Code may end up indirectly using a witness table for a Clang-imported type by inlining code that used the conformance from another module, in which case we need to ensure we have a local definition at hand in the inlining module so we can have something to link against independently. This needs to be fixed from both sides:
- During serialization, serialize not only witness tables from the current module, but from Clang-imported modules too
- During deserialization, when the SILLinker walks a loaded module, ensure that all shared conformances get deserialized, including those from ApplyInsts and inherited/associated type protocol requirements.
Fixes rdar://problem/38687726.
Code may end up indirectly using a witness table for a Clang-imported type by inlining code that used the conformance from another module, in which case we need to ensure we have a local definition at hand in the inlining module so we can have something to link against independently. This needs to be fixed from both sides:
- During serialization, serialize not only witness tables from the current module, but from Clang-imported modules too, so that their definitions can be used by other modules that inline code from the current module
- During IRGen, when we emit a reference to a SILWitnessTable or SILFunction declaration with shared linkage, attempt to deserialize the definition on demand
Fixes rdar://problem/38687726.
This is mostly intended to be used for testing at this point; in the
long run, we want to be using availability information to decide
whether to weak-link something or not. You'll notice a bunch of FIXMEs
in the test case that we may not need now, but will probably need to
handle in the future.
Groundwork for doing backward-deployment execution tests.
On many-core machines, the files that take the longest to compile are
build bottlenecks. With this change, we workaround clang scaling
problems compiling large llvm::StringSwitch expressions and bring the
"touch tools/swift/include/swift/AST/* ; time ninja swift" time down
from 1m43.24s down to 40.26s.
I've also benchmarked rebuilding Swift.o after this change. The before
and after seems to be deep in the noise.
This is going to be used for "always emit into client" functions,
such as default argument generators and stored property
initializers.
- In dead function elimination, these functions behave identically to
public functions, serving as "anchors" for the mark-and-sweep
analysis.
- There is no external variant of this linkage, because external
declarations can use HiddenExternal linkage -- the definition should
always be emitted by another translation unit in the same Swift
module.
- When deserialized, they receive shared linkage, because we want the
linker to coalesce multiple copies of the same deserialized
definition if it was deserialized from multiple translation units
in the same Swift module.
- When IRGen emits a definition with this linkage, it receives the
same LLVM-level linkage as a hidden definition, ensuring it does not
have a public entry point.
This has three principal advantages:
- It gives some additional type-safety when working
with known accessors.
- It makes it significantly easier to test whether a declaration
is an accessor and encourages the use of a common idiom.
- It saves a small amount of memory in both FuncDecl and its
serialized form.
For now these are underscored attributes, i.e. compiler internal attributes:
@_optimize(speed)
@_optimize(size)
@_optimize(none)
Those attributes override the command-line specified optimization mode for a specific function.
The @_optimize(none) attribute is equivalent to the already existing @_semantics("optimize.sil.never") attribute
This commit is mostly refactoring.
*) Introduce a new OptimizationMode enum and use that in SILOptions and IRGenOptions
*) Allow the optimization mode also be specified for specific SILFunctions. This is not used in this commit yet and thus still a NFC.
Also, fixes a minor bug: we didn’t run mandatory IRGen passes for functions with @_semantics("optimize.sil.never")
@_silgen_name and @_cdecl functions are assumed to be referenced from
C code. Public and internal functions marked as such must not be deleted
by the optimizer, and their C symbols must be public or hidden respectively.
rdar://33924873, SR-6209
This brings the capability from clang to save remarks in an external YAML files.
YAML files can be viewed with tools like the opt-viewer.
Saving the remarks is activated with the new option -save-optimization-record.
Similarly to -emit-tbd, I've only added support for single-compile mode for now.
In this case the default filename is determined by
getOutputFilenameFromPathArgOrAsTopLevel, i.e. unless explicitly specified
with -save-optimization-record-path, the file is placed in the directory of the
main output file as <modulename>.opt.yaml.
Support for @noescape SILFunctionTypes.
These are the underlying SIL changes necessary to implement the new
closure capture ABI.
Note: This includes a change to function name mangling that
primarily affects reabstraction thunks.
The new ABI will allow stack allocation of non-escaping closures as a
simple optimization.
The new ABI, and the stack allocation optimization, also require
closure context to be @guaranteed. That will be implemented as the
next step.
Many SIL passes pattern match partial_apply sequences. These all
needed to be fixed to handle the convert_function that SILGen now
emits. The conversion is now needed whenever a function declaration,
which has an escaping type, is passed into a @NoEscape argument.
In addition to supporting new SIL patterns, some optimizations like
inlining and SIL combine are now stronger which could perturb some
benchmark results.
These underlying SIL changes should be merged now to avoid conflicting
with other work. Minor benchmark discrepancies can be investigated as part of
the stack-allocation work.
* Add a noescape attribute to SILFunctionType.
And set this attribute correctly when lowering formal function types to SILFunctionTypes based on @escaping.
This will allow stack allocation of closures, and unblock a related ABI change.
* Flip the polarity on @noescape on SILFunctionType and clarify that
we don't default it.
* Emit withoutActuallyEscaping using a convert_function instruction.
It might be better to use a specialized instruction here, but I'll leave that up to Andy.
Andy: And I'll leave that to Arnold who is implementing SIL support for guaranteed ownership of thick function types.
* Fix SILGen and SIL Parsing.
* Fix the LoadableByAddress pass.
* Fix ClosureSpecializer.
* Fix performance inliner constant propagation.
* Fix the PartialApplyCombiner.
* Adjust SILFunctionType for thunks.
* Add mangling for @noescape/@escaping.
* Fix test cases for @noescape attribute, mangling, convert_function, etc.
* Fix exclusivity test cases.
* Fix AccessEnforcement.
* Fix SILCombine of convert_function -> apply.
* Fix ObjC bridging thunks.
* Various MandatoryInlining fixes.
* Fix SILCombine optimizeApplyOfConvertFunction.
* Fix more test cases after merging (again).
* Fix ClosureSpecializer. Hande convert_function cloning.
Be conservative when combining convert_function. Most of our code doesn't know
how to deal with function type mismatches yet.
* Fix MandatoryInlining.
Be conservative with function conversion. The inliner does not yet know how to
cast arguments or convert between throwing forms.
* Fix PartialApplyCombiner.
The `serialize` method can be called multiple times, but it will perform the actual serialization only the first time.
By means of this API we get the flexibility to serialize the SILModule not only after all the optimizations, but e.g. at any time during optimizations.
introduce a common superclass, SILNode.
This is in preparation for allowing instructions to have multiple
results. It is also a somewhat more elegant representation for
instructions that have zero results. Instructions that are known
to have exactly one result inherit from a class, SingleValueInstruction,
that subclasses both ValueBase and SILInstruction. Some care must be
taken when working with SILNode pointers and testing for equality;
please see the comment on SILNode for more information.
A number of SIL passes needed to be updated in order to handle this
new distinction between SIL values and SIL instructions.
Note that the SIL parser is now stricter about not trying to assign
a result value from an instruction (like 'return' or 'strong_retain')
that does not produce any.
Static initializers are now represented by a list of literal and aggregate instructions in a SILGlobalVariable.
For details see SIL.rst.
This representation is cleaner than what we did so far (point to the initializer function and do some pattern matching).
One implication of that change is that now (a subset of) instructions not necessarily have a parent function.
Regarding the generated code it's a NFC.
Also the swift module format didn't change because so far we don't serializer global variables.
This removes the function body, but preserves the SILFunction object, which may be still referenced by different kinds of meta-information e.g. debug info for inlined functions, generic specializations information, etc.
Doing this unconditionally simplifies the code and makes it less error-prone to reference SILFunctions from any kind of meta-information. It just works. No need to set any special flags, etc.
This marker can be used e.g. when a function is referenced from the generic specialization information (or any other kind of metadata) to indicate that this SIL function should not be completely eliminated by the dead function elimination pass even if it is not referenced by any function_ref instruction anymore. Such a function will be preserved in a zombie list instead if it needs to be eliminated.
- SILSerializeAll flag is now stored in the SILOptions and passed around as part of it
- Explicit SILSerializeAll/wholeModuleSerialized/makeModuleFragile API parameters are removed in many places
AccessMarkerElimination now registers a callback so that any subsequently
deserialized function bodies will have access markers stripped for optimization.
rdar:31908496 Assertion failed: (isa<X>(Val) && "cast<Ty>() argument of
incompatible type!") in SILPerformanceInliner
Previously we would drop all serialized SIL from partial swiftmodule
files generated while compiling source in non-WMO mode; all that was
missing was linking it in.
This adds a frontend flag, and a test; driver change is coming up
next.
Progress on <rdar://problem/18913977>.
This fixes a crash when referencing partially-applied methods
from @_inlineable functions.
Also, curry thunks for private methods do not need shared
linkage; private is sufficient.
Also, add a third [serializable] state for functions whose bodies we
*can* serialize, but only do so if they're referenced from another
serialized function.
This will be used for bodies synthesized for imported definitions,
such as init(rawValue:), etc, and various thunks, but for now this
change is NFC.
We didn’t do that if an optimization looked up a witness table.
Fixes rdar://problem/30544344
I couldn’t come up with an isolated test case, but this should be covered with our existing tests.
(The problem shows up when inlining of generics are enabled)
This reverts commit 1b3d29a163, reversing
changes made to b32424953e.
We're seeing a handful of issues from turning on inlining of generics,
so I'm reverting to unblock the bots.
This reverts commit a380855524, reversing
changes made to 66332140f3.
We're seeing a handful of issues from turning on inlining of generics,
so I'm reverting to unblock the bots.
In all cases the DeclCtx field was supposed to be initialized from the
SILLocation of the function, so we can save one pointer per
SILFunction.
There is one test case change where a different (more precise)
diagnostic is being generated after this change.
A witness table is dead if it is not used outside the module (private/internal) and it’s not used by any instruction or other witness table in the module.
Also the meta-type of the conforming type must not escape, because it’s possible to test any opaque type if it conforms to a protocol.
rdar://problem/23026019
