Commit Graph

7546 Commits

Author SHA1 Message Date
Mehdi Amini fd6ba1d2c0 Apply clang-tidy fixes for readability-container-size-empty in SCF.cpp (NFC) 2022-01-14 18:08:43 +00:00
Mehdi Amini 1c173c9442 Apply clang-tidy fixes for readability-identifier-naming in MathOps.cpp (NFC) 2022-01-14 18:08:42 +00:00
Mehdi Amini 3b6943e66c Apply clang-tidy fixes for performance-for-range-copy in PadOpInterchange.cpp (NFC) 2022-01-14 18:08:42 +00:00
Rahul Joshi 8067ced144 [MLIR] Introduce generic visitors.
- Generic visitors invoke operation callbacks before/in-between/after visiting the regions
  attached to an operation and use a `WalkStage` to indicate which regions have been
  visited.
- This can be useful for cases where we need to visit the operation in between visiting
  regions attached to the operation.

Differential Revision: https://reviews.llvm.org/D116230
2022-01-14 09:15:27 -08:00
Stephan Herhut aa3cabe3cb [mlir][memref] Fix memref.copy of scalar memref
Also fix a memory leak in the test while at it.

Differential Revision: https://reviews.llvm.org/D117314
2022-01-14 16:13:12 +01:00
Uday Bondhugula 05f6e93938 [MLIR] NFC. affine data copy generate utility return value cleanup
Clean up return value on affineDataCopyGenerate utility. Return the
actual success/failure status instead of the "number of bytes" which
isn't being used in the codebase in any way. The success/failure status
wasn't being sent out earlier.

Differential Revision: https://reviews.llvm.org/D117209
2022-01-14 19:37:05 +05:30
Matthias Springer 1eeffcdb7a [mlir][linalg][bufferize] Support custom insertion point for buffer copies
By default, copies are inserted right before the tensor OpOperand use. With this change, `bufferize` implementation can change the insertion point. This is needed for some ops where it would be illegal to insert a copy right before the use.

Differential Revision: https://reviews.llvm.org/D117291
2022-01-14 22:47:20 +09:00
Matthias Springer e58e401b79 [mlir][memref] Fold self copies
Fold `memref.copy %x, %x`.

Differential Revision: https://reviews.llvm.org/D117224
2022-01-14 22:44:22 +09:00
Matthias Springer 40f5f3d62d [mlir][linalg][bufferize] Use memref.copy instead of linalg.copy
Differential Revision: https://reviews.llvm.org/D117220
2022-01-14 22:29:05 +09:00
Stephan Herhut ab95ba704d [mlir][memref] Implement fast lowering of memref.copy
In the absence of maps, we can lower memref.copy to a memcpy.

Differential Revision: https://reviews.llvm.org/D116099
2022-01-14 14:22:15 +01:00
Matthias Springer 96acdfa0de [mlir][memref] Fold copy of cast
If the source/dest is a cast that does not change shape/element type, the cast can be skipped.

Differential Revision: https://reviews.llvm.org/D117215
2022-01-14 21:51:12 +09:00
Alex Zinenko bea16e72a7 [mlir] Fix invalid assertion in ModuleTranslation.cpp
LLVM dialect supports terminators with repeated successor blocks that take
different operands. This cannot be directly expressed in LLVM IR though since
it uses the number of the predecessor block to differentiate values in its PHI
nodes. Therefore, the translation to LLVM IR inserts dummy blocks to forward
arguments in case of repeated succesors with arguments. The insertion works
correctly. However, when connecting PHI nodes to their source values, the
assertion of the insertion having worked correctly was incorrect: it would only
trigger if repeated blocks were adjacent in the successor list (not guaranteed
by anything) and would not check if the successors have operands (no need for
dummy blocks in absence of operands since no PHIs are being created). Change
the assertion to only trigger in case of duplicate successors with operands,
and don't expect them to be adjacent.

Reviewed By: wsmoses

Differential Revision: https://reviews.llvm.org/D117214
2022-01-14 12:56:33 +01:00
Alexander Belyaev 436d17a8e9 [mlir] Expose a function to get vector::CombiningKind from Operation*.
Differential Revision: https://reviews.llvm.org/D117283
2022-01-14 08:28:18 +01:00
River Riddle 11067d711b [mlir] Optimize OperationName construction and usage
When constructing an OperationName, the overwhelming majority of
cases are from registered operations. This revision adds a non-locked
lookup into the currently registered operations, which prevents locking
in the common case. This revision also optimizes several uses of
RegisteredOperationName that expect the operation to be registered,
e.g. such as in OpBuilder.

These changes provides a reasonable speedup (5-10%) in some
compilations, especially on platforms where locking is expensive.

Differential Revision: https://reviews.llvm.org/D117187
2022-01-13 21:14:36 -08:00
wren romano cf35825388 [mlir][sparse] Moving the sort from factory method to the constructor.
This guarantees the preconditions of fromCOO; whereas prior to this, one could call the constructor directly with an unsorted tensor, which would cause fromCOO to misbehave.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D117167
2022-01-13 19:44:51 -08:00
Mehdi Amini cac7aabbd8 Apply clang-tidy fixes for readability-identifier-naming to MLIR (NFC) 2022-01-14 02:26:28 +00:00
Mehdi Amini e8d073951b Apply clang-tidy fixes for performance-unnecessary-value-param to MLIR (NFC) 2022-01-14 02:26:27 +00:00
Mehdi Amini bd87241c44 Apply clang-tidy fixes for modernize-use-override to MLIR (NFC) 2022-01-14 02:26:27 +00:00
Mehdi Amini 9940dcfa4a Apply clang-tidy fixes for modernize-use-equals-default to MLIR (NFC) 2022-01-14 02:26:27 +00:00
Mehdi Amini 59d8740f2f Apply clang-tidy fixes for modernize-use-default-member-init to MLIR (NFC) 2022-01-14 02:26:26 +00:00
Mehdi Amini d19f89b5d2 Apply clang-tidy fixes for llvm-qualified-auto to MLIR (NFC) 2022-01-14 02:26:26 +00:00
Rob Suderman 173fce4205 [mlir][tosa] Update default tosa-to-linalg passes
Adding the optional decompositions have been verified to improve memory
usage on common models. Added the decomposition to the default tosa to linalg
passes.

Reviewed By: NatashaKnk

Differential Revision: https://reviews.llvm.org/D117175
2022-01-13 10:17:44 -08:00
Duncan P. N. Exon Smith b77d4d54f9 mlir: Avoid SmallVector::set_size in SerializeToHsacoPass::loadLibraries
Spotted this in a final grep of projects I don't usually build before
pushing https://reviews.llvm.org/D115380, which makes
`SmallVector::set_size()` private.

Update to `truncate()`, a new-ish variant of `resize()` that asserts the
new size is not bigger and that avoids pulling in the allocation and
initialization code for growing. Doesn't really look like the perf
impact of that would matter here, but since `dirLength` is known to be a
smaller size then we might as well.

Differential Revision: https://reviews.llvm.org/D117073
2022-01-13 10:17:00 -08:00
Benoit Jacob 499703e9c0 Enable ReassociatingReshapeOpConversion with "non-identity" layouts.
Enable ReassociatingReshapeOpConversion with "non-identity" layouts.

This removes an early-return in this function, which seems unnecessary and is
preventing some memref.collapse_shape from converting to LLVM (see included lit test).

It seems unnecessary because the return message says "only empty layout map is supported"
but there actually is code in this function to deal with non-empty layout maps. Maybe
it refers to an earlier state of implementation and is just out of date?

Though, there is another concern about this early return: the condition that it actually
checks, `{src,dst}MemrefType.getLayout().isIdentity()`, is not quite the same as what the
return message says, "only empty layout map is supported". Stepping through this
`getLayout().isIdentity()` code in GDB, I found that it evaluates to `.getAffineMap().isIdentity()`
which does (AffineMap.cpp:271):

```
  if (getNumDims() != getNumResults())
    return false;
```

This seems that it would always return false for memrefs of rank greater than 1 ?

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D114808
2022-01-13 17:46:20 +00:00
Denys Shabalin a8a2ee6331 [mlir] Introduce C API for PDL dialect types
This change introduces C API helper functions to work with PDL types.
Modification closely follow the format of the https://reviews.llvm.org/D116546.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D117221
2022-01-13 15:29:01 +01:00
Javier Setoain 7c56458616 [mlir] Fix scalable type translation in splat element attr
LLVM Dialect Constant Op translations assume that if the attribute is a
vector, it's a fixed length one, generating an invalid translation for
constant scalable vector initializations.

Differential Revision: https://reviews.llvm.org/D117125
2022-01-13 11:14:41 +00:00
Ivan Butygin b65f21a498 [mlir] Provide getMixedOffsets/sizes/strides as free functions
* This is useful when you need to build mixed array from external static/dynamic arrays (e.g. from adaptor during dialect conversion)
* Also, to reduce C++ code in td and generated files

Differential Revision: https://reviews.llvm.org/D117106
2022-01-13 13:46:31 +03:00
Adrian Kuegel 159898d568 [mlir] Add missing const to cloneWith method. 2022-01-13 09:40:29 +01:00
River Riddle 0998637e6f [mlir] Add a parsePassPipeline overload that returns a new pass manager
This overload parses a pipeline string that contains the anchor operation type, and returns an OpPassManager
corresponding to the provided pipeline. This is useful for various situations, such as dynamic pass pipelines
which are not anchored within a parent pass pipeline.

fixes #52813

Differential Revision: https://reviews.llvm.org/D116525
2022-01-12 14:54:30 -08:00
natashaknk 310e9636ca [tosa][mlir] Support dynamic batch dimension for ops where the batch dim is explicit
Dynamic batch for rescale, gather, max_pool, avg_pool, conv2D and depthwise_conv2D. Split helper functions into a separate header file.

Reviewed By: rsuderman

Differential Revision: https://reviews.llvm.org/D117031
2022-01-12 14:16:50 -08:00
River Riddle 676bfb2a22 [mlir] Refactor ShapedType into an interface
ShapedType was created in a time before interfaces, and is one of the earliest
type base classes in the ecosystem. This commit refactors ShapedType into
an interface, which is what it would have been if interfaces had existed at that
time. The API of ShapedType and it's derived classes are essentially untouched
by this refactor, with the exception being the API surrounding kDynamicIndex
(which requires a sole home).

For now, the API of ShapedType and its name have been kept as consistent to
the current state of the world as possible (to help with potential migration churn,
among other reasons). Moving forward though, we should look into potentially
restructuring its API and possible its name as well (it should really have "Interface"
at the end like other interfaces at the very least).

One other potentially interesting note is that I've attached the ShapedType::Trait
to TensorType/BaseMemRefType to act as mixins for the ShapedType API. This
is kind of weird, but allows for sharing the same API (i.e. preventing API loss from
the transition from base class -> Interface). This inheritance doesn't affect any
of the derived classes, it is just for API mixin.

Differential Revision: https://reviews.llvm.org/D116962
2022-01-12 14:12:09 -08:00
River Riddle a60e83fe7c [mlir][Interfaces] Add a extraSharedClassDeclaration field
This field allows for defining a code block that is placed in both the interface
and trait declarations. This is very useful when defining a set of utilities to
expose on both the Interface class and the derived attribute/operation/type.

In non-static methods, `$_attr`/`$_op`/`$_type` (depending on the type of
interface) may be used to refer to an instance of the IR entity. In the interface
declaration, this is an instance of the interface class. In the trait declaration,
this is an instance of the concrete entity class (e.g. `IntegerAttr`, `FuncOp`, etc.).

Differential Revision: https://reviews.llvm.org/D116961
2022-01-12 14:12:08 -08:00
Rob Suderman aa1c533a4e [mlir][tosa] Expand tosa.apply_scale lowering for vectors
Apply scale may encounter scalar, tensor, or vector operations. Expand the
lowering so that it can lower arbitrary of container types.

Reviewed By: NatashaKnk

Differential Revision: https://reviews.llvm.org/D117080
2022-01-12 14:07:52 -08:00
River Riddle d4d016869d [mlir] Remove populateFuncOpTypeConversionPattern
This method simply forwards to populateFunctionLikeTypeConversionPattern,
which is more general. This also helps to remove special treatment of FuncOp from
DialectConversion.

Differential Revision: https://reviews.llvm.org/D116624
2022-01-12 14:05:35 -08:00
Kazu Hirata 44bb5cd80d [mlir] Fix a warning
This patch fixes:

  mlir/lib/Dialect/Arithmetic/Transforms/ExpandOps.cpp:161:52: error:
  'static_assert' with no message is a C++17 extension
  [-Werror,-Wc++17-extensions]
2022-01-12 12:31:43 -08:00
River Riddle 56f62fbf73 [mlir] Finish removing Identifier from the C++ API
There have been a few API pieces remaining to allow for a smooth transition for
downstream users, but these have been up for a few months now. After this only
the C API will have reference to "Identifier", but those will be reworked in a followup.

The main updates are:
* Identifier -> StringAttr
* StringAttr::get requires the context as the first parameter
  - i.e. `Identifier::get("...", ctx)` -> `StringAttr::get(ctx, "...")`

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116626
2022-01-12 11:58:23 -08:00
Christian Sigg be1aeb818c Remove NaN constant from arith.minf, arith.maxf expansion
If any of the operands is NaN, return the operand instead of a new constant.

When the rhs operand is a constant, the second arith.cmpf+select ops will be folded away.

https://reviews.llvm.org/D117010 marks the two ops commutative, which will place the constant on the rhs.

Reviewed By: herhut

Differential Revision: https://reviews.llvm.org/D117011
2022-01-12 20:56:40 +01:00
Christian Sigg f6fab68c30 Fold arith.cmpf when at least one operand is known to be NaN.
Reviewed By: herhut

Differential Revision: https://reviews.llvm.org/D117101
2022-01-12 20:48:52 +01:00
Thomas Raoux b0a309dd7a [mlir][vector] Add folding for extract + extract/insert_strided
Differential Revision: https://reviews.llvm.org/D116785
2022-01-12 11:48:35 -08:00
William S. Moses 1773dddadf [MLIR][Math] Enable constant folding of ops
Enable constant folding of ops within the math dialect, and introduce constant folders for ceil and log2

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D117085
2022-01-12 12:19:29 -05:00
Matthias Springer 57e714bcc8 [mlir][linalg][bufferize] Add pass options for `createDeallocs`
This change makes it possible to use a different buffer deallocation strategy. E.g., `-buffer-deallocation` can be used, which also works for allocations that are not in destination-passing style.

Differential Revision: https://reviews.llvm.org/D117096
2022-01-12 18:55:36 +09:00
Matthias Springer 6c654b5198 [mlir][linalg][bufferize] Support std.select bufferization
This op is an example for how to deal with ops who's OpResult may aliasing with one of multiple OpOperands.

Differential Revision: https://reviews.llvm.org/D116868
2022-01-12 17:46:44 +09:00
Uday Bondhugula fc61d07dc1 Add inliner interface for GPU dialect
Add inliner interface for GPU dialect. The interface marks all GPU
dialect ops legal to inline anywhere.

Differential Revision: https://reviews.llvm.org/D116889
2022-01-12 12:55:02 +05:30
William S. Moses d2c547342c Revert "[MLIR][Math] Enable constant folding of ops"
This reverts commit 2f8b956ab6.

There is a linker error for mlir-nvidia as seen on
https://lab.llvm.org/buildbot/#/builders/61/builds/19939.

As it's late for me here, I'm oing to rever this for now to be
investigated later.
2022-01-12 02:00:53 -05:00
William S. Moses 2f8b956ab6 [MLIR][Math] Enable constant folding of ops
Enable constant folding of ops within the math dialect, and introduce constant folders for ceil and log2

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D117085
2022-01-12 01:55:48 -05:00
William S. Moses d23fa4f2f1 [MLIR][SCF] Remove unused arguments to whileop
Canonicalize away unused arguments to the before region of a whileOp

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D117059
2022-01-11 20:18:08 -05:00
William S. Moses 97567bde5b [MLIR][SCF] Canonicalize while statement whose cmp condition is recomputed in the after region
Given a while loop whose condition is given by a cmp, don't recomputed the comparison (or its inverse) in the after region, instead use a constant since  the original condition must be true if we branched to the after region.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D117047
2022-01-11 18:34:04 -05:00
MaheshRavishankar e7cb716ef9 [mlir][Linalg] Pattern to fuse pad operation with elementwise operations.
Most convolution operations need explicit padding of the input to
ensure all accesses are inbounds. In such cases, having a pad
operation can be a significant overhead. One way to reduce that
overhead is to try to fuse the pad operation with the producer of its
source.

A sequence

```
linalg.generic -> linalg.pad_tensor
```

can be replaced with

```
linalg.fill -> tensor.extract_slice -> linalg.generic ->
tensor.insert_slice.
```

if the `linalg.generic` has all parallel iterator types.

Differential Revision: https://reviews.llvm.org/D116418
2022-01-11 13:37:25 -08:00
Mehdi Amini 75b08cce47 Apply clang-tidy fixes for readability-redundant-control-flow in OpenMPDialect.cpp (NFC) 2022-01-11 19:13:58 +00:00
Kazu Hirata 0ceb332a76 [mlir] Fix a missing override warning
This patch fixes:

  mlir/lib/Dialect/Tosa/Transforms/TosaOptionalDecompositions.cpp:28:8:
  error: 'runOnFunction' overrides a member function but is not marked
  'override' [-Werror,-Wsuggest-override]
2022-01-11 11:03:22 -08:00
Aaron DeBattista dfd070820c [mlir][tosa] Allow optional TOSA decompositions to be populated separately
Moved all TOSA decomposition patterns so that they can be optionally populated
and used by external rewrites. This avoids decomposing TOSa operations when
backends may benefit from the non-decomposed version.

Reviewed By: rsuderman, mehdi_amini

Differential Revision: https://reviews.llvm.org/D116526
2022-01-11 10:26:30 -08:00
William S. Moses 5443d2ed98 [MLIR][SCF] Simplify scf.if by swapping regions if condition is a not
Given an if of the form, simplify it by eliminating the not and swapping the regions

scf.if not(c) {
  yield origTrue
} else {
  yield origFalse
}

becomes

scf.if c {
  yield origFalse
} else {
  yield origTrue
}

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116990
2022-01-11 12:57:29 -05:00
Benjamin Kramer 7f47005dcc [mlir][linalg] Use cast instead of dyn_cast that's always dereferenced
This turns a random nullptr deref into an assertion failure in case
`tensor::registerInferTypeOpInterfaceExternalModels` isn't called.
2022-01-11 15:05:45 +01:00
Matthias Springer 2c5c5ca868 [mlir][linalg][bufferize] Fix CallOp bufferization
Previously, CallOps did not have any aliasing OpResult/OpOperand pairs. Therefore, CallOps were mostly ignored by the analysis and buffer copies were not inserted when necessary.

This commit introduces the following changes:
* Function bbArgs writable by default. A function can now be bufferized without inspecting its callers.
* Callers must introduce buffer copies of function arguments when necessary. If a function is external, the caller must conservatively assume that a function argument is modified by the callee after bufferization. If the function is not external, the caller inspects the callee to determine if a function argument is modified.

Differential Revision: https://reviews.llvm.org/D116457
2022-01-11 20:10:21 +09:00
Julian Gross 4b01968b5e [MLIR] Update allocs to memref.allocs in documentation.
Changed the remaining appearances of alloc to memref.alloc in several
documentation sections, since they lead to misunderstandings, if they
are used.

Differential Revision: https://reviews.llvm.org/D116999
2022-01-11 11:22:22 +01:00
Diego Caballero e2b658cd5d [mlir][GPU] Fix attribute name of DL specification
D115722 added a DL spec to GPU modules. It happens that the DL
default interface implementation is sensitive to the name of the
DL spec attribute. This patch is fixing the name of the attribute
to be the expected one.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116956
2022-01-11 08:30:52 +00:00
Tyler Augustine 87a9be2a74 Don't fail if unable to promote loops during unrolling
When the unroll factor is 1, we should only fail "unrolling" when the trip count also is determined to be 1 and it is unable to be promoted.

Reviewed By: bondhugula

Differential Revision: https://reviews.llvm.org/D115365
2022-01-10 22:26:21 +00:00
William S. Moses a02af37560 [MLIR] Generalize select to arithmetic canonicalization
Given a select whose result is an i1, we can eliminate the conditional in the select completely by adding a few arithmetic operations.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116839
2022-01-10 11:50:17 -05:00
James Y Knight 39f75ab6a7 Fix unused-variable warnings after d0ee094d6a. 2022-01-10 16:28:54 +00:00
Nicolas Vasilache d0ee094d6a [mlir][Bufferize] Fix incorrect bufferization of rank-reducing tensor ops.
This revision fixes SubviewOp, InsertSliceOp, ExtractSliceOp construction during bufferization
where not all offset/size/stride operands were properly specified.

A test that exhibited problematic behaviors related to incorrect memref casts is introduced.
Init tensor optimization is disabled in teh testing func bufferize pass.

Differential Revision: https://reviews.llvm.org/D116899
2022-01-10 10:14:55 -05:00
Matthias Springer 0696ad9563 [mlir][linalg][bufferize][NFC] Pass missing BufferizationState objs as const ref
These should have been updated as part of D116742.

Differential Revision: https://reviews.llvm.org/D116937
2022-01-10 23:38:04 +09:00
Nicolas Vasilache 1a2474b786 [mlir][Linalg] Disable init_tensor elimination by default
init_tensor elimination is arguably a pre-optimization that should be separated from comprehensive bufferization.
In any case it is still experimental and easily results in wrong IR with violated SSA def-use orderings.
Isolate the optimization behind a flag, separate the test cases and add a test case that would results in wrong IR.

Differential Revision: https://reviews.llvm.org/D116936
2022-01-10 09:19:18 -05:00
Serge Guelton d2cc6c2d0c Use a sorted array instead of a map to store AttrBuilder string attributes
Using and std::map<SmallString, SmallString> for target dependent attributes is
inefficient: it makes its constructor slightly heavier, and involves extra
allocation for each new string attribute. Storing the attribute key/value as
strings implies extra allocation/copy step.

Use a sorted vector instead. Given the low number of attributes generally
involved, this is cheaper, as showcased by

https://llvm-compile-time-tracker.com/compare.php?from=5de322295f4ade692dc4f1823ae4450ad3c48af2&to=05bc480bf641a9e3b466619af43a2d123ee3f71d&stat=instructions

Differential Revision: https://reviews.llvm.org/D116599
2022-01-10 14:49:53 +01:00
Matthias Springer cb64c541f9 [mlir][linalg][bufferize][NFC] Clean up bufferization entry point
Differential Revision: https://reviews.llvm.org/D116892
2022-01-10 21:50:36 +09:00
Alex Zinenko c44d521b30 [mlir] address post-commit review for D116759 2022-01-10 12:40:58 +01:00
Alex Zinenko 2f672e2ffa [mlir] Don't inline calls from dead SCCs
During iterative inlining of the functions in a multi-step call chain, the
inliner could add the same call operation several times to the worklist, which
led to use-after-free when this op was considered more than once.

Closes #52887.

Reviewed By: wsmoses

Differential Revision: https://reviews.llvm.org/D116820
2022-01-10 12:07:14 +01:00
Stephan Herhut 33cec20dbd [mlir][memref] Tighten verification of memref.reinterpret_cast
We allow the omission of a map in memref.reinterpret_cast under the assumption,
that the cast might cast to an identity layout. This change adds verification
that the static knowledge that is present in the reinterpret_cast supports
this assumption.

Differential Revision: https://reviews.llvm.org/D116601
2022-01-10 11:55:47 +01:00
Shraiysh Vaishay a8586b573e [mlir][OpenMP] Change the syntax of omp.atomic.read op
This patch changes the syntax of omp.atomic.read to take the address of
destination, instead of having the value in a result. This will allow
using omp.atomic.read operation within an omp.atomic.capture operation
thus making its implementation less complex.

Reviewed By: peixin

Differential Revision: https://reviews.llvm.org/D116396
2022-01-10 16:19:45 +05:30
Mehdi Amini 60d13b85ca Apply clang-tidy fixes for performance-move-const-arg in ReductionNode.cpp (NFC) 2022-01-10 01:05:14 +00:00
Mehdi Amini ba19fa5779 Apply clang-tidy fixes for performance-for-range-copy in ElementwiseOpFusion.cpp (NFC) 2022-01-10 01:05:14 +00:00
Mehdi Amini 0c9d59e56a Apply clang-tidy fixes for performance-for-range-copy in ModuleBufferization.cpp (NFC) 2022-01-10 01:05:14 +00:00
Mehdi Amini 6a38cbfb53 Apply clang-tidy fixes for modernize-use-equals-default in IRCore.cpp (NFC) 2022-01-10 01:05:13 +00:00
Nicolas Vasilache 8871d8236a [mlir][MemRef] NFC - Improve assertion error message 2022-01-09 14:17:37 -05:00
Nicolas Vasilache 9ba25ec92d [mlir][Bufferize] NFC - Introduce areCastCompatible assertions to catch misformed CastOp early
Differential Revision: https://reviews.llvm.org/D116893
2022-01-09 14:13:08 -05:00
Mehdi Amini 18eb681821 Fix clang-tidy readability-redundant-smartptr-get in MLIR ModuleTranslation.cpp (NFC) 2022-01-08 20:07:11 +00:00
Uday Bondhugula 9cf9ed94ed Multiple fixes to affine loop tiling return status and checks
Fix crash in the presence of yield values. Multiple fixes to affine loop
tiling pre-condition checks and return status. Do not signal pass
failure on a failure to tile since the IR is still valid. Detect index
set computation failure in checkIfHyperrectangular and return failure.
Replace assertions with proper status return. Move checks to an
appropriate place earlier in the utility before mutation happens.

Differential Revision: https://reviews.llvm.org/D116738
2022-01-08 16:50:44 +05:30
Groverkss 0e19186c82 [MLIR][NFC] Move PresburgerSet to Presburger/ directory
This patch moves PresburgerSet to Presburger/ directory. This patch is purely
mechincal, it only moves and renames functionality and tests.

This patch is part of a series of patches to move presburger functionality to
Presburger/ directory.

Reviewed By: arjunp

Differential Revision: https://reviews.llvm.org/D116836
2022-01-08 15:39:40 +05:30
Mehdi Amini 3e13c4c37c Avoid creating a ThreadPool in MlirOptMain when `--mlir-disable-threading` option is set
a32300a changed it to create a ThreadPool eagerly so that it gets reused
across buffers, however it also made it so that we create a ThreadPool
early even if we're not gonna use it later because of the command line
option `--mlir-disable-threading` is provided.

Fix #53056

Reland 45adf60802 after build fixes

Differential Revision: https://reviews.llvm.org/D116848
2022-01-08 02:26:41 +00:00
Mehdi Amini 4938949310 Revert "Avoid creating a ThreadPool in MlirOptMain when `--mlir-disable-threading` option is set"
This reverts commit 45adf60802.
Build is broken
2022-01-08 02:25:18 +00:00
Mehdi Amini 45adf60802 Avoid creating a ThreadPool in MlirOptMain when `--mlir-disable-threading` option is set
a32300a changed it to create a ThreadPool eagerly so that it gets reused
across buffers, however it also made it so that we create a ThreadPool
early even if we're not gonna use it later because of the command line
option `--mlir-disable-threading` is provided.

Fix #53056

Differential Revision: https://reviews.llvm.org/D116848
2022-01-08 02:18:19 +00:00
Groverkss 49b754b5c6 [MLIR][NFC] Move presburger functionality from FlatAffineConstraints to IntegerPolyhedron
This patch moves all presburger functionality from FlatAffineConstraints to
IntegerPolyhedron. This patch is purely mechanical, it only moves and renames
functionality and tests.

This patch is part of a series of patches to move presburger functionality to
Presburger/ directory.

Reviewed By: bondhugula

Differential Revision: https://reviews.llvm.org/D116681
2022-01-08 01:22:49 +05:30
Kazu Hirata 117422c0da [ComprehensiveBufferize] Fix a warning
This patch fixes:

  mlir/lib/Dialect/Linalg/ComprehensiveBufferize/LinalgInterfaceImpl.cpp:292:12:
  error: comparison of integers of different signs: 'int' and
  'unsigned int' [-Werror,-Wsign-compare]
2022-01-07 09:25:59 -08:00
Matthias Springer 089b910abc [mlir][linalg][bufferize][NFC] Add `analyzeOp` helper function
This function runs just the analysis of Comprehensive Bufferize, but does not bufferize the IR yet.

This is in preparation of fixing CallOp bufferization. Also needed for unifying Comprehensive Bufferize and core bufferization; the new partial bufferization can simply run bufferization without an analysis.

Differential Revision: https://reviews.llvm.org/D116456
2022-01-08 01:33:41 +09:00
Matthias Springer d9184ab1a5 [mlir][linalg][bufferize][NFC] Simplify buffer API of BufferizationState
Instead of `lookupBuffer` and `getResultBuffer`, there is now a single `getBuffer` function. This simplifies the `BufferizableOpInterface` API and is less confusing to users. They could previously have called the wrong function.

Furthermore, since `getBuffer` now takes an `OpOperand &` instead of a `Value`, users can no longer accidentally use one of the previous two functions incorrectly, which would have resulted in missing buffer copies.

Differential Revision: https://reviews.llvm.org/D116455
2022-01-08 01:12:18 +09:00
Matthias Springer 8e2b6aac32 [mlir][linalg][bufferize][NFC] Analyze OpOperands instead of OpResults
With this change, the analysis takes a look at OpOperands instead of OpResults. OpOperands can bufferize out-of-place (even if they have no aliasing OpResults). The analysis does no longer care about OpResults.

Previously, only OpResults could bufferize out-of-place, so OpOperands that have no aliasing OpResults were never copied by Comprehensive Bufferize. This does not fit wwell with the new CallOp bufferization that is introduced in a subsequent change. In essence, called FuncOps can then be treated as "black boxes" that may read/write to any bbArg, even if they do not return anything.

Differential Revision: https://reviews.llvm.org/D115706
2022-01-08 01:00:30 +09:00
Kazu Hirata e56a9c9b5b Remove redundant return statements (NFC)
Identified by readability-redundant-control-flow.
2022-01-07 07:42:35 -08:00
Matthias Springer 547b9afc54 [mlir][linalg][bufferize][NFC] Add explicit inplaceable attrs to test cases
This is in preparation of fixing CallOp bufferization. Add explicit linalg.inplaceable attrs to all bbArgs, except for the ones where inplaceability should be decided by the analysis.

Differential Revision: https://reviews.llvm.org/D115840
2022-01-08 00:25:15 +09:00
Matthias Springer b8d0753694 [mlir][linalg][bufferize] Fix copy elision in `getResultBuffer`
A buffer copy may not be elided if the to-be-bufferized op is reading the data.

Differential Revision: https://reviews.llvm.org/D116454
2022-01-08 00:19:17 +09:00
gysit cf05668c17 [mlir][OpDSL] Rename `PrimFn` to `ArithFn`.
The revision renames `PrimFn` to `ArithFn`. The name resembles the newly introduced arith dialect that implements most of the arithmetic functions. An exception are log/exp that are part of the math dialect.

Depends On D115239

Reviewed By: stellaraccident

Differential Revision: https://reviews.llvm.org/D115240
2022-01-07 12:38:03 +00:00
gysit 15757ea80a [mlir][OpDSL] Add `TypeFn` class.
This revision introduces a the `TypeFn` class that similar to the `PrimFn` class contains an extensible set of type conversion functions. Having the same mechanism for both type conversion functions and arithmetic functions improves code consistency. Additionally, having an explicit function class and function name is a prerequisite to specify a conversion or arithmetic function via attribute. In a follow up commits, we will introduce function attributes to make OpDSL operations more generic. In particular, the goal is to handle signed and unsigned computation in one operations. Today, there is a linalg.matmul and a linalg.matmul_unsigned.

The commit implements the following changes:
- Introduce the class of type conversion functions `TypeFn`
- Replace the hardwired cast and cast_unsigned ops by the `TypeFn` counterparts
- Adapt the python and C++ code generation paths to support the new cast operations

Example:
```
cast(U, A[D.m, D.k])
```
changes to
```
TypeFn.cast(U, A[D.m, D.k])
```

Depends On D115237

Reviewed By: stellaraccident

Differential Revision: https://reviews.llvm.org/D115239
2022-01-07 12:26:47 +00:00
Adrian Kuegel babad7c566 [mlir][python] Use a named object
Currently, the object would be immediately destroyed after creation.
Found by ClangTidy bugprone-unused-raii.
2022-01-07 13:22:32 +01:00
Benjamin Kramer ae29d5a484 [mlir] Provide a home for mlir::LLVM::GEPOp::kDynamicIndex
C++14 requires this and will trigger linker errors when optimizations
are disabled.
2022-01-07 12:06:04 +01:00
Adrian Kuegel 50da013406 [mlir] Use const reference for loop iteration variable. 2022-01-07 12:00:53 +01:00
Alex Zinenko f50cfc44d6 [mlir] Require struct indices in LLVM::GEPOp to be constant
Recent commits added a possibility for indices in LLVM dialect GEP operations
to be supplied directly as constant attributes to ensure they remain such until
translation to LLVM IR happens. Make this required for indexing into LLVM
struct types to match LLVM IR requirements, otherwise the translation would
assert on constructing such IR.

For better compatibility with MLIR-style operation construction interface,
allow GEP operations to be constructed programmatically using Values pointing
to known constant operations as struct indices.

Depends On D116758

Reviewed By: wsmoses

Differential Revision: https://reviews.llvm.org/D116759
2022-01-07 09:56:05 +01:00
Alex Zinenko 43ff4a6d55 [mlir] Add ConstantLike trait to LLVM::ConstantOp
This make LLVM dialect constants to work with `m_constant` matches. Implement
the folding hook for this operation as required by the trait. This in turn
allows LLVM::ConstantOp to properly participate in constant-folding.

Depends On D116757

Reviewed By: wsmoses

Differential Revision: https://reviews.llvm.org/D116758
2022-01-07 09:56:03 +01:00
Alex Zinenko cafaa35036 [mlir] Make it possible to directly supply constant values to LLVM GEPOp
In LLVM IR, the GEP indices that correspond to structures are required to be
i32 constants. MLIR models constants as just values defined by special
operations, and there is no verification that it is the case for structure
indices in GEP. Furthermore, some common transformations such as control flow
simplification may lead to the operands becoming non-constant. Make it possible
to directly supply constant values to LLVM GEPOp to guarantee they remain
constant until the translation to LLVM IR. This is not yet a requirement and
the verifier is not modified, this will be introduced separately.

Reviewed By: wsmoses

Differential Revision: https://reviews.llvm.org/D116757
2022-01-07 09:56:01 +01:00
William S. Moses 34646a2f7e [MLIR][Arith] Fold repeated xor and trunc
This patch adds two folds. One for a repeated xor (e.g. xor(xor(x, a), a)) and one for a repeated trunc (e.g. trunc(trunc(x))).

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116383
2022-01-07 03:36:10 -05:00
Kazu Hirata 410480e32b Ensure newlines at the end of files (NFC) 2022-01-06 23:44:02 -08:00
Shraiysh Vaishay 6bcb4c44de [mlir][OpenMP] Added omp.atomic.write lowering to LLVM IR
This patch adds omp.atomic.write lowering to LLVM IR.
Also, changed the syntax to have equal symbol instead of the comma to
make it more intuitive.

Reviewed By: kiranchandramohan, peixin

Differential Revision: https://reviews.llvm.org/D116416
2022-01-07 10:01:57 +05:30
Nicolas Vasilache 07c009ed55 [mlir][Linalg] Fix invalid FailureOr construction from LogicalResult::success 2022-01-06 18:43:23 -05:00
Nicolas Vasilache 9cd7e880fd [mlir][Linalg] NFC - Modernize more transformation patterns.
Differential Revision: https://reviews.llvm.org/D116763
2022-01-06 17:40:23 -05:00
Matthias Springer 42fd68b344 [mlir][linalg][bufferize] LinalgOp: Move existing region to new op
This has two advantages.

1. It is more efficient. No need to clone the entire region.
2. Recreating ops (via cloning) invalidates analysis results. Previously, an OpResult could have bufferized out-of-place, even though the analysis requested an in-place bufferization. That is because BufferizationState keeps track of OpResults for storing bufferization analysis results (and cloned ops have new OpResults).

Differential Revision: https://reviews.llvm.org/D116453
2022-01-07 07:00:24 +09:00
Matthias Springer 698896cd6c [mlir][linalg][bufferize][NFC] Change allocationFn return type to FailureOr<Value>
In addition, all functions that call `allocationFn` now return FailureOr<Value>. This resolves a few TODOs in the code base.

Differential Revision: https://reviews.llvm.org/D116452
2022-01-07 06:33:19 +09:00
Nicolas Vasilache 4a661602ef [mlir][Linalg] NFC - Modernize APIs and get rid of unnecessary tiling paterns.
Tiling patterns can be reduced to a single pattern by using interface-based patterns.

Differential Revision: https://reviews.llvm.org/D116733
2022-01-06 16:27:35 -05:00
Matthias Springer 75d65293ca [mlir][linalg][bufferize][NFC] Clean up comments and minor code refactorings
Differential Revision: https://reviews.llvm.org/D116451
2022-01-07 06:23:01 +09:00
Matthias Springer 0f5edb9d63 [mlir][linalg][bufferize] Add `createDealloc` options
If `createDealloc` is deactivated (enabled by default), newly allocated buffers are not deallocated anymore. In such a case, the missing deallocations can be inserted by the existing "BufferDeallocation" pass.

This change is needed for unifying core bufferization and Comprehensive Bufferize. Core bufferization has a separate pass for generating deallocations.

Note: In the future, this will evolve towards generating deallocation ops only for buffer allocations that do not escape block boundaries (i.e., that are in destination passing style).

Differential Revision: https://reviews.llvm.org/D116450
2022-01-07 06:13:57 +09:00
Matthias Springer bf9d8d9dfb [mlir][linalg][bufferize][NFC] Rename functions in BufferizationState
The old function names (e.g., `replaceOp`) could have been confusing to users because they sound similar to rewriter functions, but have slightly different semantics.

Differential Revision: https://reviews.llvm.org/D116449
2022-01-07 05:28:58 +09:00
MaheshRavishankar 4317a3dfad [mlir][Linalg] Disable fusion of reshape with elementwise ops for purely dynamic cases.
`tensor.collapse_shape` op when fused with a consumer elementwise
`linalg.generic` operation results in creation of tensor.expand_shape
ops. In purely dynamic cases this can end up with a dynamic dimensions
being expanded to more than one dynamic dimension. This is disallowed
by the semantics of `tensor.expand_shape` operation. (While the
transformation is itself correct, its a gap in the specification of
`tensor.expand_shape` that is the issue). So disallow fusions which
result in such a pattern.

Differential Revision: https://reviews.llvm.org/D116703
2022-01-06 10:32:24 -08:00
Matthias Springer 15c7e3ee15 [mlir][linalg][bufferize][NFC] Use RewritePatterns instead of custom traversal
This change simplifies BufferizableOpInterface and other functions. Overall, the API will get smaller: Functions related to custom IR traversal are deleted entirely. This will makes it easier to write BufferizableOpInterface implementations.

This is also in preparation of unifying Comprehensive Bufferize and core bufferization. While Comprehensive Bufferize could theoretically maintain its own IR traversal, there is no reason to do so, because all bufferize implementations in BufferizableOpInterface have to support partial bufferization anyway. And we can share a larger part of the code base between the two bufferizations.

Differential Revision: https://reviews.llvm.org/D116448
2022-01-07 00:56:54 +09:00
Matthias Springer 2975407bd4 [mlir][linalg][bufferize][NFC] Pass BufferizationState as const reference
This is mostly for documentation purposes: Passing the object as a const reference signifies that analysis decisions cannot be changed after the analysis.

Differential Revision: https://reviews.llvm.org/D116742
2022-01-07 00:18:46 +09:00
Matthias Springer cd84cf90e9 [mlir][linalg][bufferize][NFC] Do not cache bufferized function types
This does not work if BufferizationState is passed around as a const reference in most places.

Differential Revision: https://reviews.llvm.org/D116741
2022-01-07 00:04:57 +09:00
Nicolas Vasilache 2c4a56c418 [mlir][Linalg] NFC - Modernize padding pattern
Differential Revision: https://reviews.llvm.org/D116739
2022-01-06 08:59:35 -05:00
Arjun P ef8351598e [MLIR] Simplex::findPivotRow: silence spurious coverity warning
Initialize some variables to zero to avoid a warning about them possibly being
used uninitialized. In actuality, they will never be used before initialization.
2022-01-06 19:18:30 +05:30
Arjun P 0f925339e5 [MLIR] Simplex::normalizeRow: assert gcd is non-zero before dividing 2022-01-06 19:10:47 +05:30
Prashant Kumar df29318e66 [MLIR] Add division normalization by GCD in `getDivRepr` fn.
This commits adds division normalization in  the `getDivRepr` function which extracts
 the gcd from the dividend and divisor and normalizes them.

Signed-off-by: Prashant Kumar <pk5561@gmail.com>

Reviewed By: bondhugula

Differential Revision: https://reviews.llvm.org/D115595
2022-01-06 16:18:50 +05:30
Markus Böck 560972052a [mlir][LLVM] Implement mapping of phi source values of `llvm.invoke`
This patch allows the usage of the normalDestOperands and unwindDestOperands operands of llvm.invoke and have them be correctly mapped to phis in the successor when exported to LLVM IR.

Differential Revision: https://reviews.llvm.org/D116706
2022-01-06 11:27:14 +01:00
Alex Zinenko 95ddbed9b7 [mlir] Split out Python bindings for dialects into separate libs
Historically, the bindings for the Linalg dialect were included into the
"core" bindings library because they depended on the C++ implementation
of the "core" bindings. The other dialects followed the pattern. Now
that this dependency is gone, split out each dialect into a separate
Python extension library.

Depends On D116649, D116605

Reviewed By: stellaraccident

Differential Revision: https://reviews.llvm.org/D116662
2022-01-06 10:31:14 +01:00
Matthias Springer 18e08fbd01 [mlir][linalg][bufferize] Fix tiled_loop bufferization
Until now, bufferization assumed that the yieleded tensor of a linalg.tiled_loop is an output tensor. This is not necessarily the case.

Differential Revision: https://reviews.llvm.org/D116685
2022-01-06 17:51:33 +09:00
Matthias Springer 0e5f258452 [mlir][linalg][bufferize][NFC] Simplify InsertSliceOp bufferization
No need to keep track of equivalent extract_slice / insert_slice tensors during bufferization. Just emit a copy, it will fold away.

Note: The analysis still keeps track of equivalent tensors to make the correct inplace bufferization decisions.

Differential Revision: https://reviews.llvm.org/D116684
2022-01-06 17:35:45 +09:00
Vaivaswatha Nagaraj 2c384c3772 [MLIR][DataFlowAnalysis] Use a queue to maintain the worklist
Since the analysis is described to be suitable for a forward
data-flow analysis, maintaining the worklist as a queue mimics
RPO ordering of block visits, thus reaching the fixpoint earlier.

Differential Revision: https://reviews.llvm.org/D116393
2022-01-06 09:55:22 +05:30
William S. Moses 358d020017 [MLIR][LLVM] Add simple folders for bitcast/addrspacecast/gep
Add 5 simple folders
* bitcast(x : T0, T0) -> x
* addrcast(x : T0, T0) -> x
* bitcast(bitcast(x : T0, T1), T0) -> x
* addrcast(addrcast(x : T0, T1), T0) -> x
* gep %x:T, 0 -> %x:T

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116715
2022-01-05 21:17:32 -05:00
Mogball b0774e5f50 [mlir][ods] ODS ops get an `extraClassDefinition`
Extra definitions are placed in the generated source file for each op class. The substitution `$cppClass` is replaced by the op's C++ class name.

This is useful when declaring but not defining methods in TableGen base classes:

```
class BaseOp<string mnemonic>
    : Op<MyDialect, mnemonic, [DeclareOpInterfaceMethods<SomeInterface>] {
  let extraClassDeclaration = [{
    // ZOp is declared at at the bottom of the file and is incomplete here
    ZOp getParent();
  }];
  let extraClassDefinition = [{
    int $cppClass::someInterfaceMethod() {
      return someUtilityFunction(*this);
    }
    ZOp $cppClass::getParent() {
      return dyn_cast<ZOp>(this->getParentOp());
    }
  }];
}
```

Certain things may prevent defining these functions inline, in the declaration. In this example, `ZOp` in the same dialect is incomplete at the function declaration because ops classes are declared in alphabetical order. Alternatively, functions may be too big to be desired as inlined, or they may require dependencies that create cyclic includes, or they may be calling a templated utility function that one may not want to expose in a header. If the functions are not inlined, then inheriting from the base class N times means that each function will need to be defined N times. With `extraClassDefinitions`, they only need to be defined once.

Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D115783
2022-01-06 01:43:26 +00:00
wren romano ceda1ae9a7 [mlir][sparse] Strengthening first arguments of fromCOO/toCOO
Better capturing of invariants

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D116700
2022-01-05 16:15:28 -08:00
wren romano c03fd1e61f [mlir][sparse] Marking cursor parameters const
These parameters aren't modified, so we make that invariant explicit.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D116693
2022-01-05 16:14:02 -08:00
Mogball 4ca5e95c6f [mlir] Symbol DCE ignores unknown symbols
Instead of failing when it encounters a reference to an unknown symbol, Symbol DCE should ignore them. References to unknown symbols do not affect the overall function of Symbol DCE, so it should not need to fail when it encounters one.

In general, requiring that symbol references always be valid rather than only when necessary can be overly conservative.

Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D116047
2022-01-05 20:48:30 +00:00
Groverkss dde7388ad5 [MLIR] Add clearAndCopyFrom to IntegerPolyhedron
This patch adds clearAndCopyFrom to IntegerPolyhedron. This requires moving
LLVM-style RTTI from FlatAffineConstraints to IntegerPolyhedron.

This patch is part of a series of patches to move presburger math to Presburger
directory.

Reviewed By: arjunp

Differential Revision: https://reviews.llvm.org/D116533
2022-01-05 23:39:26 +05:30
Arjun P dbb2e74da3 [MLIR] Simplex::normalizeRow: early exit when gcd is one
Reviewed By: bondhugula

Differential Revision: https://reviews.llvm.org/D116672
2022-01-05 23:26:29 +05:30
Kazu Hirata afc94c0ed7 [mlir] Fix a compiler warning
This patch fixes:

  mlir/lib/Dialect/Linalg/ComprehensiveBufferize/ModuleBufferization.cpp:635:23:
  error: comparison of integers of different signs: 'int' and 'size_t'
  (aka 'unsigned long') [-Werror,-Wsign-compare]
2022-01-05 09:42:03 -08:00
Nicolas Vasilache c05db63887 [mlir] Fix for 9a7d111f4f 2022-01-05 11:40:51 -05:00
Nicolas Vasilache 9a7d111f4f [mlir][Linalg] NFC - Modernize transformation APIs.
Differential Revision: https://reviews.llvm.org/D116665
2022-01-05 11:01:40 -05:00
Matthias Springer b15b0156ca [mlir][linalg][bufferize][NFC] Simplify bufferization of CallOps
There is no need to inspect the ReturnOp of the called function.

This change also refactors the bufferization of CallOps in such a way that `lookupBuffer` is called only a single time. This is important for a later change that fixes CallOp bufferization. (There is currently a TODO among the test cases.)

Note: This change modifies a test case but is marked as NFC. There is no change of functionality, but FuncOps with empty bodies are now reported with a different error message.

Differential Revision: https://reviews.llvm.org/D116446
2022-01-06 00:28:47 +09:00
Alex Zinenko 66d4090d9b [mlir] Introduce Python bindings for the quantization dialect
So far, only the custom dialect types are exposed.

The build and packaging is same as for Linalg and SparseTensor, and in
need of refactoring that is beyond the scope of this patch.

Reviewed By: stellaraccident

Differential Revision: https://reviews.llvm.org/D116605
2022-01-05 16:26:31 +01:00
Alex Zinenko 9bcf13bf3e [mlir] Introduce C API for the Quantization dialect types
Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D116546
2022-01-05 16:20:29 +01:00
Matthias Springer a98c5a08b1 [mlir][linalg][bufferize] Fix CallOps with non-tensor operands
Such CallOps were not handled properly. When computing the new result types (and replacement values) of a CallOp, non-tensor return values were not accounted for.

Differential Revision: https://reviews.llvm.org/D116445
2022-01-06 00:19:23 +09:00
Alex Zinenko d716cfc4fa [mlir] Use public PybindAdaptors in Linalg dialect bindings
Previously, the Python bindings for the Linalg dialect relied on the internal
implementation of core bindings. Most of that functionality was moved, and the
remaining one does not need access to the implementation: it used to accept a
dialect pointer as argument, but it can always be extracted from the operation
that it also accepts; operations are available through PybindAdaptors in an
opaque way. Change the bindings in that direction.

This enables the decoupling of the Linalg dialect Python extension from the
core IR Python extension.

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D116649
2022-01-05 16:18:30 +01:00
Matthias Springer ed5e3590a3 [mlir][linalg][bufferize][NFC] Remove RewriterBase from BufferizationState
This change simplifies BufferizationState. Having `rewriter` in BufferizationState could be confusing to users because a rewriter is also passed to each `bufferize` function and it is not obvious (by looking at the API) that these two rewriters are the same.

Differential Revision: https://reviews.llvm.org/D116444
2022-01-06 00:04:43 +09:00
Nicolas Vasilache bb2f87af0a [mlir] Fix missing check on nested op values in LICM
LICM checks that nested ops depend only on values defined outside
before performing hoisting.
However, it specifically omits to check for terminators which can
lead to SSA violations.
This revision fixes the incorrect behavior.

Differential Revision: https://reviews.llvm.org/D116657
2022-01-05 09:31:23 -05:00
Nicolas Vasilache c7dd0bf41d [mlir][vector] NFC - Split out transfer split patterns
Differential Revision: https://reviews.llvm.org/D116648
2022-01-05 08:38:04 -05:00
Matthias Springer 6c6bba7436 [mlir][linalg][bufferize][NFC] Use RewriterBase instead of OpBuilder
This is in preparation of unifying core bufferization and Comprehensive Bufferize.

Differential Revision: https://reviews.llvm.org/D116102
2022-01-05 21:05:42 +09:00
Matthias Springer 46e316651f [mlir][linalg][bufferize][NFC] Refactor BufferizationOption ownership
Pass unique_ptr<BufferizationOption> to the bufferization. This allows the bufferization to enqueue additional PostAnalysisSteps. When running bufferization a second time, a new BufferizationOptions must be constructed.

Differential Revision: https://reviews.llvm.org/D116101
2022-01-05 20:24:54 +09:00
Nicolas Vasilache 11b67aaffb [mlir][scf] NFC - refactor the implementation of outlineIfOp
This revision refactors the implementation of outlineIfOp to expose
a finer-grain functionality `outlineSingleBlockRegion` that will be
reused in other contexts.

Differential Revision: https://reviews.llvm.org/D116591
2022-01-05 05:02:26 -05:00
Mehdi Amini 564bcf9d02 Align adaptor's generator accessors for attribute on the Op class
Each attribute has two accessor: one suffixed with `Attr` which returns the attribute itself
and one without the suffix which unwrap the attribute.
For example for a StringAttr attribute with a field named `kind`, we'll generate:

StringAttr getKindAttr();
StringRef getKind();

Differential Revision: https://reviews.llvm.org/D116466
2022-01-05 05:42:15 +00:00
Chuanqi Xu c75cedc237 [Coroutines] Set presplit attribute in Clang and mlir
This fixes bug49264.

Simply, coroutine shouldn't be inlined before CoroSplit. And the marker
for pre-splited coroutine is created in CoroEarly pass, which ran after
AlwaysInliner Pass in O0 pipeline. So that the AlwaysInliner couldn't
detect it shouldn't inline a coroutine. So here is the error.

This patch set the presplit attribute in clang and mlir. So the inliner
would always detect the attribute before splitting.

Reviewed By: rjmccall, ezhulenev

Differential Revision: https://reviews.llvm.org/D115790
2022-01-05 10:25:02 +08:00
wren romano c948922567 [mlir][sparse] Factoring out type-based function-name suffixes
Depends On D115010

This changes a couple of places that used to `return failure();` to now use `llvm_unreachable()` instead. However, `Transforms/Sparsification.cpp` should be doing the necessary type checks to ensure that those cases are in fact unreachable.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D115012
2022-01-04 16:17:55 -08:00
wren romano bc04a47038 [mlir][sparse] adding OverheadType::kIndex
Depends On D115008

This change opens the way for D115012, and removes some corner cases in `CodegenUtils.cpp`. The `SparseTensorAttrDefs.td` already specifies that we allow `0` bitwidth for the two overhead types and that it is interpreted to mean the architecture's native width.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D115010
2022-01-04 16:15:54 -08:00
wren romano 85b8d03e12 [mlir][sparse] Factoring out Transforms/CodegenUtils.{cpp,h}
This moves a bunch of helper functions from `Transforms/SparseTensorConversion.cpp` into `Transforms/CodegenUtils.{cpp,h}` so that they can be reused by `Transforms/Sparsification.cpp`, etc.

See also the dependent D115010 which cleans up some corner cases in this change.

Reviewed By: aartbik, rriddle

Differential Revision: https://reviews.llvm.org/D115008
2022-01-04 16:11:47 -08:00
Jacques Pienaar 9e365fe326 [mlir] Retain metadata for single loc fusedloc
If a fusedloc is created with a single location then no fusedloc
was previously created and single location returned instead. In the case
where there is a metadata associated with the location this results in
discarding the metadata. Instead only canonicalize where there is no
loss of information.

Differential Revision: https://reviews.llvm.org/D115605
2022-01-04 15:37:33 -08:00
Benjamin Kramer 41760a6b40 [mlir] Make Value's constructor constexpr. NFCI.
This allows clang to flag unused Values in more cases, so remove them.
2022-01-04 21:04:13 +01:00
Stella Laurenzo 7ee25bc56f [mlir][python] Add bindings for diagnostic handler.
I considered multiple approaches for this but settled on this one because I could make the lifetime management work in a reasonably easy way (others had issues with not being able to cast to a Python reference from a C++ constructor). We could stand to have more formatting helpers, but best to get the core mechanism in first.

Differential Revision: https://reviews.llvm.org/D116568
2022-01-04 11:04:37 -08:00
Alex Zinenko bc1df1fabb [mlir] Fix incorrect top-level comment in DialectSparseTensor.cpp 2022-01-04 18:38:30 +01:00
Markus Böck 2a0e05100c [mlir][LLVM] Set cleanup flag on `llvm.landingpad` when exporting to LLVM IR
Exporting a llvm.landingpad operation with the cleanup flag set is currently ignored by the export code.

Differential Revision: https://reviews.llvm.org/D116565
2022-01-04 08:19:26 +01:00
Nicolas Vasilache f68ecdd458 [mlir] Add CMake flags to properly enable Jit event listeners.
By default, the listeners do nothing unless linked in.
This revision allows the "Perf" and "Intel" Jit event listeners to be used.

The "OProfile" event listener is not enabled at this time,
the associated library structure is not well-isolated.

Differential Revision: https://reviews.llvm.org/D116552
2022-01-04 02:11:02 -05:00
Uday Bondhugula 80b3f08eee [MLIR[PDL] NFC. Fix unused variable warning in PDLToPDLInterp.cpp
NFC. Fix unused variable warning in PDLToPDLInterp.cpp.

Differential Revision: https://reviews.llvm.org/D116571
2022-01-04 08:25:02 +05:30
Stanislav Funiak 7de8488c3d [MLIR] Printing a null Value.
This diff adds support to printing a Value when it is null. We encounter this situation when debugging the PDL bytcode execution (where a null Value is perfectly valid). Currently, the AsmPrinter crashes (with an assert in a cast) when it encounters such Value.

We follow the same format used in other printed entities (e.g., null attribute).

Reviewed By: mehdi_amini, bondhugula

Differential Revision: https://reviews.llvm.org/D116084
2022-01-04 08:13:03 +05:30
Stanislav Funiak de6c82d6fd [MLIR][PDL] Generalize result type verification
Presently the result type verification checks if the type is used by a `pdl::OperationOp` inside the matcher. This is unnecessarily restrictive; the type could come from a `pdl::OperandOp or `pdl::OperandsOp` and still be inferrable.

Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116083
2022-01-04 08:11:46 +05:30
Stanislav Funiak b4130e9ead [MLIR][PDL] Integration test of multi-root matching and related fixes.
This diff adds an integration test to multi-root PDL matching. It consists of two subtests:
1) A 1-layer perceptron with split forward / backward operations.
2) A 2-layer perceptron with fused forward / backward operations.

These tests use a collection of hand-written patterns and TensorFlow operations to be matched. The first test has a DAG / SSA dominant resulting match; the second does not and is therefore stored in a graph region.

This diff also includes two bug fixes:
1) Mark the pdl_interp dialect as a dependent in the TestPDLByteCodePass. This is needed, because we create ops from that dialect as a part of the PDL-to-PDLInterp lowering.
2) Fix of the starting index in the liveness range for the ForEach operations (bug exposed by the integration test).

Reviewed By: Mogball

Differential Revision: https://reviews.llvm.org/D116082
2022-01-04 08:03:45 +05:30
Stanislav Funiak 138803e017 [MLIR][PDL] Make predicate order deterministic.
The tree merging of pattern predicates places the predicates in an unordered set. When the predicates are sorted, they are taken in the set order, not the insertion order. This results in nondeterministic behavior.

One solution to this problem would be to use `SetVector`. However, the value `SetVector` does not provide a `find` function for fast O(1) lookups and stores the predicates twice -- once in the set and once in the vector, which is undesirable, because we store patternToAnswer in each predicate. A simpler solution is to store the tie breaking ID (which follows the insertion order), and use this ID to break any ties when comparing predicates.

Reviewed By: Mogball

Differential Revision: https://reviews.llvm.org/D116081
2022-01-04 08:03:44 +05:30
Stanislav Funiak 2692eae574 [MLIR][PDL] Refactor the positions for multi-root patterns.
When the original version of multi-root patterns was reviewed, several improvements were made to the pdl_interp operations during the review process. Specifically, the "get users of a value at the specified operand index" was split up into "get users" and "compare the users' operands with that value". The iterative execution was also cleaned up to `pdl_interp.foreach`. However, the positions in the pdl-to-pdl_interp lowering were not similarly refactored. This introduced several problems, including hard-to-detect bugs in the lowering and duplicate evaluation of `pdl_interp.get_users`.

This diff cleans up the positions. The "upward" `OperationPosition` was split-out into `UsersPosition` and `ForEachPosition`, and the operand comparison was replaced with a simple predicate. In the process, I fixed three bugs:
1. When multiple roots were had the same connector (i.e., a node that they shared with a subtree at the previously visited root), we would generate a single foreach loop rather than one foreach loop for each such root. The reason for this is that such connectors shared the position. The solution for this is to add root index as an id to the newly introduced `ForEachPosition`.
2. Previously, we would use `pdl_interp.get_operands` indiscriminately, whether or not the operand was variadic. We now correctly detect variadic operands and insert `pdl_interp.get_operand` when needed.
3. In certain corner cases, we would trigger the "connector has not been traversed yet" assertion. This was caused by not inserting the values during the upward traversal correctly. This has now been fixed.

Reviewed By: Mogball

Differential Revision: https://reviews.llvm.org/D116080
2022-01-04 08:03:44 +05:30
Markus Böck c343c200ea [mlir][LLVM] Fix mapping of result values of `llvm.invoke` during export
The result value of a llvm.invoke operation is currently not mapped to the corresponding llvm::Value* when exporting to LLVM IR. This leads to any later operations using the result to crash as it receives a nullptr.

Differential Revision: https://reviews.llvm.org/D116564
2022-01-03 23:53:01 +01:00
Nicolas Vasilache f1f5a85af8 [mlir] NFC - Format ExecutionEngine.cpp 2022-01-03 17:17:29 -05:00
Alexander Belyaev 550ea385ab [mlir] Remove unnecessary canonicalization from Linalg Detensorize.cpp
After https://reviews.llvm.org/D115821 it became possible to create
`tensor<elem_type>` with a single `tensor.from_elements` operation without
collapsing tensor shape from `tensor<1xelem_type>` to `tensor<elem_type>`

Differential Revision: https://reviews.llvm.org/D115891
2022-01-03 16:33:45 +01:00
Uday Bondhugula e49c0e483f [MLIR] Fix confusing diagnostic during dialect conversion
Fix confusing diagnostic during partial dialect conversion. A failure to
legalize is not the same as an operation being illegal: for eg. an
operation neither explicity marked legal nor explicitly marked illegal
could have been generated and may have failed to legalize further. The
op isn't an illegal one per
https://mlir.llvm.org/docs/DialectConversion/#conversion-target
which is an op that is explicitly marked illegal.

Differential Revision: https://reviews.llvm.org/D116152
2022-01-03 20:13:23 +05:30
William S. Moses 21aa2a1b09 [MLIR] Create add of sub folder
Create folders for add(sub(a, b), b) -> a and add(b, sub(a, b)) -> a

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116471
2022-01-03 09:28:20 -05:00
Groverkss 4ca510b1d1 [MLIR] Remove dependency on IR for Simplex
This patch removes unnecessary dependency on IR for Simplex. This patch allows
users to use Presburger library without depending on MLIRIR.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116530
2022-01-03 16:23:11 +05:30
Kazu Hirata b788e352ab [Transforms] Use all_of instead of any_of (NFC) 2022-01-02 22:53:09 -08:00
Mehdi Amini 1461bd13c9 Revert "Define a `cppAccessorType` to const-ref in APFloatParameter and update ODS emitter to use it for verifier signatures"
This reverts commit 89af17c0c7.

This broke the gcc5 build.
2022-01-03 06:32:50 +00:00
William S. Moses 93c791839a [MLIR] Canonicalize/fold select %x, 1, 0 to extui
Two canonicalizations for select %x, 1, 0
  If the return type is i1, return simply the condition %x, otherwise extui %x to the return type.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116517
2022-01-03 01:26:28 -05:00
William S. Moses 834cf3be22 [MLIR][Arith] Canonicalize and/or with ext
Replace and(ext(a),ext(b)) with ext(and(a,b)). This both reduces one instruction, and results in the computation (and/or) being done on a smaller type.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116519
2022-01-03 01:25:30 -05:00
Mehdi Amini 78389de4d3 Add back missing return to non-void function
It was incorrectly removed accidentally in e4e463e747.
2022-01-03 06:18:25 +00:00
Mehdi Amini a5a24c9370 Remove misused RAII gil_scoped_release/gil_scoped_acquire: without name they don't have any effect
I'm not sure what is the right fix here, but adding a name to all these
lead to many segfaults.

Reviewed By: stellaraccident

Differential Revision: https://reviews.llvm.org/D116506
2022-01-03 06:17:04 +00:00
Groverkss 29120a5130 [MLIR][NFC] Fix clang-tidy errors in Analysis/Presburger/
This patch fixes clang-tidy errors related to different parameter names
in header file and source file in Analysis/Presburger/ directory.
2022-01-03 11:42:48 +05:30
Mehdi Amini 891a0d7ccd Remove dead return after return (NFC) 2022-01-03 06:03:34 +00:00
Mehdi Amini e4e463e747 Remove useless nesting blok and dead return statement in TosaToLinalg.cpp (NFC)
Flagged by Coverity.
2022-01-03 06:02:21 +00:00
Mehdi Amini a1e62aa75b Minor reflow of FloorDivSIOp/CeilDivSIOp folder to limit the number of APInt API calls (NFC)
Cache the result of the comparison in boolean, and check early for 0 to
leverage `(a < 0) == !(a > 0)`.
2022-01-03 05:55:23 +00:00
Groverkss 8506c8c13b [MLIR] Move LinearTransform to Presburger/
This patch moves LinearTransform to Presburger/ and makes it use
IntegerPolyhedron instead of FlatAffineConstraints. Also modifies its usage in
`FlatAffineConstraints::findIntegerSample` to support the changes.

This patch is part of a series of patches for moving presburger math functionality into Presburger directory.

Reviewed By: arjunp

Differential Revision: https://reviews.llvm.org/D116311
2022-01-03 11:11:09 +05:30
William S. Moses 1bb9f4e482 [MLIR] Create folders for extsi/extui
Create folders/canonicalizers for extsi/extui. Specifically,

extui(extui(x)) -> extui(x)
extsi(extsi(x)) -> extsi(x)
extsi(extui(x)) -> extui(x)

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116515
2022-01-03 00:11:23 -05:00
Mehdi Amini 89af17c0c7 Define a `cppAccessorType` to const-ref in APFloatParameter and update ODS emitter to use it for verifier signatures
This reduce an unnecessary amount of copy of non-trivial objects, like
APFloat.

Reviewed By: rriddle, jpienaar

Differential Revision: https://reviews.llvm.org/D116505
2022-01-03 04:57:11 +00:00
Mehdi Amini 717c4bf7e3 Apply clang-tidy fixes for misc-unused-parameters in MLIR AffineStructures.cpp (NFC) 2022-01-03 02:32:41 +00:00
William S. Moses 1a0a177965 [MLIR] Create fold for cmp of ext
This patch creates folds for cmpi( ext(%x : i1, iN) != 0) -> %x

In essence this matches patterns matching an extension of a boolean, that != 0, which is equivalent to the original condition.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116504
2022-01-02 19:48:52 -05:00
Mehdi Amini 56f5e4abb8 Replace raw-loop with llvm::any_of() in PresburgerSet.cpp (NFC)
Reported by clang-tidy.
2022-01-02 22:39:57 +00:00
Mehdi Amini 4f415216ca Apply clang-tidy fixes for performance-unnecessary-value-param to MLIR (NFC) 2022-01-02 22:37:13 +00:00
Mehdi Amini 337c937ddb Apply clang-tidy fixes for performance-move-const-arg to MLIR (NFC) 2022-01-02 22:36:56 +00:00
Mehdi Amini 7964568632 Apply clang-tidy fixes for readability-simplify-boolean-expr to MLIR NormalizeMemRefs.cpp (NFC) 2022-01-02 22:26:57 +00:00
Mehdi Amini e4853be2f1 Apply clang-tidy fixes for performance-for-range-copy to MLIR (NFC) 2022-01-02 22:19:56 +00:00
Mehdi Amini abb336d26b Apply clang-tidy fixes for modernize-use-equals-default to MLIR (NFC) 2022-01-02 22:18:36 +00:00
Mehdi Amini f829d62c21 Apply clang-tidy fixes for modernize-use-default-member-init to MLIR ReductionNode.cpp (NFC) 2022-01-02 22:18:21 +00:00
Mehdi Amini 5caee2176a Apply clang-tidy fixes for bugprone-argument-comment to MLIR ArithmeticOps.cpp (NFC) 2022-01-02 22:14:48 +00:00
Mehdi Amini 6786d7e4f5 Apply clang-tidy fixes for readability-simplify-boolean-expr to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116253
2022-01-02 01:59:31 +00:00
Mehdi Amini 5a1f6077ec Apply clang-tidy fixes for readability-container-size-empty for MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116252
2022-01-02 01:56:38 +00:00
Mehdi Amini f0fff1dfde Remove unused applyPad function from TosaToLinalg.cpp (NFC) 2022-01-02 01:53:18 +00:00
Mehdi Amini ad5d7ace34 Apply clang-tidy fixes for readability-const-return-type to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116251
2022-01-02 01:51:39 +00:00
Mehdi Amini 1fc096af1e Apply clang-tidy fixes for performance-unnecessary-value-param to MLIR (NFC)
Reviewed By: Mogball

Differential Revision: https://reviews.llvm.org/D116250
2022-01-02 01:45:18 +00:00
Mehdi Amini b11510d5df Apply clang-tidy fixes for modernize-use-using to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116357
2022-01-02 01:25:14 +00:00
Mehdi Amini 0ae2e9580c Apply clang-tidy fixes for modernize-use-override to MLIR (NFC)
Reviewed By: rriddle, jpienaar

Differential Revision: https://reviews.llvm.org/D116356
2022-01-02 01:24:02 +00:00
Mehdi Amini 513463b589 Apply clang-tidy fixes for llvm-qualified-auto to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116355
2022-01-02 01:21:40 +00:00
Mehdi Amini a86b957fd7 Apply clang-tidy fixes for bugprone-macro-parentheses to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116354
2022-01-02 01:20:01 +00:00
Mehdi Amini 89de9cc8a7 Apply clang-tidy fixes for performance-for-range-copy to MLIR (NFC)
Differential Revision: https://reviews.llvm.org/D116248
2022-01-02 01:13:42 +00:00
Mehdi Amini 3bab9d4eb0 Apply clang-tidy fixes for bugprone-copy-constructor-init to MLIR (NFC)
Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116245
2022-01-02 01:05:30 +00:00
Mehdi Amini ced8690d84 Apply clang-tidy fixes for bugprone-argument-comment to MLIR (NFC)
Differential Revision: https://reviews.llvm.org/D116244
2022-01-02 01:05:06 +00:00
Mehdi Amini 104a827ea6 Move LinalgDetensorize pass option from .cpp file to the .td declaration (NFC) 2022-01-01 21:19:31 +00:00
Mehdi Amini a978847e3a Use const reference for diagnostic in callback (NFC)
This isn't a "small" struct, flagged by Coverity.
2022-01-01 21:15:50 +00:00
Kazu Hirata 491b4e1faa [IR] Remove redundant return statements (NFC)
Identified by readability-redundant-control-flow.
2022-01-01 09:14:21 -08:00
Kazu Hirata 63846a634d [mlir] Remove unused "using" (NFC)
Identified by misc-unused-using-decls.
2022-01-01 09:14:19 -08:00
Mehdi Amini 07b264d1f0 Pass the LLVMTypeConverter by reference in UnrankedMemRefBuilder (NFC)
This is a fairly large structure (952B according to Coverity), it was
already passed by reference in most places but not consistently.
2022-01-01 02:01:41 +00:00
Mehdi Amini bb6109aae6 Pass the LLVMTypeConverter by reference in MemRefBuilder (NFC)
This is a fairly large structure (952B according to Coverity), it was
already passed by reference in most places but not consistently.
2022-01-01 01:56:50 +00:00
Mehdi Amini a9f13f8065 Fix a few unitialized class members in MLIR (NFC)
Flagged by Coverity.
2022-01-01 01:40:36 +00:00
Stella Laurenzo 5cd0b817e2 [mlir] Allow IntegerAttr to parse zero width integers.
https://reviews.llvm.org/D109555 added support to APInt for this, so the special case to disable it is no longer valid. It is in fact legal to construct these programmatically today, and they print properly but do not parse.

Justification: zero bit integers arise naturally in various bit reduction optimization problems, and having them defined for MLIR reduces special casing.

I think there is a solid case for i0 and ui0 being supported. I'm less convinced about si0 and opted to just allow the parser to round-trip values that already verify. The counter argument is that the proper singular value for an si0 is -1. But the counter to this counter is that the sign bit is N-1, which does not exist for si0 and it is not unreasonable to consider this non-existent bit to be 0. Various sources consider it having the singular value "0" to be the least surprising.

Reviewed By: lattner

Differential Revision: https://reviews.llvm.org/D116413
2021-12-30 20:33:00 -08:00
MaheshRavishankar 59442a5460 [mlir][Linalg] Change signature of `get(Parallel/Reduce/Window)Dims` method.
These method currently takes a SmallVector<AffineExpr> & as an
argument to return the dims as AffineExpr. This creation of
AffineExpr objects is unnecessary.

Differential Revision: https://reviews.llvm.org/D116422
2021-12-30 14:02:15 -08:00
Mogball 4943cda398 [mlir][arith] fixing dependencies on memref/arith 2021-12-30 20:39:22 +00:00
William S. Moses a6a583dae4 [MLIR] Move AtomicRMW into MemRef dialect and enum into Arith
Per the discussion in https://reviews.llvm.org/D116345 it makes sense
to move AtomicRMWOp out of the standard dialect. This was accentuated by the
need to add a fold op with a memref::cast. The only dialect
that would permit this is the memref dialect (keeping it in the standard dialect
or moving it to the arithmetic dialect would require those dialects to have a
dependency on the memref dialect, which breaks linking).

As the AtomicRMWKind enum is used throughout, this has been moved to Arith.

Reviewed By: Mogball

Differential Revision: https://reviews.llvm.org/D116392
2021-12-30 14:31:33 -05:00
Jacques Pienaar 4a8cef157b [mlir] Change SCF/Complex to prefixed (NFC)
See
https://llvm.discourse.group/t/psa-ods-generated-accessors-will-change-to-have-a-get-prefix-update-you-apis/4476
2021-12-30 09:57:51 -08:00
Nicolas Vasilache 2e69f4f012 [mlir][vector] Fix illegal vector.transfer + tensor.insert/extract_slice folding
vector.transfer operations do not have rank-reducing semantics.

Bail on illegal rank-reduction: we need to check that the rank-reduced
dims are exactly the leading dims. I.e. the following is illegal:
```
   %0 = vector.transfer_write %v, %t[0,0], %cst :
     vector<2x4xf32>, tensor<2x4xf32>
   %1 = tensor.insert_slice %0 into %tt[0,0,0][2,1,4][1,1,1] :
     tensor<2x4xf32> into tensor<2x1x4xf32>
```

Cannot fold into:
```
   %0 = vector.transfer_write %v, %t[0,0,0], %cst :
     vector<2x4xf32>, tensor<2x1x4xf32>
```
For this, check the trailing `vectorRank` dims of the insert_slice result
tensor match the trailing dims of the inferred result tensor.

Differential Revision: https://reviews.llvm.org/D116409
2021-12-30 14:55:16 +00:00
MaheshRavishankar 7df7586a0b [mlir][MemRef] Deprecate unspecified trailing offset, size, and strides semantics of `OffsetSizeAndStrideOpInterface`.
The semantics of the ops that implement the
`OffsetSizeAndStrideOpInterface` is that if the number of offsets,
sizes or strides are less than the rank of the source, then some
default values are filled along the trailing dimensions (0 for offset,
source dimension of sizes, and 1 for strides). This is confusing,
especially with rank-reducing semantics. Immediate issue here is that
the methods of `OffsetSizeAndStridesOpInterface` assumes that the
number of values is same as the source rank. This cause out-of-bounds
errors.

So simplifying the specification of `OffsetSizeAndStridesOpInterface`
to make it invalid to specify number of offsets/sizes/strides not
equal to the source rank.

Differential Revision: https://reviews.llvm.org/D115677
2021-12-29 11:18:29 -08:00
Johannes Doerfert 944aa0421c Reapply "[OpenMP][NFCI] Embed the source location string size in the ident_t"
This reverts commit 73ece231ee and
reapplies 7bfcdbcbf3 with mlir changes.
Also reverts commit 423ba12971 and
includes the unit test changes of
16da214004.
2021-12-29 01:10:38 -06:00
William S. Moses 99fc000c87 [MLIR] Expose atomicrmw and/or
LLVM (dialect and IR) have atomics for and/or. This patch enables atomic_rmw ops in the standard dialect for and/or that lower to these (in addition to the existing atomics such as addi, etc).

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116345
2021-12-29 00:23:28 -05:00
William S. Moses ca8997eb7f [MLIR] Add constant folder for fptosi and friends
This patch adds constant folds for FPToSI/FPToUI/SIToFP/UIToFP

Reviewed By: mehdi_amini, bondhugula

Differential Revision: https://reviews.llvm.org/D116321
2021-12-28 23:50:01 -05:00
Rob Suderman f0cb77d7d5 [mlir][tosa] Resubmit split tosa-to-linalg named ops out of pass
Includes dependency fix that resulted in canonicalizer pass not linking in.

Linalg named ops lowering are moved to a separate pass. This allows TOSA
canonicalizers to run between named-ops lowerings and the general TOSA
lowerings. This allows the TOSA canonicalizers to run between lowerings.

Differential Revision: https://reviews.llvm.org/D116057
2021-12-28 11:22:58 -08:00
Groverkss 5f22f248d8 [MLIR] Use IntegerPolyhedron in Simplex instead of FlatAffineConstraints
This patch replaces usage of FlatAffineConstraints in Simplex with
IntegerPolyhedron. This removes dependency of Simplex on FlatAffineConstraints
and puts it on IntegerPolyhedron, which is part of Presburger library.

Reviewed By: arjunp

Differential Revision: https://reviews.llvm.org/D116287
2021-12-27 19:06:35 +05:30
Groverkss 3f22d492ac [MLIR] Move `print()` and `dump()` from FlatAffineConstraints to IntegerPolyhedron.
This patch moves `FlatAffineConstraints::print` and
`FlatAffineConstraints::dump()` to IntegerPolyhedron.

Reviewed By: arjunp

Differential Revision: https://reviews.llvm.org/D116289
2021-12-27 18:40:49 +05:30
Mehdi Amini dabfefa490 Fix clang-tidy performance-move-const-arg in DLTI Dialect (NFC)
The const loop iterator was inhibiting the std::move().
2021-12-26 02:13:54 +00:00
Groverkss 5b2e611b73 [MLIR][FlatAffineConstraints][NFC] Move some static functions to be available to Presburger/
This patch moves some static functions from AffineStructures.cpp to
Presburger/Utils.cpp and some to be private members of FlatAffineConstraints
(which will later be moved to IntegerPolyhedron) to allow for a smoother
transition for moving FlatAffineConstraints math functionality to
Presburger/IntegerPolyhedron.

This patch is part of a series of patches for moving math functionality to
Presburger directory.

Reviewed By: arjunp, bondhugula

Differential Revision: https://reviews.llvm.org/D115869
2021-12-25 22:36:23 +05:30
Groverkss 27a0718ad0 Revert "[MLIR][FlatAffineConstraints][NFC] Move some static functions to be available to Presburger/"
This reverts commit 6c0eaefaf8.
2021-12-25 00:39:27 +05:30
Groverkss 6c0eaefaf8 [MLIR][FlatAffineConstraints][NFC] Move some static functions to be available to Presburger/
This patch moves some static functions from AffineStructures.cpp to
Presburger/Utils.cpp and some to be private members of FlatAffineConstraints
(which will later be moved to IntegerPolyhedron) to allow for a smoother
transition for moving FlatAffineConstraints math functionality to
Presburger/IntegerPolyhedron.

This patch is part of a series of patches for moving math functionality to
Presburger directory.

Reviewed By: arjunp, bondhugula

Differential Revision: https://reviews.llvm.org/D115869
2021-12-25 00:11:35 +05:30
Mogball 41a64338cc [mlir] Add getNumThreads to MLIRContext
Querying threads directly from the thread pool fails if there is no thread pool or if multithreading is not enabled. Returns 1 by default.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116259
2021-12-24 02:02:54 +00:00
Stella Laurenzo 8ff42766d1 [mlir] Use thread-pool's notion of thread count instead of requerying system.
The computed number of hardware threads can change over the life of the process based on affinity changes. Since we need a data structure that is at least as large as the maximum parallelism, it is important to use the value that was actually latched for the thread pool we will be dispatching work to.

Also adds an assert specifically for if it doesn't line up (I was getting a crash on an index into the vector).

Differential Revision: https://reviews.llvm.org/D116257
2021-12-23 17:04:45 -08:00
Mehdi Amini 735fe1da6b Revert "[mlir][tosa] Split tosa-to-linalg named ops out of pass"
This reverts commit 313de31fbb.

There is a missing CMake dependency, building with shared libraries is
broken:

55.509 [45/4/3061] Linking CXX shared library lib/libMLIRTosaToLinalg.so.14git
FAILED: lib/libMLIRTosaToLinalg.so.14git
...
TosaToLinalgPass.cpp: undefined reference to `mlir::createCanonicalizerPass()'
2021-12-24 00:09:15 +00:00
Rob Suderman 313de31fbb [mlir][tosa] Split tosa-to-linalg named ops out of pass
Linalg named ops lowering are moved to a separate pass. This allows TOSA
canonicalizers to run between named-ops lowerings and the general TOSA
lowerings. This allows the TOSA canonicalizers to run between lowerings.

Reviewed By: NatashaKnk

Differential Revision: https://reviews.llvm.org/D116057
2021-12-23 12:23:19 -08:00
Mogball 8f8c89f3cd [mlir] Remove spurious debug guard 2021-12-23 11:55:37 -08:00
Mehdi Amini e5639b3fa4 Fix more clang-tidy cleanups in mlir/ (NFC) 2021-12-22 20:53:11 +00:00
Mogball db68e6ab93 [mlir] Fix missing namespace (NFC) 2021-12-22 19:17:14 +00:00
Stanislav Funiak 9eb8e7b176 [MLIR][PDL] Clear up the terminology in the root ordering graph.
Previously, we defined a struct named `RootOrderingCost`, which stored the cost (a pair consisting of the depth of the connector and a tie breaking ID), as well as the connector itself. This created some confusion, because we would sometimes write, e.g., `cost.cost.first` (the first `cost` referring to the struct, the second one referring to the `cost` field, and `first` referring to the depth). In order to address this confusion, here we rename `RootOrderingCost` to `RootOrderingEntry` (keeping the fields and their names as-is).

This clarification exposed non-determinism in the optimal branching algorithm. When choosing the best local parent, we were previuosly only considering its depth (`cost.first`) and not the tie-breaking ID (`cost.second`). This led to non-deterministic choice of the parent when multiple potential parents had the same depth. The solution is to compare both the depth and the tie-breaking ID.

Testing: Rely on existing unit tests. Non-detgerminism is hard to unit-test.

Reviewed By: rriddle, Mogball

Differential Revision: https://reviews.llvm.org/D116079
2021-12-22 19:10:29 +00:00
Mogball 42ac4f3dc6 [mlir] Canonicalizer constructor should accept disabled/enabled patterns
There is no way to programmatically configure the list of disabled and enabled patterns in the canonicalizer pass, other than the duplicate the whole pass. This patch exposes the `disabledPatterns` and `enabledPatterns` options.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D116055
2021-12-22 19:08:31 +00:00
Adrian Kuegel 4a10457d33 [mlir][arith] Fix CmpIOP folding for vector types.
Previously, the folding assumed that it always operates on scalar types.

Differential Revision: https://reviews.llvm.org/D116151
2021-12-22 18:12:24 +01:00
Adrian Kuegel 6e9be9f7c1 [mlir][NFC] Use .empty() instead of .size() 2021-12-22 12:30:54 +01:00
Adrian Kuegel 35c0080333 [mlir][NFC] Fix typo in VectorOps.cpp 2021-12-22 12:24:43 +01:00
Butygin 28ab10f404 [mlir][memref] ReinterpretCast: allow static sizes/strides/offset where affine map expects dynamic
* There is no reason to forbid that case
* Also, user will get very unfriendly error like `expected result type with offset = -9223372036854775808 instead of 1`

Differential Revision: https://reviews.llvm.org/D114678
2021-12-21 16:20:01 +03:00
Mehdi Amini bb56c2b366 Fix clang-tidy issues in mlir/ (NFC)
Differential Revision: https://reviews.llvm.org/D115956
2021-12-21 00:50:07 +00:00
Mogball 8cb785cad1 [mlir][arith] Clean up ExpandOps pass 2021-12-20 21:59:11 +00:00
Mehdi Amini 02b6fb218e Fix clang-tidy issues in mlir/ (NFC)
Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D115956
2021-12-20 20:25:01 +00:00
Butygin c7f96d5ab1 [mlir][scf] Canonicalize nested scf.if's to scf.if + arith.and
Differential Revision: https://reviews.llvm.org/D115930
2021-12-20 21:53:03 +03:00
MaheshRavishankar 4142932a83 [mlir][Linalg] Move named op conversions out of canonicalizations.
These conversions are better suited to be applied at whole tensor
level. Applying these as canonicalizations end up triggering such
canonicalizations at all levels of the stack which might be
undesirable. For example some of the resulting code patterns wont
bufferize in-place and need additional stack buffers. Best is to be
more deliberate in when these canonicalizations apply.

Differential Revision: https://reviews.llvm.org/D115912
2021-12-20 10:19:05 -08:00
Jacques Pienaar 4fe5543b3c [mlir] Address compiler warning (NFC)
Mark overridden virtual function.
2021-12-20 09:33:30 -08:00
Jacques Pienaar c0342a2de8 [mlir] Switching accessors to prefixed form (NFC)
Makes eventual prefixing flag flip smaller change.
2021-12-20 08:03:43 -08:00
Mehdi Amini 7f9e9c7fc3 Move getAsmBlockArgumentNames from OpAsmDialectInterface to OpAsmOpInterface
This method is more suitable as an opinterface: it seems intrinsic to
individual instances of the operation instead of the dialect.
Also remove the restriction on the interface being applicable to the entry block only.

Differential Revision: https://reviews.llvm.org/D116018
2021-12-20 07:18:01 +00:00
Arjun P 4fa96b7eca [MLIR] Simplex: split some basic functionality out into a SimplexBase class
This is a purely mechanical patch moving some functionality out from the
`Simplex` class out into a `SimplexBase` class. This pavees the way for
a future patch adding support for lexicographic optimization with a class
`LexSimplex`, which will inherit from `SimplexBase`. Inheriting directly
from `Simplex` would bring many additional functions that would not work in
`LexSimplex` because it operates slighty differently from `Simplex`. So We
split out only the basic functionality it needs to inherit into `SimplexBase`.

Reviewed By: Groverkss

Differential Revision: https://reviews.llvm.org/D115831
2021-12-19 22:24:40 +05:30
bakhtiyar ec0e4545ca Make AsyncParallelForRewrite parameterizable with a cost model which drives deciding the parallelization granularity.
Reviewed By: ezhulenev, mehdi_amini

Differential Revision: https://reviews.llvm.org/D115423
2021-12-19 08:41:01 -08:00
Mehdi Amini cc4781464f Fix warning "comparison of integers of different signs" (NFC) 2021-12-18 20:38:32 +00:00
Aaron DeBattista 64f694acaf [mlir][tosa] Move tosa canonicalizers to optional optimization pass
TOSA's canonicalizers that change dense operations should be moved to a
seperate optimization pass to avoid canonicalizing to operations not supported
for relevant backends.

Reviewed By: rsuderman

Differential Revision: https://reviews.llvm.org/D115890
2021-12-16 23:33:54 -08:00
Rob Suderman 0763f12213 [mlir][tosa] Handle rescale case where shift > 63
It is possible for the shift value to exceed the number of bits. In these
cases we can just multiply by zero. This is relatively rare occurence but
should be handled.

Reviewed By: not-jenni

Differential Revision: https://reviews.llvm.org/D115779
2021-12-16 15:30:48 -08:00
not-jenni f9cefc7b90 [mlir][tosa] Add tosa.max_pool2d as no-op canonicalization
When the input and output of a pool2d op are both 1x1, it can be canonicalized to a no-op

Reviewed By: rsuderman

Differential Revision: https://reviews.llvm.org/D115908
2021-12-16 15:27:26 -08:00