EGL contexts in GRDirectContext.MakeGL, mark2 #303
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
See python/cpython#128161 "Defining iterator in a separate class no longer works in 3.13" We have iterator for SkTextBlob defined by SkTextBlob::Iter(textblob), which is the c++/pybind11 equivalent of the same situation. Following the suggestion: python/cpython#128161 (comment) Also see actions/runner-images#11241 Fixes kyamagu#295
…o it earlier. See https://cibuildwheel.pypa.io/en/stable/options/#linux-image and pypa/cibuildwheel#1772 Fixes kyamagu#284 and kyamagu#175 The newer Linux has /usr/bin/ninja and does not have python
Comments about recovering from auditwheel repair changing the checksums to match current build
…L contexts with EGL There are two ways of enabling EGL support on Linux AFAIK: `skia_use_egl=true` switches over to X11 completely. Or we just add `GrGLInterfaces::MakeEGL` to GLX builds. This is the 2nd approach. `skia_use_egl=true` is simpler, but modifies existing X11 behavior. Fixes kyamagu#287
gn is really written for clang. Later version of gn has this fixed, but won't build on older g++ (12) for another problem. Sigh. Staying in this version of gn for the main builds. Fedora ships gn actually, though it is a lot more change to adapt to this.
On EL7, EGL/egl.h is in mesa-libEGL-devel instead of libglvnd-devel (later Redhat) Missing GLES2/gl2.h - in mesa-libGLES-devel It is not likely that the runtime (libgGL*) libraries are missing if the devel are installed, but you never know. Include them also.
Move the EGL test before the others to avoid messing the others up
See pypa/cibuildwheel#2118 Excluding libEGL.so.1 is enough.
|
Opencv has a setup for headless builds: https://github.com/opencv/opencv-python/blob/255564a37c4a23275485e827abbe19031347d458/.github/workflows/build_wheels_linux.yml#L129 - maybe we can do something similar. |
Native aarch64 linux CI hosts are fast enough. These workarounds are not needed anymore. Fixes kyamagu#315
… available. 3.11/3.12/3.13 (and 3.14 dev) are apparently available via setup-python. Another attempt at kyamagu#318
…40.0-231-g5b8860c391)
…; and pybind11/stl.h for optional
m138 adds a 2nd overload taking SkRecorder*: virtual bool isValid(GrRecordingContext* context) const = 0; virtual bool isValid(SkRecorder*) const = 0;
…heData m138 adds an overload taking size_t: void storeVkPipelineCacheData(); void storeVkPipelineCacheData(size_t maxSize);
Tagging v136 beta 13 to build.
… and removed in Python 3.13
…le on epel-release; hope this is an option on older yum
Merged
…--skip-unavailable option
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Tidied up, identical to and replaces #294 . @Swarzox
I don't think this should be merged as is, mainly for its additional dependency which most linux users don't need, as well as the fact that ideally, EGL builds should be built on the platform against the GPU vendor's libEGL libraries provided.
That said, since I added a CI stanza for building and testing on EGL headless fedora, I probably can make a mark 3 and just provide the difference as a patch, patch it in just before testing, and maintain such a patch indefinitely (until the skia folks change the surrounding GL code...).
So some of this might end up in m134 or later, just not in this form.