Expand Trunc in B3 to support Double to Float
https://bugs.webkit.org/show_bug.cgi?id=163809
Source/JavaScriptCore:
Reviewed by Geoffrey Garen.
This feature is useful for passing floating point arguments via registers.
Currently, ArgumentRegValue returns a the full 64-bit register. Thus, we
need a way to indicate to B3 that we only want the low 32-bits.
* b3/B3Common.h:
(JSC::B3::isIdentical):
* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::lower):
* b3/B3Opcode.h:
* b3/B3ReduceStrength.cpp:
* b3/B3Validate.cpp:
* b3/B3Value.cpp:
(JSC::B3::Value::typeFor):
* b3/testb3.cpp:
(JSC::B3::testAddFPRArgsFloat):
(JSC::B3::testCeilArgWithEffectfulDoubleConversion):
(JSC::B3::testFloorArgWithEffectfulDoubleConversion):
(JSC::B3::testDoubleProducerPhiToFloatConversionWithDoubleConsumer):
(JSC::B3::run):
Websites/webkit.org:
Update the docs to explain that Trunc now works with fleating point numbers.
Reviewed by Geoffrey Garen.
* docs/b3/intermediate-representation.html:
git-svn-id: http://svn.webkit.org/repository/webkit/trunk@207696 268f45cc-cd09-0410-ab3c-d52691b4dbfc
10 files changed
