Although I don't plan to bring over new assertions wholesale
into the current qualification branch, it's entirely possible
that various minor changes in main will use the new assertions;
having this basic support in the release branch will simplify that.
(This is why I'm adding the includes as a separate pass from
rewriting the individual assertions)
getVarInfo() now always returns a variable with a location and scope.
To opt out of this change, getVarInfo(false) returns an incomplete variable.
This can be used to work around bugs, but should only really be used for
printing.
The complete var info will also contain the type, except for debug_values,
as its type depends on another instruction, which may be inconsistent if
called mid-pass.
All locations in debug variables are now also stripped of flags, to avoid
issues when comparing or hashing debug variables.
For years, optimizer engineers have been hitting a common bug caused by passes
assuming all SILValues have a parent function only to be surprised by SILUndef.
Generally we see SILUndef not that often so we see this come up later in
testing. This patch eliminates that problem by making SILUndef uniqued at the
function level instead of the module level. This ensures that it makes sense for
SILUndef to have a parent function, eliminating this possibility since we can
define an API to get its parent function.
rdar://123484595
This function is clearly returning the opposite of what its name says:
```
const SILDebugScope *SILBasicBlock::getScopeOfFirstNonMetaInstruction() {
for (auto &Inst : *this)
if (Inst.isMetaInstruction())
return Inst.getDebugScope();
return begin()->getDebugScope();
}
```
Looking at the PR history (sadly GH doesn't preserve old versions of the patch...)
https://github.com/apple/swift/pull/15575
There was this snippet of code:
```
// Find the correct debug scope for alloc stack. We want to give to the
// expanded sequence the correct debug scope so we skip over instructions
// that aren't lowered to anything real (e.g. debug_value).
static const SILDebugScope *findAllocStackDebugScope(SILBasicBlock &BB) {
auto It = BB.begin();
while (It != BB.end()) {
if (!isMaintenanceInst(&*It))
```
We don't know what used to be after that line but, based on the comments, it
must have been a `return It->getDebugScope()`.
Adrian then asked the author to make this a different helper function
`getScopeOfFirstNonMetaInstruction`, and the subsequence force-push had the code
we see today. So maybe it was in this conversion that the author made the
mistake?
Fixing the implementation doesn't cause any tests to fail (sadly the original PR
did not add any SIL->SIL tests, which would have been ideal).
Introduce two modes of bridging:
* inline mode: this is basically how it worked so far. Using full C++ interop which allows bridging functions to be inlined.
* pure mode: bridging functions are not inlined but compiled in a cpp file. This allows to reduce the C++ interop requirements to a minimum. No std/llvm/swift headers are imported.
This change requires a major refactoring of bridging sources. The implementation of bridging functions go to two separate files: SILBridgingImpl.h and OptimizerBridgingImpl.h.
Depending on the mode, those files are either included in the corresponding header files (inline mode), or included in the c++ file (pure mode).
The mode can be selected with the BRIDGING_MODE cmake variable. By default it is set to the inline mode (= existing behavior). The pure mode is only selected in certain configurations to work around C++ interop issues:
* In debug builds, to workaround a problem with LLDB's `po` command (rdar://115770255).
* On windows to workaround a build problem.
This is phase-1 of switching from llvm::Optional to std::optional in the
next rebranch. llvm::Optional was removed from upstream LLVM, so we need
to migrate off rather soon. On Darwin, std::optional, and llvm::Optional
have the same layout, so we don't need to be as concerned about ABI
beyond the name mangling. `llvm::Optional` is only returned from one
function in
```
getStandardTypeSubst(StringRef TypeName,
bool allowConcurrencyManglings);
```
It's the return value, so it should not impact the mangling of the
function, and the layout is the same as `std::optional`, so it should be
mostly okay. This function doesn't appear to have users, and the ABI was
already broken 2 years ago for concurrency and no one seemed to notice
so this should be "okay".
I'm doing the migration incrementally so that folks working on main can
cherry-pick back to the release/5.9 branch. Once 5.9 is done and locked
away, then we can go through and finish the replacement. Since `None`
and `Optional` show up in contexts where they are not `llvm::None` and
`llvm::Optional`, I'm preparing the work now by going through and
removing the namespace unwrapping and making the `llvm` namespace
explicit. This should make it fairly mechanical to go through and
replace llvm::Optional with std::optional, and llvm::None with
std::nullopt. It's also a change that can be brought onto the
release/5.9 with minimal impact. This should be an NFC change.
"reborrow" flag on the SILArgument avoids transitive walk over the phi operandsi
to determine if it is a reborrow in multiple utilities.
SIL transforms must keep the flag up-to-date by calling SILArgument::setReborrow.
SILVerifier checks to ensure the flag is not invalidated.
Currently "escaping" is not used anywhere.
I am adding this to make it easy to determine if a SILFunction that is not inout
aliasable is captured. This is useful when emitting certain types of
diagnostics like I need to emit with move only.
Instead of setting the parent pointer to null, set the `lastInitializedBitfieldID` to -1.
This allows to keep the parent block information, even when an instruction is removed from it's list.
It makes no sense to operate on the block's instruction list without
also including SILBasicBlock.h anyway. Similarly, it doesn't make
sense to query the entry block without including SILFunction.h.
Now we can call these simple getters in critical-path loops assuming
they are as cheap as a load. I was avoiding calling these in critical
path code, which resulted in less readability and consistency across
the code base.
It's used to implement `InstructionSet` and `ValueSet`: sets of SILValues and SILInstructions.
Just like `BasicBlockSet` for basic blocks, the set is implemented by setting bits directly in SILNode.
This is super efficient because insertion and deletion to/from the set are basic bit operations.
The cost is an additional word in SILNode. But this is basically negligible: it just adds ~0.7% of memory used for SILInstructions.
In my experiments, I didn't see any relevant changes in memory consumption or compile time.
This subclass of SILArgument should be eliminated--it's not always a
phi, and whether it is a "phi argument" has nothing whatsoever to do
with the opcode. That is a property of a value's uses, not a property of the
value.
Until then, provide a logical and useful API within the type. This
often avoids the need to explicitly cast to a SILPhiArgument type and
avoids a lot of boilerplate in code that deals with phis.
Note: PhiOperand and PhiValue are improved abstractions on top of this
API. But the SILArgument-level API is still an important bridge
between SILArgument and other phi abstractions.
SROA and Mem2Reg now can leverage DIExpression -- op_fragment, more
specifically -- to generate correct debug info for optimized SIL. Some
important highlights:
- The new swift::salvageDebugInfo, similar to llvm::salvageDebugInfo,
tries to restore / transfer debug info from a deleted instruction.
Currently I only implemented this for store instruction whose
destination is an alloc_stack value.
- Since we now have source-variable-specific SIL location inside a
`debug_value` instruction (and its friends), this patch teaches
SILCloner and SILInliner to remap the debug scope there in addition
to debug scope of the instruction.
- DCE now does not remove `debug_value` instruction whose associating
with a function argument SSA value that is not used elsewhere. Since
that SSA value will not disappear so we should keep the debug info.
This is the initial version of a buildable SIL definition in libswift.
It defines an initial set of SIL classes, like Function, BasicBlock, Instruction, Argument, and a few instruction classes.
The interface between C++ and SIL is a bridging layer, implemented in C.
It contains all the required bridging data structures used to access various SIL data structures.
It's not needed anymore with delayed instruction deletion.
It was used for two purposes:
1. For analysis, which cache instructions, to avoid dangling instruction pointers
2. For passes, which maintain worklists of instructions, to remove a deleted instructions from the worklist. This is now done by checking SILInstruction::isDeleted().
When an instruction is "deleted" from the SIL, it is put into the SILModule::scheduledForDeletion list.
The instructions in this list are eventually deleted for real in SILModule::flushDeletedInsts(), which is called by the pass manager after each pass run.
In other words: instruction deletion is deferred to the end of a pass.
This avoids dangling instruction pointers within the run of a pass and in analysis caches.
Note that the analysis invalidation mechanism ensures that analysis caches are invalidated before flushDeletedInsts().
Instead, put the archetype->instrution map into SIlModule.
SILOpenedArchetypesTracker tried to maintain and reconstruct the mapping locally, e.g. during a use of SILBuilder.
Having a "global" map in SILModule makes the whole logic _much_ simpler.
I'm wondering why we didn't do this in the first place.
This requires that opened archetypes must be unique in a module - which makes sense. This was the case anyway, except for keypath accessors (which I fixed in the previous commit) and in some sil test files.
This removes the ambiguity when casting from a SingleValueInstruction to SILNode, which makes the code simpler. E.g. the "isRepresentativeSILNode" logic is not needed anymore.
Also, it reduces the size of the most used instruction class - SingleValueInstruction - by one pointer.
Conceptually, SILInstruction is still a SILNode. But implementation-wise SILNode is not a base class of SILInstruction anymore.
Only the two sub-classes of SILInstruction - SingleValueInstruction and NonSingleValueInstruction - inherit from SILNode. SingleValueInstruction's SILNode is embedded into a ValueBase and its relative offset in the class is the same as in NonSingleValueInstruction (see SILNodeOffsetChecker).
This makes it possible to cast from a SILInstruction to a SILNode without knowing which SILInstruction sub-class it is.
Casting to SILNode cannot be done implicitly, but only with an LLVM `cast` or with SILInstruction::asSILNode(). But this is a rare case anyway.
This removes the ambiguity when casting from a SingleValueInstruction to SILNode, which makes the code simpler. E.g. the "isRepresentativeSILNode" logic is not needed anymore.
Also, it reduces the size of the most used instruction class - SingleValueInstruction - by one pointer.
Conceptually, SILInstruction is still a SILNode. But implementation-wise SILNode is not a base class of SILInstruction anymore.
Only the two sub-classes of SILInstruction - SingleValueInstruction and NonSingleValueInstruction - inherit from SILNode. SingleValueInstruction's SILNode is embedded into a ValueBase and its relative offset in the class is the same as in NonSingleValueInstruction (see SILNodeOffsetChecker).
This makes it possible to cast from a SILInstruction to a SILNode without knowing which SILInstruction sub-class it is.
Casting to SILNode cannot be done implicitly, but only with an LLVM `cast` or with SILInstruction::asSILNode(). But this is a rare case anyway.
This makes it easier to understand conceptually why a ValueOwnershipKind with
Any ownership is invalid and also allowed me to explicitly document the lattice
that relates ownership constraints/value ownership kinds.
When merging many blocks to a single block (in the wrong order), instructions are getting moved over and over again.
This is quadratic and can result in very long compile times for large functions.
To fix this, always move the instruction to smaller block to the larger block.
rdar://problem/56268570
Specifically, I split it into 3 initial categories: IR, Utils, Verifier. I just
did this quickly, we can always split it more later if we want.
I followed the model that we use in SILOptimizer: ./lib/SIL/CMakeLists.txt vends
a macro (sil_register_sources) to the sub-folders that register the sources of
the subdirectory with a global state variable that ./lib/SIL/CMakeLists.txt
defines. Then after including those subdirs, the parent cmake declares the SIL
library. So the output is the same, but we have the flexibility of having
subdirectories to categorize source files.
