- Add widget.wayland_dmabuf_backend.enabled
- Expose it to Wayland backend by WaylandDisplay object
- Add more logging to Wayland DMABuf backend
Differential Revision: https://phabricator.services.mozilla.com/D33919
--HG--
extra : moz-landing-system : lando
To run task_for_pid() on child processes, we need the child task port for
security reasons. This port can be obtained via a Mach IPC exchange.
This is what GeckoChildProcessHost::GetChildTask() provides, so we use it
in cocoa's version of GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D25927
--HG--
extra : moz-landing-system : lando
- Add widget.wayland_dmabuf_backend.enabled
- Expose it to Wayland backend by WaylandDisplay object
- Add more logging to Wayland DMABuf backend
Differential Revision: https://phabricator.services.mozilla.com/D33919
--HG--
extra : moz-landing-system : lando
- Don't block all drawing until whole screen was painted, clear the screen and
allow partial drawing instead.
- Detect window resize and allow partial drawing directly to wayland buffer in this case.
- Remove backbuffer copy - wait to buffer release instead.
Depends on D33910
Differential Revision: https://phabricator.services.mozilla.com/D33911
--HG--
extra : moz-landing-system : lando
To run task_for_pid() on child processes, we need the child task port for
security reasons. This port can be obtained via a Mach IPC exchange.
This is what GeckoChildProcessHost::GetChildTask() provides, so we use it
in cocoa's version of GetProcInfo()
Differential Revision: https://phabricator.services.mozilla.com/D25927
--HG--
extra : moz-landing-system : lando
Because of the way the canvas 2D implementation juggles between path builders and paths, we have to keep track of current and first points in the path, in case a builder is created out of it.
Differential Revision: https://phabricator.services.mozilla.com/D33757
--HG--
extra : source : c79b3af2395894b449731fccdd9eb006b4d0189e
gfxPlatform::GetSourceSurfaceForSurface and CreateDrawTargetForSurface
are both static methods that we sometimes use via the pattern
gfxPlatform::GetPlatform()->... This is problematic because this forces
gfxPlatform to be initialized in the process, and in the GPU process, we
don't support this. It should be safe to call these methods without
initializing gfxPlatform, so this patch removes the GetPlatform() call.
GetSourceSurfaceForSurface may end up initializing gfxPlatform anyways,
depending on whether or not a DrawTarget was given. This should not be a
concern for the crashes observed in bug 1435586.
Differential Revision: https://phabricator.services.mozilla.com/D33785
Make nsIWebBrowserPrint included unconditionally for addressing nsIDocShell.rs build failure.
Remove direct_call of PPrinting and PRemotePrintJob. Their ipdl are built unconditionally, but their derived classes are not built with --disable-printing.
Differential Revision: https://phabricator.services.mozilla.com/D33391
--HG--
extra : moz-landing-system : lando
- Implement WaylandDMABufSurface class which provides basic 2D surfaces located at GPU memory.
- Implement ability to bind it as wl_buffer for direct wayland rendering and R/W mapping for CPU
Differential Revision: https://phabricator.services.mozilla.com/D33597
--HG--
extra : moz-landing-system : lando
- Implement nsGbmLib object which binds symbols run-time from libgbm.so library.
- Load available DRM pixel formats from Wayland registry and store related modifiers at WaylandDisplay object.
- Implement Gbm device setup at WaylandDisplay.
Differential Revision: https://phabricator.services.mozilla.com/D32633
--HG--
extra : moz-landing-system : lando
Only gtk returns failure ever, and nobody checks the result anyway.
Use an enum class so that it's clear from the caller what it means.
Differential Revision: https://phabricator.services.mozilla.com/D32353
--HG--
extra : moz-landing-system : lando
When transparent popup window is rendered clear window background to clear
visual artifacts and bold shadows over the popup.
Differential Revision: https://phabricator.services.mozilla.com/D33155
--HG--
extra : moz-landing-system : lando
Only gtk returns failure ever, and nobody checks the result anyway.
Use an enum class so that it's clear from the caller what it means.
Differential Revision: https://phabricator.services.mozilla.com/D32353
--HG--
extra : moz-landing-system : lando
- Rename recent WindowBackBuffer class to WindowBackBufferShm to clearly state that it uses Shm memory to store the pixel buffer.
- Implement WindowBackBufferDMABuf which stores pixel data in GPU memory in WaylandDMABufSurface object.
- Use WaylandDMABufSurface as a wayland backend when DMABuf is available and gfx.wayland_dmabuf_backend.enabled is set.
- Implement WindowImageSurface which temporary stores front buffer pixel data. It's used when front buffer is used by compositor and we want to draw.
Instead of the front/back buffer flip and read-back data from front buffer, don't flip, store the drawing and draw the pixels when compositor
releases the front buffer.
Differential Revision: https://phabricator.services.mozilla.com/D32635
--HG--
extra : moz-landing-system : lando
Ship Wayland DMABuf protocol headers. This is a non-offical extension thus it's missing in stable wayland headers.
Differential Revision: https://phabricator.services.mozilla.com/D32632
--HG--
extra : moz-landing-system : lando
Recently we track and position all popups on Wayland which leads to
wrong placement when parent popup window is hidden.
In this patch we use plain gtk_window_move() for hidden popups and
we don't track them so they can't be used as a parent window.
Also implement nsWindow::HideWaylandPopupAndAllChildren() to clearly
hide a popup window with and its children windows.
Differential Revision: https://phabricator.services.mozilla.com/D31056
--HG--
extra : moz-landing-system : lando
This fix wrong rendering of radio and checkbox indicators in menus by
correcting the dimensions acording to used theme.
Differential Revision: https://phabricator.services.mozilla.com/D32316
--HG--
extra : moz-landing-system : lando
If OnKeysChanged was called after keyboard_handle_keymap, then our newly configured keymap would be immediately cleared again. In order to avoid such scenario, use only keyboard_handle_keymap on wayland.
Differential Revision: https://phabricator.services.mozilla.com/D32599
--HG--
extra : moz-landing-system : lando
Use standard Gtk/Gnome way to set Firefox Gtk theme. Gtk theme of web content can be still configured
by widget.content.allow-gtk-dark-theme and widget.content.gtk-theme-override.
Depends on D29823
Differential Revision: https://phabricator.services.mozilla.com/D30919
--HG--
extra : moz-landing-system : lando
We previously disabled system dark themes for Firefox chrome and content by gtk-application-prefer-dark-theme settings.
That option is no longer preferred by gnome project and was removed from tweaks tools. Theme makers are encouraged to use
a different name for the dark theme variants, like Adwaita-dark, Yaru-dark and so on. This option also does not
work when the GTK theme is missing the light variant completely.
To address that, this patch implements heuristics based on https://www.w3.org/TR/AERT/#color-contrast
to check if the system GTK theme has good contrast/visibility with default HTML colors
(white background and black text).
If widget.content.gtk-theme-override is empty and the system theme fails the test with gtk-application-prefer-dark-theme set to false, then Adwaita:light theme is used.
This patch was tested with some distro default light themes (Ambiance, Radiance, Yaru - Ubuntu, Arc - KDE, Menta - MATE)
and dark/light themes are recognized correctly.
Differential Revision: https://phabricator.services.mozilla.com/D29823
--HG--
extra : moz-landing-system : lando
This should be an idempotent patch. The way to come up with this patch has been:
* Run the first script attached to the bug and pipe it to xclip, then paste it
in color.rs
* Add the relevant #[derive] annotations and remove the color.mako.rs
definition.
* Reorder the values to match the ColorID definition, on which some widget
prefs and caching stuff relies on.
* Manually port some documentation from nsLookAndFeel.h
* Run `rg 'eColorID_' | cut -d : -f 1 | sort | uniq >files`
* Run the second script attached to the bug.
* Manually fix usage of `LAST_COLOR` (adding the `End` variant), and adding
casts to integer as needed.
* Add an static assert so that people remember to update the prefs, rather than
a comment on the definition :)
Differential Revision: https://phabricator.services.mozilla.com/D32610
--HG--
extra : moz-landing-system : lando
- Follow Gtk and get theme button text color directly from "button" CSS node instead of "button label"
- Provide new -moz-gtk-buttonactivetext color for active/pressed button text color
- Replace ButtonText color with -moz-gtk-buttonactivetext when it's appropriate
Differential Revision: https://phabricator.services.mozilla.com/D30566
--HG--
extra : moz-landing-system : lando
Martin Stransky
Boris Zbarsky
Tarek Ziadé
Dorel Luca
Sebastian Hengst
Mihai Alexandru Michis
Nicolas Silva
Andrew Osmond
sotaro
Emilio Cobos Álvarez
Jonathan Watt
Jan Horak
Brindusan Cristian
Kenny Levinsen
Coroiu Cristina
Razvan Maries
myfreeweb
arthur.iakab
Sylvestre Ledru
Lee Salzman