This feature had been marked as `TODO` in the `tensor.splat`
documentation for a while. This MR includes:
- Support for dynamically shaped tensors in the return type of
`tensor.splat` with the syntax suggested in the `TODO` comment.
- Updated op documentation.
- Bufferization support.
- Updates in op folders affected by the new feature.
- Unit tests for valid/invalid syntax, valid/invalid folding, and
lowering through bufferization.
- Additional op builders resembling those available in `tensor.empty`.
When the concatenated dim is statically sized but the inputs are
dynamically sized, reifyResultShapes must return the static shape. Fixes
the implementation of the interface for tensor.concat in such cases.
The folder for `tensor.extract` is not operating correctly when it is
consuming the result of a `tensor.from_elements` operation.
The existing unit test named `@extract_from_tensor.from_elements_3d` in
`mlir/test/Dialect/Tensor/canonicalize.mlir` seems an attempt to stress
this code. However, this unit tests creates a `tensor.from_elements` op
exclusively from constants, which gets folded away into a single
constant tensor. Therefore, the buggy code was never executed in unit
tests.
I have added a new unit test named
`@extract_from_tensor.from_elements_variable_3d` that makes sure the
`tensor.from_elements` op is not folded away by having its input
operands come directly from function arguments. The original folder code
would have made this test fail.
This bug was notably affecting the lowering of the `tosa.pad` op in the
`tosa-to-tensor` pass, where the generated code is likely to contain a
`tensor.from_elements` + `tensor.extract` op sequence.
Op verifiers should verify only local properties of an op. The dynamic
sizes of a `tensor.generate` op should not be verified. Dynamic sizes
that have a negative constant value should not prevent the
`tensor.generate` op from verifying.
Also share some code between the `tensor.empty` and `tensor.generate`
"dynamic dim -> static dim" canonicalization patterns.
Remove the `invalid-canonicalize.mlir` file and move the test case to
`canonicalize.mlir`. Canonicalization no longer produces IR that does
not verify (and leaves the op as is).
Without folding the result of the initial tensor.dim, the
ReifyResultShapes implementation would be incorrect because it would
return a dynamic shape for a static result shape.
This adds an operation for concatenating ranked tensors along a static
dimension, as well as a decomposition mirroring the existing lowering
from TOSA to Tensor. This offers a convergence point for "input" like
dialects that include various lowerings for concatenation operations,
easing later analysis. In the future, this op can implement the
necessary interfaces for tiling, as well as potentially add conversions
to some kind of linalg and/or memref counterpart.
This patch adds the op, the decomposition, and some basic
folding/canonicalization. Replacing lowerings with the op (such as the
TOSA lowering) will come as a follow up.
See
https://discourse.llvm.org/t/rfc-tensor-add-a-tensor-concatenate-operation/74858
This commit fixes a crash of the canonicalizer when there are slice ops
with offset/size SSA values that have a negative constant value. Such
ops are invalid if they are reachable and their offsets/sizes should not
be folded to static integer values. (But such ops may appear in
non-reachable block.)
This commit fixes#71150.
In #71153, the `memref.subview` canonicalizer crashes due to a negative
`size` being passed as an operand. During `SubViewOp::verify` this
negative `size` is not yet detectable since it is dynamic and only
available after constant folding, which happens during the
canonicalization passes. As discussed in
<https://discourse.llvm.org/t/rfc-more-opfoldresult-and-mixed-indices-in-ops-that-deal-with-shaped-values/72510>,
the verifier should not be extended as it should "only verify local
aspects of an operation".
This patch fixes#71153 by not folding in aforementioned situation.
Also, this patch adds a basic offset and size check in the
`OffsetSizeAndStrideOpInterface` verifier.
Note: only `offset` and `size` are checked because `stride` is allowed
to be negative
(54d81e49e3).
This patch fixes two checks where a `SmallBitVector` containing the
potential dropped dims of a SubView/ExtractSlice operation was queried
via `empty()` instead of `none()`.
Fixes https://github.com/llvm/llvm-project/issues/60656.
This patch implements a basic fold for various reshape/resize tensor
operations. Specifically, the folding removes tensor reshape/resize ops
when they are applied to a constant tensor. For example, the following
function:
```mlir
func.func @main(%dest : tensor<8x16x8x32xf32>) -> tensor<8x16x8x32xf32> {
%cst = arith.constant dense<1.000000e-01> : tensor<64x128xf32>
%0 = tensor.pack %cst outer_dims_perm = [1, 0] inner_dims_pos = [0, 1]
inner_tiles = [8, 32] into %dest : tensor<64x128xf32> -> tensor<8x16x8x32xf32>
return %0 : tensor<8x16x8x32xf32>
}
```
will be changed into the following with `mlir-opt -canonicalize`:
```mlir
func.func @main(%arg0: tensor<8x16x8x32xf32>) -> tensor<8x16x8x32xf32> {
%cst = arith.constant dense<1.000000e-01> : tensor<8x16x8x32xf32>
return %cst : tensor<8x16x8x32xf32>
}
```
As a side-note, this patch is essentially an extension of
f79f430d4b.
The destination operand of the `tensor.unpack` operation is only needed
to carry shape information. So if the producer of the destination
operand implements the `DestinationStyleOpInterface`, then fold it into
the `tensor.unpack` operation by replacing the destination operand with
the destination for the source.
The majority of subset ops operate on hyperrectangular subsets. This
commit adds a new optional interface method
(`getAccessedHyperrectangularSlice`) that can be implemented by such
subset ops. If implemented, the other `operatesOn...` interface methods
of the `SubsetOpInterface` do not have to be implemented anymore.
The comparison logic for hyperrectangular subsets (is
disjoint/equivalent) is implemented with `ValueBoundsOpInterface`. This
makes the subset hoisting more powerful: simple cases where two
different SSA values always have the same runtime value can now be
supported.
There is currently an op interface for subset insertion ops
(`SubsetInsertionOpInterface`), but not for subset extraction ops. This
commit adds `SubsetExtractionOpInterface` to `mlir/Interfaces`, as well
as a common dependent op interface: `SubsetOpInterface`.
- `SubsetOpInterface` is for ops that operate on tensor subsets. It
provides interface methods to check if two subset ops operate on
equivalent or disjoint subsets. Ops that implement this interface must
implement either `SubsetExtractionOpInterface` or
`SubsetInsertionOpInterface`.
- `SubsetExtractionOpInterface` is for ops that extract from a tensor at
a subset. E.g., `tensor.extract_slice`, `tensor.gather`,
`vector.transfer_read`. Current implemented only on
`tensor.extract_slice`.
- `SubsetInsertionOpInterface` is for ops that insert into a destination
tensor at a subset. E.g., `tensor.insert_slice`,
`tensor.parallel_insert_slice`, `tensor.scatter`,
`vector.transfer_write`. Currently only implemented on
`tensor.insert_slice`, `tensor.parallel_insert_slice`.
Other changes:
- Rename `SubsetInsertionOpInterface.td` to `SubsetOpInterface.td`.
- Add helper functions to `ValueBoundsOpInterface.cpp` for checking
whether two slices are disjoint.
The new interfaces will be utilized by a new "loop-invariant subset
hoisting"
transformation. (This new transform is roughly
what `Linalg/Transforms/SubsetHoisting.cpp` is doing, but in a generic
and interface-driven way.)
`SubsetInsertionOpInterface` is an interface for ops that insert into a
destination tensor at a subset. It is currently used by the
bufferization framework to support efficient
`tensor.extract_slice/insert_slice` bufferization and to drive "empty
tensor elimination".
This commit moves the interface to `mlir/Interfaces`. This is in
preparation of adding a new "loop-invariant subset hoisting"
transformation to
`mlir/Transforms/Utils/LoopInvariantCodeMotionUtils.cpp`, which will
utilize `SubsetInsertionOpInterface`. (This new transform is roughly
what `Linalg/Transforms/SubsetHoisting.cpp` is doing, but in a generic
and interface-driven way.)
Two `OpOperand`s are the same if they belong to the same owner and have
the same operand number. There are currently no comparison operators
defined on `OpOperand` and we work around this in multiple places by
comparing pointers.
Note: `OpOperand`s are stored in an op, so it is valid to compare their
pointers to determine if they are the same operand. E.g.,
`getOperandNumber` is also implemented via pointer arithmetics.
* `tensor.collapse_shape` may bufferize to a memory read because the op
may have to reallocate the source buffer.
* `tensor.reshape` should not use `bufferization.clone` for
reallocation. This op has requirements wrt. the order of buffer
writes/reads. Use `memref.alloc` and `memref.copy` instead. Also fix a
bug where the memory space of the source buffer was not propagated to
the reallocated buffer.
`BufferizableOpInterface::bufferizesToAllocation` is queried when
forming equivalence sets during bufferization. It is not really needed
for ops like `tensor.empty` which do not have tensor operands, but it
should be added for consistency.
This change should have been part of #68080. No test is added because
the return value of this function is irrelevant for ops without tensor
operands. (However, this function acts as a form documentation,
describing the bufferization semantics of the op.)
The TableGen code generator now generates C++ code that returns a single
`OpOperand &` for `get...Mutable` of operands that are not variadic and
not optional. `OpOperand::set`/`assign` can be used to set a value (same
as `MutableOperandRange::assign`). This is safer than
`MutableOperandRange` because only single values (and no longer
`ValueRange`) can be assigned.
E.g.:
```
// Assignment of multiple values to non-variadic operand.
// Before: Compiles, but produces invalid op.
// After: Compilation error.
extractSliceOp.getSourceMutable().assign({v1, v2});
```
Make `tensor.empty` bufferizable, so that the
`-empty-tensor-to-alloc-tensor` pass becomes optional. This makes the
bufferization easier to use. `tensor.empty` used to be non-bufferizable,
so that there two separate ops, one that can be optimized away
(`tensor.empty`) and one that is guaranteed to bufferize to an
allocation (`bufferization.alloc_tensor`). With the recent improvements
of "empty tensor elimination" this is no longer needed and
`bufferization.alloc_tensor` can be phased out.
`scf::ForallOp` has `shared_outs` tensor operands which are used to
insert partial results into in the parallel terminator. The
`scf::ForallOp` returns one tensor for each `shared_out` which then
contains the combined result from all threads. Since the parallel
terminator cannot change the shape of the `shared_out`, ForallOp is a
`DestinationStyleOp` and this patch implements the interface by
declaring the `outputs` operands as `inits` in the language of the DPS
interface.
For this change to work, we need to add an exception to the Pattern that
folds `tensor.cast` Ops into DPS Ops because `scf::Forall` needs special
handling of its `BlockArgument` Type during this folding.
Bufferization of tensor.reshape generates a memref.reshape operation.
memref.reshape requires the source memref to have an identity layout.
The bufferization process may result in the source memref having a
non-identity layout, resulting in a verification failure.
This change causes the bufferization interface for tensor.reshape to
copy the source memref to a new buffer when the source has a
non-identity layout.
This commit removes the deallocation capabilities of
one-shot-bufferization. One-shot-bufferization should never deallocate
any memrefs as this should be entirely handled by the
ownership-based-buffer-deallocation pass going forward. This means the
`allow-return-allocs` pass option will default to true now,
`create-deallocs` defaults to false and they, as well as the escape
attribute indicating whether a memref escapes the current region, will
be removed. A new `allow-return-allocs-from-loops` option is added as a
temporary workaround for some bufferization limitations.
`operator[]` returns `OpOperand &` instead of `Value`.
* This allows users to get OpOperands by name instead of "magic" number.
E.g., `extractSliceOp->getOpOperand(0)` can be written as
`extractSliceOp.getSourceMutable()[0]`.
* `OperandRange` provides a read-only API to operands: `operator[]`
returns `Value`. `MutableOperandRange` now provides a mutable API:
`operator[]` returns `OpOperand &`, which can be used to set operands.
Note: The TableGen code generator could be changed to return `OpOperand
&` (instead of `MutableOperandRange`) for non-variadic and non-optional
arguments in a subsequent change. Then the `[0]` part in the above
example would no longer be necessary.
This commit generalizes empty tensor elimination to operate on subset
ops. No new test cases are added because all current subset ops were
already supported previously. From this perspective, this change is NFC.
A new interface method (and a helper method) are added to
`SubsetInsertionOpInterface` to build the subset of the destination
tensor.
This commit generalizes the special
tensor.extract_slice/tensor.insert_slice bufferization rules to tensor
subset ops.
Ops that insert a tensor into a tensor at a specified subset (e.g.,
tensor.insert_slice, tensor.scatter) can implement the
`SubsetInsertionOpInterface`.
Apart from adding a new op interface (extending the API), this change is
NFC. The only ops that currently implement the new interface are
tensor.insert_slice and tensor.parallel_insert_slice, and those ops were
are supported by One-Shot Bufferize.
This is the first commit in a series with the goal to rework the
BufferDeallocation pass. Currently, this pass heavily relies on copies
to perform correct deallocations, which leads to very slow code and
potentially high memory usage. Additionally, there are unsupported cases
such as returning memrefs which this series of commits aims to add
support for as well.
This first commit removes the deallocation capabilities of
one-shot-bufferization.One-shot-bufferization should never deallocate any
memrefs as this should be entirely handled by the buffer-deallocation pass
going forward. This means the allow-return-allocs pass option will
default to true now, create-deallocs defaults to false and they, as well
as the escape attribute indicating whether a memref escapes the current region,
will be removed.
The documentation should w.r.t. these pass option changes should also be
updated in this commit.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D156662
This patch addresses a crash that occurs when negative dynamic sizes are
provided in tensor.emptyOp by adding a check to ensure that dynamic
sizes are non-negative.
Fixes#64064
This enables canonicalization to fold away unnecessary tensor.dim ops
which in turn enables folding away of other operations, as can be seen
in conv_tensors_dynamic where affine.min operations were folded away.
This commit provides a default implementation for all ops that implement
the `DestinationStyleOpInterface`. Result values of such ops are tied to
operand, and those have the same type.
Fixes an issue where `isCastLikeExtractSliceOp` did not account for the fact
that `tensor.extract_slice` may drop non-unit dimensions. This change makes the
utility function behave inline with its name/description. The only user of this
function is in the `FindPayloadReplacementOpInterface` for the
`tensor::ExtractSliceOp`. This can potentially cause downstream projects to have
more "listener could not find replacement op" errors when interpreting Transform
IR, but the behavior is inline with the documented conservative behavior of the
Transform dialect's TrackingListener.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D158635
`reifyResultShapes` should return an `Attribute` if and only if the respective dimension is static.
This fixes#64256.
Differential Revision: https://reviews.llvm.org/D158166
We are able to fuse the pack op only if inner tiles are not tiled or
they are fully used. Otherwise, it could generate a sequence of
non-trivial ops.
Differential Revision: https://reviews.llvm.org/D157932
`getBufferType` computes the bufferized type of an SSA value without bufferizing any IR. This is useful for predicting the bufferized type of iter_args of a loop.
To avoid endless recursion (e.g., in the case of "scf.for", the type of the iter_arg depends on the type of init_arg and the type of the yielded value; the type of the yielded value depends on the type of the iter_arg again), `fixedTypes` was used to fall back to "fixed" type. A simpler way is to maintain an "invocation stack". `getBufferType` implementations can then inspect the invocation stack to detect repetitive computations (typically when computing the bufferized type of a block argument).
Also improve error messages in case of inconsistent memory spaces inside of a loop.
Differential Revision: https://reviews.llvm.org/D158060
This revision is needed to support bufferization of `cf.br`/`cf.cond_br`. It will also be useful for better analysis of loop ops.
This revision generalizes `getAliasingOpResults` to `getAliasingValues`. An OpOperand can now not only alias with OpResults but also with BlockArguments. In the case of `cf.br` (will be added in a later revision): a `cf.br` operand will alias with the corresponding argument of the destination block.
If an op does not implement the `BufferizableOpInterface`, the analysis in conservative. It previously assumed that an OpOperand may alias with each OpResult. It now assumes that an OpOperand may alias with each OpResult and each BlockArgument of the entry block.
Differential Revision: https://reviews.llvm.org/D157957
* Move `foldDynamicIndexList` to `DialectUtils` and simplify function.
* Move `OpWithOffsetSizesAndStridesConstantArgumentFolder` to `ViewLikeInterface` and add documentation.
Differential Revision: https://reviews.llvm.org/D156581
In https://reviews.llvm.org/D151611, a check was added to the tensor verifier to
emit an error on negative tensor dimensions. This check allowed for dynamic
dimensions, hence negative dimensions were still able to get through the verifier.
This is a problem in situations such as #60558, where the dynamic dimension is
converted to a static (and possibly negative) dimension by another pass in the
compiler. This patch fixes that by doing another check during the
`StaticTensorGenerate` conversion, and return a failure if the dimension is
negative.
As a side-note, I have to admit that I do not know why returning a failure in
`StaticTensorGenerate` gives a nice "tensor dimensions must be non-negative"
error. I suspect that the verifier runs again when `return failure()` is called,
but I am not sure.
Fixes#60558.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D155728
Remove patterns that fold tensor subset ops into vector transfer ops from the vector dialect. These patterns already exist in the tensor dialect.
Differential Revision: https://reviews.llvm.org/D154932
