Commit Graph

964 Commits (main)

Author SHA1 Message Date
Jiawei Wu 026e8db2e4
[Stablehlo] add converter for aten.scatter.src op (#2295) 2023-07-24 10:14:45 +08:00
Alexandre Rames 1e468e8294 Fix canonicalization of `torch.prim.TupleUnpack`. 2023-07-20 20:08:46 +02:00
Alexandre Rames a20422ce65 Support `DerefineOp` in `RefinePublicReturn`. 2023-07-20 20:08:46 +02:00
Alexandre Rames 4847563bed Clean up verification of calling conventions.
The implementation at this place was a remnent of the times the pipeline was
run only once.
Rely instead on the backend verification, after optimizations have had an
opportunity to resolve some uncertainties. (e.g. `!torch.optional`).
2023-07-20 20:08:46 +02:00
Matthias Gehre 64d7626a52
Fixes for split tensor and slice (#2314)
* RecomposeComplexOps: Remove dead slice op

* lib/Dialect/Torch/IR/TorchOps.cpp: Fold slice ops even when they are on non-value tensors

* lib/Conversion/TorchToTosa/TorchToTosa.cpp: Fix slice start/end out of range/none

* lib/Dialect/Torch/IR/TorchOps.cpp: AtenSliceTensorOp::fold: Fold slices that go from 0:int_max

* More tests for aten.split.Tensor
2023-07-20 09:53:54 +02:00
Jiawei Wu 3f843c8fd9
[torch-dialect] fix aten.type_as op's folder (#2283)
[torch-dialect] fix torch.type_as op's folder by decomposing it to prim.dtype + aten.to_dtype
2023-07-20 09:51:58 +08:00
Ramiro Leal-Cavazos 718f53ff8a
Fix handling of `!torch.number` in abstract interpretation library (#2309)
In PyTorch, the `NumberType` is equal to `Union[int, float,
complex]`. However, the abstract interpretation library was treating
the `NumberType` as `Union[int, float]`, resulting in type mismatches
when reifying certain dtype functions. This commit fixes the type
inconsistency by having the abstract interpretation functions take as
an input a `Union[int, float, complex]` for the ops that take
`!torch.number` inputs.
2023-07-17 09:52:04 -07:00
Jiawei Wu c7fa42b7d3
[Torch Dialect] Add canonicalizer for aten.to.other op (#2273)
Canonicalize aten.to.other to prim.device + prim.dtype + aten.to.device
Co-authored-by: wujiawei.aml <wujiawei.aml@bytedance.com>
2023-06-30 09:43:08 +08:00
Yuanqiang Liu 449cfb8375
[Torch Dialect] add more scalar op folders (#2265) 2023-06-29 10:37:13 +08:00
Yuanqiang Liu 1ea2b57ab7
[Torch Dialect] add folder for aten.add (#2264)
* [Torch Dialect] add folder for aten.add

* update

* update

* update
2023-06-27 10:55:28 +08:00
Yuanqiang Liu 96b14e952e
[Torch Dialect] Support aten.device.with_index (#2254) 2023-06-23 01:07:14 +08:00
Vivek Khandelwal f6a6cfea4e
[MLIR][TORCH] Add support for negative index values for index.Tensor op (#2233)
This commit adds the support for index.Tensor op when the index values
are negative. This commit wraps around the index values by checking
their values at run time.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2023-06-16 14:21:04 -05:00
Yuanqiang Liu 7c6961bcbf
[Torch Dialect] Support aten.cuda and add canonicalizer for aten.cuda (#2231) 2023-06-14 09:56:39 +08:00
Maksim Levental 0caaf8d32a
Bump LLVM (#2176)
* Bump LLVM

---------

Co-authored-by: Matthias Gehre <matthias.gehre@xilinx.com>
2023-06-13 16:17:23 +02:00
Yuanqiang Liu ddea56a832
[Torch Dialect] fix torch.uint8's dtype infer (#2227) 2023-06-13 10:38:20 +08:00
Christopher McGirr b461daa06e
fix(TorchToTosa.cpp): adjust torch->tosa div conversion (#2200)
check the return type of the division to figure out whether to use
the floating point implementation of a division or to use the integer.

the issue rose from the fact that the inputs are all integer but the
result was casted to floating point. The conversion then chose to
use the integer implementation of division which is not legal in tosa
when all the inputs get casted to floating point.

fix(TorchToLinalg): AtenDivScalarOp

upcast self operand as well if applicable, the self operand must also
be casted to float as it can be an integer.
2023-06-12 11:18:38 +02:00
Matthias Gehre 27a3d09917
Torch: Fold RuntimeAssertOp when condition is true (#2198) 2023-06-09 19:06:25 +08:00
Yuanqiang Liu 5a7bf4e4cb
[Torch Dialect] Add canonicalize pattern for aten.is_floating_point (#2194)
* [Torch Dialect] Add canonicalize pattern for aten.is_floating_point

* implement as fold

* add lit test
2023-06-07 17:05:31 +08:00
Vivek Khandelwal da886280fe
[MLIR][TORCH] Add E2E support for aten.tril op (#2202)
Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2023-06-05 16:17:01 -07:00
Ramiro Leal-Cavazos dff3405d5a
Add alias analysis for cast-like ops to maximize-value-semantics (#2160)
When `use_tracing=True` is used to import a model into Torch-MLIR,
several casts get inserted in the IR to bridge the untyped inputs and
outputs with the typed body of the computation. These casts create
extra aliases of tensors that cause the current analysis in
`maximize-value-semantics` to fail.

In particular, the `maximize-value-semantics` analysis assumes that the
only valid alias right after an overwrite is the overwritten
alias. So, if there is a use of a casted version of the overwritten
alias after the overwrite, the analysis fails.

This commit improves the analysis by identifying all cast-like aliases
of the overwritten alias and allowing such aliases to be used after an
overwrite.

Because this issue only arises when using tracing, it cannot be
currently tested e2e, so only lit test is added.
2023-05-25 17:05:41 +00:00
Zhekun Zhang f0b7b63be0
[Stablehlo] Add aten.uniform lowering (#2101)
* add uniform stablehlo lowering

* add unit test

* new line

* rm redundant file

* Empty commit, trigger test

* fix include

* address comments

---------

Co-authored-by: zhekun.zhang <zhekun.zhang@bytedance.com>
2023-05-25 10:32:55 +08:00
TatWai Chong ed4ecb072f
[tosa] support lowering basic torch binary ops with mixed dtypes (#2122)
Lowering torch operations that allow different compatible data types
in its operands to tosa end up generating invalid tosa IR with mixed
data types. In tosa spec, certain operations (generally element-wise
operations) require all operands to have the same data type.

Add wrapper functions for those element-wise tosa ops to perform op
creation with type conversion if necessary.
2023-05-18 17:12:18 -07:00
Ramiro Leal-Cavazos de02b56e17
Replace RefineTypes with dtype functions (#2105)
This commit adds dtype functions for all the torch ops that did not
previously have one and removes the pass `RefineTypes`, since the
abstract interpretation library now takes care of all the dtype
propagation.

All dtype functions added are tested except for
- `aten.embedding`
- `aten._embedding_bag`
- `aten.embedding_bag`

These functions need a change to the testing framework to allow
specifying the actual data inside the tensor used for testing. I will
fix this in a follow up patch.

Co-authored-by: Jiahao Li <liplus17@163.com>
2023-05-12 13:40:45 -07:00
Sean Silva d7614c261d Integrate LLVM
LLVM: 26ee8947702d79ce2cab8e577f713685a5ca4a55
MHLO: 4805d8498dfb81566076f56f52273b426c1cc5bf

Per: https://github.com/llvm/torch-mlir/issues/1178#issuecomment-1538492185
2023-05-09 10:14:27 -07:00
Chi_Liu 51e0a2c933
[Stablehlo] Add stablehlo support for aten.abs (#2068)
Co-authored-by: AmosLewis <Amos_Lewsi@foxmail.com>
2023-05-08 22:13:00 -07:00
Yuanqiang Liu ef6dae6ae2
[Linalg] fix lowering reduce max with -inf (#2097) 2023-05-08 09:17:49 -07:00
Yuanqiang Liu 0096ceae2f
[Stablehlo] fix reduce max init_value with -inf (#2064)
* [Stablehlo] fix reduce max init_value with -inf

* update
2023-05-06 12:05:51 -07:00
Zhekun Zhang 0cf9ee340b
[Torch Dialect] Add to.dtype_layout canonicalize patterns (#2062)
* add to.dtype_layout canonicalize patterns

* update comment

---------

Co-authored-by: zhekun.zhang <zhekun.zhang@bytedance.com>
2023-05-02 20:06:02 -07:00
Ramiro Leal-Cavazos 96d662647f
Fix import of constant bool tensor parameters (#2047)
Bool tensors are represented in TorchScript as an array of
`int8_t`s. However, when importing them into Torch-MLIR, the importer
was assuming the array had `int32_t` elements, leading to the importer
reading into memory that was out of bounds. This commit fixes the
casting of the bool tensor.
2023-04-20 18:38:48 -07:00
Chi_Liu f3d1eda09f
[TOSA] Add aten.abs support (#2032) 2023-04-14 08:43:39 -07:00
Zhekun Zhang 1bd5747ca3
[StableHlo] Fix transposed convolution conversion (#2026)
* fix conv bwd

* fix

* fix group case

* clean up

---------

Co-authored-by: zhekun.zhang <zhekun.zhang@bytedance.com>
2023-04-13 11:24:39 -07:00
Yuanqiang Liu 3e83a86354
[Torch Dialect] fix isValidSubtype with dynamic dim (#2018) 2023-04-11 01:02:18 -07:00
Vivek Khandelwal 98747d09a8 [MLIR][TORCH] Add support for prims::view_of op
This op does nothing and just returns the input operand as the
result of the op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2023-04-11 07:58:10 +05:30
Vivek Khandelwal e90ea3d7ab [MLIR][TORCH] Extend implementation of aten._index_put_impl op.
This commits adds the support for cases for index_put_op:
1.) where index is a 2-d tensor.
2.) where indices is a list of tensors and none, with exactly
2 non none tensors along the consecutive dimensions.

This commit also adds a utility to compute the broadcast shape
given the two input tensors.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2023-04-05 14:04:30 +05:30
Alexandre Rames d24fa71368
Minor fixes for `ConvertTorchConversionToMLProgram`. (#1991)
* Only create the global seed variable if it does not exist already.
* Make the pass a module pass. A func pass may not modify its parent op.
2023-04-04 09:09:58 -07:00
Yuanqiang Liu c86f46bd70
[test] rename TorchToMhlo to TorchToStablehlo (#1995) 2023-04-03 18:41:25 -07:00
Ramiro Leal-Cavazos e0f301c890
Add `extra_library` kwarg to `torch_mlir.compile` (#1986)
This commit adds the ability to specify extra abstract interpretation
functions in `torch_mlir.compile` to use during type refinement. This
allows users to easily add custom ops without having to interact with
MLIR or C++ directly.
2023-03-30 09:20:19 -07:00
Ramiro Leal-Cavazos d803ab4eeb
Cast `number` to `float` when shape function takes Scalar arg (#1978)
To keep things simple in shape functions, `Scalar` inputs are
considered `float`s. This means that when inserting the shape
functions into the IR, we must cast any `!torch.number`s into `float`s
so that the operand type matches the expected type in the shape
function. This commit adds the cast from `Scalar` to `float`.
2023-03-28 09:30:31 -07:00
Maksim Levental 953ea39cb5
handles 2,3,4 from https://github.com/llvm/torch-mlir/issues/1963 (#1964) 2023-03-24 21:50:01 -05:00
Michael Feliz 2389729fb9
Add support for aten_remainder in TorchToTosa (#1966) 2023-03-23 17:55:58 -07:00
Ramiro Leal-Cavazos eae3ff7f1c
Change dtype functions interface to take ints tuple for each tensor (#1965)
The original design for the dtype functions outlined in
https://github.com/llvm/torch-mlir/issues/1462 was unable to properly
handle ops that take optional tensors as an input when the optional
tensor has a value of None. By the time the op gets imported into
torch-mlir, if an optional value is None, all information about the
original type is lost from the op type signature, preventing
torch-mlir from knowing if a value of None was from an optional tensor
or not, which was crucial in the original design since each tensor
argument must be turned into two separate arguments for the dtype
function.

This commit changes the interface to dtype functions such that each
tensor turns into a tuple of two ints, the first representing the rank
of the tensor and the second the dtype of the tensor. Since now there
is a one-to-one correspondence between the operands of an op and the
operands of its dtype function, there is no ambiguity about which
operand of the op corresponds with which operand of the dtype
function.

To test the implementation, this commit defines dtype function for
convolution op, which takes one optional tensor as an argument.
2023-03-23 11:05:39 -07:00
Sean Silva c319a20828 Update to LLVM 029313cc979ae71877b65794b1063d4e51184cc8
- mergeBlockBefore -> inlineBlockBefore
- move tosa-to-tensor pass ordering

https://github.com/llvm/torch-mlir/issues/1178#issuecomment-1476217922
2023-03-21 04:16:20 -07:00
Matthias Gehre aa5bcb3cf2
LowerToBackendContract: Explicitly error out on unimplemented operator (#1947)
* LowerToBackendContract: Explicitly error out on unimplemented operator

But only reject torch.operator when results are invalid.
Otherwise it might be a custom op that the backend supports.
2023-03-20 16:27:08 +01:00
Ramiro Leal-Cavazos d310bb12bd
Expand definition of tensor subtype to include shape/dtype info (#1929)
Currently, the op `torch.tensor_static_info_cast` will not get
canonicalized away if the result type has any shape or dtype
information. This is because `isValidSubtype` only returns true when
the tensor types being compared are exactly the same or the supertype
has no shape and dtype information. Being unable to canonicalize away
the `torch.tensor_static_info_cast` gets in the way of further
optimizations, such as shape propagation.

This commit improves `isValidSubtype` by adding logic that compares
the shapes and dtypes of the two tensor types to determine of one type
is indeed a valid subtype of the other.

Fixes https://github.com/llvm/torch-mlir/issues/1926
2023-03-10 16:43:57 -08:00
Ziheng Jiang dca2b8a40a
[TORCH] Improve type refinement for aten.cat. (#1908)
* [TORCH] Fix type refinement for aten.cat.

* Add test.

* Address comments.

* Update.

* Update.

* Update.

* Update.

* Update.

---------

Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2023-03-09 16:17:35 -08:00
Zhekun Zhang 1d3a7419c5
[Torch Dialect] add RSub, ScalarImplicit canonicalize (#1899)
* add rsub, scalarimplit canonicalizer

* reformat

* address comments

* fix bug

* fix test

* Update elementwise.py

* resolve merge conflict

* change to 3

* change to 3

* real fix

* fix name

* add torchdynamo fail test

---------

Co-authored-by: zhekun.zhang <zhekun.zhang@bytedance.com>
2023-03-06 17:38:27 -08:00
Ramiro Leal-Cavazos d30af8772b
Handle uninitialized lattice elements in RefineTypes (#1911)
The data-flow analysis does not always propagate information to the
entire graph. This results in some lattice elements being
uninitialized. Currently the lattice elements are not checked to see
if they are uninitialized before rewriting the graph, potentially
resulting in invalid IR (see
https://github.com/llvm/torch-mlir/issues/1896).

This commit adds handling for uninitialized lattice elements.
2023-03-03 08:55:58 -08:00
Yuanqiang Liu 7a8304f935
[Torch Dialect] add folder for aten.sub.float (#1871) 2023-03-02 09:07:33 -08:00
Yuanqiang Liu fc1e091d6a
[Torch Dialect] add aten.pow.int_float op and it's folder (#1872) 2023-02-28 09:36:05 -08:00
Maksim Levental 2eddb3fde7
WIP: No PyTorch dep (#1854) 2023-02-13 14:21:06 -06:00
Yuanqiang Liu 6ab990e1e8
[Torch Dialect] add folder for aten.Int.float (#1863) 2023-02-10 13:59:03 -08:00
Yuanqiang Liu 2f6fdb7f0b
[Torch Dialect] add folder for prim.min.int (#1864) 2023-02-10 13:58:15 -08:00
Ashay Rane 711646d095
mhlo: migrate conversion to stablehlo (#1840)
This patch replaces all MHLO operations with their StableHLO
counterparts and adds a validation pass to ensure that no MHLO operations
remain before translating all Stablehlo operations to the MHLO dialect
for further lowering to the Linalg dialect.

This patch also updates all lit tests so that they refer to the
`convert-torch-to-stablehlo` pass and so that they check for StableHLO
operations.
2023-02-02 07:29:47 -06:00
Chi_Liu 00fc14a6e1
[TOSA] Add to.dtype i1 to i64 (#1830) 2023-01-27 09:21:06 -08:00
Gleb Kazantaev 3930588a7e
Enable VerifyBackendContract in LTC backend (#1798)
* Enable VerifyBackendContract in LTC backend

* Update VerifyBackendContract pass

* Move convert_scalar_implicit to jit_utils

* Rename VerifyBackendContract to VerifyBackendContractNoDecompositions

* Update verify-backend-contract-error.mlir test
2023-01-24 22:14:17 -05:00
Ramiro Leal-Cavazos 6c86bec04f
build: update llvm tag to 9acc2f37 (#1828)
This commit makes the following changes:

- Update dialects to use fold API `kEmitFoldAdaptorFolder` and update
signature of `fold` methods (see PSA
https://discourse.llvm.org/t/psa-new-improved-fold-method-signature-has-landed-please-update-your-downstream-projects/67618)
- Replace `makeArrayRef` with `ArrayRef` (see
https://reviews.llvm.org/D140896)
- Remove `TypeRange{}` arg from `b.create<scf::IfOp>` since builder no
longer takes that argument
- Make `func`s in `Torch/invalid.mlir` private, since symbol
declarations cannot be public. (see https://discourse.llvm.org/t/rfc-symbol-definition-declaration-x-visibility-checks/2140)
2023-01-25 01:29:42 +00:00
Maksim Levental 8696752eb6
Expose metadata of torch-mlir types (plus verify DictType key and value types). (#1785) 2023-01-16 10:25:02 -06:00
Ashay Rane 4e4a571104
[TOSA] Add LeakyReLU conversion pass (#1790)
* feat(TorchToTOSA): LeakyReLU legalization

* test(LeakyReLU): Add LIT test and enable e2e test

Co-authored-by: Philipp Braun <philipp.braun@amd.com>
2023-01-10 21:42:07 -08:00
Ashay Rane 0faba6d2fc
build: update llvm tag to de3f0f7f (#1789)
Credit to @vivekkhandelwal1 for finding the necessary changes.

Summary of changes:

 - Switch Tosa_IntArrayAttr[N], Tosa_IntArrayAttrUpto[N] to DenseI64ArrayAttr.

 - Replace kNoIterationLimit with kNoLimit. (https://reviews.llvm.org/D140525)

 - Add dependency on MhloPasses when MHLO is enabled

 - Specify result type when using mhlo::DotOp
2023-01-10 17:07:19 -06:00
Raghavan Raman 0979df6589
Fix unsqueeze in Torch to Tosa conversion (#1780) 2023-01-10 11:09:58 -08:00
Ramiro Leal-Cavazos 273664ded6
[custom op] Replace `tanh` dtype function with `expm1` (#1769)
This commit replaces the `tanh` dtype function, which was being used
to test the implementation of dtype functions in
a710237437, with a dtype function for
`expm1`. The dtype function for `expm1` is identical to the `tanh`
one, so the same level of testing is maintained.

Currently, there are ops getting dtype information from the
`RefineTypes` pass and ops getting dtype information from the
`TorchDtypeRefinementPipeline`. Since each pass can only propagete
dtype information for the ops it knows how to handle, some models with
many ops handled in both passes require the two dtype propagation
passes to execute many times, reaching the iteration limit set in the
`LowerToBackendContractPass`. To temporarily avoid this issue while
the migration to `TorchDtypeRefinementPipeline` is finished, this
commit switches `tanh` to `expm1`, since the latter is used a lot less
in large models.
2023-01-03 14:18:26 -08:00
Ashay Rane ac780529b4
Revert e2e support for aten logical or/and/xor/not ops (#1757)
This reverts commit eaab9be207, since it
is causing the post-merge CI tests to fail, causing subsequent PRs to be
blocked.  Specifically, the tests
`ElementwiseAtenLogicalAndOpPromoteBroadcastModule_basic` and
`ElementwiseAtenLogicalXorOpPromoteBroadcastModule_basic` fail because
the oracle does not match the computed result.  This patch reverts the
commit to make the post-merge builds green again.
2022-12-29 21:01:06 -06:00
Jiahao Li eaab9be207
Add e2e support for aten logical or/and/xor/not ops (#1752) 2022-12-26 10:23:38 +08:00
Jiahao Li 49071f86e6
[MHLO] Evaluate RuntimeAssertOp at compile time (#1732) 2022-12-22 17:12:52 +08:00
Tanyo Kwok 297fd3aa47
Revert "reimplement linear lowering torchToMhlo (#1524)" (#1744)
This reverts commit 50b524546f.
2022-12-21 21:24:07 -08:00
zzp_miracle 50b524546f
reimplement linear lowering torchToMhlo (#1524) 2022-12-22 10:15:16 +08:00
Jiahao Li 15b249777b
[Torch][MHLO] Decompose aten.copy op. Lower aten.rsqrt & sigmoid to mhlo. (#1734) 2022-12-22 10:13:59 +08:00
Chi_Liu 9dc09ac8c5
[TOSA] Add aten.gather support for tosa (#1680) 2022-12-21 11:04:07 -08:00
Chi_Liu b2cefc0b64
[TOSA] Add aten.masked_fill.Tensor/Scalar support (#1735) 2022-12-21 08:56:07 -08:00
ataheridezfouli-groq 17ee643aeb
[TORCH] Add Complex Number support (#1673)
Add Complex number dtype support to torch tensors. Add
aten.fft_fft op to test complex numbers.
2022-12-15 21:40:01 +00:00
Ramiro Leal-Cavazos 60db793feb
Pass op legality info to `verifyBackendContractPass` (#1705)
In order to verify if a given IR satisfies the backend contract, the
verifier needs to know if decompositions took place, and if so, which
ops were decomposed and which were not.

This commit adds two arguments to `verifyBackendContractPass` to
specify if decompositions took place and which ops to consider backend
legal, similar to the arguments of `LowerToBackendContractPass`.
2022-12-15 08:32:52 -08:00
Ahmed S. Taei b1f6832849
Add aten.slice.Tensor & aten.cat folders (#1691) 2022-12-13 13:02:47 -08:00
Ramiro Leal-Cavazos a710237437
[custom op] Generalize shape library logic to work with dtypes (#1594)
* [custom op] Generalize shape library logic to work with dtypes

This commit generalizes the shape library logic, so that dtype rules
for ops can also be expressed using the same mechanism. In other
words, each op can now have a shape function and a dtype function
specified in Python that is imported during lowering to calculate the
shapes and dtypes throught a program. For more information about how
to specify a dtype function, see the updated
`docs/adding_a_shape_and_dtype_function.md`.

For those not familiar with how the shape library works, the file
`docs/calculations_lib.md` provides an overview.
2022-12-13 08:25:41 -08:00
Chi_Liu 163d19cce6
[TOSA] Add aten.add/sub.Scalar/Tensor si64 type support (#1604) 2022-12-12 12:13:07 -08:00
Ramiro Leal-Cavazos a54b334578
Allow running DecomposeComplexOps more than once (#1671)
The current implementation of `DecomposeComplexOps` fails if an op
expected to be decomposed does not get decomposed in the first
iteration of the `createTorchSimplificationPipeline` in
`LowerToBackendContractPass`. However, some graphs require multiple
iterations of `createTorchSimplificationPipeline` to fully propagate
all statically knowable information, such as dtypes and shapes, to the
entire graph, sometimes resulting in the need to run
`DecomposeComplexOps` more than once.

This commit changes `DecomposeComplexOps` to use a greedy algorithm
for pattern application and moves the legalization check of ops to the
`LowerToBackendContractPass` to allow for the `DecomposeComplexOps` to
run more than once.
2022-12-08 09:26:38 -08:00
Ramiro Leal-Cavazos 76190e8a3f
Remove unnecessary decompose-complex-ops tests (#1693)
This commit removes lit tests from the `decompose-complex-ops` that
are essentially testing a macro expansion, in accordance with
https://github.com/llvm/torch-mlir/blob/main/docs/architecture.md#dos-and-donts-for-unit-vs-end-to-end-testing .
2022-12-08 08:22:08 -08:00
Ramiro Leal-Cavazos dd35488da5
build: update llvm tag to 798fa4b4 (#1684)
- Support for non-prefixed accessors has been removed. See:
  https://reviews.llvm.org/D136727
- Rename `operands` to `methodOperands` in `prim.CallMethod` since the
  name `operands` overlaps with a builtin method name. See:
  https://reviews.llvm.org/D136727
- Add passes in refbackend to lower memref.subview. See:
  https://reviews.llvm.org/D136377
- Replace `CopyToValueTensorOps` first in `RewriteViewLikeSubgraph` in
  maximize-value-semantics.

  The current implementation of the `RewriteViewLikeSubgraph` pass in
  maximize-value-semantics creates temporarily invalid IR. In
  particular, given a forward slice starting from a
  `CopyToNonValueTensorOp` and ending in `CopyToValueTensorOp`s, the
  pass first replaces all uses of the `CopyToNonValueTensorOp` with
  its operand, which results in all the `CopyToValueTensorOp` users
  having their operand have type `!torch.vtensor`, which is invalid.

  The correct way to do things is to first replace all the
  `CopyToValueTensorOp`s with their operand, and then replace all uses
  of the `CopyToNonValueTensorOp` with its operand.

  This only started failing now because the generated accessor
  `getOperand` for the `CopyToValueTensorOp` now returns a
  `TypedValue<NonValueTensorType>`, which has an assert checking that
  the value returned is of the expected type.
2022-12-07 12:20:41 -08:00
Vivek Khandelwal f416953600 [MLIR][TORCH] Add TorchConversionToMLProgram and MLProgramBufferize pass
This commit changes the `InsertRngGlobalsPass` to `TorchConversionToMLProgram`
pass. This commit also adds the `MLProgramBufferize` pass for the
bufferization of ml_program dialect ops to run on refbackend.

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-12-02 13:20:46 +05:30
Vivek Khandelwal e7edcc62fd build: update llvm tag to 147fe9de
Summary of changes:
- Replace call to `MemoryEffectOpInterface::hasNoEffect`
  with `isMemoryEffectFree`.
- Make fix for the dynamic dims, since
  `kDynamicSize` value changed to
  `std::numeric_limits<int64_t>::min()` from `-1` in llvm
- `makeShapeLLVMCompatible` and `makeShapeTorchCompatible`
  utilities convert shapes in order to remain consistent
  with the Torch and MLIR semantics.
- Update tags
  llvm: 147fe9de29dc13c14835127b35280c4d95c8e8ba
  mhlo: 1944b5fa6062ec4c065d726c9c5d64f1487ee8c5

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-12-01 13:36:50 +05:30
Ramiro Leal-Cavazos 0983a7f93a
Fix modulus calculation in LCG algorithm of refbackend (#1658)
The current implementation sets the `nextSeed` value to `temp & 127`,
which is wrong. The last step of the LCG algorithm for the multiplier
and increment chosen should be `temp % 2^{64} = temp & (1 <<
63)`. However, because we are dealing with i64 values, the modulus
operation happens automatically, so it is not needed.

See Donald Knuth's values for LCG here:
https://en.wikipedia.org/wiki/Linear_congruential_generator
2022-11-30 08:46:52 -08:00
Tanyo Kwok bbcdb38d99
Revert "Decompose torch.slice_scatter (#1622)" (#1659)
This reverts commit f3f2f10030.
2022-11-30 12:47:13 +08:00
Vivek Khandelwal d9cbf01d1e Revert "build: update llvm tag to 147fe9de"
This reverts commit e45ad313d4.
2022-11-25 12:41:56 +05:30
Vivek Khandelwal e45ad313d4 build: update llvm tag to 147fe9de
Summary of changes:
- Update call to `hasNoEffect` utility
- `KDynamicSize` value changed to
  `std::numeric_limits<int64_t>::min()` from `-1`
- Update tags
  llvm: 147fe9de29dc13c14835127b35280c4d95c8e8ba
  mhlo: 1944b5fa6062ec4c065d726c9c5d64f1487ee8c5

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-11-24 12:44:43 +05:30
Tanyo Kwok f3f2f10030
Decompose torch.slice_scatter (#1622)
* Decompose torch.slice_scatter

* fix compilation error

* update file check

* fix ci

* fix i64 torch.tensor dtype
2022-11-23 18:14:12 +08:00
Vivek Khandelwal da8fdc9f96 [MLIR][TORCH] Fix refine types crash
This commit fixes https://github.com/llvm/torch-mlir/issues/1599.

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-11-23 15:17:37 +05:30
Tanyo Kwok 4aad5ccf39
fix #1626 return type mismatch (#1634) 2022-11-23 15:02:41 +08:00
Vivek Khandelwal 55c7e66aa7 [MLIR][TORCH] Fix mean and mean.dim op for large-sized inputs
This commit fixes the aten.mean and aten.mean.dim op decomposition
for supporting large-sized inputs.
This commit also fixes the formatting for the file stats.py

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-11-22 08:38:51 +05:30
Vivek Khandelwal 4cbd3927d7 [MLIR][TORCH] Add aten.sort.int op
Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-11-20 19:00:41 +05:30
Chi_Liu 29c8f47723
[TOSA] Add aten.clamp op tosa support (#1609)
Co-authored-by: AmosLewis <Amos_Lewsi@foxmail.com>
2022-11-18 13:32:13 -08:00
Sean Silva 39de4d6265 [cleanup] Make diagnostics better
Also remove some unused imports.
2022-11-17 02:09:54 -08:00
Sambhav Jain fc4c8d4ed9
Enable torch-mlir LIT tests in Bazel (#1585)
Adds support to run `.mlir` LIT tests in bazel. 

```
bazel test @torch-mlir//test/...
```

Follow-on PR will contain these updates:
- Add tests to GHA CI workflow
- Add `.py` LIT tests to bazel
2022-11-15 14:02:19 -08:00
Chi_Liu dfe7513a45
[MLIR][TORCH] Fix aten.unsqueeze op (#1578)
The range of the unsqueeze dim is: [-input.dim() - 1, input.dim() + 1), the bug forgets to add 1.
2022-11-14 09:09:15 -08:00
Daniel Ellis a7ac0def45
Move single-tensor-tuple-return test to mlir unit test.
Also, add multiple return test.
2022-11-10 09:23:53 -05:00
Xiafei Qiu 4f173c6e0f
update llvm tag to a2620e00. (#1567)
- also update MHLO to 57ba12a2(branch greencommit/2022-11-07-a2620e00)
- change -pass-pipeline format to make tests pass.
2022-11-10 18:39:28 +08:00
Tanyo Kwok 17bc7c89cc
build: update llvm tag to 74fb770d (#1539)
* build: update llvm tag to 74fb770d

This commit makes the following changes needed to update bump LLVM:

+ replace usages of `tensor::createPadScalarOp`, see https://reviews.llvm.org/D136493
+ Update file checks
2022-11-01 15:27:09 +08:00
Ramiro Leal-Cavazos b723186983
Remove all but one of valsem ops + move fill.Scalar to elementwise (#1531)
This commit removes almost all of the valsem ops, since the value
semantics version of the ops now exist in PyTorch. The only op missing
is `aten.bernoulli_.float`. In addition, this commit also simplifies
the implementation of `aten.fill.Scalar` by moving it to the pattern
that converts elementwise ops.
2022-10-28 15:06:11 +00:00
Ashay Rane a11ea93877
build: update llvm tag to f8b84268 (#1528)
The only change required was to update a test to reflect the changes
in https://reviews.llvm.org/D136541.
2022-10-26 15:33:53 -05:00
Vivek Khandelwal ca87033d2f [MLIR][TORCH] Add E2E support for aten.mse_loss op
This commit adds decomposition for the `aten.mse_loss` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-10-25 21:06:58 +05:30
Chi_Liu ad6f5848cb
[MLIR][TORCH] Add TorchToTosa lowering for aten.where.self op (#1454) 2022-10-18 09:39:39 -07:00
Ramiro Leal-Cavazos 82a3860e25
build: update llvm tag to 4546397e (#1502)
This commit makes the following changes needed to update bump LLVM:

- Replace `linalg.init_tensor` with `tensor.empty` (see:
https://reviews.llvm.org/D135129)
- Replace `NoSideEffect` with `Pure` (see
https://reviews.llvm.org/D135505)
- Replace `body` region accessor for `ReduceOp` and `ReduceWindowOp`
with `getBody`
- Fix incorrect use of `tosa::ReduceSumOp` in `AtenNativeLayerNormOp`
conversion pattern. The result type of `tosa::ReduceSumOp` must have
the same rank as the input type. (see:
https://www.mlplatform.org/tosa/tosa_spec.html#_reduce_sum)

Co-authored-by: Ashay Rane <ashay@users.noreply.github.com>

Co-authored-by: Ashay Rane <ashay@users.noreply.github.com>
2022-10-18 04:22:53 +00:00
Gaurav Shukla da90a25f90 [MLIR][TORCH] Add E2E support for `aten.[div.int|bitwise_or.Tensor]` ops
This commit adds lowering of `aten.div.int` and `aten.bitwise_or.Tensor`
ops. Both these ops are required in order to support bloom_560m model.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-10-10 22:28:51 +05:30
Vivek Khandelwal d3cc3f1aff [tosa] Add lowering for aten.to.dtype and aten._to_copy op
This commit adds the TorchToTosa lowering for `aten.to.dtype` and
`aten._to_copy` op.

Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-10-06 12:00:25 +05:30
Vivek Khandelwal 56f9a9b5de [tosa] Add TorchToTosa lowering for torch.prim.NumToTensor.Scalar op
Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-10-06 12:00:25 +05:30
Gleb Kazantaev 708fa346a6
Fix Base Lazy Backend Type Conversion (#1412)
* Fix c10::prim::Constant conversion; Added CAPI for passes; Added passes to base lazy backend

* Update ivalue_importer to use ImportOptions; Added tests for non-value/value tensor types

* Added tests for scalar Constant import; Updated MB::importFunction to use ImportOptions

* Test updates

* Move back module variable name

* Remove RefineTypes from TorchMlirLoweringContext::Build()

* Rename pass; Remove passes from base lazy backend

* Rename pass to VerifyBackendContractPass

* Aligned cmd pass name; Fixed TorchConversion passes registration
2022-10-04 15:53:28 -07:00
Vivek Khandelwal 9dd5ae8239
[tosa] Add TorchToTosa lowering for aten.arange.start_step op (#1442) 2022-09-30 07:33:41 -07:00
AmosLewis 940959589b [MLIR][TORCH] Add Byte and Char Dtype support 2022-09-30 13:19:31 +05:30
Ashay Rane 0b46462528
Miscellaneous fixes for Windows builds (#1376)
* test: allow spaces in path to Python executable

On Windows, the path to the Python binary may contain spaces, so this
patch adds quotes around the path to the python executable.

Thanks to @sstamenova for suggesting the fix!

* python: remove header file that causes Windows build failures

Similar to https://reviews.llvm.org/D125284, we can safely remove this
header file without affecting the build on either Linux.  It is
necessary to remove this header file on Windows builds since otherwise
it causes build errors.

* python: drop `TORCH_API` from function defined in Torch-MLIR

`TORCH_API` should apply to functions that are either exported by
libtorch.so or ones that are imported from libtorch.so by its downstream
consumers (like Torch-MLIR).  Neither case applies to the
`importJitFunctionAsFuncOp()` function, since it is defined in
Torch-MLIR (and thus outside libtorch.so).  This patch fixes the problem
by dropping `TORCH_API` from that function's declaration.

* python: make output of class anotations deterministic

The `class-annotator-repr.py` test checks for class annotations in a
specific order, but prior to this patch, the order was
non-deterministic, since the code iterated on an _unordered_ map.

This patch makes the iteration order deterministic through two changes:
1. using a sorted map
2. using the class qualified name instead of the address of the class in
memory

* test: use Python3_EXECUTABLE as interpreter path for consistency

This ensures that tests use the Python3 version that was detected using
CMake, instead of whichever python version that happens to be in the
PATH variable when invoking the test.

* test: fix RUN string

The parenthesis syntax does not run on Windows (the shell interprets the
`(` character as part of the path).  Moreover, the ODR violation in the
comment no longer seems to apply.

* python: port parallel test framework to Windows

Since Windows does not support `fork` natively, Python's
`multiprocessing` module needs to use `spawn` on Windows.  However, to
use `spawn`, the multiprocessing module serializes (or pickles) the
worker function and its arguments.  Sadly, the multiprocessing module
(both the default one in Python and the one that is extended in PyTorch)
is unable to serialize lambda functions (see
https://stackoverflow.com/a/19985580) for detals.

Unfortunately, given how our tests are structured, we require that the
function under test is passed as an argument to another function, so we
cannot sidestep our use of lambda functions.

To resolve this problem, this patch makes use of the `multiprocess` and
`dill` Python modules, which together offers a multiprocessing mechanism
that can serialize lambda functions.  The multiprocess module also
offers a process pool, which simplifies the code for our parallel
testing framework.
2022-09-29 12:07:43 -05:00
Vivek Khandelwal bce00c8ed1 [tosa] Fix torch.vtensor.literal lowering
Signed-Off By: Vivek Khandelwal<vivek@nod-labs.com>
2022-09-29 17:03:10 +05:30
JakopinA 8ef0c874c2
Implement Expand/Collapse Functionality for Aten.View (#1353) 2022-09-27 11:08:14 -07:00
Eric Kunze cb1b8796a2
Convert torch si literals into signless for TOSA (#1421) 2022-09-26 16:54:27 -07:00
武家伟 c03aa63325
[MLIR] Add canonicalizer for aten.slice.t op (#1413)
* [MLIR] Add canonicalizer for aten.slice.t op

* Add mlir tests and strength the canonicalizer

* rename variable

Co-authored-by: Vremold <xremold@gamil.com>
2022-09-26 14:35:50 -07:00
Tanyo Kwok 16dd7e2e5f
Fix dynamic shapes type verifications (#1409)
* Fix dynamic shapes type verifications
2022-09-23 20:50:29 +08:00
Tanyo Kwok 061a97c3f2
Replace empty_like && empty_memory_format with full/full_like (#1398)
* Replace empty_like && empty_memory_format with full/full_like

* fix broadcast rank0 tensor
2022-09-23 10:24:36 +08:00
long.chen 797feaf129
[torch-mlir][Tosa] fix during torch.max.dim lower to tosa the reshape's new shape attr mismatch reshape's result type (#1378) 2022-09-16 21:29:56 -07:00
Ashay Rane a9e1014fc7
python: add `CHECK-LABEL` statements to localize checks (#1363)
It seems as though an upstream change in PyTorch has caused the module
dump to include not just the module being tested, but also several
seemingly unrelated functions in the `torch._decom.decompositions`
namespace.  The presence of these new functions caused lit to match
variables against incorrect statements (i.e.  statements in the
unrelated functions instead of the module under test).

This patch inserts `CHECK-LABEL` statements in the failing tests so that
lit ignores these unrelated functions and only checks the statements at
or after the test module definition.
2022-09-13 15:44:13 -05:00
gpetters94 48418b9c22
Fold away type_as (#1358) 2022-09-12 18:59:12 -04:00
Ashay Rane e52e886845
build: update llvm tag to 00d648bd (#1307)
- Update MHLO commit to build with LLVM commit hash 00d648bd
 - Update TorchToMhlo code to work with Stablehlo
 - Re-enabled two failing TOSA tests, thus resolving Github Issue #1231
2022-08-30 14:44:00 -05:00
Sean Silva 0e3ddbac91 Remove VerifyInvariantsBeforeBackendLowering
LowerToBackendContract now checks all this consistently.
2022-08-26 10:24:43 -07:00
Tanyo Kwok 3d0e18bbe7
Add decomposition for aten.roll (#1170)
* Add decomposition for aten.roll

* add e2e unittest

* refine type of torch.roll

* fix aten::cat output type
2022-08-24 08:36:05 +08:00
Tanyo Kwok 9176b5ed29
Add decomposition for aten.flatten.using_ints (#1161) 2022-08-23 11:52:54 +08:00
Sean Silva 01290d134a Add a way for backends to control which ops are legal for them.
We were already hitting many cases where backends different in terms of
the legal ops that they wanted. This caused unnecessary coupling between
the backends. Examples:
- https://github.com/llvm/torch-mlir/pull/1161
- https://github.com/llvm/torch-mlir/pull/862

This PR centralizes all compilation to go through `torch_mlir.compile`
so that we can keep the logic centralized there. We should move these
lists closer to each backend. Especially cases like
https://github.com/llvm/torch-mlir/pull/862 where blocking a
decomposition is necessary to avoid a crash emphasize that the set of
decompositions is tightly coupled to the backend, and should be
"controlled by the backend" and not something arbitrarily tweakable.

Also:
- Fix a small bug in the way we passed through the backendLegalOps
  option.
- Add better error messages in `torch_mlir.compile` for import errors.
2022-08-22 14:16:13 -07:00
武家伟 99fb4c8637
Add folder for ToF64Op and FromF64Op (#1257) 2022-08-22 09:49:39 +08:00
Vivek Khandelwal 65d811e267 [MLIR][TORCH] Fix dynamic cases for aten.index.Tensor 2022-08-19 12:13:20 +05:30
武家伟 7bd173a1c4
[MHLO] Eliminate explicit dynamic output shape generating in converting AtenSliceTensorOp (#1245)
[MHLO] Eliminate explicit dynamic output shape generating in converting AtenSliceTensorOp
2022-08-19 10:14:57 +08:00
Ramiro Leal-Cavazos 9bc606c384
Add support for returning more than one copy of the same tensor (#1228)
One of the simplifications made by the pass `RefinePublicReturn`
currently only happens if the tensor in question only has one
user. However, the current method of checking this does not correctly
handle the case of a user having multiple uses of the same
tensor. This commit makes sure only unique users are considered.
2022-08-18 22:41:45 +00:00
Sean Silva 283e0f141a Add a concept of "backend legal ops".
This is a first step towards formalizing the set of ops in our backend
contract. The goal is to eventually formalize `torch` dialect ops into 3
categories:
1. Legal in backend contract
2. Illegal in backend contract
3. Conditionally legal in backend contract

The "conditionally legal" set are the ops that we can optionally
decompose for backends.

This patch adds relevant pass options for this throughout the compiler,
in preparation for a new set of traits which will formalize this
classification.
2022-08-18 11:46:50 -07:00
Sean Silva 57681f7947 Iteratively run the main simplification pipeline.
This introduces a new pass LowerToBackendContract (better name very
welcome) which performs the bulk of the simplifications that we do,
such as
- shape refinement
- dtype refinement
- maximizing value semantics
- inlining global slots
- decomposing complex ops

The key difference from before is that it iterates the set of
transformations, which can help to break a number of "catch-22" issues
where one simplification depends on another, the latest example being
here:
https://github.com/llvm/torch-mlir/issues/1131

This also exposed that RefineTypes was sometimes crashing/asserting for
certain inputs. This commit hardens it a bit.
2022-08-17 14:54:33 -07:00
Yan Xu 9be8997536
Revert "add native_dropout and related ops pattern (#1211)" (#1230)
This reverts commit c935795086.
2022-08-17 13:48:10 +08:00
武家伟 11a5b5ac52
[MHLO] Add AtenRSubScalarOp conversion pattern to MHLO (#1233)
* [MHLO] Add AtenRSubScalarOp conversion pattern
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-17 09:07:36 +08:00
Ashay Rane 84d345c650
build: update llvm tag to 2dde4ba6 (#1229)
Summary of changes:
 - Tensor dialect now sets `emitAccessorPrefix` to prefixed, thus
   requring updates to methods that retrieve arguments
   [https://reviews.llvm.org/D131361]
 - Update MHLO to build with LLVM commit hash 2dde4ba6
 - Replace `AbsOp` with `AbsFOp` [https://reviews.llvm.org/D131325]
 - Replace deprecated `getValue()` with `value()`
   [https://reviews.llvm.org/D131349]
 - Remove `AnalysisState::defaultInitialize()`
   [https://reviews.llvm.org/D131746]
 - Update MHLO MLIR tests to use the updated assembly format
 - Disabled two failing TOSA tests (Github Issue link:
   https://github.com/llvm/torch-mlir/issues/1231)
2022-08-15 23:54:45 -07:00
武家伟 3b3cb99ef8
Generalize canonicalization pattern for more aten.sub/div/mul/add op (#1209)
Generalize canonicalization pattern for more sub/div/mul/add op, but for AtenDivTensorModeOp in 'trunc' rounding mode, we try to fold it.
2022-08-16 13:24:08 +08:00
Yan Xu c935795086
add native_dropout and related ops pattern (#1211) 2022-08-15 09:28:47 +08:00
Ramana Radhakrishnan 738f4fe96a
Rename TorchToStd pass as TorchToArith (#1163)
All the converters in this pass appear to create ops from the
arith dialect. Hence the full rename.

Fix GH Issue #409.
2022-08-10 20:12:51 +01:00
武家伟 87562773f8
[MHLO] Add AtenCatOp conversion pattern to MHLO (#1208)
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
Co-authored-by: Vremold <xremold@gamil.com>
2022-08-09 22:12:34 -07:00
Ashay Rane bb47c166a0
llvm: update tag to 061e0189 (#1180)
Summary of changes:
 - Switch to C++17 (similar to https://reviews.llvm.org/D131348)
 - Update MHLO to build with LLVM commit hash 061e0189
 - Replace deprecated `hasValue()` and `getValue()` with `has_value()`
   and `value()` respectively (https://reviews.llvm.org/D131349)
 - Use `TypedAttr` (https://reviews.llvm.org/D130092)
 - Use updated assembly format of `mhlo.compare` op (commit
   d03ef01e70fbf9afd0fa1976fbb7ed31838929b3 in MHLO repo)
2022-08-08 20:17:35 -07:00
武家伟 351f15424e
[MHLO] Add transposed convolution conversion pattern (#1171)
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-09 09:50:07 +08:00
Sean Silva 504de5e701 Rework how global slot initializers work.
Rather than a per-global-slot initializer region, we now have one for
the whole module. For example, it might look like this:

```
torch.global_slot "private" @tensor : !torch.tensor
torch.global_slot "private" @list : !torch.list<tensor>
torch.global_slot.module_initializer {
  %0 = torch.tensor.literal(dense<0.0> : tensor<f32>) : !torch.tensor
  %1 = torch.prim.ListConstruct %0 : (!torch.tensor) -> !torch.list<tensor>
  torch.initialize.global_slots [
    @tensor(%0 : !torch.tensor)
    @list(%1 : !torch.list<tensor>)
  ]
}
```

This new structure allows GlobalizeObjectGraph to create the initializer in a
much simpler way, avoiding the need to reason about whether different slots
alias each other. Reasoning about whether slots alias each other now is the
responsibility of InlineGlobalSlots, which has to do a much more complicated
analysis, implemented using MLIR's dataflow analysis framework.

Recommended review order:
- Check out the new IR constructs in the .mlir files of various passes
- Op definitions (*.td)
- Changes to GlobalizeObjectGraph pass.
- InlineGlobalSlots pass (~total rewrite)
- Misc changes:
  - Moving torchMlirAdjustStaticInformation for sharing with C++ code.
  - EraseModuleInitializer pass

To make this a bit nicer, it would be good to have a `torch.module` op
with an initializer region attached. That would be more invasive though.

This change has highlighted certain aspects of our project layering
which are worth calling out. None of our backends can handle global
slots, so we enforce that there are no global slots before backend
lowering. At an earlier stage in the project, we had aspirations of
transparently handling mutable global state and such, but for reasons
described below, that is no longer a goal. So really global slots should
be seen as a progressive lowering step as part of inlining all the
IValue's in the original program (GlobalizeObjectGraph is also one such
step).

Over time, with insights from work like IREE-JAX, it has become clear
that there isn't a reliable programming model we can compile for users
where we just transparently handle mutable global state (and some other
things, like lists and dictionaries). There is a need for an "outer
program" that orchestrates more restricted subroutines of the kind we
can handle in our compile flow here. The benefit of that is that it
decouples considerations like shapes, dtypes, etc. from the program
constructs used in the outer program. As long as the outer program can
efficiently invoke (pipelining/async/etc.) high-performance
data-parallel numerical subroutines of the kind we compile in our flow
here, then there is a complete programming model. This is also
consistent with the direction of upstream PyTorch which is becoming more
tracing-based (which inherently loses a lot of program structure, which
then has to be applied back with an "outer program" orchestrating the
traced subroutines).
2022-08-08 18:12:06 -07:00
Tanyo Kwok 290d7755fb
importer: add initial support for loading Float16 tensors (#1169)
follow up #761:

    This patch updates the `torch_mlir::convertTensorToMlirElementsAttr()`
    method to enable the creation of tensors whose base type is Float16.
    This patch also adds a test to validate the IR generation, and it
    updates the test for importing tensors of various types.
2022-08-08 12:37:31 +08:00
Tanyo Kwok 1ee865983b
[MHLO] fix tensor mode aten.div op pattern (#1160)
* [MHLO] fix tensor mode aten.div op pattern

See RFC #999
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-06 23:38:06 +08:00
武家伟 c94431f71c
[MHLO] Add convolution op pattern (#1152)
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-04 00:41:35 -07:00
武家伟 d030591df9
[MHLO] Init MHLO pooling-like op conversion (#1141)
* [MHLO] Init MHLO pooling-like op conversion and remove 'op' suffix in filenames

Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>

See RFC #999
2022-08-04 12:34:22 +08:00
Tanyo Kwok f0a24f59f6
[MHLO] Init MHLO linear op patterns (#1132)
See RFC https://github.com/llvm/torch-mlir/issues/999

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com
2022-08-03 19:10:54 -07:00
武家伟 636f5acb10
[MHLO] Init MHLO reduce-like op conversion (#1133)
* [MHLO] init reduce-like op conversion from Torch to MHLO
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-03 10:47:52 +08:00
Tanyo Kwok 0b23af27d3
[MHLO] support non-constant torch scalar in BasicOps (#1134)
See RFC https://github.com/llvm/torch-mlir/issues/999

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com
2022-08-03 08:16:31 +08:00
Yan Xu 704efdc259
[MHLO] add aten::gelu op pattern (#1127)
add aten::gelu op pattern, and moved some unit tests from basic.mlir to elementwise.mlir
2022-08-02 15:01:30 +08:00
武家伟 76c976682c
[MHLO] Support for dynamic shape in basic op conversion by introducing CHLO dialect (#1123)
* [MHLO] Support for dynamic shape in basic op conversion by introducing CHLO dialect
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>

* [MHLO] Support I32 as shape tensor dtype

* [NFC] Add a 'TODO' annotation
2022-08-02 12:53:24 +08:00
Jae Hoon (Antonio) Kim 425362263b Clean up Autogen (#1112)
* Remove unnecessary sed in autogen

* Remove .pyc files frrom VCS
2022-07-30 09:40:02 -04:00
Jae Hoon (Antonio) Kim 1bde00c73d Fix LTC Decoupling (#815)
* Initial changes

* Fix up native functions

* Further fix decoupling

* Remove unnecessary ops

* Formatting and copyright banners:

* Add pytorch submodule
2022-07-30 09:40:02 -04:00
PhaneeshB 8b5631d4c5 [MLIR][TORCH] Add decomposition for aten.std.dim Op
Signed-Off By: Phaneesh Barwaria <phaneesh@nod-labs.com>
2022-07-29 23:52:54 +05:30
Vivek Khandelwal d386b8f9e5 [MLIR][TORCH] Add decomposition for aten.var.correction op
This commit adds the decomposition for `aten.var.correction` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com
2022-07-29 11:08:57 +05:30
Vivek Khandelwal 7247c6a3a7 [MLIR][TORCH] Add E2E support for aten.ge.int op
This commit adds lowering of `aten.ge.int` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-07-29 11:08:57 +05:30
Quinn Dawkins 11a8901078
[MLIR][TORCH] Add support for multiple indexing tensors for aten.index.Tensor (#1097)
- Includes a canonicalizer for `aten.add.t`needed for successfully lowering the shape function
 - Only offers support for statically sized index tensors when there is more than one
 - Dynamic shape support remains for single indexing tensors
2022-07-28 19:00:02 -04:00
武家伟 052d2f84dc
[MHLO] Init MHLO basic op conversion (#1092)
* [MHLO] Init MHLO basic Op Conversion
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>

* [NFC] Remove 'from @llvm-project' annotation

Co-authored-by: wujiawei.jw <wujiawei.jw@bytedance.com>
2022-07-27 13:07:51 +08:00
Kevin Kiningham e8f327cc00 Add lowering to linalg for softplus and log1p
Follows existing conventions for unary operators.
2022-07-25 21:25:57 +05:30
Tanyo Kwok 44ead68772
[MHLO] Init MHLO gather op patterns (#1104)
See RFC https://github.com/llvm/torch-mlir/issues/999

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com
2022-07-25 23:47:46 +08:00
Tanyo Kwok f50d7013cd
[MHLO] Add [un]squeeze op patterns (#1099)
* [MHLO] Add [un]squeeze op patterns

* Conform to llvm coding standard

* minor update
2022-07-25 23:28:48 +08:00
Tanyo Kwok b80ce79b9f
[MHLO] Init MHLO view like op patterns (#1090)
* [MHLO] Init MHLO view like op patterns

See RFC: https://github.com/llvm/torch-mlir/issues/999

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com

* update filecheck test cases

* rebase, remove chlo and clang-format
2022-07-22 15:18:18 +08:00
Tanyo Kwok a02dbb2d5e
[MHLO] Init MHLO slice like op patterns (#1091)
See RFC: https://github.com/llvm/torch-mlir/issues/999

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com
2022-07-22 11:32:45 +08:00
Ramiro Leal-Cavazos f271e6a88c
Add verifiers for ToBuiltinTensorOp and FromBuiltinTensorOp (#1089)
This commit adds verifiers to the ops `ToBuiltinTensorOp` and
`FromBuiltinTensorOp` that make sure that the input and output have
the same shape and data type.
2022-07-21 21:41:45 +00:00
powderluv 31fd812acf
Add linux and macOS source builds in CI (#1070)
This enables building Pytorch from source in the CI.
The build should mostly hit the ccache.
Release builds will follow once we have some runtime on the CI.
2022-07-21 14:16:03 -07:00
Ashay Rane 72dd04cdb3
Revert "python: trim registration and loading of dialects and passes" (#1093)
This reverts commit ad283c1043, since it's
causing nightly build failures for all platforms.
2022-07-21 09:35:42 -07:00
Ashay Rane ad283c1043
python: trim registration and loading of dialects and passes (#1084)
In the interest of merging upstream LLVM quickly, a previous patch
(7f08169) updated the torch-mlir build to register all dialects and
passes through Python bindings.  This patch limits the dialects and
passes to only those that are used in torch-mlir.

Key to this change are the removal of
`MLIRPythonExtension.RegisterEverything` and the introduction of a new
Python module (`_mlir_libs/_site_initialize_0.py`), where we register
the dialects and passes used by torch-mlir.
2022-07-20 18:34:17 -07:00
Ziheng Jiang c61c99e887
[MHLO] Init MHLO integration. (#1083)
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-07-20 16:18:16 -07:00
Vivek Khandelwal 4c25878e64 [MLIR][TORCH] Add canonicalization pattern for prim.ListUnpack op
This commit adds the canonicalization pattern for the `prim.ListUnpack` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-07-18 13:51:25 +05:30
Vivek Khandelwal 3589134d31 [MLIR][TORCH] Add decomposition for aten.var.dim op
This commit adds the decomposition for `aten.var.dim` op.
This commit also make changes in the decomposition for `aten.var` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-07-15 09:53:42 +05:30
Ashay Rane 29bc48aedb
torch: add pass to catch non-value tensors (#1052)
This patch adds a new pass `torch-verify-conversion-to-value-semantics`,
which looks for non-value semantics tensors to catch such tensors early
during compilation.

This pass requires `torch-refine-public-return` pass to ensure that
return operations are updated to use value tensors, followed by the
canonicalize pass to remove any dead ops that may use or produce
non-value tensors.
2022-07-13 17:11:15 -07:00
Ashay Rane 64c04bd5f6
canonicalizer: [nfc] update LIT variable names for consistency (#1051)
A previous patch used lowercase names for LIT variables.  This patch
replaces them with uppercase names to maintain consistency with other
variables.
2022-07-13 12:28:25 -07:00
Ashay Rane ac4d7d10e0
canonicalizer: propagate type information across copy and cast ops (#1030)
Prior to this patch, the canonicalizers for `AtenSizeOp` and
`AtenSizeIntOp` succeeded only if the tensor operand's type information
included the size of the requested dimension(s).  We can extend the set
of optimizable cases by propagating types across operations whose result
type matches the input tensor type.

Specifically, this patch enables the canonicalizers for `AtenSizeOp` and
`AtenSizeIntOp` to see past `tensor_static_info_cast`,
`copy.to_vtensor`, and `copy.to_tensor` ops until it reaches the first
op whose result type contains size information for the requested
dimensions, with a maximum bound of 6 parent lookups to avoid indefinite
compilation times.  All other encountered ops cause the canonicalizer to
give up.
2022-07-12 12:38:37 -07:00
Sean Silva e5e11e214b GlobalizeObjectGraph: Clean up handling of unused slots
The way we did it previously still created the slot and copied the
initializer even if unused.
2022-07-12 10:47:28 -07:00
Ashay Rane 9017be9e9e
torch: copy uses to prevent iterator invalidation (#1033)
Prior to this patch, the code in the `torch-simplify-shape-calculations`
pass iterated on the uses of an op's result while also modifying the
value.  This caused the iterator to get invalidated, thus terminating
the loop early and producing incorrect IR.  This patch makes use of
`llvm::make_early_inc_range()` to ensure that the iterator is not
invalidated while executing the loop body.
2022-07-11 18:47:04 -07:00
Ramiro Leal-Cavazos 11148e60d6
Undo shape lib changes + update function signature of sum + zero (#1035)
This commit does three things:
  1. Reverts some of the shape lib changes merged in
  https://github.com/llvm/torch-mlir/pull/844
  2. Updates the signature of `aten.sum_dim_IntList` that was recently
  updated in
  23bdb570cf
  3. Replaces `aten.zero.functional` with `aten.zero`, updated in 960758b0b7
2022-07-11 10:56:12 -07:00
Prateek Gupta 2d75654b2c [TORCH][MLIR] Add lowering of `aten.slice_scatter` and
`aten.select_scatter` op.

This commit adds:
1.  Lowering of `aten.slice_scatter` op into `tensor.insert_slice`
op.
2. Decomposes the `aten.select_scatter` op into `aten.slice_scater`
op.

Signed-Off-By: Prateek Gupta <gprateek93@gmail.com>
2022-07-11 14:07:21 +05:30
Ashay Rane 340d8af28a
torch: handle `torch.prim.dtype` ops during type refinement (#1013)
The canonicalizer converts `torch.prim.dtype` ops into integer constants
for valid types, but the type may not be known until type refinement is
complete.  However, type refinement cannot make progress until
`torch.prim.dtype` ops have been resolved to their corresponding integer
constants, thus creating a circular dependency.

This patch creates a tight coupling between type refinement and the
lowering of `torch.prim.dtype` ops by handling such ops as they are
encountered during type refinement.  The unit test in this patch aims to
check whether the type refinement pass can now handle chains of
operations that alternate between type construction and type refinement.
2022-07-08 16:38:51 -07:00
Ramiro Leal-Cavazos 6a72ab4502
Add basic support for list of optional tensors in reduce-op-variants (#971)
This commit adds support for lists of type `list<optional<tensor>>`
where each element in the list is either a `!torch.tensor` or a
`!torch.none`.
2022-07-08 11:12:15 -07:00
Quinn Dawkins f0c3b5a7ed
Add E2E support for aten.len.str (#969) 2022-07-07 10:41:55 -07:00
Ashay Rane 88316b3b4e
torch: fold prim.dtype(bf16) to integer constant 15 (#1012)
A prior patch (63538de2) that added support for bfloat16 type did not
add the canonicalization pattern to fold `torch.prim.dtype` operations
on bfloat16 tensors into the integer constant 15.  This patch fixes the
problem.
2022-07-06 18:21:43 -07:00
Tanyo Kwok d4f1f41435
[MLIR][TORCH] Add decomposition of aten.repeat (#932)
* [MLIR][TORCH] Add decomposition of aten.repeat

* refine & rebase

* refine static shapes

* add e2e test

* Rebase and Refine naming style
2022-07-01 13:02:31 +08:00
Ashay Rane f947443f98
python: lower `prim::{Load,Store,Enter,Exit}` nodes to torch dialect (#983)
TorchScript nodes like `prim::Load` and `prim::Store` aren't supported
in torch-mlir because they can't be lowered to backends, but such nodes
can occur in the TorchScript IR.

This patch adds a rudimentary translation from such nodes to
corresponding ops in the Torch dialect.  Since we expected such nodes to
go away during lowering because of the SymbolDCE pass, this patch does
not add code to lower these ops beyond the Torch dialect.
2022-06-30 13:17:35 -07:00
Sean Silva 227dea7b2e Add support for ScalarType::QUInt8
I ran into this while poking around at
https://github.com/llvm/torch-mlir/issues/959
2022-06-29 15:33:28 -07:00
Ashay Rane 163fa57cde
torch: allow torch dialect ops after running drop-shape pass (#979)
In the `pyhpc_turbulent_kinetic_energy` TorchBench benchmark, the shape
calculation occurs inside loops, but because `DropShapeCalculationsPass`
does not explicitly mark the Torch dialect as legal, the pass execution
fails.

This patch adds Torch to the list of legal dialects, and adds a test to
validate the translation.
2022-06-25 07:27:47 -07:00
Ashay Rane 234fc7fe0c
linalg: lower `aten.triu` op to `linalg.generic` (#965)
Prior to this patch, the torch dialect included `AtenTriuOp` for
computing the upper triangular part of the input matrix, but there was
no code for lowering the op to the linalg dialect.

This patch adds code to generate a `linalg.generic` operation that
compares indices (computed using `linalg.index`) to choose between zero
or the original value (using `arith.select`).  The lowering fails if the
number of dimensions are less than two.  This patch also adds a few
end-to-end tests.
2022-06-23 22:45:48 -07:00
Tanyo Kwok 143a7bcb76
[MLIR][TORCH] Add folder for torch_c.from_i64 & torch_c.to_i64 (#933)
* [MLIR][TORCH] Add folder for torch_c.from_i64 & torch_c.to_i64

* add unit tests for each individual fold

* fix failure of NumelZeroRankModule & TestMultipleTensorAndPrimitiveTypesReturn
2022-06-24 09:34:39 +08:00
erman-gurses 5cff40c88a Add canonicalization for aten.add.tensor op 2022-06-23 17:24:59 -04:00
Maksim Levental 829717c96e
Bump LLVM (#958) 2022-06-22 22:23:46 -05:00
Vivek Khandelwal 77ab31641f [MLIR][TORCH] Add decomposition of aten.numpy_T op
This commit adds the decomposition of `aten.numpy_T` op into
`aten.t` or `aten.permute` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-06-16 00:01:22 +05:30
Vivek Khandelwal 33fa8e7761 [MLIR][TORCH] Add decomposition of aten.floor_divide op
This commit adds the decomposition of `aten.floor_divide` op into
`aten.div.Tensor_mode` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-06-14 08:56:25 +05:30
Vivek Khandelwal a11ef674a7 [MLIR][TORCH] Add E2E support for aten.baddbmm op
This commit decomposes `aten.baddbmm` op into `aten.bmm`,
`aten.mul.Scalar`, and `aten.add.Tensor` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-06-07 22:26:28 +05:30
Vivek Khandelwal 2718b4d838 [MLIR][TORCH] Add E2E support for aten.clamp_[min|max] op
This commit decomposes `aten.clamp_min` and `aten.clamp_max` op
into `aten.clamp` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-06-06 11:52:29 +05:30
Henry Tu abf5c94a1b
Replace valsem.aten.zero with aten.zero.functional (#893) 2022-06-03 16:27:31 -04:00
Vivek Khandelwal 06750815d1 [tosa] Support for AtenAvgPool2d op
Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-27 07:56:37 +05:30
Vivek Khandelwal 6f548fc3ad [MLIR][TORCH] Add decomposition of aten.adaptive_avg_pool2d op
This commit adds the decomposition of `aten.adaptive_avg_pool2d` op into
`aten.avg_pool2d` op. The current decomposition only supports cases where
input size is equal to the output size.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-27 07:56:37 +05:30
Vivek Khandelwal 56e77d4213 [MLIR][TORCH] Add E2E support for aten.Bool.[float|int] op
This commit adds lowering of `aten.Bool.float` and `aten.Bool.int` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-24 21:18:34 +05:30
Vivek Khandelwal 014a6d16c7 [MLIR][TORCH] Add E2E support for aten.any.bool op
This commit adds lowering of `aten.any.bool` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-24 17:24:28 +05:30
Vivek Khandelwal bc9b2156e3 [MLIR][TORCH] Add E2E support for aten.sqrt.int op
This commit adds lowering of `aten.sqrt.int` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-24 16:50:39 +05:30
Sean Silva 3fb54cba4c torch.prim.TupleIndex: Adjust tensor types when folding.
In cases where a refinement/derefinement was needed, we didn't fold.

Fixes https://github.com/llvm/torch-mlir/issues/863
2022-05-19 09:36:27 -07:00
Ashay Rane bb52a460cb
mlir: bump llvm tag to 5380e3 (#856)
In addition to updating the llvm-project submodule, this patch also:

1. updates shape functions and tests so that `func` and `call`
   operations refer to the `func` dialect
2. avoid duplicate registration of dialects
2022-05-16 12:54:35 -07:00
Ramiro Leal-Cavazos 96f90efd16
Add shape info to `rand_like` + support for `dtype` flag (#851)
The op `aten.rand_like` was missing a shape function, unit tests, and
the `dtype` argument was being ignored in its decomposition. This
commit fixes all three things.
2022-05-12 16:00:59 -07:00
Yi Zhang 28be6511d2 Fix type promotion code for scalar only operations
Fix the type promotion code for scalar only operation to return
TorchType which is the type tracked in ValueKnowledge.scalarType.

- Fix `getPromotedResultScalarType` to return Torch type.
- Add `getBuiltInTypeForTorchScalar` helper to convert scalar type
to builtin type before passing to the next level type promotion
helper `updateResultTypeState`.
- Add `setScalarType` helper to make setting ValueKnowledge.scalarType
  easier.
2022-05-07 10:37:21 -04:00
Vivek Khandelwal 96fabc0036 [MLIR][TORCH] E2E support for [ge|ceil].float, [ge|ne|gt].float_int op
This commit adds lowering of `aten.ge.float`, `aten.ge.float_int`,
`aten.ne.float_int`, `aten.gt.float_int` and `aten.ceil.float` op.
This commit also fixes formatting for the file scalar.py and scalar_comparison.py.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-05 21:48:35 +05:30
Yi Zhang 9f7264a7a4 Add support for scalar type propagation
The main changes are:
- Added `ValueKnowledge.scalarType` to track scalar type information.
- Added `ValueKnowledge.kind` to indicate the value kind.
- Modified the meet and join helper functions. The ValueKnowledge has
slightly more complicated state now so the meet and join function need
to look at the `kind` field in addition to just the type field.
2022-05-04 16:57:56 -04:00
Sean Silva 32159c4e54 Fix TupleIndex canonicalizer.
It would change the result type.
2022-05-03 09:08:49 -07:00
Sean Silva ab5ad7af09 Add tracing suport to `torch_mlir.compile`.
This also has a fix for the adjustment of types of TupleConstruct
inputs, which I found when using this new functionality on a model.

Some scenarios in tracing create situations where the output of
TupleConstruct has a more refined type than the inputs.

This introduces a helper `adjustStaticInformationForValues` which
subsumes the `derefineValues` helper and the tensor static information
adjustment we were doing.
2022-05-03 09:08:40 -07:00
Vivek Khandelwal c0634bc996 [MLIR][TORCH] Add E2E support for aten.to.dtype_layout op
This commit decomposes `aten.to.dtype_layout` op into `aten.to.dtype` op.
This commit also fixes the formatting for the file type_conversion.py.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-05-03 12:48:58 +05:30
Prateek Gupta 81ee5bb58c [TORCH][MLIR] Fix ConstantPad2dStaticModule test.
This commit fixes the `ConstantPad2dStaticModule` test case by adding
the lowering of `aten.pad` operation. Previously the test case
mapped to `aten.constant_pad_nd` operation.
The `aten.pad` now decomposes into `aten.constant_pad_nd` operation.

Signed-Off-By: Prateek Gupta <prateek@nod-labs.com>
2022-04-29 21:57:01 +05:30
Ashay Rane 809f240f01
importer: add initial support for loading BFloat16 tensors (#761)
This patch updates the `torch_mlir::convertTensorToMlirElementsAttr()`
method to enable the creation of tensors whose base type is BFloat16.
This patch also adds a test to validate the IR generation, and it
updates the test for importing tensors of various types.
2022-04-29 09:01:49 -07:00
Prateek Gupta e1db318a3c [TORCH][MLIR]Add lowering for control flow operations.
1. This commit adds lowering of "while-like" prim loop to scf.while
operation.
2. Adds lowering of "for-like" prim loops to scf.for operation.

Signed-Off-By: Prateek Gupta <prateek@nod-labs.com>
2022-04-29 16:25:58 +05:30
Sean Silva 44c7b181d3 Revert "[MLIR][TORCH] Add E2E support for aten.ge.float op"
This reverts commit 564734b2d7.
2022-04-28 07:49:58 -07:00
Sean Silva eff144c0b7 Revert "[MLIR][TORCH] Add E2E support for aten.ge.float_int op"
This reverts commit 1f102cc400.
2022-04-28 07:49:58 -07:00
Sean Silva 7669ee4e4a Revert "[MLIR][TORCH] Add E2E support for aten.ne.float_int op"
This reverts commit 51dd462592.
2022-04-28 07:49:58 -07:00
Sean Silva 5ef9f501fa Revert "[MLIR][TORCH] Add E2E support for aten.ceil.float op"
This reverts commit 78f5747568.
2022-04-28 07:49:58 -07:00
Vivek Khandelwal 78f5747568 [MLIR][TORCH] Add E2E support for aten.ceil.float op
This commit adds lowering of `aten.ceil.float` op.
This commit also fixes formatting for the file scalar.py.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-28 11:49:35 +05:30
Vivek Khandelwal 51dd462592 [MLIR][TORCH] Add E2E support for aten.ne.float_int op
This commit adds lowering of `aten.ne.float_int` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-27 21:16:48 +05:30
Vivek Khandelwal 1f102cc400 [MLIR][TORCH] Add E2E support for aten.ge.float_int op
This commit adds lowering of `aten.ge.float_int` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-27 21:16:48 +05:30
Vivek Khandelwal 564734b2d7 [MLIR][TORCH] Add E2E support for aten.ge.float op
This commit adds lowering of `aten.ge.float` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-27 21:16:48 +05:30
Vivek Khandelwal f5b6c4b601 [MLIR][TORCH] Add E2E support for aten.div.float op
This commit adds lowering of `aten.div.float` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-27 21:16:48 +05:30
Sean Silva 73cc2ac152 Ensure that imported function input type and block arg types are consistent.
I wasn't able to find exactly what frontend situation created it, but
`torch.jit.trace` will sometimes create functions where the
`jit::Block`'s param node has refined tensor types. So we need to adjust
the function's formal param types to those refined types.
2022-04-27 08:01:23 -07:00
Ashay Rane 9208bf0eb6
llvm: bump tag to e1318078 (#781)
The updated LLVM code includes a patch to create bfloat16 array
attributes, thus enabling a different patch to torch-mlir to flesh out
support for the bfloat16 type.
2022-04-26 12:27:51 -07:00
Ashay Rane 9ec4712516
types: allow bf16 as result type for various tensor ops (#798)
Prior to this patch, the result type for several tensor operations could
only be float32, float64, or null.  This patch adds bf16 to the list of
allowed result types.
2022-04-26 11:55:58 -07:00
Vivek Khandelwal 769f3a8870 [MLIR][TORCH] Add E2E support for max_pool2d_with_indices op
This commit adds lowering of `max_pool2d_with_indices` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-04-18 21:05:19 +05:30
Ashay Rane a893c7d5cf
Add shape transfer function and lowering to linalg for aten.neg (#759)
* shape: add shape transfer function for aten.neg

Prior to this patch, the list of shape transfer functions did not
include `aten.neg`, which resulted in errors like below.

```
error: unsupported by backend lowering: tensor with unknown rank or dtype
note: see current operation: %0 = "torch.aten.neg"(%arg0) :
  (!torch.vtensor<[256,256],f32>) -> !torch.vtensor<*,f32>
note: this is likely due to a missing shape transfer function in shape_lib_gen.py
```

This patch fixes the problem by adding a shape transfer function to
reflect the point-wise nature of this operation.

* linalg: add translation of aten.neg operation

This patch adds a translation rule to lower `aten.neg` operations on
tensors to an `arith.negf` operation wrapped inside a `linalg.generic`
operation.  This patch also adds a rudimentary test.
2022-04-15 11:11:22 -07:00
Sean Silva e7721fb784 Fix error message.
RefineTypes doesn't handle shape refinement anymore.
2022-04-07 14:46:44 -07:00
Sean Silva c17c0a6ba2 Fix for 0-size dim inferred incorrectly.
The issue was in the canonicalizer for torch.aten.ge.int -- in cases
where the operands were swapped, it would miscompile. This issue is
fixed and folding support generalized to `torch.aten.size.int < 0` as
well.

Fixes #716
2022-03-30 16:36:15 -07:00
Gaurav Shukla 969785d1b6 [LINALG] Add E2E support for `aten.where.[Scalar|ScalarSelf|ScalarOther]` ops
This commit decomposes different variants of `aten.where.*` op into
`aten.where.Self` op. It covers `aten.where.Scalar`,
`aten.where.ScalarSelf` and `aten.where.ScalarOther` ops.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-03-30 20:36:48 +05:30
Vivek Khandelwal 2597c481f6 [MLIR][TORCH] Add E2E support for aten.new_empty op
This commit decomposes `aten.new_empty` op into `aten.empty.memory_format` op.

This commit also made a dtype fix to the constant tensor allocation like ops.
Earlier the dtype for the result was inferred from the result type; now, it's
being evaluated as per the original definition of the op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-30 13:21:01 +05:30
Sean Silva 140babd952 Add minimal support for Union types.
A recent PyTorch commit made ConstantPad2d call a helper function with a
`Union[int, float]` type annotated. This commit adds minimal support for
representing and dealing with that.
https://github.com/pytorch/pytorch/pull/73287

Changes:
- Adding support for `!torch.union<T1, T2, T3>`/`Torch::UnionType`,
  along with the importer and CAPI code.
- Add support in isValidSubtype for union types.
- Adding a canonicalizer for `torch.derefine` to help simplify some code
  that derefines to a UnionType (this also fixes #664).

There is still more work to do for really supporting UnionType well,
such as canonicalizing UnionType's so that they can be compared with
pointer equality.
2022-03-29 17:45:48 -07:00
Liam Fitzpatrick f2269ced80
Improve list index normalization SimplifyShapeCalculations. (#710)
The reified code to compute the shape of torch.aten.constant_pad_nd
uses negative indices when setting list elements. This was not
converted to a positive offset in one place in SimplifyShapeCalculations
which prevented computation of the static shape.
2022-03-29 22:21:47 +02:00
Maksim Levental 25ba51b2af
This commit decomposes aten._reshape_alias op into aten.view op. (#690) 2022-03-28 23:54:28 -05:00
Sean Silva 776426ea4e [SimplifyShapeCalculations] Fix AbstractlyInterpretListOpsWithinABlock
The logic in the rewriting phase had a bug in case of a read-only op
coming before mutation ops. The logic would use the op itself as the
"latest literal", but that is not correct, because later on we replace
the op itself with the *final* "latest literal", assuming that all uses
of the op have been rewritten -- that was working in general, except for
any read-only ops at the beginning.

Big thanks to @ljfitz for the tiny reproducer!

Fixes #704
2022-03-28 13:18:35 -07:00
Anup Gangwar 5d7a6c2976
[tosa] Support for Aten[Unsqueeze|Contiguous|Dropout|Reshape|View] ops (#700) 2022-03-25 14:15:07 -07:00
Gaurav Shukla 02b6d04eb4 [LINALG] Add E2E support for `aten.zero_` op
This commit adds decomposition of `aten.zero_` op.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-03-25 12:46:50 +05:30
Gaurav Shukla 7c3ba25238 [LINALG] Add decomposition of `aten.dropout` op
- This commit adds decomposition of `aten.dropout` op. It also covers the
  training mode of the same op.
- It also adds lowering of `aten.sub.float` op.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-03-22 13:14:49 +05:30
Vivek Khandelwal 5b9bdfaf3f [MLIR][TORCH] Add E2E support for aten._to_copy op
This commit decomposes `aten._to_copy` op into
`valsem.aten.copy` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-21 19:12:37 +05:30
Vivek Khandelwal 13383b03b8 [MLIR][TORCH] Add value tensor variant to aten::copy_ op
This commit adds the op `ValsemVariantAtenCopyOp` that represents
`AtenCopy_Op` without the underscore. This is needed to make sure
that the `ReduceOpVariants` pass turns the in-place op into an op
that takes value tensors as inputs, otherwise the
`MaximizeValueSemantics` pass will not be able to add value
semantics correctly.

This commit also adds the lowering of `ValsemVariantAtenCopyOp`.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-21 19:12:37 +05:30
Vivek Khandelwal 4c0cd5c23d [MLIR][TORCH] Add E2E support for aten.expand_as op
This commit decomposes `aten.expand_as` op into `aten.broadcast_to` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-21 12:47:39 +05:30
Vigilans 63fb1e5aad Bump LLVM at 8361c5da30588d3d4a48eae648f53be1feb5cfad 2022-03-18 13:16:14 -04:00
Ramiro Leal-Cavazos 218b4875d5
Make conditions for type refinement of static cast less strict (#680)
This commit adds support for type refinement when
`torch.tensor_static_info_cast`s are involved, even when there are
users of the casted tensor that don't allow type refinements.

Originally the canonicalization pattern for
`torch.tensor_static_info_cast` would check if all the users of the
casted tensor allowed type refinements before making any changes. This
means that if at least one of the users did not allow type
refinements, the pattern would fail. This becomes an issue when doing
shape calculations because the calculations need the shape information
of each input tensor to be available before the calculation can be
simplified.
2022-03-18 09:10:12 -07:00
Vivek Khandelwal 8da7d90611 [MLIR][TORCH] Add E2E support for aten.index_put op
This commit decomposes `aten.index_put` op into
`valsem.aten.index_put_impl` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-16 22:02:02 +05:30
Vivek Khandelwal 3d95c3d6c9 [MLIR][TORCH] Add value tensor variant to aten::_index_put_impl_
This commit adds the op `ValsemVariantAtenIndexPutImplOp` that represents
`Aten_IndexPutImpl_Op` without the underscore. This is needed to
make sure that the `ReduceOpVariants` pass turns the in-place op
into an op that takes value tensors as inputs, otherwise the
`MaximizeValueSemantics` pass will not be able to add value
semantics correctly.

This commit also adds the lowering of `ValsemVariantAtenIndexPutImplOp` op.

This commit also updates the `torch.bincount` op test cases.
2022-03-16 22:02:02 +05:30
Ramiro Leal-Cavazos 0bcc6d1075
Add maximize-value-semantics support for multiple non-value tensor inputs (#659)
This commit adds value semantics support for ops such as
`aten.view_as` and `aten.expand_as` that take two non-value 
tensors as input.
2022-03-15 18:13:45 -07:00
Sean Silva 92da4988f0 Improve "pseudo" op terminology.
The term "pseudo" is very vague and was getting confusing (I felt I had
to explain it in every comment referencing it). Instead, rework the
"pseudo" ops to instead be named:

- MLIR Syntax: `torch.valsem.*`
- C++ / ODS: `ValsemVariant*Op`

This makes it clear what the concept is, and avoids confusion with other
things that might be called "pseudo", since these are very specific and
should be 100% consistently named w.r.t. the non-valsem-variant ops that
they correspond to.
2022-03-15 17:57:52 -07:00
Sean Silva 84a9693006 Elide `!torch.` prefix in nested dialect types.
This leads to much more succinct types in many cases:

```
!torch.list<!torch.int>
!torch.list<int>

!torch.tuple<!torch.list<!torch.int>, !torch.list<!torch.int>>
!torch.tuple<list<int>, list<int>>

!torch.optional<!torch.list<!torch.int>>
!torch.optional<list<int>>

!torch.list<list<list<tensor>>>
!torch.list<!torch.list<!torch.list<!torch.tensor>>>
```

I would like to take this further and allow omitting the `!torch.`
prefix in all cases, but that's harder -- for example, we currently use
`FuncOp` for functions, and so I don't think we can customize the
printing there. It seems like it will be a longer road to getting that
level of customization.
2022-03-15 17:24:08 -07:00
Sean Silva a5fe0cf063 Introduce new shape library design.
See the documentation in `docs/shape_lib.md` and
`docs/adding_a_shape_function.md` for an overview of the system.

This completely overhauls how we represent shape functions. In
particular, RefineTypes does not infer shapes anymore (only dtypes).
Shape functions are now written in (TorchScript'able) Python.

Recommended review order:

1. Read `docs/shape_lib.md` and `docs/adding_a_shape_function.md`.
1. Code and tests for ReifyShapeCalculations, DropShapeCalculations.
1. Code and tests for SimplifyShapeCalculations.
1. shape_lib_gen.py
1. Code and tests for new RefineTypes pass.
1. Random folders/canonicalizers in TorchOps.cpp and associated test in
   `canonicalize.mlir`.
1. New ReadOnly trait inferred from the registry.
1. Any miscellaneous remaining stuff.

Example `-print-ir-after-all` for ElementwiseUnaryModule:
[IR lowering dump](https://gist.github.com/silvasean/e4dc8cbc8d00aac7819602e3cbd8e212).

Example `-print-ir-after-all` for ElementwiseBinaryModule:
[IR lowering dump](https://gist.github.com/silvasean/daf6860ecced732af3568af6b1899113).
2022-03-15 12:41:58 -07:00
Ramiro Leal-Cavazos 51e267aa37
Combine maximize-value-semantics rewrite patterns into one pattern (#642)
This commit replaces the two rewrite patterns of
maximize-value-semantics with a single pattern that captures the
behavior of both as well as other edge cases previously not
supported. The new pattern works by first performing alias analysis on
a subgraph to see if pattern is applicable, then rewriting all
non-value tensors to value tensors in a single go.
2022-03-10 09:36:52 -08:00
Gaurav Shukla e57d3f9774 [LINALG] Fix `aten.bernoulli` op lowering
- This commit adds E2E support for `aten.rand_like` and
  `aten.bernoulli_.Tensor` ops.
- The `aten.bernoulli(x)` was implemented as:
  `aten.bernoulli(x) = rand_like(x) < 0.5`, assuming 0.5 as default
  probability, whereas according to the pytorch documentation:
  https://pytorch.org/docs/stable/generated/torch.bernoulli.html#torch.bernoulli
  the input x in `aten.bernoulli(x)` is itself a tensor containing
  probabilities to be used for drawing the binary random number.
- So this commit fixes the `aten.bernoulli(x)` implementation as:
  `aten.bernoulli(x) = rand_like(x) < x`.
- It also fixes the case where the input to `aten.bernoulli_.float` is
  an integer tensor. In this case the input must be casted to float type
  before passing it as operand to `aten.rand_like` op.
  `aten.bernoulli_.float(x, p) = rand_like(float(x)) < p`.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-03-05 09:38:22 +05:30
Vivek Khandelwal af551bd9cd [MLIR][TORCH] Add E2E support for aten.full_like op
This commit decomposes `aten.full_like` op into `aten.empty_like`
and `aten.fill` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-04 21:58:23 +05:30
Vivek Khandelwal d61ae92eee [MLIR][TORCH] Add E2E support for aten.full op
This commit decomposes `aten.full` op into `aten.empty` and
`aten.fill` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-04 21:58:23 +05:30
Ramiro Leal-Cavazos 9ce62473f9
Add static type information support to `aten.bmm` (#636)
This commit adds static type information support to `aten.bmm`. This
is needed for the forward pass of Bert training.
2022-03-03 13:01:17 -08:00
Ramiro Leal-Cavazos 5ec70c175d
[LINALG] Add torch-to-linalg lowering for `TensorStaticInfoCastOp` (#634)
This commit adds a lowering for `TensorStaicInfoCastOp` that simply
replaces the op with the `tensor::CastOp`.
2022-03-02 13:35:26 -08:00
Yi Zhang 1d285f0153 Add aten.hardtanh e2e support. 2022-03-02 12:28:06 -05:00
Prashant Kumar 819f29316f Decompose aten.silu op
Decomposition of aten.silu.op is added as silu(x) = x * sigmoid(x).
2022-03-01 23:24:19 +05:30
Vivek Khandelwal ddd45d6068 [MLIR][TORCH] Add E2E support for aten.new_zeros, aten.new_ones op
This commit adds lowering of `aten.new_zeros` and `aten.new_ones` op

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-03-01 22:09:47 +05:30
Prashant Kumar 7c637eebc3 [LINALG] Decompose aten_hardswish op.
`aten.hardswish` op is decomposed into (x/6) * Relu6(x+3).
2022-02-25 21:59:27 +05:30
Gaurav Shukla 056cd2078d Revert "[LINALG] Decompose `aten.batch_norm` into `aten.native_batch_norm`"
This reverts commit 442ff4605c.
2022-02-25 15:46:55 +05:30
Ramiro Leal-Cavazos ba29d4f250
Add operand type invariant to `torch.overwrite.tensor.contents` (#606)
This commit adds the invariant to the op `torch.overwrite.tensor.contents` that
both of its operands have the same shape and size. In order to
maintain the invariant, special handling of this op is added to the
`RefineTypes` pass.
2022-02-22 11:41:46 -08:00
Ramiro Leal-Cavazos ea371a9bf2
Fix handling of view-like ops in `maximize-value-semantics` (#611)
This commit adds handling to the `maximize-value-semantics` pass for
the case where a view-like op depends on a tensor that has been
overwritten by a value tensor. The approach for removing the
dependency is to change the input to the view-like op to be a copy of
the value tensor that is being used to overwrite.

This commit also removes `AtenFill_ScalarOp` and
`AtenBernoulli_FloatOp` from the list of view-like ops, since these
ops now have a corresponding op with value semantics into which they
get converted in the `reduce-op-variants` pass.
2022-02-18 10:19:07 -08:00
Ramiro Leal-Cavazos 2823277f7c
Add static type information support to `aten.mm` (#602)
This commit adds static type information support to `aten.mm`. This is
needed for the forward pass of Bert training.
2022-02-18 09:56:48 -08:00
Nirvedh f8cb32faf0 LLVM bump
Major changes: opTrait changed to Trait, selectOp moved to arith dialect
assertOp moved to cf dialect
2022-02-16 15:28:13 -05:00
Gaurav Shukla 442ff4605c [LINALG] Decompose `aten.batch_norm` into `aten.native_batch_norm`
- This commit decomposes the `aten.batch_norm` op into the
  `aten.native_batch_norm` op, instead of lowering it to the
  `linalg.generic` op.
- It also adds run-time asserts in the `aten.native_batch_norm` lowering
  to make sure that the shape of the weight, bias, running_mean, and
  running_var must match the num of features.
- Since the `aten.native_batch_norm` op is not supported at TOSA backend,
  all the modules that are dependent on the `aten.native_batch_norm` op
  will fail and therefore they should be removed from the TOSA `passing`
  set.
- It also moves `checkNotNone` to utility.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-02-16 23:41:38 +05:30
Anup Gangwar c60468f141
[tosa] Support for Aten[Zeros|Ones|Fill_Scalar] ops (#604)
Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>

Co-authored-by: Anup Gangwar <anup.gangwar@arm.com>
2022-02-16 09:53:51 -08:00
Prashant Kumar 8b79b5f48f Modify aten._log_softmax op decomposition for numerical stability.
`aten.log_softmax` is decomposed to be more numerically stable.
2022-02-16 12:26:17 +05:30
Gaurav Shukla cd21dda867 [LINALG] Add E2E support for `aten.Hardsigmoid` op
This commit adds lowering of `aten.Hardsigmoid` op.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-02-16 02:35:18 +05:30
Ramiro Leal-Cavazos 00a6e9c1bb
[LINALG] Add value tensor variant to `fill_.Scalar` (#600)
This commit adds the op `PseudoAtenFillScalarOp` that represents
`AtenFill_ScalarOp` without the underscore. The approach is the same
as in commit dd998fa4d4.

Adding this op allows for a simpler and more consistent version of the
`empty` and `empty_like` op e2e tests.
2022-02-15 11:58:03 -08:00
Ramiro Leal-Cavazos 413e6000d2
[LINALG] Add value tensor variant to `bernoulli_.float` (#597)
This commit adds the op `PseudoAtenBernoulliFloatOp` that represents
`AtenBernoulli_FloatOp` without the underscore. This is needed to make
sure that the `ReduceOpVariants` pass turns the in-place op into an op
that takes value tensors as inputs, otherwise the
`MaximizeValueSemantics` pass will not be able to add value semantics
correctly.
2022-02-14 18:58:48 -08:00
Gaurav Shukla 78c7844c6c [LINALG] Add E2E support for `aten.eq.int` op
- This commit adds lowering of `aten.eq.int` op as a part of
  `convert-torch-to-std` pass.
- It also refactors the code for binary comparison ops lowering.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-02-15 01:37:35 +05:30
Gaurav Shukla f00d1686c8 [LINALG] Add E2E support for `aten.[Bool.Tensor|Float.Tensor]` op
- This commit adds lowering of `aten.Bool.Tensor` and
  `aten.Float.Tensor` op as a part of `convert-torch-to-linalg` pass.
- It also adds support for returning bool types.
- It also fixes lowering of the `aten.Int.Tensor` op for non-zero rank
  input tensors.
- If a scalar number is converted to a 0-d tensor and passed on to the
  `aten.Float.Tensor` op, it folds to the scalar number.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-02-14 23:09:20 +05:30
Yi Zhang 9e7b6cab08 Add folder for aten.gt/lt.float 2022-02-14 12:34:01 -05:00
Henry Tu 73ac9a7e2e Added support for importing node prim::Constant with list type
Prior to this commit, importing a `prim::Constant` node with list type would result in an error since it was not supported. `ivalue_importer::importIValue` was modified to return the MlirValue corresponding to the root so its parent operation could be extracted.
2022-02-11 20:54:06 -05:00
Yi Zhang ce4d6d1f83 Remove hacky aten.select.int lowering code 2022-02-11 18:14:58 -05:00
Anup Gangwar 756b75fb2d
[tosa] Support for some ops and fix for Issue #532 (#575)
* [tosa] Support for AtenNe[Tensor|Scalar]Op, AtenLog2Op,
AtenBitwiseAndTensorOp, AtenSquareOp and AtenThresholdOp
* Fix for Issue #532 - Mixed input types for few ops and updated few
tests to use i32 instead of i64

Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>

Co-authored-by: Anup Gangwar <anup.gangwar@arm.com>
2022-02-11 12:30:02 -08:00
Prashant Kumar 258660deb6 Add aten.bernoulli decomposition.
aten.bernoulli is decomposed to aten.gtTensor(aten.uniform(x), x).
2022-02-11 00:35:33 +05:30
Prashant Kumar 102c497c4c Add decomposition of _log_softmax op.
Decompose _log_softmax into log(softmax(x)).
2022-02-10 23:17:26 +05:30
Prateek Gupta 318946a650 [TORCH][MLIR] Add E2E support for `aten._unsafe_view` op.
This commit adds decomposition of `aten._unsafe_view` op into
`aten.view` op.

Signed-Off-By: Prateek Gupta<prateek@nod-labs.com>
2022-02-10 22:28:58 +05:30
Gaurav Shukla bd177bdfc7 [TORCH][MLIR] Add run-time assert support in Torch-dialect
- This commit adds `aten.assert` op in the Torch dialect.
- The `aten.assert` op is lowered to `mlir::Assert` op.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2022-02-09 12:03:01 -05:00
Anup Gangwar f9f97ea184 * [tosa] Support for AtenNativeLayerNormOp
* [tosa] Support for AtenPermuteOp

Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>
2022-02-04 14:46:31 -05:00
Prashant Kumar 68acc8696e Modify softmax decomposition to be more numerically stable.
The softmax decomposition is modified according to https://github.com/pytorch/functorch/blob/main/functorch/_src/decompositions.pytorch
to account for numerical stability. Also, modified aten.argmax lowering
to handle negative dimension.
2022-02-03 21:20:36 +05:30
Suraj Sudhir 1b505cbac5
RefineTypes fixes for TOSA backend (#557)
Handles Linear, Adaptive_AvgPool2D and FlattenUsintInts
Adds ResNet18 static model for TOSA

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm.com>
2022-02-01 14:08:54 -08:00
Yi Zhang 0cb216a1ad [Torch][Linalg] Add basic support for RNG
This PR include the following pieces:
- Add torch `Generator` type. `Generator` type is converted to i64 in
refbackend type converter.
- Add seed managment support for the default global generator.
`torch_c.getNextSeed` op is used to get the seed. On refbackend, the
`torch_c.getNextSeed` is lowered to load/store from [0] of global
variable `default_generator` memref<i64> in `InsertRngGlobals` pass.
- Add `aten.uniform_` and testing as an example op for RNG ops. Add
`torch.pseudo.aten.uniform` op. It has the same operands and return as
the `aten.uniform_` from the op registry except for value semantics.
2022-01-31 18:56:42 -05:00
Yi Zhang 5d9a15263a [TORCH] Add aten.std e2e support 2022-01-31 15:17:49 -05:00
Prashant Kumar e58b66bc3b Add lowering of `aten.max.dim` op.
Lowering of `aten.max.dim` op has been added.
2022-01-31 21:41:22 +05:30
Anup Gangwar 454fa9d123
* [tosa] Support for AtenFlattenUsingIntsOp (#548) 2022-01-28 21:38:56 -08:00
Liam Fitzpatrick 8bc028af05 Fold __is__ and unchecked_cast of derefine
The added e2e maxpool testcase from #545 was not getting a static shape
due to an unfolded prim.If when RefineTypes was called. This was because
of unfolded torch.iaten.__is__ and torch.prim.unchecked_cast operators
with torch.derefine operands.
2022-01-28 17:54:40 -05:00
Anup Gangwar 7a5736facd
* [tosa] Support for AtenReshapeOp (#543)
* [tosa] Support for AtenBatchNormOp

Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>

Co-authored-by: Anup Gangwar <anup.gangwar@arm.com>
2022-01-27 14:38:59 -08:00
stephenneuendorffer 3fd9b7789e
Bump LLVM to 881ff4e4ebe8cc0cc045c7c167cffb01f94f27f8 (#539) 2022-01-25 22:16:30 -08:00
Anup Gangwar f8080bd1c5
* [tosa] Support for AtenRsubScalarOp for scalar constants (#531)
* [tosa] Support for AtenCeilOp and AtenReciprocalOp
* [tosa] Support for comparator ops, Aten[Gt|Lt|Eq][Tensor|Scalar]Op with scalar constant
* [tosa] Support for Scalar variants of Aten[Mul|Div|Add|Sub] Ops with scalar constants

Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>

Co-authored-by: Anup Gangwar <anup.gangwar@arm.com>
2022-01-20 10:58:30 -08:00
Vivek Khandelwal 6fe70c7794 [MLIR][TORCH] Add E2E support for aten.index.Tensor op
This commit adds lowering of `aten.index.Tensor` op

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-01-19 13:37:56 +05:30
dan 3745f54489 Update external/llvm-project
- Add `qualified` to ods because of
https://reviews.llvm.org/D113873 and https://reviews.llvm.org/D116905
- Needed to revert https://github.com/llvm/torch-mlir/pull/520 as it
was based on an old torch version.
https://github.com/llvm/torch-mlir/pull/527 will bring this back with
a better design.
- Change ConvertAtenCatOp to use more accurate tensor shape info and
as much static info as possible to pass `tensor.insert_slice`
verification code added by https://reviews.llvm.org/D114715
- Other minor fixes
2022-01-18 13:25:42 -05:00
Anup Gangwar d69d29b7a6 * [tosa] Support for AtenPowTensorScalarOp with constant Scalar as input
Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>
2022-01-11 22:55:54 -05:00
Liam Fitzpatrick 077e55d756 Add support for constant_pad_nd
Note that to enable folding of the code coming from an example
like the ConstantPad2dStaticModule e2e test, support for other
operations had to be added/improved:
- aten::neg.int
- aten::eq.float
- aten::eq.str
- prim::Uninitialized
2022-01-11 10:25:25 -05:00
Vivek Khandelwal 35cf8d18f7 Add support for two return values
This commit adds support for two return values of type
memref f32 and i64.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2022-01-11 11:07:10 +05:30
Yi Zhang 732a76f45c Make broadcasting result shape more static
This involes the following 2 parts:
- Change refine type to propagate more static shape info.
- Get as much static shape info as possible when creating the result
tensor when converting to linalg.
2022-01-06 18:39:27 -05:00
Liam Fitzpatrick ccfdfd1b80 Refine static shapes for conv2d and maxpool2d 2022-01-03 11:09:23 -06:00
Vivek Khandelwal 4486de5ef3 [MLIR][TORCH] Add E2E support for torch.arange op
This commit adds lowering of `aten.arange.start_step` op.
This commit decomposes `aten.arange` and `aten.arange.start` into
`aten.arange.start_step` op.

Signed-Off By: Vivek Khandelwal <vivek@nod-labs.com>
2021-12-27 22:45:48 +05:30
Gaurav Shukla a83004c806 [TORCH][MLIR] Fold trivial cases of `aten.to.dtype` and `aten.view` op
- It folds `aten.to.dtype` when the input tensor type and result type
  are exactly same.
- It folds `aten.view` when the rank of both the input tensor type and
  result type is unity.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2021-12-24 13:32:34 +05:30
xndcn 5eed562e19 add aten.sub.int/aten.mul.int lowering in TorchToStd 2021-12-17 10:35:15 -08:00
Anup Gangwar a6c3050dd0 * [tosa] Support for Maximum and Minimum
Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>
2021-12-15 11:58:19 -08:00
Prashant Kumar ab81f871e4 Add aten.tensor.int and aten.tensor.float op lowerings.
Add the required lowerings and correct test cases.
These op produce zero-d tensors and it was incorrectly mentioned in
refine types to produce 1d tensor of size 1.
2021-12-15 17:21:34 +05:30
Anup Gangwar cce490d71d
* [tosa] Support for Rsqrt legalization (#480)
Signed-off-by: Anup Gangwar <anup.gangwar@arm.com>

Co-authored-by: Anup Gangwar <anup.gangwar@arm.com>
2021-12-14 10:03:58 -08:00
Gaurav Shukla 5a47f92390 [TORCH][MLIR] Add E2E support for `aten.squeeze.dim` op
This commit adds lowering of `aten.squeeze.dim` op into
`linalg.TensorCollapseShape` op. Here, the dim(th) dimension of the
input tensor is not supposed to be dynamic.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2021-12-10 17:01:20 +05:30
Suraj Sudhir c9c9b68d1f [tosa] Add Torch reduction operators
- Supports variants with multiple dims, one dim, all dime
- Leverages legalize_common and legalize_utils code from
TensorFlow-TOSA work

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm.com>
2021-12-03 09:01:48 -08:00
Daniel Garvey a52aded0b9
Add lowering for slice and selectInt (#398) 2021-12-02 22:09:21 -06:00
Gaurav Shukla 73b27b32dc [MLIR][TORCH] Add E2E support for `aten.squeeze` op
This commit adds lowering of `aten.Squeeze` op into
`linalg.TensorCollapseShape` op. The size 1 dynamic dimensions are not
handled as a part of this commit.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2021-11-30 23:00:28 +05:30
Yi Zhang 5d28549c2c Add folder for torch.aten.Int.Tensor
This is to fold the common pattern from Bert inference like:
```
%111 = torch.prim.NumToTensor.Scalar %110 : !torch.int ->
    !torch.vtensor<[],si64>
%112 = torch.aten.Int.Tensor %111 : !torch.vtensor<[],si64> ->
    !torch.int
```
2021-11-30 21:55:48 +05:30
dan 03fdf56f21 add aten.add.int lowering in TorchToStd 2021-11-29 13:22:50 -05:00
Yi Zhang 0fe70994e5 Add support for multiple return values
This change is to unblock the work of some backprop ops returning more
than one tensors. We will need to think of a more scalable approach
in the future if more flexible return types combinations are needed.
2021-11-16 21:07:45 -05:00
Suraj Sudhir 628a21bb13
[mlir][tosa] Refactor conversions to use templates (#416)
- Remove use of conversion construction macros
- Add mul and div op conversions
- Add corresponding tests

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm.com>
2021-11-11 16:15:58 -08:00
Suraj Sudhir 1019ddf5a0 [tosa] Add structure for eltwise ops
Add a bunch of op legalizations.

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm.com>
2021-11-11 11:03:24 -08:00
Yi Zhang 3bd9d2a4c7 Add e2e support for aten._softmax_backward_data.
Decompose aten._softmax_backward_data into aten math ops. Also decompose
`aten.size` to facilitate decomposing _softmax_backward_data.
2021-11-09 13:09:30 +05:30
Yi Zhang 05c4dd8e39 Add convertScalarToDtype helper.
This is to facilitate scalar type conversion in the TorchToLinalg. As
part of adding the helper, this PR also:
- Updated `AtenAddTensorOp`, `AtenSubTensorOp` to use the helpers to
support more type variants.
- Added e2e type promotion testing.
- Added i32 memref return/arg type to support e2e testing.
2021-11-08 17:50:52 -05:00
George Petterson f41958037a Add NumToTensor 2021-11-08 15:56:52 -05:00
Prateek Gupta 18e8806b14 [TORCH][MLIR] Add E2E support for aten::to.dtype.
This commit adds end to end support for AtenToDtypeOp from aten
to linalg.

Signed-Off-By: Prateek Gupta <prateek@nod-labs.com>
2021-11-08 12:56:03 -05:00
Prashant Kumar fd505db2c6 Adding support for returning elemental types.
Support for returning elemental types. Previously, only
memref types as returning types was supported. All the hacky ways
to write tests which return elemental types should be taken care of.

Signed-off-by: Prashant Kumar <prashant@nod-labs.com>
2021-11-08 22:20:48 +05:30
Prashant Kumar 53b4275ef5 Add lowering of `aten.Int.Tensor` op.
The lowering of `aten.Int.Tensor` op has been added.
The changes has been made as a part of `convert-torch-to-linalg` pass.

Signed-off-by: Prashant Kumar <prashant@nod-labs.com>
2021-11-01 21:58:08 +05:30
Yi Zhang 752abc8d01 Add type promotion code to refine types.
The types have different levels of categories: where
complex > floating > integral > boolean (> means left hand
side has higher category).

The operands have different levels of priorities where:
dimensioned tensor > 0-dim tensor > scalar == wrapped 0-dim tensor.
This is represented by the `ResultTypeState.dimResult`,
`ResultTypeState.zeroResult` and `ResultTypeState..wrappedResult` in
the source code.

For operands of the same priorities, the result type should be the
highest categories with sufficient width to hold all operands.

By default, only the highest priority operands participate in the type
promotion logic. Lower priority operands participate if they are in
a higher category than any higher priority operands.

For example, <[],f32> (lower priority) and <[1], si64> tensor would
result in <[?],f32> tensor because floating > integeral. Another example
<[],f64> (lower priority) and <[1], f32> tensor would result in
<[?], f32> tensor because f32 and f64 are the same category.

The ScalarType enum definition, type promotion table, ResultTypeState
struct definition and some helpers are copied from
aten/src/ATen/native/TypeProperties.*
Other references:
- https://pytorch.org/docs/stable/tensor_attributes.html#type-promotion-doc
- https://github.com/pytorch/pytorch/issues/9515

Other minor changes:
1. Fix `visitExpandLikeOp` to consider cases where the given sizes list
size is larger than the input rank.
2. Add back the somehow deleted `torch.aten.softmax.int` tests in
decompose-complex-ops.mlir.
2021-10-29 11:17:39 -04:00
Sean Silva eb6996d557 Update llvm-project to 6f9c25167d16acff3ff8e4f54a8c14a2a175fc59
- Changes to dialect conversion that result in no-op materializations
  not being created.
2021-10-28 17:43:04 -07:00
Suraj Sudhir 7e4ef74774
[tosa] Add Torch.sigmoid fp32 to TOSA (#386)
* [tosa] Add Torch.sigmoid fp32 to TOSA

Signed-off-by: Suraj Sudhir <suraj.sudhir@arm.com>
2021-10-28 10:09:12 -07:00
Prashant Kumar 5009cbf55c Add lowering of aten.matmul op.
Lowering of `aten.matmul` op is added from torch to linalg dialect.
The different cases correspond to
https://pytorch.org/docs/stable/generated/torch.matmul.html.
TODO: Broadcasting in case of batch-matmul is yet to be taken care of.

Signed-off-by: Prashant Kumar <prashant@nod-labs.com>
2021-10-26 12:45:09 -04:00
Yi Zhang abfaf8c577 Add aten.ne.bool to make CI pass 2021-10-21 14:45:41 -04:00
Yi Zhang a459e09ab7 E2e support for aten.softmax.int and aten.embedding
- Added a DecomposeComplexOps pass to decompose complex torchOps.
- Refactored `visitAtenArgmaxOp` and `visitAtenAnyDimOp` to
`visitReductionAlongDimIntOp`.
- Moved some helper functions into
torch-mlir/Dialect/Torch/Utils/Utils.h to be shared by multiple files.
- Added support for f64 tensor as argument and return types.
2021-10-18 17:57:45 -04:00
Yi Zhang 0902438882 Update llvm-project to a54f4eae0e1d0ef5adccdcf9f6c2b518dc1101aa
This brings in https://reviews.llvm.org/D110797. PRs that are in
progress will need to use scripts provided by
https://llvm.discourse.group/t/psa-removed-arithmetic-ops-from-standard/4455.
2021-10-18 13:36:42 -04:00
Sean Silva 19e9fc4ef1 Bring some more order to the e2e error reporting situation.
- Move `run_pipeline_with_repro_report` to a more common place, and use it
  consistently
- Attach a `torch.debug_module_name` to the enclosing `builtin.module`
  op to allow for self-contained error reporting (not needing to pass
  the names around.
- Remove redundant error reporting in linalg_on_tensors_backend.py and
  tosa_backend.py (their respective backend abstract base classes now
  take care of the error reports themselves)
- Save off original value of sys.stderr, rather than always resetting to
  `sys.__stderr__`. This is just more hygienic, and allows nesting if
  desired.
2021-10-08 13:00:12 -07:00
Sean Silva 0c5c84d63d Add a basic TOSA E2E backend.
We lower through linalg-on-tensors and use RefBackend to run it.
This adds enough support for a "tanh" op. Adding more ops should be
fairly mechanical now that things are wired up. Run with:
```
./tools/torchscript_e2e_test.sh -c tosa
```

The backend structure is very similar to linalg-on-tensors based E2E
backends and is a nice parallel (see `tosa_backend.py`). Actually, this
forced a nice refactoring to the layering here. We removed
`torchscript-module-to-linalg-on-tensors-backend-pipeline` and instead
require separately running
```
torchscript-function-to-torch-backend-pipeline,torch-backend-to-linalg-on-tensors-backend-pipeline
```
This highlights the step that lowers to the "torch backend contract"
of cleaned up `torch` dialect ops is a critical step in the lowering.
Going forward, that is the key load-bearing contract of the torch-mlir
project, not the linalg-on-tensors backend contract.

Recommended review order:
- `TorchToTosa.cpp` / `TorchToTosa/basic.mlir`
- `python/torch_mlir_e2e_test/torchscript/configs/tosa_backend.py` and
  the new `utils.py` file there.
- `python/torch_mlir_e2e_test/tosa_backends/linalg_on_tensors.py` and
  `abc.py` in that directory for the TOSA backend e2e interface.
- other misc mechanical changes
2021-10-08 09:59:45 -07:00
Stephen Neuendorffer 00d42ccaee Add tool substitutions to support out-of-tree builds 2021-10-07 21:16:43 -07:00
dan 2e1498ad11 add i64 support to refbackend 2021-10-05 15:12:44 -04:00
Sean Silva 5b6902e31c Dual license the torch-mlir project.
This commit (with approval from all contributors) dual licenses
the torch-mlir project under both the standard LLVM license and the
standard PyTorch license. This will facilitate moving code between
torch-mlir and the two upstream projects.

The standard file comment is now:

```
// This file is licensed under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// Also available under a BSD-style license. See LICENSE.
```

See `LICENSE` in the project root for the terms of both licenses.
2021-10-01 10:46:08 -07:00
Sean Silva 4fad753073 Move external/torch-mlir to the root of the repo. 2021-09-27 17:11:08 -07:00
Sean Silva a99cbeeb7e Move TorchConversion dialect and TorchTo* into torch-mlir 2021-09-23 21:39:31 -07:00
Sean Silva 2213584c4f VerifyBackendContract -> VerifyLinalgOnTensorsBackendContract
This moves it into TorchConversion since it is only needed there.

This removes the Backend/ directory.
2021-09-23 21:39:31 -07:00
Yi Zhang 603e068e45 E2e implementation for `aten.cat`,`aten.gather`, `aten.bmm`
Also contains the following changes:
- Remove derefineOp canonicalizer because it's not safe.
- Support for optional tensor and list tensors in reduceOpVariant. This
only works for some special detected and easy to handle cases. For list,
it covers the case list is got from a `ListConstruct`. For optional, it
covers the case optional is constructed from a `DerefineOp`.
- Remove the `inferReturnTypes` for `FromBuiltinTensorOp` because it's
not safe to deduce types from the input. For example, a built-in tensor
of i8 could be converted to si8 or ui8. It's better to let the user
specify the return type explicitly.
2021-09-22 19:15:01 -04:00
Sean Silva 1a0b953ea7 Eliminate almost all mentions of IREE.
A few remain in examples/docs that will be naturally be updated in due
time.

This regresses the list support and the general direction of more widely
supported control flow, lists/dicts/globals that we were going for with
the TorchScript path. The idea is that we are deferring that work to
make torch-mlir a very clean standalone thing. We will reboot it,
probably using some of the tools of iree_pydm to make it simpler, and in
a more natural place (such as an iree-torch repo that depends on IREE and
torch-mlir to build a working PyTorch frontend solution for IREE -- it
was really weird that npcomp depended on IREE).
2021-09-22 16:06:38 -07:00
Sean Silva a25163fbfa Remove old RefBackend
It is superceded by the new one.
2021-09-22 15:33:28 -07:00
Sean Silva f9c48d0b89 Bring up new RefBackend.
`tools/torchscript_e2e_test.sh` is all green.

This needs a few passes I put into torch-mlir/lib/RefBackend (not to be
confused with `npcomp/lib/RefBackend`, which will soon be deleted).

For the sake of review, since this brings together a lot of things, I
split this into its own commit. I temporarily commented out some "list"
stuff that we are going to remove as part of the torch-mlir refocus.
2021-09-22 14:20:22 -07:00
Sean Silva 68fefe7e1f Remove NPCOMP_ENABLE_IREE CMake flag.
Our new dependency management solution relies:
- on the C++ side with the public iree-dialects project, which we
  include and are using as representative of some missing upstream
  ops (so we treat them "as if" they were upstream, with the hope of
  upstreaming them after some codevelopment has happened)
- on the Python side, with simple PYTHONPATH manipulation or installed
  Python packages. No CMake stuff required.
2021-09-17 09:27:49 -07:00
Sean Silva b6be96d722 [torch-mlir earthmoving (2/N)] Python code movement.
This moves the bulk of the Python code (including the Torch interop)
from `frontends/pytorch` into `torch-mlir/TorchPlugin`. This also
required reconciling a bunch of other Python-related stuff, like the
`torch` dialects.

As I did this, it was simpler to just remove all the old numpy/basicpy
stuff because we were going to delete it anyway and it was faster than
debugging an intermediate state that would only last O(days) anyway.

torch-mlir has two top-level python packages (built into the
`python_packages` directory):

- `torch_mlir_dialects`: `torch` dialect Python bindings (does not
  depend on PyTorch). This also involves building the aggregate CAPI for
  `torch-mlir`.
- `torch_mlir`: bindings to the part of the code that links against
  PyTorch (or C++ code that transitively does).

Additionally, there remain two more Python packages in npcomp (but
outside `torch-mlir`):

- `npcomp_torch`: Contains the e2e test framework and testing configs
  that plug into RefBackend and IREE.
- `npcomp_core`: Contains the low-level interfaces to RefBackend and
  IREE that `npcomp_torch` uses, along with its own
  `MLIR_PYTHON_PACKAGE_PREFIX=npcomp.` aggregation of the core MLIR
  python bindings. (all other functionality has been stripped out)

After all the basicpy/numpy deletions, the `npcomp` C++ code is now very
tiny. It basically just contains RefBackend and the `TorchConversion`
dialect/passes (e.g. `TorchToLinalg.cpp`).

Correspondingly, there are now 4 main testing targets paralleling the
Python layering (which is reflective of the deeper underlying dependency
structure)

- `check-torch-mlir`: checks the `torch-mlir` pure MLIR C++ code.
- `check-torch-mlir-plugin`: checks the code in `TorchPlugin` (e.g.
  TorchScript import)
- `check-frontends-pytorch`: Checks the little code we have in
  `frontends/pytorch` -- mainly things related to the e2e framework
  itself.
- `check-npcomp`: Checks the pure MLIR C++ code inside npcomp.

There is a target `check-npcomp-all` that runs all of them.
The `torch-mlir/build_standalone.sh` script does a standalone build of
`torch-mlir`.

The e2e tests (`tools/torchscript_e2e_test.sh`) are working too.

The update_torch_ods script now lives in
`torch-mlir/build_tools/update_torch_ods.sh` and expects a standalone
build.

This change also required a fix upstream related to cross-shlib Python
dependencies, so we also update llvm-project to
8dca953dd39c0cd8c80decbeb38753f58a4de580 to get
https://reviews.llvm.org/D109776 (no other fixes were needed for the
integrate, thankfully).

This completes most of the large source code changes. Next will be
bringing the CI/packaging/examples back to life.
2021-09-15 13:40:30 -07:00
Sean Silva 28a7738189 [torch-mlir earthmoving (1/N)] C/C++ code movement.
This creates the `external/torch-mlir` directory as an
LLVM_EXTERNAL_PROJECTS-compatible project (analogous to
`iree-dialects`) and completes movement/rename of all pure MLIR C/C++
compiler code into there. The next step will be to move all the Python
code / code that links/includes PyTorch C++ code (which currently lives
in `frontends/pytorch`) into a subdirectory here.

I call this "earthmoving" because it is mostly mechanical changes and
renames. As a quick summary (we can change this down the road easily)
- C++ `mlir::NPCOMP::Torch -> mlir::torch::Torch`
- CAPI `npcompTorchListTypeGet -> torchMlirTorchListTypeGet`
- preprocessor `#ifndef NPCOMP_ -> #ifndef TORCHMLIR_`
- CMake `NPCOMPFoo -> TorchMLIRFoo`

The goal of this is to create a standalone project creating a center of
mass for entry into the MLIR ecosystem from PyTorch, suitable in scope
for eventual inclusion/ownership in PyTorch. The idea is that
`external/torch-mlir` will some day be pulled out into its own
repository, and then npcomp will simply pull it in as a submodule.

Layering-wise, what lives in `torch-mlir` lowers code from PyTorch
(currently TorchScript, but TorchFX or pytorch/xla-style tracing are
possible extensions) down to what we have been calling the "Torch
backend contract" which is cleaned up IR (inlining, simplifcation,
conversion to value tensors, ...) entirely in the `torch` dialect. This
is the branching off point for further lowering, of which npcomp takes
one opinion (outside `torch-mlir` of course!), namely the
`TorchConversion` dialect/transforms which lower to IR suitable for IREE
and other linalg-on-tensors based lower-level compilers.

Summary of changes:
- move `{include,lib,test}/Dialect/Torch` into `torch-mlir`
- move relevant parts of CAPI into `torch-mlir`.
- leave a few things related to the `torch-mlir` Python build commented
  out, which should be resolved in a subsequent change.
2021-09-10 21:44:37 -07:00
Sean Silva a7252f9a06 Add basic support for lists.
This plumbs through a vertical slice of support for lists.

The main chunk of new code here is AnnotateABIPass which captures the
program signature at the Torch backend contract layer, right before we
start `TorchConversion`. The `TorchConversion` lowering process is lossy
w.r.t. types, so it's necessary to do this for all targets in general.
Like using `!iree.list` directly, we use IREE's ABI annotation
representation for this, although there is nothing very IREE-specific
about it (see
https://github.com/google/iree/blob/main/docs/developers/design_docs/function_abi.md)

We change `ListLiteralModule_basic` to use `!torch.int` because IREE
doesn't support f64 yet (and we don't yet have a way for users to say
that they want `!torch.float` to lower as f32).

Recommended review order:
- AnnotateABIPass and tests
- Arg marshaling in npcomp_backend.py and `iree.py`
- Updates to `list_programs.py` / `xfail_sets.py`
- Moving DeleteDeadIREEListsPass to Backend/Common, so that backends
  that don't support lists can use it. RefBackend uses that pass, for
  example.
2021-09-09 20:48:55 -07:00
Yi Zhang 73d553e168 MT model compilation minor changes
This contains the following changes:
 - Fix optional knowledge propagation. The initial knowledge should
 always be NotNone for the operations we implemented.
 - Add Folder for `prim.dtype`
2021-09-09 19:02:48 -04:00
Sean Silva ed2afe43e7 Fix TorchToIREE lowering.
We needed to resize the list, not just reserve capacity.
2021-09-03 23:57:54 +00:00
Sean Silva 1dec561cfd Update llvm-project to 830c0b9023cd0cf91955900e0d96283e7a8c3711
- builder.getSymbolRefAttr is gone.
- OpAsmOpInterface's getAsmResultNames method needs explicit override
- a bunch of churn for builtin.func needing to be made explicit (and
  sometimes implicit?)
- operation printers no longer need to print the operation name
  themselves.
- snuck in beneficial trivial addition to TmpDeleteDeadIREEListsPass to
  test a particular upstream change e2e with my local patchset.
2021-09-03 14:16:38 -07:00
Yi Zhang 3b0e5910a8 Refine types continue.
This should cover all the ops that are left in MT.
2021-09-02 14:39:28 -04:00
Sean Silva 29e1b2fe89 Delete RestrictedCanonicalizer
It doesn't work properly with the new dialect registration framework.
This was latent and only was exposed when running through npcomp-opt.
Not worth investing the brainpower to fix now.
2021-08-27 19:09:29 +00:00
Yi Zhang d6b9709fa5 Changes to refine types
- Add `!torch.optional` knowledge tracking
- Changes to improve type propagation for branches and terminators. See
examples in `refine-types-branch.mlir`
- Refator to separate handling of different ops from `visitOperation`
- Add refine types for a few new ops
2021-08-27 11:42:00 -04:00
Yi Zhang bc5eae41ca Add more folders to fold away branches
Added folders to a few binary computing ops, `TupleUnpack`,
`__contains__.str` and `__getitem__.Dict_str`.
2021-08-26 17:37:49 -04:00
Stella Laurenzo 4148f88576 Merge npcomp and mlir python namespaces.
* Now the parts of the MLIR API are directly exported under the npcomp module (i.e. `npcomp.ir`, etc).
* Has required fixes for https://reviews.llvm.org/D108489
* Deletes npcomp.tracing vs fixing it because it was a very early experiment that will not be carried forward.
* This makes the npcomp python distribution completely standalone and separate from an mlir installation.
* Makes most of npcomp itself relocatable for future use as a library.
* Most things are a namespace package now. In the future we can s/torch_mlir/npcomp.frontends.torch/ and have it layer properly.
2021-08-22 21:00:42 -07:00
Sean Silva cab8d922ec Add TorchToIREE and factor out TorchConversion dialect.
This converts a basic list op (torch.prim.ListConstruct) to the IREE
dialect.

```
    def forward(self, x: float):
            return [x, x]
```

turns into:

```
builtin.func @forward(%arg0: !torch.float) -> !torch.list<!torch.float> {
  %0 = torch.prim.ListConstruct %arg0, %arg0 : (!torch.float, !torch.float) -> !torch.list<!torch.float>
  return %0 : !torch.list<!torch.float>
}
```

which turns into:

```
builtin.func @forward(%arg0: f64) -> !iree.list<f64> {
  %c1 = constant 1 : index
  %c0 = constant 0 : index
  %c2 = constant 2 : index
  %0 = iree.list.create %c2 : !iree.list<f64>
  iree.list.set %0[%c0], %arg0 : !iree.list<f64>, f64
  iree.list.set %0[%c1], %arg0 : !iree.list<f64>, f64
  return %0 : !iree.list<f64>
}
```

As part of doing this, I realized that it was time to formalize the IR
form that we reach right before running TorchTo{Linalg,Std,...}. We now
call it the "Torch backend contract". We then lower the "Torch backend
contract" to the "npcomp backend contract", which involves the new
TorchConversion (`torch_c`) dialect, which holds ops that need to
operate on both the npcomp backend types (e.g. builtin tensors, i1, IREE
list, etc.) and the `!torch` types.

This made more sense, as I realized that if I didn't factor out
`torch_c` then the Torch dialect would have a dependency on IREE
dialect (we previously didn't notice this was an issue because we only
depended on `builtin` types), which seemed wrong to me.

Recommended review order:
- TorchToIREE.cpp / `TorchToIREE/basic.mlir`
- Look at the new structure of createTorchScriptToNpcompBackendPipeline.
  It now lives in TorchConversion/Transforms/Passes.cpp and cleanly
  calls into `Torch::createTorchScriptToTorchBackendPipeline` for the
  frontend lowering to the Torch backend contract.
- Mechanical change extracting
  `torch_c.{to,from}_{i1,i64,f64,builtin_tensor,iree_list}` into a new
  TorchConversion dialect, and a few passes specific to the lowering
  from the Torch backend contract to the npcomp backend contract.
- Minor fixes to TorchToLinalg.cpp to use unconverted operands (now that
  we convert lists as part of operand materialization, we need to use
  the original operands). Also added test for AtenMaxPool2dOp and fixed
  m_TorchConstantIntList.
- TmpDeleteDeadIREELists pass. Temporary pass for deleting dead IREE lists that
  are created as part of operand materialization for conv/max pool/avg pool ops
  in TorchToLinalg.
2021-08-16 15:01:58 -07:00
Yi Zhang 85ff8b692b Fix compilation errors from MT model
With the following changes the compilation can continue until
RefineTypes pass:

- Add operators without ODS into `torch_ods_gen.py`
- Add some new optional and list types in `TorchTypes.td`
- Add some folders for aten int type comparator ops
- Modify GlobalizeObjectGraph.cpp. For global slots that's not used,
dont check if an aliased value is stored in more than one of global
slots. This can work around a failure where the same tensor is stored
in multiple "version" slots which are not used.
2021-08-16 16:37:23 -04:00
Sean Silva a3bfd115ee Remove npcomp-iree-backend-lower-linkage pass.
This is no longer needed by IREE.
2021-08-09 15:28:02 -07:00
Yi Zhang 0342b73bf1 Add torch.aten.flatten.using_ints and aten.MaxPool2d linalg lowering
- torch.aten.flatten.using_ints to linalg lowering
- torch.aten.max_pool2d to linalg lowering
- Support torch.aten.conv2d for more flexible dilation and strides values
2021-08-04 12:00:43 -04:00
Sean Silva 496051163f Rename npcomp-run-mlir to refback-run
This better represents its limited scope. This was causing confusion --
people were feeding it higher level ops that require frontend lowering.
2021-08-03 18:24:24 -07:00
Sean Silva f168cacd6d Remove TCF and TCP.
These were legacy concepts that are now superceded by direct Torch to
linalg-on-tensors lowering. These were based on some very early thinking
related to the layering of frontends vs codegen, which is now obsolete
because:
- We expected a lot more centralization at the frontend (TCF) level. It
  turns out that frontend needs really vary a lot, and there is no grand
  unifying TCF dialect plausible. The additional layer isn't worth it.
- Linalg-on-tensors obsoletes the primary need for TCP. There are still
  a few things not representable with linalg-on-tensors, but the support
  is growing and the whole "not included in linalg-on-tensors" direction
  needs to be rethought. Our TCP dialect didn't cover any of the
  actually important things in this space (such as sort, FFT, top-k,
  etc.).

See historical [slides](https://drive.google.com/file/d/1iljcpTQ5NPaMfGpoPDFml1XkYxjK_6A4/view) / [recording](https://drive.google.com/file/d/1jSPa8TwPKUt0WuLquGc8OgSUVYJHMvWZ/view)
for more details on the origin story here.

Their presence was confusing users too
[bug](https://github.com/llvm/mlir-npcomp/issues/248).

Also,
- Trim down npcomp-run-mlir testing. It was testing TCF to TCP
  lowering for the most part. The essential stuff is retained and
  rephrased with linalg-on-tensors. (we should probably rename it
  "refback-run" or something, as it is just a way to invoke RefBackend)
- test/Python/Backend/RefJIT/simple_invoke_numpy.py is XFAIL'ed. Our
  "anti-framework" direction seems to be the likely future path.
2021-08-02 12:08:39 -07:00
Stella Laurenzo 445472c51e Build packages for npcomp-torch.
* Adds a minimal setup.py for frontends/pytorch
* Makes npcomp-core export its headers and libraries
* Adds a script to build packages.
* Adds CI step to package and smoke test.
* Will need some more tweaks and coordination prior to deploying (version locking etc).
2021-07-29 19:58:59 -07:00
Stella Laurenzo cd44a35177
Bump llvm-project to 5b2e7f50a6798fd9b9c79d9d62fdebcd9e78525b. (#260) 2021-07-29 12:26:54 -07:00
Stella Laurenzo 2dbab50444
Rework the python build to a static assembly of MLIR+NPCOMP (#251)
* Adapt to python build system updates.

* Bump llvm to 310c9496d80961188e8d8f8ad306cdf44bd7541f (includes python build updates)
* Adds refback C-API.
* Re-layers all python builds.
* Rework CI.
2021-07-27 16:10:10 -07:00
Stella Laurenzo 2ecbcbf8c7
Bump llvm-project to a085c23aa3c8f91866d7f4588d4f683407dc775d. (#250)
* Added additional *ToLLVM conversion patterns (they were disaggregated from standard).
* Misc renames.
* Spelling change on ConvNCHW op, and it now expects strides and dilations attributes.
2021-07-23 14:13:19 -07:00
Sean Silva 83b5b5456d Bump llvm-project to da289a174fc6617c7be37be2947480510fd4f02a
- Build adjustments for `.cpp.inc` dialect files.
- Renaming of `memref.dim` to `tensor.dim` for tensor case.

Minor changes:
- Renaming of `mlir::linalg::ReassociationIndices` to
  `mlir::ReassociationIndices`.
- Adjust command line option parsing in npcomp-run-mlir.
2021-07-07 13:57:29 -07:00
Sean Silva 79928cd2dd Generalize support for elementwise ops.
We plumb through e2e a fair number of interesting cases:
- unary, binary, ternary elementwise ops
- ops like `torch.aten.add.Tensor` that also take a scalar parameter
- static size-1 broadcasting

We allow the static size-1 broadcasting case, but emit a runtime error
in the case of dynamic size-1 broadcasting. This seems like a sweet spot
subset of things that can be lowered directly to linalg, while not being
overly constraining to users. This is consistent with what IREE is doing
for CHLO->Linalg lowering as well
([code](50bf7a87e4/iree/compiler/InputConversion/MHLO/BroadcastingToLinalgPatterns.cpp (L1))).

To test the static size-1 case, we added support for the
`torch.aten.unsqueeze` op and lowering for it through
`linalg.tensor_expand_shape`. This involved a generalization of
`MaximizeValueSemantics` able to handle it (the solution there also
works for `torch.aten.flatten.using_ints` which we need for ResNet
anyway)

Also, a few minor additional changes:
- Add `VerifyInvariantsBeforeBackendLowering` pass, which catches a
  large class of errors before we get to backend lowering (now that we
  are doing dialect conversion, the errors are way nicer if we just emit
  them up front rather than in the guts of a random pattern).
- Minor change to RefBackend to allow `linalg.tensor_expand_shape`.

Recommended review order:
- e2e tests in elementwise.py
- `ConvertElementwiseOp` in TorchToLinalg.cpp + elementwise.mlir test
- `ConvertAtenUnsqueezeOp` in TorchToLinalg.cpp + unsqueeze.mlir test
- RefineTypes.cpp + tests
- MaximizeValueSemantics changes + test
- VerifyInvariantsBeforeBackendLowering pass + test
2021-06-28 13:28:38 -07:00
Sean Silva 145d4ae23c Bump llvm-project to a37cf17834d39411ed1d669098b428f8374c5b45
Changes:
- Change to operand ordering of `linalg.fill`.
2021-06-23 10:03:29 -07:00
Sean Silva 90c6c64fd6 Make torch.constant.float print a little nicer.
This printing is chosen to be similar to how MLIR prints the values by
default.
2021-06-23 08:07:45 -07:00
Sean Silva 60a947b4a7 Add CastOpInterface to torch.prim.unchecked_cast.
This allows it to fold away in trivial cases.
2021-06-23 08:07:45 -07:00
Yi Zhang 45f2edfc7a Add TorchToSCF pass.
1. Add TorchToSCF pass.
2. Convert prim.If and prim.If.yield.
2021-06-23 08:06:43 -07:00
Yi Zhang 5ad144c4fe More folding for aten.gt.int, aten.ne.int and Aten__Getitem__TOp.
- Fold more for aten.gt.int, aten.ne.int and Aten__Getitem__TOp
- Some format cleaning up
2021-06-23 08:06:37 -07:00
Sean Silva 79aade33da Make MaximizeValueSemantics a bit smarter.
This adds a pattern to MaximizeValueSemantics which does a simple
abstract interpretation within a block, which handles simple cases of
`torch.overwrite_tensor`, enough to remove all the unnecessary uses of
non-value tensors in ResNet right now.

Before/after IR:
[gist](https://gist.github.com/silvasean/a3e1ef625b19dfc63579f73cd3b543b6)

Also,
- Split `torch.copy.tensor` into `torch.copy.to_tensor` and
  `torch.copy.to_vtensor` which convert between value and non-value
  semantic tensors. This is a much cleaner factorization as they have
  very separate use cases and properties (e.g. different side effects)
- Remove the various canonicalization patterns they had, which were
  confusing because they resulted in limited forms of maximizing value
  semantics throughout the pipeline. We should structure our compilation
  pipeline such that only MaximizeValueSemantics should be maximizing
  value semantics.
- Adjust pass pipeline to only run MaximizeValueSemantics once.
- Make OverwriteTensorOp `$value` always be a value tensor and
  `$overwritten` be a non-value tensor.
2021-06-22 16:48:57 -07:00
Yi Zhang e6adecac83 Convert Torch constant ops to std.constant 2021-06-18 12:22:47 -07:00
Sean Silva 78d2cc0818 Make `torch.copy.tensor` canonicalization a bit smarter.
This removes most of the trivial cases that MaximizeValueSemantics needs
to handle, making it easier to see the nontrivial cases.
2021-06-17 18:11:58 -07:00
Sean Silva 333e07a74e Add `torch.vtensor.literal` op.
This op is much better behaved than the `torch.tensor.literal` op
(which is the new name of the `torch.tensor` op). In particular
`torch.tensor.literal`:
- always has a maximally refined type.
- always has value semantics.
- can be constant folded / CSE'd.

ReduceOpVariants is changed to perform the transformation from
`torch.tensor.literal` to `torch.vtensor.literal` (which in general
involves static information casts and copies.

This new op also allowed tightening up `torch.tensor.literal` to only
accept NonValueTensorType (instead of any tensor type).

This new ".literal" name is more descriptive. It was getting too
confusing seeing an op called just `torch.tensor` (we originally called
it that because that's the name of the similar function in the Torch
Python API, but it just doesn't fit here).
2021-06-17 14:37:04 -07:00
Sean Silva 4a0eb44d17 Add a !torch.float type.
This removes the dependence of the `torch` dialect on the low-level
builtin types.
Now the `torch` dialect is a standalone layer, suitable for targeting
from higher-level Python abstractions without any premature lowering to
primitive types.
2021-06-17 09:24:18 -07:00
Sean Silva f49ebf1690 Add `!torch.int` type.
This replaces the ad-hoc use of `i64` throughout the Torch layer, and
helps to keep it crystal clear the distinction between `!torch.int`
(which is modeling the Python `int` type) and the various types that
serve as dtypes of tensors, which are a totally different type universe.

Changes:
- `!torch.int` type and C bindings.
- Change `torch.constant.int` parser to not need the `: i64` at the end.
- `m_TorchConstantInt` matcher to aid with matching constants.
- BackendTypeConversion changes for `!torch.int` -> `i64` type
  conversion.
- Refactor finalizing patterns in FinalizingBackendTypeConversionPass
  (they were getting very repetitive).
- Mechanical rewriting of `!torch.int` to `i64` in all the tests, and
  `AnyTorchIntType` to `Torch_IntType` in the `.td` files.
2021-06-17 07:28:23 -07:00
Sean Silva 224afb186e Add folders for torch.aten.gt.int / torch.aten.ne.int
This fixes a "regression" on ResNet where we weren't folding away all
the control flow. For now, our policy is to "optimize hard enough" to
make that control flow go away, because we don't yet have a way to lower
to the backend the stuff guarded by the control flow (RaiseException,
string operations, etc.).

It remains to be seen how much optimization we decide to do at this
level in the fullness of time -- the torch op set is not particularly
well-designed (at least not idiomatically for MLIR) for general
optimization. Ideally, with really good backend support for various
features, all the heavy optimization will happen at that layer on `std`
ops and `scf` control flow. But I have a suspicion we might end up
needing more optimization earlier in the pipeline.
2021-06-16 14:04:31 -07:00
Sean Silva 8860b5c55d Add `torch.prim.If`
This removes the use of `scf.if`, which required laundering back and
forth between `i1` and `!torch.bool` in the frontend. We will eventually
lower this op to `scf.if`, but this results in a cleaner IR and layering
at the frontend.
2021-06-16 14:04:31 -07:00
Sean Silva 784156a998 Add `!torch.bool` type.
This finishes removing the dependence on the basicpy dialect!

Changes:
- Add `!torch.bool` type and replace use of `!basicpy.BoolType` in
  Torch-related code.
- Rename BuiltinTensorize to BackendTypeConversion since now it handles
  bool conversions (and, when we add !torch.int and !torch.float, it
  will handle those as well), and generalize the related utilities (I
  also moved them to Torch/Transforms since they aren't really part of
  Torch/IR).
  - Add `torch.to_i1` and `torch.from_i1` ops for materializations
- [cleanup] Reorganize `torch.constant.*` ops in TorchOps.td
- Remove dependency of `torch` dialect on `basicpy` dialect and also
  `std` dialect. For `std`, we use some call related ops, but the
  `torch` dialect itself never produces them (we have passes that do
  though).

This is fairly mechanical. Recommended review order:
- New stuff in Torch/IR
- New BuiltinTypeConversion files.
- Mechnical fixups elsewhere.
2021-06-16 13:22:00 -07:00
Sean Silva 3ccf6002af Add `torch.constant.int` and `torch.constant.float`.
- This removes reliance on basicpy.numeric_constant.
- Also, add OpAsmOpInterface to the `torch.constant.none` and
  `torch.constant.str` ops.
2021-06-15 15:29:42 -07:00
Sean Silva 2e850ecb72 Add !torch.str type.
- Remove dependence on `!basicpy.BytesType`.
- Add `torch.constant.str "s"` analogous to `torch.constant.none`.
2021-06-15 10:10:59 -07:00
Sean Silva 92ee0fa98f Add `!torch.tuple<T1, T2>` type.
This further eliminates the need for the `basicpy` dependency.

This required adding `torch.prim.TupleConstruct` to replace
`basicpy.build_tuple`.
2021-06-15 08:15:22 -07:00
Sean Silva 6b2424512b Make C API files more consistent
- Make consistent with MLIR Core
  - Use `//` or `///` comments.
  - Use `bool` type for booleans
  - No duplicated comments in .cpp files
- Split types into separate files `{Basicpy,Numpy,Torch}Types.h`
- Add dialect prefix consistently to C API symbols. We have lots of
  similarly named types (e.g. "list" type in basicpy and torch).
2021-06-14 15:34:43 -07:00
Sean Silva db282fd1b4 Introduce native `!torch.none` type.
- Add `torch.constant.none` op to construct it (naming is chosen to be
  analogous to Torch's representation of a prim::Constant with
  NoneType, rather than using the "singleton" terminology of Basicpy).
2021-06-14 13:30:58 -07:00
Sean Silva 0b6516c7cc Bump llvm-project to cbd0054b9eb17ec48f0702e3828209646c8f5ebd
Changes:
- MLIR_BINDINGS_PYTHON_ENABLED -> MLIR_ENABLE_BINDINGS_PYTHON
- canonicalizer constant insertion order
- EDSC is gone now
2021-06-10 16:26:45 -07:00
Yi Zhang e0ff5248fb Add TorchList type and prim::ListConstruct #218 2021-06-10 14:31:35 -07:00
Sean Silva 370e3270ab Introduce `!torch.tensor` / `!torch.vtensor` types.
This removes our reliance on the numpy dialect and avoids our off-label
use of the builtin tnesor type for modeling unknown dtypes.  The
`!torch.vtensor` (`ValueTensorType`) type is a value-semantic tensor.
The `!torch.tensor` (`NonValueTensorType`) type is a non-value-semantic
tensor. The new types look as follows syntactically:

```
// Least-static-information, non-value-semantic tensor.
!torch.tensor
// Explicit form of least-static-information variant.
!torch.tensor<*,unk>
// Least-static-information, value-semantic tensor.
!torch.vtensor
// Explicit form of least-static-information variant.
!torch.vtensor<*,unk>
// Fixed-set of allowable element types, with first-class support for
// Torch's frontend signedness semantics.
!torch.tensor<*,si32>
// First-class support for unknown dtypes.
!torch.tensor<[?,?,?],unk>
// Standard MLIR representation of `?` for unknown dimensions.
!torch.tensor<[?,2,?,4],unk>
// Statically shaped / dtyped example.
!torch.vtensor<[1,2,3,4],f32>
```

This required fairly significant changes throughout the compiler, but
overall it is a big cleanup. We now have a much clearer layering of "the
Torch frontend lowering" vs "lowering to std + linalg + etc.".

At the C++ level, there is `ValueTensorType`, `NonValueTensorType`.
We also have a helper `BaseTensorType` (kind of like ShapedType) which
interoperates with those two.

Included changes:
- New `torch.tensor(dense<0.0> : tensor<5xf32>) : !torch.tensor` op for
  creating torch tensor literals in the frontend.
- Consistently use signedness for the types (except i1 which I didn't
  touch -- we need to sort out the situation with !basicpy.BoolType
  there anyway so will be attending to that soon)
- Frontend can annotate whether an argument to the function has value
  semantics. We currently require this, as our backend contract does not
  currently allow us to even model the non-value-semantic case. Before,
  the value-semantic assumption was randomly injected in the middle of
  the pass pipeline.
- Move ArrayToTensor (now called MaximizeValueSemantics) and
  RefinePublicReturn passes to torch dialect.
- The TorchToStd and TorchToLinalg passes are now type conversions from
  `!torch.vtensor` to `tensor` and use the dialect conversion infra.
  The overall conversion pipeline is set up following the best practices
  of the "Type Conversions the Not-So-Hard Way" talk. This required
  introducing `torch-func-builtin-tensorize` and
  `torch-finalizing-builtin-tensorize` passes analogous to the upstream
  bufferization passes with the corresponding names (mostly just
  copypasta from there).
- Misc Torch-level canonicalizations -- we now cleanly layer the
  lowering to std later in the pipeline, so we are gradually lessening
  our reliance on random std constant folding before we get to that
  point.

Recommended review order:
- New types in TorchTypes.td/TorchTypes.h/TorchDialect.cpp
- New ops in TorchOps.td / TorchOps.cpp
- Less important / more mechanical stuff
  - Frontend changes.
  - Pass changes/additions in `Torch/Transforms` and `Conversion/`
2021-06-10 10:56:48 -07:00
Sean Silva d66e8fe1f8 Get simple quantized model importing.
This is enough to import the program and get it through the compilation
pipeline. It of course fails at the VerifyBackendContract pass since
there is a lot missing, but the final IR for a simple quantized MLP is
looking pretty decent already:
[IR](https://gist.github.com/silvasean/f76bccd76e9b193d396cfb2f9a11f54d)

Main changes:
- Add support for importing torch quantized tensors, including
  `torch.per_tensor_affine.create` op and `!torch.qint8` element type.
- Add support for importing `LinearPackedParamsBase` (basically a weight
  + optional bias, but requires `torch.linear_params.create` op +
  `!torch.LinearParams` type to model it). This was less painful than I
  expected, as it has the necessary methods to opaquely unpack itself. I
  factored things so it should be easy to extend to other custom classes
  like `ConvPackedParamsBase`.
- Add minimal boilerplate for importing `quantized::*` ops, with
  `quantized::linear` being a motivating example.
- Add e2e test with simple quantized MLP (courtesy of @phoenix-meadowlark).

This is somewhat of an abuse of `!numpy.ndarray` / `tensor`, as
really the proper semantics of `!torch.qint8` dtype on a Torch tensor is
"check the quantizer object of the tensor for side data (scale/offset,
possibly per-channel) that defines the full semantics of the tensor". We
don't have any such notion of "side data" for `!numpy.ndarray` /
`tensor`, let alone anything that would have the associated behavior of
keying off the dtype to determine if the side data is present.
This will be fixed by a proper `!torch.tensor` type.
2021-05-20 11:28:20 -07:00
Sean Silva 2efda323ff Significantly restructure torch/aten import design.
This is a really major and invasive restructuring of the way we get
torch operators (`torch::jit::Operator` / `c10::OperatorHandle`) into
MLIR. Please forgive the challenging review, but due to the sheer
invasiveness, it wasn't really practical do do it in sane smaller
pieces.

This fully replaces everything that was already working on the
TorchScript path (actually, more -- we added tanh support to
TorchToLinalg in order to delete the older code paths). Additionally,
I've kept the lights on for the acap path too, including what little e2e
stuff was working before (for expediency I made a few tiny compromises
along the way that will be easy to undo when we give that path proper
attention).

Overview of the new design:
- The torch operator `somens::someunqualname.someoverloadname` is
  imported as `torch.somens.someunqualname.someoverloadname` (skip the
  last dotted part if the overload name is empty), OR, if we don't have
  such an op registered, it is imported as
  `torch.operator "somens.someunqualname.someoverloadname" (...) : ...`.
  - The addition of the "overload name" is a critical element here, as
    the `(ns,unqual,overload)` triple is unique, which solves a lot of
    problems we were having.
  - This involves having separate MLIR ops for the `trailing_` and
    `.out` variants and all the different overloads. This seemed
    necessary, because the set of overloads is so wild and varied and
    unstructured. The previous design was leaning into some underlying
    structure that just isn't there -- the default situation is
    the "random overload that we want to manage on the MLIR side",
    rather than that being an exception. E.g.  `aten::ne` (not-equal)
    has 21 overloads, only 4 of which are c10 dispatcher ops see
    [gist](https://gist.github.com/silvasean/190ba918c550c956260e21254e1b8aa1),
    and the "out" variant is really called `.Tensor_out` instead of
    `.out` as it frequently is for other ops.
  - Rationale for all being in `torch` namespace: the set of operators
    are so varied and unstructured that "dialect per namespace"
    doesn't result in anything resembling the typical MLIR dialect
    boundary expectations. We could maybe draw the boundary at
    dispatcher ops vs non-dispatcher ops, but that doesn't seem to
    really result in very much useful structure at this point in time.
  - Note: within the torch operator registry, we effectively have a
    mini-basicpy subdialect (already type-resolved), which is reasonably
    structured.
  - The existing Torch op interfaces are also removed -- now that we
    track the overload name, we can losslessly find the original
    operator.
- Instead of `ATenRecognizeKernelsPass`, we now have a
  `ReduceOpVariantsPass` that keys off certain traits (and perhaps
  eventually interfaces) to reduce variants of ops to a smaller set,
  ideally operating on immutable tensors and using surrounding ops to
  model the mutability/aliasing aspects.
  - Note: `torch.ns.unqual.overload` ops allow both immutable and
    mutable tensors (unlike the previous hard distinction in the common
    case). This is a premonition for a future change that will introduce a
    bona fide `!torch.tensor` type that will clean up a bunch of stuff.
- `TorchToLinalg` / `TorchToStd` supercede the existing
  "ATen->TCF->TCP->Linalg" path.
- The new `torch_ods_gen.py` supercedes `torch_signature_ods_gen.py`.
  It should look somewhat familiar, but the benefit of hindsight has
  allowed a lot of simplifications.

The overall trend seems to be to make the `torch` dialect a nice layer
independent of anything else. It feels like as a natural result of
various future changes we will be removing the reliance on basicpy+numpy
dialects and have a nice self-contained type system too that properly
models the TorchScript type system (including proper subtyping,
mutable/immutable tensors, optional dtype, etc.).

Recommended review order:
- Start at some of the new import IR, e.g. in
  `frontends/pytorch/test/node_import/prim.py`,
  `frontends/pytorch/test/acap_export/test_export_add3.py`, and other
  tests.
- `frontends/pytorch/python/torch_mlir_utils/codegen/torch_ods_gen.py`
  and associated generated files:
  - `include/npcomp/Dialect/Torch/IR/GeneratedAtenOps.td`
  - `include/npcomp/Dialect/Torch/IR/GeneratedPrimOps.td`
- Inspect `ReduceOpVariants.cpp` / `reduce-op-variants.mlir` and the new
  traits in `include/npcomp/Dialect/Torch/IR/TorchTraits.h`
- Various code changes in the import path in
  `frontends/pytorch/csrc/builder`. Probably most interesting is the new
  code in `torch_to_mlir_utils.cpp` that has the logic to create the
  `torch.operator` ops or `torch.ns.unqual.overload` ops.

This is the [new ResNet IR](https://gist.github.com/silvasean/5407aafb710d07612b7b5b92eabecebe),
just to be able to look at a substantial sample of IR in the new style.
2021-05-19 13:37:39 -07:00
Sean Silva 133bdf4b31 [cleanup] Add materializer for basicpy.singleton
This allows the canonicalizer to coalesce it like other constants.
2021-05-03 09:54:44 -07:00
Sean Silva 3d08c83580 Add flatten op recognition + shape refinement.
This op has complex aliasing semantics, so it is kept mutable for now.

With this, we reduce ResNet18 to a single BB with all aten operators
having rank + dtype:
https://gist.github.com/silvasean/2fcb1c6e4d4ae27461204a43ae9c5031
2021-05-03 09:54:44 -07:00
Sean Silva 122cae2ee3 Add aten::len.t, aten::size, and aten::gt.int primitive ops
Also add some canonicalizations that finally reduce ResNet down to a
single block.
2021-04-30 10:57:02 -07:00
Sean Silva ec6d06aa86 Add some more ResNet ops.
- aten::relu_, aten::max_pool2d, aten::adaptive_avg_pool2d, aten::batch_norm, aten::conv2d

No aten-to-linalg conversion for the latter ones, as they are fairly
substantial. At this point, I'm trying to get shape inference and stuff
working for them and the IR cleaned up.
2021-04-30 10:57:02 -07:00
Sean Silva 9257457d8a Add AllowsTypeRefinement trait and use it to improve RefineTypes
This trait lets us model the semantics of various aten/torch/numpy ops
that are insensitive to type refinements. This replaces
hardcoded/inconsistent checks for this property.

To show usage of this new trait, we fix up some old uses, and improve
RefineTypes to be smarter about rewriting with this trait.
2021-04-30 10:57:02 -07:00
Sean Silva 1c832604d2 Remove old aten-to-std / ATenLowering pass.
It was confusing now that we have `convert-aten-to-std`.
2021-04-30 10:57:02 -07:00
Sean Silva 55c3cc6624 Add recognition/folder/lowering for aten::__is__, aten::ne.int, and aten::dim
Interestingly, TorchScript has its own op (`torch::jit::Operator`)
registry separate from the dispatcher (it is a superset of the
dispatcher).

This is where the "prim" ops and some "aten" ops (that should probably
be renamed to "prim") live. In particular, `aten::__is__` is in that
latter category of "aten but really prim". This registry is also the
source of truth for what the TorchScript interpreter calls into when it
executes.

The bulk of the "not part of the dispatcher" ops live in
09feb5f579/torch/csrc/jit/runtime/register_prim_ops.cpp (L82)

And the registry itself lives in:
09feb5f579/torch/csrc/jit/runtime/operator.cpp (L196)

This fold further reduces the IR of ResNet by folding away some
more not-taken branches. These not-taken branches in ResNet require
first-class handling of the list type which we don't yet have on any
backend.
2021-04-30 10:57:02 -07:00
Sean Silva 7eb36b4ae7 Constant fold through basicpy.bool_cast.
This is the start of a push to getting ResNet running.

This involves throwing in the towel on an O0 pipelinie for now. See note
in the code. We keep an options struct with `optimize` flag, but it
default to true for now.
2021-04-30 10:57:02 -07:00
River Riddle 4678a7fedd Refactor RefineTypes to use the upstream ForwardDataFlowAnalysis engine
This removes the need for defining all of the custom propagation logic,
and also adds support for propagating value knowledge across branches,
through regions, and across calls.
2021-04-27 13:17:56 -07:00
Sean Silva 179105ca3e Add basic MLP's to the e2e curriculum.
These tests pass on the reference backend.

- Add aten.linear op + shape xfer function + ATen->Linalg lowering.
 - Note: this needs to be more automated, and needs to cover more cases.
 - Current not implemented caveats:
  - size-1 broadcasting for bias vector (either static-size-1 or ? case)
  - higher-rank aten.linear ops (not produced by torch.nn.Linear though)
  - type promotion (still don't even know the exact rules here)
- Add folder for torch.derefine op. Now the inliner can clean it up as
  it inlines. (call boundaries are a main place we need to insert
  torch.derefine) This is brittle -- the other important case is control
  flow which will need to be handled via an extension to
  RefineTypes.cpp (as will more robust call handling). River has an
  in-flight patch to update it to the new dataflow framework so I didn't
  want to do anything intrusive here.
    - Also adjust torch.derefine syntax to use the keyword `to` instead of
      `->`, as most type-only, cast-like ops do.
2021-04-27 12:18:54 -07:00
Sean Silva 9ba77c6e13 Add InlineGlobalSlots pass.
This inlines global slots if possible. This allows them to participate
in folding, canonicalization, shape inference, etc.

Example use cases:
- inlining weights and biases that are readonly during inference
- inlining the "training" bool to allow stuff to fold away

For training use cases (especially internal training loop), we will need
something smarter to get good performance. That would look like an "SSA
formation" which promotes the global slots to tensors in the program,
flushing them back to the slots at the minimal number of necessary
places. We might want to let backends do that transformation though.
This also interacts with shape inference (type bounds on the slots to
even lower them to backends in the first place).
2021-04-27 12:18:54 -07:00
Sean Silva b1c49ae648 Move GlobalizeObjectGraph tests to their own directory 2021-04-27 12:18:54 -07:00
Sean Silva c4123d4d4d Add npcomp-verify-backend-contract pass.
This pass verifies that a given module satisfies the contract that we
have for backends. This is phrased as an "allowlist", because we want to
keep this interface tight. Also, this gives much better diagnostics than
a backend randomly crashing or failing to compile would (though they
could still be improved).

This was especially painful because if we had
`tensor<?x!numpy.any_dtype>` slip through, at some point RefBackend
would convert it to a memref type and trip the "verify type invariants"
assertion which gives no location or anything and crashed the process,
which was very unpleasant.

We implement this with the dialect conversion framework, which works
reasonably well and was quick to put together and familiar, but is still
very "op oriented". We probably want to make this hand-rolled
eventually, especially the error reporting (the most useful kind of
error for a dialect conversion user is not necessarily the best for this
use case). Also, in production, these error will go to users, and need
to be surfaced carefully such as "the compiler needs a type annotation
on this function parameter" which in general requires some special
analysis, wordsmithing, and overall awareness of the e2e use case (such
as how much we can lean into certain source locations) to provide a
meaningful user-level diagnostic.

Also, add `inline` to the current frontend lowering pass pipeline to
allow slightly more complicated programs that otherwise would fail on
shape inference.
2021-04-20 12:00:35 -07:00
Sean Silva f5dfa02523 Add `aten.mm` to linalg lowering.
This is our first op with error semantics, and stresses the system.

There are a few design notes of special interest:
- RefineTypes.cpp's note about shape inference in the presence of code
  that dynamically produces and error, and it is provable statically.
- ATenToLinalg.cpp's notes about future automation of the ATen->linalg
  path.
- The notes in Passes.td about using low-tech `std.assert` ops instead
  of `shape.assuming`.

Note: Doesn't work on IREE yet due to the `std.assert` op (needs to be
lowered to `vm.fail` on the IREE side).
2021-04-16 12:03:31 -07:00
Sean Silva 28a0f02746 Add support for compiling through IREE.
Recommended review order:
- Changes in frontends/pytorch/examples/
- Changes in python/npcomp/compiler/pytorch/backend/
- Boilerplate for the `npcomp-iree-backend-lower-linkage` pass.

This change separates out a
`npcomp.compiler.pytorch.backend.frontend_lowering` module that does the
common lowering for all backends. The individual compiler backends
`npcomp.compiler.pytorch.backend.{refjit,iree}` now accept a loosely
defined "TCP + scalar code" IR mix that will be formalized in the
future as the interface to codegen backends.

This also required adding a small pass
`npcomp-iree-backend-lower-linkage` which adds `iree.module.export` onto
functions, and layering that into the frontend flow. The pass doesn't
require a C++-level dependency on IREE, which is nice for now. TBD how
we are going to handle lists (we hope we can get away with sneakerneting
some td files and relying on loose IR compatibility).

Running through IREE requires the ability to import `iree.compiler` and
`iree.runtime`, which can be obtained as follows:
```
python3 -m pip install iree-compiler-snapshot iree-runtime-snapshot -f https://github.com/google/iree/releases/tag/snapshot-20210406.200
PYTHONPATH="${PYTHONPATH}:${MY_IREE_BUILD}/bindings/python/"
```

This patch makes it painfully clear that we don't have any e2e testing
harness to really plug into, and also don't have a usable Python API to
our compiler stack (something usable in a jupyter notebook).
That will be addressed in subsequent commits. We've been flying by the
seat of our pants with this `examples` directory that isn't subject to
any kind of testing or real usability concerns.
2021-04-09 13:15:07 -07:00
Sean Silva 927546b3c5 Add RefinePublicReturn pass.
This pass allows shape information to be propagated to return types,
which is nontrivial and cannot be cleanly put anywhere else as it
changes the public ABI, which is a concern that we want to keep
concentrated in one place.
2021-04-07 11:06:34 -07:00
Sean Silva 1e357ae680 Add simple type refinement pass.
Currently implemented as a simple intraprocedural dataflow analysis over
a standard ShapedType lattice (hasRank, sizes, and elementType).

It currently hardcodes a few key pieces of information:
- shape transfer functions
- whether it is legal to update the operand type of an op

This needs to be made pluggable obviously and the core propagation logic
moved somewhere agnostic.
2021-04-07 11:06:34 -07:00
Sean Silva 6431b0f11f Add primitive ArrayToTensor (numpy-array-to-tensor) pass.
The current implementation is just sufficient to do a unary aten.tanh
from the e2e spike, and just applies some local rewrite patterns.  I've
sketched out the more full explanation of where this pass eventually
need to go in the pass docs.

Adding this required adding `numpy.tensor_static_info_cast`, which is
the tensor analog of `numpy.static_info_cast`. This op encapsulates the
same numpy-specific "no runtime code" casting semantics, in particular
the interpretation of `!numpy.any_dtype`. The
`numpy.tensor_static_info_cast` I see in practice now are "information
erasing" and will be removed by a later pass that exploits the fact that
aten ops are agnostic to the static info in the operand types (so
substituting a type with more static info is fine).

Side note: we *need* to do dtype and rank inference before aten->tcf
(which will eventually mostly be aten->linalg+guards), because each aten
op is idiosyncratically overloaded based on dtype and rank. Without
copying that idiosyncratic overloading into lower layers (layering
violation), we cannot really lower it to anything until we do that.
2021-04-05 17:56:35 -07:00
Sean Silva 30356c41c8 Add torch-adjust-calling-conventions pass.
This pass incorporates torch.type_bound info and also removes NoneType
returns (eventually it will rewrite tuple types too, but can't yet
because !basicpy.TupleType doesn't track element types).

Recommend looking at adjust-calling-conventions.mlir first to see what
it is doing, and holding your nose for the implementation of the pass.
I decided to implement this with the conversion framework, because it
gives us *some* goodies for type conversion -- mainly avoiding large
amounts of tricky RAUW dances. Unfortunately, the conversion framework
isn't a perfect fit for a couple reasons:
- the incorporation of torch.type_bound is a context-sensitive rewrite
  (requires looking at the arg attr, not just the type).
- NoneType conversion is 1->0, which requires some special handling
- (not implemented yet) 1->N tuple type conversions require special
  handling.
It's a little bit scary, but on balance doing it the other way would
have its own downsides.
2021-04-05 17:56:35 -07:00
Sean Silva e749074bae Basic infra for annotate shapes and dtypes on arguments.
These allow users to annotate a known "type bound" on the argument,
which can seed shape/dtype inference. We don't rewrite the function
types as part of the import process (it will happen in a
yet-to-be-written pass) because:

1. We would need to interprocedurally rewrite all calls to keep the IR
   consistent. Currently, we have a place after GlobalizeObjectGraph but
   before we convert to tensors where this is convenient to do. Ideally,
   we would do this on the object graph representation.

1. We don't necessarily know that adjusting the function type is a legal
   calling convention change. The pass will have blessed knowledge (by
   the pass pipeline author) that adjusting the argument type based on
   the type bound is safe (which it frequently is).

2. Note that in principle, a type bound could be a fairly general thing
   (such as maximum sizes of dimensions, unions of multiple concrete
   types, etc.). The pass will in principle have logic to interpret the
   type bounds and to determine a suitable "best" (and legal) argument
   type.
2021-04-01 18:40:03 -07:00
Sean Silva c6d56fed8a Add unary tanh lowering. 2021-03-30 16:39:49 -07:00
Sean Silva 99178a167d Bump llvm-project to 0524a09cc7e1a0797982feacf505825231efbee7
- renames of OwningRewritePatternList -> RewritePatternSet
  - also `insert` to `add`
- RewritePatternSet holds a context now
- memref dialect split from std
2021-03-23 14:29:05 -07:00
Bryce Arden 4591884d06 [refbackrt] Scalar arg support
* Adds f32 scalar argument support across the ABI boundary.
* Adds support for passing input type / shape information
  across the ABI boundary
* Adds support for parsing / creating input FloatAttr's in
  `npcomp-run-mlir`
2021-03-23 13:16:44 -07:00
Sean Silva 703428eff4 Add support for "trailing_" and "out" variants of various ops.
We already had the `promoteTrailingOutTensor` flag, but weren't using
it. A inplaceVariantKernelName flag needed to be added.

This change is a little dissatisfying, as the conversions done by the
RecognizeKernelsPass are currently non-orthogonal. In particular,
`kDropResultAndAliasArg0` probably won't work as intended if mixed with
these (we probably need to promote kDropResultAndAliasArg0 to not be an
arg-level thing anyway, as we have done with promoteTrailingOutTensor).

This involved adding a new op `numpy.overwrite_array`.

```
numpy.overwrite_array %arg2 overwrites %arg0 : tensor<2x3xf32>, !numpy.ndarray<[2,3]:f32>
```

This models the destructive update behavior. Note that in the above op,
we cannot simply RAUW %arg0 with a suitably conveted %arg2 (for example,
%arg0 might have uses that are not dominated by %arg2, or might have an
alias relation with some other array in the program). In general, we
need a pass analogous to "SSA-formation" which knows how to see through
these to uncover an underlying tensor program.

Also, add tanh_out_e2e.py/div_inplace_e2e.py and fix some bitrot in
refjit.py which is my running example I'm trying to get working.
2021-03-19 10:34:50 -07:00
Aaron Arthurs 4fd9b4afb5
Import ATen conv2d conversion and test (#180)
* Import ATen conv2d conversion and test

This is a first attempt at expanding ATen-to-TCF conversion for the
conv2d operator. Eventually, this will come in use when lowering a
high-level conv-based model.
2021-03-12 17:21:16 -08:00
Sean Silva 58c7030104 Support multiple instances of a class in GlobalizeObjectGraph.
This happens in practice with e.g. ResNet from torchvision (multiple
instances of the same BatchNorm class).

The key observation is that for this program, and the expected set of
programs, we can convert the program to the same globalized form with a
bit more static analysis and effort to suitably monomorphize the
program. Though what we are doing here is fairly annoying to implement,
it saves any nontrivial later pass from having to do similar analyses
(or worse). E.g. shape inference would need to be object-graph aware,
mutation/lifetime analyses would have to be aware, etc. Additionally, it
would make us front-load what it means to have a !torch.nn.Module type
on an ABI boundary, which we are just not ready to handle.

I'm really, really hoping that in practice we can get away with
this, otherwise it's going to be really rough designing a representation
(and implementing everything to back it) that is convenient to transform
and gracefully scales from full object graph (in the most dynamic case)
down to a fixed set of global slots like we have here (in the most
static case, which we presume a lot of practical programs fall into).

This also involved introducing a
`torch-prepare-for-globalize-object-graph` pass that does a minimal set of
lowerings to simplify the IR into a more orthogonal and analyzable form,
and a `torch-globalize-pipeline` helper.

Recommended review order:
- updated documentation in Passes.td
- new tests in `globalize-object-graph-multiple-instances*.mlir`
- implementation of GlobalizeObjectGraph.cpp
- PrepareForGlobalizeObjectGraph.cpp + prepare-for-globalize-object-graph.mlir
- misc stuff like torch-globalize-pipeline pipeline definition.

With this, we can import, globalize, and inline resnet18 from
torchvision:
https://gist.github.com/silvasean/821586afc19b67d9fb72030b2e0adeb8
2021-03-11 19:21:07 -08:00
Sean Silva c837dbb077 Properly import the entire torch::jit::CompilationUnit
This primarily unlocks proper handling of free functions (that is,
functions that are not methods of any torch.nn.Module).

Recommended review order:
- `ivalue_importer.cpp` + `ivalue_import/functions*.py`
- `GlobalizeObjectGraph.cpp` + test case
- misc other stuff

The `torch::jit::CompilationUnit` is basically a backing store or
"context" holding all the possible functions in the program. The
previous code was not explicitly accessing this data structure, since it
just imported the `torch::jit::Function`'s that it saw attached to
methods.

Subtly, any time a TorchScript module called into a free function, the
free function gets incorporated into the torch::jit::CompilationUnit,
but doesn't show up anywhere when dumping the module, except in the
curious pattern:

```
%5 : Function = prim::Constant[name="adaptive_avg_pool2d"]()
%6 : Tensor = prim::CallFunction(%5, %input.1, %4)
```

That is, calls are indirect calls, and are accessed via `prim::Constant`
materializing a function object. Even stranger, the `name` attribute here
doesn't really even tell the full story -- it doesn't correspond to
anything. It turns out that the c10::FunctionType itself actually holds
a pointer to the `torch::jit::Function` in the compilation unit
directly (so there is actually no indirection in prim::CallMethod,
because any two values of the same FunctionType call the same
function!). E.g. when converting the IR to bytecode, the "name" is
ignored [code link](1d6bd15790/torch/csrc/jit/runtime/interpreter.cpp (L937)).
We do import `prim::CallFunction` as a `std.call_indirect` though
because it's more braindead to do it that way (it gets canonicalized to
a direct call easily).
2021-03-01 12:08:01 -08:00
Sean Silva 79a3f639bf Give torch.global_slot an initializer region.
This is a much simpler representation than the ad-hoc initializer
function we had before. It is also less general, but given the rationale
in Passes.td it seems like the right tradeoff right now.

We can probably carry this representation for quite a while, and when we
can't, it likely means that TorchScript has fixed their object identity
bug and we probably need to just upgrade to a more general object graph
modeling (more general than GlobalizeObjectGraph).

In particular, we don't want to deal with defining and carrying around
this initializer function concept until we need it. For example, if we
want to constant-fold the global slots into uses, this is a much better
representation, and it plays better with symbol-dce (the initializer
function counts as a "use" of the symbol).

(the alternative would have been to write a pass that converts the
initializer function to this form when possible, but I realized that
lots of information had been lost which made that fairly annoying -- it
was all self-inflicted anyway, so best to just go to the source
(GlobalizeObjectGraph) before the information is lost)

Now symbol-dce works nicely (no more "training" bools)
```
pt_util ~/tmp/classifier.pt --import --exported-name forward \
| npcomp-opt -torch-globalize-object-graph -inline -symbol-dce
```
IR: https://gist.github.com/silvasean/8abe63d70d24e29d6db9170ccc8d512b
2021-02-26 16:24:19 -08:00
Sean Silva a375ccf9da Add ability to annotate TorchScript classes.
The first use case is to annotate certain program constructs as either
exported or private. In this commit we plumb it down to
GlobalizeObjectGraph which makes use of this information.

Recommended review order:
1. class_annotator.h/.cpp + `test/module_import/annotations/*`
    - New abstractions to communicate with Python code and annotate.
2. IR changes in TorchOps.td
    - Adding "private" attribute to various things.
3. ivalue_import.cpp changes
    - Module + ClassAnnotator = annotated IR
4. GlobalizeObjectGraph.cpp + tests
    - use new "private" attributes to create "private" IR.
    - also, tweak some of the op deleting mechanics, which was triggering
      some memory errors / assertions

With this, we can run the classifier through and inline it as follows:
```
frontends/pytorch/utils/pt_util.py --import --exported-name forward ~/tmp/classifier.pt \
| npcomp-opt -torch-globalize-object-graph -inline
```
IR: https://gist.github.com/silvasean/32dcad9f6270557f412094a77cecdd69
2021-02-25 11:28:34 -08:00
Sean Silva c424c24ed8 Bump llvm-project to c68d2895a1f4019b387c69d1e5eec31b0eb5e7b0
- dialect registration
- StringAttr::get: order of context arg
- math dialect
- LogicalResult nodiscard
- error message for invalid broadcast
2021-02-22 12:23:24 -08:00
Sean Silva 1b769f7841 Extend GlobalizeObjectGraph to handle torch.prim.GetAttr returning NnModuleType
This happens in practice. With this, we can globalize slots for the
non-trivial classifier layer obtained from
https://github.com/NVIDIA/NeMo/blob/main/tutorials/nlp/Joint_Intent_and_Slot_Classification.ipynb

This also adds support for tuple return types, which were needed by that
model.
2021-02-19 10:23:25 -08:00
Sean Silva 158c5c484d Implement GlobalizeObjectGraph transformation.
This required restructuring of how we model TorchScript on import. The
main difference is that now we split out a `torch.class_type` that holds
methods and declarations of the types of each slot. This is more
consistent with TorchScript (our previous representation was
"denormalized").

Recommended reading order:
1. check out the description of `torch.class_type` in `TorchOps.td` and
   look at `test/Dialect/Torch/ops.mlir` and
   `frontends/pytorch/test/module_import/` to familiarize with the new
   representation.
   - Just look at the new IR. The diff between the old names and new
     names is confusing.
2. check out `test/Dialect/Torch/globalize-object-graph*.mlir`
   and read along with the pass description in
   `include/npcomp/Dialect/Torch/Transforms/Passes.td`
3. Read the code in `GlobalizeObjectGraph.cpp` and miscellaneous changes
   in `ivalue_importer.cpp`, `TorchOps.cpp`, etc.
2021-02-18 18:18:47 -08:00
Sean Silva 7f7bf39551 Add prim::Print and fix prim::CallMethod
For now, we are treating strings as bytes.
2021-02-10 15:15:56 -08:00
Aaron J Arthurs 63ee4f268a Import basic TCP pad test 2021-01-28 12:01:35 -08:00
Aaron J Arthurs c0e14da888 Fix TensorFromElementsOp reference 2021-01-28 12:01:35 -08:00
Aaron J Arthurs fc650c9447 Import TCP pad 2021-01-28 12:01:35 -08:00
Sean Silva 689b40c7a6 Add initial TorchScript module importer
It turns out that this was easiest to structure as a general IValue
importer, since torch module are just one of the possible IValue's.

We import the IValue object graph in a braindead fashion into basicpy
ops and a new `torch.nn_module` op that is used to model the
attributes/methods of a torch::jit::Module IValue. See `Torch/ops.mlir`
for an example, and also check out the .py import tests in
`frontends/pytorch/test/module_import`.

As part of this change, a few housekeeping tasks:
- extract some helpers from graph_importer.cpp
- more helpers around the C API
- misc touchups
2021-01-28 11:55:17 -08:00
mikeurbach 0f6a65a1c5
Enable building using LLVM_EXTERNAL_PROJECTS. (#152)
This allows building NPCOMP as an external project of LLVM, similar to
how CIRCT can be built: https://github.com/llvm/circt/pull/227.

The CMake options to use this build style look like this:

```
  -DLLVM_EXTERNAL_PROJECTS=npcomp \
  -DLLVM_EXTERNAL_NPCOMP_SOURCE_DIR=/path/to/mlir-npcomp \
```
2021-01-26 11:43:43 -07:00
Sean Silva 3f4161635c Bump llvm-project to be7352c00d51f4358db3a23ed6a077f7cb48eafd
- TensorFromElementsOp -> tensor::FromElementsOp
- `cmpi "eq", ...` -> `cmpi eq, ...`. Same for `cmpf`
- syntax change for private func ops
- some changes to the python bindings
2021-01-21 11:16:55 -08:00
Sean Silva 6351474382 Bump llvm-project to bc556e5685c0f97e79fb7b3c6f15cc5062db8e36
- `let typeDesription` -> `let description`
- LLVMIntegerType -> IntegerType
2021-01-08 14:18:09 -08:00
Stella Laurenzo 3f706473fd NFC: Delete npcomp python API and switch to upstream.
* Most updates are mechanical except:
  * python/npcomp/__init__.py and python/NpcompModule.cpp: New init/registration bits to replace some automatic things being done in the old bindings. Also an annoying linkage hack that I'll need to triage next.
  * NpcompModule.cpp: New python helpers for custom types and other hard to reach items (for the new bindings).
  * PybindUtils.h: Extended type casting so that the local extension can directly exchange Mlir* C types.
  * python/npcomp/dialects/*: Build support and ODS bindings for local dialects.
  * mlir_utils.py: Defines an ImportContext to replace the old/bad "Helper" class that tracked locations, and insertion points. This has a number of methods on it that would be good candidates to think about better ways to do them upstream.
* Also hoisted a few stand-alone samples to dedicated unit tests as they covered important things.
* More cleanup can be done, but keeping this patch as mechanical as possible to stay in NFC land (this is big enough).
2021-01-08 10:46:24 -08:00
Sean Silva 97d6d04d41 Bump llvm-project to 16c6e9c58e9ae50a775945e6b407f1891f353d2f
Changes:
- linalg init tensor change (outs+init -> just outs)
- IntegerType::get and other builtin types now take the context as the
  first arg
- LLVMType::* is gone. Now LLVM Types are just regular Type's.
2021-01-05 16:12:11 -08:00
Aaron Arthurs 85898aaf10
Add TCF convolutional op with bias addition (#137) 2020-12-15 12:53:12 -08:00
Sean Silva b2077738ca Bump llvm-project to 444822d77a7fea28aa49edf24533c987efa1b2ee
Fixes:
- renames StandardTypes -> BuiltinTypes
- std.extract_element -> tensor.extract
2020-12-11 14:43:38 -08:00
Sean Silva 46aa6d0a24 [RefBackend] Fix leaks related to ABI boundaries.
Best as I can tell (e.g. from LeakSanitizer), this fixes all the leaks
except for those due to buffers created internally to the codegenned
code itself (up next I'll add the buffer deallocation pass to fix
those).

The main change is that instead of attempting to pass `refbackrt::Tensor`
to the codegenned function directly, we make all the ABI types be
UnrankedMemRef which gets passed awkwardly (but workably) as a
`{size_t rank, void *ptrToDescriptor}` on the ABI. The reason why
refbackrt::Tensor wasn't workable is that is that MLIR doesn't really
have a way to deal with the lifetime of unranked memref descriptors that
happen inside the function, which is inevitably what would happen in the
old code that would emit runtime calls to
`refbackrt.to_memref/refbackrt.from_memref` to convert back and forth to
`refbackrt::Tensor` inside the codegenned code.

So, instead of the `refbackrt.to_memref/refbackrt.from_memref` with no
real sound basis for valid lifetime management, we now have a lovely
piece of code in `refbackrt::invoke` in `Runtime.cpp` that just barely
seems to be sound. We rely on the codegenned code having these
properties, which it seems to have:

- it won't free memref descriptors or their backing buffer for arguments
  of UnrankedMemRef type.

- it will allocate a separate memref descriptor for each result
  UnrankedMemRef (which is ensured by having a separate memref_cast for
  each)

- we can sniff the `allocatedPtr`'s (i.e. the backing buffer pointers)
  to avoid double-freeing in the case of aliasing of the backing buffer
  (including backing buffers for arguments feeding into results)

- to catch the case of statically allocated data (which we need to avoid
  passing to `free`) , check if the `allocatedPtr` is (no joke) equal to
  `0xDEADBEEF`, because there is otherwise no way to distinguish
  statically allocated from malloc'ed data...  (std.global_memref lowering
  to LLVM by happenstance sets the allocatedPtr equal to `0xDEADBEEF`,
  presumably mainly as a debugging thing)

Even with all this, we *still* need to (internally to refbackrt::invoke)
make copies of all inputs/outputs! And the details of how the LLVM-level
ABI gets laid out for e.g. function arguments/returns is still super
tricky.

This really highlights how deficient memref is as the general runtime
type for our use case. It's stewing in my mind how best to improve the
situation. My general gut feeling is that IREE's abstractions for this
are "right", but I need to think more how to distill those aspects of
IREE's design in a "reference" way for RefBackend.

Some implementation notes:

- In terms of how this is implemented, this did catch a bug in our ABI
  wrapper functions in LowerToLLVM.cpp, which I had to fix (it happened to
  work before through some combination of npcomprt::Tensor being passed as
  a single pointer + probably me infinite-monkey-ing it until it worked)

- This actually removes 2 out of the 3 compiler runtime functions (the
  only one left is "abort_if". (most of the memref descriptor code moved
  from CopmilerRuntime.cpp to Runtime.cpp)

  - this also means deleting `refbackrt.from_memref` and
  `refbackrt.to_memref`
2020-11-25 13:09:58 -08:00
Stella Laurenzo 3937dd14cb Add basicpy.numeric_constant op.
* Going through TODOs on the PyTorch side, this is a big cause of them (not being able to have constants for signed/unsigned).
* Added complex while in here since we're at the phase where it is better to just have things complete than partially done.
2020-11-24 16:44:40 -08:00
Stella Laurenzo bea0af419d NFC: Prefactor some basicpy ops in advance of more type work.
* Organizes the BasicPyOps.td file by function.
* Renamed `to_boolean` -> `as_predicate_value` (trying to consistently use "predicate" to refer to i1/low-level types and Bool/Boolean to refer to Python bool types).
2020-11-24 15:49:37 -08:00
Sean Silva 358159a6eb [RefBackend] Open-code shape.get_extent as extract_element
It was annoying that we were creating shape.get_extent in the middle of
the bufferization pipeline, as it required running convert-shape-to-std
at an awkward place. To make that cleaner, just open-code the
extract_element ops that shape.get_extent expands into.

This is a little gross, but it helps with the macroscopic pipeline
ordering issues. Anyway, the train is long-gone of trying to treat
shapes as some special data type that should only be operated on with
shape ops.

Also,
- reorder tensor constant bufferize (which is a module pass) to bracket
all the bufferization function passes, to make the parallelism
opportunities there clearer. Now we have a very clean little
bufferization segment of our pipeline construction.
2020-11-17 11:00:38 -08:00
Stella Laurenzo a7ff87a922 Sever C++ level depend on IREE and rebase on exe and python interface.
* IREE doesn't have proper install support, so there is some temporary hoaky hacking in our CMakeLists.txt to shuttle some symlinks around.
* Reworked the original numpy e2e with IREE test to pipe through iree-translate.
* Removed all of the C++-level dependencies.
* Will generalize and apply to the PyTorch backend in a followup.
2020-11-16 21:32:56 -08:00
Sean Silva 5227d52c26 [RefBackend] Use std.global_memref instead of homegrown thing
This vastly simplifies our code, allowing deleting multiple ops,
simplifying multiple passes, and removing a whole pass.

Now `refback` dialect is down to one op (refback.alloc_memref, which
simplifies allocations to just take a shape instead of individual
extents).
2020-11-13 18:43:50 -08:00
Stella Laurenzo d1488c8572 Move existing npcomp.compiler -> npcomp.compiler.numpy.
* Makes room for the pytorch compiler.
* Some common things can be hoisted from the numpy side but some more consolidation needs to happen first.
2020-11-10 19:26:40 -08:00
Sean Silva 1c7c362e29 [TCP] Replace tcp.matmul with linalg.matmul.
This involved adding a `tcp.splatted` op to splat a dynamically sized
init tensor. See rationale in TCPOps.td docs.

One interesting observation is that when lowering tcf.matmul to
linalg.matmul, we need to both 1) create the error checks and 2)
calculate a shape transfer function to create the init tensors.
Previously, 2) was deferred to bufferizing tcp.matmul later. I'm not
sure if this is a conflation of concerns or not. For now, it's not a big
burden.
2020-11-10 18:58:28 -08:00
Sean Silva 0427aacb0b [TCP] Replace elementwise ops with std elementwise ops. 2020-11-10 18:58:28 -08:00
Sean Silva ceab22cf90 Bump llvm-project to 53a0d45db6d0f33dfbb724c99ce2560ae25473c2
Date:   Wed Oct 28 13:25:48 2020 -0700

- fixup for func syntax change
2020-11-10 15:22:46 -08:00
Stella Laurenzo e60dc2470e Add aten.maximum op and conversions from aten->tcf.
* Conversions are very simple, suporting mul, maximum and add (alpha=1 only).
* Example added with pass pipeline needed to run.
* Much missing off of the golden path but sufficient for such simple cases.
2020-11-04 17:20:54 -08:00
Stella Laurenzo 6c702b149f Add a number of kernels and new patterns.
* convolution, convolution_backward, _log_softmax, _log_softmax_backward_data, nll_loss_forward, nll_loss_backward, nll_loss2d_forward, nll_loss2d_backward, copy_
* Extends the recognition logic and metadata for handling inplace transformations, optional tensors, ints, lists and dropped args.
* The kernel_calls generated by test_conv_nllloss_grads.py now convert to ATen.
* The result *almost* comes out as a pure tensor program with the exception of the copy_ op, which I will do some followup work to deal with.
* More progress on #97
2020-11-04 14:36:59 -08:00
Sean Silva 57e58b9272 [RefBackend] Use upstream func-bufferize pass.
Now, the only bufferization we have left is lowering tensor constants to
memref, which will hopefully proceed soon after Rahul's new
std.global_memref lands + the lowering to LLVM IR. Then I'll port
LowerConstantTensorsToMemref to upstream and we'll be 100% upstream
bufferization, except for our local TCP dialect (which will probably go
away and be replaced by std elementwise + linalg named ops on tensors :)
).
2020-11-02 17:38:33 -08:00
Sean Silva f9c2f8eb0d [RefBackend] Use upstream SCF bufferization pass. 2020-10-30 18:12:41 -07:00
Sean Silva 0761df9f58 Bump llvm-project to 72ddd559b8aafef402091f8e192e025022e4ebef
- Fixup to OpBuilderDAG
- Update for affine map naming
2020-10-30 18:12:41 -07:00
Aaron J Arthurs 29c715b6b1 Add TCP mul test 2020-10-30 15:11:52 -07:00
Stella Laurenzo c08935a418 Rewrite ATen ODS code generator to be based on new op registry and new signature recognition system.
* Deletes prior code generator from previous attempt (moved some of it into this one).
* Renames old generated tablegen source to "Legacy".
* Generates ODS and import rules for most binary and unary arithmetic ops.
* Removes old generated ops and integration tests that were testing details of the prior setup.
2020-10-28 10:37:37 -07:00
Aaron J Arthurs 94ea6f7c92 [RefBackend] Support element-wise multiply op
Register the following for the multiply op:
- tcf.mul
- tcp.mul
- TCP->TCP lowering
- Shape transfer, broadcasted multiplicands
- Lower to standard `MulFOp` op
2020-10-27 19:41:23 -07:00
Stella Laurenzo 510f226df2 Expose signature metadata to ops and implement ATenRecognizeKernelsPass pass.
* Two op interfaces, one for querying instance metadata and one for getting static data needed to construct an op from a generic form.
* For torch.generic_kernel ops, metadata is splatted in during capture from Torch (it comes from the op registry, which will work for either device capture or graph import).
* Moved the 'add' out of the generated set so I can experiment on it. It implements the TorchBuildableKernelOpInterface interface which provides its metadata.
* The ATenRecognizeKernelsPass pass generically lowers from a torch.generic_kernel to recognized ops that implement the TorchBuildableKernelOpInterface, handling the various types of transformations that we allow at this stage.
2020-10-26 20:31:45 -07:00
Sean Silva 14470f9ff6 [RefBackend] Use upstream std bufferization.
It now subsumes the one we had.
2020-10-21 16:46:56 -07:00
Stella Laurenzo 029815152e Add remaining pieces to capture full example models.
* Adds Basicpy List, Tuple, Dict types and plumbs through C API.
* Started debugging the issues around aten::conv2d capture, but a PyTorch bug is suspected.
* Was able to manually verify that the basic conv2d forward test captures correctly with a workaround.
* Need to resolve some printing issues upstream and move these tests to an integration test target (they take ~seconds to run).
2020-10-19 22:16:59 -07:00
Sean Silva 06a8ba6900 [RefBackend] Use more idiomatic bufferize pattern for TCP.
The time has come for BypassShapes/LowerShapedResultsToMemref to go away :(
For the reference backend, being consistent with upstream conventions is
the name of the game now.

This is a step down in a number of ways, e.g. test clarity and
separation of concerns. But it is fewer files and fewer tests, and
*does* address the "TODO: This is really fragile". It also eliminates two
more ops from the refback dialect (sadly, they are the
shaped_results/yield that we were getting kind of fond of, but alas).
2020-10-15 20:15:53 -07:00
Sean Silva b6bdc8cc4f [RefBackend] Use upstream BufferizeTypeConverter
Now that it has grown source/target materialization capabilities
(spelled with ops tensor_load/tensor_to_memref), we can use it. We can
also now delete refback.memref_to_tensor/refback.tensor_to_memref.

This is also a first step to reducing the downstream functionality
needed in the refback dialect.
2020-10-15 15:58:51 -07:00
Stella Laurenzo 5c5b8db70f Update test configuration to import mlir from LLVM install location.
* Also adds two lit tests to verify that all of our extensions load without fireworks, which is a good indication that the shared library deps are sane.
* Bumps llvm-project version to use D89167.
2020-10-12 15:25:07 -07:00
Sean Silva 93fc21dad0 [RefBackend] Split out TCF->TCP conversion.
Now the reference backend is cleanly accepts "TCP"+scalar ops.

We introduce tcf-refback-lowering-pipeline which also does TCF->TCP
conversion for convenience until we have a "target interface".
2020-10-12 11:56:39 -07:00
Stella Laurenzo af4edb63ae Start reworking towards a shared library build.
* Need to have a dag of shared library deps in order to interop across python extensions (as presented in ODM).
* Introduced add_npcomp_library and friends to mirror the MLIR setup.
* Adds a libNPCOMP.so shared library.
* Redirects tools and extensions to link against libNPCOMP.so (instead of static libs).
* Moves all libraries to lib/, all binaries to bin/ and all python extensions to python/. The invariant is that the rpaths are setup to have a one level directory structure.
* Reworks the _torch_mlir extension to build like the others (still need to come up with a consolidated rule to do this instead of open coded).
* Includes an upstream version bump to pick up needed changes.

Sizes with dynamic linking (stripped, release, asserts enabled):
  libNPCOMP.so: 43M (includes much of the underlying LLVM codegen deps)
  libMLIR.so: 31M
  _npcomp.so: 1.6M (python extension)
  _torch_mlir.so: 670K (python extension)
  npcomp-capi-ir-test: 6.3K
  npcomp-opt: 351K
  npcomp-run-mlir: 461K
  mnist-playground: 530K

Still more can be done to normalize and optimize but this gets us structurally to the starting point.
2020-10-09 16:02:58 -07:00
Sean Silva 7edb5f3641 [RefBackend] Rename RefBackend dialect to Refback
I now realize that VerboseCamelCase is not the best choice for dialect
directory/file names and C++ identifiers (take e.g. "Linalg", "Basicpy",
etc. as prior art here; not LinearAlgebra or BasicPython). If I had to
name the convention it seems to be "Shortword" (or of course just
acronym dialects like LLVM, SCF, etc.).

This rename also has the side benefit of differentiating RefBackend
directories, which now refer to the actual backend itself, from
Refback/Refbackrt, which are the dialects which happen to be used by
that backend.
2020-10-08 09:07:00 -07:00
Sean Silva bf99a82832 [RefBackend] Rename Npcomprt dialect to Refbackrt. 2020-10-08 09:07:00 -07:00
Sean Silva ddc2e9de5d [RefBackend] Rename test/E2E.
I missed this in the previous commits.
2020-10-07 15:52:11 -07:00
Sean Silva 21255d5f8e [RefBackend] Rename "E2E" to RefBackend. 2020-10-07 10:29:48 -07:00
Sean Silva 5017430dc7 [RefBackend] Split out RefBackend (refback) dialect from TCP.
This is the first in a patch series that is refactoring the
constellation of things variously called or associated with "E2E",
"RefE2E", "npcomprt", and "TCP" into a more cleanly layered result.

Concretely, this first patch fixes the fact that TCP was basically
acting like a dumping ground needed by the reference backend. This
splits it out, which is fairly mechanical, but touches a lot of lines of
code (basically replacing `tcp` with `refback` and `TCP` with
`RefBackend).

Now, the RefBackend dialect is that dumping ground, which
is slighly better, as it starts allowing TCP to become a nice clean
middle layer that is not related per se to the reference backend.

The previous name RefE2E or "reference e2e flow" was super confusing.
Now that we are seeing more clearly where the "backend" distinction
lies, the [RefBackend] commit tag is born :)
2020-10-07 10:29:48 -07:00
Sean Silva 8022dfaf1a [RefE2E] Initialize the linalg matmul accumulator buffer.
I was seeing some miscompiles due to the uninitialized data read here
before. Interestingly, this was masked in some of our previous test
cases, since the uninitialized data "always" was so small that it would
present as a rounding error for the 1.0-10.0 sized values that the
matmul was computing on.
2020-10-02 16:24:52 -07:00
Stella Laurenzo e5433e314f Add capture function arguments.
* Adds at::Tensor -> MlirValue tracking.
* Adds conversions for tensor and scalar types to MLIR types.
* Adds npcomp C APIs for constructing custom types.
* Reworks pybind include so as to get Torch pybind helpers (needed to pass at::Tensor type from Python->C++).
2020-10-01 18:59:58 -07:00
Stella Laurenzo 3d74337be0 Add a torch.kernel_call op and associated predicates. 2020-09-29 15:10:38 -07:00
Stella Laurenzo 2c9ca79c89 Add boilerplate for Torch dialect. 2020-09-28 15:26:17 -07:00
Sean Silva 16c26ef57e [RefE2E] Use upstream shape constraint conversion pass.
Now that we upstreamed our pass, we can remove it.
The final pass that landed upstream doesn't do the shape.assuming
canonicalization to legalize that op away, so added a
restricted-canonicalizer pass that allowed to run just shape dialect
canonicalizations, which deletes the shape.assuming.
The pass ended up kind of ugly. See the TODO's on it for some potential
cleaner directions.
2020-09-28 09:34:44 -07:00
Sean Silva 6ea37cfed6 Bump llvm-project to 9ed1e5873c19eb817fb9e36d0262c7effee5d35e
Date:   Fri Sep 18 13:55:52 2020 -0700

- Update to linalg syntax
- New generated builders are better. Custom builder for
tcp.shaped_results is now redundant.
2020-09-28 09:34:44 -07:00
Sean Silva f9b37c55b7 [RefE2E] Add support for unary ops exp and tanh
This is fairly mechanical.
2020-09-24 18:41:30 -07:00
Sean Silva c69e9fabc5 [RefE2E] Add support for "max".
This cleans up the lowering pipeline to easily allow extending to
multiple binary ops. It looks fairly repetitive at multiple levels, but
I don't want to prematurely generalize. I think that in principle we
could derive a large swatch of TCF + TCP from a single linalg-style
specification. Another direction is to use an OpInterface (something
like "buildLinalgGenericBody"). I'm keeping my eye on it.

In a subsequent commit, I'll mechanically add a set of binary ops
modeled off of the std arithmetic ops.
2020-09-22 18:38:32 -07:00
Sean Silva 7b7f35744b [RefE2E] Add interesting control flow example.
This also required adding a lowering for ForOp in our tensor->memref
conversion.
2020-09-21 12:25:24 -07:00
Sean Silva 276f5b80ea [RefE2E] Add assemblyFormat for TCF and TCP ops and tidy up. 2020-09-18 15:03:53 -07:00
Sean Silva dc8afc9271 [RefE2E] Refactor how tcf.add is lowered.
It was previously going through this awkward route that prematurely
created linalg.generic ops, which was an annoying layering problem since
we can't compute a shape transfer function for linalg.generic in the
general case. Now we pass it through the same path as tcp.matmul, with
the shape transfer function being defined for tcp.add.

This also removed the need for TCPToLinalg (now deleted). The equivalent
of that is happening in lower-shaped-results-to-memref. One interesting
outcome of this: we're basically using linalg as a "Buffer TCP". We
might want to look into using named structured ops for more of TCP, but
that would be a big velocity hit since then any change to the ODS /
verification for those ops would be a change to the upstream structured
op ODS generator. After we have more experience defining this manually,
we should re-evaluate rebasing TCP on generated named linalg ops.
2020-09-18 15:03:53 -07:00
Sean Silva d8675f8ad2 [RefE2E] Add support for matmul.
I'm pretty happy with how this turned out. It looks pretty much like it
should -- one change at each layer. This particular op bottoms out on
linalg which takes care of the rest.

- Add tcf.matmul
- Add tcp.matmul
- Add TCF->TCP lowering
- Add tcp.matmul shape transfer function (BypassShapes.cpp)
- Add tcp.matmul -> linalg.matmul lowering (LowerShapedResultsToMemref.cpp)
- Add support to LowerShapeConstraints for lowering the new
shape.cstr_require

This matmul op is pretty limited in its capabilities. There is no
batching and no multidimensional contraction. Certainly more design work
will be needed to find the right abstractions that aren't too general
but also help to canonicalize many cases from frontends. This is mainly
to show that adding a new op needn't be very "scary" once we have the
e2e infra in place.

Also,
- this clears out some exploratory cruft from the TCF dialect now that
this is starting to become real.
2020-09-18 11:31:01 -07:00
Sean Silva 62738d3641 [RefE2E] Fix nul-termination bug.
I was seeing some of the error messages come out with some garbage at
the end. This fixes it.
2020-09-18 11:31:01 -07:00
Sean Silva 75f57b461e
Totally rework RefE2E tensor to memref flow. (#42)
This now gets the overall "RefE2E" compilation stack to a point that I'm
fairly happy with. We simplify it by mostly embracing the "descriptor"
view of the world.

The overall flow is best understood by reading through the
createE2ELoweringPipeline function in lib/E2E/E2E.cpp
That function creates a pass pipeline that lowers from "TCF" (which is
~numpy level of abstraction) down to LLVM IR.

A brief high-level summary of what happens there:

1. TCF to TCP conversion. This involves reifying error handling in the
form of shape constraints. See test/Conversion/TCFToTCP/basic.mlir

2. Lowering shape constraints. This converts shape constraints into
eager error-handling code. See test/E2E/lower-shape-constraints.mlir
This pass will soon go upstream.
Because this lowers to std.assert, some later passes like
LowerToNpcomprtABI and LowerToLLVM are updated to properly plumb this
through e2e.
See test/npcomp-run-mlir/invalid-broadcast.mlir for an execution test
that properly aborts in case of an error.

3. Lowering tensors to memrefs. This is done via a series of passes
rather than an single mega conversion. Unlike the previous code that
mixed in the npcomprt ABI stuff here, it's now a very clean "pure
memref" conversion.
See test/E2E/lower-*-to-memref.mlir and
lib/E2E/TensorToMemref/
Most of the changes are concentrated here.

4. As part of the above, we use the upstream ConvertShapeToStandard for
lowering shapes.

5. We lower linalg to loops and lower loops to CFG using upstream
passes.

6. Rewrite the "ABI" boundaries of the program to npcomprt data
structures (LowerToNpcomprtABI). This mainly affects ABI boundaries and
how global tensor constants are represented. One of the major
improvements in this commit is that now it's a very clean rewrite that
just replaces memrefs on ABI boundaries with !npcomprt.tensor (before
there was a get_extent function that is not needed).
See test/E2E/lower-to-npcomprt-abi.mlir

7. Lower to LLVM with upstream mlir patterns + some patterns for the
npcomprt lowerings.

One aspect here that is still a remnant of a non-descriptor-based tensor
to memref flow is the BypassShapes + LowerShapedResultsToMemref.
BypassShapes wraps the "tensor compute" ops in a tcp.shaped_results
(basically a "tie_shape" kind of op), and then
LowerShapedResultsToMemref uses those annotations to allocate output
buffers while lowering the "tensor compute ops". Note that there are
very few "tensor compute" ops currently supported (tcp.add +
tcp.broadcast_to), so we just hardcode them in both passes.
Realistically, I expect this to go away as we fully embrace the
descriptor-based approach for simplicity, so don't look too deep into
it.
2020-09-16 17:31:40 -07:00
Stella Laurenzo 97d83f786a Bump submodule versions.
* llvm-project: b5924a8e27536d19dd5c4d302db29fb6163d5faa
* mhlo: 848ca244d20f045b7921da55a98a04d95ef94f0e
* Multiple breakages that need to be fixed.

Fixes:
* Refactor dialect registration
* Remove all kindof methods (Casting functionality has been added upstream and is implicitly
available, see https://llvm.discourse.group/t/removing-kinds-from-attributes-and-types/1547.)
* Update dialect registration to comply with https://reviews.llvm.org/D85495.
* Remove type kinds and update some changed dialect signatures.
* Upgrade ATen dialect to match upstream needs.
  * Move dialect registration to tablegen.
  * Register the ListType in tablegen.
  * Change dialect initialization signature.
* Use TypeSwitch in MlirIr location printer.
* Remove global registry depends from npcomp-opt.
* Change LowerToLLVM to pass an MLIRContext vs an LLVMDialect for type creation.
* Remove dep on MLIREDSCInterface that is removed upstream.
* Thread through the DialectRegistry for opt and python-like tools.
* Modernize pass registration (This was forced because the GEN_PASS_REGISTRATION code now generates inline functions vs literal pass registration statements)

Co-authored-by: Marius Brehler <marius.brehler@iml.fraunhofer.de>
2020-09-08 13:26:42 -07:00
stephenneuendorffer bb668e6e26
Add ATen Dialect (#16)
This patch adds a dialect intended to be used as a frontend dialect
to facilitate lowering from "A Tensor Library" in torch/pytorch.

This patch includes several passes that are useful in conjuction with the
dialect:

--aten-layer-name: Generates layer names for each operation, which are not
  present in the original pytorch.
--aten-to-std: Lower the ATen dialect into standard dialect function calls.
--return-elimination-pass: convert functions (primarily the toplevel function)
  to pass return values by reference.  This simplifies pytorch integration.
--aten-op-report: generate a textual report about the model
--liveness-report

Future patches will implement actual integration with the pytorch jit to
intercept and generates MLIR in this dialect, then lower the resulting MLIR
into function calls through aten-layer-name -> aten-to-std ->
return-elimination -> std-to-llvm. The result would then jitted using the LLVM
jit, linked against a runtime library which makes calls back into pytorch to
implement all the layers.

Co-authored-by: Jeff Fifield <jeff.fifield@xilinx.com>

Co-authored-by: Jeff Fifield <jeff.fifield@xilinx.com>
2020-08-12 19:28:04 -07:00
stephenneuendorffer 5beaf4cc01
Fix build again (#14)
The RuntimeShlib.so now lives in /lib.
2020-08-07 08:36:03 -07:00
Stella Laurenzo 38abe99805 Collapse python_native/ into python/.
* These were separated originally for layering reasons that no longer apply.
* Most of the python extension code is under lib/ with just the module setup in python/.
2020-08-03 17:46:34 -07:00
Stella Laurenzo 571c8b448a Collapse different top level test directories into test/.
* Uses local configs and unsupported annotation to disable optional tests.
* This separation was just an artifact of having initial trouble getting lit setup.
2020-08-03 17:41:16 -07:00
Stella Laurenzo fc484d1bd8 Rework reference shape lowering based on upstream shape dialect changes.
* Primarily, the upstream shape dialect now uses tensor<?xindex> for non-erroring, immediate shape calculations (and will return this for shape_of of a tensor or memref).
* In addition, upstream passes do not yet exist for fully lowering to standard ops, so the passes here need to be extended to handle this new convention.
* This should be seen as an intermediate state, necessary to integrate a new LLVM version and needs more work and cleanup for generality.
* There is a good deal of awkwardness in these conversions. The hope is that additional upstream work will yield better defined conversion paths once out of this intermediate state.
2020-08-03 13:43:49 -07:00
Sean Silva 3f3dcad871 Make input file to npcomp-run-mlir be positional.
This makes command lines more succinct.
2020-07-13 16:02:19 -07:00
Sean Silva df0d3fcaff Consolidate LLVM definitions of runtime data structures.
This required making module descriptors hold a FuncDescriptor* instead
of a pointer to array of FuncDescriptors as it previously did, which is
innocuous (just requires an llvm.bitcast after the llvm.mlir.addressof).
2020-07-10 17:50:55 -07:00
Sean Silva e228aa4b11 npcomprt: add support for constants
- create tcp.global + tcp.get_global_memref
- create npcomprt.global + npcomprt.get_global
- LLVM lowering for new npcomprt ops
- Runtime:
 - GlobalDescriptor struct emitted by LLVM lowering
 - implement __npcomp_compiler_rt_get_global

Also,
- cleanly isolate all runtime data structure definitions shared by the
compiler and runtime into lib/runtime/CompilerDataStructures.h
2020-07-10 17:31:24 -07:00
Stella Laurenzo efbcf0aa44 Add NumpyPublicFunctionsToTensor pass.
* Rewrites public function signatures to operate on tensors (vs ndarray).
* Most of our backends presume immutable tensors at public function boundaries.
2020-07-08 22:51:54 -07:00
Stella Laurenzo 5ceb37c19b Add NumpyToTCF conversion.
* Just for numpy.add right now.
2020-07-08 21:03:57 -07:00
Sean Silva f18014f60c LowerRankedShapes: support shape.const_shape op.
Also, the previous code had a special case for deleting this op when it
had no uses. This is subsumed by the change in this commit since now
shape.const_shape is properly lowered.

With this change, the included test case with multiple serially
dependent ops works!
This specific issue was related to the scalar argument to that
function. We needed to compute a broadcast of a scalar shape (which is a
shape.const_shape) with another shape.
2020-07-08 20:12:40 -07:00
Sean Silva b4f0cea8fa Rework e2e flow to use new "npcomprt"
This ~totally reworks the existing "runtime" stuff to be more
principled and usable, such as from Python. It's still not fully
production-quality, mainly in the department of memory management (e.g.
it currently leaks memory; we need to figure out "who frees memrefs" +
the analysis and transformation needed to do that (maybe use upstream
buffer allocation pass?)).

The user API is in include/npcomp/runtime/UserAPI.h, though
include/npcomp/JITRuntime/JITModule.h is a friendlier wrapper.

The stuff under {include,lib}/runtime is totally firewalled from the
compiler and tiny (<6kB, though no attention has gone into optimizing
that size). For example, we don't link in libSupport into the runtime,
instead having our own bare bones replacements for basics like ArrayRef
(the JITRuntime helps with bridging that gap, since it *can* depend on
all common LLVM utilities).

The overall features of npcomprt is that it exposes a module that
with multiple function entry points. Each function has arguments and
results that are tensor-valued, and npcomprt::Tensor is the runtime type
that is used to interact with that (and a npcomprt::Ref<T>
reference-counting wrapper is provided to wrap npcomprt::Tensor in the
common case).

From an implementation perspective, an npcomprt module at the
LLVM/object/binary level exposes a single module descriptor struct that
has pointers to other metadata (currently just a list of function
metadata descriptors). All interactions with the npcomp runtime are
keyed off of that module descriptor, including function lookups and
dispatching. This is done to dodge platform ABI issues and also allow
enough reflection to e.g. verify provided arguments.

Most of the compiler-side work here was in LowerToNpcomprtABI and
LowerToLLVM.

Also,
- Rename npcomp_rt/NpcompRt to npcomprt/Npcomprt; it was getting
annoying to type the underscores/caps.
- misc improvements to bash_helpers.sh
2020-07-08 19:36:19 -07:00
Stella Laurenzo 5aa2f0f9f6 Add a trivial copy elision canonicalization on ndarray->tensor.
* This elides the very common code the compiler adds for chaining otherwise tensor-related numpy ops together.
* More aggressive canonicalizations would require more advanced analysis.
2020-07-05 18:09:43 -07:00
Stella Laurenzo fae15ec5e7 Allow the ndarray type to carry a shape. 2020-07-05 17:34:03 -07:00
Stella Laurenzo 046751254f Refactor old tracing tests and remove deprecated ops.
* Old doctests to run under lit.
* Old custom filecheck tests -> pytest directory (under lit).
* Rename some old ufunc ops in the tracer.
2020-06-29 16:19:03 -07:00
Stella Laurenzo b2708e4687 Add test case for !numpy.ndarray. 2020-06-28 17:41:21 -07:00
Stella Laurenzo 7bd5733d38 Add "template function" ops and importer code.
* This starts to lay down the infra for reasoning about calls
* Adds the importer code to generate IR for function calls of compiler recognized static functions.
2020-06-26 18:36:36 -07:00
Stella Laurenzo 529873d13c Wire up IREE compilation and runtime in a new backend test.
* Adds python bindings for invoking flow, HAL, and VM lowering pipelines.
* Adds pythong bindings for translating to VM module flatbuffer.
* Adds a new backend_test/iree directory and configure lit to find the IREE python rt bindings.
* Open code a simple_invoke.py that exercises the whole pipeline (need real APIs for a lot of this).
* Fails when invoking the function because I never implemented argument marshaling for scalars :(
* Plenty of stuff to do tomorrow.
2020-06-19 00:30:34 -07:00
Stella Laurenzo 308a54c3d0 Bump llvm-project to 52cae05e087b3d4fd02849fc37c387c720055ffb (2020/6/10).
* Fixes compile errors from upstream.
* XFAIL several tests that are now failing to legalize (will hand off to Sean).
2020-06-11 16:10:05 -07:00
Stella Laurenzo 8280b86c05 Aggregate all lit test targets under check-npcomp. 2020-06-07 14:35:58 -07:00
Stella Laurenzo af4466197e Add lit test suite for python compiler.
* Adds a test for simple constants and fixes issues.
2020-06-07 14:29:39 -07:00
Sean Silva 92e45703ad Remove XFAIL.
This test seems to be passing, after a clean rebuild of everything
(including MLIR).
2020-06-03 20:52:16 -07:00
Sean Silva cd7258dbd4 Enable warnings by default.
The secret here is LLVM_ENABLE_WARNINGS=ON.

I also fixed a couple warnings, which gets us to be warning-clean.

I noticed also that npcomp-run-mlir/basic.mlir seems to be failing.
Maybe something since the latest integrate. My next commit (introduce
npcomp mini runtime) will largely rewrite it though, so it'll get fixed
then.
2020-06-03 20:39:34 -07:00
Sean Silva e62e2e2915 Rename tests that run e2e-lowering-pipeline.
This makes it more clear that they are grouped together. A directory
seemed too heavyweight.
2020-06-01 19:34:44 -07:00
Sean Silva 7b9f0c3364 Add ability to run without optimizations.
The default is to only do the bare minimum needed for correctness, since
that stresses the layering of the system maximally.
2020-06-01 19:33:59 -07:00
Sean Silva e8b1a07ef4 Initial NpcompRt (npcomp_rt) dialect boilerplate. 2020-06-01 19:07:53 -07:00
Sean Silva e7b5a2b8a3 Make LowerRankedShapes clean up shape.from_extents ops.
We were previously relying on a later canonicalization pass to clean
them up, but it is a cleaner invariant if the pass gets rid of them
itself.
2020-05-29 18:00:35 -07:00
Sean Silva ccd5754b88 Rename `check-npcomp-opt` to just `check-npcomp`.
It runs npcomp-run-mlir as well now, so having `-opt` in the name is
confusing.
2020-05-29 16:12:10 -07:00
Sean Silva ea822968fa Add bare-bones npcomp-run-mlir.
The code isn't super clean, but is a useful incremental step
establishing most of the boilerplate for future enhancements.
We can't print or return tensors yet so correctness TBD, but I've
stepped into the running code in the debugger so I know it definitely is
running.

This is the first step to building out an npcomp mini-runtime. The
mini-runtime doesn't have to be fancy or complex, but it should at least
be layered nicely (which this code and the current compiler interaction
with the "runtime" code is not). Now that we have boilerplate for e2e
execution in some form, we can build that out.
2020-05-28 18:37:11 -07:00
Sean Silva 3a09455540 Use upstream shape.from_extents
Replace our local `tcp.shape_from_extents` op with the upstream
`shape.from_extents` op.
2020-05-21 14:51:01 -07:00
Sean Silva 1d3dbd9d5c Lower to LLVM dialect.
With this commit, we finish conversion to LLVM dialect, and should be
ready for subsequent commits to convert to an LLVM module and let LLVM
codegen to native machine code.

This required a custom "lower to LLVM" pass to support lowering
tcp.abort_if to a runtime call. In the future, this pass will grow to do
type conversions for our own runtime types as we add those.
2020-05-20 18:56:10 -07:00