Bug 1568889 enabled the rustfmt lint checks in continuous integration,
and while they did not cause any errors, this is a follow-up patch
to fix the non-fatal warnings.
Differential Revision: https://phabricator.services.mozilla.com/D39484
--HG--
extra : moz-landing-system : lando
If the subdocument display item gets created when the subdocument doesn't have a root frame it'll be based on the subdocument frame in the parent document. This means that the prescontext pointer on the document viewer of the subdocument can go away without the subdocument display item getting any kind of invalidation. So we make sure the subdocument display item gets rebuilt if that happens. The reason the prescontext pointer is important is that is what nsSubDocumentFrame::GetSubdocumentPresShellForPainting uses (via nsDocShell::GetPresShell) to get the subdocument presshell. So it will fail to get a presshell and RetainedDisplayListBuilder::IncrementSubDocPresShellPaintCount will crash.
Differential Revision: https://phabricator.services.mozilla.com/D39310
--HG--
extra : moz-landing-system : lando
If copying selected password is allowed by `TextEditor`, we should allow to
drag it too because web apps usually request to input new password twice,
but if users used password generator, they may want to use drag and drop the
new password rather than copy and paste.
Differential Revision: https://phabricator.services.mozilla.com/D39279
--HG--
extra : moz-landing-system : lando
First, we need to make `nsCopySupport::FireClipboardEvent()` keep handling
`eCopy` and `eCut` event even in password field, only if `TextEditor` allows
them.
Then, we need to make `nsPlainTextSerializer::AppendText()` not expose
masked password for making users safer. Although `TextEditor` does not allow
`eCopy` nor `eCut` when selection is not in unmasked range. Fortunately,
retrieving masked and unmasked password from `nsTextFragment` has already
been implemented in `ContentEventHandler.cpp`. This patch moves it into
`EditorUtils` and makes `ContentEventHandler.cpp` and `nsPlaintextSerializer`
share it.
Differential Revision: https://phabricator.services.mozilla.com/D39000
--HG--
extra : moz-landing-system : lando
It does not make sense to copy masked password with mask characters.
Therefore, we should allow users to copy/cut in password fields only when
selected range is in unmasked range.
Note that for web-compat, copy/cut are always enabled in HTML/XHTML document
in content. Therefore this patch changes the behavior only in chrome's
password fields.
Additionally, only the test uses `nsIEditor.canDelete()`. Therefore, this
removes it and make the test use `nsIDocShell.isCommandEnabled()` instead.
Unfortunately, `nsIEditor.canCopy()` and `nsIEditor.canCut()` are used by
BlueGriffon, therefore, we cannot get rid of them for now.
Differential Revision: https://phabricator.services.mozilla.com/D38999
--HG--
rename : editor/libeditor/tests/test_bug1067255.html => editor/libeditor/tests/test_cut_copy_delete_command_enabled.html
extra : moz-landing-system : lando
The code callng action.updateRequest, in FirefoxDataProvider, expects the updateRequest action
to be processed once the returned promise is resolved. Otherwise it may spawn
duplicated requestData requests.
Differential Revision: https://phabricator.services.mozilla.com/D39153
--HG--
extra : moz-landing-system : lando
Toolchain path for Windows version is `<NDK ROOT>/toolchains/llvm/prebuilt/windows-x86_64` etc, so it isn't '`winnt`.
So we has to replace `host.kernel.lower()` with `windows`.
Differential Revision: https://phabricator.services.mozilla.com/D39474
--HG--
extra : moz-landing-system : lando
When we use SVG layout, and transform-box is fill-box, the current
implementation makes the object not align with offset-path because there
may be an offset (SVG x and y values) between the anchor point and offset-path.
We have to tweak the anchor point to fix the alignment.
Differential Revision: https://phabricator.services.mozilla.com/D39419
--HG--
rename : testing/web-platform/tests/css/motion/offset-anchor-transform-box-fill-box.html => testing/web-platform/tests/css/motion/offset-anchor-transform-box-fill-box-001.html
rename : testing/web-platform/tests/css/motion/offset-anchor-transform-box-fill-box.html => testing/web-platform/tests/css/motion/offset-anchor-transform-box-fill-box-002.html
extra : moz-landing-system : lando
`layout.reflow.synthMouseMove` was added by Fennec/Maemo era (bug 657844) since this was low-end device. Since `layout.reflow.synthMouseMove` is false even if now, sampling rate of GeckoView's mouse event is still very sparse.
Since today is 2019, so we should change this to match sampling rate of mouse event with desktop.
Differential Revision: https://phabricator.services.mozilla.com/D39278
--HG--
extra : moz-landing-system : lando
IDBTransaction generates IDB_LOG base on the mAbortCode, but it fires event
according to the argument "aResult" of FireCompleteOrAbortEvents(). It's
possible to have a situation that IDBTransaction hasn't been aborted yet, but
the transaction on the parent process is aborted and propagate the reason to
IDBTransaction. In this case, IDBTransaction would still fire an abort event as
expected. However, the IDB_LOG would show that it fired a complete event.
Therefore, this patch corrects the behavior and adds an assertion to ensure it's
only possible that parent process tells child process to abort at the last
minate rather than the case of parent is canceling the abort.
Differential Revision: https://phabricator.services.mozilla.com/D36900
--HG--
extra : moz-landing-system : lando
This commit adds the text of the spec as inline comments and refactors
the code such that it directly corresponds to the spec's steps. This
makes it easier to understand how the spec's algorithm is implemented.
Differential Revision: https://phabricator.services.mozilla.com/D28976
--HG--
extra : moz-landing-system : lando
We retrieve the preference in the hudservice, where
WebConsoles and BrowserConsoles are created from.
If the pref is set to true, we assign a different
title to the Browser Console window so it will be
easier to spot.
The preference is then passes to the BrowserConsole,
WebConsole, WebConsoleUI and finally WebConsoleConnectionProxy
instances.
Later, we'll check this pref to connect to different
targets and listen to new ones.
Depends on D39646
Differential Revision: https://phabricator.services.mozilla.com/D39647
--HG--
extra : moz-landing-system : lando
1. Inline "Learn more" link with the main description text.
2. Remove margin-inline-start from labels
3. Remove textbox styling of the URL
Differential Revision: https://phabricator.services.mozilla.com/D39572
--HG--
extra : moz-landing-system : lando
Now ensure that we know whether malloc memory was freed by the collector or the mutator. Most of the time we can get the |wasSwept| state from from the FreeOp, so mostly this is passing the FreeOp through in more places. That's not always possible though since some code uses the runtime's default FreeOp, so this sets the approriate flag on the default free while we're sweeping too.
Differential Revision: https://phabricator.services.mozilla.com/D39393
--HG--
extra : moz-landing-system : lando
Now we're calculating retained size here we can remove the same calculation from Statistics. I had to add a flag to say whether a zone was collected in the last collection so that we know which zones to add to add up. I checked that this produced the same result as the original calculation.
Differential Revision: https://phabricator.services.mozilla.com/D39392
--HG--
extra : moz-landing-system : lando
The main idea here is to pass a |wasSwept| parameter to memory tracking methods that remove memory. We know what to pass for GC heap size because this always goes through HeapSize::removeGCArena(). Getting this right for malloc memory is the subject of a later patch.
Differential Revision: https://phabricator.services.mozilla.com/D39387
--HG--
extra : moz-landing-system : lando
`AddVarCache()` has a `bool aSkipAssignment` parameter. This patch removes that
parameter by splitting the function in two: `AddVarCache()` and
`AddVarCacheNoAssignment()`. (The former calls the latter.)
There are also tons of `Add*VarCache()` functions with `aSkipAssignment`
parameters that default to `false`. These defaults are never overridden, so
this patch removes the unnecessary arguments.
Differential Revision: https://phabricator.services.mozilla.com/D39433
--HG--
extra : moz-landing-system : lando