This utility is generally a horrible idea but even worse the
callers were not doing anything to ensure the required
invariants actually held.
Add a new canReplaceLoadSequence() method and chek it in the
right places.
yields in generalized accessors: _read and _modify, which are
yield-once corountines. This pass is based on the existing SIL verifier
checks but diagnoses only those errors that can be introduced by programmers
when using yields.
<rdar://43578476>
I also ported the constant_propagation.sil tests over for ownership and updated
a few parts of the cast optimizer so that those tests pass with and without
ownership. I purposely only updated the parts of the cast optimizer that crashed
with ownership in the relevant test so that I can add new sil code coverage for
those uncovered code paths.
replace value uses action to make sure that we properly notify passes that we
made the change.
This fixes a latent bug where we were not notifying SILCombine about this
replacement.
Specifically:
- auto replaceCast = [&](SingleValueInstruction *NewCast) {
- assert(Ty.getAs<AnyMetatypeType>()->getRepresentation()
- == NewCast->getType().getAs<AnyMetatypeType>()->getRepresentation());
- MCI->replaceAllUsesWith(NewCast);
- EraseInstAction(MCI);
- return NewCast;
+ auto replaceCast = [&](SILValue newValue) -> SILValue {
+ assert(ty.getAs<AnyMetatypeType>()->getRepresentation() ==
+ newValue->getType().getAs<AnyMetatypeType>()->getRepresentation());
+ ReplaceValueUsesAction(mci, newValue);
+ EraseInstAction(mci);
+ return newValue;
};
Notice how we use MCI->replaceAllUsesWith instead of one of our replace call
backs. SILCombine hooks these to know if it should re-run users.
NOTE: I changed all places that the CastOptimizer is created to just pass in
nullptr for now so this is NFC.
----
Right now the interface of the CastOptimizer is muddled and confused. Sometimes
it is returning a value that should be used by the caller, other times it is
returning an instruction that is meant to be reprocessed by the caller.
This series of patches is attempting to clean this up by switching to the
following model:
1. If we are optimizing a cast of a value, we return a SILValue. If the cast
fails, we return an empty SILValue().
2. If we are optimizing a cast of an address, we return a boolean value to show
success/failure and require the user to use the SILBuilderContext to get the
cast if they need to.
This normalizes the creation of pass pipelines by ensuring that all pass
pipelines take a SILOption instead of only some. It also makes it so that we do
not need to propagate options through various pipeline creation helpers.
Beside fixing the compiler crash, this change also improves the stack-nesting correction mechanisms in the inliners:
* Instead of trying to correct the nesting after each inlining of a callee, correct the nesting once when inlining is finished for a caller function.
This fixes a potential compile time problem, because StackNesting iterates over the whole function.
In worst case this can lead to quadratic behavior in case many begin_apply instructions with overlapping stack locations are inlined.
* Because we are doing it only once for a caller, we can remove the complex logic for checking if it is necessary.
We can just do it unconditionally in case any coroutine gets inlined.
The inliners iterate over all instruction of a function anyway, so this does not increase the computational complexity (StackNesting is roughly linear with the number of instructions).
rdar://problem/47615442
This will let me strip ownership from non-transparent functions at the beginning
of the perf pipeline. Then after we serialize, I will run OME on the transparent
functions. Otherwise, we can not perform mandatory inlining successfully since
we can not inline ossa into non-ossa functions.
Instead of some special treatment of unreachable blocks, model unreachable as implicitly deallocating all alive stack locations at that point.
This requires an additional forward-dataflow pass. But it now correctly models the problem and fixes a compiler crash.
rdar://problem/47402694
I discovered while updating PMO for ownership that for ~5 years there has been a
bug where we were treating copy_addr of trivial values like an "Assign" (in PMO
terminology) of a non-trivial value and thus stopping allocation
elimination. When I fixed this I discovered that this caused us to no longer
emit diagnostics in a predictable way. Specifically, consider the following
swift snippet:
var _: UInt = (-1) >> 0
Today, we emit a diagnostic that -1 can not be put into a UInt. This occurs
since even though the underlying allocation is only stored into, the copy_addr
assign keeps it alive, causing the diagnostics pass to see the conversion. With
my fix though, we see that we are only storing into the allocation, causing the
allocation to be eliminated before the constant propagation diagnostic pass
runs, causing the diagnostic to no longer be emitted.
We should truly not be performing this type of DCE before we emit such
diagnostics. So in this commit, I split the pass into two parts:
1. A load promotion pass that performs the SSA formation needed for SSA based
diagnostics to actually work.
2. An allocation elimination passes that run /after/ SSA based diagnostics.
This should be NFC since the constant propagation SSA based diagnostics do not
create memory operations so the output should be the same.
Improves SwiftSyntax release build speed by 4x.
Limit the size of the sets tracked by inter procedural
AccessedStorageAnalysis. There is a lot of leeway in this limit since
"normal" code doesn't come close to hitting it and SwiftSyntax compile
time isn't noticeably affected until 10x this limit.
This change also avoids reanalyzing function bodies once results have
bottomed out and avoids copying sets in the common case.
Fixes <rdar://problem/46905624> Release build time regression, a lot of time spent on AccessEnforcementOpts::run().
This is just a band aid. The fundamental problem is really:
<rdar://problem/47195282> Recomputing BottomUpIPAnalysis takes most of
the SwiftSyntax compile time.
SwiftSyntax compile time is still at least an order of magnitude
longer than it should be.
Generalizes the ConcreteExistentialInfo abstraction so it can be used
both by the ExistentialSpecializer and SILCombine, allowing redundant
code in ExistentialSpecializer.cpp to be deleted.
Splits OpenedArchetypeInfo from ConcreteExistentialInfo. Adds a
ConcreteOpenedArchetypeInfo convenience wrapper around them both, for
use wherever we were originally using ConcreteExistentialInfo.
Splits getAddressOfStackInit into getStackInitInst, This is cleaner and
allows both the ExistentialSpecializer and SILCombine to handle more
interesting cases in the future, like unconditional_checked_cast.
Creates utilities, initializeSubstitutionMap, and
initializeConcreteTypeDef to simplify an generalize
ConcreteExistentialInfo.
While rewriting ExistentialSpecializer to use the new
abstraction, I fixed a latent bug in which is was using a SIL
argument index as a function type parameter index (this would
have broken up if/when we decide to enable calls with indirect
results).
This cannot be correctly done as a SILCombine because it must create
new instructions at a previous location. Move the optimization into
CastOptimizer. Insert the new metatype instructions in the correct
spot. And manually do the replaceAllUsesWith and eraseInstruction
steps.
Fixes <rdar://problem/46746188> crash in swift_getObjCClassFromObject.
We've been running doxygen with the autobrief option for a couple of
years now. This makes the \brief markers into our comments
redundant. Since they are a visual distraction and we don't want to
encourage more \brief markers in new code either, this patch removes
them all.
Patch produced by
for i in $(git grep -l '\\brief'); do perl -pi -e 's/\\brief //g' $i & done
This allows Swift code to implement a fast path via a protocol type
check as follows:
if let existentialVal = genericVal as? SomeProtocol {
// do something fast.
}
Fixes <rdar://problem/46322928> Failure to devirtualize a protocol
method applied to an opened existential blocks implemention of
DataProtocol.
Note: the approach of devirtualization via backward pattern matching
is fundamentally wrong and will never be fully general. It should be a
forward type propagation.
This is in preparation for verifying that when ownership verification is enabled
that only enums and trivial values can have any ownership. I am doing this in
preparation for eliminating ValueOwnershipKind::Trivial.
rdar://46294760
Implements a constant interpreter that can deal with basic integer operations.
Summary of the features that it includes:
* builtin integer values, and builtin integer insts
* struct and tuple values, and insts that construct and extract them (necessary to use stdlib integers)
* function referencing and application (necessary to call stdlib integer functions)
* error handling data structures and logic, for telling you why your value is not evaluatable
* metatype values (not necessary for integers, but it's only a few extra lines, so I thought it would be more trouble than it's worth to put them in a separate PR)
* conditional branches (ditto)
Inlining has always been quadratic for no good reason. There was a
special hack for single-block callees that allowed linear inlining.
Instead, the now iterates over blocks and instructions in reverse,
splitting blocks as it inlines. There no longer needs to be special
case for single block callees, and the inliner is linear for all kinds
of callees.
This further simplifies and cleans up the code. There are just a few
basic invariants that the common inliner needs to provide about how
blocks are split and laid out. We can do this if we don't add hacks
for special cases within the inliner. Those invariants allow the
inliner clients to be much simpler and more efficient.
PerformanceInliner still needs to be fixed.
Fixes SR-9223: Inliner exhibits slow compilation time with a large
static array.
A recent SILCloner rewrite removed a special case hack for single
basic block callee functions:
commit c6865c0dff
Merge: 76e6c4157e9e440d13a6
Author: Andrew Trick <atrick@apple.com>
Date: Thu Oct 11 14:23:32 2018
Merge pull request #19786 from atrick/silcloner-cleanup
SILCloner and SILInliner rewrite.
Instead, the new inliner simply merges trivial unconditional branches
after inlining the return block. This way, the CFG is always in
canonical state after inlining. This is more robust, and avoids
interfering with subsequent SIL passes when non-single-block callees
are inlined.
The problem is that inlining a series of calls within a large block
could result in interleaved block splitting and merging operations,
which is quadratic in the block size. This showed up when inlining the
tens of thousands of array subscript calls emitted for a large array
initialization.
The first half of the fix is to simply defer block merging until all
calls are inlined. We can't expect SimplifyCFG to run immediately
after inlining, nor would we want to do that, *especially* for
mandatory inlining. This fix instead exposes block merging as a
trivial utility.
Note: by eliminating some unconditional branches, this change could
reduce the number of debug locations emitted. This does not
fundamentally change any debug information guarantee, and I was unable
to observe any behavior difference in the debugger.
In the included, test case, the optimization was sinking
releases past is_escaping_closure.
Rewrite the isBarrier logic to be conservative and define the
mayCheckRefCount property in SIL/InstructionUtils. Properties that may
need to be updated when SIL changes belong there.
Note that it is particularly bad behavior if the presence of access
markers in the code cause miscompiles unrelated to access enforcement.
Fixes <rdar://problem/45846920> TestFoundation, TestProcess, closure
argument passed as @noescape to Objective-C has escaped.
General case:
begin_access A
...
strong_release / release_value / destroy
end_access
The release instruction can be sunk below the end_access instruction,
This extends the lifetime of the released value, but, might allow us to
Mark the access scope as no nested conflict.
Rewrite the SILCLoners used in SimplifyCFG. For convenience, there is
now simply a BasicBlockCloner and a SILFunctionCloner. It's pretty
obvious what they do and almost impossible to use incorrectly.
This is worthwhile on its own just to make the usage clear, but the
real reason is that after this cleanup, it will be possible to remove
many extraneous calls to global critical edge splitting related to
cloning.
General case:
—
begin_access A (may or may not have no_nested_conflict)
load/store
end_access
apply // may have a scoped access that conflicts with A
begin_access A [no_nested_conflict]
load/store
end_access A
—
The second access scope does not need to be emitted.
NOTE: KeyPath access must be identified at the top-level, non-inlinable stdlib entry point.
As such, The sodlib entry pointed is annotated by a new @_semantics that is equivalent to inline(never)
Introduce an "early redundant load elimination", which does not optimize loads from arrays.
Later array optimizations, like ABCOpt, get confused if an array load in a loop is converted to a pattern with a phi argument.
This problem was introduced with accessors.
rdar://problem/44184763
