width,
unsigned int height,
CoglTextureFlags flags,
CoglPixelFormat format,
CoglPixelFormat internal_format,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int rowstride,
const guint8 *data)
{
CoglBitmap bitmap;
if (format == COGL_PIXEL_FORMAT_ANY)
return COGL_INVALID_HANDLE;
if (data == NULL)
return COGL_INVALID_HANDLE;
/* Rowstride from width if not given */
if (rowstride == 0)
rowstride = width * _cogl_get_format_bpp (format);
/* Wrap the data into a bitmap */
bitmap.width = width;
bitmap.height = height;
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
bitmap.data = (guint8 *) data;
bitmap.format = format;
bitmap.rowstride = rowstride;
return cogl_texture_new_from_bitmap (&bitmap, flags, internal_format);
}
CoglHandle
[cogl] Remove max_waste argument from Texture ctors The CoglTexture constructors expose the "max-waste" argument for controlling the maximum amount of wasted areas for slicing or, if set to -1, disables slicing. Slicing is really relevant only for large images that are never repeated, so it's a useful feature only in controlled use cases. Specifying the amount of wasted area is, on the other hand, just a way to mess up this feature; 99% the times, you either pull this number out of thin air, hoping it's right, or you try to do the right thing and you choose the wrong number anyway. Instead, we can use the CoglTextureFlags to control whether the texture should not be sliced (useful for Clutter-GST and for the texture-from-pixmap actors) and provide a reasonable value for enabling the slicing ourself. At some point, we might even provide a way to change the default at compile time or at run time, for particular platforms. Since max_waste is gone, the :tile-waste property of ClutterTexture becomes read-only, and it proxies the cogl_texture_get_max_waste() function. Inside Clutter, the only cases where the max_waste argument was not set to -1 are in the Pango glyph cache (which is a POT texture anyway) and inside the test cases where we want to force slicing; for the latter we can create larger textures that will be bigger than the threshold we set. Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com> Signed-off-by: Robert Bragg <robert@linux.intel.com> Signed-off-by: Neil Roberts <neil@linux.intel.com>
2009-05-23 14:18:18 -04:00
cogl_texture_new_from_bitmap (CoglHandle bmp_handle,
CoglTextureFlags flags,
CoglPixelFormat internal_format)
{
CoglHandle tex;
/* First try putting the texture in the atlas */
if ((tex = _cogl_atlas_texture_new_from_bitmap (bmp_handle,
flags,
internal_format)))
return tex;
/* If that doesn't work try a fast path 2D texture */
if ((tex = _cogl_texture_2d_new_from_bitmap (bmp_handle,
flags,
internal_format)))
return tex;
/* Otherwise create a sliced texture */
return _cogl_texture_2d_sliced_new_from_bitmap (bmp_handle,
flags,
internal_format);
}
CoglHandle
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
cogl_texture_new_from_file (const char *filename,
CoglTextureFlags flags,
CoglPixelFormat internal_format,
GError **error)
{
CoglHandle bmp_handle;
CoglBitmap *bmp;
CoglHandle handle = COGL_INVALID_HANDLE;
g_return_val_if_fail (error == NULL || *error == NULL, COGL_INVALID_HANDLE);
bmp_handle = cogl_bitmap_new_from_file (filename, error);
if (bmp_handle == COGL_INVALID_HANDLE)
return COGL_INVALID_HANDLE;
bmp = (CoglBitmap *) bmp_handle;
/* We know that the bitmap data is solely owned by this function so
we can do the premult conversion in place. This avoids having to
copy the bitmap which will otherwise happen in
_cogl_texture_prepare_for_upload */
internal_format = _cogl_texture_determine_internal_format (bmp->format,
internal_format);
if (!_cogl_texture_needs_premult_conversion (bmp->format, internal_format) ||
_cogl_bitmap_convert_premult_status (bmp, bmp->format ^ COGL_PREMULT_BIT))
handle = cogl_texture_new_from_bitmap (bmp, flags, internal_format);
cogl_handle_unref (bmp);
return handle;
}
CoglHandle
cogl_texture_new_from_foreign (GLuint gl_handle,
GLenum gl_target,
GLuint width,
GLuint height,
GLuint x_pot_waste,
GLuint y_pot_waste,
CoglPixelFormat format)
{
return _cogl_texture_2d_sliced_new_from_foreign (gl_handle,
gl_target,
width,
height,
x_pot_waste,
y_pot_waste,
format);
}
CoglHandle
cogl_texture_new_from_sub_texture (CoglHandle full_texture,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int sub_x,
int sub_y,
int sub_width,
int sub_height)
{
2010-01-18 04:22:04 -05:00
return _cogl_sub_texture_new (full_texture, sub_x, sub_y,
sub_width, sub_height);
}
CoglHandle
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
cogl_texture_new_from_buffer_EXP (CoglHandle buffer,
unsigned int width,
unsigned int height,
CoglTextureFlags flags,
CoglPixelFormat format,
CoglPixelFormat internal_format,
unsigned int rowstride,
const unsigned int offset)
{
CoglHandle texture;
CoglBuffer *cogl_buffer;
CoglPixelBuffer *pixel_buffer;
g_return_val_if_fail (cogl_is_buffer (buffer), COGL_INVALID_HANDLE);
if (format == COGL_PIXEL_FORMAT_ANY)
return COGL_INVALID_HANDLE;
cogl_buffer = COGL_BUFFER (buffer);
pixel_buffer = COGL_PIXEL_BUFFER (buffer);
/* Rowstride from CoglBuffer or even width * bpp if not given */
if (rowstride == 0)
rowstride = pixel_buffer->stride;
if (rowstride == 0)
rowstride = width * _cogl_get_format_bpp (format);
/* use the CoglBuffer height and width as last resort */
if (width == 0)
width = pixel_buffer->width;
if (height == 0)
height = pixel_buffer->height;
if (width == 0 || height == 0)
{
/* no width or height specified, neither at creation time (because the
* buffer was created by cogl_pixel_buffer_new()) nor when calling this
* function */
return COGL_INVALID_HANDLE;
}
#if !defined (COGL_HAS_GLES)
if (cogl_features_available (COGL_FEATURE_PBOS))
{
CoglBitmap bitmap;
/* Wrap the data into a bitmap */
bitmap.width = width;
bitmap.height = height;
bitmap.data = GUINT_TO_POINTER (offset);
bitmap.format = format;
bitmap.rowstride = rowstride;
_cogl_buffer_bind (cogl_buffer, GL_PIXEL_UNPACK_BUFFER);
texture = cogl_texture_new_from_bitmap (&bitmap, flags, internal_format);
_cogl_buffer_bind (NULL, GL_PIXEL_UNPACK_BUFFER);
}
else
#endif
{
texture = cogl_texture_new_from_data (width,
height,
flags,
format,
internal_format,
rowstride,
cogl_buffer->data);
}
return texture;
}
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int
cogl_texture_get_width (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return 0;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_width (tex);
}
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int
cogl_texture_get_height (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return 0;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_height (tex);
}
CoglPixelFormat
cogl_texture_get_format (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return COGL_PIXEL_FORMAT_ANY;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_format (tex);
}
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int
cogl_texture_get_rowstride (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return 0;
/* FIXME: This function should go away. It previously just returned
the rowstride that was used to upload the data as far as I can
tell. This is not helpful */
tex = COGL_TEXTURE (handle);
/* Just guess at a suitable rowstride */
return (_cogl_get_format_bpp (cogl_texture_get_format (tex))
* cogl_texture_get_width (tex));
}
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int
cogl_texture_get_max_waste (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return 0;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_max_waste (tex);
}
gboolean
cogl_texture_is_sliced (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return FALSE;
tex = COGL_TEXTURE (handle);
return tex->vtable->is_sliced (tex);
}
/* Some CoglTextures, notably sliced textures or atlas textures when repeating
* is used, will need to divide the coordinate space into multiple GL textures
* (or rather; in the case of atlases duplicate a single texture in multiple
* positions to handle repeating)
*
* This function helps you implement primitives using such textures by
* invoking a callback once for each sub texture that intersects a given
* region specified in texture coordinates.
*/
void
_cogl_texture_foreach_sub_texture_in_region (CoglHandle handle,
float virtual_tx_1,
float virtual_ty_1,
float virtual_tx_2,
float virtual_ty_2,
CoglTextureSliceCallback callback,
void *user_data)
{
CoglTexture *tex = COGL_TEXTURE (handle);
tex->vtable->foreach_sub_texture_in_region (tex,
virtual_tx_1,
virtual_ty_1,
virtual_tx_2,
virtual_ty_2,
callback,
user_data);
}
/* If this returns FALSE, that implies _foreach_sub_texture_in_region
* will be needed to iterate over multiple sub textures for regions whos
* texture coordinates extend out of the range [0,1]
*/
gboolean
_cogl_texture_can_hardware_repeat (CoglHandle handle)
{
CoglTexture *tex = (CoglTexture *)handle;
return tex->vtable->can_hardware_repeat (tex);
}
/* NB: You can't use this with textures comprised of multiple sub textures (use
* cogl_texture_is_sliced() to check) since coordinate transformation for such
* textures will be different for each slice. */
void
_cogl_texture_transform_coords_to_gl (CoglHandle handle,
float *s,
float *t)
{
CoglTexture *tex = COGL_TEXTURE (handle);
tex->vtable->transform_coords_to_gl (tex, s, t);
}
2010-01-18 04:22:04 -05:00
gboolean
_cogl_texture_transform_quad_coords_to_gl (CoglHandle handle,
float *coords)
{
CoglTexture *tex = COGL_TEXTURE (handle);
return tex->vtable->transform_quad_coords_to_gl (tex, coords);
}
GLenum
_cogl_texture_get_gl_format (CoglHandle handle)
{
CoglTexture *tex = COGL_TEXTURE (handle);
return tex->vtable->get_gl_format (tex);
}
gboolean
cogl_texture_get_gl_texture (CoglHandle handle,
GLuint *out_gl_handle,
GLenum *out_gl_target)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return FALSE;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_gl_texture (tex, out_gl_handle, out_gl_target);
}
[cogl] Move the texture filters to be a property of the material layer The texture filters are now a property of the material layer rather than the texture object. Whenever a texture is painted with a material it sets the filters on all of the GL textures in the Cogl texture. The filter is cached so that it won't be changed unnecessarily. The automatic mipmap generation has changed so that the mipmaps are only generated when the texture is painted instead of every time the data changes. Changing the texture sets a flag to mark that the mipmaps are dirty. This works better if the FBO extension is available because we can use glGenerateMipmap. If the extension is not available it will temporarily enable automatic mipmap generation and reupload the first pixel of each slice. This requires tracking the data for the first pixel. The COGL_TEXTURE_AUTO_MIPMAP flag has been replaced with COGL_TEXTURE_NO_AUTO_MIPMAP so that it will default to auto-mipmapping. The mipmap generation is now effectively free if you are not using a mipmap filter mode so you would only want to disable it if you had some special reason to generate your own mipmaps. ClutterTexture no longer has to store its own copy of the filter mode. Instead it stores it in the material and the property is directly set and read from that. This fixes problems with the filters getting out of sync when a cogl handle is set on the texture directly. It also avoids the mess of having to rerealize the texture if the filter quality changes to HIGH because Cogl will take of generating the mipmaps if needed.
2009-06-04 11:04:57 -04:00
void
_cogl_texture_set_filters (CoglHandle handle,
GLenum min_filter,
GLenum mag_filter)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
[cogl] Move the texture filters to be a property of the material layer The texture filters are now a property of the material layer rather than the texture object. Whenever a texture is painted with a material it sets the filters on all of the GL textures in the Cogl texture. The filter is cached so that it won't be changed unnecessarily. The automatic mipmap generation has changed so that the mipmaps are only generated when the texture is painted instead of every time the data changes. Changing the texture sets a flag to mark that the mipmaps are dirty. This works better if the FBO extension is available because we can use glGenerateMipmap. If the extension is not available it will temporarily enable automatic mipmap generation and reupload the first pixel of each slice. This requires tracking the data for the first pixel. The COGL_TEXTURE_AUTO_MIPMAP flag has been replaced with COGL_TEXTURE_NO_AUTO_MIPMAP so that it will default to auto-mipmapping. The mipmap generation is now effectively free if you are not using a mipmap filter mode so you would only want to disable it if you had some special reason to generate your own mipmaps. ClutterTexture no longer has to store its own copy of the filter mode. Instead it stores it in the material and the property is directly set and read from that. This fixes problems with the filters getting out of sync when a cogl handle is set on the texture directly. It also avoids the mess of having to rerealize the texture if the filter quality changes to HIGH because Cogl will take of generating the mipmaps if needed.
2009-06-04 11:04:57 -04:00
return;
tex = COGL_TEXTURE (handle);
tex->vtable->set_filters (tex, min_filter, mag_filter);
}
void
[cogl] Move the texture filters to be a property of the material layer The texture filters are now a property of the material layer rather than the texture object. Whenever a texture is painted with a material it sets the filters on all of the GL textures in the Cogl texture. The filter is cached so that it won't be changed unnecessarily. The automatic mipmap generation has changed so that the mipmaps are only generated when the texture is painted instead of every time the data changes. Changing the texture sets a flag to mark that the mipmaps are dirty. This works better if the FBO extension is available because we can use glGenerateMipmap. If the extension is not available it will temporarily enable automatic mipmap generation and reupload the first pixel of each slice. This requires tracking the data for the first pixel. The COGL_TEXTURE_AUTO_MIPMAP flag has been replaced with COGL_TEXTURE_NO_AUTO_MIPMAP so that it will default to auto-mipmapping. The mipmap generation is now effectively free if you are not using a mipmap filter mode so you would only want to disable it if you had some special reason to generate your own mipmaps. ClutterTexture no longer has to store its own copy of the filter mode. Instead it stores it in the material and the property is directly set and read from that. This fixes problems with the filters getting out of sync when a cogl handle is set on the texture directly. It also avoids the mess of having to rerealize the texture if the filter quality changes to HIGH because Cogl will take of generating the mipmaps if needed.
2009-06-04 11:04:57 -04:00
_cogl_texture_ensure_mipmaps (CoglHandle handle)
{
CoglTexture *tex;
[cogl] Move the texture filters to be a property of the material layer The texture filters are now a property of the material layer rather than the texture object. Whenever a texture is painted with a material it sets the filters on all of the GL textures in the Cogl texture. The filter is cached so that it won't be changed unnecessarily. The automatic mipmap generation has changed so that the mipmaps are only generated when the texture is painted instead of every time the data changes. Changing the texture sets a flag to mark that the mipmaps are dirty. This works better if the FBO extension is available because we can use glGenerateMipmap. If the extension is not available it will temporarily enable automatic mipmap generation and reupload the first pixel of each slice. This requires tracking the data for the first pixel. The COGL_TEXTURE_AUTO_MIPMAP flag has been replaced with COGL_TEXTURE_NO_AUTO_MIPMAP so that it will default to auto-mipmapping. The mipmap generation is now effectively free if you are not using a mipmap filter mode so you would only want to disable it if you had some special reason to generate your own mipmaps. ClutterTexture no longer has to store its own copy of the filter mode. Instead it stores it in the material and the property is directly set and read from that. This fixes problems with the filters getting out of sync when a cogl handle is set on the texture directly. It also avoids the mess of having to rerealize the texture if the filter quality changes to HIGH because Cogl will take of generating the mipmaps if needed.
2009-06-04 11:04:57 -04:00
if (!cogl_is_texture (handle))
return;
tex = COGL_TEXTURE (handle);
tex->vtable->ensure_mipmaps (tex);
}
2010-01-18 04:22:04 -05:00
void
_cogl_texture_ensure_non_quad_rendering (CoglHandle handle)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return;
tex = COGL_TEXTURE (handle);
tex->vtable->ensure_non_quad_rendering (tex);
2010-01-18 04:22:04 -05:00
}
gboolean
cogl_texture_set_region (CoglHandle handle,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int src_x,
int src_y,
int dst_x,
int dst_y,
unsigned int dst_width,
unsigned int dst_height,
int width,
int height,
CoglPixelFormat format,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int rowstride,
const guint8 *data)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return FALSE;
tex = COGL_TEXTURE (handle);
return tex->vtable->set_region (tex,
src_x, src_y,
dst_x, dst_y,
dst_width, dst_height,
width, height,
format,
rowstride,
data);
}
/* Reads back the contents of a texture by rendering it to the framebuffer
* and reading back the resulting pixels.
*
* It will perform multiple renders if the texture is larger than the
* current glViewport.
*
* It assumes the projection and modelview have already been setup so
* that rendering to 0,0 with the same width and height of the viewport
* will exactly cover the viewport.
*
* NB: Normally this approach isn't normally used since we can just use
* glGetTexImage, but may be used as a fallback in some circumstances.
*/
static void
do_texture_draw_and_read (CoglHandle handle,
CoglBitmap *target_bmp,
GLint *viewport)
{
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int bpp;
float rx1, ry1;
float rx2, ry2;
float tx1, ty1;
float tx2, ty2;
int bw, bh;
CoglBitmap rect_bmp;
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int tex_width, tex_height;
bpp = _cogl_get_format_bpp (COGL_PIXEL_FORMAT_RGBA_8888);
tex_width = cogl_texture_get_width (handle);
tex_height = cogl_texture_get_height (handle);
ry2 = 0;
ty2 = 0;
/* Walk Y axis until whole bitmap height consumed */
for (bh = tex_height; bh > 0; bh -= viewport[3])
{
/* Rectangle Y coords */
ry1 = ry2;
ry2 += (bh < viewport[3]) ? bh : viewport[3];
/* Normalized texture Y coords */
ty1 = ty2;
ty2 = (ry2 / (float) tex_height);
rx2 = 0;
tx2 = 0;
/* Walk X axis until whole bitmap width consumed */
for (bw = tex_width; bw > 0; bw-=viewport[2])
{
/* Rectangle X coords */
rx1 = rx2;
rx2 += (bw < viewport[2]) ? bw : viewport[2];
/* Normalized texture X coords */
tx1 = tx2;
tx2 = (rx2 / (float) tex_width);
/* Draw a portion of texture */
cogl_rectangle_with_texture_coords (0, 0,
rx2 - rx1,
ry2 - ry1,
tx1, ty1,
tx2, ty2);
/* Read into a temporary bitmap */
rect_bmp.format = COGL_PIXEL_FORMAT_RGBA_8888;
rect_bmp.width = rx2 - rx1;
rect_bmp.height = ry2 - ry1;
rect_bmp.rowstride = bpp * rect_bmp.width;
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
rect_bmp.data = g_malloc (rect_bmp.rowstride *
rect_bmp.height);
_cogl_texture_driver_prep_gl_for_pixels_download (rect_bmp.rowstride,
bpp);
GE( glReadPixels (viewport[0], viewport[1],
rect_bmp.width,
rect_bmp.height,
GL_RGBA, GL_UNSIGNED_BYTE,
rect_bmp.data) );
/* Copy to target bitmap */
_cogl_bitmap_copy_subregion (&rect_bmp,
target_bmp,
0,0,
rx1,ry1,
rect_bmp.width,
rect_bmp.height);
/* Free temp bitmap */
g_free (rect_bmp.data);
}
}
}
/* Reads back the contents of a texture by rendering it to the framebuffer
* and reading back the resulting pixels.
*
* NB: Normally this approach isn't normally used since we can just use
* glGetTexImage, but may be used as a fallback in some circumstances.
*/
gboolean
_cogl_texture_draw_and_read (CoglHandle handle,
CoglBitmap *target_bmp,
GLuint target_gl_format,
GLuint target_gl_type)
{
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int bpp;
CoglHandle framebuffer;
[draw-buffers] First pass at overhauling Cogl's framebuffer management Cogl's support for offscreen rendering was originally written just to support the clutter_texture_new_from_actor API and due to lack of documentation and several confusing - non orthogonal - side effects of using the API it wasn't really possible to use directly. This commit does a number of things: - It removes {gl,gles}/cogl-fbo.{c,h} and adds shared cogl-draw-buffer.{c,h} files instead which should be easier to maintain. - internally CoglFbo objects are now called CoglDrawBuffers. A CoglDrawBuffer is an abstract base class that is inherited from to implement CoglOnscreen and CoglOffscreen draw buffers. CoglOffscreen draw buffers will initially be used to support the cogl_offscreen_new_to_texture API, and CoglOnscreen draw buffers will start to be used internally to represent windows as we aim to migrate some of Clutter's backend code to Cogl. - It makes draw buffer objects the owners of the following state: - viewport - projection matrix stack - modelview matrix stack - clip state (This means when you switch between draw buffers you will automatically be switching to their associated viewport, matrix and clip state) Aside from hopefully making cogl_offscreen_new_to_texture be more useful short term by having simpler and well defined semantics for cogl_set_draw_buffer, as mentioned above this is the first step for a couple of other things: - Its a step toward moving ownership for windows down from Clutter backends into Cogl, by (internally at least) introducing the CoglOnscreen draw buffer. Note: the plan is that cogl_set_draw_buffer will accept on or offscreen draw buffer handles, and the "target" argument will become redundant since we will instead query the type of the given draw buffer handle. - Because we have a common type for on and offscreen framebuffers we can provide a unified API for framebuffer management. Things like: - blitting between buffers - managing ancillary buffers (e.g. attaching depth and stencil buffers) - size requisition - clearing
2009-09-25 09:34:34 -04:00
int viewport[4];
CoglBitmap alpha_bmp;
CoglHandle prev_source;
CoglMatrixStack *projection_stack;
CoglMatrixStack *modelview_stack;
_COGL_GET_CONTEXT (ctx, FALSE);
bpp = _cogl_get_format_bpp (COGL_PIXEL_FORMAT_RGBA_8888);
framebuffer = _cogl_get_framebuffer ();
/* Viewport needs to have some size and be inside the window for this */
_cogl_framebuffer_get_viewport4fv (framebuffer, viewport);
if (viewport[0] < 0 || viewport[1] < 0 ||
viewport[2] <= 0 || viewport[3] <= 0)
return FALSE;
/* Setup orthographic projection into current viewport (0,0 in bottom-left
* corner to draw the texture upside-down so we match the way glReadPixels
* works)
*/
projection_stack = _cogl_framebuffer_get_projection_stack (framebuffer);
[draw-buffers] First pass at overhauling Cogl's framebuffer management Cogl's support for offscreen rendering was originally written just to support the clutter_texture_new_from_actor API and due to lack of documentation and several confusing - non orthogonal - side effects of using the API it wasn't really possible to use directly. This commit does a number of things: - It removes {gl,gles}/cogl-fbo.{c,h} and adds shared cogl-draw-buffer.{c,h} files instead which should be easier to maintain. - internally CoglFbo objects are now called CoglDrawBuffers. A CoglDrawBuffer is an abstract base class that is inherited from to implement CoglOnscreen and CoglOffscreen draw buffers. CoglOffscreen draw buffers will initially be used to support the cogl_offscreen_new_to_texture API, and CoglOnscreen draw buffers will start to be used internally to represent windows as we aim to migrate some of Clutter's backend code to Cogl. - It makes draw buffer objects the owners of the following state: - viewport - projection matrix stack - modelview matrix stack - clip state (This means when you switch between draw buffers you will automatically be switching to their associated viewport, matrix and clip state) Aside from hopefully making cogl_offscreen_new_to_texture be more useful short term by having simpler and well defined semantics for cogl_set_draw_buffer, as mentioned above this is the first step for a couple of other things: - Its a step toward moving ownership for windows down from Clutter backends into Cogl, by (internally at least) introducing the CoglOnscreen draw buffer. Note: the plan is that cogl_set_draw_buffer will accept on or offscreen draw buffer handles, and the "target" argument will become redundant since we will instead query the type of the given draw buffer handle. - Because we have a common type for on and offscreen framebuffers we can provide a unified API for framebuffer management. Things like: - blitting between buffers - managing ancillary buffers (e.g. attaching depth and stencil buffers) - size requisition - clearing
2009-09-25 09:34:34 -04:00
_cogl_matrix_stack_push (projection_stack);
_cogl_matrix_stack_load_identity (projection_stack);
_cogl_matrix_stack_ortho (projection_stack,
0, (float)(viewport[2]),
0, (float)(viewport[3]),
(float)(0),
(float)(100));
modelview_stack = _cogl_framebuffer_get_modelview_stack (framebuffer);
[draw-buffers] First pass at overhauling Cogl's framebuffer management Cogl's support for offscreen rendering was originally written just to support the clutter_texture_new_from_actor API and due to lack of documentation and several confusing - non orthogonal - side effects of using the API it wasn't really possible to use directly. This commit does a number of things: - It removes {gl,gles}/cogl-fbo.{c,h} and adds shared cogl-draw-buffer.{c,h} files instead which should be easier to maintain. - internally CoglFbo objects are now called CoglDrawBuffers. A CoglDrawBuffer is an abstract base class that is inherited from to implement CoglOnscreen and CoglOffscreen draw buffers. CoglOffscreen draw buffers will initially be used to support the cogl_offscreen_new_to_texture API, and CoglOnscreen draw buffers will start to be used internally to represent windows as we aim to migrate some of Clutter's backend code to Cogl. - It makes draw buffer objects the owners of the following state: - viewport - projection matrix stack - modelview matrix stack - clip state (This means when you switch between draw buffers you will automatically be switching to their associated viewport, matrix and clip state) Aside from hopefully making cogl_offscreen_new_to_texture be more useful short term by having simpler and well defined semantics for cogl_set_draw_buffer, as mentioned above this is the first step for a couple of other things: - Its a step toward moving ownership for windows down from Clutter backends into Cogl, by (internally at least) introducing the CoglOnscreen draw buffer. Note: the plan is that cogl_set_draw_buffer will accept on or offscreen draw buffer handles, and the "target" argument will become redundant since we will instead query the type of the given draw buffer handle. - Because we have a common type for on and offscreen framebuffers we can provide a unified API for framebuffer management. Things like: - blitting between buffers - managing ancillary buffers (e.g. attaching depth and stencil buffers) - size requisition - clearing
2009-09-25 09:34:34 -04:00
_cogl_matrix_stack_push (modelview_stack);
_cogl_matrix_stack_load_identity (modelview_stack);
/* Direct copy operation */
if (ctx->texture_download_material == COGL_INVALID_HANDLE)
{
ctx->texture_download_material = cogl_material_new ();
cogl_material_set_blend (ctx->texture_download_material,
"RGBA = ADD (SRC_COLOR, 0)",
NULL);
}
prev_source = cogl_handle_ref (ctx->source_material);
cogl_set_source (ctx->texture_download_material);
cogl_material_set_layer (ctx->texture_download_material, 0, handle);
cogl_material_set_layer_combine (ctx->texture_download_material,
0, /* layer */
"RGBA = REPLACE (TEXTURE)",
NULL);
do_texture_draw_and_read (handle, target_bmp, viewport);
/* Check whether texture has alpha and framebuffer not */
/* FIXME: For some reason even if ALPHA_BITS is 8, the framebuffer
still doesn't seem to have an alpha buffer. This might be just
a PowerVR issue.
GLint r_bits, g_bits, b_bits, a_bits;
GE( glGetIntegerv (GL_ALPHA_BITS, &a_bits) );
GE( glGetIntegerv (GL_RED_BITS, &r_bits) );
GE( glGetIntegerv (GL_GREEN_BITS, &g_bits) );
GE( glGetIntegerv (GL_BLUE_BITS, &b_bits) );
printf ("R bits: %d\n", r_bits);
printf ("G bits: %d\n", g_bits);
printf ("B bits: %d\n", b_bits);
printf ("A bits: %d\n", a_bits); */
if ((cogl_texture_get_format (handle) & COGL_A_BIT)/* && a_bits == 0*/)
{
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
guint8 *srcdata;
guint8 *dstdata;
guint8 *srcpixel;
guint8 *dstpixel;
int x,y;
/* Create temp bitmap for alpha values */
alpha_bmp.format = COGL_PIXEL_FORMAT_RGBA_8888;
alpha_bmp.width = target_bmp->width;
alpha_bmp.height = target_bmp->height;
alpha_bmp.rowstride = bpp * alpha_bmp.width;
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
alpha_bmp.data = g_malloc (alpha_bmp.rowstride *
alpha_bmp.height);
/* Draw alpha values into RGB channels */
cogl_material_set_layer_combine (ctx->texture_download_material,
0, /* layer */
"RGBA = REPLACE (TEXTURE[A])",
NULL);
do_texture_draw_and_read (handle, &alpha_bmp, viewport);
/* Copy temp R to target A */
srcdata = alpha_bmp.data;
dstdata = target_bmp->data;
for (y=0; y<target_bmp->height; ++y)
{
for (x=0; x<target_bmp->width; ++x)
{
srcpixel = srcdata + x*bpp;
dstpixel = dstdata + x*bpp;
dstpixel[3] = srcpixel[0];
}
srcdata += alpha_bmp.rowstride;
dstdata += target_bmp->rowstride;
}
g_free (alpha_bmp.data);
}
/* Restore old state */
_cogl_matrix_stack_pop (modelview_stack);
_cogl_matrix_stack_pop (projection_stack);
/* restore the original material */
cogl_set_source (prev_source);
cogl_handle_unref (prev_source);
return TRUE;
}
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
int
cogl_texture_get_data (CoglHandle handle,
CoglPixelFormat format,
cogl: improves header and coding style consistency We've had complaints that our Cogl code/headers are a bit "special" so this is a first pass at tidying things up by giving them some consistency. These changes are all consistent with how new code in Cogl is being written, but the style isn't consistently applied across all code yet. There are two parts to this patch; but since each one required a large amount of effort to maintain tidy indenting it made sense to combine the changes to reduce the time spent re indenting the same lines. The first change is to use a consistent style for declaring function prototypes in headers. Cogl headers now consistently use this style for prototypes: return_type cogl_function_name (CoglType arg0, CoglType arg1); Not everyone likes this style, but it seems that most of the currently active Cogl developers agree on it. The second change is to constrain the use of redundant glib data types in Cogl. Uses of gint, guint, gfloat, glong, gulong and gchar have all been replaced with int, unsigned int, float, long, unsigned long and char respectively. When talking about pixel data; use of guchar has been replaced with guint8, otherwise unsigned char can be used. The glib types that we continue to use for portability are gboolean, gint{8,16,32,64}, guint{8,16,32,64} and gsize. The general intention is that Cogl should look palatable to the widest range of C programmers including those outside the Gnome community so - especially for the public API - we want to minimize the number of foreign looking typedefs.
2010-02-09 20:57:32 -05:00
unsigned int rowstride,
guint8 *data)
{
CoglTexture *tex;
if (!cogl_is_texture (handle))
return FALSE;
tex = COGL_TEXTURE (handle);
return tex->vtable->get_data (handle, format, rowstride, data);
}