mutter/cogl/cogl-texture-2d-sliced.c

1477 lines
51 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/>.
*
*
*
* 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-debug.h"
#include "cogl-private.h"
#include "cogl-util.h"
#include "cogl-bitmap.h"
#include "cogl-bitmap-private.h"
#include "cogl-texture-private.h"
#include "cogl-texture-2d-gl.h"
#include "cogl-texture-2d-private.h"
#include "cogl-texture-2d-sliced-private.h"
#include "cogl-texture-gl-private.h"
#include "cogl-texture-driver.h"
#include "cogl-context-private.h"
#include "cogl-object-private.h"
#include "cogl-spans.h"
#include "cogl-journal-private.h"
cogl: rename CoglMaterial -> CoglPipeline This applies an API naming change that's been deliberated over for a while now which is to rename CoglMaterial to CoglPipeline. For now the new pipeline API is marked as experimental and public headers continue to talk about materials not pipelines. The CoglMaterial API is now maintained in terms of the cogl_pipeline API internally. Currently this API is targeting Cogl 2.0 so we will have time to integrate it properly with other upcoming Cogl 2.0 work. The basic reasons for the rename are: - That the term "material" implies to many people that they are constrained to fragment processing; perhaps as some kind of high-level texture abstraction. - In Clutter they get exposed by ClutterTexture actors which may be re-inforcing this misconception. - When comparing how other frameworks use the term material, a material sometimes describes a multi-pass fragment processing technique which isn't the case in Cogl. - In code, "CoglPipeline" will hopefully be a much more self documenting summary of what these objects represent; a full GPU pipeline configuration including, for example, vertex processing, fragment processing and blending. - When considering the API documentation story, at some point we need a document introducing developers to how the "GPU pipeline" works so it should become intuitive that CoglPipeline maps back to that description of the GPU pipeline. - This is consistent in terminology and concept to OpenGL 4's new pipeline object which is a container for program objects. Note: The cogl-material.[ch] files have been renamed to cogl-material-compat.[ch] because otherwise git doesn't seem to treat the change as a moving the old cogl-material.c->cogl-pipeline.c and so we loose all our git-blame history.
2010-10-27 13:54:57 -04:00
#include "cogl-pipeline-opengl-private.h"
#include "cogl-primitive-texture.h"
Adds CoglError api Although we use GLib internally in Cogl we would rather not leak GLib api through Cogl's own api, except through explicitly namespaced cogl_glib_ / cogl_gtype_ feature apis. One of the benefits we see to not leaking GLib through Cogl's public API is that documentation for Cogl won't need to first introduce the Glib API to newcomers, thus hopefully lowering the barrier to learning Cogl. This patch provides a Cogl specific typedef for reporting runtime errors which by no coincidence matches the typedef for GError exactly. If Cogl is built with --enable-glib (default) then developers can even safely assume that a CoglError is a GError under the hood. This patch also enforces a consistent policy for when NULL is passed as an error argument and an error is thrown. In this case we log the error and abort the application, instead of silently ignoring it. In common cases where nothing has been implemented to handle a particular error and/or where applications are just printing the error and aborting themselves then this saves some typing. This also seems more consistent with language based exceptions which usually cause a program to abort if they are not explicitly caught (which passing a non-NULL error signifies in this case) Since this policy for NULL error pointers is stricter than the standard GError convention, there is a clear note in the documentation to warn developers that are used to using the GError api. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit b068d5ea09ab32c37e8c965fc8582c85d1b2db46) Note: Since we can't change the Cogl 1.x api the patch was changed to not rename _error_quark() functions to be _error_domain() functions and although it's a bit ugly, instead of providing our own CoglError type that's compatible with GError we simply #define CoglError to GError unless Cogl is built with glib disabled. Note: this patch does technically introduce an API break since it drops the cogl_error_get_type() symbol generated by glib-mkenum (Since the CoglError enum was replaced by a CoglSystemError enum) but for now we are assuming that this will not affect anyone currently using the Cogl API. If this does turn out to be a problem in practice then we would be able to fix this my manually copying an implementation of cogl_error_get_type() generated by glib-mkenum into a compatibility source file and we could also define the original COGL_ERROR_ enums for compatibility too. Note: another minor concern with cherry-picking this patch to the 1.14 branch is that an api scanner would be lead to believe that some APIs have changed, and for example the gobject-introspection parser which understands the semantics of GError will not understand the semantics of CoglError. We expect most people that have tried to use gobject-introspection with Cogl already understand though that it is not well suited to generating bindings of the Cogl api anyway and we aren't aware or anyone depending on such bindings for apis involving GErrors. (GnomeShell only makes very-very minimal use of Cogl via the gjs bindings for the cogl_rectangle and cogl_color apis.) The main reason we have cherry-picked this patch to the 1.14 branch even given the above concerns is that without it it would become very awkward for us to cherry-pick other beneficial patches from master.
2012-08-31 14:28:27 -04:00
#include "cogl-error-private.h"
#include "cogl-texture-gl-private.h"
#include <string.h>
#include <stdlib.h>
#include <math.h>
static void _cogl_texture_2d_sliced_free (CoglTexture2DSliced *tex_2ds);
COGL_TEXTURE_DEFINE (Texture2DSliced, texture_2d_sliced);
static const CoglTextureVtable cogl_texture_2d_sliced_vtable;
typedef struct _ForeachData
{
CoglMetaTextureCallback callback;
void *user_data;
float x_normalize_factor;
float y_normalize_factor;
} ForeachData;
static void
re_normalize_sub_texture_coords_cb (CoglTexture *sub_texture,
const float *sub_texture_coords,
const float *meta_coords,
void *user_data)
{
ForeachData *data = user_data;
/* The coordinates passed to the span iterating code were
* un-normalized so we need to renormalize them before passing them
* on */
float re_normalized_coords[4] =
{
meta_coords[0] * data->x_normalize_factor,
meta_coords[1] * data->y_normalize_factor,
meta_coords[2] * data->x_normalize_factor,
meta_coords[3] * data->y_normalize_factor
};
data->callback (sub_texture, sub_texture_coords, re_normalized_coords,
data->user_data);
}
static void
_cogl_texture_2d_sliced_foreach_sub_texture_in_region (
CoglTexture *tex,
float virtual_tx_1,
float virtual_ty_1,
float virtual_tx_2,
float virtual_ty_2,
CoglMetaTextureCallback callback,
void *user_data)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglSpan *x_spans = (CoglSpan *)tex_2ds->slice_x_spans->data;
CoglSpan *y_spans = (CoglSpan *)tex_2ds->slice_y_spans->data;
CoglTexture **textures = (CoglTexture **)tex_2ds->slice_textures->data;
float un_normalized_coords[4];
ForeachData data;
/* NB: its convenient for us to store non-normalized coordinates in
* our CoglSpans but that means we need to un-normalize the incoming
* virtual coordinates and make sure we re-normalize the coordinates
* before calling the given callback.
*/
data.callback = callback;
data.user_data = user_data;
data.x_normalize_factor = 1.0f / tex->width;
data.y_normalize_factor = 1.0f / tex->height;
un_normalized_coords[0] = virtual_tx_1 * tex->width;
un_normalized_coords[1] = virtual_ty_1 * tex->height;
un_normalized_coords[2] = virtual_tx_2 * tex->width;
un_normalized_coords[3] = virtual_ty_2 * tex->height;
/* Note that the normalize factors passed here are the reciprocal of
* the factors calculated above because the span iterating code
* normalizes by dividing by the factor instead of multiplying */
_cogl_texture_spans_foreach_in_region (x_spans,
tex_2ds->slice_x_spans->len,
y_spans,
tex_2ds->slice_y_spans->len,
textures,
un_normalized_coords,
tex->width,
tex->height,
COGL_PIPELINE_WRAP_MODE_REPEAT,
COGL_PIPELINE_WRAP_MODE_REPEAT,
re_normalize_sub_texture_coords_cb,
&data);
}
static uint8_t *
_cogl_texture_2d_sliced_allocate_waste_buffer (CoglTexture2DSliced *tex_2ds,
CoglPixelFormat format)
{
CoglSpan *last_x_span;
CoglSpan *last_y_span;
uint8_t *waste_buf = NULL;
/* If the texture has any waste then allocate a buffer big enough to
fill the gaps */
last_x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan,
tex_2ds->slice_x_spans->len - 1);
last_y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan,
tex_2ds->slice_y_spans->len - 1);
if (last_x_span->waste > 0 || last_y_span->waste > 0)
{
int bpp = _cogl_pixel_format_get_bytes_per_pixel (format);
CoglSpan *first_x_span
= &g_array_index (tex_2ds->slice_x_spans, CoglSpan, 0);
CoglSpan *first_y_span
= &g_array_index (tex_2ds->slice_y_spans, CoglSpan, 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
unsigned int right_size = first_y_span->size * last_x_span->waste;
unsigned int bottom_size = first_x_span->size * last_y_span->waste;
waste_buf = g_malloc (MAX (right_size, bottom_size) * bpp);
}
return waste_buf;
}
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
static CoglBool
_cogl_texture_2d_sliced_set_waste (CoglTexture2DSliced *tex_2ds,
CoglBitmap *source_bmp,
CoglTexture2D *slice_tex,
uint8_t *waste_buf,
CoglSpan *x_span,
CoglSpan *y_span,
CoglSpanIter *x_iter,
CoglSpanIter *y_iter,
int src_x,
int src_y,
int dst_x,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
int dst_y,
CoglError **error)
{
CoglBool need_x, need_y;
CoglContext *ctx = COGL_TEXTURE (tex_2ds)->context;
/* If the x_span is sliced and the upload touches the
rightmost pixels then fill the waste with copies of the
pixels */
need_x = x_span->waste > 0 &&
x_iter->intersect_end - x_iter->pos >= x_span->size - x_span->waste;
/* same for the bottom-most pixels */
need_y = y_span->waste > 0 &&
y_iter->intersect_end - y_iter->pos >= y_span->size - y_span->waste;
if (need_x || need_y)
{
int bmp_rowstride = cogl_bitmap_get_rowstride (source_bmp);
CoglPixelFormat source_format = cogl_bitmap_get_format (source_bmp);
int bpp = _cogl_pixel_format_get_bytes_per_pixel (source_format);
uint8_t *bmp_data;
const uint8_t *src;
uint8_t *dst;
unsigned int wy, wx;
CoglBitmap *waste_bmp;
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
bmp_data = _cogl_bitmap_map (source_bmp, COGL_BUFFER_ACCESS_READ, 0, error);
if (bmp_data == NULL)
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
return FALSE;
if (need_x)
{
src = (bmp_data + ((src_y + (int) y_iter->intersect_start - dst_y) *
bmp_rowstride) +
(src_x + (int)x_span->start + (int)x_span->size -
(int)x_span->waste - dst_x - 1) * bpp);
dst = waste_buf;
for (wy = 0;
wy < y_iter->intersect_end - y_iter->intersect_start;
wy++)
{
for (wx = 0; wx < x_span->waste; wx++)
{
memcpy (dst, src, bpp);
dst += bpp;
}
src += bmp_rowstride;
}
waste_bmp = cogl_bitmap_new_for_data (ctx,
x_span->waste,
y_iter->intersect_end -
y_iter->intersect_start,
source_format,
x_span->waste * bpp,
waste_buf);
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (!_cogl_texture_set_region_from_bitmap (COGL_TEXTURE (slice_tex),
0, /* src_x */
0, /* src_y */
x_span->waste, /* width */
/* height */
y_iter->intersect_end -
y_iter->intersect_start,
waste_bmp,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
/* dst_x */
x_span->size - x_span->waste,
y_iter->intersect_start -
y_span->start, /* dst_y */
0, /* level */
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
error))
{
cogl_object_unref (waste_bmp);
_cogl_bitmap_unmap (source_bmp);
return FALSE;
}
cogl_object_unref (waste_bmp);
}
if (need_y)
{
unsigned int copy_width, intersect_width;
src = (bmp_data + ((src_x + (int) x_iter->intersect_start - dst_x) *
bpp) +
(src_y + (int)y_span->start + (int)y_span->size -
(int)y_span->waste - dst_y - 1) * bmp_rowstride);
dst = waste_buf;
if (x_iter->intersect_end - x_iter->pos
>= x_span->size - x_span->waste)
copy_width = x_span->size + x_iter->pos - x_iter->intersect_start;
else
copy_width = x_iter->intersect_end - x_iter->intersect_start;
intersect_width = x_iter->intersect_end - x_iter->intersect_start;
for (wy = 0; wy < y_span->waste; wy++)
{
memcpy (dst, src, intersect_width * bpp);
dst += intersect_width * bpp;
for (wx = intersect_width; wx < copy_width; wx++)
{
memcpy (dst, dst - bpp, bpp);
dst += bpp;
}
}
waste_bmp = cogl_bitmap_new_for_data (ctx,
copy_width,
y_span->waste,
source_format,
copy_width * bpp,
waste_buf);
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (!_cogl_texture_set_region_from_bitmap (COGL_TEXTURE (slice_tex),
0, /* src_x */
0, /* src_y */
copy_width, /* width */
y_span->waste, /* height */
waste_bmp,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
/* dst_x */
x_iter->intersect_start -
x_iter->pos,
/* dst_y */
y_span->size - y_span->waste,
0, /* level */
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
error))
{
cogl_object_unref (waste_bmp);
_cogl_bitmap_unmap (source_bmp);
return FALSE;
}
cogl_object_unref (waste_bmp);
}
_cogl_bitmap_unmap (source_bmp);
}
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
return TRUE;
}
static CoglBool
_cogl_texture_2d_sliced_upload_bitmap (CoglTexture2DSliced *tex_2ds,
CoglBitmap *bmp,
CoglError **error)
{
CoglSpan *x_span;
CoglSpan *y_span;
CoglTexture2D *slice_tex;
int x, y;
uint8_t *waste_buf;
CoglPixelFormat bmp_format;
bmp_format = cogl_bitmap_get_format (bmp);
waste_buf = _cogl_texture_2d_sliced_allocate_waste_buffer (tex_2ds,
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
bmp_format);
/* Iterate vertical slices */
for (y = 0; y < tex_2ds->slice_y_spans->len; ++y)
{
y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan, y);
/* Iterate horizontal slices */
for (x = 0; x < tex_2ds->slice_x_spans->len; ++x)
{
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 slice_num = y * tex_2ds->slice_x_spans->len + x;
CoglSpanIter x_iter, y_iter;
x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan, x);
/* Pick the gl texture object handle */
slice_tex = g_array_index (tex_2ds->slice_textures,
CoglTexture2D *, slice_num);
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (!_cogl_texture_set_region_from_bitmap (COGL_TEXTURE (slice_tex),
x_span->start, /* src x */
y_span->start, /* src y */
x_span->size -
x_span->waste, /* width */
y_span->size -
y_span->waste, /* height */
bmp,
0, /* dst x */
0, /* dst y */
0, /* level */
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
error))
{
if (waste_buf)
g_free (waste_buf);
return FALSE;
}
/* Set up a fake iterator that covers the whole slice */
x_iter.intersect_start = x_span->start;
x_iter.intersect_end = (x_span->start +
x_span->size -
x_span->waste);
x_iter.pos = x_span->start;
y_iter.intersect_start = y_span->start;
y_iter.intersect_end = (y_span->start +
y_span->size -
y_span->waste);
y_iter.pos = y_span->start;
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (!_cogl_texture_2d_sliced_set_waste (tex_2ds,
bmp,
slice_tex,
waste_buf,
x_span, y_span,
&x_iter, &y_iter,
0, /* src_x */
0, /* src_y */
0, /* dst_x */
0,
error)) /* dst_y */
{
if (waste_buf)
g_free (waste_buf);
return FALSE;
}
}
}
if (waste_buf)
g_free (waste_buf);
return TRUE;
}
static CoglBool
_cogl_texture_2d_sliced_upload_subregion (CoglTexture2DSliced *tex_2ds,
int src_x,
int src_y,
int dst_x,
int dst_y,
int width,
int height,
CoglBitmap *source_bmp,
CoglError **error)
{
CoglTexture *tex = COGL_TEXTURE (tex_2ds);
CoglSpan *x_span;
CoglSpan *y_span;
CoglSpanIter x_iter;
CoglSpanIter y_iter;
CoglTexture2D *slice_tex;
int source_x = 0, source_y = 0;
int inter_w = 0, inter_h = 0;
int local_x = 0, local_y = 0;
uint8_t *waste_buf;
CoglPixelFormat source_format;
source_format = cogl_bitmap_get_format (source_bmp);
waste_buf =
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
_cogl_texture_2d_sliced_allocate_waste_buffer (tex_2ds, source_format);
/* Iterate vertical spans */
for (source_y = src_y,
_cogl_span_iter_begin (&y_iter,
(CoglSpan *)tex_2ds->slice_y_spans->data,
tex_2ds->slice_y_spans->len,
tex->height,
dst_y,
dst_y + height,
COGL_PIPELINE_WRAP_MODE_REPEAT);
!_cogl_span_iter_end (&y_iter);
_cogl_span_iter_next (&y_iter),
source_y += inter_h )
{
y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan,
y_iter.index);
/* Iterate horizontal spans */
for (source_x = src_x,
_cogl_span_iter_begin (&x_iter,
(CoglSpan *)tex_2ds->slice_x_spans->data,
tex_2ds->slice_x_spans->len,
tex->width,
dst_x,
dst_x + width,
COGL_PIPELINE_WRAP_MODE_REPEAT);
!_cogl_span_iter_end (&x_iter);
_cogl_span_iter_next (&x_iter),
source_x += inter_w )
{
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 slice_num;
x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan,
x_iter.index);
/* Pick intersection width and height */
inter_w = (x_iter.intersect_end - x_iter.intersect_start);
inter_h = (y_iter.intersect_end - y_iter.intersect_start);
/* Localize intersection top-left corner to slice*/
local_x = (x_iter.intersect_start - x_iter.pos);
local_y = (y_iter.intersect_start - y_iter.pos);
slice_num = y_iter.index * tex_2ds->slice_x_spans->len + x_iter.index;
/* Pick slice texture */
slice_tex = g_array_index (tex_2ds->slice_textures,
CoglTexture2D *, slice_num);
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (!_cogl_texture_set_region_from_bitmap (COGL_TEXTURE (slice_tex),
source_x,
source_y,
inter_w, /* width */
inter_h, /* height */
source_bmp,
local_x, /* dst x */
local_y, /* dst y */
0, /* level */
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
error))
{
if (waste_buf)
g_free (waste_buf);
return FALSE;
}
if (!_cogl_texture_2d_sliced_set_waste (tex_2ds,
source_bmp,
slice_tex,
waste_buf,
x_span, y_span,
&x_iter, &y_iter,
src_x, src_y,
dst_x, dst_y,
error))
{
if (waste_buf)
g_free (waste_buf);
return FALSE;
}
}
}
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
if (waste_buf)
g_free (waste_buf);
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
static int
_cogl_rect_slices_for_size (int size_to_fill,
int max_span_size,
int max_waste,
GArray *out_spans)
{
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 n_spans = 0;
CoglSpan span;
/* Init first slice span */
span.start = 0;
span.size = max_span_size;
span.waste = 0;
/* Repeat until whole area covered */
while (size_to_fill >= span.size)
{
/* Add another slice span of same size */
if (out_spans)
g_array_append_val (out_spans, span);
span.start += span.size;
size_to_fill -= span.size;
n_spans++;
}
/* Add one last smaller slice span */
if (size_to_fill > 0)
{
span.size = size_to_fill;
if (out_spans)
g_array_append_val (out_spans, span);
n_spans++;
}
return n_spans;
}
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
static int
_cogl_pot_slices_for_size (int size_to_fill,
int max_span_size,
int max_waste,
GArray *out_spans)
{
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 n_spans = 0;
CoglSpan span;
/* Init first slice span */
span.start = 0;
span.size = max_span_size;
span.waste = 0;
/* Fix invalid max_waste */
if (max_waste < 0)
max_waste = 0;
while (TRUE)
{
/* Is the whole area covered? */
if (size_to_fill > span.size)
{
/* Not yet - add a span of this size */
if (out_spans)
g_array_append_val (out_spans, span);
span.start += span.size;
size_to_fill -= span.size;
n_spans++;
}
else if (span.size - size_to_fill <= max_waste)
{
/* Yes and waste is small enough */
/* Pick the next power of two up from size_to_fill. This can
sometimes be less than the span.size that would be chosen
otherwise */
span.size = _cogl_util_next_p2 (size_to_fill);
span.waste = span.size - size_to_fill;
if (out_spans)
g_array_append_val (out_spans, span);
return ++n_spans;
}
else
{
/* Yes but waste is too large */
while (span.size - size_to_fill > max_waste)
{
span.size /= 2;
g_assert (span.size > 0);
}
}
}
/* Can't get here */
return 0;
}
static void
_cogl_texture_2d_sliced_gl_flush_legacy_texobj_wrap_modes (CoglTexture *tex,
GLenum wrap_mode_s,
GLenum wrap_mode_t,
GLenum wrap_mode_p)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
int i;
/* Pass the set wrap mode on to all of the child textures */
for (i = 0; i < tex_2ds->slice_textures->len; i++)
{
CoglTexture2D *slice_tex = g_array_index (tex_2ds->slice_textures,
CoglTexture2D *,
i);
_cogl_texture_gl_flush_legacy_texobj_wrap_modes (COGL_TEXTURE (slice_tex),
wrap_mode_s,
wrap_mode_t,
wrap_mode_p);
}
}
static CoglBool
_cogl_texture_2d_sliced_setup_spans (CoglContext *ctx,
CoglTexture2DSliced *tex_2ds,
int width,
int height,
int max_waste,
CoglPixelFormat internal_format,
CoglError **error)
{
int max_width;
int max_height;
int n_x_slices;
int n_y_slices;
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 (*slices_for_size) (int, int, int, GArray*);
/* Initialize size of largest slice according to supported features */
if (cogl_has_feature (ctx, COGL_FEATURE_ID_TEXTURE_NPOT))
{
max_width = width;
max_height = height;
slices_for_size = _cogl_rect_slices_for_size;
}
else
{
max_width = _cogl_util_next_p2 (width);
max_height = _cogl_util_next_p2 (height);
slices_for_size = _cogl_pot_slices_for_size;
}
/* Negative number means no slicing forced by the user */
if (max_waste <= -1)
{
CoglSpan span;
/* Check if size supported else bail out */
if (!ctx->driver_vtable->texture_2d_can_create (ctx,
max_width,
max_height,
internal_format))
{
_cogl_set_error (error,
COGL_TEXTURE_ERROR,
COGL_TEXTURE_ERROR_SIZE,
"Sliced texture size of %d x %d not possible "
"with max waste set to -1",
width,
height);
return FALSE;
}
n_x_slices = 1;
n_y_slices = 1;
/* Init span arrays */
tex_2ds->slice_x_spans = g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan),
1);
tex_2ds->slice_y_spans = g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan),
1);
/* Add a single span for width and height */
span.start = 0;
span.size = max_width;
span.waste = max_width - width;
g_array_append_val (tex_2ds->slice_x_spans, span);
span.size = max_height;
span.waste = max_height - height;
g_array_append_val (tex_2ds->slice_y_spans, span);
}
else
{
/* Decrease the size of largest slice until supported by GL */
while (!ctx->driver_vtable->texture_2d_can_create (ctx,
max_width,
max_height,
internal_format))
{
/* Alternate between width and height */
if (max_width > max_height)
max_width /= 2;
else
max_height /= 2;
if (max_width == 0 || max_height == 0)
{
/* Maybe it would be ok to just g_warn_if_reached() for this
* codepath */
_cogl_set_error (error,
COGL_TEXTURE_ERROR,
COGL_TEXTURE_ERROR_SIZE,
"No suitable slice geometry found");
return FALSE;
}
}
/* Determine the slices required to cover the bitmap area */
n_x_slices = slices_for_size (width,
max_width, max_waste,
NULL);
n_y_slices = slices_for_size (height,
max_height, max_waste,
NULL);
/* Init span arrays with reserved size */
tex_2ds->slice_x_spans = g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan),
n_x_slices);
tex_2ds->slice_y_spans = g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan),
n_y_slices);
/* Fill span arrays with info */
slices_for_size (width,
max_width, max_waste,
tex_2ds->slice_x_spans);
slices_for_size (height,
max_height, max_waste,
tex_2ds->slice_y_spans);
}
return TRUE;
}
static void
_cogl_texture_2d_sliced_slices_free (CoglTexture2DSliced *tex_2ds)
{
if (tex_2ds->slice_textures != NULL)
{
int i;
for (i = 0; i < tex_2ds->slice_textures->len; i++)
{
CoglTexture2D *slice_tex =
g_array_index (tex_2ds->slice_textures, CoglTexture2D *, i);
cogl_object_unref (slice_tex);
}
g_array_free (tex_2ds->slice_textures, TRUE);
}
}
static void
_cogl_texture_2d_sliced_free (CoglTexture2DSliced *tex_2ds)
{
_cogl_texture_2d_sliced_slices_free (tex_2ds);
if (tex_2ds->slice_x_spans != NULL)
g_array_free (tex_2ds->slice_x_spans, TRUE);
if (tex_2ds->slice_y_spans != NULL)
g_array_free (tex_2ds->slice_y_spans, TRUE);
/* Chain up */
_cogl_texture_free (COGL_TEXTURE (tex_2ds));
}
static CoglBool
_cogl_texture_2d_sliced_init_base (CoglContext *ctx,
CoglTexture2DSliced *tex_2ds,
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
int width,
int height,
int max_waste,
CoglPixelFormat internal_format,
CoglError **error)
{
CoglTexture *tex = COGL_TEXTURE (tex_2ds);
_cogl_texture_init (tex, ctx, width, height, &cogl_texture_2d_sliced_vtable);
tex_2ds->max_waste = max_waste;
tex_2ds->internal_format = internal_format;
return _cogl_texture_2d_sliced_setup_spans (ctx, tex_2ds,
width, height,
max_waste,
internal_format,
error);
}
CoglTexture2DSliced *
cogl_texture_2d_sliced_new_with_size (CoglContext *ctx,
int width,
int height,
int max_waste,
CoglPixelFormat internal_format)
{
CoglTexture2DSliced *tex_2ds;
CoglError *ignore_error = NULL;
/* Since no data, we need some internal format */
if (internal_format == COGL_PIXEL_FORMAT_ANY)
internal_format = COGL_PIXEL_FORMAT_RGBA_8888_PRE;
/* Init texture with empty bitmap */
tex_2ds = g_new0 (CoglTexture2DSliced, 1);
if (!_cogl_texture_2d_sliced_init_base (ctx,
tex_2ds,
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
width, height,
max_waste,
internal_format,
&ignore_error))
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
{
/* In this case we failed to find any suitable slicing geometry
* for the given texture size.
*
* We don't need to do anything with the error here since it
* will be picked up on later when trying to allocate the
* texture.
*/
cogl_error_free (ignore_error);
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
}
/* NB: We need to be sure that cogl_texture_is_sliced() will work
* correctly before returning since
* cogl_framebuffer_allocate() uses this api to determine
* if a texture can be rendered to which may be before the
* slices have been allocated.
*/
return _cogl_texture_2d_sliced_object_new (tex_2ds);
}
static CoglBool
_cogl_texture_2d_sliced_allocate (CoglTexture *tex,
CoglError **error)
{
CoglContext *ctx = tex->context;
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
int n_x_slices;
int n_y_slices;
int n_slices;
int x, y;
CoglPixelFormat format = tex_2ds->internal_format;
CoglSpan *x_span;
CoglSpan *y_span;
if (!tex_2ds->slice_x_spans || !tex_2ds->slice_y_spans)
{
_cogl_set_error (error,
COGL_TEXTURE_ERROR,
COGL_TEXTURE_ERROR_SIZE,
"Couldn't find suitable slicing geometry "
"for given size");
return FALSE;
}
n_x_slices = tex_2ds->slice_x_spans->len;
n_y_slices = tex_2ds->slice_y_spans->len;
n_slices = n_x_slices * n_y_slices;
tex_2ds->slice_textures = g_array_sized_new (FALSE, FALSE,
sizeof (CoglTexture2D *),
n_slices);
/* Allocate each slice */
for (y = 0; y < n_y_slices; ++y)
{
y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan, y);
for (x = 0; x < n_x_slices; ++x)
{
CoglTexture *slice;
x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan, x);
COGL_NOTE (SLICING, "CREATE SLICE (%d,%d)\tsize (%d,%d)",
x, y,
(int)(x_span->size - x_span->waste),
(int)(y_span->size - y_span->waste));
slice = COGL_TEXTURE (
cogl_texture_2d_new_with_size (ctx,
x_span->size, y_span->size,
format));
g_array_append_val (tex_2ds->slice_textures, slice);
if (!cogl_texture_allocate (slice, error))
{
_cogl_texture_2d_sliced_slices_free (tex_2ds);
return FALSE;
}
}
}
return TRUE;
}
CoglTexture2DSliced *
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
_cogl_texture_2d_sliced_new_from_bitmap (CoglBitmap *bmp,
int max_waste,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
CoglPixelFormat internal_format,
CoglBool can_convert_in_place,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
CoglError **error)
{
CoglContext *ctx;
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
CoglTexture2DSliced *tex_2ds;
CoglBitmap *upload_bmp;
int width, height;
_COGL_RETURN_VAL_IF_FAIL (cogl_is_bitmap (bmp), NULL);
ctx = _cogl_bitmap_get_context (bmp);
width = cogl_bitmap_get_width (bmp);
height = cogl_bitmap_get_height (bmp);
/* Create new texture and fill with loaded data */
tex_2ds = g_new0 (CoglTexture2DSliced, 1);
internal_format =
_cogl_texture_determine_internal_format (cogl_bitmap_get_format (bmp),
internal_format);
upload_bmp = _cogl_bitmap_convert_for_upload (bmp,
internal_format,
can_convert_in_place,
error);
if (upload_bmp == NULL)
{
_cogl_texture_2d_sliced_free (tex_2ds);
return NULL;
}
/* NB: we may fail to find any suitable slicing geometry for the
* given texture size. */
if (!_cogl_texture_2d_sliced_init_base (ctx,
tex_2ds,
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
width, height,
max_waste,
internal_format,
error))
goto error;
if (!cogl_texture_allocate (COGL_TEXTURE (tex_2ds), error))
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
goto error;
if (!_cogl_texture_2d_sliced_upload_bitmap (tex_2ds,
upload_bmp,
error))
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
goto error;
cogl_object_unref (upload_bmp);
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
return _cogl_texture_2d_sliced_object_new (tex_2ds);
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
error:
cogl_object_unref (upload_bmp);
cogl-bitmap: Encapsulate the CoglBitmap even internally The CoglBitmap struct is now only defined within cogl-bitmap.c so that all of its members can now only be accessed with accessor functions. To get to the data pointer for the bitmap image you must first call _cogl_bitmap_map and later call _cogl_bitmap_unmap. The map function takes the same arguments as cogl_pixel_array_map so that eventually we can make a bitmap optionally internally divert to a pixel array. There is a _cogl_bitmap_new_from_data function which constructs a new bitmap object and takes ownership of the data pointer. The function gets passed a destroy callback which gets called when the bitmap is freed. This is similar to how gdk_pixbuf_new_from_data works. Alternatively NULL can be passed for the destroy function which means that the caller will manage the life of the pointer (but must guarantee that it stays alive at least until the bitmap is freed). This mechanism is used instead of the old approach of creating a CoglBitmap struct on the stack and manually filling in the members. It could also later be used to create a CoglBitmap that owns a GdkPixbuf ref so that we don't necessarily have to copy the GdkPixbuf data when converting to a bitmap. There is also _cogl_bitmap_new_shared. This creates a bitmap using a reference to another CoglBitmap for the data. This is a bit of a hack but it is needed by the atlas texture backend which wants to divert the set_region virtual to another texture but it needs to override the format of the bitmap to ignore the premult flag.
2010-07-07 13:44:16 -04:00
_cogl_texture_2d_sliced_free (tex_2ds);
return NULL;
}
CoglTexture2DSliced *
cogl_texture_2d_sliced_new_from_bitmap (CoglBitmap *bmp,
int max_waste,
CoglPixelFormat internal_format,
CoglError **error)
{
return _cogl_texture_2d_sliced_new_from_bitmap (bmp,
max_waste,
internal_format,
FALSE,
error);
}
CoglTexture2DSliced *
_cogl_texture_2d_sliced_new_from_foreign (CoglContext *ctx,
unsigned int gl_handle,
unsigned int gl_target,
int width,
int height,
int x_pot_waste,
int y_pot_waste,
CoglPixelFormat format,
CoglError **error)
{
/* NOTE: width, height and internal format are not queriable
* in GLES, hence such a function prototype.
*/
int gl_width = 0;
int gl_height = 0;
CoglTexture2DSliced *tex_2ds;
CoglTexture *tex;
CoglSpan x_span;
CoglSpan y_span;
CoglTexture2D *tex_2d;
/* This should only be called when the texture target is 2D. If a
rectangle texture is used then _cogl_texture_new_from_foreign
will create a cogl_texture_rectangle instead */
_COGL_RETURN_VAL_IF_FAIL (gl_target == GL_TEXTURE_2D, NULL);
gl_width = width + x_pot_waste;
gl_height = height + y_pot_waste;
/* Validate pot waste */
if (x_pot_waste < 0 || x_pot_waste >= width ||
y_pot_waste < 0 || y_pot_waste >= height)
return NULL;
tex_2d = cogl_texture_2d_new_from_foreign (ctx,
gl_target,
gl_width,
gl_height,
format,
error);
if (!tex_2d)
return NULL;
/* The texture 2d backend may use a different pixel format if it
queries the actual texture so we'll refetch the format it
actually used */
format = cogl_texture_get_format (COGL_TEXTURE (tex_2d));
/* Create new texture */
tex_2ds = g_new0 (CoglTexture2DSliced, 1);
tex = COGL_TEXTURE (tex_2ds);
_cogl_texture_init (tex, ctx, gl_width, gl_height,
&cogl_texture_2d_sliced_vtable);
tex_2ds->max_waste = 0;
tex_2ds->internal_format = format;
/* Create slice arrays */
tex_2ds->slice_x_spans =
g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan), 1);
tex_2ds->slice_y_spans =
g_array_sized_new (FALSE, FALSE,
sizeof (CoglSpan), 1);
tex_2ds->slice_textures =
g_array_sized_new (FALSE, FALSE,
sizeof (CoglTexture2D *), 1);
/* Store info for a single slice */
x_span.start = 0;
x_span.size = gl_width;
x_span.waste = x_pot_waste;
g_array_append_val (tex_2ds->slice_x_spans, x_span);
y_span.start = 0;
y_span.size = gl_height;
y_span.waste = y_pot_waste;
g_array_append_val (tex_2ds->slice_y_spans, y_span);
g_array_append_val (tex_2ds->slice_textures, tex_2d);
_cogl_texture_set_allocated (COGL_TEXTURE (tex_2ds), TRUE);
return _cogl_texture_2d_sliced_object_new (tex_2ds);
}
CoglTexture2DSliced *
cogl_texture_2d_sliced_new_from_data (CoglContext *ctx,
int width,
int height,
int max_waste,
CoglPixelFormat format,
CoglPixelFormat internal_format,
int rowstride,
const uint8_t *data,
CoglError **error)
{
CoglBitmap *bmp;
CoglTexture2DSliced *tex_2ds;
_COGL_RETURN_VAL_IF_FAIL (format != COGL_PIXEL_FORMAT_ANY, NULL);
_COGL_RETURN_VAL_IF_FAIL (data != NULL, NULL);
/* Rowstride from width if not given */
if (rowstride == 0)
rowstride = width * _cogl_pixel_format_get_bytes_per_pixel (format);
/* Wrap the data into a bitmap */
bmp = cogl_bitmap_new_for_data (ctx,
width, height,
format,
rowstride,
(uint8_t *) data);
tex_2ds = cogl_texture_2d_sliced_new_from_bitmap (bmp, max_waste,
internal_format,
error);
cogl_object_unref (bmp);
return tex_2ds;
}
CoglTexture2DSliced *
cogl_texture_2d_sliced_new_from_file (CoglContext *ctx,
const char *filename,
int max_waste,
CoglPixelFormat internal_format,
CoglError **error)
{
CoglBitmap *bmp;
CoglTexture2DSliced *tex_2ds = NULL;
_COGL_RETURN_VAL_IF_FAIL (error == NULL || *error == NULL, NULL);
bmp = _cogl_bitmap_from_file (ctx, filename, error);
if (bmp == NULL)
return NULL;
tex_2ds = _cogl_texture_2d_sliced_new_from_bitmap (bmp,
max_waste,
internal_format,
TRUE, /* can convert in-place */
error);
cogl_object_unref (bmp);
return tex_2ds;
}
static CoglBool
_cogl_texture_2d_sliced_is_foreign (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglTexture2D *slice_tex;
/* Make sure slices were created */
if (tex_2ds->slice_textures == NULL)
return FALSE;
/* Pass the call on to the first slice */
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, 0);
return _cogl_texture_is_foreign (COGL_TEXTURE (slice_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
static int
_cogl_texture_2d_sliced_get_max_waste (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
return tex_2ds->max_waste;
}
static CoglBool
_cogl_texture_2d_sliced_is_sliced (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
/* It's possible that we failed to calculate valid slicing geometry
* when initializing the texture due to the max_waste size and in
* this case we report that the texture is not sliced.
*
* In this case though we know that we will be throwing an error
* when this texture is later allocated so it shouldn't really
* matter what we report here since the texture won't be used in the
* end.
*/
if (!tex_2ds->slice_x_spans || !tex_2ds->slice_y_spans)
return FALSE;
if (tex_2ds->slice_x_spans->len != 1 ||
tex_2ds->slice_y_spans->len != 1)
return TRUE;
else
return FALSE;
}
static CoglBool
_cogl_texture_2d_sliced_can_hardware_repeat (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglTexture2D *slice_tex;
CoglSpan *x_span;
CoglSpan *y_span;
/* If there's more than one texture then we can't hardware repeat */
if (tex_2ds->slice_textures->len != 1)
return FALSE;
/* If there's any waste then we can't hardware repeat */
x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan, 0);
y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan, 0);
if (x_span->waste > 0 || y_span->waste > 0)
return FALSE;
/* Otherwise pass the query on to the single slice texture */
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, 0);
return _cogl_texture_can_hardware_repeat (COGL_TEXTURE (slice_tex));
}
static void
_cogl_texture_2d_sliced_transform_coords_to_gl (CoglTexture *tex,
float *s,
float *t)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglSpan *x_span;
CoglSpan *y_span;
CoglTexture2D *slice_tex;
g_assert (!_cogl_texture_2d_sliced_is_sliced (tex));
/* Don't include the waste in the texture coordinates */
x_span = &g_array_index (tex_2ds->slice_x_spans, CoglSpan, 0);
y_span = &g_array_index (tex_2ds->slice_y_spans, CoglSpan, 0);
*s *= tex->width / (float)x_span->size;
*t *= tex->height / (float)y_span->size;
/* Let the child texture further transform the coords */
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, 0);
_cogl_texture_transform_coords_to_gl (COGL_TEXTURE (slice_tex), s, t);
}
static CoglTransformResult
2010-01-18 04:22:04 -05:00
_cogl_texture_2d_sliced_transform_quad_coords_to_gl (CoglTexture *tex,
float *coords)
{
CoglBool need_repeat = FALSE;
int i;
/* This is a bit lazy - in the case where the quad lies entirely
* within a single slice we could avoid the fallback. But that
* could likely lead to visual inconsistency if the fallback involves
* dropping layers, so this might be the right thing to do anyways.
*/
2010-01-18 04:22:04 -05:00
if (_cogl_texture_2d_sliced_is_sliced (tex))
return COGL_TRANSFORM_SOFTWARE_REPEAT;
for (i = 0; i < 4; i++)
if (coords[i] < 0.0f || coords[i] > 1.0f)
need_repeat = TRUE;
if (need_repeat && !_cogl_texture_2d_sliced_can_hardware_repeat (tex))
return COGL_TRANSFORM_SOFTWARE_REPEAT;
2010-01-18 04:22:04 -05:00
_cogl_texture_2d_sliced_transform_coords_to_gl (tex, coords + 0, coords + 1);
_cogl_texture_2d_sliced_transform_coords_to_gl (tex, coords + 2, coords + 3);
return (need_repeat
? COGL_TRANSFORM_HARDWARE_REPEAT : COGL_TRANSFORM_NO_REPEAT);
2010-01-18 04:22:04 -05:00
}
static CoglBool
_cogl_texture_2d_sliced_get_gl_texture (CoglTexture *tex,
GLuint *out_gl_handle,
GLenum *out_gl_target)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglTexture2D *slice_tex;
if (tex_2ds->slice_textures == NULL)
return FALSE;
if (tex_2ds->slice_textures->len < 1)
return FALSE;
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, 0);
return cogl_texture_get_gl_texture (COGL_TEXTURE (slice_tex),
out_gl_handle, out_gl_target);
}
static void
_cogl_texture_2d_sliced_gl_flush_legacy_texobj_filters (CoglTexture *tex,
GLenum min_filter,
GLenum mag_filter)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglTexture2D *slice_tex;
int i;
_COGL_RETURN_IF_FAIL (tex_2ds->slice_textures != NULL);
/* Apply new filters to every slice. The slice texture itself should
cache the value and avoid resubmitting the same filter value to
GL */
for (i = 0; i < tex_2ds->slice_textures->len; i++)
{
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, i);
_cogl_texture_gl_flush_legacy_texobj_filters (COGL_TEXTURE (slice_tex),
min_filter, mag_filter);
}
}
static void
_cogl_texture_2d_sliced_pre_paint (CoglTexture *tex,
CoglTexturePrePaintFlags flags)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
int i;
_COGL_RETURN_IF_FAIL (tex_2ds->slice_textures != NULL);
/* Pass the pre-paint on to every slice */
for (i = 0; i < tex_2ds->slice_textures->len; i++)
{
CoglTexture2D *slice_tex = g_array_index (tex_2ds->slice_textures,
CoglTexture2D *, i);
_cogl_texture_pre_paint (COGL_TEXTURE (slice_tex), flags);
}
}
2010-01-18 04:22:04 -05:00
static void
_cogl_texture_2d_sliced_ensure_non_quad_rendering (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
int i;
_COGL_RETURN_IF_FAIL (tex_2ds->slice_textures != NULL);
/* Pass the call on to every slice */
for (i = 0; i < tex_2ds->slice_textures->len; i++)
{
CoglTexture2D *slice_tex = g_array_index (tex_2ds->slice_textures,
CoglTexture2D *, i);
_cogl_texture_ensure_non_quad_rendering (COGL_TEXTURE (slice_tex));
}
2010-01-18 04:22:04 -05:00
}
static CoglBool
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
_cogl_texture_2d_sliced_set_region (CoglTexture *tex,
int src_x,
int src_y,
int dst_x,
int dst_y,
int dst_width,
int dst_height,
int level,
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
CoglBitmap *bmp,
CoglError **error)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglBitmap *upload_bmp;
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
CoglBool status;
upload_bmp = _cogl_bitmap_convert_for_upload (bmp,
cogl_texture_get_format (tex),
FALSE, /* can't convert in
place */
error);
if (!upload_bmp)
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
return FALSE;
status = _cogl_texture_2d_sliced_upload_subregion (tex_2ds,
src_x, src_y,
dst_x, dst_y,
dst_width, dst_height,
upload_bmp,
error);
cogl_object_unref (upload_bmp);
Allow propogation of OOM errors to apps This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors.
2012-11-08 12:54:10 -05:00
return status;
}
static CoglPixelFormat
_cogl_texture_2d_sliced_get_format (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
return tex_2ds->internal_format;
}
static GLenum
_cogl_texture_2d_sliced_get_gl_format (CoglTexture *tex)
{
CoglTexture2DSliced *tex_2ds = COGL_TEXTURE_2D_SLICED (tex);
CoglTexture2D *slice_tex;
/* Assert that we've allocated our slices at this point */
cogl_texture_allocate (tex, NULL); /* (abort on error) */
/* Pass the call on to the first slice */
slice_tex = g_array_index (tex_2ds->slice_textures, CoglTexture2D *, 0);
return _cogl_texture_gl_get_format (COGL_TEXTURE (slice_tex));
}
static CoglTextureType
_cogl_texture_2d_sliced_get_type (CoglTexture *tex)
{
return COGL_TEXTURE_TYPE_2D;
}
static const CoglTextureVtable
cogl_texture_2d_sliced_vtable =
{
FALSE, /* not primitive */
_cogl_texture_2d_sliced_allocate,
_cogl_texture_2d_sliced_set_region,
NULL, /* get_data */
_cogl_texture_2d_sliced_foreach_sub_texture_in_region,
_cogl_texture_2d_sliced_get_max_waste,
_cogl_texture_2d_sliced_is_sliced,
_cogl_texture_2d_sliced_can_hardware_repeat,
_cogl_texture_2d_sliced_transform_coords_to_gl,
2010-01-18 04:22:04 -05:00
_cogl_texture_2d_sliced_transform_quad_coords_to_gl,
_cogl_texture_2d_sliced_get_gl_texture,
_cogl_texture_2d_sliced_gl_flush_legacy_texobj_filters,
_cogl_texture_2d_sliced_pre_paint,
2010-01-18 04:22:04 -05:00
_cogl_texture_2d_sliced_ensure_non_quad_rendering,
_cogl_texture_2d_sliced_gl_flush_legacy_texobj_wrap_modes,
_cogl_texture_2d_sliced_get_format,
_cogl_texture_2d_sliced_get_gl_format,
_cogl_texture_2d_sliced_get_type,
_cogl_texture_2d_sliced_is_foreign,
NULL /* set_auto_mipmap */
};