For example:
%0 = string_literal "abc"
%1 = integer_literal 0x8000000000000000
%2 = builtin "stringObjectOr_Int64" (%0, %1)
%3 = integer_literal 0x4000000000000000
%4 = builtin "and_Int64" (%2, %3)
In this case we know that %4 is 0.
Certain patterns of directly applied partial_apply's were not being
inlined. This can happen when a closure is defined in the
implementation of a generic witness method. I noticed this issue while
debugging SR-6254: "Fix (or explain) strange ways to make code >3
times faster/slower".
After the witness method is devirtualization and specialized, the
closure application looks like this:
%pa = partial_apply [callee_guaranteed] %fn() : $@convention(thin) (@in Int) -> @out Int
%cvt = convert_escape_to_noescape %pa
apply %cvt(...)
SILCombine already removes the partial_apply and convert_escape_to_noescape generating:
%thick = thin_to_thick %fn
apply %thick(...)
However, surprisingly, neither the inliner nor SILCombine can handle this.
I cleaned up the code in SILCombine's apply visitor that handles
thin_to_thick and generalized it to handle this and other patterns.
I also added a restriction to this code so that it doesn't break SIL
ownership in the future, as I discussed with Arnold.
Certain patterns of directly applied partial_apply's were not being
inlined. This can happen when a closure is defined in the
implementation of a generic witness method. I noticed this issue while
debugging SR-6254: "Fix (or explain) strange ways to make code >3
times faster/slower".
After the witness method is devirtualization and specialized, the
closure application looks like this:
%pa = partial_apply [callee_guaranteed] %fn() : $@convention(thin) (@in Int) -> @out Int
%cvt = convert_escape_to_noescape %pa
apply %cvt(...)
SILCombine already removes the partial_apply and convert_escape_to_noescape generating:
%thick = thin_to_thick %fn
apply %thick(...)
However, surprisingly, neither the inliner nor SILCombine can handle this.
I cleaned up the code in SILCombine's apply visitor that handles
thin_to_thick and generalized it to handle this and other patterns.
I also added a restriction to this code so that it doesn't break SIL
ownership in the future, as I discussed with Arnold.
Local.cpp was ~3k lines of which 1.5k (i.e. 1/2) was the cast optimizer. This
commit extracts the cast optimizer into its own .cpp and .h file. It is large
enough to stand on its own and allows for Local.cpp to return to being a small
group of helper functions.
I am making some changes in this area due to the change in certain function
conventions caused by the +0-normal-arg work. I am just trying to leave the area
a little cleaner than before.
* Reduce array abstraction on apple platforms dealing with literals
Part of the ongoing quest to reduce swift array literal abstraction
penalties: make the SIL optimizer able to eliminate bridging overhead
when dealing with array literals.
Introduce a new classify_bridge_object SIL instruction to handle the
logic of extracting platform specific bits from a Builtin.BridgeObject
value that indicate whether it contains a ObjC tagged pointer object,
or a normal ObjC object. This allows the SIL optimizer to eliminate
these, which allows constant folding a ton of code. On the example
added to test/SILOptimizer/static_arrays.swift, this results in 4x
less SIL code, and also leads to a lot more commonality between linux
and apple platform codegen when passing an array literal.
This also introduces a couple of SIL combines for patterns that occur
in the array literal passing case.
* Implement a few silcombine transformations for arrays
- Useless existential_ref <-> class conversions.
- mark_dependence_inst depending on uninteresting instructions.
- release(init_existential_ref(x)) -> release(x) when hasOneUse(x)
- Update COWArrayOpt to handle the new forms generated by this.
these aren't massive performance wins, but do shrink the size of SIL when
dealing with arrays.
* Generalize testcase to work on linux and on mac when checking stdlib is enabled.
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.
It is important to de-serialize the devirtualized function (and its callees), especially because we must make sure that all transparent functions are de-serialized.
SILCombine did not do that. But as we have the same optimization in the Devirtualizer, it's not needed to duplicate the code in SILCombine.
The only reason we had this peephole in SILCombine is that the Devirtualizer pass could not handle partial applies.
So with this change the Devirtualizer can now also handle partial applies.
Fixes rdar://problem/30544344 (again, after my first attempt failed)
Not sure why but this was another "toxic utility method".
Most of the usages fell into one of three categories:
- The base value was always non-null, so we could just call
getCanonicalType() instead, making intent more explicit
- The result was being compared for equality, so we could
skip canonicalization and call isEqual() instead, removing
some boilerplate
- Utterly insane code that made no sense
There were only a couple of legitimate uses, and even there
open-coding the conditional null check made the code clearer.
Also while I'm at it, make the SIL open archetypes tracker
more typesafe by passing around ArchetypeType * instead of
Type and CanType.
Till now there was no way in SIL to explicitly express a dependency of an instruction on any opened archetypes used by it. This was a cause of many errors and correctness issues. In many cases the code was moved around without taking into account these dependencies, which resulted in breaking the invariant that any uses of an opened archetype should be dominated by the definition of this archetype.
This patch does the following:
- Map opened archetypes to the instructions defining them, i.e. to open_existential instructions.
- Introduce a helper class SILOpenedArchetypesTracker for creating and maintaining such mappings.
- Introduce a helper class SILOpenedArchetypesState for providing a read-only API for looking up available opened archetypes.
- Each SIL instruction which uses an opened archetype as a type gets an additional opened archetype operand representing a dependency of the instruction on this archetype. These opened archetypes operands are an in-memory representation. They are not serialized. Instead, they are re-constructed when reading binary or textual SIL files.
- SILVerifier was extended to conduct more thorough checks related to the usage of opened archetypes.
Till now there was no way in SIL to explicitly express a dependency of an instruction on any opened archetypes used by it. This was a cause of many errors and correctness issues. In many cases the code was moved around without taking into account these dependencies, which resulted in breaking the invariant that any uses of an opened archetype should be dominated by the definition of this archetype.
This patch does the following:
- Map opened archetypes to the instructions defining them, i.e. to open_existential instructions.
- Introduce a helper class SILOpenedArchetypesTracker for creating and maintaining such mappings.
- Introduce a helper class SILOpenedArchetypesState for providing a read-only API for looking up available opened archetypes.
- Each SIL instruction which uses an opened archetype as a type gets an additional opened archetype operand representing a dependency of the instruction on this archetype. These opened archetypes operands are an in-memory representation. They are not serialized. Instead, they are re-constructed when reading binary or textual SIL files.
- SILVerifier was extended to conduct more thorough checks related to the usage of opened archetypes.
Till now there was no way in SIL to explicitly express a dependency of an instruction on any opened archetypes used by it. This was a cause of many errors and correctness issues. In many cases the code was moved around without taking into account these dependencies, which resulted in breaking the invariant that any uses of an opened archetype should be dominated by the definition of this archetype.
This patch does the following:
- Map opened archetypes to the instructions defining them, i.e. to open_existential instructions.
- Introduce a helper class SILOpenedArchetypesTracker for creating and maintaining such mappings.
- Introduce a helper class SILOpenedArchetypesState for providing a read-only API for looking up available opened archetypes.
- Each SIL instruction which uses an opened archetype as a type gets an additional opened archetype operand representing a dependency of the instruction on this archetype. These opened archetypes operands are an in-memory representation. They are not serialized. Instead, they are re-constructed when reading binary or textual SIL files.
- SILVerifier was extended to conduct more thorough checks related to the usage of opened archetypes.
As there are no instructions left which produce multiple result values, this is a NFC regarding the generated SIL and generated code.
Although this commit is large, most changes are straightforward adoptions to the changes in the ValueBase and SILValue classes.
While I'm in this code, generalize it to propagate the original
type information in more cases, including when the original type
is still dependent, and teach it to handle existential metatypes.
rdar://24114020
The main idea here is that we really, really want to be
able to recover the protocol requirement of a conformance
reference even if it's abstract due to the conforming type
being abstract (e.g. an archetype). I've made the conversion
from ProtocolConformance* explicit to discourage casual
contamination of the Ref with a null value.
As part of this change, always make conformance arrays in
Substitutions fully parallel to the requirements, as opposed
to occasionally being empty when the conformances are abstract.
As another part of this, I've tried to proactively fix
prospective bugs with partially-concrete conformances, which I
believe can happen with concretely-bound archetypes.
In addition to just giving us stronger invariants, this is
progress towards the removal of the archetype from Substitution.
This is another bug exposed by changing the instruction allocation.
Depending on the allocated address of new SILBuilder instructions, those instructions were added to the SILCombiner worklist or not.
This bug didn't cause any crashes or miscombiles, but resulted in a non-deterministic result of SILCombine.
(libraries now)
It has been generally agreed that we need to do this reorg, and now
seems like the perfect time. Some major pass reorganization is in the
works.
This does not have to be the final word on the matter. The consensus
among those working on the code is that it's much better than what we
had and a better starting point for future bike shedding.
Note that the previous organization was designed to allow separate
analysis and optimization libraries. It turns out this is an
artificial distinction and not an important goal.
