Igalia / meta-webkit

Yocto / OpenEmbedded layer for WebKit based engines and browsers
MIT License
125 stars 69 forks source link

wpewebkit,webkitgtk: Bump up version to 2.36.3 #375

Closed psaavedra closed 2 years ago

psaavedra commented 2 years ago

wpewebkit:

webkitgtk:

aperezdc commented 2 years ago

Beware of https://bugs.webkit.org/show_bug.cgi?id=241109 though! You may want to import the patch once I have it ready.

psaavedra commented 2 years ago

Beware of https://bugs.webkit.org/show_bug.cgi?id=241109 though! You may want to import the patch once I have it ready.

Thanks for the advice. We don't have a CI job for raspberrry 3 or 4 in 64 bits. I just created one for the model 4 in https://github.com/Igalia/meta-webkit/pull/376. I will integrate this MR before this in order to ensure the build error is not happening here.

aperezdc commented 2 years ago

Beware of https://bugs.webkit.org/show_bug.cgi?id=241109 though! You may want to import the patch once I have it ready.

Thanks for the advice. We don't have a CI job for raspberrry 3 or 4 in 64 bits. I just created one for the model 4 in #376. I will integrate this MR before this in order to ensure the build error is not happening here.

That sounds like a good plan 👌🏼. It can happen that with the build settings used the issue won't show up if the unified build compilation units get arranged in a different way than my test builds for Buildroot. The safest would be to apply the patch, but of course that can be done later in a follow-up PR.

psaavedra commented 2 years ago

Beware of https://bugs.webkit.org/show_bug.cgi?id=241109 though! You may want to import the patch once I have it ready.

Thanks for the advice. We don't have a CI job for raspberrry 3 or 4 in 64 bits. I just created one for the model 4 in #376. I will integrate this MR before this in order to ensure the build error is not happening here.

That sounds like a good plan 👌🏼. It can happen that with the build settings used the issue won't show up if the unified build compilation units get arranged in a different way than my test builds for Buildroot. The safest would be to apply the patch, but of course that can be done later in a follow-up PR.

just for curious, do you have -DUSE_ATSPI=ON in the buildroot cmake config line?

github-actions[bot] commented 2 years ago

Pull Request is being checked for https://github.com/Igalia/meta-webkit/pull/375/commits

aperezdc commented 2 years ago

Beware of https://bugs.webkit.org/show_bug.cgi?id=241109 though! You may want to import the patch once I have it ready.

Thanks for the advice. We don't have a CI job for raspberrry 3 or 4 in 64 bits. I just created one for the model 4 in #376. I will integrate this MR before this in order to ensure the build error is not happening here.

That sounds like a good plan 👌🏼. It can happen that with the build settings used the issue won't show up if the unified build compilation units get arranged in a different way than my test builds for Buildroot. The safest would be to apply the patch, but of course that can be done later in a follow-up PR.

just for curious, do you have -DUSE_ATSPI=ON in the buildroot cmake config line?

The USE_ATSPI option is enabled by default when ENABLE_ACCESSIBILITY is enabled. Both Source/cmake/OptionsGTK.cmake and Source/cmake/OptionsWPE.cmake contain:

SET_AND_EXPOSE_TO_BUILD(USE_ATSPI ${ENABLE_ACCESSIBILITY})

For Buildroot's WebKitGTK package accessibility is enabled (which imples USE_ATSPI), while for the WPE WebKit package it's disabled.

psaavedra commented 2 years ago

The new build test for rpi4-64 passes. https://gitlab.com/browsers/meta-webkit/-/jobs/2526168429#L255. I will merge this PR and I will move your proposal to https://github.com/Igalia/meta-webkit/issues/377