* Moved `random` methods out of FloatingPoint.swift
* Added `random` methods to FloatingPointRandom.swift
* Added FloatingPointRandom.swift to CMakeLists.txt
* Added FloatingPointRandom.swift to GroupInfo.json
* Moved filename within CMakeLists.txt
We can't actually check for NaN (because the notion doesn't exist for Comparable), but we can require that the endpoint is non-exceptional by checking if it equals itself.
The "Result" type is expected to have no overhead when used.
This requires the member functions to be inlinable.
Which is probably okay for library evolution, because it's unlikely that those methods change in an incompatible way.
To be able to constant fold string interpolation, the right semantic attributes must be in place.
Also, the interpolation's write function must be inlinable.
For the _typeName constant folding, a semantic attribute is required.
Instead of doing the sanity checks by default (with an assert-build of the stdlib), only do the checks if the environment variable SWIFT_DEBUG_ENABLE_COW_SANITY_CHECKS is set to true.
The checks can give false alarms in case a binary is built against a no-assert stdlib but run with an assert-stdlib.
Therefore only do the checks if it's explicitly enabled at runtime.
rdar://problem/65475776
Inlining those operations does not really give any benefit.
Those operations are complex and most likely, cannot be folded into simpler patterns after inlining, anyway.
Reduces some code size for cases where those functions were inlined (due to a not perfect inlining heuristics).
LLVM doesn't have a stable ABI for Float16 on x86 yet; we're working with Intel to get that fixed, but we don't want to make the type available on macOS until a stable ABI is actually available, because we'd break binaries compiled before any calling convention changes if we do.
This function walks all the fields of a struct, class, or tuple, and calls
`body` with the name, offset, and type of each field. `body` can perform
any required work or validation, returning `true` to continue walking fields
or `false` to stop immediately.
We ultimately want to explicitly change standard library uses of #file to #fileID, but once we do, previous compilers won’t be able to build the standard library. So instead, we will temporarily build the standard library with -enable-experimental-concise-pound-file, which should have the same effect, but will back-deploy to compilers going back several months.
Clean up a few general patterns that are now obviated by canImport
This aligns more generally with the cleanup that the Swift Package
Manager has already done in their automated XCTest-plumbing tool in
apple/swift-package-manager#1826.
Previously these always went through the FloatingPoint-provided default implementation, which is not particularly efficient. Also try removing inlinable from the generic _convert hooks, since we probably never want to actually inline them.
