The way this test was written didn't guarantee a deterministic message
order, resulting in different output in Chromium and Firefox. This
change slightly rearranges the message order to make it deterministic.
This change is necessary as a prepartion for upcoming change that makes
MessagePort post messages from a separate thread, which would've
revealed the non-deterministic message order.
Reimplements c3121c9d at the transport layer, allowing us to solve the
same problem once, in a single place, for both the LibIPC connection and
MessagePort. This avoids exposing a workaround for a macOS specific Unix
domain socket issue to higher abstraction layers.
We don't need the [[Fields]] and [[PrivateMethods]] slots for most ESFO
instances, so let's reduce their impact on class size!
This shrinks ESFO from 200 bytes to 160 bytes, allowing more allocations
before we have to collect garbage.
We were doing way too much computation every time an ESFO was
instantiated. This was particularly sad, since the results of these
computations were identical every time!
This patch adds a new SharedFunctionInstanceData object that gets
shared between all instances of an ESFO instantiated from some kind of
AST FunctionNode.
~5% speedup on Speedometer 2.1 :^)
These generate what seems to be nonsense warnings on Function and
ByteBuffer; they *should* be investigated at some point, but they don't
provide anything useful at this point.
This is an editorial change in the ECMA-402 spec. See:
a2beb66
We implement this change by introducing a virtual interface that all
Intl "service" objects must implement. A future patch will make use of
the virtualized RelevantExtensionKeys and ResolutionOptionDescriptors
accessors, and we will need to be able to use those slots from a generic
instance type.
This is an editorial change in the ECMA-402 spec. See:
e3f7260
Note the other changes in this commit do not apply to our implementation
as we defer to ICU for the affected steps.
There's a bit of a UTF-8 assumption with this change. But nearly every
caller of these methods were immediately creating a String from the
resulting ByteString anyways.
Don't use a Vector to form the transformed string. We can construct the
string immediately and store the result in its buffer, and thus avoid a
double allocation.
In a synthetic benchmark, lowercasing a 500 character ASCII string
1 million times reduced from 550ms to 65ms on my machine.
* Use `any_of` instead of manual loops
* Don't check if a code point is upper/lowercase twice. The check we
are using is already present inside the case converter.
* Move StringImpl's implementation into ByteString. ByteString is its
only user, so let's avoid some jumping around. The other ASCII case
methods on StringImpl will soon also be removed.
With this change, the responsibility for prepending messages with their
size and ensuring the entire message is received before returning it to
the caller is moved to TransportSocket. This removes the need to
duplicate this logic in both LibIPC and MessagePort.
Another advantage of reducing message granularity at IPC::Transport
layer is that it will make it easier to support alternative transport
implementations (like Mach ports, which unlike Unix domain sockets are
not stream oriented).
We were not properly handling the case that prefix code point was the
empty string (which we represent as an OptionalNone). While this
still resulted in the correct pattern string being generated, an
incorrect regular expression was being generated causing matching
to fail.
This has no functional difference as far as I can tell, but for
clarity explicitly do not attempt to do this, which has the nice
side effect of not checking for whitespace known to not exist.
It turns out that the problem here was simply that we were trimming
trailing whitespace when we did not need to, which was meaning that
the port number of '80 ' was being converted to the empty string
per URLPattern elision as the port matches the http scheme.
We cached the length identifier for GetLength, but not
GetLengthWithThis. This caused an `has_value()` verification failure
when accessing super.length. Found by Fuzzilli.
This commit disallows "default" as a font-family name, when the name is
not quoted because unquoted names are treated as custom-idents, for
which the name "default" is not allowed.
