electron/patches/chromium/disable_compositor_recycling.patch
electron-roller[bot] 8c89137ecd
chore: bump chromium to 122.0.6194.0 (main) (#40750)
* chore: bump chromium in DEPS to 122.0.6180.0

* chore: update render_widget_host_view_base.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5079532

minor manual changes to match upstream changes

* chore: update printing.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5094777

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5054299

Update patch to upstream changes in print_view_manager_base.cc.
CC @codebytere for 👀

* chore: update feat_add_onclose_to_messageport.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5083156

no manual changes; patch applied with fuzz 1

* chore: update build_do_not_depend_on_packed_resource_integrity.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5087072

minor manual changes to match upstream changes

* chore: update port_autofill_colors_to_the_color_pipeline.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/4960481

no manual changes; patch applied with fuzz 2

* chore: update revert_api_dcheck-fail_when_we_reenter_v8_while_terminating.patch

Xref: https://chromium-review.googlesource.com/c/v8/v8/+/5076902
Xref: https://chromium-review.googlesource.com/c/v8/v8/+/5096253

The `bailout_value` variable, which allowed us to fail gracefully in
these macros, has been removed upstream. This raises the temperature
on https://github.com/electron/electron/issues/40741 ...

Since we don't have a bailout value, return a default-constructed
instance of whatever the return type is of the function that invokes
the macro.

* Remove performActionOnRemotePage

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5068017

It appears it is not used by anything within Chromium or outside.

* perf: in SubmenuButton ctor, move() the PressedCallback

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5066872

The copy construtor is more expensive and has been removed upstream

* chore: update constructor call for ElectronExtensionsRendererClient::dispatcher_

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5055146

* chore: handle CAPTURED_SURFACE_CONTROL in Converter<blink::PermissionType>::ToV8()

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5096614

* chore: e patches all

* chore: bump chromium to 122.0.6181.0

* chore: e patches all

* refactor: Add lookup for per-RenderFrame SpellCheckProvider

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5075094

Previously we could pass a routing ID into the spellcheck methods.
This ID arg has been replaced with a SpellCheckHost& arg.

We can get the SpellCheckHost from an accessor in SpellCheckProvider,
which was already being created on a per-render-frame basis in
RenderClientBase.

This commit adds a RenderFrame-to-SpellCheckProvider accessor
in RenderClientBase so that it can be used to get that needed
SpellCheckHost.

* chore: make lint happy

* chore: restore the will_cause_resize arg to WebContents::ExitFullscreen

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5074445

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5031312

upstream reverted this change

* chore: bump chromium in DEPS to 122.0.6182.0

* chore: update patches

* chore: fix virtual function names that were renamed upstream

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5115241

* chore: handle new blink::PermissionType::SMART_CARD in PermissionType converter

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5049989

* chore: node script/gen-libc++-filenames.js

* fixup! chore: update build_do_not_depend_on_packed_resource_integrity.patch

* chore: remove revert_api_dcheck-fail_when_we_reenter_v8_while_terminating.patch

The previous experiment of removing the DCHECK in 35093783 was a huge
failure. We either need to either:

A) use upstream as-is and fix the DCHECK failure, or
B) revert https://chromium-review.googlesource.com/c/v8/v8/+/5076902 too
C) find some other workaround

This commit is the first half of option A, removing the patch to use
upstream as-is.

* chore: bump chromium in DEPS to 122.0.6184.0

* chore: bump chromium in DEPS to 122.0.6186.0

* chore; update mas_avoid_private_macos_api_usage.patch.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5085619

minor manual changes to match upstream changes

* chore: update build_do_not_depend_on_packed_resource_integrity.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5120922

minor manual changes to match upstream changes

* chore: update fix_move_autopipsettingshelper_behind_branding_buildflag.patch

Xref: https://chromium-review.googlesource.com/c/chromium/src/+/5041595

no manual changes; patch applied with fuzz 1

* chore: e patches all

* chore: add refactor_update_node-platform_to_match_v8-platform_api_changes.patch

* chore: bump chromium in DEPS to 122.0.6188.0

* chore: bump chromium in DEPS to 122.0.6190.0

* chore: bump chromium in DEPS to 122.0.6192.0

* chore: bump chromium in DEPS to 122.0.6194.0

* Remove an unnecessary DCHECK which causes Lacros to crash.

https://chromium-review.googlesource.com/c/chromium/src/+/5124880

* [FixIt Week] Add shared regular profile function that includes ash

https://chromium-review.googlesource.com/c/chromium/src/+/5100709

* Simplify the PrintManagerHost::ScriptedPrint() call

https://chromium-review.googlesource.com/c/chromium/src/+/5019703

* [local compile hints] Add tests

https://chromium-review.googlesource.com/c/chromium/src/+/5024173

* chore: update patches

* fixup! chore: add refactor_update_node-platform_to_match_v8-platform_api_changes.patch

* chore: node script/gen-libc++-filenames.js

* Remove //ui/ozone/buildflags.h

https://chromium-review.googlesource.com/c/chromium/src/+/5119048

* fixup! Remove //ui/ozone/buildflags.h

* fixup! Remove //ui/ozone/buildflags.h

* chore: update patches

* fixup! refactor: Add lookup for per-RenderFrame SpellCheckProvider

* fixup! Remove //ui/ozone/buildflags.h

* Add permission policy verification in PermissionControllerImpl

https://chromium-review.googlesource.com/c/chromium/src/+/5059112

* fix: ensure spellcheck interface is registered for frame

* [exceptions] Unify pending and scheduled exceptions

https://chromium-review.googlesource.com/c/v8/v8/+/5050065

* fix: dchecks in node.js tests

* fix: try adding termination check

* chore: document breaking iframe change

---------

Co-authored-by: electron-roller[bot] <84116207+electron-roller[bot]@users.noreply.github.com>
Co-authored-by: Charles Kerr <charles@charleskerr.com>
Co-authored-by: PatchUp <73610968+patchup[bot]@users.noreply.github.com>
Co-authored-by: deepak1556 <hop2deep@gmail.com>
Co-authored-by: Shelley Vohr <shelley.vohr@gmail.com>
2024-01-03 09:52:49 +01:00

24 lines
1.7 KiB
Diff

From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Andy Locascio <andy@slack-corp.com>
Date: Wed, 21 Aug 2019 12:09:10 -0700
Subject: fix: disabling compositor recycling
Compositor recycling is useful for Chrome because there can be many tabs and spinning up a compositor for each one would be costly. In practice, Chrome uses the parent compositor code path of browser_compositor_view_mac.mm; the NSView of each tab is detached when it's hidden and attached when it's shown. For Electron, there is no parent compositor, so we're forced into the "own compositor" code path, which seems to be non-optimal and pretty ruthless in terms of the release of resources. Electron has no real concept of multiple tabs per window, so it should be okay to disable this ruthless recycling altogether in Electron.
diff --git a/content/browser/renderer_host/render_widget_host_view_mac.mm b/content/browser/renderer_host/render_widget_host_view_mac.mm
index fae7865cda633ae96df60b2d8d6beb1384031969..e0a3dbccbab10f7936910fc131750f6fbc56b00e 100644
--- a/content/browser/renderer_host/render_widget_host_view_mac.mm
+++ b/content/browser/renderer_host/render_widget_host_view_mac.mm
@@ -557,7 +557,11 @@
return;
host()->WasHidden();
- browser_compositor_->SetRenderWidgetHostIsHidden(true);
+ // Consider the RWHV occluded only if it is not attached to a window
+ // (e.g. unattached BrowserView). Otherwise we treat it as visible to
+ // prevent unnecessary compositor recycling.
+ const bool unattached = ![GetInProcessNSView() window];
+ browser_compositor_->SetRenderWidgetHostIsHidden(unattached);
}
void RenderWidgetHostViewMac::SetSize(const gfx::Size& size) {