Add a diagnostic pass that emits errors when a violation of the "Law of
Exclusivity" is detected at compile time. The Law of Exclusivity requires
that the access duration of any access to an address not overlap
with an access to the same address unless both accesses are reads.
This pass relies on 'begin_access' and 'end_access' SIL instruction
markers inserted by SILGen to determine when an access to an address begins and
ends. It models the in-progress accesses with a map from storage locations to
the counts of read and write-like accesses in progress for that location.
At some point, pass definitions were heavily macro-ized. Pass
descriptive names were added in two places. This is not only redundant
but a source of confusion. You could waste a lot of time grepping for
the wrong string. I removed all the getName() overrides which, at
around 90 passes, was a fairly significant amount of code bloat.
Any pass that we want to be able to invoke by name from a tool
(sil-opt) or pipeline plan *should* have unique type name, enum value,
commend-line string, and name string. I removed a comment about the
various inliner passes that contradicted that.
Side note: We should be consistent with the policy that a pass is
identified by its type. We have a couple passes, LICM and CSE, which
currently violate that convention.
I improved all of the pass names so they are minimal but fully descriptive.
The PASS macros contained wordy, unhelpful descriptions, which served
no functional purpose. Those descriptions are permanently preseved in
git history. If anyone thinks they contain useful information, then
they should be added to the *definition* of the pass (where passes are
actually described).
Long-winded pass comments have no place in Passes.def. The purpose of
those macros is to associate a unique type name, enum value,
command-line-option name, and pretty name with each pass for tooling
purposes. Any one of those entities needs to be sufficient for looking
up the others. That's not where anyone should look read a description
of the pass.
I am going to run it very early and use it to ensure that extra copies due to my
refactoring of SILGenPattern do not cause COW copies to be introduced.
For now, it does a very simple optimization, namely, it eliminates a copy_value,
with only a destroy_value user on a guaranteed parameter.
It is now disabled behind a flag.
Hoist alloc_stack instructions of 'generic' or resilient type to the entry
block. At the same time also perform a very simple stack coloring analysis.
This does not use a true liveness-analysis yet but rather employs some simple
conservative checks to see whether the live ranges of two alloc_stacks might
interfere.
AllocStackHoisting is an IRGen SIL pass. This allows for using IRGen's type
lowering information. Furthermore, hoisting and merging the alloc_stack
instructions this late does not interfere with SIL optimizations because the
resulting SIL never gets serialized.
This pipeline is run as part of IRGen and has access to the IRGenModule.
Passes that run as part of this pipeline can query for the IRGenModule.
We will use it for the AllocStackHoisting pass. It wants to know if a type is of
non-fixed size.
To break the cyclic dependency between IRGen -> SILOptimizer -> IRGen that would
arise from the SILPassManager having to know about the createIRGENPASS()
function IRGen passes instead of exposing this function dynamically have to add
themselves to the pass manager.
This pass works by blowing up if it finds an apply that calls a function
specified via the cl command line option 'bug-reducer-tester-target-func'. This
makes it easy to test sil-bug-reducer.
This is a hidden option. It should be used like: -assume-single-threaded
When this function is provided, the compiler assumes that the code will be executed in the single threaded mode. It then performs certain optimizations that can benefit from it, e.g. it marks as non-atomic all reference counting instructions in the user code being compiled.
Often times SILGen wants to hold onto values that have been copied. This causes
an issue, when due to Cleanups firing, SILBuilder inserts destroys and destroys
the copy that produced the value that SILGen held onto. This will then cause
SILGen to emit incorrect code.
There really is no reason to introduce such complexity into SILBuilder when a
small simple guaranteed pass can perform the same work. Thus the introduction of
this pass.
In a later commit, I am going to eliminate the SILBuilder entry points.
rdar://28685236
radar rdar://problem/28434323
SILGen has no reason to insert shadow copies for inout parameters any more. They cannot be captured. We still emit these copies. Sometimes deshadowing removes them, but sometimes it does not.
In this PR we just avoid emitting the copies and remove the deshadowing pass.
This PR chery-picked some of @dduan work and built on top of it.
This consists of 3 parts:
1) Extend CallerAnalysis to also provide information if a function is partially applied
2) A new DeadArgSignatureOpt pass, similar to FunctionSignatureOpts, which just specializes for dead arguments of partially applied functions.
3) Let CapturePropagation eliminate such partial_apply instructions and replace them with a thin_to_thick conversion of the specialized functions.
This optimzation improves benchmarks where static struct or class functions are passed as a closure (e.g. -20% for SortStrings).
Such functions have a additional metatype parameter. We used to create a partial_apply in this case, which allocates a context, etc.
But this is not necessary as the metatype parameter is not used in most cases.
rdar://problem/27513085
This commit fixes a bug where we were not checking that all predecessors had the
cond_fail block as its only successor. This occured since we were bailing early
when we saw a constant. So if we saw a predecessor with a constant before a
predecessor that had multiple successors, we would optimize even though we would
be introducing an extra cond_fail along a path.
I added a new utility pass to test this code since so much is going on in
SimplifyCFG that it is difficult to construct a test case running the full
pass.
Really this code should be in a different pass (properly SIL Code Motion TBH).
But for now, this commit just fixes the bug.
rdar://26904047
For details see the comment in ConditionForwarding.cpp.
This optimization pass helps to optimize loops iterating over closed ranges, e.g. for i in 0...n { }
We can remove the retain/release pair preceeding the builtins based on the
knowledge that the lifetime of the reference is guaranteed by someone hanging on
to the reference elsewhere.
Eventually, we decided to do this
1. Have the function signature opts (used to be called the cloner to create
the optimized function.
2. Mark the thunk as always_inline
3. Rely on the inliner to inline the thunk to get the benefit of calling optimized
function directly.
We decided to use the inliner to rewrite the caller's callsites.
And eventually I will turn FunctionSignatureAnalysis into a Utility.
As its data should only be used and kept in the cloner pass.
This was mistakenly reverted in an attempt to fix buildbots.
Unfortunately it's now smashed into one commit.
---
Introduce @_specialize(<type list>) internal attribute.
This attribute can be attached to generic functions. The attribute's
arguments must be a list of concrete types to be substituted in the
function's generic signature. Any number of specializations may be
associated with a generic function.
This attribute provides a hint to the compiler. At -O, the compiler
will generate the specified specializations and emit calls to the
specialized code in the original generic function guarded by type
checks.
The current attribute is designed to be an internal tool for
performance experimentation. It does not affect the language or
API. This work may be extended in the future to add user-visible
attributes that do provide API guarantees and/or direct dispatch to
specialized code.
This attribute works on any generic function: a freestanding function
with generic type parameters, a nongeneric method declared in a
generic class, a generic method in a nongeneric class or a generic
method in a generic class. A function's generic signature is a
concatenation of the generic context and the function's own generic
type parameters.
e.g.
struct S<T> {
var x: T
@_specialize(Int, Float)
mutating func exchangeSecond<U>(u: U, _ t: T) -> (U, T) {
x = t
return (u, x)
}
}
// Substitutes: <T, U> with <Int, Float> producing:
// S<Int>::exchangeSecond<Float>(u: Float, t: Int) -> (Float, Int)
---
[SILOptimizer] Introduce an eager-specializer pass.
This pass finds generic functions with @_specialized attributes and
generates specialized code for the attribute's concrete types. It
inserts type checks and guarded dispatch at the beginning of the
generic function for each specialization. Since we don't currently
expose this attribute as API and don't specialize vtables and witness
tables yet, the only way to reach the specialized code is by calling
the generic function which performs the guarded dispatch.
In the future, we can build on this work in several ways:
- cross module dispatch directly to specialized code
- dynamic dispatch directly to specialized code
- automated specialization based on less specific hints
- partial specialization
- and so on...
I reorganized and refactored the optimizer's generic utilities to
support direct function specialization as opposed to apply
specialization.
This split the function signature module pass into 2 functin passes.
By doing so, this allows us to rewrite to using the FSO-optimized
function prior to attempting inlining, but allow us to do a substantial
amount of optimization on the current function before attempting to do
FSO on that function.
And also helps us to move to a model which module pass is NOT used unless
necesary.
I do not see regression nor improvement for on the performance test suite.
functionsignopts.sil and functionsignopt_sroa.sil are modified because the
mangler now takes into account of information in the projection tree.
Temporarily reverting @_specialize because stdlib unit tests are
failing on an internal branch during deserialization.
This reverts commit e2c43cfe14, reversing
changes made to 9078011f93.
This change follows up on an idea from Michael (thanks!).
It enables debugging and profiling on SIL level, which is useful for compiler debugging.
There is a new frontend option -gsil which lets the compiler write a SIL file and generated debug info for it.
For details see docs/DebuggingTheCompiler.rst and the comments in SILDebugInfoGenerator.cpp.
This pass finds generic functions with @_specialized attributes and
generates specialized code for the attribute's concrete types. It
inserts type checks and guarded dispatch at the beginning of the
generic function for each specialization. Since we don't currently
expose this attribute as API and don't specialize vtables and witness
tables yet, the only way to reach the specialized code is by calling
the generic function which performs the guarded dispatch.
In the future, we can build on this work in several ways:
- cross module dispatch directly to specialized code
- dynamic dispatch directly to specialized code
- automated specialization based on less specific hints
- partial specialization
- and so on...
I reorganized and refactored the optimizer's generic utilities to
support direct function specialization as opposed to apply
specialization.
We were giving special handling to ApplyInst when we were attempting to use
getMemoryBehavior(). This commit changes the special handling to work on all
full apply sites instead of just AI. Additionally, we look through partial
applies and thin to thick functions.
I also added a dumper called BasicInstructionPropertyDumper that just dumps the
results of SILInstruction::get{Memory,Releasing}Behavior() for all instructions
in order to verify this behavior.
This is similar and yet different from epilogue release matcher. Particularly
how retain is found and when to bail. Therefore this is put into a different
class than ConsumedArgToEpilogueReleaseMatcher
This is currently a NFC other than some basic testing using the epilogue dumper.
This allows for the RCIdentityAnalysis to be tested independent of other
passes.
Also add some initial tests for RCIdentity. I am stepping through "strip by
strip" but I did not have time to finish the coverage.
Add back a stand-alone devirtualizer pass, running prior to generic
specialization. As with the stand-alone generic specializer pass, this
may add functions to the pass manager's work list.
This is another step in unbundling these passes from the performance
inliner.
Begin unbundling devirtualization, specialization, and inlining by
recreating the stand-alone generic specializer pass.
I've added a use of the pass to the pipeline, but this is almost
certainly not going to be the final location of where it runs. It's
primarily there to ensure this code gets exercised.
Since this is running prior to inlining, it changes the order that some
functions are specialized in, which means differences in the order of
output of one of the tests (one which similarly changed when
devirtualization, specialization, and inlining were bundled together).
This makes it easy to use -sil-verify-all to verify that both type of info are
created correctly and that analyses are properly updating them. I am going to
use this to harden testing of the loop canonicalizer.
This enables array value propagation in array literal loops like:
for e in [2,3,4] {
r += e
}
Allowing us to completely get rid of the array.
rdar://19958821
SR-203
This reverts commit 82ff59c0b9.
Original commit message:
This allows us to compile the function:
func valueArray() -> Int{
var a = [1,2,3]
var r = a[0] + a[1] + a[2]
return r
}
Down to just a return of the value 6. And should eventually allow us to remove
the overhead of vararg calls.
rdar://19958821
(Headers first)
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.
