This fixes an issue in One-Shot Bufferize that could lead to missing buffer copies in the future. This bug can currently not be triggered because of the order in which ops are analyzed (always bottom-to-top). However, if we consider different traversal orders for the analysis in the future, this bug can cause subtle issues that are difficult to debug.
Example:
```
%0 = ...
%1 = tensor.insert ... into %0
%2 = tensor.extract_slice %0
tensor.extract %2[...]
```
In case of a top-to-bottom analysis of the above IR, the `tensor.insert` is analyzed before the `tensor.extract_slice`. In that case, the `tensor.insert` will bufferize in-place because %2 is not yet known to become an alias of %0 (and therefore causing a conflict).
With this change, the `tensor.insert` will bufferize out-of-place, regardless of the traversal order.
Differential Revision: https://reviews.llvm.org/D135049
This fixes a bug where a required buffer copy was not inserted.
Not only written aliases, but also read aliases should be taken into account when computing common enclosing repetitive regions. Furthermore, for writing ops, it does not matter where the destination tensor is defined, but where the op itself is located.
Differential Revision: https://reviews.llvm.org/D135420
Inserting a tensor into an equivalent tensor is a no-op after bufferization. No alloc is needed.
Differential Revision: https://reviews.llvm.org/D132662
Bufferization already makes the assumption that buffers pass function
boundaries in the strided form and uses the corresponding affine map layouts.
Switch it to use the recently introduced strided layout instead to avoid
unnecessary casts when bufferizing further operations to the memref dialect
counterparts that now largely rely on the strided layout attribute.
Depends On D133947
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D133951
This method allows to declare regions as "repetitive" even if the parent op does not implement the RegionBranchOpInterface.
This is needed to support loop-like ops that have parallel semantics but do not branch between regions.
Differential Revision: https://reviews.llvm.org/D133113
The patch introduces the required changes to update the pass declarations and definitions to use the new autogenerated files and allow dropping the old infrastructure.
Reviewed By: mehdi_amini, rriddle
Differential Review: https://reviews.llvm.org/D132838
The patch introduces the required changes to update the pass declarations and definitions to use the new autogenerated files and allow dropping the old infrastructure.
Reviewed By: mehdi_amini, rriddle
Differential Review: https://reviews.llvm.org/D132838
Even though iter_arg and init_arg of an scf.for loop may have the same tensor type, their bufferized memref types are not necessarily equal. It is sometimes necessary to insert a cast in case of differing layout maps.
Differential Revision: https://reviews.llvm.org/D132860
This change generalizes getBufferType. This function can be used to predict the buffer type of any tensor value (not just BlockArguments) without changing any IR. It also subsumes getMemorySpace. This is useful for loop bufferization, where the precise buffer type of an iter_arg cannot be known without examining the loop body.
Differential Revision: https://reviews.llvm.org/D132859
It's only used from there, and this lets us remove the dependency from Analysis
to the Arith dialect.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D132928
tensor.pad is lowered to tensor.generate + tensor.insert_slice during bufferization. For best performance with constant padding values, users should vectorize the IR before bufferizing it.
This change also relaxes tje restriction that no new ops that bufferize to a memory write should be added during bufferization. Since bufferization has been split into two steps a while ago (tensor copy insertion + bufferization), it is reasonable to allow this now.
Differential Revision: https://reviews.llvm.org/D132355
bufferization.to_memref ops are not supported in One-Shot Analysis. They often trigger a failed assertion that can be confusing. Instead, scan for to_memref ops before running the analysis and immediately abort with a proper error message.
Differential Revision: https://reviews.llvm.org/D132027
AllocTensorElimination does currently not support chains where the type is
changing. AllocTensorElimination used to generate invalid IR for such
inputs. With this commit, AllocTensorElimination does no longer apply to
such inputs. (It can be extended to support such IR if needed.)
Differential Revision: https://reviews.llvm.org/D131880
This reland includes changes to the Python bindings.
Switch variadic operand and result segment size attributes to use the
dense i32 array. Dense integer arrays were introduced primarily to
represent index lists. They are a better fit for segment sizes than
dense elements attrs.
Depends on D131801
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D131803
Using a loop init_arg inside of the loop is not supported. This change adds a pre-processing pass that resolves such IR with copies.
Differential Revision: https://reviews.llvm.org/D131689
Switch variadic operand and result segment size attributes to use the
dense i32 array. Dense integer arrays were introduced primarily to
represent index lists. They are a better fit for segment sizes than
dense elements attrs.
Depends on D131738
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D131702
In the Transform dialect extensions, provide the separate mechanism to
declare dependent dialects (the dialects the transform IR depends on)
and the generated dialects (the dialects the payload IR may be
transformed into). This allows the Transform dialect clients that are
only constructing the transform IR to avoid loading the dialects
relevant for the payload IR along with the Transform dialect itself,
thus decreasing the build/link time.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D130289
This op used to belong to the sparse dialect, but there are use cases for dense bufferization as well. (E.g., when a tensor alloc is returned from a function and should be deallocated at the call site.) This change moves the op to the bufferization dialect, which now has an `alloc_tensor` and a `dealloc_tensor` op.
Differential Revision: https://reviews.llvm.org/D129985
bufferization.writable is used in most cases instead. All remaining test cases are updated. Some code that is no longer needed is deleted.
Differential Revision: https://reviews.llvm.org/D129739
This change removes the partial bufferization passes from the sparse compilation pipeline and replaces them with One-Shot Bufferize. One-Shot Analysis (and TensorCopyInsertion) is used to resolve all out-of-place bufferizations, dense and sparse. Dense ops are then bufferized with BufferizableOpInterface. Sparse ops are still bufferized in the Sparsification pass.
Details:
* Dense allocations are automatically deallocated, unless they are yielded from a block. (In that case the alloc would leak.) All test cases are modified accordingly. E.g., some funcs now have an "out" tensor argument that is returned from the function. (That way, the allocation happens at the call site.)
* Sparse allocations are *not* automatically deallocated. They must be "released" manually. (No change, this will be addressed in a future change.)
* Sparse tensor copies are not supported yet. (Future change)
* Sparsification no longer has to consider inplacability. If necessary, allocations and/or copies are inserted during TensorCopyInsertion. All tensors are inplaceable by the time Sparsification is running. Instead of marking a tensor as "not inplaceable", it can be marked as "not writable", which will trigger an allocation and/or copy during TensorCopyInsertion.
Differential Revision: https://reviews.llvm.org/D129356
This is a partial revert of D128615.
to_memref(to_tensor(x)) always be folded to x. But to_tensor(to_memref(x)) cannot be folded in the general case because writes to the intermediary memref may go unnoticed.
Differential Revision: https://reviews.llvm.org/D129354
The `unknownTypeConversion` bufferization option (enum) is now a type converter function option. Some logic of `getMemRefType` is now handled by that function.
This change makes type conversion more controllable. Previously, there were only two options when generating memref types for non-bufferizable ops: Static identity layout or fully dynamic layout. With this change, users of One-Shot Bufferize can provide a function with custom logic.
Differential Revision: https://reviews.llvm.org/D129273
The result shape of a rank-reducing subview cannot be inferred in the general case. Just the result rank is not enough. The only thing that we can infer is the layout map.
This change also improves the bufferization patterns of tensor.extract_slice and tensor.insert_slice to fully support rank-reducing operations.
Differential Revision: https://reviews.llvm.org/D129144
This change updates all remaining bufferization patterns (except for scf.while) and the remaining bufferization infrastructure to infer the memory space whenever possible instead of falling back to "0". (If a default memory space is set in the bufferization options, we still fall back to that value if the memory space could not be inferred.)
Differential Revision: https://reviews.llvm.org/D128423
Add a failure return value and bufferization options argument. This is to keep a subsequent change smaller.
Differential Revision: https://reviews.llvm.org/D128278
This is useful because the result type of an op can sometimes be inferred from its body (e.g., `scf.if`). This will be utilized in subsequent changes.
Also introduces a new `getBufferType` interface method on BufferizableOpInterface. This method is useful for computing a bufferized block argument type with respect to OpOperand types of the parent op.
Differential Revision: https://reviews.llvm.org/D128420
