- Introduced the io.activity.enabled pref, so IOActivityMonitor can run without a timer
- Added IOActivityMonitor::NotifyActivities() to trigger notifications manually
- Added ChromeUtils.requestIOActivity() so we can trigger it via JS
MozReview-Commit-ID: 9JA2rLaM496
- Introduced the io.activity.enabled pref, so IOActivityMonitor can run without a timer
- Added IOActivityMonitor::NotifyActivities() to trigger notifications manually
- Added ChromeUtils.requestIOActivity() so we can trigger it via JS
MozReview-Commit-ID: 9JA2rLaM496
This was done automatically replacing:
s/mozilla::Move/std::move/
s/ Move(/ std::move(/
s/(Move(/(std::move(/
Removing the 'using mozilla::Move;' lines.
And then with a few manual fixups, see the bug for the split series..
MozReview-Commit-ID: Jxze3adipUh
PerformanceCounters are currently disabled in two ways:
- a preference that's off by default "dom.performance.enable_scheduler_timing"
- calls made only for nightly using #ifndef RELEASE_OR_BETA
In order to simplify the code, let's remove the #ifndef and rely only on the pref.
That will also allows us to use the feature in every version going forward.
The performance will not be impacted since the current code is already using
the (cached) pref value to determine if the counters are used.
MozReview-Commit-ID: 47t2M1O13aH
Chromeutils.RequestPerformanceMetrics() is now composed of two parts:
- calls content processes via IPDL to get their counters
- directly dispatch counters from the parent process
MozReview-Commit-ID: HlgcEOzkyAq
There's no standard way to create a JS error with full stack and location
information from a saved stack. Since we need this functionality in order to
reject promises with useful Error objects, this patch adds a simple helper to
make that possible.
MozReview-Commit-ID: FyGuo4UjfsQ
Most WebExtension APIs are async, and have fairly complicated error reporting
semantics. As a result, when we report an error, the current JS stack has very
little to do with the JS caller that triggered the error, which makes it
difficult to diagnose.
In order to improve the situation, we need to store the location of the caller
at the start of an async operation, so we can tie the error to some marginally
useful location. We don't have a reasonable way to do that now other than
proactively creating an error object when the API is called, or creating a
promise with a full async stack, both of which are too expensive.
This helper instead returns a single SavedStack frame with a given principal,
which should be considerably cheaper, and likely good enough to give a
starting point for debugging cryptic errors.
MozReview-Commit-ID: BTxhpZK9Fdz
This helper makes it easy to lazily import a JavaScript module the first time
one of its exports is required. It is intended to replace
XPCOMUtils.defineLazyModuleGetter, which has similar functionality but is much
less efficient.
MozReview-Commit-ID: 2zxXYwrn3Dr
Looking up and copying exported properties each time a module is loaded is
fairly expensive at the best of times. It's even more expensive when we only
want to export a subset of symbols, which generally requires creating a
temporary object to hold the exports, or fetching them directly from the
returned global.
Aside from making the general case a bit faster, storing exports on an object
allows us to optimize lazy module imports by fetching imported symbols
directly from the exports object with very little additional overhead.
MozReview-Commit-ID: C9PGoXPNmsh
This is similar to Services.tm.idleDispatchToMainThread, but provides an
IdleDeadline argument to its callbacks, the same way that
Window.requestIdleCallback does.
The IdleDeadline argument was necessary for my first attempt at this bug. It's
not necessary for the current version, but I suspect it will be useful in
other areas, and it also avoids some XPConnect overhead, so it's probably
worth keeping.
MozReview-Commit-ID: FtrbNkE7Vz5
This will allow us to verify the entire detection pipeline in real nightly
builds, which will give us confidence that real heap corruption will be
detected and reported properly.
MozReview-Commit-ID: 43Fp2HT8RYy
As part of the normalization process for WebExtension API calls, we need to
extract and validate the full set of value properties (including properties
X-rays would normally deny access to) from cross-compartment objects. This
currently involves waiving X-rays, enumerating property descriptors, and
unwaiving X-rays - all through X-ray wrappers and waivers - and generating a
lot of expensive and short-lived wrappers in-between.
This helper reads out the list of safe properties from within the object's
compartment, and then copies them over to an object in the target compartment,
without any X-ray overhead, or any unnecessary intermediate wrappers or
compartment switches. It cuts about 40% off the overhead of our normalization
code.
MozReview-Commit-ID: H582oAYocaX
In the code that I'm profiling, the XPC WrappedNative overhead of calling
these functions adds up to about a quarter of the time spent executing the
code. The overhead of the WebIDL versions is negligible.
MozReview-Commit-ID: 30qJy5RtP9d
This change renames OriginAttributes.mInBrowser to mInIsolatedMozBrowser and
nsIPrincipal::GetIsInBrowserElement to GetIsInIsolatedMozBrowserElement. Other
methods that pass these values around also have name changes.
Tokens such as "inBrowser" have previously been serialized into cache keys, used
as DB column names, stored in app registries, etc. No changes are made to any
serialization formats. Only runtime method and variable names are updated.
No behavior changes are made in this patch, so some renamed methods may have
nonsensical implementations. These are corrected in subsequent patches
focused on behavior.
MozReview-Commit-ID: 66HfMlsXFLs
