Since Uint32 can't be represented in a MIRType_Int32, this function should
return a MIRType_Double.
Allow MSimdExtractElement(Uint32x4) to return a MIRType_Int32 too. It will work
like the double version followed by MTruncateToInt32 which bitcasts the Uint32
value range into the Int32 value range.
Add a MSimdBinaryComp::AddLegalized function which expands unsigned compares on
target platforms that don't support them directly. The early expansion exposes
the constants to MIR optimizations.
Unsigned comparison is expressed in terms of signed comparison by offsetting
both sides by INT_MIN.
The conversion from Uint32x4 to Float32x4 is not available as an SSE
instruction, so we need to expand into a larger instruction sequence lifted
from LLVM. Make this expansion early when generating MIR so that it can be
exposed to LICM and GVN optimizations.
The conversion from Float32x4 to Uint32x4 can throw a RangeError. It is handled
similarly to LFloat32x4ToInt32x4. This expansion depends on the details of the
cvttps2dq instruction that can't be expressed in MIR, so it can't be expanded
early.
Add a new InlinableNative::SimdUint32x4 enumerator, and emit the corresponding
JSJitInfo objects in SIMD.cpp.
Start producing template objects for Uint32x4 operations in BaselineIC.cpp.
Add a new SimdSign enum class to SIMD.h which will be used to distinguish
between signed and unsigned integers in the few places where it matters.
Map the SIMD.Uint32x4 type to the existing MIRType_Int32x4 + SimdSign::Unsigned.
Map SIMD.Int32x4 to MITType_Int32x4 + SimdSign::Signed.
Add a 'SimdSign sign' argument to those inlineSimd...() functions that care.
Some MIR instructions will get similar fields in the following commits.
For now, abort inlining if unsigned vectors are actually encountered. These cases
will be fixed in the following commits.
Extract the code that generates template objects for SIMD operations, and
rewrite it to use the JSJitInfo nativeOp encoding.
This avoids the native function pointer comparisons, and it makes it simpler to
add new SIMD types and operations.
The extractLane(), anyTrue(), and allTrue() SIMD functions produce scalar
values, and so they don't need a template object. The canInlineSimd() function
was rejecting these functions because of the missing template object.
At the same time, explicitly avoid inlining any SIMD operations if the JIT does
not support SIMD. This was previously controlled by the absense of the template
object.
This saves some code size in a cold function, and it makes it possible to pass
in the SIMD type as a dynamic argument.
Also detemplatize the static CreateSimdType() to save some code size.
Replace all of the Get*TypeRepr() self-hosting functions with a single
GetSimdTypeDescr() which takes one of the JS_SIMDTYPEREPR_* constants as an
argument instead.
Total code shrink ~ 32 KB.
Remove obsolete SIZE_CHANGED event and its handler in nsWindow. Also
remove some other supporting code (such as gAndroidBounds and the
FORCED_RESIZE event) that should be unnecessary by now.
Convert the SIZE_CHANGED event to a native method in GLController, and
carry over the SIZE_CHANGED implementation to the new implementation in
GLController. Some other changes were made for correctness in handling
size changes.
Right now, we rely on browser.js sending Gecko:Ready, to set the
GeckoThread state to RUNNING upon receiving Gecko:Ready. This patch
gets rid of this dependency on browser.js and Gecko:Ready.
Convert the old prefs code in browser.js to use observer notifications
that are sent by the new PrefsHelper implementation, in order to handle
pseudo-prefs.
Implement the PrefsHelper native methods. The previous browser.js
implementation supported "pseudo-prefs" that did not exist as actual
prefs, but was accessible through PrefsHelper. In order to accommodate
these pseudo-prefs, we send observer notifications in order to
communicate with browser.js about prefs that we don't support.
Introduce a new implementation for PrefsHelper that does not use events
or rely on browser.js for getting and/or setting prefs. Also add an
addObserver method to better match the removeObserver method.
* add Jetpack definition
* e10s mochitests were not using --e10s flag [1]
* disable mochitest e10s since they're not green anymore
* increase Marionette max runtime
[1]
Unfortunately, we had two payload entries defined for e10s mochitests.
That would cause the first paylod (with --e10s) to be ignored.
This patch makes use of the security reporter component (which hasn't landed yet
- see blockers) to allow automatic TLS error reports to be sent directly from
nsHttpChannel.cpp rather than sending them from browser.js. This allows failed
connections for subresources to be reported.
Some of the report sending from browser.js was retained to allow reports to be
sent at the time a user enables reporting. This has been modified to also make
use of the component.
Since the patient is on the table, I've also taken the opportunity to
remove the retry and status bits from aboutCertError.xhtml and
aboutNetError.xhtml - which removes a bunch of code and simplifies things a bit.
The mochitests have been modified to cope with the fact that the UI does not
update with report sending / report failures - instead, success and failure are
determined by examining the response seen from the server from within the test.
