And unbreak the LLDB testsuite.
This patch fixes three problems with the original implementation:
- Use SILBuilderWithScope instead of SILBuilder to avoid holes in the
lexical scopes.
- Use an artificial location for stores to the alloca to avoid the debugger
stopping before the variable is initialized.
- Recognize debug_value_addr instructions referring to an alloc_stack
instruction to avoid introducing an extra indirection in the debug info.
rdar://problem/31975108
The debug location is now unconditionally being set for each
instruction, so resetting the location at the beginning of each basic
block is redundant.
<rdar://problem/25772716>
The previous change to the DebugInfo/linetable-cleanups.swift test
introduced FileCheck patterns with references to named IR values. In a
release build, the values are anonymous, so those patterns do not match.
rdar://problem/35639381
Switch statements generate at least one anonymous match variable per
case, which consumes both a lot of stack space and an explosion of
range extension depencies due to the way case statements are scoped.
rdar://problem/34326355
Before this change, stepping through the code
1 foo(x,
2 f(a)
3 f(b)
4 )
would visit the code in the order 2, 3, 4, with the function call
being on line 4. After this patch the order is 2, 3, 1 with the
function call being on line 1. This is both closer to what clang
generates for simialar C code and more useful to the programmer since
it is easier to understand which function is being called in a nested
expression.
rdar://problem/35430708
Previously, we just always promoted destroy_addr. With ownership, this does not
work as well since we need to be able to reason about the take operation that we
are performing. In the general case, this would require adding code to
pred-memopts for tracking reads and for compensation store code since we would
need to eliminate the store of the taken value to prevent a double use.
I am going to loop back around later in the year and add back this code once the
time is available. I filed SR-6341. In case any third party contributor is
interested in looking at re-enabling this optimization before I get back to it.
rdar://31521023
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.
... using an inline namespace as the parent of the outermost
declaration(s) that have private or fileprivate accessability. Once
LLDB supports this we can retire the existing hack of storing it as a
fake command line argument.
rdar://problem/18296829
I noticed in a follow-up patch that if you just swiftc without passing Onone
these flags are not set and sometimes happen to default to right thing ... or
not; as can be seen by the test cases modified. For example, at Onone we are
supposed to include an extra swift module "SwiftOnoneSupport".
The filecheck pattern in DebugInfo/liverange-extension-vector.swift
was failing because llvm.dbg.value now has two metadata arguments. The
"[[P]]" pattern was matching "%4, metadata !60" instead of only the "%4".
conversions that reverse an implicit conversion done to align
foreign declarations with their imported types.
For example, consider an Objective-C method that returns an NSString*:
- (nonnull NSString*) foo;
This will be imported into Swift as a method returning a String:
func foo() -> String
A call to this method will implicitly convert the result to String
behind the scenes. If the user then casts the result back to NSString*,
that would normally be compiled as an additional conversion. The
compiler cannot simply eliminate the conversion because that is not
necessarily semantically equivalent.
This peephole recognizes as-casts that immediately reverse a bridging
conversion as a special case and gives them special power to eliminate
both conversions. For example, 'foo() as NSString' will simply return
the original return value. In addition to call results, this also
applies to call arguments, property accesses, and subscript accesses.
as the variable is in. If ASAN inserts additional basic blocks to sanitize
the store, it will reuse the location and scope. If the scope is not a
descendant of the scope of the variable an optimization in the live debug
values pass will kill all DEBUG_VALUEs and thus prevent them from being
propagated.
rdar://problem/30433661
