* chore: bump chromium in DEPS to 145.0.7572.0
* chore: update patches (trivial only)
* chore(patch-conflict): feat_filter_out_non-shareable_windows_in_the_current_application_in.patch
Polished the edits and formatted the result. No real changes.
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7232079 "7232079: Use WindowsToExclude to exclude PiP window in macOS screencapture device"
* chore(patch-conflict): feat_add_support_for_missing_dialog_features_to_shell_dialogs.patch
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7214586 "7214586: Refactor SelectFileDialogLinuxPortal to request XDG portal on demand"
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7237910 "7237910: Remove g_gtk_ui global"
* 7228586: Migrate SystemMemoryInfo from ByteCount to ByteSize
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7228586
* 7207583: GlobalRenderFrameHostId to ChildProcessId
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7207583
* 7205548: Remove uses of BodyAsStringCallbackDeprecated (extensions/)
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7205548
* chore: bump chromium in DEPS to 145.0.7574.0
* chore: update libc++ filenames
* chore: update patches (trivial only)
* chore(patch-conflict): feat_filter_out_non-shareable_windows_in_the_current_application_in.patch
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7246150 "7246150: Fix crash in ScreenCaptureKitDeviceMac due to null PIPScreenCaptureCoordinator"
* fixup! chore(patch-conflict): feat_add_support_for_missing_dialog_features_to_shell_dialogs.patch
* fixup! chore(patch-conflict): feat_filter_out_non-shareable_windows_in_the_current_application_in.patch
* 7239572: [OOPIF PDF] Enable OOPIF PDF by default on Windows/macOS/Linux
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7239572
* chore: bump chromium in DEPS to 145.0.7576.0
* fixup! chore(patch-conflict): feat_add_support_for_missing_dialog_features_to_shell_dialogs.patch
* chore: update patches (trivial only)
* chore: add note to keep patch that was upstreamed but reverted
fix_restore_original_resize_performance_on_macos.patch was organically upstreamed but that change got reverted just after the current roll's cutoff.
I've added a note in the patch contents so the patch sticks around and so we can keep it after the revert is included.
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7255334 "7255334: Revert 'Fix jank when resizing browser window'"
* chore: bump chromium in DEPS to 145.0.7577.0
* chore: update patches (trivial only)
* chore(patch-conflict): keep patch after revert
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7255334 "7255334: Revert 'Fix jank when resizing browser window'"
* 7237910: Remove g_gtk_ui global
I tried to find a way to avoid the patch, but other approaches seemed complex and involved.
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7237910
* 7251900: Reland "Remove callback_helpers.h include from is_callback.h (try 5)"
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7251900
* 7170174: [LNA] Retry requests for cached local resources
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/7170174
* fix: PDFs use OOPIF (behavior change)
* fixup! 7237910: Remove g_gtk_ui global
* fixup! 7251900: Reland "Remove callback_helpers.h include from is_callback.h (try 5)"
---------
Co-authored-by: electron-roller[bot] <84116207+electron-roller[bot]@users.noreply.github.com>
Co-authored-by: clavin <clavin@electronjs.org>
* refactor: make api::Clipboard::GetClipboardBuffer() private
* refactor: move GetClipboadBuffer() into anonymous namespace
* refactor: use gin::Arguments in StopRecording()
* refactor: use gin::Arguments in ImageView::New()
* refactor: use gin::Arguments in AppendSwitch()
* refactor: use gin::Arguments WebContentsView::New()
* refactor: make gin::Arguments arg const in WrappableBase::InitWithArgs()
This makes explicit that we are using it for wrapper + isolate, not the args values
* refactor: remove gin_helper::Arguments arg from ExposeAPI()
refactor: remove gin_helper::Arguments arg from ExposeAPIInWorld()
* refactor: remove gin_helper::Arguments arg from ElectronBindings::GetSystemMemoryInfo()
* refactor: remove gin_helper::Arguments arg from preload_utils::GetBinding()
* refactor: use gin::Arguments in OpenExternal()
* refactor: use gin::Arguments in ExecuteInWorld()
* refactor: use gin::Arguments in ExecuteJavaScript()
* refactor: use gin::Arguments in InvokeNew()
* refactor: use gin::Arguments in ExecuteJavaScriptInIsolatedWorld()
* refactor: remove unused GetNextArgument() marshaller for gin_helper::Arguments
* refactor: remove unused #include gin_helper/arguments.h
* chore: remove unused gin_helper::Arguments
* fixup! refactor: use gin::Arguments in ExecuteJavaScriptInIsolatedWorld()
Xref: https://github.com/electron/electron/pull/48447
* fix: Optimize the value of memory.free in the return data of getSystemMemoryInfo().
* fix: Improve the value of memory in the return data of getSystemMemoryInfo().
* fix: complete API doc.
* Update docs/api/process.md
Co-authored-by: Will Anderson <will@itsananderson.com>
* fix: update name to fileBacked.
* fix: fix with code conflict
---------
Co-authored-by: Will Anderson <will@itsananderson.com>
* Remove microtasks_scope.h and microtasks_scope.cc
* Use v8::MicrotasksScope when ignoring browser checkpoint
These call always skip the browser checkpoint, so they are equivalent to using v8::MicrotasksScope directly (modulo the optional wrapper behavior).
* Remove MicrotasksScope from node_bindings.cc
This code seems contradictory: it explicitly specifies "do not run microtasks" yet runs a microtask checkpoint in the browser process.
Looking at its history, it [was introduced][1] with the intention to not run microtasks, but a [subtle C++ language behavior][2] caused it to do the opposite later in the same roll. Since the original intention was to not run microtasks, and since that is also the simplest explanation, we can assume `ignore_browser_checkpoint` should be true and migrate this to `v8::MicrotasksScope` as it is equivalent (modulo the optional wrapper behavior).
[1]: a4ea80dd47 (diff-efe58cf03c97028f37f801db044d396a5f428686da6595d2c692f1c052bbd09c)
[2]: https://github.com/electron/electron/pull/43185
* Migrate gin_helper/promise.h and gin_helper/promise.cc to v8::MicrotasksScope
Restores the [original][1] behavior of running the microtask checkpoint at destruction, but preserves the behavior of running microtasks in the browser process. This had last changed in the migration to gin_helper::MicroTasks.
[1]: https://github.com/electron/electron/pull/16401
* fix: do not run microtasks in V8Serializer in browser process
* Remove no-op MicrotasksScope in `shell/browser/api/electron_api_auto_updater.cc`
This call was added in https://github.com/electron/electron/pull/40576 as an expansion of `gin_helper::EmitEvent`.
Since this only runs in the browser process and `bool ignore_browser_checkpoint = true` this code is a no-op.
Node should perform a microtask checkpoint if necessary in `node::MakeCallback`.
* Remove no-op MicrotasksScope in `shell/common/api/electron_bindings.cc`
This method is only called by the browser process. The containing function, `ElectronBindings::DidReceiveMemoryDump`, is only used in two places:
* `ElectronBindings::GetProcessMemoryInfo` in the same file, which has a `CHECK` that it's running in the browser process at the top.
* From `shell/browser/api/electron_api_web_contents.cc`, which is only run in the browser process.
Added a DCHECK for clarity and validation.
* Replace `gin_helper::MicrotasksScope` with `v8::MicrotasksScope` in `shell/renderer/`
The browser check is unnecessary in the renderer. Since `gin_helper::MicrotasksScope` will always act exactly like `v8::MicrotasksScope`, it's clear to just use the v8 object directly. This also brings them in line with the many other uses of `v8::MicrotasksScope` in `shell/renderer/`.
* refactor: move scope scaffolding into SettletScope
idea stolen from SpellCheckScope
* refactor: move impl of PromiseBase::RejectPromise() to the cc file
* chore: remove unused #include
This allows apps to measure their CPU usage over any given period
without worrying about other calls affecting the output,
as they would with `percentCPUUsage`.
* refactor: use std::erase() in WebContentsPreferences::~WebContentsPreferences()
* refactor: use std::erase() in WindowList::RemoveWindow()
* refactor: use std::erase() in ElectronBindings::EnvironmentDestroyed()
* refactor: use std::erase() in CleanedUpAtExit::~CleanedUpAtExit()
* refactor: use std::erase_if() in ElectronBrowserContext::RevokeDevicePermission()
* refactor: use std::erase_if() in UsbChooserController::GotUsbDeviceList()
* refactor: move DoesDeviceMatch() out of class into anonymous namespace
* refactor: use base::Contains() in KeyWeakMap::Has()
* refactor: use base::Contains() in WebRequest::RequestFilter::MatchesType()
* refactor: use base::Contains() in BaseWindow::AddBrowserView()
* refactor: use base::Contains() in DeepFreeze()
* refactor: use base::Contains() in Clipboard::Read()
* Revert "refactor: use base::Contains() in BaseWindow::AddBrowserView()"
This reverts commit 60152359d3978451ebdd7c8eed602c2fb8a9cafa.
* refactor: use base::Contains() in BaseWindow::AddBrowserView()
* refactor: use base::Contains() in IsDevToolsFileSystemAdded()
* refactor: use base::Contains() in MessagePort::DisentanglePorts()
* refactor: use base::Contains() in PowerSaveBlocker::IsStarted()
* refactor: use base::Contains() in SpellCheckClient::OnSpellCheckDone()
* refactor: use base::Contains() in ShowTaskDialogWstr()
* refactor: use base::Contains() in PrintViewManagerElectron::ScriptedPrint()
* refactor: use base::Contains() in PrintViewManagerElectron::DidGetPrintedPagesCount()
* refactor: use base::Contains() in NativeWindow::AddDraggableRegionProvider()
* refactor: use base::Contains() in ElectronBindings::ActivateUVLoop()
* refactor: use base::Contains() in NativeWindowViews::IsVisibleOnAllWorkspaces()
* refactor: use base::Contains() in HidChooserController::OnDeviceAdded()
* refactor: use base::Contains() in ElectronSandboxedRendererClient::WillReleaseScriptContext()
* refactor: use base::Contains() in ElectronRendererClient::WillDestroyWorkerContextOnWorkerThread()
* refactor: use base::Contains() in GlobalShortcut::OnKeyPressed()
* refactor: change defined(MAS_BUILD) to IS_MAS_BUILD()
This is missing-definition safe and thus allows us to move the definition of this macro away from "all compilation targets" to "just the compilation targets that depend on this macro".
In turn this makes the rebuild time changing from mas <-> darwin only 80 seconds on my machine, instead of the 12-15 minutes it used to take. This will also allow us in the future to build both MAS and darwin on the same CI machine. Costing us ~2 minutes on one machine but saving us anywhere from 30 minutes to an hour of CI time on other parts of the matrix.
* build: always define IS_MAS_BUILD even on non-mac builds
* build: use extra_configs
* refactor: add a wrapper for wrangling uv handles.
Part 1 of a fix for #25248, #22069.
Place the uv_asyncs owned by NodeBindings, ElectronBindings inside a new
UvHandle wrapper class which manages uv_handles' need for their closed()
callback to be invoked before the handles' memory can be freed.
* chore: make lint happy
* refactor: use DCHECK_EQ() instead of DCHECK()
* refactor: fix oops
* fix: use Node.js isolate setup logic in bindings
* Flags should be more process-specific
* Remove redundant isolate function setting
* Remove old SetFatalErrorHandler call
* fix: let Node.js perform microtask checkpoint in the main process
* fix: don't specify v8::MicrotasksScope for explicit policy
* fix: remove checkpoint from some call-sites
We already perform checkpoint at the end of a task,
either through MicrotaskRunner or through NodeBindings.
There isn't a need to add them again when calling into JS
except when dealing with promises.
* fix: remove checkpoint from some call-sites
We already perform checkpoint at the end of a task,
either through MicrotaskRunner or through NodeBindings.
There isn't a need to add them again when calling into JS
except when dealing with promises.
* fix incorrect specs
* default constructor arguments are considered for explicit mark
* add regression spec
* refactor: use gin in Promise
* refactor: separate Promise impl that returns nothing
* refactor: use Promise<void> for promise that returns nothing
* fix: methods should be able to run on both browser and renderer process
* fix: should not pass base::StringPiece across threads
* refactor: no more need to use different ResolvePromise for empty Promise
* refactor: move Promise to gin_helper
* refactor: make util::Promise type safe when chaining in native
* fixup! refactor: make util::Promise type safe when chaining in native
* chore: remove spare brackets
* chore: bump chromium in DEPS to 675d7dc9f3334b15c3ec28c27db3dc19b26bd12e
* chore: update patches
* chore: bump chromium in DEPS to dce3562696f165a324273fcb6893f0e1fef42ab1
* chore: const interfaces are being removed from //content
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1631749
Bug: https://bugs.chromium.org/p/chromium/issues/detail?id=908139
* chore: update patches
* chore: blink::MediaStreamType is now consistent and deduplicated
* chore: update patches and printing code for ref -> uniq
* chore: bridge_impl() --> GetInProcessNSWindowBridge
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1642988
* fixme: TotalMarkedObjectSize has been removed
* chore: fix linting
* chore: bump chromium in DEPS to 9503e1a2fcbf17db08094d8caae3e1407e918af3
* chore: fix slightly broken printing patch
* chore: update patches for SiteInstanceImpl changes
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1612025
* chore: update patches for SiteInstanceImpl changes
* chore: bump chromium in DEPS to 6801e6c1ddd1b7b73e594e97157ddd539ca335d7
* chore: update patches
* chore: bump chromium in DEPS to 27e198912d7c1767052ec785c22e2e88b2cb4d8b
* chore: remove system_request_context
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1647172
* chore: creation of FtpProtocolHandler needs an auth cache
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1639683
* fixme: disable marked spec
* chore: bump chromium in DEPS to 3dcd7fe453ad13a22b114b95f05590eba74c5471
* chore: bump chromium in DEPS to bdc24128b75008743d819e298557a53205706e7c
* chore: bump chromium in DEPS to 7da330b58fbe0ba94b9b94abbb8085bead220228
* update patches
* remove TotalMarkedObjectSize
https://chromium-review.googlesource.com/c/chromium/src/+/1631708
* add libvulkan.so to dist zip manifest on linux
* chore: bump chromium in DEPS to 1e85d0f45b52649efd0010cc9dab6d2804f24443
* update patches
* add angle features to gpuinfo
https://chromium-review.googlesource.com/c/chromium/src/+/1638658
* mark 'marked' property as deprecated
* disable webview resize test
* FIXME: disable vulkan on 32-bit arm
* chore: bump chromium in DEPS to cd0297c6a83fdd2b1f6bc312e7d5acca736a3c56
* Revert "FIXME: disable vulkan on 32-bit arm"
This reverts commit 5c1e0ef302a6db1e72231d4e823f91bb08e281af.
* backport from upstream: fix swiftshader build on arm
https://swiftshader-review.googlesource.com/c/SwiftShader/+/32768/
* update patches
* viz: update OutputDeviceWin to new shared memory api
https://chromium-review.googlesource.com/c/chromium/src/+/1649574
* base::Contains{Key,Value} => base::Contains
https://chromium-review.googlesource.com/c/chromium/src/+/1649478
* fixup! viz: update OutputDeviceWin to new shared memory api
* stub out StatusIconLinuxDbus-related delegate methods
https://chromium-review.googlesource.com/c/chromium/src/+/1638180
* chore: bump chromium in DEPS to 964ea3fd4bdc006d62533f5755043076220181f1
* Remove the BrowserContext methods to create URLRequestContexts for main/media partitions when a partition_domain is specified
https://chromium-review.googlesource.com/c/chromium/src/+/1655087
* fixup! stub out StatusIconLinuxDbus-related delegate methods
* add remote_cocoa to chromium_src deps
https://chromium-review.googlesource.com/c/chromium/src/+/1657068
* fixup! stub out StatusIconLinuxDbus-related delegate methods
* attempt at fix linux-debug build
* add swiftshader/libvulkan.so to arm manifest
* chore: bump chromium in DEPS to 28688f76afef27c36631aa274691e333ddecdc22
* update patches
* chore: bump chromium in DEPS to fe7450e1578a9584189f87d59d0d1a8548bf6b90
* chore: bump chromium in DEPS to f304dfd682dc86a755a6c49a16ee6876e0db45fb
* chore: bump chromium in DEPS to f0fd4d6c365aad9edd83bdfff9954c47d271b75c
* Update patches
* Remove no longer needed WOA patch
* Put back IOThread in BrowserProcess
We need this until we enable the network service.
* move atom.ico to inputs
* Update to latest LKGR to fix no template named 'bitset' in namespace 'std'
* fixup! Put back IOThread in BrowserProcess
* chore: bump chromium in DEPS to dcf9662dc9a896a175d791001350324167b1cad3
* Update patches
content_allow_embedder_to_prevent_locking_scheme_registry.patch is no longer necessary as it was upstreamed via https://chromium-review.googlesource.com/c/chromium/src/+/1637040
* Fix renamed enum
* Use newer docker container
Contains updated dependencies
* Try to track down arm test failures
* Fix arm tests
* chore: bump chromium in DEPS to 8cbceef57b37ee14b9c4c3405a3f7663922c5b5d
* Update patches
* Add needed dependencies for testing 32-bit linux
* Remove arm debugging.
* Remove additional debugging
* Fix compiler errors
* Handle new macOS helper
* Fix compile error on Linux
* chore: bump chromium in DEPS to 66a93991ddaff6a9f1b13d110959947cb03a1860
* Add new helper files to manifests
* fix BUILD.gn for macOS
* Fix compile errors
* Add patch to put back colors needed for autofill/datalist
* chore: bump chromium in DEPS to e89617079f11e33f33cdb3924f719a579c73704b
* Updated patches
* Remove no longer needed patch
* Remove no longer needed patch
* Fix compile error with patch
* Really fix the patch
* chore: bump chromium in DEPS to c70f12476a45840408f1d5ff5968e7f7ceaad9d4
* chore: bump chromium in DEPS to 06d2dd7a8933b41545a7c26349c802f570563fd5
* chore: bump chromium in DEPS to b0b9ff8f727deb519ccbec7cf1c8d9ed543d88ab
* Update patches
* Fix compiler errors
* Fix removed ChromeNetLog
* Revert "Fix removed ChromeNetLog"
This reverts commit 426dfd90b5ab0a9c1df415d71c88e8aed2bd5bbe.
* Remove ChromeNetLog.
https://chromium-review.googlesource.com/c/chromium/src/+/1663846
* chore: bump chromium in DEPS to fefcc4926d58dccd59ac95be65eab3a4ebfe2f29
* Update patches
* Update v8 patches
* Fix lint error
* Fix compile errors
* chore: bump chromium in DEPS to 4de815ef92ef2eef515506fe09bdc466526a8fd9
* Use custom protocol to test baseURLForDataURL
* Use newer SDK (10.0.18362) for Windows
* Update patches
* Update arm manifest since swiftshader reenabled.
* Don't delete dir that isn't ever there.
* Fix compile errors.
* Need src dir created
* Update for removed InspectorFrontendAPI.addExtensions
* Revert "Use newer SDK (10.0.18362) for Windows"
This reverts commit 68763a0c88cdc44b971462e49662aecc167d3d99.
* Revert "Need src dir created"
This reverts commit 7daedc29d0844316d4097648dde7f40f1a3848fb.
* Revert "Don't delete dir that isn't ever there."
This reverts commit bf424bc30ffcb23b1d9a634d4df410342536640e.
* chore: bump chromium in DEPS to 97dab6b0124ea53244caf123921b5d14893bcca7
* chore: bump chromium in DEPS to c87d16d49a85dc7122781f6c979d354c20f7f78b
* chore: bump chromium in DEPS to 004bcee2ea336687cedfda8f8a151806ac757d15
* chore: bump chromium in DEPS to 24428b26a9d15a013b2a253e1084ec3cb54b660b
* chore: bump chromium in DEPS to fd25914e875237df88035a6abf89a70bf1360b57
* Update patches
* Update node to fix build error
* Fix compile errors
* chore: bump chromium in DEPS to 3062b7cf090f1d9522c04ca8fa0a906f88ababe9
* chore: update node ref for pushed tags
* chore: update patches for new chromium
* chore: fix printing patches
* Use new (10.0.18362) Windows SDK
* roll node to fix v8 build issues in debug build
* Add support for plugin helper
* fix: add patch to fix gpu info enumeration
Can be removed once CL lands upstream.
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1685993
* spec: navigator.requestMIDIAccess now requires a secure origin
This test requires a secure origin so we fake one.
Refs: https://chromium-review.googlesource.com/c/chromium/src/+/1657952
* FIXME: temporarily disable SharedWorker tests
* use released version of node-abstractsocket
* fix abstract-socket
