colin
/
mlir-hlo
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
mlir-hlo/lib/Dialect/mhlo/transforms
History
Jacques Pienaar 95ba03534f Allow variadic operands/result in MHLO while 
This just adds support for it in the op, but keeps the production/uses as is (e.g., single tensor or tuple) matching what XLA export requires. In follow up here, would be to add pass for export to retuple and then the canonical form could be changed. Tuple'ing given control flow via regions & multi-result operations does not add representational power and all the get_tuple_element ops obscure the computation.

The old form allowed single tensor or tuple. The new variadic number of tensor or tuples as tuples may be nested, so the input could have (Tensor<..>, Tuple<Tensor<...>, Tuple<...>, ...>, Tensor<...>) and HLO_Tensor doesn't allow Tuples.

PiperOrigin-RevId: 378934388
 		2021-06-11 13:08:28 -07:00
..
CMakeLists.txt [MLIR][HLO] Rename `move-up-dynamic-broadcasts-for-fusion` to `broadcast-propagation` 2021-06-08 01:51:10 -07:00
broadcast_propagation.cc [MLIR][HLO] Extend broadcast propagation pass to enable more fusion 2021-06-09 03:03:37 -07:00
chlo_legalize_to_hlo.cc Fix Cosh approximation for F16. 2021-06-10 06:16:44 -07:00
chlo_legalize_to_hlo_pass.cc Separate CHLO transforms for expanding compositions and lowering broadcasts. 2021-05-18 13:33:59 -07:00
chlo_legalize_to_hlo_patterns.td Fix Cosh approximation for F16. 2021-06-10 06:16:44 -07:00
hlo_legalize_to_lhlo.cc PR #50100: [MLIR][DISC] Bufferize DynamicIotaOp and DynamicPadOp 2021-06-10 14:20:45 -07:00
legalize_control_flow.cc Allow variadic operands/result in MHLO while 2021-06-11 13:08:28 -07:00
legalize_gather_to_torch_index_select.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
legalize_tensor_load_op.cc PR #49598: [MLIR][DISC] legalize tensor_load inserted during hlo-to-lhlo conversion 2021-06-01 16:27:56 -07:00
legalize_to_linalg.cc [MHLO:linalg] Be more aggressive about turning mhlo.const into std.constant 2021-06-07 11:58:23 -07:00
legalize_to_standard.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
legalize_to_standard_patterns.td Lowering for mhlo.ceil to std.ceil 2020-08-12 16:15:35 -07:00
legalize_trigonometric_to_approximation.cc Fix tanh lowering for NaN input. 2021-03-24 06:34:36 -07:00
lhlo_fuse_linalg.cc Update lhlo to use the new structured op interface. 2021-06-07 03:11:03 -07:00
lhlo_legalize_to_affine.cc PR #47315: [MLIR] Add concatenateOp lowering from lmhlo to Affine. 2021-04-14 11:06:38 -07:00
lhlo_legalize_to_gpu.cc [MLIR][NFC] Rename ReduceOp operands() => inputs(). 2021-04-14 12:08:23 -07:00
lhlo_legalize_to_parallel_loops.cc [MLIR][NFC] Rename ReduceOp operands() => inputs(). 2021-04-14 12:08:23 -07:00
lower_complex.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
lower_complex_patterns.td Generate Equal and NotEqual kernels for complex types. 2021-04-15 00:35:52 -07:00
lower_general_dot.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
materialize_broadcasts.cc More cleanup in mlir-hlo to prepare for the standalone build 2020-08-03 19:28:00 -07:00
materialize_broadcasts_pass.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
mhlo_control_flow_to_scf.cc Allow variadic operands/result in MHLO while 2021-06-11 13:08:28 -07:00
mhlo_fusion.cc Integrate LLVM at llvm/llvm-project@37e1458128 2021-04-22 22:57:08 -07:00
optimize_mhlo.cc More cleanup in mlir-hlo to prepare for the standalone build 2020-08-03 19:28:00 -07:00
optimize_mhlo_pass.cc Integrate LLVM at llvm/llvm-project@b24436ac96 2021-03-23 12:20:17 -07:00
rank_specialization.cc [MLIR][HLO] Find shape equivalences and use them for better rank specialization 2021-06-11 04:00:26 -07:00
sink_constants_to_control_flow.cc Enable fallback legalization for MaxPoolGradGrad and MaxPool3DGradGrad ops 2020-11-13 13:49:05 -08:00
test_infer_shaped_type_pass.cc [mlir][hlo] Avoid dyn_cast_or_null when called with getDefiningOp result (NFC) 2021-05-27 00:20:42 -07:00
transform_unranked_hlo.cc Replace !any_of with none_of 2021-05-04 01:55:36 -07:00
unfuse_batch_norm.cc Integrate LLVM at llvm/llvm-project@678241795c 2021-03-16 13:33:00 -07:00
unfuse_batch_norm_pass.cc mlir-hlo-opt: set preloadDialectsInContext to false. 2021-03-30 01:07:14 -07:00