Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | | Remove a panic due to missing video support | Anthony Ramine | 2018-09-14 | 1 | -2/+4 | |
| | | ||||||
* | | Remove erroneous check from CopyTexImage2D | Anthony Ramine | 2018-09-13 | 1 | -16/+0 | |
| | | | | | | | | | | The internal format of the bound texture doesn't matter, what matters is which components can be found in the framebuffer. | |||||
* | | Properly support PACK_ALIGNMENT in WebGL 1 | Anthony Ramine | 2018-09-12 | 1 | -83/+112 | |
| | | ||||||
* | | Simplify WebGLRenderingContext::PixelStorei | Anthony Ramine | 2018-09-12 | 1 | -25/+8 | |
| | | ||||||
* | | Auto merge of #21461 - jdm:webgltmp2, r=nox | bors-servo | 2018-09-12 | 1 | -32/+31 | |
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Various webgl fixes for framebuffer attachment test These changes resolve all panics on macOS when running framebuffer-object-attachment.html in headless and headful testing. --- - [x] `./mach build -d` does not report any errors - [x] `./mach test-tidy` does not report any errors - [x] There are tests for these changes OR - [x] Fixes #13710. Fixes #20570. <!-- Reviewable:start --> --- This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/21461) <!-- Reviewable:end --> | |||||
| * | | webgl: Ensure that depth and stencil attachments are rebound after messing ↵ | Josh Matthews | 2018-09-10 | 1 | -26/+8 | |
| | | | | | | | | | | | | with DEPTH_STENCIL attachments. | |||||
| * | | webgl: Move framebuffer initialization logic to WebGL thread. | Josh Matthews | 2018-09-10 | 1 | -2/+15 | |
| | | | ||||||
| * | | webgl: Differentiate between missing colour attachments and incomplete ↵ | Josh Matthews | 2018-09-10 | 1 | -4/+8 | |
| | | | | | | | | | | | | attachments. | |||||
| * | | webgl: Ensure that framebuffers have a color attachment before reading or ↵ | Josh Matthews | 2018-09-10 | 1 | -1/+1 | |
| | | | | | | | | | | | | writing. | |||||
* | | | Reuse input buffer for RGB/UNSIGNED_BYTE in rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-10 | 1 | -7/+8 | |
| | | | ||||||
* | | | Reuse input vector for LUMINANCE_ALPHA/HALF_FLOAT | Anthony Ramine | 2018-09-10 | 1 | -5/+6 | |
| | | | ||||||
* | | | Fix the remaining LUMINANCE* cases in rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-10 | 1 | -18/+5 | |
|/ / | ||||||
* | | Fix the UNSIGNED_SHORT_4_4_4_4 case of premultiply_pixels | Anthony Ramine | 2018-09-09 | 1 | -9/+9 | |
| | | ||||||
* | | Fix the UNSIGNED_SHORT_5_5_5_1 case of premultiply_pixels | Anthony Ramine | 2018-09-09 | 1 | -2/+3 | |
| | | ||||||
* | | Remove some useless argument mutability | Anthony Ramine | 2018-09-09 | 1 | -2/+2 | |
| | | | | | | | | I wonder why this doesn't trigger a warning. | |||||
* | | Fix the LUMINANCE of rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-09 | 1 | -4/+1 | |
| | | | | | | | | | | The internal format and format values match, thus the canvas or image input must be interpreted as already in LUMINANCE format. | |||||
* | | Fix the LUMINANCE_ALPHA case of rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-09 | 1 | -1/+1 | |
| | | | | | | | | | | The internal format and format values match, thus the canvas or image input must be interpreted as already in LUMINANCE_ALPHA format. | |||||
* | | Reuse the input vector in more cases of rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-08 | 1 | -56/+70 | |
| | | ||||||
* | | Fix the ALPHA/FLOAT case of rgba8_image_to_tex_image_data | Anthony Ramine | 2018-09-08 | 1 | -1/+1 | |
| | | | | | | | | I'm pretty sure this is supposed to write the ALPHA component. | |||||
* | | Reuse input of rgba8_image_to_tex_image_data in some cases | Anthony Ramine | 2018-09-08 | 1 | -9/+8 | |
| | | ||||||
* | | Make rgba8_image_to_tex_image_data not overallocate in some case | Anthony Ramine | 2018-09-08 | 1 | -1/+1 | |
| | | ||||||
* | | Make rgba8_image_to_tex_image_data a free-standing function | Anthony Ramine | 2018-09-08 | 1 | -201/+202 | |
| | | ||||||
* | | Make remove_premultiplied_alpha mutate its input | Anthony Ramine | 2018-09-08 | 1 | -16/+15 | |
| | | ||||||
* | | Make premultiply_pixels mutate its input | Anthony Ramine | 2018-09-08 | 1 | -46/+28 | |
| | | ||||||
* | | Use byte channels to send textures to the WebGL thread | Anthony Ramine | 2018-09-08 | 1 | -10/+10 | |
| | | ||||||
* | | Use ipc::bytes_channel in ReadPixels | Anthony Ramine | 2018-09-08 | 1 | -3/+3 | |
| | | ||||||
* | | Simplify WebGLRenderingContext::get_image_data | Anthony Ramine | 2018-09-08 | 1 | -8/+8 | |
| | | ||||||
* | | Use a bytes channel in BufferData | Anthony Ramine | 2018-09-07 | 1 | -14/+23 | |
| | | | | | | | | This means we don't need to copy the input ArrayBuffer at all on the DOM side. | |||||
* | | Simplify WebGLBuffer::buffer_data | Anthony Ramine | 2018-09-07 | 1 | -14/+5 | |
| | | | | | | | | | | There is no need to pass the target to that buffer method, given the buffer has been retrieved by looking up the one bound to that target in the context. | |||||
* | | Make validate_framebuffer return a WebGLResult<()> | Anthony Ramine | 2018-09-06 | 1 | -40/+15 | |
| | | ||||||
* | | Use WebGLResult for returns of instanced draw methods | Anthony Ramine | 2018-09-06 | 1 | -46/+27 | |
| | | ||||||
* | | Support unions of objects in overloads | Anthony Ramine | 2018-08-30 | 1 | -35/+14 | |
| | | | | | | | | Part of #20513, implementing the parts useful for WebGL. | |||||
* | | Fix gl.isBuffer for buffers that are marked for deletion but still attached | Anthony Ramine | 2018-08-30 | 1 | -1/+1 | |
| | | ||||||
* | | Always emit INVALID_OPERATION on null element buffers in drawElements | Anthony Ramine | 2018-08-30 | 1 | -7/+9 | |
| | | ||||||
* | | Fix the error for invalid arrays passed to gl.vertexAttrib*v() | Anthony Ramine | 2018-08-30 | 1 | -4/+8 | |
| | | ||||||
* | | Fix the build for NLL | Simon Sapin | 2018-08-24 | 1 | -3/+3 | |
| | | | | | | | | | | | | Test with `RUSTFLAGS="-Zborrowck=mir -Ztwo-phase-borrows" cargo build` https://internals.rust-lang.org/t/help-us-get-non-lexical-lifetimes-nll-over-the-finish-line/7807/7 | |||||
* | | Properly check limit in gl.activeTexture() | Anthony Ramine | 2018-08-23 | 1 | -139/+153 | |
| | | ||||||
* | | Revert "Fix the build for NLL" | Josh Matthews | 2018-08-07 | 1 | -3/+3 | |
| | | | | | | | | This reverts commit d1733aa5029c5b97390a236d94eed916ddb64577. | |||||
* | | Fix the build for NLL | Simon Sapin | 2018-08-08 | 1 | -3/+3 | |
|/ | | | | | | Test with `RUSTFLAGS="-Zborrowck=mir -Ztwo-phase-borrows" cargo build` https://internals.rust-lang.org/t/help-us-get-non-lexical-lifetimes-nll-over-the-finish-line/7807/7 | |||||
* | Merge code from Draw* and Draw*Instanced methods | Anthony Ramine | 2018-08-02 | 1 | -122/+22 | |
| | | | | | This made me realise we weren't supporting OES_element_index_uint in the ANGLE_instanced_arrays extension. | |||||
* | Auto merge of #21324 - servo:webgl, r=emilio | bors-servo | 2018-08-02 | 1 | -9/+18 | |
|\ | | | | | | | | | | | | | | | Properly set initial values for WebGL texture filters <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/21324) <!-- Reviewable:end --> | |||||
| * | Use the DOM cache for gl.getTexParameter(gl.TEXTURE_*_FILTER) | Anthony Ramine | 2018-08-02 | 1 | -9/+18 | |
| | | | | | | | | Part of #20596. | |||||
* | | Auto merge of #21313 - jdm:glstuff, r=nox | bors-servo | 2018-08-02 | 1 | -3/+16 | |
|\ \ | |/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Framebuffer and renderbuffer fixes This commits address two separate panics that occur when running the framebuffer-object-attachment.html test. The test still panics due to another framebuffer completion status problem, so the overall test results don't demonstrate any improvement. --- - [x] `./mach build -d` does not report any errors - [x] `./mach test-tidy` does not report any errors - [x] These changes fix #21252 - [x] There are tests for these changes <!-- Reviewable:start --> --- This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/21313) <!-- Reviewable:end --> | |||||
| * | webgl: Emulate some renderbuffer formats in non-GLES. | Josh Matthews | 2018-08-02 | 1 | -3/+16 | |
| | | ||||||
* | | Always use a WebGLVertexArrayObject to handle vertex attribs | Anthony Ramine | 2018-08-02 | 1 | -316/+148 | |
|/ | | | | This lets us clean up how buffers are reference-counted. | |||||
* | Properly check for context ownership of objects passed to gl.is*() | Anthony Ramine | 2018-07-31 | 1 | -6/+14 | |
| | ||||||
* | Simplify gl.getShaderParameter() (fixes #20562) | Anthony Ramine | 2018-07-31 | 1 | -12/+8 | |
| | | | | | | The expectation change is due to the shader now using its DOM-side compile status. It is actually a bug for the shader to think it actually compiled successfully, but at least it does so consistently now. | |||||
* | Fix program and shader lifetime cycle | Anthony Ramine | 2018-07-31 | 1 | -9/+20 | |
| | ||||||
* | Cache which capabilities are enabled in the context (fixes #20534) | Anthony Ramine | 2018-07-25 | 1 | -25/+75 | |
| | | | | This is needed for #20555. | |||||
* | Properly check for GL object ownership (fixes #21133) | Anthony Ramine | 2018-07-24 | 1 | -16/+81 | |
| |