The new utility folds patterns like
TOP:
// %borrowee is some owned value
%lifetime = begin_borrow %borrowee
BOTTOM:
// %copy is some transitive copy of %borrowee
apply %f(%copy)
end_borrow %lifetime
destroy_value %borrowee
into
TOP:
%move = move_value [lexical] %borrowee
%lifetime begin_borrow [lexical] %move
BOTTOM:
end_borrow %lifetime
apply %f(%move)
It is intended to be run after ShrinkBorrowScope moves the end_borrow up
to just before a relevant apply and after CanonicalizeOSSALifetime moves
destroy_value instructions up to just after their last guaranteed use,
at which point these patterns will exist.
Pulled out a simple check--that CanonicalizeOSSALifetime now uses to
determine whether to continue hoisting a destroy_value instruction--into
a predicate that can be used by LexicalDestroyFolding.
Flow-isolation is a diagnostic SIL pass that finds
unsafe accesses to properties in initializers and
deinitializers that cannot gain isolation to otherwise
protect those accesses from concurrent modifications.
See SE-327 for more details about how and why it exists.
This commit includes changes and features like:
- The removal of the escaping-use restriction
- Flow-isolation that works properly with `defer` statements
- Flow-isolation with an emphasis on helpful diagnostics.
It also includes known issues like:
- Local / nonescaping functions are not analyzed by
flow-isolation, despite it being technically possible.
The main challenge in supporting it efficiently is that
such functions do not have a single exit-point, like
a `defer`. In particular, arbitrary functions can throw
so there are points where nonisolation should _not_ flow
out of the function at a call-site in the initializer, etc.
- The implementation of the flow-isolation pass is not
particularly memory efficient; it relies on BitDataflow
even though the particular flow problem is simple.
So, a more efficient implementation would be specialized for
this particular problem, etc.
There are also some changes to the Swift language itself: defer
will respect its context when deciding its property access kind.
Previously, a defer in an initializer would always access a stored
property through its accessor methods, instead of doing so directly
like its enclosing function might. This inconsistency is unfortunate,
so for Swift 6+ we make this consistent. For Swift 5, only a defer
in a function that is a member of the following kinds of types
will gain this consistency:
- an actor type
- any nominal type that is actor-isolated, excluding UnsafeGlobalActor.
These types are still rather new, so there is much less of a chance of
breaking expected behaviors around defer. In particular, the danger is
that users are relying on the behavior of defer triggering a property
observer within an init or deinit, when it would not be triggering it
without the defer.
This optimizes keypath-closures, like
```
a.map { \.x }
```
It results in a significant performance improvement for such code patterns.
rdar://87968067
Squashed commit of the following:
commit e5a05ffe44
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 27 17:45:31 2022 +0900
cleanup
commit 1f751cea5a
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 27 14:50:33 2022 +0900
cleanups
commit c632f3215d
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 27 14:01:09 2022 +0900
add test for generic from actor decl
commit 09b8bd50a7
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 27 14:00:58 2022 +0900
cleanups
commit 31f4d0cffd
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 27 11:40:51 2022 +0900
fix test
commit ad4db2fb6c
Merge: 97227edcca07e2dfda56
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 26 23:31:41 2022 +0900
Merge branch 'main' into wip-zzz
commit 97227edcca
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 26 21:01:25 2022 +0900
remove @_dynamic methods!
fix tests
commit 1c79344dbb
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 19 12:51:09 2022 +0900
cleanup
wip
stuck
fixed the stack cleanups
cleanups pretty good now
weird load
rki
works
remove hack
add take + throw + return
fix test
more tests fixed
more tests fixed
more tests fixed
commit 3ed494c175
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Tue Jan 18 21:09:28 2022 +0900
stack issues in SIL verification
commit 5cf43a7f86
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Tue Jan 18 09:19:51 2022 +0900
about to call the remoteCall
goot to return, but missing subs
commit df8e47122a
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 13 14:09:49 2022 +0900
[Distributed] Refactor Invocation to Decoder/Encoder
getting there
done-recording
working on the string init
stuck trying to get String initializer SILFunction
created the remote call target
commit fc7bd62f32
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 23:01:14 2022 +0900
[Distributed] Pass arguments from Invocation to HBuffer
commit cafc2cc058
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 22:08:47 2022 +0900
works
commit a7d01837ff
Author: Pavel Yaskevich <pyaskevich@apple.com>
Date: Tue Jan 11 15:48:58 2022 -0800
[Distributed] Adjust interface of `swift_distributed_execute_target`
Since this is a special function, `calleeContext` doesn't point to
a direct parent but instead both parent context (uninitialized)
and resume function are passed as last arguments which means that
`callContext` has to act as an intermediate context in call to accessor.
commit c1f830be27
Author: Pavel Yaskevich <pyaskevich@apple.com>
Date: Tue Jan 11 17:00:08 2022 -0800
[Distributed] Drop optionality from result buffer in `_executeDistributedTarget`
`RawPointer?` is lowered into a two arguments since it's a struct,
to make it easy let's just allocate an empty pointer for `Void` result.
commit c83c2c37b6
Author: Pavel Yaskevich <pyaskevich@apple.com>
Date: Tue Jan 11 17:02:45 2022 -0800
[Distributed] NFC: Update _remoteCall test-case to check multiple different result types
commit 29e7cf50e4
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 21:32:37 2022 +0900
wip
commit 9128ecc6f8
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 20:46:20 2022 +0900
wip
commit a6b2a62a67
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 20:38:22 2022 +0900
wip
commit 8b188f0d43
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 16:55:10 2022 +0900
wip
commit 3796bec2b9
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Jan 12 16:55:02 2022 +0900
wip
commit 0ffc68b5ef
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Tue Jan 11 21:44:58 2022 +0900
[Distributed] Implementing ad-hoc protocol requirements
commit 78862575e4
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 6 18:03:54 2022 +0900
cleanup
commit 5f4ab89e25
Merge: 24a628e7c0fdda6f2ee4
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 6 15:51:39 2022 +0900
Merge branch 'main' into wip-impl-execute-swift
commit 24a628e7c0
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Thu Jan 6 15:33:21 2022 +0900
wip
commit 69e7fed09d
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Dec 22 06:36:45 2021 +0900
[Distributed] comment out distributed_actor_remoteCall for now
commit 376733a9f6
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Tue Dec 21 16:00:06 2021 +0900
reimplement distributed get type info impls
commit 74ab47886a
Author: Konrad `ktoso` Malawski <konrad_malawski@apple.com>
Date: Wed Dec 15 21:37:08 2021 +0900
[Distributed] Implement func metadata and executeDistributedTarget
dont expose new entrypoints
able to get all the way to calling _execute
The `run-unit-tests` is a "pseudo" pass which is invoked from sil-opt and runs all the unit tests, implemented in Swift.
This is done from the `swift-unit-tests.sil` lit test.
* add `BasicBlockSet`
* add `BasicBlockWorklist`
* add `BasicBlockRange`, which defines a range of blocks from a common dominating “begin” block to a set of “end” blocks.
* add `InstructionRange`, which is similar to `BasicBlockRange`, just on instruction level. It can be used for value lifetime analysis.
* rename `StackList` -> `Stack` and move it to `Optimizer/DataStructures`
* rename `PassContext.passContext` to `PassContext._bridged`
* add notify-functions to PassContext
Introduce a new instruction `dealloc_stack_ref ` and remove the `stack` flag from `dealloc_ref`.
The `dealloc_ref [stack]` was confusing, because all it does is to mark the deallocation of the stack space for a stack promoted object.
And a few other small related changes:
* remove libswiftPassInvocation from SILInstructionWorklist (because it's not needed)
* replace start/finishPassRun with start/finishFunction/InstructionPassRun
NFC
Extract and rewrite the destroy hoisting algorithm originally from
CopyForwarding (in 2014).
This is now a light-weight utility for hoisting destroy_addr
instructions. Shrinking an object's memory lifetime can allow removal
of copy_addr and other optimization.
This is extremely low-overhead and can run at any optimization level
without dependency on any analysis.
This algorithm is:
- Incremental
- SSA-based
- Canonical
- Free from alias analysis
See file-level comments.
The immediate purpose is to specify and test the constraints
introduced by adding lexical variable lifetimes to SIL semantics. It
can be used as a template for end_borrow hoisting.
Ultimately, this utility can be invoked within any pass that needs to
optimize a particular uniquely identified address. It will be used to
remove much of the complexity from CopyForwarding.
Pessimistic, non-iterative data flow for analyzing backward reachability
from a set of last uses to their dominating def or nearest barrier.
Meet: ReachableEnd(predecessor) = intersection(ReachableBegin, successors)
Intended for frequently called utilities where minimizing the cost of
data flow is more important than analyzing reachability across
loops. Expected to visit very few blocks because barriers often occur
close to a last use.
Note: this does not require initializing bitsets for all blocks in the
function for each SSA value being analyzed.
* rename the CrossModuleSerializationSetup pass to simply CrossModuleOptimization
* remove the CMO specific serializer pass. Instead run the CrossModuleSerializationSetup pass directly before the standard serializer pass.
* correctly handle shared functions (e.g. specializations)
* refactoring
So that CopyPropagation and other clients can react accordingly, pass
back a list of copy_value instructions that were rewritten by
ShrinkBorrowScope. In CopyPropagation, add each modified copy to the
copy worklist.
* [Distributed] towards DistributedActorSystem; synthesize the id earlier, since Identifiable.id
* Fix execute signature to what Pavel is working with
* funcs are ok in sil
* fixed lifetime of id in inits
* fix distributed_actor_deinit
* distributed_actor_local
* update more tests
fixing tests
fix TBD test
fix Serialization/distributed
fix irgen test
Fix null pointer crashes
* prevent issues with null func ptrs and fix Distributed prorotocol test
* fix deinit sil test
To give a bit more information, currently the way the move function is
implemented is that:
1. SILGen emits a builtin "move" that is called within the function _move in the
stdlib.
2. Mandatory Inlining today if the final inlined type is address only, inlines
builtin "move" as mark_unresolved_move_addr. Otherwise, if the inlined type
is loadable, it performs a load [take] + move [diagnostic] + store [init].
3. In the diagnostic pipeline before any mem optimizations have run, we run the
move checker for addresses. This eliminates /all/ mark_unresolved_move_addr
as part of emitting diagnostics. In order to make this work, we perform a
small optimization before the checker runs that moves the
mark_unresolved_move_addr from being on temporary alloc_stacks to the true
base underlying address we are trying to move. This optimization is necessary
since _move is generic and often times SILGen will emit this temporary that
we do not want.
4. Then after we have run the guaranteed mem optimizations, we run the object
based move checker emitting diagnostics.
This PR changes the scheme above to the following:
1. SILGen emits a builtin "move" that is called within the function _move in the
stdlib.
2. Mandatory Inlining inlines builtin "move" as mark_unresolved_move_addr.
3. In the diagnostic pipeline before we have run any mem optimizations and
before we have run the actual move address checker, we massage the IR as we
do above but in a separate pass where in addition we try to match this pattern:
```
%temporary = alloc_stack $LoadableType
store %1 to [init] %temporary : $*LoadableType
mark_unresolved_move_addr %temporary to %otherAddr : $*LoadableType
destroy_addr %temporary : $*LoadableType
```
and transform it to:
```
%temporary = alloc_stack $LoadableType
%2 = move_value [allows_diagnostics] %1 : $*LoadableType
store %2 to [init] %temporary : $*LoadableType
destroy_addr %temporary : $*LoadableType
```
ensuring that the object move checker will handle this.
4. Then after we have run the guaranteed mem optimizations, we run the object
based move checker emitting diagnostics.
During copy propagation (for which -enable-copy-propagation must still
be passed), also try to shrink borrow scopes by hoisting end_borrows
using the newly added ShrinkBorrowScope utility.
Allow end_borrow instructions to be hoisted over instructions that are
not deinit barriers for the value which is borrowed. Deinit barriers
include uses of the value, loads of memory, loads of weak references
that may be zeroed during deinit, and "synchronization points".
rdar://79149830
Rerun RLE with cutting off the base address of loads at `ref_element/tail_addr [immutable]`. This increases the chance of catching loads of immutable COW class properties or elements.
* Replace the uniqueness result of a begin_cow_mutation of an empty Array/Set/Dictionary singleton with zero.
* Remove empty begin_cow_mutation - end_cow_mutation pairs
* Remove empty end_cow_mutation - begin_cow_mutation pairs
Use the include search path for the header search rather than
constructing the path using relative paths. This is important for
getting the interop to work properly as well.
Fix isScopeAffectingInstructionDead to use this new API. A stdlib
assert, which has "program_termination" semantics, should not be
considered read-only.
isReadOnlyConstantEvaluableCall API:
/// Return true iff the \p applySite is constant-evaluable and read-only.
///
/// Functions annotated as "constant_evaluable" are assumed to be "side-effect
/// free", unless their signature and substitution map indicates otherwise. A
/// constant_evaluable function call is read only unless it:
/// (1) has generic parameters
/// (2) has inout parameters
/// (3) has indirect results
///
/// Read-only constant evaluable functions can do only the following and
/// nothing else:
/// (1) The call may read any memory location.
/// (2) The call may destroy owned parameters i.e., consume them.
/// (3) The call may write into memory locations newly created by the call.
/// (4) The call may use assertions, which traps at runtime on failure.
/// (5) The call may return a non-generic value.
///
/// Essentially, these are calls whose "effect" is visible only in their return
/// value or through the parameters that are destroyed. The return value
/// is also guaranteed to have value semantics as it is non-generic and
/// reference semantics is not constant evaluable.
NOTE: This pass is disabled when -enable-experimental-lexical-lifetimes is
enabled.
When that flag is disabled, this removes the lexical flag from begin_borrow and
alloc_stack. This ensures that we can begin using begin_borrow [lexical] and
friends to emit diagnostics without impacting performance. I am going to be
preparing a subsequent patch that causes us to emit lexical lifetimes by
default. Due to this pass, I am not expecting any issues around perf.
This is just an initial prototype for people to play with. It is as always
behind the -enable-experimental-move-only flag.
NOTE: In this PR I implemented this only for 'local let' like things (local
lets/params). I did not implement in this PR support for local var and haven't
done anything with class ivars or globals.
rdar://83957028
