They are non-standard aliases for "DragEvent" and "KeyboardEvent" that
are not supported by any other UA, and we would like to drop support.
So first let's stop using them ourselves, so we can use telemetry to see
if any sites are using them.
MozReview-Commit-ID: ICC33ORa2st
The reason this works is because the window object stays the same between location
changes, but the document object changes.
Since stylesheets are applied to documents, not windows, we need to make sure to
inject a stylesheet for each document, otherwise the find toolbar highlights will
look unstyled and wonky.
In bug 1279695 I made a change - to make tests pass and patch landable - to keep
track of all the things (state, mainly range sets) mapped to window objects, because
I hadn't thought of the fact that the content scripts and JSMs that are loaded in
its scope will see many different window objects. As many as there are tabs open,
actually (not counting frames).
The mistake I made there was to keep track of the state flag 'did I already install
the stylesheet?' on a per-window basis, unlike I did before. This patch corrects
this back to the way it was and should be, really.
MozReview-Commit-ID: 2t0mer2s10x
--HG--
extra : rebase_source : 6e793b695a4a86874898ca470639b60d5a91a001
Throw some process selectors in the Histograms and Keyed Histograms sections
to allow users to choose which process type's histograms they'd like to see.
Rewrite the categorical histogram accumulation code to use the common path.
This way it gets remote accumulation for cheap.
MozReview-Commit-ID: 3q6gdSvBix
I saw a one-off crash on try in internal_GetHistogramByEnumId. Not reproducible
but maybe possible if we're trying to accumulate using an invalid ID. So let's
guard against that.
MozReview-Commit-ID: Ei6eTlV91mJ
On content process shutdown we send a content process ping to ensure we have
up-to-date data from the content process before it goes away. Now we need to
also flush the batched telemetry accumulations to the parent so that it can be
present in the ping.
No attempt is made to synchronize access to IPCTimerFired. It is safe to
re-enter.
No attempt is made to cancel the timer as its firing is benign.
MozReview-Commit-ID: 1gjNH9IPhKf
Clear isn't generally called at all, and isn't dispatched to the parent process
for child telemetry aggregation. Clear should only be called on the parent
process.
MozReview-Commit-ID: stIutvAO6h
The JS histograms, too, need to dispatch their accumulations from child to
parent.
JSHistograms_Add now only supports histograms that are in gHistogramsMap or
that were created in the parent process. After bug 1288745, maybe we'll be able
to change this to be less convoluted.
MozReview-Commit-ID: 3qTH89YKbGP
Take the opportunity presented through changing child telemetry accumulation
to bring the ping form closer to the ideas expressed in bug 1281795.
childPayloads still exists, but without histograms or keyedHistograms which are
now at root.processes.content.{keyedH|h}istograms. This will require coordinated
changes in the aggregator and moztelemetry libraries.
MozReview-Commit-ID: AqG2jmBBC2W
To simplify using child telemetry from the parent process, only allow child
telemetry payloads to be generated once per child process, on shut down.
This will allow us to use the child telemetry's subsession information to leave
childPayloads the way it currently is.
Will need to update test_ChildHistograms.js as it is the only consumer.
MozReview-Commit-ID: 2qSztg0QHV5
I originally thought we'd be able to avoid the previous implementation's waste
of a map full of every kind of keyed histogram. Unfortunately, other code
(TelemetrySession at the very least) depends on this (and will throw if a keyed
histogram isn't present, even if it is empty)
MozReview-Commit-ID: 8MCGVa595UB
The original commit didn't properly support subsession histograms, so rectify
that lapse by adding support for stripping out the base name of a histogram
when trying to determine its id.
MozReview-Commit-ID: LvUek6f5WUx
Batch the accumulations to only transmit every so often, so we don't incur
too much in the way of IPC overhead penalties.
What this doesn't do:
* remove or restructure child telemetry code to adapt to the new way
* send the telemetry anywhere
* allow for the child process to clear child histograms
* support anything but histograms (but this is expected and okay)
MozReview-Commit-ID: JnUkcmN3Ya7
Throw some process selectors in the Histograms and Keyed Histograms sections
to allow users to choose which process type's histograms they'd like to see.
nsTimer fires on the thread that created the timer. An nsTimer instance should
only be manipulated on its target thread (it isn't threadsafe). IPC using
PContent must be on the main thread.
Thus, everything to do with the gIPCTimer must be on the main thread.
This also takes care of bug 1299312.
MozReview-Commit-ID: IcVRYsoX2R9
Throw some process selectors in the Histograms and Keyed Histograms sections
to allow users to choose which process type's histograms they'd like to see.
Rewrite the categorical histogram accumulation code to use the common path.
This way it gets remote accumulation for cheap.
MozReview-Commit-ID: 3q6gdSvBix
I saw a one-off crash on try in internal_GetHistogramByEnumId. Not reproducible
but maybe possible if we're trying to accumulate using an invalid ID. So let's
guard against that.
MozReview-Commit-ID: Ei6eTlV91mJ
On content process shutdown we send a content process ping to ensure we have
up-to-date data from the content process before it goes away. Now we need to
also flush the batched telemetry accumulations to the parent so that it can be
present in the ping.
No attempt is made to synchronize access to IPCTimerFired. It is safe to
re-enter.
No attempt is made to cancel the timer as its firing is benign.
MozReview-Commit-ID: 1gjNH9IPhKf
Clear isn't generally called at all, and isn't dispatched to the parent process
for child telemetry aggregation. Clear should only be called on the parent
process.
MozReview-Commit-ID: stIutvAO6h
The JS histograms, too, need to dispatch their accumulations from child to
parent.
JSHistograms_Add now only supports histograms that are in gHistogramsMap or
that were created in the parent process. After bug 1288745, maybe we'll be able
to change this to be less convoluted.
MozReview-Commit-ID: 3qTH89YKbGP
Take the opportunity presented through changing child telemetry accumulation
to bring the ping form closer to the ideas expressed in bug 1281795.
childPayloads still exists, but without histograms or keyedHistograms which are
now at root.processes.content.{keyedH|h}istograms. This will require coordinated
changes in the aggregator and moztelemetry libraries.
MozReview-Commit-ID: AqG2jmBBC2W
To simplify using child telemetry from the parent process, only allow child
telemetry payloads to be generated once per child process, on shut down.
This will allow us to use the child telemetry's subsession information to leave
childPayloads the way it currently is.
Will need to update test_ChildHistograms.js as it is the only consumer.
MozReview-Commit-ID: 2qSztg0QHV5
I originally thought we'd be able to avoid the previous implementation's waste
of a map full of every kind of keyed histogram. Unfortunately, other code
(TelemetrySession at the very least) depends on this (and will throw if a keyed
histogram isn't present, even if it is empty)
MozReview-Commit-ID: 8MCGVa595UB
The original commit didn't properly support subsession histograms, so rectify
that lapse by adding support for stripping out the base name of a histogram
when trying to determine its id.
MozReview-Commit-ID: LvUek6f5WUx
Batch the accumulations to only transmit every so often, so we don't incur
too much in the way of IPC overhead penalties.
What this doesn't do:
* remove or restructure child telemetry code to adapt to the new way
* send the telemetry anywhere
* allow for the child process to clear child histograms
* support anything but histograms (but this is expected and okay)
MozReview-Commit-ID: JnUkcmN3Ya7
Services.logins.removeAllLogins is already done in registerCleanupFunction
MozReview-Commit-ID: 8FGF3kJmvj0
--HG--
extra : rebase_source : fa3dc1d635620f0285009685ff7d0a281a8ba49b
More splitting of test_prompt.html to make these tests easier to maintain/digest.
MozReview-Commit-ID: DuakmtkJ41I
--HG--
rename : toolkit/components/passwordmgr/test/mochitest/test_prompt_promptAuth.html => toolkit/components/passwordmgr/test/mochitest/test_prompt_promptAuth_proxy.html
extra : rebase_source : b0056d757df10befd6d662bdff29e52343835fa9
This switches getPrompt to use a ChromeWindow which matches what the shipping code does for e10s.
MozReview-Commit-ID: DcVSr6JfmdK
--HG--
extra : rebase_source : 35178483a1223c71c0dd2142479799225ccde1bb
This introduces a new NLP (Natural Language Processing) module with only one
method: 'levenstein'. We're using it to allow the highlighter to keep running
when the it starts the iterator with a word that's one edit distance behind the
value in the findField.
MozReview-Commit-ID: K8oeiXoiLUe
This fix is not related to the referenced bug but came up during review.
MozReview-Commit-ID: IjrxWzkLIq1
--HG--
extra : rebase_source : d53179af98106049bcf1a12efed74c4527ac62c1
And change `this.global.Object.create(null)` to
`Cu.createObjectIn(this.global)`. The tests pass either way, but
`Cu.createObjectIn` is more explicit.
MozReview-Commit-ID: LmL6rTru5zZ
--HG--
extra : rebase_source : 4cf7b1463bf8b6882b6fd453657eae0ff43ed64d
Split the `shouldInject` method into separate methods:
- `shouldInject` to determine whether the API (or namespace)
should be injected.
- `getImplementation` to return the actual implementation.
Introduced `SchemaAPIInterface` for documentation purposes, and
two concrete implementations `LocalAPIImplementation` and
`ProxyAPIImplementation` which provide the functionality to run a local
and remote implementation of the API for which the schema API is
generated, respectively. These classes store the necessary details for
the invocation, so the methods that were formerly in the `Context` in
Schemas.jsm no longer get the `pathObj`, `path` or `name` parameters.
And merge the `path` and `name` in the implementation of remote APIs
because there is no need for having them separate, as the callers and
callees often did redundant pre/post-processing on `data.path` because
of the way it was implemented.
MozReview-Commit-ID: isbG9i9pNP
--HG--
extra : rebase_source : 22cdc3ab3d14c6381f9f540739d6750281ae8c71
The API implementation is already available upfront when the schema API
is generated, so `pathObj` has the implementation and can be used
instead of looking up the implementation over and over again with
`findPathInObject`.
MozReview-Commit-ID: FnCIyoaxgA4
--HG--
extra : rebase_source : 440b25fcfb4a0438b1ff8680ad770930e7427de7
- This was the last non-schema-generated API in content scripts.
MozReview-Commit-ID: FaIOCHoircf
--HG--
extra : rebase_source : 7bab2249a7462a581e493f7aa937df45cb895107
- Use schema-generated runtime API for content scripts instead of
untyped API.
- Move logic that cannot be run in the main process to a new file.
Together with the previous patch that migrated the i18n API, this
concludes the fix for bug 1286712.
MozReview-Commit-ID: A3yG0x1kjwx
--HG--
extra : rebase_source : 19efe95149c423c0f9284bb70e289a282fb758c1
Use schema-generated i18n API for both content script and
addon contexts, instead of just the addon context.
MozReview-Commit-ID: AKDAFxNv6Zs
--HG--
extra : rebase_source : ea66e8bab7b4713d971614ce27e8c1d9350a0a14
- By default, schema APIs are not injected in content scripts unless
the JSON schema sets the "restrictions" attribute to `["content"]`.
- Added the "restrictions" attribute to the storage and test schemas.
Other APIs will follow in subsequent commits and make use of the
primitives introduced in this commit.
MozReview-Commit-ID: 1rNjQap0BiM
--HG--
extra : rebase_source : bf2c7fb70a38fbf13cdfd4337cb3cb6fef7e67e8
Replace test_disabled_addon_can_be_enabled_after_reload in test_reload.js with test_reload_to_invalid_version_fails.
MozReview-Commit-ID: 9OEBnbwNplC
--HG--
extra : transplant_source : %F8T%13%BAZ%9C%9D%A2%23%E7n%FC%2CL%81%90%05g%0D%C3
The patch also changes RemoteOpenFileChild::OpenNSPRFileDesc() so that it
cannot succeed with a null fd, so that checking just the return value is
sufficient.
--HG--
extra : rebase_source : cc40bbcf2a9991edc9d3da3fb624d27db50b4996
- Moved Management logic to ExtensionUtils (as SchemaAPIManager) so that
the logic can be used by addon and content processes.
- Remove the `context.extension.hasPermission(api.permission)` check in
`generateAPIs` because the only user (`registeredPrivilegedAPI`) was
removed before in bug 1295082.
- Add new category "webextension-scripts-content", intended for
registering the few scripts that must be loaded in a content process.
MozReview-Commit-ID: 81nhblV8YE6
--HG--
extra : rebase_source : d2eee0298ad5038225ea56c2e752532af0bc6fb7
This only supports flac with a STREAMINFO header.
MozReview-Commit-ID: FaT9N6xJDPY
--HG--
extra : rebase_source : ba4e1f78f3a3ca9b4062edfc84445d444c323d80
This patch also makes sure that we correctly grab a weak reference to the
related window instead of just setting a "relatedBrowser" property directly on
the JS object (which shadows the XBL property getter).
MozReview-Commit-ID: 97VQyCoY1Cj
- Add `envType` to BaseContext.
- Pass an explicit envType to all `registerSchemaAPI` invocations.
- The factories passed to `registerSchemaAPI` will be split up later, so
that content scripts (`content_child`) and addon pages can share
common implementations.
- The factories that implement the addon API will also be split up,
to separate code running in the main process (`addon_parent`) from
code running in a child process (`addon_child`).
- Remove the use of a hardcoded list of `namespaces` from ProxyContext.
Now `envType` is used to specify whether an API should be activated.
MozReview-Commit-ID: Jiff8HIwG92
--HG--
extra : rebase_source : 946a3c0009a4e3223c2d10044b3099a94c845394
Local wrappers currently look up the API object over and over again
whenever a schema API is invoked. This can be optimized by re-using
the lookup result from a `shouldInject` invocation, which is passed
as the `pathObj` parameter to the wrapper methods.
This commit adds the necessary changes and tests to allow this to
happen, but does not modify the wrapper in Extension.jsm yet.
Also, this construction allows the `ChildAPIManager` to use a local
implementation if available and fall back to a remote implementation
otherwise.
MozReview-Commit-ID: C9gm7A9Zppb
--HG--
extra : rebase_source : e2ab1e40de72fdea23acc9a68f3cf7d77e3da66f
Currently there is a tight coupling between registered APIs because they
share the same global scope, and the dependencies between the modules
that use these globals are not explicit. Consequently, it would be
possible for APIs to break when the registered APIs run in separate
processes, because then there are separate global scopes.
To mitigate this issue, this patch isolates the global namespaces of
API registrations in different environments, starting with the "chrome"
process. Content and addon processes will follow later.
A new JSM is introduced to avoid hidden dependencies between ext-*.js
and the script loader. ExtensionUtils.jsm would be a natural choice for
this shared utility method, but cannot be used because its local
`EventEmitter` implementation conflicts with the `EventEmitter` import
in ext-tabs.js.
So, this patch provides isolation of global variables declared through
`globals.XXX = ...`, but does not provide isolation for `Cu.import`-ed
logic. Ideally `Cu.import` should always use its second argument to
prevent inadvertent namespace pollution.
MozReview-Commit-ID: 1DTZaKOaeSE
--HG--
extra : rebase_source : 1376a1325fd9bf186e09e1dbe83467f3ad94516f
The presence of these globals interfere with the attempt to get ext-*.js
scripts to load in a content process because these globals are only
available in the main process.
MozReview-Commit-ID: 7syjAGcuUnu
--HG--
extra : rebase_source : d614424d50bbb2a5dea6c23e4dae1586b9efe4fb
The main motive for this patch is to remove the use of the GlobalManager
global (which was used to see if an extension ID is valid, which was
specifically added in order to create thebrowser_ext_lastError.js test).
To preserve test coverage I implemented a full validation of the
runtime.sendMessage method.
Now the error for a non-existent extension is identical in both the
content script and background pages. Note that this also fixes a
minor privacy leak: Previously extensions could see whether another
extension is installed by sending a message to the specified extension
and using the different responses to see whether another extension is
installed.
MozReview-Commit-ID: 82R97Ei25Xr
--HG--
extra : rebase_source : 900a65e695afd6db83d5102f515dc29d46d001f1
Robert Strong
William Lachance
Kris Maglione
Aryeh Gregor
Wes Kocher
Bobby Holley
Sebastian Hengst
Ryan VanderMeulen
Iris Hsiao
Mike de Boer
Makoto Kato
Andrzej Hunt
Tom Schuster
Michael Layzell
Chris H-C
Robert Helmer
Eitan Isaacson
Sebastian Kaspari
Gijs Kruitbosch
André Bargull
Sunny
Fred Lin
Thomas Nguyen
Tim Huang
Thomas Wisniewski
Johann Hofmann
Marco Bonardo
Jared Wein
Rutuja Surve
Rob Wu
Alastor Wu
Mark Hammond
Shubheksha Jalan
Fischer.json
Julian Descottes
Matthew Noorenberghe
Justin Dolske
Yoshi Huang
Dão Gottwald
Kearwood (Kip) Gilbert
Bob Silverberg
Shane Caraveo
David Anderson
Paolo Amadini
Jim Mathies
Andrea Marchesini
Christoph Kerschbaumer
Nicholas Nethercote
Mike Hommey
Jean-Yves Avenard
Blake Kaplan
Andrew Swan
Yura Zenevich
dimi