This is a futile attempt to discourage future use of getType() by
giving it a "scary" name.
We want people to use getInterfaceType() like with the other decl kinds.
Introduce the notion of "semantic result parameter". Handle differentiation of inouts via semantic result parameter abstraction. Do not consider non-wrt semantic result parameters as semantic results
Fixes#67174
All value ownership instructions (begin_borrow, copy_value, move_value)
need to be handled in the same way. Factored out a common function for
that handling in preparation to visit move_value instructions.
Reformatting everything now that we have `llvm` namespaces. I've
separated this from the main commit to help manage merge-conflicts and
for making it a bit easier to read the mega-patch.
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.
Adjoint buffers of projections (e.g. obtained via begin_access) are same as adjoint buffer of underlying struct value. As a result, when propagating adjoint values to pullback successor blocks we tend to produce lots of identical copies (essentially for every struct access and in every basic block) of adjoint buffers.
These copy_addrs instructions are then lowered down to plain loads and stores and while the redundant copies are usually optimized away by subsequent optimization passes, presence of such copies leads to elevated memory consumption and compilation time as one needs to track liveness of these values being copied.
Track the values being propagated and simply do not generate extra copies if the same value was already propagated.
One step towards #61773
Linear maps are captured in vjp routine via callee-guaranteed partial apply and are passed as @owned references to the enclosing pullback that finally consumes them. Necessary retains are inserted by a partial apply forwarder.
However, this is not the case when the function being differentiated contains loops as heap-allocated context is used and bare pointer is captured by the pullback partial apply. As a result, partial apply forwarder does not retain the linear maps that are owned by a heap-allocated context, however, they are still treated as @owned references and therefore are released in the pullback after the first call. As a result, subsequent pullback calls release linear maps and we'd end with possible use-after-free.
Ensure we retain values when we load values from the context.
Reproducible only when:
* Loops (so, heap-allocated context)
* Pullbacks of thick functions (so context is non-zero)
* Multiple pullback calls
* Some cleanup while there
Fixes#64257
This essentially passes the members of a linear map tuple as individual arguments. It yields few nice simplifications:
* No linear map tuples at all for getters / setters
* No tuple formation / deconstruction around pullbacks
* Pullbacks with loops still use heap-allocated tuples
The changes are intentionally were made close to the original implementation w/o possible simplifications to ease the review
Fixes#63207, supersedes #63379 (and fixes#63234)
Single input variable might yield multiple adjoint buffers if control flow is involved. Therefore we cannot simply transfer debug info from the input variable: it will be invalid as we will end with multiple locations for a single "source" variable, and, even worse, might end with conflicting debug info as different buffers might be optimized
differently.
We do:
- Drop input argument number. This must be unique and we're not
- Correct variable name
`getValue` -> `value`
`getValueOr` -> `value_or`
`hasValue` -> `has_value`
`map` -> `transform`
The old API will be deprecated in the rebranch.
To avoid merge conflicts, use the new API already in the main branch.
rdar://102362022
Apparently #60467 changed the semantics of store_borrow as it started to produce a value. This change was not documented in SIL spec and not all places were updated to new semantics.
Now the adjoint of store_borrow should be generated for the value of instruction itself, not the destination address
We had two notions of canonical types, one is the structural property
where it doesn't contain sugared types, the other one where it does
not contain reducible type parameters with respect to a generic
signature.
Rename the second one to a 'reduced type'.
Andy some time ago already created the new API but didn't go through and update
the old occurences. I did that in this PR and then deprecated the old API. The
tree is clean, so I could just remove it, but I decided to be nicer to
downstream people by deprecating it first.
This is yet another optional-related bug. It could be triggered if `unchecked_take_enum_data_addr` is inside a loop. In such case the accumulated value of adjoint buffer would be re-used in the subsequent loop iterations producing wrong results.
Fixes#58353 (SR16094)
Apparently, the parameter index calculation in #37861 was not always correct in presence of other pullback parameters (e.g. captures and non-differentiated args). Collect all inout parameters and collect inout parameter adjoints correctly.
Resolves SR-15891
* Attempt at a patch for SR-14053 and SR-14218, based on Dan Zheng's initial fix.
* Incorporated Dan's cleanup suggestions.
* Converting a bool SmallVector to a SmallBitVector.
* Testing if Windows issues with this test are due to runtime support.
* Simplifying test case.
Fixes:
[[SR-14290]: SIL verification fails when differentiating a function of [[Double]]](rdar://74876596)
[[SR-14298]: [AutoDiff] SIL verification failed: non-contiguous lexical scope at -Onone](rdar://75032457)
Refactor SILGen's ApplyOptions into an OptionSet, add a
DoesNotAwait flag to go with DoesNotThrow, and sink it
all down into SILInstruction.h.
Then, replace the isNonThrowing() flag in ApplyInst and
BeginApplyInst with getApplyOptions(), and plumb it
through to TryApplyInst as well.
Set the flag when SILGen emits a sync call to a reasync
function.
When set, this disables the SIL verifier check against
calling async functions from sync functions.
Finally, this allows us to add end-to-end tests for
rdar://problem/71098795.
- Properly clone and use debug scopes for all instructions in pullback functions.
- Emit `debug_value` instructions for adjoint values.
- Add debug locations and variable info to adjoint buffer allocations.
- Add `TangentBuilder` (a `SILBuilder` subclass) to unify and simplify special emitter utilities for tangent vector code generation. More simplifications to come.
Pullback variable inspection example:
```console
(lldb) n
Process 50984 stopped
* thread #1, queue = 'com.apple.main-thread', stop reason = step over
frame #0: 0x0000000100003497 main`pullback of foo(x=0) at main.swift:12:11
9 import _Differentiation
10
11 func foo(_ x: Float) -> Float {
-> 12 let y = sin(x)
13 let z = cos(y)
14 let k = tanh(z) + cos(y)
15 return k
Target 0: (main) stopped.
(lldb) fr v
(Float) x = 0
(Float) k = 1
(Float) z = 0.495846391
(Float) y = -0.689988375
```
Resolves rdar://68616528 / SR-13535.
In derivatives of loops, no longer allocate boxes for indirect case payloads. Instead, use a custom pullback context in the runtime which contains a bump-pointer allocator.
When a function contains a differentiated loop, the closure context is a `Builtin.NativeObject`, which contains a `swift::AutoDiffLinearMapContext` and a tail-allocated top-level linear map struct (which represents the linear map struct that was previously directly partial-applied into the pullback). In branching trace enums, the payloads of previously indirect cases will be allocated by `swift::AutoDiffLinearMapContext::allocate` and stored as a `Builtin.RawPointer`.
`SILBuilder::createAllocStack` expects a debug variable when the location is a `VarDecl`. Since we are in pullbacks, there's no debug variables so we pass an empty one.
General support for debug-info-in-pullbacks will be added as part of SR-13535.
Also add a negative test for SR-13866.
Resolves SR-13865.
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.
Fix SIL pullback function type calculation: remap original `unowned` results to
the correct pullback parameter convention depending on the `TangentVector` type
lowering.
Make `PullbackCloner` visitors for the following instructions check and handle
tangent value categories: `struct`, `struct_extract`, `tuple`, `destructure_tuple`.
Add differentiation tests for `Optional` struct and class stored properties,
exercising the instruction visitors above.
Resolves SR-13430.
