Commit Graph

20 Commits

Author SHA1 Message Date
Joe Groff
69a7ef6e2d Revert "SILGen: Tweak key path computed property lowering to produce a consistent runtime-callable ABI."
This reverts commit 4522cd09aa. This was a failed approach I shouldn't have committed. Fixes rdar://problem/32201589.
2017-05-15 16:10:55 -07:00
Roman Levenstein
60d6fb33b7 Use AnyObject instead of _Class when dumping the layout constraints 2017-05-04 15:54:50 -07:00
Slava Pestov
71f2a2617d Demangling: Update NodePrinter for primitive AnyObject 2017-05-01 17:33:02 -07:00
Joe Shajrawi
0d0cac357a retain_value_addr and release_value_addr SIL instructions: take as an input an address, load the value inside it and call retain_value and release_value respectively 2017-04-30 10:23:55 -07:00
Joe Groff
517c45aa3d Followups from merging master 2017-04-19 20:58:52 -07:00
Joe Groff
595e0e4ede Merge branch 'master' into keypaths 2017-04-19 18:38:24 -07:00
Joe Groff
4522cd09aa SILGen: Tweak key path computed property lowering to produce a consistent runtime-callable ABI. 2017-04-18 11:44:39 -07:00
Joe Groff
ee66ff35c5 SILGen: Lower computed property key path components. 2017-04-14 17:13:38 -07:00
Slava Pestov
0d0dc207c0 AST: Mangling for primitive AnyObject
A protocol composition with an explicit 'AnyObject' member is
now mangled as <protocol list> 'Xl'. For subclass existentials,
I changed the mangling from <protocol list> <class> 'XE' to
<protocol list> <class> 'Xl'.

Not used for anything just yet.
2017-04-13 21:17:07 -07:00
Huon Wilson
22c9e20a9d [IRGen] Completely remove witness table offsets.
These are unused.
2017-04-13 11:44:01 -07:00
Erik Eckstein
789646a15b Demangling: Make demangled names more readable and further reduce the size of the simplified demangled names
The goal here is to make the short demangling as short and readable as possible, also at the cost of omitting some information.
The assumption is that whenever the short demangling is displayed, there is a way for the user to also get the full demangled name if needed.

*) omit <where ...> because it does not give useful information anyway

Deserializer.deserialize<A where ...> () throws -> [A]
--> Deserializer.deserialize<A> () throws -> [A]

*) for multiple specialized functions only emit a single “specialized”

specialized specialized Constructible.create(A.Element) -> Constructible<A>
--> specialized Constructible.create(A.Element) -> Constructible<A>

*) Don’t print function argument types:

foo(Int, Double, named: Int)
--> foo(_:_:named:)

This is a trade-off, because it can lead to ambiguity if there are overloads with different types.

*) make contexts of closures, local functions, etc. more readable by using “<a> in <b>” syntax
This is also done for the full and not only for the simplified demangling.

Renderer.(renderInlines([Inline]) -> String).(closure #1)
--> closure #1 in Renderer.renderInlines

*) change spacing, so that it matches our coding style:

foo <A> (x : A)
--> foo<A>(x: A)
2017-04-13 08:43:28 -07:00
Slava Pestov
374db66e8b Merge pull request #8685 from slavapestov/subclass-existentials-silgen
Preliminary SILGen support for subclass existentials (SE-0156)
2017-04-12 11:44:23 -07:00
Slava Pestov
b5ae8c4de6 AST: Add mangling for class-constrained existentials 2017-04-12 00:07:29 -07:00
practicalswift
5e255e07d7 [gardening] Remove redundant logic 2017-04-11 23:04:55 +02:00
Erik Eckstein
cd3da4a218 Demangling: Print varargs correctly.
Instead of [ Swift.Int]..., print  Swift.Int...
2017-03-27 10:52:53 -07:00
Erik Eckstein
e427ded6c9 Mangling: refactoring: clean up tuple nodes
Replace VariadicTuple and NonVariadicTuple with a single Tuple node.
The variadic property is now part of the tuple element and not of the whole tuple.
2017-03-25 19:07:52 -07:00
Erik Eckstein
327017a65b Mangling: support mangling of varargs even if it's not the last argument.
Example:
func a(arr: Int..., n: String)

fixes SR-1076
2017-03-25 19:07:52 -07:00
Slava Pestov
e912f27944 AST/SILGen: New mangling for vtable thunks
Simply mangling the derived method is no longer sufficient. Now also
mangle the base method, so that eventually we handle this sort of
scenario:

class Base {
  // introduces: Base.method
  func method(_: Int, _: Int) {}
}

class First : Base {
  // overrides: Base.method
  // introduces: First.method
  override func method(_: Int?, _: Int) {}
}

class Second : First {
  // overrides: Base.method, First.method
  // introduces: Second.method
  override func method(_: Int?, _: Int?) {}
}

Here, the override of Base.method by Second.method and the
override of First.method by Second.method require distinct
manglings even though the derived method (Second.method) is
the same in both cases.

Note that while the new mangling is longer, vtable thunks are
always emitted with private linkage, so with the exception of
the standard library which is built with -sil-serialize-all
they will not affect the size of dylibs.

The standard library itself has very few classes so it doesn't
matter there either.

This patch doesn't actually add any support to introduce new
vtable entries for methods that override; this is coming up
next.
2017-03-23 23:40:52 -07:00
Roman Levenstein
e7503b1502 [mangling] Define a mangling for _Class and _NativeClass layout constraints
I forgot to define the mangling when I introduced these two new layout constraints.
2017-03-22 23:36:40 -07:00
Erik Eckstein
5e80555c9b demangler: put the demangler into a separate library
Previously it was part of swiftBasic.

The demangler library does not depend on llvm (except some header-only utilities like StringRef). Putting it into its own library makes sure that no llvm stuff will be linked into clients which use the demangler library.

This change also contains other refactoring, like moving demangler code into different files. This makes it easier to remove the old demangler from the runtime library when we switch to the new symbol mangling.

Also in this commit: remove some unused API functions from the demangler Context.

fixes rdar://problem/30503344
2017-03-09 13:42:43 -08:00