Introduce a new transform operation to replace `tensor.empty` with
`alloc_tensor` operations. The operation is a pass-through if the target
operation is already a `alloc_tensor`; otherwise, it expects a
`tensor.empty` as a target. Currently, it does not return any results.
The operation is expected to run before `one_shot_bufferize` as
`one_shot_bufferize` rejects `tensor.empty`.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D140026
External functions have no body, so they cannot be analyzed. Assume conservatively that each tensor bbArg may be aliasing with each tensor result. Furthermore, assume that each function arg is read and written-to after bufferization. This default behavior can be controlled with `bufferization.access` (similar to `bufferization.memory_layout`) in test cases.
Also fix a bug in the dialect attribute verifier, which did not run for region argument attributes.
Differential Revision: https://reviews.llvm.org/D139517
While it is unlikely to matter in practice, there is no reason
for this value to be larger than it should be. 64 bytes is the
size of a cache line in most machines, and we can fit a full
512-bit vector in it.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D139434
TensorCopyInsertion should not have been exposed as a pass. This was a flaw in the original design. It is a preparation step for bufferization and certain transforms (that would otherwise be legal) are illegal between TensorCopyInsertion and actual rewrite to MemRef ops. Therefore, even if broken down as two separate steps internally, they should be exposed as a single pass.
This change affects the sparse compiler, which uses `TensorCopyInsertionPass`. A new `SparsificationAndBufferizationPass` is added to replace all passes in the sparse tensor pipeline from `TensorCopyInsertionPass` until the actual bufferization (rewrite to memref/non-tensor). It is generally unsafe to run arbitrary passes in-between, in particular passes that hoist tensor ops out of loops or change SSA use-def chains along tensor ops.
Differential Revision: https://reviews.llvm.org/D138915
The previous error message was confusing. Also improve code documentation and some minor code cleanups.
Differential Revision: https://reviews.llvm.org/D138902
Adds RegionBranchOpInterface for AffineIf Op and tests it
using buffer deallocation pass.
Reviewed By: bondhugula
Differential Revision: https://reviews.llvm.org/D130962
MemRef has been accepting a general Attribute as memory space for
a long time. This commits updates bufferization side to catch up,
which allows downstream users to plugin customized symbolic memory
space. This also eliminates quite a few `getMemorySpaceAsInt`
calls, which is deprecated.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D138330
Expose `function-boundary-type-conversion` in `OneShotBufferizeOp`. To
reuse options between passes and transform operations, create a
`BufferizationEnums.td`.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D137833
tensor.empty op elimination is an optimization that brings IR in a more bufferization-friendly form. E.g.:
```
%0 = tensor.empty()
%1 = linalg.fill(%cst, %0) {inplace = [true]}
%2 = tensor.insert_slice %1 into %t[10][20][1]
```
Is rewritten to:
```
%0 = tensor.extract_slice %t[10][20][1]
%1 = linalg.fill(%cst, %0) {inplace = [true]}
%2 = tensor.insert_slice %1 into %t[10][20][1]
```
This optimization used to operate on bufferization.alloc_tensor ops. This is not correct because the documentation of bufferization.alloc_tensor says that it always bufferizes to an allocation. Instead, this optimization should operate on tensor.empty ops, which can then be lowered to bufferization.alloc_tensor ops (if they don't get eliminated).
Differential Revision: https://reviews.llvm.org/D137162
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
Many tests wrap the piece of the IR related to the transform dialect
into `transform.with_pdl_patterns` without actually using PDL patterns
inside. Some of these are leftovers from migration to `structured.match`
and some others are cargo cult, both are useless and pollute the tests.
Reviewed By: guraypp
Differential Revision: https://reviews.llvm.org/D135661
Use the recently introduced TransformTypeInterface instead of hardcoding
the PDLOperationType. This will allow the operations to use more
specific transform types to express pre/post-conditions in the future.
It requires the syntax and Python op construction API to be updated.
Dialect extensions will be switched separately.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D135584
tensor.empty/linalg.init_tensor produces an uninititalized tensor that can be used as a destination operand for destination-style ops (ops that implement `DestinationStyleOpInterface`).
This change makes it possible to implement `TilingInterface` for non-destination-style ops without depending on the Linalg dialect.
RFC: https://discourse.llvm.org/t/rfc-add-tensor-from-shape-operation/65101
Differential Revision: https://reviews.llvm.org/D135129
One of the test cases matched IR from a subsequent test case. For this reason, the test case appeared to pass while it is actually broken.
This change does not fix the test case itself. It will be fixed when we overhaul the buffer deallocation implementation. (The memory leak in this test case is an edge case.)
Differential Revision: https://reviews.llvm.org/D135046
All relevant operations have been switched to primarily use the strided
layout, but still support the affine map layout. Update the relevant
tests to use the strided format instead for compatibility with how ops
now print by default.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D134045
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
The three following ops in the memref dialect: transpose, expand_shape,
collapse_shape, have been originally designed to operate on memrefs with
strided layouts but had to go through the affine map representation as the type
did not support anything else. Make these ops produce memref values with
StridedLayoutAttr instead now that it is available.
Depends On D133938
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D133947
Memref subview operation has been initially designed to work on memrefs with
strided layouts only and has never supported anything else. Port it to use the
recently added StridedLayoutAttr instead of extracting the strided from
implicitly from affine maps.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D133938
Introduce a new attribute to represent the strided memref layout. Strided
layouts are omnipresent in code generation flows and are the only kind of
layouts produced and supported by a half of operation in the memref dialect
(view-related, shape-related). However, they are internally represented as
affine maps that require a somewhat fragile extraction of the strides from the
linear form that also comes with an overhead. Furthermore, textual
representation of strided layouts as affine maps is difficult to read: compare
`affine_map<(d0, d1, d2)[s0, s1] -> (d0*32 + d1*s0 + s1 + d2)>` with
`strides: [32, ?, 1], offset: ?`. While a rudimentary support for parsing a
syntactically sugared version of the strided layout has existed in the codebase
for a long time, it does not go as far as this commit to make the strided
layout a first-class attribute in the IR.
This introduces the attribute and updates the tests that using the pre-existing
sugared form to use the new attribute instead. Most memref created
programmatically, e.g., in passes, still use the affine form with further
extraction of strides and will be updated separately.
Update and clean-up the memref type documentation that has gotten stale and has
been referring to the details of affine map composition that are long gone.
See https://discourse.llvm.org/t/rfc-materialize-strided-memref-layout-as-an-attribute/64211.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D132864
Currently, buffer deallocation considers arith.select to be
non-aliasing, which results in deallocs being inserted incorrectly. Since
arith.select doesn't implement any useful interfaces, this change just handles
it explicitly. Eventually this should probably be fixed properly, if this pass
is going to be used long term.
Reviewed By: springerm
Differential Revision: https://reviews.llvm.org/D132460
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
Introduce two different failure propagation mode in the Transform
dialect's Sequence operation. These modes specify whether silenceable
errors produced by nested ops are immediately propagated, thus stopping
the sequence, or suppressed. The latter is useful in end-to-end
transform application scenarios where the user cannot correct the
transformation, but it is robust enough to silenceable failures. It
can be combined with the "alternatives" operation. There is
intentionally no default value to avoid favoring one mode over the
other.
Downstreams can update their tests using:
S='s/sequence \(%.*\) {/sequence \1 failures(propagate) {/'
T='s/sequence {/sequence failures(propagate) {/'
git grep -l transform.sequence | xargs sed -i -e "$S"
git grep -l transform.sequence | xargs sed -i -e "$T"
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D131774
This operation is a NavigationOp that simplifies the writing of transform IR.
Since there is no way of refering to an interface by name, the current implementation uses
an EnumAttr and depends on the interfaces it supports.
In the future, it would be worthwhile to remove this dependence and generalize.
Differential Revision: https://reviews.llvm.org/D130267
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 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
This allows for better type inference during bufferization and is in preparation of supporting memory spaces.
Differential Revision: https://reviews.llvm.org/D128581
This attribute is currently supported on AllocTensorOp only. Future changes will add support to other ops. Furthermore, the memory space is not propagated properly in all bufferization patterns and some of the core bufferization infrastructure. This will be addressed in a subsequent change.
Differential Revision: https://reviews.llvm.org/D128274
All bufferizable ops that bufferize to an allocation receive a `bufferization.escape` attribute during TensorCopyInsertion.
Differential Revision: https://reviews.llvm.org/D128137
This change changes the bufferization so that it utilizes the new TensorCopyInsertion pass. One-Shot Bufferize no longer calls the One-Shot Analysis. Instead, it relies on the TensorCopyInsertion pass to make the entire IR fully inplacable. The `bufferize` implementations of all ops are simplified; they no longer have to account for out-of-place bufferization decisions. These were already materialized in the IR in the form of `bufferization.alloc_tensor` ops during the TensorCopyInsertion pass.
Differential Revision: https://reviews.llvm.org/D127652
If `create-deallocs=0`, mark all bufferization.alloc_tensor ops as escaping. (Unless they already have an `escape` attribute.) In the absence of analysis information, check SSA use-def chains to see if the value may be yielded.
Differential Revision: https://reviews.llvm.org/D127302
There are various shortcuts in `BufferizationState::getBuffer` that avoid a buffer copy when we just need an allocation (and no initialization). This change adds those shortcuts to the TensorCopyInsertion pass, so that `getBuffer` can be simplified in a subsequent change.
Differential Revision: https://reviews.llvm.org/D126821
This pass runs the One-Shot Analysis to find out which tensor OpOperands must bufferize out-of-place. It then rewrites those tensor OpOperands to explicit allocations with a copy in the form of `bufferization.alloc_tensor`. The resulting IR can then be bufferized without having to care about read-after-write conflicts.
This change makes it possible to connect One-Shot Analysis to other bufferizations such as the sparse compiler.
Differential Revision: https://reviews.llvm.org/D126573
If `copy` is specified, the newly allocated buffer is initialized with the given contents. Also add an optional `escape` attribute to indicate whether the buffer of the tensor may be returned from the parent block (aka. "escape") after bufferization.
This change is in preparation of connecting One-Shot Bufferize to the sparse compiler.
Differential Revision: https://reviews.llvm.org/D126570
This simplifies the bufferization itself and is in preparation of connecting with the sparse compiler.
Differential Revision: https://reviews.llvm.org/D126814
Users should explicitly run `-buffer-results-to-out-params` instead.
The purpose of this change is to remove `finalizeBuffers`, which made it difficult to extend the bufferization to custom buffer types.
Differential Revision: https://reviews.llvm.org/D126253
The buffer deallocation pass must now be run explicitly when `allow-return-alloc` is set.
This results in a few extra buffer copies in unoptimized test cases. The proper way to avoid such copies is to relax the OpOperand/OpResult aliasing contract on ops such as scf.for. Some of these copies can also be avoided by improving the buffer deallocation pass.
Differential Revision: https://reviews.llvm.org/D126252
This commit allows for One-Shot Bufferize to be used through the transform dialect. No op handle is currently returned for the bufferized IR.
Differential Revision: https://reviews.llvm.org/D125098
Find writability conflicts (writes to buffers that are not allowed to be written to) by checking SSA use-def chains. This is better than the current writability analysis, which is too conservative and finds false positives.
Differential Revision: https://reviews.llvm.org/D127256
Also fixes integration of the pass into One-Shot Bufferize and adds additional test cases.
BufferResultsToOutParams can be used with "identity-layout-map" and "fully-dynamic-layout-map". "infer-layout-map" is not supported.
Differential Revision: https://reviews.llvm.org/D125636
No longer pass static dim sizes as an attribute. This was redundant and required extra checks in the verifier. This change also makes the op symmetrical to memref::AllocOp.
Differential Revision: https://reviews.llvm.org/D126178
This change adds a new op `alloc_tensor` to the bufferization dialect. During bufferization, this op is always lowered to a buffer allocation (unless it is "eliminated" by a pre-processing pass). It is useful to have such an op in tensor land, because it allows users to model tensor SSA use-def chains (which drive bufferization decisions) and because tensor SSA use-def chains can be analyzed by One-Shot Bufferize, while memref values cannot.
This change also replaces all uses of linalg.init_tensor in bufferization-related code with bufferization.alloc_tensor.
linalg.init_tensor and bufferization.alloc_tensor are similar, but the purpose of the former one is just to carry a shape. It does not indicate a memory allocation.
linalg.init_tensor is not suitable for modelling SSA use-def chains for bufferization purposes, because linalg.init_tensor is marked as not having side effects (in contrast to alloc_tensor). As such, it is legal to move linalg.init_tensor ops around/CSE them/etc. This is not desirable for alloc_tensor; it represents an explicit buffer allocation while still in tensor land and such allocations should not suddenly disappear or get moved around when running the canonicalizer/CSE/etc.
BEGIN_PUBLIC
No public commit message needed for presubmit.
END_PUBLIC
Differential Revision: https://reviews.llvm.org/D126003
