mlir-hlo

e422ce5a9a Integrate LLVM at llvm/llvm-project@8086f9d87e A. Unique TensorFlower 2021-05-27 18:18:18 -0700
402b74ed7f Fix type bug in mhlo.dynamic-update-slice lowering. Hanhan Wang 2021-05-27 17:52:38 -0700
b5f444232f Integrate LLVM at llvm/llvm-project@d47dd11071 A. Unique TensorFlower 2021-05-27 16:04:58 -0700
6ea8f1ddf8 Integrate LLVM at llvm/llvm-project@eba69b59d1 Benjamin Kramer 2021-05-27 08:35:16 -0700
e319f87857 Integrate LLVM at llvm/llvm-project@c467585682 Benjamin Kramer 2021-05-27 06:23:16 -0700
5deed55dc3 Integrate LLVM at llvm/llvm-project@d82f2a123f Benjamin Kramer 2021-05-27 04:24:22 -0700
b32f885ad7 [MLIR][KernelGen] Enable cluster rank specialization A. Unique TensorFlower 2021-05-27 02:43:43 -0700
a4fa6afa07 [mlir][hlo] Avoid dyn_cast_or_null when called with getDefiningOp result (NFC) Adrian Kuegel 2021-05-27 00:18:35 -0700
d939a156d8 Integrate LLVM at llvm/llvm-project@5852582532 A. Unique TensorFlower 2021-05-26 16:59:38 -0700
28c411606f Add support for lowering mhlo.dynamic-update-slice ops to Linalg and std ops. Hanhan Wang 2021-05-26 15:30:09 -0700
1c7415ba0b Integrate LLVM at llvm/llvm-project@c5c1ec7945 A. Unique TensorFlower 2021-05-26 13:18:43 -0700
af01c08ce6 Allow tuple as results of mhlo.custom_call op (NFC) Mehdi Amini 2021-05-26 12:55:03 -0700
26a0053d7d Remove linalg.indexed_generic from mhlo lowerings to linalg Robert Suderman 2021-05-26 12:23:11 -0700
ca09dabf1a Integrate LLVM at llvm/llvm-project@ea91a8cbab Benjamin Kramer 2021-05-26 11:13:57 -0700
4ebcebf31c [MLIR][HLO] Exploit scalar properties in rank specialization lowering A. Unique TensorFlower 2021-05-26 09:58:25 -0700
2f8f3d692c Integrate LLVM at llvm/llvm-project@01120fe5b3 Benjamin Kramer 2021-05-26 07:21:01 -0700
edf5ec8084 Integrate LLVM at llvm/llvm-project@cb65419b1a Benjamin Kramer 2021-05-26 04:46:41 -0700
4b3e9373e6 Integrate LLVM at llvm/llvm-project@bc56620b8b A. Unique TensorFlower 2021-05-25 17:13:32 -0700
1439ec7cae Integrate LLVM at llvm/llvm-project@0bbb502daa A. Unique TensorFlower 2021-05-25 13:11:08 -0700
cb46298a07 [MLIR][HLO] Support all smaller ranks in rank specialization cases A. Unique TensorFlower 2021-05-25 08:37:51 -0700
c5af02fd8d Integrate LLVM at llvm/llvm-project@f3c577ed38 Benjamin Kramer 2021-05-25 07:52:10 -0700
a847109ac7 Support complex types when converting HLO multiply op. Adrian Kuegel 2021-05-25 04:34:24 -0700
5816920258 Support complex types when converting HLO divide op. Adrian Kuegel 2021-05-25 03:42:25 -0700
8e28008e38 Integrate LLVM at llvm/llvm-project@6359842bc0 Benjamin Kramer 2021-05-25 03:29:03 -0700
758ae7da6b Support complex types when converting HLO compare op (EQ/NE). Adrian Kuegel 2021-05-25 01:53:00 -0700
5504f82f11 Integrate LLVM at llvm/llvm-project@c2f819af73 A. Unique TensorFlower 2021-05-24 17:51:22 -0700
af6092ef13 Integrate LLVM at llvm/llvm-project@d65c32fb41 A. Unique TensorFlower 2021-05-24 12:56:31 -0700
b93e54d8a4 PR #49454: [MLIR][DISC] Upgrade to use the new `reifyReturnTypeShapes` interface. wyzhao 2021-05-24 10:10:21 -0700
28c4112f35 Downgrade some emitErrors in patterns to notifyMatchFailure. Stella Laurenzo 2021-05-21 14:08:22 -0700
1ba4c714c9 Add support for lowering mhlo.scatter ops to Linalg. Hanhan Wang 2021-05-21 12:16:14 -0700
f5dc73ab60 Integrate LLVM at llvm/llvm-project@eb6429d0fb A. Unique TensorFlower 2021-05-21 06:29:21 -0700
1f996ab796 Integrate LLVM at llvm/llvm-project@b9076d119a A. Unique TensorFlower 2021-05-21 04:07:37 -0700
97e6103933 [MLIR][HLO] Reshape to scalars in rank specialization A. Unique TensorFlower 2021-05-21 03:11:11 -0700
c3147d76d5 Integrate LLVM at llvm/llvm-project@e3cf7c88c4 A. Unique TensorFlower 2021-05-21 02:09:27 -0700
3daf65578a [MLIR][HLO] Add scalar cases for binary rank specialization A. Unique TensorFlower 2021-05-21 01:34:45 -0700
a7884196f5 PR #49228: [MLIR][DISC] porting dynamic shape related OPs to mhlo and lmhlo dialect Feiwen 2021-05-20 23:15:58 -0700
cd8f585cf7 [MHLO:Linalg] Add support for lowering torch_index_select of unsigned tensors Hanhan Wang 2021-05-20 17:02:05 -0700
a1445aa0fa Integrate LLVM at llvm/llvm-project@cb82e8ea33 A. Unique TensorFlower 2021-05-20 16:49:28 -0700
41f663ce47 [HLO] Adopt custom syntax for convolution dimensions and window attributes (HLO) Rahul Joshi 2021-05-20 12:12:57 -0700
fc88cf1ff4 [HLO] Adopt custom syntax for convolution dims and window attributes for LMHLO_GPU Rahul Joshi 2021-05-20 09:40:28 -0700
57aeb5ab16 Integrate LLVM at llvm/llvm-project@0316f3e649 A. Unique TensorFlower 2021-05-20 06:08:26 -0700
c62fd89663 [MLIR][HLO] Add equal shapes case to rank specialization A. Unique TensorFlower 2021-05-19 05:37:49 -0700
71394fb301 Properly handle if DynamicBroadcastInDimOp shape is not of index type. Stella Laurenzo 2021-05-18 21:11:13 -0700
0fe07e3814 Separate CHLO transforms for expanding compositions and lowering broadcasts. Stella Laurenzo 2021-05-18 13:32:58 -0700
e0d9e9bffd Integrate LLVM at llvm/llvm-project@6381664580 A. Unique TensorFlower 2021-05-18 09:29:18 -0700
1aa6eb24b0 Integrate LLVM at llvm/llvm-project@560b709abe A. Unique TensorFlower 2021-05-18 08:10:25 -0700
dedc129c8c Integrate LLVM at llvm/llvm-project@f8dbd61074 A. Unique TensorFlower 2021-05-18 04:56:43 -0700
623b630fee Integrate LLVM at llvm/llvm-project@3bc2b97b34 A. Unique TensorFlower 2021-05-18 03:16:45 -0700
6af3d2df91 [MLIR][HLO] Add rank specialization with multiple non-scalar operands A. Unique TensorFlower 2021-05-18 03:01:20 -0700
0168484eed Integrate LLVM at llvm/llvm-project@e2e1a78abc A. Unique TensorFlower 2021-05-18 00:37:35 -0700
577ab8bb81 Integrate LLVM at llvm/llvm-project@ab5ff154ab A. Unique TensorFlower 2021-05-17 12:00:07 -0700
3f4e8bc658 Integrate LLVM at llvm/llvm-project@8854b27b19 A. Unique TensorFlower 2021-05-17 09:54:39 -0700
474e419729 [MLIR][HLO] Generalize rank specialization with single operand A. Unique TensorFlower 2021-05-17 08:11:59 -0700
b86b18489c Integrate LLVM at llvm/llvm-project@50511df32e A. Unique TensorFlower 2021-05-17 08:07:12 -0700
c514c73390 [MLIR][HLO] Extend rank specialization clustering pass A. Unique TensorFlower 2021-05-17 07:30:45 -0700
b82bbf4dd1 Integrate LLVM at llvm/llvm-project@7c16f93c44 A. Unique TensorFlower 2021-05-17 06:03:59 -0700
ccd70d5717 [MLIR][HLO] Add `rank-specialization-to-scf` pass A. Unique TensorFlower 2021-05-17 03:55:32 -0700
295ef229d6 Integrate LLVM at llvm/llvm-project@a11117a7c0 A. Unique TensorFlower 2021-05-17 02:32:23 -0700
804d1b50d6 Integrate LLVM at llvm/llvm-project@88a8965a7d A. Unique TensorFlower 2021-05-16 17:23:41 -0700
b06baae910 Fixing nondeterminism in pattern application. The ReduceRegion* patterns are matching on the same ops as the PointwiseToLinalg* patterns and on certain toolchains (MSVC) the order can be wrong. If the pointwise runs first then it converts the op *within* the reduction before the reduction one runs, leading to nested linalg op weirdness. Ben Vanik 2021-05-14 12:56:46 -0700
a361253e4f [HLO] Add custom print/parse for window attributes of convolutions (in LMHLO) Rahul Joshi 2021-05-14 09:46:42 -0700
e4caaaf921 Integrate LLVM at llvm/llvm-project@e51ef7f070 Adrian Kuegel 2021-05-14 06:11:01 -0700
76341f3720 [MLIR][HLO] Add mixed test for `rank-specialization-cluster` pass A. Unique TensorFlower 2021-05-14 04:38:10 -0700
9248f0a182 Integrate LLVM at llvm/llvm-project@c82a0ae70e Adrian Kuegel 2021-05-14 02:34:08 -0700
b1b6e81913 Integrate LLVM at llvm/llvm-project@7c57a9bd7d A. Unique TensorFlower 2021-05-14 00:55:07 -0700
d2cc74317c Implement constant folding for mhlo.Sign. A. Unique TensorFlower 2021-05-13 03:53:14 -0700
d764806c1e [MHLO:Linalg] Add support for lowering reshape of unsigned tensors Hanhan Wang 2021-05-12 15:13:20 -0700
a2c9b3c9d7 [mlir] update gentbl() Bazel macro Alex Zinenko 2021-05-12 10:58:02 -0700
420c42a0a1 [MLIR][HLO] Support CHLO unary operations in rank specialization clustering A. Unique TensorFlower 2021-05-12 10:20:03 -0700
596918a6f1 [MLIR][HLO] Allow rank specialization clustering with `chlo.broadcast_select` op A. Unique TensorFlower 2021-05-12 08:55:45 -0700
e260aa771c [HLO] Add custom print/parse for convolution dimension numbers (in LMHLO) Rahul Joshi 2021-05-12 08:51:40 -0700
30779f0c2f Added build rules for HLO and LHLO ops MD. Itai Zukerman 2021-05-12 08:22:41 -0700
8895262330 Integrate LLVM at llvm/llvm-project@b7a11274f9 Adrian Kuegel 2021-05-12 06:03:01 -0700
875803e5e1 [MLIR][HLO] Add more tests for `rank-specialization-cluster` pass A. Unique TensorFlower 2021-05-12 04:45:06 -0700
b95162f182 Integrate LLVM at llvm/llvm-project@34c098b780 Adrian Kuegel 2021-05-12 03:46:37 -0700
313d24bc8f [MLIR][HLO] Add `rank-specialization-cluster` pass A. Unique TensorFlower 2021-05-12 03:45:09 -0700
7f84779868 Integrate LLVM at llvm/llvm-project@6351993da7 Adrian Kuegel 2021-05-12 02:33:23 -0700
5ec03ba8a5 Integrate LLVM at llvm/llvm-project@23596fece0 A. Unique TensorFlower 2021-05-12 00:53:50 -0700
b783b3272e Integrate LLVM at llvm/llvm-project@3f8be15f29 A. Unique TensorFlower 2021-05-11 22:53:27 -0700
3ccf3938fa Integrate LLVM at llvm/llvm-project@fec2945998 A. Unique TensorFlower 2021-05-11 18:35:02 -0700
2ea9470515 Remove BASE_HLO_ConvOp to remove coupling between MHLO and LMHLO conv ops Jacques Pienaar 2021-05-11 11:53:40 -0700
a4db6c57aa Removed all (most) BASE_HLO_* ops. Itai Zukerman 2021-05-11 09:47:40 -0700
7f7a86ad0d [MLIR][HLO] Implement `RegionBranchOpInterface` for rank specialization cluster A. Unique TensorFlower 2021-05-11 09:01:41 -0700
1432db02e7 Integrate LLVM at llvm/llvm-project@df47368d40 Adrian Kuegel 2021-05-11 06:22:29 -0700
96a47345cc [MLIR][HLO] Add `rank_specialization_cluster` op to CHLO A. Unique TensorFlower 2021-05-11 05:17:01 -0700
86b7eb434c [MHLO] Don't crash trying to constant fold mhlo.convert on complex Benjamin Kramer 2021-05-11 05:14:54 -0700
3cce7017fc Integrate LLVM at llvm/llvm-project@88a48999d2 Adrian Kuegel 2021-05-11 03:57:42 -0700
6191b3e528 [MLIR][HLO] Use `#` operator for list concatenation in C/HLO definitions A. Unique TensorFlower 2021-05-11 02:52:35 -0700
920f261262 Integrate LLVM at llvm/llvm-project@888307ee62 Adrian Kuegel 2021-05-11 01:17:02 -0700
1de55d2620 Integrate LLVM at llvm/llvm-project@1e11616a07 A. Unique TensorFlower 2021-05-10 23:32:40 -0700
7f86dd9f7e Constant fold compare EQ if one of the operands is true and compare NE if one of the operands is false. A. Unique TensorFlower 2021-05-10 18:53:02 -0700
2d88f2f601 Integrate LLVM at llvm/llvm-project@bf812ea484 A. Unique TensorFlower 2021-05-10 18:31:48 -0700
2af1796194 Integrate LLVM at llvm/llvm-project@5c7b43aa82 A. Unique TensorFlower 2021-05-10 15:45:29 -0700
2a4c63d949 Integrate LLVM at llvm/llvm-project@9ba661f912 Adrian Kuegel 2021-05-10 03:55:58 -0700
9d4c194c32 Integrate LLVM at llvm/llvm-project@5b158093e2 Adrian Kuegel 2021-05-10 00:14:53 -0700
39195ba15c Integrate LLVM at llvm/llvm-project@808bc11d9e A. Unique TensorFlower 2021-05-08 10:36:05 -0700
807d306229 Integrate LLVM at llvm/llvm-project@793b4b2603 Stephan Herhut 2021-05-07 06:38:53 -0700
2fe6189703 Integrate LLVM at llvm/llvm-project@911a541620 Stephan Herhut 2021-05-07 03:38:38 -0700
2d7278ee64 Integrate LLVM at llvm/llvm-project@44ee974e2f Stephan Herhut 2021-05-07 01:06:19 -0700
ce4c76314e [NFC] Remove all_gather_dimension from all-to-all in the unit test Rahul Joshi 2021-05-06 18:13:46 -0700

Commit Graph Select branches Hide Pull Requests master Mono Color

Commit Graph

Select branches

Hide Pull Requests

master