Commit Graph

157 Commits

Author SHA1 Message Date
MaheshRavishankar b62f9f4407 [mlir][Linalg] Add pattern to fold linalg.tensor_reshape that add unit extent dims.
A sequence of two reshapes such that one of them is just adding unit
extent dims can be folded to a single reshape.

Differential Revision: https://reviews.llvm.org/D88057
2020-09-23 00:01:58 -07:00
Nicolas Vasilache ed229132f1 [mlir][Linalg] Uniformize linalg.generic with named ops.
This revision allows representing a reduction at the level of linalg on tensors for generic ops by uniformizing with the named ops approach.
2020-09-22 04:13:22 -04:00
Ahmed S. Taei 9b47525824 Reorder linalg.conv indexing_maps loop order
Change the indexing map to iterate over the (b, x0, x1, z0, z1, q, k) instead of (b, x0, x1, k, q, z0, z1) to evaluate the convolution expression:
Y[b, x0, x1, k] = sum(W[z0, z1, q, k] * X[b, x0 + z0, x1 + z1, q], z0, z1, q)

This allows llvm auto vectorize to work and has better locality resulting significant performance improvments

Differential Revision: https://reviews.llvm.org/D87781
2020-09-22 04:53:57 +00:00
Nicolas Vasilache 93fd30bac3 [mlir][Linalg] Evolve named ops to use assembly form and support linalg on tensors.
This revision allows representing a reduction at the level of linalg on tensors for named ops. When a structured op has a reduction and returns tensor(s), new conventions are added and documented.

As an illustration, the syntax for a `linalg.matmul` writing into a buffer is:

```
  linalg.matmul ins(%a, %b : memref<?x?xf32>, tensor<?x?xf32>)
               outs(%c : memref<?x?xf32>)
```

, whereas the syntax for a `linalg.matmul` returning a new tensor is:

```
  %d = linalg.matmul ins(%a, %b : tensor<?x?xf32>, memref<?x?xf32>)
                    init(%c : memref<?x?xf32>)
                      -> tensor<?x?xf32>
```

Other parts of linalg will be extended accordingly to allow mixed buffer/tensor semantics in the presence of reductions.
2020-09-18 06:14:30 -04:00
Hanhan Wang f16abe5f84 [mlir][Vector] Add a folder for vector.broadcast
Fold the operation if the source is a scalar constant or splat constant.

Update transform-patterns-matmul-to-vector.mlir because the broadcast ops are folded in the conversion.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D87703
2020-09-17 08:54:51 -07:00
Eugene Zhulenev 8c0dc1e38b Enable inlining for Linalg dialect
Enable inlining for Linalg dialect.

Differential Revision: https://reviews.llvm.org/D87567
2020-09-16 10:19:13 -04:00
Nicolas Vasilache e6f2f17f05 [mlir][Linalg] Refactor StructuredOpInterface - NFC
This revision refactors and cleans up a bunch of things to simplify StructuredOpInterface
before work can proceed on Linalg on tensors:
- break out pieces of the StructuredOps trait that are part of the StructuredOpInterface,
- drop referenceIterators and referenceIndexingMaps that end up being more confusing than useful,
- drop NamedStructuredOpTrait
2020-09-11 07:53:12 -04:00
Benjamin Kramer a0e0d30a29 [mlir][Linalg] Print both types for linalg.transpose
Previously only the input type was printed, and the parser applied it to
both input and output, creating an invalid transpose. Print and parse
both types, and verify that they match.

Differential Revision: https://reviews.llvm.org/D87462
2020-09-11 11:16:51 +02:00
Jakub Lichman 53ffeea6d5 [mlir][Linalg] Reduction dimensions specified in TC definition of ConvOps.
This commit specifies reduction dimensions for ConvOps. This prevents
running reduction loops in parallel and enables easier detection of kernel dimensions
which we will need later on.

Differential Revision: https://reviews.llvm.org/D87288
2020-09-09 15:17:07 +00:00
Jakub Lichman 8d35080ebb [mlir][Linalg] Wrong tile size for convolutions fixed
Sizes of tiles (subviews) are bigger by 1 than they should. Let's consider
1D convolution without batches or channels. Furthermore let m iterate over
the output and n over the kernel then input is accessed with m + n. In tiling
subview sizes for convolutions are computed by applying requested tile size
together with kernel size to the above mentioned expression thus let's say
for tile size of 2 the subview size is 2 + size(n), which is bigger by one
than it should since we move kernel only once. The problem behind it is that
range is not turned into closed interval before the composition. This commit
fixes the problem by turning ranges first into closed intervals by substracting
1 and after the composition back to half open by adding 1.

Differential Revision: https://reviews.llvm.org/D86638
2020-09-03 06:01:21 +00:00
Hanhan Wang eb4efa8832 [mlir][Linalg] Enhance Linalg fusion on generic op and tensor_reshape op.
The tensor_reshape op was only fusible only if it is a collapsing case. Now we
propagate the op to all the operands so there is a further chance to fuse it
with generic op. The pre-conditions are:

1) The producer is not an indexed_generic op.
2) All the shapes of the operands are the same.
3) All the indexing maps are identity.
4) All the loops are parallel loops.
5) The producer has a single user.

It is possible to fuse the ops if the producer is an indexed_generic op. We
still can compute the original indices. E.g., if the reshape op collapses the d0
and d1, we can use DimOp to get the width of d1, and calculate the index
`d0 * width + d1`. Then replace all the uses with it. However, this pattern is
not implemented in the patch.

Reviewed By: mravishankar

Differential Revision: https://reviews.llvm.org/D86314
2020-08-28 01:55:49 -07:00
Kazuaki Ishizaki a23d055912 [mlir] NFC: fix trivial typo under test and tools
Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D86648
2020-08-27 15:37:42 +09:00
MaheshRavishankar 5ccac05d43 [mlir][Linalg] Modify callback for getting id/nprocs in
LinalgDistribution options to allow more general distributions.

Changing the signature of the callback to send in the ranges for all
the parallel loops and expect a vector with the Value to use for the
processor-id and number-of-processors for each of the parallel loops.

Differential Revision: https://reviews.llvm.org/D86095
2020-08-18 14:04:40 -07:00
MaheshRavishankar a65a50540e [mlir][Linalg] Canonicalize tensor_reshape(splat-constant) -> splat-constant.
When the operand to the linalg.tensor_reshape op is a splat constant,
the result can be replaced with a splat constant of the same value but
different type.

Differential Revision: https://reviews.llvm.org/D86117
2020-08-18 08:17:09 -07:00
MaheshRavishankar 41d4120017 [mlir][Linalg] Allow distribution `scf.parallel` loops generated in
Linalg to processors.

This changes adds infrastructure to distribute the loops generated in
Linalg to processors at the time of generation. This addresses use
case where the instantiation of loop is done just to distribute
them. The option to distribute is added to TilingOptions for now and
will allow specifying the distribution as a transformation option,
just like tiling and promotion are specified as options.

Differential Revision: https://reviews.llvm.org/D85147
2020-08-10 14:52:17 -07:00
Nicolas Vasilache 3110e7b077 [mlir] Introduce AffineMinSCF folding as a pattern
This revision adds a folding pattern to replace affine.min ops by the actual min value, when it can be determined statically from the strides and bounds of enclosing scf loop .

This matches the type of expressions that Linalg produces during tiling and simplifies boundary checks. For now Linalg depends both on Affine and SCF but they do not depend on each other, so the pattern is added there.
In the future this will move to a more appropriate place when it is determined.

The canonicalization of AffineMinOp operations in the context of enclosing scf.for and scf.parallel proceeds by:
  1. building an affine map where uses of the induction variable of a loop
  are replaced by `%lb + %step * floordiv(%iv - %lb, %step)` expressions.
  2. checking if any of the results of this affine map divides all the other
  results (in which case it is also guaranteed to be the min).
  3. replacing the AffineMinOp by the result of (2).

The algorithm is functional in simple parametric tiling cases by using semi-affine maps. However simplifications of such semi-affine maps are not yet available and the canonicalization does not succeed yet.

Differential Revision: https://reviews.llvm.org/D82009
2020-08-07 14:30:38 -04:00
Nicolas Vasilache 54fafd17a7 [mlir][Linalg] Introduce canonicalization to remove dead LinalgOps
When any of the memrefs in a structured linalg op has a zero dimension, it becomes dead.
This is consistent with the fact that linalg ops deduce their loop bounds from their operands.

Note however that this is not the case for the `tensor<0xelt_type>` which is a special convention
that must be lowered away into either `memref<elt_type>` or just `elt_type` before this
canonicalization can kick in.

Differential Revision: https://reviews.llvm.org/D85413
2020-08-06 06:08:46 -04:00
Alex Zinenko ec1f4e7c3b [mlir] switch the modeling of LLVM types to use the new mechanism
A new first-party modeling for LLVM IR types in the LLVM dialect has been
developed in parallel to the existing modeling based on wrapping LLVM `Type *`
instances. It resolves the long-standing problem of modeling identified
structure types, including recursive structures, and enables future removal of
LLVMContext and related locking mechanisms from LLVMDialect.

This commit only switches the modeling by (a) renaming LLVMTypeNew to LLVMType,
(b) removing the old implementaiton of LLVMType, and (c) updating the tests. It
is intentionally minimal. Separate commits will remove the infrastructure built
for the transition and update API uses where appropriate.

Depends On D85020

Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D85021
2020-08-04 14:29:25 +02:00
Jakub Lichman eef1bfb2d2 [mlir][Linalg] Conv {1,2,3}D ops defined with TC syntax
Replaced definition of named ND ConvOps with tensor comprehension
syntax which reduces boilerplate code significantly. Furthermore,
new ops to support TF convolutions added (without strides and dilations).

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D84628
2020-07-31 13:20:17 +02:00
Jakub Lichman 1aaf8aa53d [mlir][Linalg] Conv1D, Conv2D and Conv3D added as named ops
This commit is part of a greater project which aims to add
full end-to-end support for convolutions inside mlir. The
reason behind having conv ops for each rank rather than
having one generic ConvOp is to enable better optimizations
for every N-D case which reflects memory layout of input/kernel
buffers better and simplifies code as well. We expect plain linalg.conv
to be progressively retired.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D83879
2020-07-29 16:39:56 +02:00
lorenzo chelini 946be75b9e [MLIR][Linalg] Retire C++ DotOp in favor of a linalg-ods-gen'd op
- replace DotOp, now that DRR rules have been dropped.

- Capture arguments mismatch in the parser. The number of parsed arguments must
  equal the number of expected arguments.

Reviewed By: ftynse, nicolasvasilache

Differential Revision: https://reviews.llvm.org/D82952
2020-07-28 12:34:19 +02:00
MaheshRavishankar 8f6e84ba7b [mlir][Linalg] Enable fusion of std.constant (producer) with
linalg.indexed_generic (consumer) with tensor arguments.

The implementation of fusing std.constant producer with a
linalg.indexed_generic consumer was already in place. It is exposed
with this change. Also cleaning up some of the patterns that implement
the fusion to not be templated, thereby avoiding lot of conditional
checks for calling the right instantiation.

Differential Revision: https://reviews.llvm.org/D84566
2020-07-27 09:51:20 -07:00
MaheshRavishankar 4ff48db68d [mlir][Linalg] Fixing bug in subview size computation in Linalg tiling.
The `makeTiledViews` did not use the sizes of the tiled views based on
the result of the loop bound inference computation. This manifested as
an error in computing tile sizes with convolution where not all the
result expression of concatenated affine maps are simple
AffineDimExpr.

Differential Revision: https://reviews.llvm.org/D84366
2020-07-23 11:09:55 -07:00
Jakub Lichman 919922b0c2 [mlir] Added verification check for linalg.conv to ensure memrefs are of rank > 2
linalg.conv does not support memrefs with rank smaller than 3 as stated here:
https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/nn/convolution

However it does not verify it and thus crashes with "LLVM ERROR: out of memory"
error for 1D case and "nWin > 0 && "expected at least one window dimension"" assertion
for 2D case. This commit adds check for that in the verification method.

Differential Revision: https://reviews.llvm.org/D84317
2020-07-23 12:27:05 +02:00
Jakub Lichman e4dd964df0 [mlir] Loop bounds inference in linalg.generic op improved to support bounds for convolution
Loop bound inference is right now very limited as it supports only permutation maps and thus
it is impossible to implement convolution with linalg.generic as it requires more advanced
loop bound inference. This commits solves it for the convolution case.

Depends On D83158

Differential Revision: https://reviews.llvm.org/D83191
2020-07-23 11:01:54 +02:00
Thomas Raoux a1b9fb220f [mlir][linalg] Add vectorization transform for CopyOp
CopyOp get vectorized to vector.transfer_read followed by vector.transfer_write

Differential Revision: https://reviews.llvm.org/D83739
2020-07-22 12:40:42 -07:00
Benjamin Kramer bf561dd2eb [mlir][Vector] Vectorize integer matmuls
The underlying infrastructure supports this already, just add the
pattern matching for linalg.generic.

Differential Revision: https://reviews.llvm.org/D84335
2020-07-22 19:39:56 +02:00
Jakub Lichman f9c8febc52 [mlir] Added support for symbols inside linalg.generic and map concatenation
This commit adds functionality needed for implementation of convolutions with
linalg.generic op. Since linalg.generic right now expects indexing maps to be
just permutations, offset indexing needed in convolutions is not possible.
Therefore in this commit we address the issue by adding support for symbols inside
indexing maps which enables more advanced indexing. The upcoming commit will
solve the problem of computing loop bounds from such maps.

Differential Revision: https://reviews.llvm.org/D83158
2020-07-20 19:20:47 +02:00
Nicolas Vasilache 512da70be7 [mlir][Vector] Degrade masking information when forwarding linalg.copy to vector.transfer
Summary:
linalg.copy + linalg.fill can be used to create a padded local buffer.
The `masked` attribute is only valid on this padded buffer.
When forwarding to vector.transfer ops, the attribute must be reset
conservatively.

Differential Revision: https://reviews.llvm.org/D83782
2020-07-15 02:32:45 -04:00
Thomas Raoux 6d5aeb0dce [mlir][linalg] Improve aliasing approximation for hoisting transfer read/write
Improve the logic deciding if it is safe to hoist vector transfer read/write
out of the loop. Change the logic to prevent hoisting operations if there are
any unknown access to the memref in the loop no matter where the operation is.
For other transfer read/write in the loop check if we can prove that they
access disjoint memory and ignore them in this case.

Differential Revision: https://reviews.llvm.org/D83538
2020-07-10 14:55:04 -07:00
Nicolas Vasilache 56c638b5c1 [mlir][Linalg] Generalize Vectorization of Linalg contractions
This revision adds support for vectorizing named and generic contraction ops to vector.contract. Cases in which the memref is 0-D are special cased to emit std.load/std.store instead of vector.transfer. Relevant tests are added.

Differential revision: https://reviews.llvm.org/D83307
2020-07-10 10:28:34 -04:00
River Riddle 9db53a1827 [mlir][NFC] Remove usernames and google bug numbers from TODO comments.
These were largely leftover from when MLIR was a google project, and don't really follow LLVM guidelines.
2020-07-07 01:40:52 -07:00
Uday Bondhugula 6d6d5db251 [MLIR][Linalg] Generate the right type of load/store when lowering max/min pooling ops
While lowering min/max pooling ops to loops, generate the right kind of
load/stores (std or affine) instead of always generating std
load/stores.

Differential Revision: https://reviews.llvm.org/D83080
2020-07-04 14:55:02 +05:30
Nicolas Vasilache 7d9518c800 [mlir][Linalg] Add an option to use Alloca instead of malloc/free pairs.
Summary: A relevant test is also added.

Subscribers: mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, stephenneuendorffer, Joonsoo, grosul1, Kayjukh, jurahul, msifontes

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D82959
2020-07-01 09:44:01 -04:00
lorenzo chelini e31e8f1ed5 [MLIR][Linalg] Retire C++ MatvecOp in favor of a linalg-ods-gen'd op
Replace C++ MatvecOp, now that DRR rules have been dropped.

Differential Revision: https://reviews.llvm.org/D82007
2020-06-18 11:36:49 +02:00
Nicolas Vasilache eae76faeea [mlir][Linalg] Retire C++ MatmulOp in favor of a linalg-ods-gen'd op.
Summary:
This revision replaces MatmulOp, now that DRR rules have been dropped.
This revision also fixes minor parsing bugs and a plugs a few holes to get e2e paths working (e.g. library call emission).

During the replacement the i32 version had to be dropped because only the EDSC operators +, *, etc support type inference.

Deciding on a type-polymorphic behavior, and implementing it, is left for future work.

Reviewers: aartbik

Subscribers: mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul, msifontes

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D81935
2020-06-16 10:46:35 -04:00
Kirill Bobyrev 9b72b47ed6 Revert "[mlir][Linalg] Retire C++ MatmulOp in favor of a linalg-ods-gen'd op."
This reverts commit 8c6c49f293.

As discussed offline, this patch breaks internal builds and tests so I'm
reverting it for now.
2020-06-16 11:02:28 +02:00
Nicolas Vasilache 8c6c49f293 [mlir][Linalg] Retire C++ MatmulOp in favor of a linalg-ods-gen'd op.
This revision replaces MatmulOp, now that DRR rules have been dropped.
This revision also fixes minor parsing bugs and a plugs a few holes to get e2e paths working (e.g. library call emission).

During the replacement the i32 version had to be dropped because only the EDSC operators +, *, etc support type inference.

Deciding on a type-polymorphic behavior, and implementing it, is left for future work.

Differential Revision: https://reviews.llvm.org/D79762
2020-06-15 18:14:15 -04:00
Mehdi Amini 95371ce9c2 Enable FileCheck -enable-var-scope by default in MLIR test
This option avoids to accidentally reuse variable across -LABEL match,
it can be explicitly opted-in by prefixing the variable name with $

Differential Revision: https://reviews.llvm.org/D81531
2020-06-12 00:43:09 +00:00
Frederik Gossen 904f91db5f [MLIR][Standard] Make the `dim` operation index an operand.
Allow for dynamic indices in the `dim` operation.
Rather than an attribute, the index is now an operand of type `index`.
This allows to apply the operation to dynamically ranked tensors.
The correct lowering of dynamic indices remains to be implemented.

Differential Revision: https://reviews.llvm.org/D81551
2020-06-10 13:54:47 +00:00
Mehdi Amini d31c9e5a46 Change filecheck default to dump input on failure
Having the input dumped on failure seems like a better
default: I debugged FileCheck tests for a while without knowing
about this option, which really helps to understand failures.

Remove `-dump-input-on-failure` and the environment variable
FILECHECK_DUMP_INPUT_ON_FAILURE which are now obsolete.

Differential Revision: https://reviews.llvm.org/D81422
2020-06-09 18:57:46 +00:00
Nicolas Vasilache 6953cf6502 [mlir][Linalg] Add a hoistRedundantVectorTransfers helper function
This revision adds a helper function to hoist vector.transfer_read /
vector.transfer_write pairs out of immediately enclosing scf::ForOp
iteratively, if the following conditions are true:
   1. The 2 ops access the same memref with the same indices.
   2. All operands are invariant under the enclosing scf::ForOp.
   3. No uses of the memref either dominate the transfer_read or are
   dominated by the transfer_write (i.e. no aliasing between the write and
   the read across the loop)

To improve hoisting opportunities, call the `moveLoopInvariantCode` helper
function on the candidate loop above which to hoist. Hoisting the transfers
results in scf::ForOp yielding the value that originally transited through
memory.

This revision additionally exposes `moveLoopInvariantCode` as a helper in
LoopUtils.h and updates SliceAnalysis to support return scf::For values and
allow hoisting across multiple scf::ForOps.

Differential Revision: https://reviews.llvm.org/D81199
2020-06-05 06:50:24 -04:00
Uday Bondhugula 0f6999af88 [MLIR] Update linalg.conv lowering to use affine load in the absence of padding
Update linalg to affine lowering for convop to use affine load for input
whenever there is no padding. It had always been using std.loads because
max in index functions (needed for non-zero padding if not materializing
zeros) couldn't be represented in the non-zero padding cases.

In the future, the non-zero padding case could also be made to use
affine - either by materializing or using affine.execute_region. The
latter approach will not impact the scf/std output obtained after
lowering out affine.

Differential Revision: https://reviews.llvm.org/D81191
2020-06-05 12:28:30 +05:30
Nicolas Vasilache 3463d9835b [mlir][Linalg] Add a hoistViewAllocOps helper function
This revision adds a helper function to hoist alloc/dealloc pairs and
alloca op out of immediately enclosing scf::ForOp if both conditions are true:
   1. all operands are defined outside the loop.
   2. all uses are ViewLikeOp or DeallocOp.

This is now considered Linalg-specific and will be generalized on a per-need basis.

Differential Revision: https://reviews.llvm.org/D81152
2020-06-04 18:59:03 -04:00
Hanhan Wang 27fca57546 [mlir][Linalg] Add support for fusion between indexed_generic ops and tensor_reshape ops
Summary:
The fusion for tensor_reshape is embedding the information to indexing maps,
thus the exising pattenr also works for indexed_generic ops.

Depends On D80347

Differential Revision: https://reviews.llvm.org/D80348
2020-06-03 14:59:47 -07:00
Hanhan Wang cc11ceda16 [mlir][Linalg] Add support for fusion between indexed_generic ops and generic ops on tensors.
Summary:
Different from the fusion between generic ops, indices are involved. In this
context, we need to re-map the indices for producer since the fused op is built
on consumer's perspective. This patch supports all combination of the fusion
between indexed_generic ops and generic ops, which includes tests case:
  1) generic op as producer and indexed_generic op as consumer.
  2) indexed_generic op as producer and generic op as consumer.
  3) indexed_generic op as producer and indexed_generic op as consumer.

Differential Revision: https://reviews.llvm.org/D80347
2020-06-03 14:58:43 -07:00
Nicolas Vasilache e349fb70a2 [mlir][Linalg] NFC - Make markers use Identifier instead of StringRef
Summary: This removes string ownership worries by putting everything into the context and allows more constructing identifiers programmatically.

Reviewers: ftynse

Subscribers: mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D81027
2020-06-03 05:52:32 -04:00
Nicolas Vasilache 9534192c3b [mlir][Linalg] Make contraction vectorization use vector transfers
This revision replaces the load + vector.type_cast by appropriate vector transfer
operations. These play more nicely with other vector abstractions and canonicalization
patterns and lower to load/store with or without masks when appropriate.

Differential Revision: https://reviews.llvm.org/D80809
2020-05-29 15:03:46 -04:00
Nicolas Vasilache 1ee114322c [mlir][Linalg][Vector] Add forwarding patterns between linalg.copy and vector.transfer
This revision adds custom rewrites for patterns that arise during linalg structured
ops vectorization. These patterns allow the composition of linalg promotion,
vectorization and removal of redundant copies.

The patterns are voluntarily limited and restrictive atm.
More robust behavior will be implemented once more powerful side effect modeling and analyses are available on view/subview.

On the transfer_read side, the following pattern is rewritten:
```
   %alloc = ...
   [optional] %view = std.view %alloc ...
   %subView = subview %allocOrView ...
   [optional] linalg.fill(%allocOrView, %cst) ...
   ...
   linalg.copy(%in, %subView) ...
   vector.transfer_read %allocOrView[...], %cst ...
```
into
```
   [unchanged] %alloc = ...
   [unchanged] [optional] %view = std.view %alloc ...
   [unchanged] [unchanged] %subView = subview %allocOrView ...
   ...
   vector.transfer_read %in[...], %cst ...
```

On the transfer_write side, the following pattern is rewriten:
```
   %alloc = ...
   [optional] %view = std.view %alloc ...
   %subView = subview %allocOrView...
   ...
   vector.transfer_write %..., %allocOrView[...]
   linalg.copy(%subView, %out)
```

Differential Revision: https://reviews.llvm.org/D80728
2020-05-29 08:08:34 -04:00
MaheshRavishankar 2b0c8546ac [mlir][Linalg] Add pass to remove unit-extent dims from tensor
operands of Generic ops.

Unit-extent dimensions are typically used for achieving broadcasting
behavior. The pattern added (along with canonicalization patterns
added previously) removes the use of unit-extent dimensions, and
instead uses a more canonical representation of the computation.  This
new pattern is not added as a canonicalization for now since it
entails adding additional reshape operations. A pass is added to
exercise these patterns, along with an API entry to populate a
patterns list with these patterns.

Differential Revision: https://reviews.llvm.org/D79766
2020-05-28 11:06:47 -07:00