magnum

Commit Graph

Author	SHA1	Message	Date
Vladimír Vondruš	a97384839b	Use privately Corrade Optional instead of the problematic std::optional. Not in the public APIs yet (Trade etc.), as I need to solve backwards compatibility first.	8 years ago
Vladimír Vondruš	623fa97970	Updated copyright year.	8 years ago
Vladimír Vondruš	32613a592a	Implemented glClearBuffer() APIs from GL 3.0 / GLES 3.0. According to my browser, glClear() can't be used to clear integer framebuffers, so this APIs are needed.	9 years ago
Vladimír Vondruš	1c433486cd	Consistently use macOS instead of OSX.	9 years ago
Vladimír Vondruš	2145ae9b96	Dropped NaCl support.	9 years ago
Vladimír Vondruš	0371ebd9a3	SVGA3D: work around broken glTex[ture][Sub]ImageD() for arrays. SVGA3D has broken handling of glTex[ture][Sub]ImageD() for 1D arrays, 2D arrays, 3D textures and cube map textures where it uploads just the first slice in the last dimension. This is only with copies from host memory, not with buffer images. Seems to be fixed in Mesa 13, but I have no such system to verify that on. Relevant commit in the Mesa sources: `2aa9ff0cda` This is one of the uglier workarounds -- I had to reintroduce multiple code paths for glTexImage() which were removed when implementing ARB_DSA and the workaround consists of basically a bunch of functions that are slicing the image and calling the original implementations with each slice.	9 years ago
Vladimír Vondruš	2148d2c235	Properly include a header to avoid crashes on MSVC GLES build. Did I mention that this is a serious fucking compiler misfeature/bug?	9 years ago
Vladimír Vondruš	e255a01ce3	SVGA3D: disable GL_ARB_get_texture_sub_image altogether. In some cases it causes OOB writes, corrupting application memory. Nasty.	9 years ago
Vladimír Vondruš	1f94067981	Implementation: doc++, minor cleanup.	9 years ago
Vladimír Vondruš	5d3e2e4223	Work around a bug with DSA glNamedBufferData() in SVGA3D driver. For some reason a buffer allocated with this function caused Mesh::draw() to draw nothing. I gave up on investigating the root cause -- using non-DSA glBufferData() "just works", so in case of this driver the "svga3d-broken-dsa-bufferdata" workaround uses the non-DSA code path.	9 years ago
Vladimír Vondruš	0455800111	Detecting VMware SVGA3D driver. The SVGA3D driver is used on Windows as well, which means Mesa is now detected on Windows too.	9 years ago
Vladimír Vondruš	4b589e10ac	More consistent naming in Context::DetectedDriver. DetectedDriver::AMD is now DetectedDriver::Amd, DetectedDriver::ProbablyAngle is now just DetectedDriver::Angle. The old names are still present, but deprecated and will be removed in the future.	9 years ago
Vladimír Vondruš	6faf6c009a	On the other hand, most driver detection makes no sense on WebGL.	9 years ago
Vladimír Vondruš	cb2d590a33	Properly detect AMD and Intel drivers on ES builds as well. (Some) AMD and Intel cards / drivers support WGL/GLX_create_context_es2_profile as well.	9 years ago
Vladimír Vondruš	031da4f8a2	Bind some default VAO in case we are on core profile with ARB_VAO disabled. Otherwise it's not possible to render anything. The original behavior was accepted by NVidia, but it's not accepted by Mesa, e.g.	9 years ago
Vladimír Vondruš	fe1a5413c5	Added Context::isCoreProfile(). It's funny when implementing a new feature immediately involves doing a NV-specific workaround. Sigh.	9 years ago
Vladimír Vondruš	c870a1a6e8	Ability to detect Mesa drivers.	9 years ago
Alice Margatroid	58270fe9f0	push_back -> emplace_back	9 years ago
Vladimír Vondruš	7a16273e5d	Updated copyright year.	9 years ago
Vladimír Vondruš	9a7d3a1bec	Added TransformFeedback::maxVertexStreams() limit query.	10 years ago
Vladimír Vondruš	7bd41d2b93	Implemented support for desktop GLES on Windows. Because the library still links to the old crappy opengl32.dll, we need to load all symbols above OpenGL 1.1, not just those that are above OpenGL ES 2.0/3.0.	10 years ago
Vladimír Vondruš	c93cc0ba08	Added nv-windows-dangling-transform-feedback-varying-names workaround. On Windows NVidia drivers the glTransformFeedbackVaryings() does not make a copy of its char* arguments so it fails at link time when the original char arrays are not in scope anymore. Enabling synchronous debug output circumvents this bug. Can be triggered by running TransformFeedbackGLTest with GL_KHR_debug extension disabled.	10 years ago
Vladimír Vondruš	42d9679ac0	Exclude unreachable code from code coverage.	10 years ago
Vladimír Vondruš	fab154290c	Fix compilation on not-exactly-recent updates of MSVC 2015.	10 years ago
Vladimír Vondruš	4988c1ccf7	Ability to control the amount of console output in engine initialization.	10 years ago
Vladimír Vondruš	f68b40c5d9	Implemented *Framebuffer::copyImage() and copySubImage(). Default, ARB_DSA and EXT_DSA paths.	10 years ago
Vladimír Vondruš	f96b22e117	Platform: no-forward-compatible-core-context also on Intel/Windows. Detected in WindowlessWglApplication, added it to both Sdl2Application and WindowlessWglApplication. Renamed the workaround to reflect that it is for all three major vendors.	10 years ago
Vladimír Vondruš	01e0292361	Extracted CubeMapTexture::Coordinate enum into CubeMapCoordinate. Allows to break the dependency on the <Magnum/CubeMapTexture.h> header in Framebuffer, TextureState and elsewhere. The old CubeMapTexture::Coordinate enum is now just an alias, is marked as deprecated and will be removed in a future release.	10 years ago
Vladimír Vondruš	2f97617bb6	Minor cleanup and documentation.	10 years ago
Vladimír Vondruš	23646e6690	Hell, this project is big.	10 years ago
Vladimír Vondruš	d9b25c6057	Implemented ARB_shader_image_load_store.	10 years ago
Vladimír Vondruš	a1cae889df	Use Containers::Array instead of std::vector for non-resizable storage.	10 years ago
Vladimír Vondruš	3f0737fa98	Added missing ARB_compute_shader limit queries.	10 years ago
Vladimír Vondruš	b52d0e345d	Fix false-positive AMD/NVidia driver detection on OSX. Apple has its own drivers but still names the vendor the same way as if it were the Windows/Linux binary drivers. Weird and/or sneaky.	10 years ago
Vladimír Vondruš	395450cd5b	Fix compilation on Windows.	10 years ago
Vladimír Vondruš	d1714bfd1e	Context::current() now returns reference instead of pointer. 103% of use cases use the returned value directly without checking, so we might as well do the check ourselves. Added new function hasCurrent() and added deprecated backward-compatibility conversion and -> operators. Wow, that creeped to a lot of places. Last dinosaur from the pointer age.	10 years ago
Vladimír Vondruš	d5dae1a211	Added workaround for NVidia reporting compressed block size in bits. Instead of bytes as per specification of glGetInternalformativ().	10 years ago
Vladimír Vondruš	f746eaaf7d	Workaround broken full cube map compressed image query on NVidia. The query returns only data of the first cube map face and not the others, worked around that by querying each face separately.	10 years ago
Vladimír Vondruš	6bb6c4c398	Add workaround for inconsistent cubemap compressed image size on NVidia. The affected test case now passes again.	10 years ago
Vladimír Vondruš	9715e0c4b8	Add NVidia to list of detected drivers. I hoped this commit will never need to happen.	10 years ago
Vladimír Vondruš	7d0a5d9047	Make it possible to get predictable compressed cubemap image sizes. I expect the drivers to return size of one face when I'm querying that particular face using the pre-DSA or EXT_DSA API and size of all six faces when I'm querying whole texture using DSA API. One can dream, eh? It appears that, at least on my NVidia, the returned value does not depend on whether I'm querying all faces or a single one, but RATHER is based on whether the texture is immutable or not. How's that predictable at all?! Workaround in the next commit.	10 years ago
Vladimír Vondruš	27aea0ccd5	Don't spread cubemap texture level parameter query workarounds elsewhere. Pre-DSA code paths need to specify for which face we are querying the level parameters, which meant that all other calls had to specify the (implicit) target too. I'm also preparing to put a cubemap-specific workaround in the level parameter query and that really shouldn't be present in the generic implementation for all texture types. The other place where a specific target is needed is in setImage() implementations, but these are rather big chunks of code and I don't feel like copying these verbatim to cubemap implementation just to isolate the workaround in one place.	10 years ago
Vladimír Vondruš	b777135116	MSVC 2015 Update 1: in-class constexpr initialization works now.	11 years ago
Vladimír Vondruš	ea65ff39fb	Print an error if the user tries to disable unknown workaround.	11 years ago
Vladimír Vondruš	7b9e5184b7	Removed old amd-explicit-location-crash workaround. Seems to be not crashing anymore.	11 years ago
Vladimír Vondruš	3d9d70f153	Support for layered texture attachments in Framebuffer. I.e. rendering to a layer addressed by `gl_Layer` in geometry shader.	11 years ago
Vladimír Vondruš	291b5bd3a8	Fix Framebuffer::attachCubeMapTexture() behaving as a layered attachment. The DSA function does not accept any texture target parameter so the cube map texture was bound as a whole (and thus behaving as a layered attachment) instead of just a single face. Thanks to @chpatrick for the report.	11 years ago
Vladimír Vondruš	ded598128a	Recognize some more driver-specific workarounds.	11 years ago
Vladimír Vondruš	b0fd865e52	Fix build with CORRADE_NO_ASSERT. The variable is needed always.	11 years ago
Vladimír Vondruš	cbf3e427a1	Recognizing `amd-nv-no-forward-compatible-core-context` workaround.	11 years ago

1 2 3 4

186 Commits (56db82dc118d98e92676022e2177d56980bc7cab)