ic
/
magnum
mirror of https://github.com/mosra/magnum.git
6
0
Fork 0
Code Issues Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
10401 Commits
21 Branches
19 Tags
70 MiB
Tree: 332bcb62fa
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '332bcb62fa'
${ noResults }
magnum/src/Magnum/Test/PixelFormatTest.cpp

592 lines
24 KiB
Raw Normal View History Unescape

Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
/*
This file is part of Magnum.
Updated copyright year.
6 years ago
Copyright © 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019,
Updated copyright year.
4 years ago
2020, 2021, 2022 Vladimír Vondruš <mosra@centrum.cz>
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and associated documentation files (the "Software"),
to deal in the Software without restriction, including without limitation
the rights to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.
*/
#include <sstream>
#include <Corrade/TestSuite/Tester.h>
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#include <Corrade/TestSuite/Compare/Numeric.h>
Adapted to Corrade changes. Mostly missing includes now.
7 years ago
#include <Corrade/Utility/DebugStl.h>
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
#include <Corrade/Utility/Configuration.h>
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
#include "Magnum/PixelFormat.h"
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#include "Magnum/Math/Vector3.h"
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Wrap all tests in unnamed namespaces. This makes the compiler warn about test cases that aren't ever used, which is a good thing.
7 years ago
namespace Magnum { namespace Test { namespace {
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
struct PixelFormatTest: TestSuite::Tester {
explicit PixelFormatTest();
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
void mapping();
void compressedMapping();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void size();
Reserve zero [Compressed]PixelFormat for an invalid value. Catches accidents better -- just look at the diff. Might catch accidents in your code now as well, so beware!
7 years ago
void sizeInvalid();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void sizeImplementationSpecific();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
void channelFormatCount();
void channelFormatCountInvalid();
void channelFormatCountDepthStencilImplementationSpecific();
void isSrgb();
void isSrgbInvalid();
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
void isSrgbDepthStencilImplementationSpecific();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
void isDepthOrStencil();
void isDepthOrStencilInvalid();
void isDepthOrStencilImplementationSpecific();
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
void compressedBlockSize();
void compressedBlockSizeInvalid();
void compressedBlockSizeImplementationSpecific();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void isImplementationSpecific();
void wrap();
void wrapInvalid();
void unwrap();
void unwrapInvalid();
void compressedIsImplementationSpecific();
void compressedWrap();
void compressedWrapInvalid();
void compressedUnwrap();
void compressedUnwrapInvalid();
void debug();
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void debugPacked();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void debugImplementationSpecific();
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void debugImplementationSpecificPacked();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void compressedDebug();
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void compressedDebugPacked();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void compressedDebugImplementationSpecific();
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void compressedDebugImplementationSpecificPacked();
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
void configuration();
void compresedConfiguration();
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
};
PixelFormatTest::PixelFormatTest() {
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
addTests({&PixelFormatTest::mapping,
&PixelFormatTest::compressedMapping,
&PixelFormatTest::size,
Reserve zero [Compressed]PixelFormat for an invalid value. Catches accidents better -- just look at the diff. Might catch accidents in your code now as well, so beware!
7 years ago
&PixelFormatTest::sizeInvalid,
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
&PixelFormatTest::sizeImplementationSpecific,
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
&PixelFormatTest::channelFormatCount,
&PixelFormatTest::channelFormatCountInvalid,
&PixelFormatTest::channelFormatCountDepthStencilImplementationSpecific,
&PixelFormatTest::isSrgb,
&PixelFormatTest::isSrgbInvalid,
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
&PixelFormatTest::isSrgbDepthStencilImplementationSpecific,
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
&PixelFormatTest::isDepthOrStencil,
&PixelFormatTest::isDepthOrStencilInvalid,
&PixelFormatTest::isDepthOrStencilImplementationSpecific,
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
&PixelFormatTest::compressedBlockSize,
&PixelFormatTest::compressedBlockSizeInvalid,
&PixelFormatTest::compressedBlockSizeImplementationSpecific,
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
&PixelFormatTest::isImplementationSpecific,
&PixelFormatTest::wrap,
&PixelFormatTest::wrapInvalid,
&PixelFormatTest::unwrap,
&PixelFormatTest::unwrapInvalid,
&PixelFormatTest::compressedIsImplementationSpecific,
&PixelFormatTest::compressedWrap,
&PixelFormatTest::compressedWrapInvalid,
&PixelFormatTest::compressedUnwrap,
&PixelFormatTest::compressedUnwrapInvalid,
&PixelFormatTest::debug,
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
&PixelFormatTest::debugPacked,
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
&PixelFormatTest::debugImplementationSpecific,
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
&PixelFormatTest::debugImplementationSpecificPacked,
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
&PixelFormatTest::compressedDebug,
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
&PixelFormatTest::compressedDebugPacked,
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
&PixelFormatTest::compressedDebugImplementationSpecific,
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
&PixelFormatTest::compressedDebugImplementationSpecificPacked,
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
&PixelFormatTest::configuration,
&PixelFormatTest::compresedConfiguration});
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
void PixelFormatTest::mapping() {
/* This goes through the first 16 bits, which should be enough. Going
through 32 bits takes 8 seconds, too much. */
UnsignedInt firstUnhandled = 0xffff;
UnsignedInt nextHandled = 1; /* 0 is an invalid format */
for(UnsignedInt i = 1; i <= 0xffff; ++i) {
const auto format = PixelFormat(i);
/* Each case verifies:
Test: fix false sense of security in pixel/mesh enum mapping tests. The stripped-down versions didn't actually test for correct order because there was no external API result to check against expected output. That's fixed now, plus clarified some comments and removed copypasted nonsense.
6 years ago
- that the entries are ordered by number by comparing a function to
expected result (so insertion here is done in proper place)
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
- that there was no gap (unhandled value inside the range)
- that channel count times size of a channel equals to size of the
format, unless it's a depth/stencil type */
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
#define _c(format) \
case PixelFormat::format: \
Test: fix false sense of security in pixel/mesh enum mapping tests. The stripped-down versions didn't actually test for correct order because there was no external API result to check against expected output. That's fixed now, plus clarified some comments and removed copypasted nonsense.
6 years ago
CORRADE_COMPARE(Utility::ConfigurationValue<PixelFormat>::toString(PixelFormat::format, {}), #format); \
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
CORRADE_COMPARE(nextHandled, i); \
CORRADE_COMPARE(firstUnhandled, 0xffff); \
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
if(!isPixelFormatDepthOrStencil(PixelFormat::format)) \
CORRADE_COMPARE(pixelFormatChannelCount(PixelFormat::format)*pixelFormatSize(pixelFormatChannelFormat(PixelFormat::format)), pixelFormatSize(PixelFormat::format)); \
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
++nextHandled; \
continue;
#include "Magnum/Implementation/pixelFormatMapping.hpp"
#undef _c
}
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
#pragma GCC diagnostic pop
#endif
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
/* Not handled by any value, remember -- we might either be at the end
of the enum range (which is okay) or some value might be unhandled
here */
firstUnhandled = i;
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
CORRADE_COMPARE(firstUnhandled, 0xffff);
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
void PixelFormatTest::compressedMapping() {
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
/* This goes through the first 16 bits, which should be enough. Going
through 32 bits takes 8 seconds, too much. */
UnsignedInt firstUnhandled = 0xffff;
UnsignedInt nextHandled = 1; /* 0 is an invalid format */
for(UnsignedInt i = 1; i <= 0xffff; ++i) {
const auto format = CompressedPixelFormat(i);
/* Each case verifies:
Test: fix false sense of security in pixel/mesh enum mapping tests. The stripped-down versions didn't actually test for correct order because there was no external API result to check against expected output. That's fixed now, plus clarified some comments and removed copypasted nonsense.
6 years ago
- that the entries are ordered by number by comparing a function to
expected result (so insertion here is done in proper place)
Document what's all tested in this crazy macro.
4 years ago
- that there was no gap (unhandled value inside the range)
- that the block size table entry matches
- that the block data size is whole bytes and at most 16 bytes
- that the block size is at most 16x16x16 */
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
#define _c(format, width, height, depth, size) \
case CompressedPixelFormat::format: \
Test: fix false sense of security in pixel/mesh enum mapping tests. The stripped-down versions didn't actually test for correct order because there was no external API result to check against expected output. That's fixed now, plus clarified some comments and removed copypasted nonsense.
6 years ago
CORRADE_COMPARE(Utility::ConfigurationValue<CompressedPixelFormat>::toString(CompressedPixelFormat::format, {}), #format); \
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_COMPARE(nextHandled, i); \
CORRADE_COMPARE(firstUnhandled, 0xffff); \
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
CORRADE_COMPARE(compressedPixelFormatBlockSize(CompressedPixelFormat::format), (Vector3i{width, height, depth})); \
CORRADE_COMPARE(compressedPixelFormatBlockDataSize(CompressedPixelFormat::format), size/8); \
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_COMPARE(size % 8, 0); \
CORRADE_COMPARE_AS(width, 16, TestSuite::Compare::LessOrEqual); \
CORRADE_COMPARE_AS(height, 16, TestSuite::Compare::LessOrEqual); \
CORRADE_COMPARE_AS(depth, 16, TestSuite::Compare::LessOrEqual); \
CORRADE_COMPARE_AS(size/8, 16, TestSuite::Compare::LessOrEqual); \
++nextHandled; \
continue;
#include "Magnum/Implementation/compressedPixelFormatMapping.hpp"
#undef _c
}
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#pragma GCC diagnostic pop
#endif
/* Not handled by any value, remember -- we might either be at the end
of the enum range (which is okay) or some value might be unhandled
here */
firstUnhandled = i;
}
CORRADE_COMPARE(firstUnhandled, 0xffff);
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
void PixelFormatTest::size() {
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
CORRADE_COMPARE(pixelFormatSize(PixelFormat::R8I), 1);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::R16UI), 2);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGB8Unorm), 3);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGBA8Snorm), 4);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGB16I), 6);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGBA16F), 8);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGB32UI), 12);
CORRADE_COMPARE(pixelFormatSize(PixelFormat::RGBA32F), 16);
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
}
void PixelFormatTest::sizeInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
std::ostringstream out;
Error redirectError{&out};
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
pixelFormatSize(PixelFormat{});
pixelFormatSize(PixelFormat(0xdead));
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
CORRADE_COMPARE(out.str(),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"pixelFormatSize(): invalid format PixelFormat(0x0)\n"
"pixelFormatSize(): invalid format PixelFormat(0xdead)\n");
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
}
void PixelFormatTest::sizeImplementationSpecific() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
std::ostringstream out;
Error redirectError{&out};
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
pixelFormatSize(pixelFormatWrap(0xdead));
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
CORRADE_COMPARE(out.str(), "pixelFormatSize(): can't determine size of an implementation-specific format 0xdead\n");
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
}
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
void PixelFormatTest::channelFormatCount() {
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::R8Unorm), PixelFormat::R8Unorm);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RG8Snorm), PixelFormat::R8Snorm);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGB8Srgb), PixelFormat::R8Srgb);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGBA8UI), PixelFormat::R8UI);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RG8I), PixelFormat::R8I);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RG16Unorm), PixelFormat::R16Unorm);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGBA16Snorm), PixelFormat::R16Snorm);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RG16UI), PixelFormat::R16UI);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGBA16I), PixelFormat::R16I);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGB32UI), PixelFormat::R32UI);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RG32I), PixelFormat::R32I);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGB16F), PixelFormat::R16F);
CORRADE_COMPARE(pixelFormatChannelFormat(PixelFormat::RGB32F), PixelFormat::R32F);
CORRADE_COMPARE(pixelFormatChannelCount(PixelFormat::R16UI), 1);
CORRADE_COMPARE(pixelFormatChannelCount(PixelFormat::RG8Unorm), 2);
CORRADE_COMPARE(pixelFormatChannelCount(PixelFormat::RGB16I), 3);
CORRADE_COMPARE(pixelFormatChannelCount(PixelFormat::RGBA16F), 4);
}
void PixelFormatTest::channelFormatCountInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
pixelFormatChannelFormat(PixelFormat{});
pixelFormatChannelFormat(PixelFormat(0xdead));
pixelFormatChannelCount(PixelFormat{});
pixelFormatChannelCount(PixelFormat(0xdead));
CORRADE_COMPARE(out.str(),
"pixelFormatChannelFormat(): invalid format PixelFormat(0x0)\n"
"pixelFormatChannelFormat(): invalid format PixelFormat(0xdead)\n"
"pixelFormatChannelCount(): invalid format PixelFormat(0x0)\n"
"pixelFormatChannelCount(): invalid format PixelFormat(0xdead)\n");
}
void PixelFormatTest::channelFormatCountDepthStencilImplementationSpecific() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
pixelFormatChannelFormat(pixelFormatWrap(0xdead));
pixelFormatChannelFormat(PixelFormat::Depth16Unorm);
pixelFormatChannelCount(pixelFormatWrap(0xdead));
pixelFormatChannelCount(PixelFormat::Depth16Unorm);
CORRADE_COMPARE(out.str(),
"pixelFormatChannelFormat(): can't determine channel format of an implementation-specific format 0xdead\n"
"pixelFormatChannelFormat(): can't determine channel format of PixelFormat::Depth16Unorm\n"
"pixelFormatChannelCount(): can't determine channel count of an implementation-specific format 0xdead\n"
"pixelFormatChannelCount(): can't determine channel count of PixelFormat::Depth16Unorm\n");
}
void PixelFormatTest::isSrgb() {
CORRADE_VERIFY(isPixelFormatSrgb(PixelFormat::RG8Srgb));
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
CORRADE_VERIFY(!isPixelFormatSrgb(PixelFormat::RG8Snorm));
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
CORRADE_VERIFY(!isPixelFormatSrgb(PixelFormat::RGB16F));
}
void PixelFormatTest::isSrgbInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
isPixelFormatSrgb(PixelFormat{});
isPixelFormatSrgb(PixelFormat(0xdead));
CORRADE_COMPARE(out.str(),
"isPixelFormatSrgb(): invalid format PixelFormat(0x0)\n"
"isPixelFormatSrgb(): invalid format PixelFormat(0xdead)\n");
}
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
void PixelFormatTest::isSrgbDepthStencilImplementationSpecific() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
isPixelFormatSrgb(pixelFormatWrap(0xdead));
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
isPixelFormatSrgb(PixelFormat::Depth16Unorm);
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
CORRADE_COMPARE(out.str(),
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
"isPixelFormatSrgb(): can't determine colorspace of an implementation-specific format 0xdead\n"
"isPixelFormatSrgb(): can't determine colorspace of PixelFormat::Depth16Unorm\n");
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
}
void PixelFormatTest::isDepthOrStencil() {
CORRADE_VERIFY(!isPixelFormatDepthOrStencil(PixelFormat::RG8Srgb));
CORRADE_VERIFY(!isPixelFormatDepthOrStencil(PixelFormat::RGB16F));
CORRADE_VERIFY(isPixelFormatDepthOrStencil(PixelFormat::Stencil8UI));
}
void PixelFormatTest::isDepthOrStencilInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
isPixelFormatDepthOrStencil(PixelFormat{});
isPixelFormatDepthOrStencil(PixelFormat(0xdead));
CORRADE_COMPARE(out.str(),
"isPixelFormatDepthOrStencil(): invalid format PixelFormat(0x0)\n"
"isPixelFormatDepthOrStencil(): invalid format PixelFormat(0xdead)\n");
}
void PixelFormatTest::isDepthOrStencilImplementationSpecific() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
std::ostringstream out;
Error redirectError{&out};
isPixelFormatDepthOrStencil(pixelFormatWrap(0xdead));
CORRADE_COMPARE(out.str(),
"isPixelFormatDepthOrStencil(): can't determine type of an implementation-specific format 0xdead\n");
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
void PixelFormatTest::compressedBlockSize() {
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
CORRADE_COMPARE(compressedPixelFormatBlockSize(CompressedPixelFormat::Etc2RGB8A1Srgb), (Vector3i{4, 4, 1}));
CORRADE_COMPARE(compressedPixelFormatBlockDataSize(CompressedPixelFormat::Etc2RGB8A1Srgb), 8);
CORRADE_COMPARE(compressedPixelFormatBlockSize(CompressedPixelFormat::Astc5x4RGBAUnorm), (Vector3i{5, 4, 1}));
CORRADE_COMPARE(compressedPixelFormatBlockDataSize(CompressedPixelFormat::Astc5x4RGBAUnorm), 16);
CORRADE_COMPARE(compressedPixelFormatBlockSize(CompressedPixelFormat::Astc12x10RGBAUnorm), (Vector3i{12, 10, 1}));
CORRADE_COMPARE(compressedPixelFormatBlockDataSize(CompressedPixelFormat::Astc12x10RGBAUnorm), 16);
CORRADE_COMPARE(compressedPixelFormatBlockSize(CompressedPixelFormat::PvrtcRGBA2bppUnorm), (Vector3i{8, 4, 1}));
CORRADE_COMPARE(compressedPixelFormatBlockDataSize(CompressedPixelFormat::PvrtcRGBA2bppUnorm), 8);
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
/* The rest tested in compressedMapping() */
}
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
void PixelFormatTest::compressedBlockSizeInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
std::ostringstream out;
Error redirectError{&out};
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
compressedPixelFormatBlockSize(CompressedPixelFormat{});
compressedPixelFormatBlockSize(CompressedPixelFormat(0xdead));
compressedPixelFormatBlockDataSize(CompressedPixelFormat{});
compressedPixelFormatBlockDataSize(CompressedPixelFormat(0xdead));
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_COMPARE(out.str(),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockSize(): invalid format CompressedPixelFormat(0x0)\n"
"compressedPixelFormatBlockSize(): invalid format CompressedPixelFormat(0xdead)\n"
"compressedPixelFormatBlockDataSize(): invalid format CompressedPixelFormat(0x0)\n"
"compressedPixelFormatBlockDataSize(): invalid format CompressedPixelFormat(0xdead)\n");
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
}
void PixelFormatTest::compressedBlockSizeImplementationSpecific() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
std::ostringstream out;
Error redirectError{&out};
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
compressedPixelFormatBlockSize(compressedPixelFormatWrap(0xdead));
compressedPixelFormatBlockDataSize(compressedPixelFormatWrap(0xdead));
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_COMPARE(out.str(),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockSize(): can't determine size of an implementation-specific format 0xdead\n"
"compressedPixelFormatBlockDataSize(): can't determine size of an implementation-specific format 0xdead\n");
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void PixelFormatTest::isImplementationSpecific() {
constexpr bool a = isPixelFormatImplementationSpecific(PixelFormat::RGBA8Unorm);
constexpr bool b = isPixelFormatImplementationSpecific(PixelFormat(0x8000dead));
CORRADE_VERIFY(!a);
CORRADE_VERIFY(b);
}
void PixelFormatTest::wrap() {
constexpr PixelFormat a = pixelFormatWrap(0xdead);
CORRADE_COMPARE(UnsignedInt(a), 0x8000dead);
}
void PixelFormatTest::wrapInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
std::ostringstream out;
Error redirectError{&out};
pixelFormatWrap(0xdeadbeef);
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
CORRADE_COMPARE(out.str(), "pixelFormatWrap(): implementation-specific value 0xdeadbeef already wrapped or too large\n");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
void PixelFormatTest::unwrap() {
constexpr UnsignedInt a = pixelFormatUnwrap(PixelFormat(0x8000dead));
CORRADE_COMPARE(a, 0xdead);
}
void PixelFormatTest::unwrapInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
std::ostringstream out;
Error redirectError{&out};
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
pixelFormatUnwrap(PixelFormat::R8Snorm);
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
CORRADE_COMPARE(out.str(), "pixelFormatUnwrap(): PixelFormat::R8Snorm isn't a wrapped implementation-specific value\n");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
void PixelFormatTest::compressedIsImplementationSpecific() {
CORRADE_VERIFY(!isPixelFormatImplementationSpecific(PixelFormat::RGBA8Unorm));
CORRADE_VERIFY(isPixelFormatImplementationSpecific(pixelFormatWrap(0xdead)));
}
void PixelFormatTest::compressedWrap() {
CORRADE_COMPARE(UnsignedInt(pixelFormatWrap(0xdead)), 0x8000dead);
}
void PixelFormatTest::compressedWrapInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
std::ostringstream out;
Error redirectError{&out};
compressedPixelFormatWrap(0xdeadbeef);
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
CORRADE_COMPARE(out.str(), "compressedPixelFormatWrap(): implementation-specific value 0xdeadbeef already wrapped or too large\n");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
void PixelFormatTest::compressedUnwrap() {
CORRADE_COMPARE(UnsignedInt(compressedPixelFormatUnwrap(CompressedPixelFormat(0x8000dead))), 0xdead);
}
void PixelFormatTest::compressedUnwrapInvalid() {
Test: make use of CORRADE_SKIP_IF_NO_ASSERT().
4 years ago
CORRADE_SKIP_IF_NO_ASSERT();
Make this compile and test cleanly with CORRADE_NO_ASSERT defined.
6 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
std::ostringstream out;
Error redirectError{&out};
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
compressedPixelFormatUnwrap(CompressedPixelFormat::EacR11Snorm);
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
CORRADE_COMPARE(out.str(), "compressedPixelFormatUnwrap(): CompressedPixelFormat::EacR11Snorm isn't a wrapped implementation-specific value\n");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
void PixelFormatTest::debug() {
std::ostringstream out;
Debug{&out} << PixelFormat::RG16Snorm << PixelFormat(0xdead);
CORRADE_COMPARE(out.str(), "PixelFormat::RG16Snorm PixelFormat(0xdead)\n");
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void PixelFormatTest::debugPacked() {
std::ostringstream out;
/* Last is not packed, ones before should not make any flags persistent */
Debug{&out} << Debug::packed << PixelFormat::RG16Snorm << Debug::packed << PixelFormat(0xdead) << PixelFormat::RGBA8Unorm;
CORRADE_COMPARE(out.str(), "RG16Snorm 0xdead PixelFormat::RGBA8Unorm\n");
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void PixelFormatTest::debugImplementationSpecific() {
std::ostringstream out;
Debug{&out} << pixelFormatWrap(0xdead);
CORRADE_COMPARE(out.str(), "PixelFormat::ImplementationSpecific(0xdead)\n");
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void PixelFormatTest::debugImplementationSpecificPacked() {
std::ostringstream out;
/* Second is not packed, the first should not make any flags persistent */
Debug{&out} << Debug::packed << pixelFormatWrap(0xdead) << PixelFormat::RGBA8Unorm;
CORRADE_COMPARE(out.str(), "ImplementationSpecific(0xdead) PixelFormat::RGBA8Unorm\n");
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void PixelFormatTest::compressedDebug() {
std::ostringstream out;
Debug{&out} << CompressedPixelFormat::Bc3RGBAUnorm << CompressedPixelFormat(0xdead);
CORRADE_COMPARE(out.str(), "CompressedPixelFormat::Bc3RGBAUnorm CompressedPixelFormat(0xdead)\n");
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void PixelFormatTest::compressedDebugPacked() {
std::ostringstream out;
/* Last is not packed, ones before should not make any flags persistent */
Debug{&out} << Debug::packed << CompressedPixelFormat::Bc3RGBAUnorm << Debug::packed << CompressedPixelFormat(0xdead) << CompressedPixelFormat::Astc10x10RGBAF;
CORRADE_COMPARE(out.str(), "Bc3RGBAUnorm 0xdead CompressedPixelFormat::Astc10x10RGBAF\n");
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
void PixelFormatTest::compressedDebugImplementationSpecific() {
std::ostringstream out;
Debug{&out} << compressedPixelFormatWrap(0xdead);
CORRADE_COMPARE(out.str(), "CompressedPixelFormat::ImplementationSpecific(0xdead)\n");
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
void PixelFormatTest::compressedDebugImplementationSpecificPacked() {
std::ostringstream out;
/* Second is not packed, the first should not make any flags persistent */
Debug{&out} << compressedPixelFormatWrap(0xdead) << CompressedPixelFormat::Astc10x10RGBAF;
CORRADE_COMPARE(out.str(), "CompressedPixelFormat::ImplementationSpecific(0xdead) CompressedPixelFormat::Astc10x10RGBAF\n");
}
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
void PixelFormatTest::configuration() {
Utility::Configuration c;
c.setValue("format", PixelFormat::RGB8Srgb);
CORRADE_COMPARE(c.value("format"), "RGB8Srgb");
CORRADE_COMPARE(c.value<PixelFormat>("format"), PixelFormat::RGB8Srgb);
c.setValue("zero", PixelFormat(0));
CORRADE_COMPARE(c.value("zero"), "");
CORRADE_COMPARE(c.value<PixelFormat>("zero"), PixelFormat{});
c.setValue("invalid", PixelFormat(0xdead));
CORRADE_COMPARE(c.value("invalid"), "");
CORRADE_COMPARE(c.value<PixelFormat>("invalid"), PixelFormat{});
}
void PixelFormatTest::compresedConfiguration() {
Utility::Configuration c;
c.setValue("format", CompressedPixelFormat::Astc3x3x3RGBASrgb);
CORRADE_COMPARE(c.value("format"), "Astc3x3x3RGBASrgb");
CORRADE_COMPARE(c.value<CompressedPixelFormat>("format"), CompressedPixelFormat::Astc3x3x3RGBASrgb);
c.setValue("zero", CompressedPixelFormat(0));
CORRADE_COMPARE(c.value("zero"), "");
CORRADE_COMPARE(c.value<CompressedPixelFormat>("zero"), CompressedPixelFormat{});
c.setValue("invalid", CompressedPixelFormat(0xdead));
CORRADE_COMPARE(c.value("invalid"), "");
CORRADE_COMPARE(c.value<CompressedPixelFormat>("invalid"), CompressedPixelFormat{});
}
Wrap all tests in unnamed namespaces. This makes the compiler warn about test cases that aren't ever used, which is a good thing.
7 years ago
}}}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
CORRADE_TEST_MAIN(Magnum::Test::PixelFormatTest)