Using the unknown-sized Builtin.Word types complicates producing
compile-time overflow diagnostics. If we don't know the target Word
size, we don't know if there is an overflow. But SIL optimizer does not
know the size of Word, this is the point of having the Word type in the
first place.
Also, this opens up more possibilities for optimizations.
rdar://17604532
Swift SVN r24788
/Users/buildslave/jenkins/workspace/swift_tools-RA_stdlib-RD/megaclang/src/tools/swift/validation-test/stdlib/NumericDiagnostics.swift.gyb:29:33: error: expected diagnostic not produced
x1_UInt8 * x2_Float80 // expected-error{{ }}
Swift SVN r24238
Replace hard-coded lists of real-number types, and of arithmetic and logical operators I’m testing, with lists added to SwiftIntTypes.py, which becomes a bit of a misnomer.
Separate them into ones which only work on integer types, and those which also allow reals.
Do a double loop, over just the integers and then the reals as well.
Change the run line to allow importing StdlibUnittest with help from Dmitri, which I won't actually need until I do result type checking.
Approved by Dmitri.
Swift SVN r24029
