Commit Graph

175 Commits

Author SHA1 Message Date
Adrian Kuegel 0f36979e2c Integrate LLVM at llvm/llvm-project@220de1f32a
Updates LLVM usage to match
[220de1f32add](https://github.com/llvm/llvm-project/commit/220de1f32add)

PiperOrigin-RevId: 337270545
2020-10-15 02:54:24 -07:00
Rahul Joshi f6b4e6758a Add GPU specific LMHLO level ops
- Introduce operations in a new lmhlo_gpu dialect that map to GPU library function calls
  in the XLA:GPU backend.
- Add basic unit tests as well.

PiperOrigin-RevId: 337132166
2020-10-14 11:23:55 -07:00
A. Unique TensorFlower 15fe5ca737 Integrate LLVM at llvm/llvm-project@6713332fdd
Updates LLVM usage to match
[6713332fddb7](https://github.com/llvm/llvm-project/commit/6713332fddb7)

PiperOrigin-RevId: 336988895
2020-10-13 17:21:14 -07:00
Adrian Kuegel 53d16d9f8e Integrate LLVM at llvm/llvm-project@93377888ae
Updates LLVM usage to match
[93377888ae89](https://github.com/llvm/llvm-project/commit/93377888ae89)

PiperOrigin-RevId: 336669842
2020-10-12 08:42:31 -07:00
Tres Popp f6af1fc134 Support hlo to lhlo buffer placement through shape.assuming ops.
PiperOrigin-RevId: 336287728
2020-10-09 07:14:12 -07:00
Stephan Herhut d986bd7ad7 Use tensor_cast instead of mhlo::reshape in the lowering of unranked binary operations.
We know that the value already is a scalar and we just want to update the type, so no need to reshape anything.

PiperOrigin-RevId: 336252315
2020-10-09 01:46:48 -07:00
A. Unique TensorFlower 3736c5542f [MLIR][KernelGen] Fix unranked codegeneration in kernel generator
PiperOrigin-RevId: 335847086
2020-10-07 05:39:55 -07:00
Lucy Fox 84277ed784 Remove unused `using` statements.
PiperOrigin-RevId: 335517711
2020-10-05 15:53:59 -07:00
A. Unique TensorFlower a33441907c [MLIR][KernelGen] Add E2E test for `tf.Acos`
PiperOrigin-RevId: 335419310
2020-10-05 08:14:36 -07:00
A. Unique TensorFlower bae0815ef0 [MLIR][KernelGen] Legalize `atan` to approximation
PiperOrigin-RevId: 335417836
2020-10-05 08:05:52 -07:00
A. Unique TensorFlower 7f84a86cf5 [MLIR][KerneGen] Lower `tf.Atan` all the way to LLVM
PiperOrigin-RevId: 335394668
2020-10-05 05:07:13 -07:00
Alexander Belyaev d927e32451 [HLO] Clean-up dynamic allocation in hlo-legalize-to-lhlo pass.
PiperOrigin-RevId: 335385243
2020-10-05 03:55:24 -07:00
A. Unique TensorFlower 7367eac074 Add folder for mhlo::remainder
PiperOrigin-RevId: 335372628
2020-10-05 02:20:01 -07:00
Adrian Kuegel 3eb767b43d Add missing lowering step for IsFiniteOp.
Also add a BUILD target for generating the GPU kernel.

PiperOrigin-RevId: 334993362
2020-10-02 03:09:00 -07:00
Chao Xie 5f303440da [MLIR][KerneGen] Lower `tf.Atan` all the way to LLVM
PiperOrigin-RevId: 334843070
2020-10-01 10:25:51 -07:00
A. Unique TensorFlower 458e861254 [MLIR][KerneGen] Lower `tf.Atan` all the way to LLVM
PiperOrigin-RevId: 334810730
2020-10-01 07:38:40 -07:00
A. Unique TensorFlower 049ca060a1 [MLIR][KernelGen] Legalize `atan2` to approximation
Legalize `atan2` analogously to XLA.  `atan2` is first reduced to `atan` on the
interval [-1, 1] and subsequently approximated.  This CL also adds e2e tests for
trigonometric approximations.

PiperOrigin-RevId: 334794336
2020-10-01 05:34:48 -07:00
A. Unique TensorFlower 4b1809784a Support collapse_slice_dims in the mhlo.gather->mhlo.slice canonicalizer
PiperOrigin-RevId: 334774763
2020-10-01 02:46:49 -07:00
Rahul Joshi bce128b070 Introduce CustomCall operation in LHLO Dialect
- And add conversion from MHLO CustomCall to LHLO CustomCall
- According to XLA documentation, the called function should not be side effecting,
  so marking the argument MemRefs as MemRead.

PiperOrigin-RevId: 334737196
2020-09-30 20:56:18 -07:00
Benjamin Kramer dfe64d3958 Implement InferShapedTypeOpInterface for mhlo.complex
Binary companion for 8bcd33e4b7

PiperOrigin-RevId: 334651523
2020-09-30 12:14:15 -07:00
Benjamin Kramer c8919f8419 Implement InferShapedTypeOpInterface and use inferReturnTypes for mhlo.imag and mhlo.real
This makes the lhlo lowering work with dynamic shapes.

PiperOrigin-RevId: 334553472
2020-09-30 02:02:26 -07:00
Ahmed S. Taei 39389587d2 {mhlo.is_finite, lmhlo.is_finite} -> linalg.generic conversion
PiperOrigin-RevId: 334414295
2020-09-29 10:50:42 -07:00
Benjamin Kramer 6459f12235 Lower mhlo.not to a xor with all ones
PiperOrigin-RevId: 334361499
2020-09-29 05:59:40 -07:00
Robert Suderman 26ac5baae4 Make mhlo.sort return variadic results instead of a tuple
Tuple is only used on XLA's sort to return multiple inputs. MLIR supports
multiple inputs, switch to a tuple return.

PiperOrigin-RevId: 334226937
2020-09-28 13:32:23 -07:00
Ahmed S. Taei 9c6640cbb6 Only apply GeneralDotOpLoweringPatterns for static shaped inputs
PiperOrigin-RevId: 333439680
2020-09-23 21:42:01 -07:00
Robert Suderman 233f1a8a1a Folders for mhlo.compare
Constant evaluation of compare for the case where inputs are either the same
variable or the values are constant.

PiperOrigin-RevId: 333342328
2020-09-23 12:03:48 -07:00
A. Unique TensorFlower 08e0d09463 [MLIR][KernelGen] Rename `legalize-tanh-to-approximation` to `legalize-trigonometric-to-approximation`
To add more approximation lowerings in the future, generalize the pass name.

PiperOrigin-RevId: 333340075
2020-09-23 11:53:45 -07:00
Benjamin Kramer dd92c8ef61 Integrate LLVM at llvm/llvm-project@7e78d89052
Updates LLVM usage to match
[7e78d89052b1](https://github.com/llvm/llvm-project/commit/7e78d89052b1)

PiperOrigin-RevId: 333090785
2020-09-22 09:08:23 -07:00
Mehdi Amini 7abd557a61 Add a header for table-gen generated pass for MHLO and use it in SinkConstantsToControlFlowPass
A non globally registered pass should define `getName()` in order to generate correct crash reproducers.
This is something we get "for free" when using the TableGen generated base class.
We should also migrate the other passes to the same mechanism and remove the static
global registration.

PiperOrigin-RevId: 332976907
2020-09-21 18:01:57 -07:00
A. Unique TensorFlower 4002077261 [MLIR][KernelGen] Lower `tf.Sinh` to MLHLO
PiperOrigin-RevId: 332425724
2020-09-18 04:27:07 -07:00
A. Unique TensorFlower 2fbbbe9cf1 [MLIR][KernelGen] Lower `tf.Acos` to LMHLO.
- Add ranked code generation for `mhlo.compare/select`
- Add bufferization for `tensor_cast`
- Add lowerings for `Atan2Op`

PiperOrigin-RevId: 332407734
2020-09-18 01:40:18 -07:00
A. Unique TensorFlower b1fd4d27cf [MLIR][KernelGen] Implement InferShapedTypeOpInterface for `mhlo.compare/select`
PiperOrigin-RevId: 332227340
2020-09-17 07:10:10 -07:00
Hanhan Wang b29dd5ef8f Fix a bug in the case check of reshape op lowering.
PiperOrigin-RevId: 332044191
2020-09-16 11:06:16 -07:00
A. Unique TensorFlower 69b80d8deb [MLIR] Extend unranked transformation to CHLO dialect
PiperOrigin-RevId: 332026604
2020-09-16 09:49:18 -07:00
Stephan Herhut 2aa07b0091 Insert explicit casts to model extra shape knowledge for unranked chlo transform
When transforming unranked binary operations from CHLO to HLO, we insert `shape.broadcast` operations. Due to context, we know that the result of the `shape.broadcast` operation has a static shape. Instead of modelling this in the type of the broadcast operation itself, which is illegal,  we now use an explicit cast.

PiperOrigin-RevId: 331989879
2020-09-16 06:14:49 -07:00
A. Unique TensorFlower 1880f87737 Integrate LLVM at llvm/llvm-project@e1669843f2
Updates LLVM usage to match
[e1669843f2aa](https://github.com/llvm/llvm-project/commit/e1669843f2aa)

PiperOrigin-RevId: 331987679
2020-09-16 05:58:08 -07:00
Hanhan Wang 1800f44a29 Add sqrt folder.
PiperOrigin-RevId: 331974344
2020-09-16 04:04:09 -07:00
A. Unique TensorFlower a6fdebdc6c [MLIR] Lower `chlo.constant_like` to MHLO
Lower `chlo.constant_like` to a constant and, if needed, a broadcast.

PiperOrigin-RevId: 331964137
2020-09-16 02:41:44 -07:00
A. Unique TensorFlower da43c8596b [MLIR] Simplify and generalize `transform-unranked-hlo`
This refactoring allows to support a wider range of n-ary operations in future
changes.

PiperOrigin-RevId: 331953362
2020-09-16 01:13:23 -07:00
Robert Suderman dff98f8cd5 Folder for mhlo.negate
PiperOrigin-RevId: 331592342
2020-09-14 11:38:44 -07:00
A. Unique TensorFlower 48022987ce [XLA][MLIR] Lower `tf.Tan` and `tf.Sin` to MLHLO
Add `tan` op and lowering to CHLO dialect, move CHLO lowerings to
`chlo_legalize_to_hlo_patterns` and extend missing patterns.

PiperOrigin-RevId: 331506094
2020-09-14 02:34:52 -07:00
Ehsan Toosi ce1c8a1ebc [MLIR][LHLO] Replace lhlo-copy-removal pass with mlir-copy-removal pass
Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/43137

This PR removes lhlo-copy-removal pass entirely and replace its usages with ```mlir::createCopyRemovalPass()```.

--
7ce1a06f507c8db46c6d7b43c7870cf56002e18e by Ehsan Toosi <ehsan.nadjaran_toosi@dfki.de>:

[mlir][lhlo] Replace lhlo-copy-removal pass with mlir-copy-removal pass

COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/tensorflow/pull/43137 from dfki-ehna:using_mlir_copy_removal 7ce1a06f507c8db46c6d7b43c7870cf56002e18e
PiperOrigin-RevId: 331498501
2020-09-14 01:22:19 -07:00
A. Unique TensorFlower a7a7184eb6 [XLA][MLIR] Lower `tf.Tan` and `tf.Sin` to MLHLO
Add `tan` op and lowering to CHLO dialect, move CHLO lowerings to
`chlo_legalize_to_hlo_patterns` and extend missing patterns.

PiperOrigin-RevId: 331128170
2020-09-11 05:05:58 -07:00
A. Unique TensorFlower 90927f6b53 [XLA][MLIR] Lower `tf.Tan` and `tf.Sin` to MLHLO
Add `tan` op and lowering to CHLO dialect, move CHLO lowerings to
`chlo_legalize_to_hlo_patterns` and extend missing patterns.

PiperOrigin-RevId: 331125286
2020-09-11 04:39:28 -07:00
Robert Suderman d0c8d17373 Update mhlo.concatenate inferred return type for dynamics
MHLO concatenate should support dynamic inputs. Its possible that the output
shape can be inferred from a dimension in one input that is not dynamic in
another.

PiperOrigin-RevId: 331054181
2020-09-10 17:46:11 -07:00
Robert Suderman 6eefb07767 Fix zero attr mistake
PiperOrigin-RevId: 331050555
2020-09-10 17:21:39 -07:00
A. Unique TensorFlower b22f2f0eea Integrate LLVM at llvm/llvm-project@52f0837778
Updates LLVM usage to match
[52f0837778b6](https://github.com/llvm/llvm-project/commit/52f0837778b6)

PiperOrigin-RevId: 330939173
2020-09-10 08:14:23 -07:00
Ehsan Toosi d599485e06 PR #43069: [hlo] Unbreak hlo-legalize-to-lhlo test
Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/43069

The legalization of mlhlo.ReturnOp to lhlo.TerminatorOp by using BufferAssignmentReturnOpConverter fails since the Memref typed results (or the Memref typed operands of Return operation) are set to stay as results after legalization but lhlo.TerminatorOp doesn't accept any operands. Therefore, BufferAssignmentReturnOpConverter must be replaced with a manual conversion that removes all operands of mlhlo.ReturnOp and inserts copy operations in their places.
Copybara import of the project:

--
8be0435b0147263c3872bedec58fd215f784b450 by Ehsan Toosi <ehsan.nadjaran_toosi@dfki.de>:

[hlo] Unbreak hlo-legalize-to-lhlo test

COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/tensorflow/pull/43069 from dfki-ehna:fix_hlo_legalize_to_lhlo_test 8be0435b0147263c3872bedec58fd215f784b450
PiperOrigin-RevId: 330907602
2020-09-10 04:09:03 -07:00
A. Unique TensorFlower f46ba09653 Integrate LLVM at llvm/llvm-project@4964d75d70
Updates LLVM usage to match
[4964d75d7078](https://github.com/llvm/llvm-project/commit/4964d75d7078)

PiperOrigin-RevId: 330713009
2020-09-09 06:50:26 -07:00
Robert Suderman 81d51d810b Bitwise and/or/xor folders
Includes both and/or/xor on same inputs, constant all ones/zeros single arg
folder, and constant input folders.

PiperOrigin-RevId: 330610858
2020-09-08 16:27:13 -07:00
Mehdi Amini 9d4273b5a7 Delete mhlo static dialect registration
This is obsolete now.

PiperOrigin-RevId: 330605210
2020-09-08 15:57:36 -07:00
Alexander Belyaev ebc7992d31 [MLIR][KERNEL_GEN] Add a library to lower kernels with the host side.
* Unified TF->Cubin and TF->Kernel_with_host side lowering in `kernel_creator.h|cc`
* Added a pass that attaches GPU binary blob to GPUModuleOp
* Refactored most of the code.
* Added tf_to_kernel binary that emits obj file

PiperOrigin-RevId: 330494488
2020-09-08 06:06:29 -07:00
A. Unique TensorFlower 73b4861f2c Canonicalize mhlo.gather to mhlo.slice if it has a single set of constant indices
PiperOrigin-RevId: 330380755
2020-09-07 07:28:46 -07:00
Eugene Burmako dde1ed56cc Add support for legalizing mhlo.slice to lmhlo.slice
PiperOrigin-RevId: 330153599
2020-09-04 21:43:01 -07:00
Eugene Burmako 2a9d7ac084 Add support for legalizing lmhlo.transpose to linalg.generic
PiperOrigin-RevId: 330130704
2020-09-04 15:37:42 -07:00
Eugene Burmako f5d12604ed Add support for legalizing mhlo.transpose to lmhlo.transpose
PiperOrigin-RevId: 330127758
2020-09-04 14:59:08 -07:00
A. Unique TensorFlower 801262688c Integrate LLVM at llvm/llvm-project@1426ac0482
Updates LLVM usage to match
[1426ac048295](https://github.com/llvm/llvm-project/commit/1426ac048295)

PiperOrigin-RevId: 329839640
2020-09-02 20:03:41 -07:00
A. Unique TensorFlower 65b0613491 Integrate LLVM at llvm/llvm-project@202766947e
Updates LLVM usage to match
[202766947edb](https://github.com/llvm/llvm-project/commit/202766947edb)

PiperOrigin-RevId: 329673065
2020-09-02 02:27:52 -07:00
Robert Suderman 7c93352a40 Scalar / Trivial folding for mhlo.select
This covers the case where the predicate is a splat or the on_true/on_false
values are the same.

PiperOrigin-RevId: 329622785
2020-09-01 18:34:12 -07:00
A. Unique TensorFlower a622bf479b Fix mhlo::SliceOp::fold to not crash on unknown shapes
PiperOrigin-RevId: 329504383
2020-09-01 07:37:36 -07:00
Benjamin Kramer 049f034116 Lower mhlo.floor to lmhlo to linalg
PiperOrigin-RevId: 329304882
2020-08-31 08:16:36 -07:00
Jacques Pienaar 344c500fca [mhlo] Add legalize to SCF pass
Start of pass to legalize MHLO control flow to SCF for further optimization in common form. The current version just matches a very simple instance (which also happens to occur a few times). Exposes some further canonicalization opportunities that aren't yet addressed.

PiperOrigin-RevId: 329017723
2020-08-28 15:11:58 -07:00
Mehdi Amini 36ddbeb6b2 Remove the dependency on global dialect registry from mlir-hlo
PiperOrigin-RevId: 328457105
2020-08-25 20:30:42 -07:00
Uday Bondhugula 94296bb7ec PR #42509: [MLIR] Add folder for mhlo get_dimension_size
Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/42509

Add folder for mhlo GetDimensionSizeOp (`mhlo.get_dimension_size`).
`get_dimension_size` folds to a constant when the corresponding tensor
dimension size is statically known / constant.
Copybara import of the project:

--
5994915525ec2e932125aa1f133ce2260ba100af by Uday Bondhugula <uday@polymagelabs.com>:

[MLIR] Add folder for mhlo get_dimension_size

Add folder for mhlo GetDimensionSizeOp. get_dimension_size folds to a
constant when the corresponding tensor dimension size is statically
known / constant.

COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/tensorflow/pull/42509 from polymage-labs:get_dimension_size_fold 5994915525ec2e932125aa1f133ce2260ba100af
PiperOrigin-RevId: 328222517
2020-08-24 15:36:30 -07:00
Mehdi Amini 73b5a44f33 Update third_party/tensorflow/compiler/mlir/tensorflow/utils/... to not depend on the global Dialect Registry (NFC)
PiperOrigin-RevId: 328171679
2020-08-24 10:59:22 -07:00
Uday Bondhugula 282dba6d38 PR #42508: [MLIR] Erase dead lmhlo.constant ops
Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/42508

An lmhlo.constant op on an memref that is locally allocated and with
no users other than dealloc's can be deleted. Add a canonicalization
pattern for this.
Copybara import of the project:

--
8758c409a15f567e7cb8e1077faa020f5705c85a by Uday Bondhugula <uday@polymagelabs.com>:

[MLIR] Erase dead lmhlo.constant ops

An lmhlo.constant op on an memref that is locally allocated and with
no other users (other than dealloc's) can be deleted. Add a
canonicalization patter for this.

COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/tensorflow/pull/42508 from polymage-labs:lhlo_constant_erase 8758c409a15f567e7cb8e1077faa020f5705c85a
PiperOrigin-RevId: 328042416
2020-08-23 12:28:54 -07:00
Hanhan Wang bfd629ecb0 Enhance lowering reshape op to Linalg.
Handle non-expansion and non-collapsion cases by rewriting it to two reshape
ops.

PiperOrigin-RevId: 327926863
2020-08-21 23:27:34 -07:00
Richard Uhler 9a232c7012 Fix failing segment_reduction_ops_mlir_bridge_test
By adding support for complex types to GetScalarOfType and using appropriate
choice of limits for initial values in the unsorted segment reduction ops.

PiperOrigin-RevId: 327061577
2020-08-17 11:28:20 -07:00
Robert Suderman 5252aeae8f Lowering for mhlo.ceil to std.ceil
PiperOrigin-RevId: 326335301
2020-08-12 16:15:35 -07:00
Jacques Pienaar 5dac76f4af Add chlo.constant_like op which splats a constant to shape of operand
This allows specifying a constant whose shape is only known when operand shape is. Also use it to update tf.Acos legalization.

PiperOrigin-RevId: 325860604
2020-08-11 14:54:48 -07:00
Alexander Belyaev 843af36e05 [MLIR] Add e2e test for unranked unary TF op, lowered and run with CPU runner.
PiperOrigin-RevId: 325665428
2020-08-09 02:38:00 -07:00
A. Unique TensorFlower e6fa003bf2 Integrate LLVM at llvm/llvm-project@b6d9add71b
Updates LLVM usage to match
[b6d9add71b1a](https://github.com/llvm/llvm-project/commit/b6d9add71b1a)

PiperOrigin-RevId: 325589103
2020-08-08 04:35:25 -07:00
Andy Ly 53fdda7f3e Update mhlo.constant to use a custom assembly format instead of a custom printer and parser (NFC).
PiperOrigin-RevId: 325560779
2020-08-07 22:23:06 -07:00
Lucy Fox d742477c02 Verify that MHLO DynamicUpdateSlice start indices have matching element types.
HLO requires that the element types match for all start index parameters. Right now we don't catch this invalid case until export, so adding a check in the verifier so that we catch this sooner.

This also requires a small tweak to the TF InplaceUpdate op lowering.

PiperOrigin-RevId: 325463796
2020-08-07 22:21:40 -07:00
A. Unique TensorFlower e00565b544 Integrate LLVM at llvm/llvm-project@9dbdaea9a0
Updates LLVM usage to match
[9dbdaea9a0e6](https://github.com/llvm/llvm-project/commit/9dbdaea9a0e6)

PiperOrigin-RevId: 325354353
2020-08-07 22:20:06 -07:00
Rahul Joshi a6978cf4ab [MLIR][NFC] Adopt FuncOp/Region argument API's.
- Use FuncOp::getArguments() and Region::getArguments() and friends where possible
  instead of going through the front() block.

PiperOrigin-RevId: 325352975
2020-08-07 22:19:21 -07:00
Lucy Fox cd22ecd136 Relax DynamicBroadcastInDim verifier when dimensions are dynamic.
For input and output dimensions which must match, we shouldn't fail in the case where one dim is dynamic and the other is static. This is insufficient information to conclude a dimension mismatch.

PiperOrigin-RevId: 325344738
2020-08-07 22:18:38 -07:00
A. Unique TensorFlower a68a16cdc7 [MLIR][XLA] Allow for choice of safe/unsafe variant in broadcast utils
Create safe or unsafe variants of `shape.broadcast` depending on the context.
The representation by means of an extent tensor is only legal if the operands
are known to be broadcastable. Currently, there is no use in a safe context in
the codebase but it will be used for shape inference eventually.

PiperOrigin-RevId: 325228073
2020-08-07 22:16:11 -07:00
Alexander Belyaev bc3293a05f [MLIR] Remove Affine->STD and SCF->STD patterns from lhlo->llvm pass.
PiperOrigin-RevId: 325219360
2020-08-07 22:15:24 -07:00
Mehdi Amini 701312720c Add CMake files and lit configurations, enough for `ninja check-mlir-hlo` to pass on all the tests
PiperOrigin-RevId: 325172984
2020-08-07 22:14:34 -07:00
Andy Ly c340367702 Add canonicalization for unpacking and repacking the same tuple (e.g. tuple -> get_tuple_element -> tuple).
These unpacking and repacking of tuples may be generated when modifying tuple arguments or results.

PiperOrigin-RevId: 325162694
2020-08-05 21:38:02 -07:00
A. Unique TensorFlower 4372124362 [MLIR][XLA] Allow for choice of safe/unsafe variant in broadcast utils
Create safe or unsafe variants of `shape.broadcast` depending on the context.
The representation by means of an extent tensor is only legal if the operands
are known to be broadcastable. Currently, there is no use in a safe context in
the codebase but it will be used for shape inference eventually.

PiperOrigin-RevId: 325079842
2020-08-05 12:43:29 -07:00
A. Unique TensorFlower 5d3cc2105e [MLIR][HLO] Remove redundant casts from unranked to ranked transformation
The transformation of unranked to ranked operations no longer generates cast
operations for shapes and sizes. Instead, we use the newly introduced support
for extent tensor and index types directly.

PiperOrigin-RevId: 325057440
2020-08-05 11:11:43 -07:00
A. Unique TensorFlower 37c36a4389 [MLIR][XLA] Allow for choice of safe/unsafe variant in broadcast utils
Create safe or unsafe variants of `shape.broadcast` depending on the context.
The representation by means of an extent tensor is only legal if the operands
are known to be broadcastable. Currently, there is no use in a safe context in
the codebase but it will be used for shape inference eventually.

PiperOrigin-RevId: 325056915
2020-08-05 11:09:23 -07:00
Stephan Herhut 6584c2ab1f Remove optional static registration for hlo dialects again.
Instead, we invoke multiple test tools in a row in end to end tests now. For hlo dialects and passes, we use mlir-hlo-opt explicitly.

PiperOrigin-RevId: 324989884
2020-08-05 03:53:17 -07:00
Mehdi Amini 6c7d1a7c7e Internal change
PiperOrigin-RevId: 324753700
2020-08-03 22:59:44 -07:00
Smit Hinsu 3fe9a7d2db Legalize TensorFlow NonMaxSuppressionV4 and SelfAdjointEigV2Op ops to HLO
Added support for HLO ops bitcast-convert, sort and while in MlirHloBuilder and enabled tests for NonMaxSuppressionV4 and SelfAdjointEigV2Op using these ops.

PiperOrigin-RevId: 324360651
2020-08-03 19:30:54 -07:00
A. Unique TensorFlower 7809320a5e Integrate LLVM at llvm/llvm-project@b7cfa6ca92
Updates LLVM usage to match
[b7cfa6ca9283](https://github.com/llvm/llvm-project/commit/b7cfa6ca9283)

PiperOrigin-RevId: 324331764
2020-08-03 19:30:43 -07:00
Smit Hinsu 577a81a66d Sink standard dialect constants in sink_constants_to_control_flow pass
This is required before exporting HLO dialect ops with standard dialect constant to XLA.

Also, sink constants for sort op as well. Added a TODO to generalize this pass to handle more ops and non-const values defined outside.

PiperOrigin-RevId: 324301911
2020-08-03 19:30:29 -07:00
Stephan Herhut 734b9b25fd Add optional static registration for mhlo/lmhlo passes.
PiperOrigin-RevId: 324190465
2020-08-03 19:29:40 -07:00
Thomas Joerg 735ae2838c Integrate LLVM at llvm/llvm-project@cd4e8d7f6f
Updates LLVM usage to match
[cd4e8d7f6f5e](https://github.com/llvm/llvm-project/commit/cd4e8d7f6f5e)

PiperOrigin-RevId: 324173542
2020-08-03 19:29:29 -07:00
Tres Popp cce4bddf4b Remove unnecessary conversions between Shape and ExtentTensor.
PiperOrigin-RevId: 323981215
2020-08-03 19:29:04 -07:00
Tres Popp ffef8d6593 Support CHLO->LHLO lowering for broadcasting operations with both inputs unranked.
PiperOrigin-RevId: 323960733
2020-08-03 19:28:52 -07:00
Stephan Herhut 1b0eb4baa7 Do not mandate the result type of shape computations but have it be inferred from context.
The computation of a broadcasted shape forced the use of the shape type unnecessarily, which blocked further canonicalizations.

PiperOrigin-RevId: 323783998
2020-08-03 19:28:12 -07:00
Mehdi Amini cd01bb4c4e More cleanup in mlir-hlo to prepare for the standalone build
Shuffle files around, use TableGen to register passes, and introduce
a `mlir-hlo-opt.cpp` file to hold the main entry point of the -opt tool
and stop relying on static registration for dialect/passes.

PiperOrigin-RevId: 323674455
2020-08-03 19:28:00 -07:00
Stephan Herhut effd3fb4f9 Extend unranked to ranked pattern for hlo operations to all unary and binary ops.
As this is essentially always the same pattern, only one operation is tested.

PiperOrigin-RevId: 323525418
2020-08-03 19:27:49 -07:00
Thomas Joerg 739758f9cc Integrate LLVM at llvm/llvm-project@eed333149d
Updates LLVM usage to match
[eed333149d17](https://github.com/llvm/llvm-project/commit/eed333149d17)

PiperOrigin-RevId: 323354988
2020-08-03 19:27:25 -07:00
Robert Suderman 8023baa959 Modified HLOAbsOp lowering for differing types.
PiperOrigin-RevId: 323082107
2020-08-03 19:27:12 -07:00
Hanhan Wang 8f262ae8f5 Add support for lowering mhlo.iota to Linalg.
PiperOrigin-RevId: 322799853
2020-07-30 22:34:45 +00:00
Tres Popp 4251630426 Support CHLO broadcasting operations between scalar and unranked tensors.
This is done through reshaping the unranked tensor into a 1D ranked tensor which will result in a safe broadcast/indexing logic when the other operand is a scalar.

PiperOrigin-RevId: 322553661
2020-07-30 22:34:40 +00:00