experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
STT-tensorflow/tensorflow/compiler/xla/service/gpu
History
Deven Desai 868395d7d0 Updating XLA code to account for the device lib files location change in ROCm 3.9 
The location of the ROCm device lib files is changing in ROCm 3.9

Current (ROCm 3.8 and before) location is $ROCM_PATH/lib

```
root@ixt-rack-04:/opt/rocm-3.8.0# find . -name *.bc
./lib/oclc_isa_version_701.amdgcn.bc
./lib/ocml.amdgcn.bc
./lib/oclc_daz_opt_on.amdgcn.bc
./lib/oclc_isa_version_700.amdgcn.bc
./lib/oclc_isa_version_810.amdgcn.bc
./lib/oclc_unsafe_math_off.amdgcn.bc
./lib/oclc_wavefrontsize64_off.amdgcn.bc
./lib/oclc_isa_version_803.amdgcn.bc
./lib/oclc_isa_version_1011.amdgcn.bc
./lib/oclc_isa_version_1012.amdgcn.bc
./lib/opencl.amdgcn.bc
./lib/oclc_unsafe_math_on.amdgcn.bc
./lib/oclc_isa_version_1010.amdgcn.bc
./lib/oclc_finite_only_off.amdgcn.bc
./lib/oclc_correctly_rounded_sqrt_on.amdgcn.bc
./lib/oclc_daz_opt_off.amdgcn.bc
./lib/oclc_isa_version_802.amdgcn.bc
./lib/ockl.amdgcn.bc
./lib/oclc_isa_version_906.amdgcn.bc
./lib/oclc_isa_version_1030.amdgcn.bc
./lib/oclc_correctly_rounded_sqrt_off.amdgcn.bc
./lib/hip.amdgcn.bc
./lib/oclc_isa_version_908.amdgcn.bc
./lib/oclc_isa_version_900.amdgcn.bc
./lib/oclc_isa_version_702.amdgcn.bc
./lib/oclc_wavefrontsize64_on.amdgcn.bc
./lib/hc.amdgcn.bc
./lib/oclc_isa_version_902.amdgcn.bc
./lib/oclc_isa_version_801.amdgcn.bc
./lib/oclc_finite_only_on.amdgcn.bc
./lib/oclc_isa_version_904.amdgcn.bc
```

New (ROCm 3.9 and above) location is $ROCM_PATH/amdgcn/bitcode
```
root@ixt-hq-99:/opt/rocm-3.9.0-3703# find -name *.bc
./amdgcn/bitcode/oclc_isa_version_700.bc
./amdgcn/bitcode/ocml.bc
./amdgcn/bitcode/oclc_isa_version_1030.bc
./amdgcn/bitcode/oclc_isa_version_1010.bc
./amdgcn/bitcode/oclc_isa_version_904.bc
./amdgcn/bitcode/hip.bc
./amdgcn/bitcode/hc.bc
./amdgcn/bitcode/oclc_daz_opt_off.bc
./amdgcn/bitcode/oclc_wavefrontsize64_off.bc
./amdgcn/bitcode/oclc_wavefrontsize64_on.bc
./amdgcn/bitcode/oclc_isa_version_900.bc
./amdgcn/bitcode/oclc_isa_version_1012.bc
./amdgcn/bitcode/oclc_isa_version_702.bc
./amdgcn/bitcode/oclc_daz_opt_on.bc
./amdgcn/bitcode/oclc_unsafe_math_off.bc
./amdgcn/bitcode/ockl.bc
./amdgcn/bitcode/oclc_isa_version_803.bc
./amdgcn/bitcode/oclc_isa_version_908.bc
./amdgcn/bitcode/oclc_isa_version_802.bc
./amdgcn/bitcode/oclc_correctly_rounded_sqrt_off.bc
./amdgcn/bitcode/oclc_finite_only_on.bc
./amdgcn/bitcode/oclc_isa_version_701.bc
./amdgcn/bitcode/oclc_unsafe_math_on.bc
./amdgcn/bitcode/oclc_isa_version_902.bc
./amdgcn/bitcode/oclc_finite_only_off.bc
./amdgcn/bitcode/opencl.bc
./amdgcn/bitcode/oclc_isa_version_906.bc
./amdgcn/bitcode/oclc_isa_version_810.bc
./amdgcn/bitcode/oclc_isa_version_801.bc
./amdgcn/bitcode/oclc_correctly_rounded_sqrt_on.bc
./amdgcn/bitcode/oclc_isa_version_1011.bc
```

Also not the change in the filename(s)

This commit updates the XLA code, that has the device lib path + filename(s) hardcoded, to account for the change in location / filename
2020-12-04 02:26:58 +00:00
..
data
Replace instances of "blacklist" with "denylist" where possible. See Google Developer guidelines at https://developers.google.com/style/word-list#blacklist for more information.
2020-07-20 16:05:25 -07:00
ir
Integrate LLVM at llvm/llvm-project@a2291a58bf
2020-10-09 23:22:38 -07:00
llvm_gpu_backend
Updating XLA code to account for the device lib files location change in ROCm 3.9
2020-12-04 02:26:58 +00:00
tests
Change scatter emission to go through MLIR
2020-10-21 10:54:29 -07:00
alias_passthrough_params_test.cc
[XLA] Unify aliasing types
2020-07-07 16:07:00 -07:00
alias_passthrough_params.cc
[XLA] Unify aliasing types
2020-07-07 16:07:00 -07:00
alias_passthrough_params.h
…
amdgpu_compiler_registration.cc
…
amdgpu_compiler.cc
[XLA/GPU] Make Thunk::Initialize() happen at compile-time, not run-time. This simplifies GpuExecutable for MLIR transition.
2020-05-12 15:55:36 -07:00
amdgpu_compiler.h
[XLA/GPU] Make Thunk::Initialize() happen at compile-time, not run-time. This simplifies GpuExecutable for MLIR transition.
2020-05-12 15:55:36 -07:00
backend_configs.proto
…
buffer_allocations.cc
[XLA] Perform "copy protection" for aliasing on CPU and GPU
2020-06-24 23:29:08 -07:00
buffer_allocations.h
[XLA] Perform "copy protection" for aliasing on CPU and GPU
2020-06-24 23:29:08 -07:00
buffer_comparator_test.cc
…
buffer_comparator.cc
Merge pull request #42683 from nouiz:upstream_master_grid_size
2020-10-06 04:04:24 -07:00
buffer_comparator.h
…
BUILD
Merge pull request #43964 from trentlo:horizontal_input_fusion_again
2020-10-15 01:40:44 -07:00
cholesky_thunk.cc
[NFC] Eliminate references to HLO instr in the Convolution/Cholesky Thunk constructor.
2020-10-02 15:04:22 -07:00
cholesky_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
collective_permute_thunk.cc
[NFC] Eliminate references to HLO Inst from CollectivePermute Thunk.
2020-10-06 09:09:55 -07:00
collective_permute_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
conditional_thunk.cc
[NFC] Eliminate reference to HLO Inst from ConditionalThunk
2020-10-06 10:54:24 -07:00
conditional_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
convolution_thunk.cc
[NFC] Eliminate references to HLO instr in the Convolution/Cholesky Thunk constructor.
2020-10-02 15:04:22 -07:00
convolution_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
copy_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
copy_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
cublas_gemm_pad_for_tensor_cores_test.cc
…
cublas_gemm_pad_for_tensor_cores.cc
…
cublas_gemm_pad_for_tensor_cores.h
…
cudnn_batchnorm_rewriter.cc
Remove executable bit from several source files.
2020-04-14 04:20:59 -07:00
cudnn_batchnorm_rewriter.h
…
cudnn_batchnorm_runner.cc
[NFC] Eliminate references to HLO insts from CudnnBatchNorm Thunks.
2020-10-06 09:24:36 -07:00
cudnn_batchnorm_runner.h
[NFC] Eliminate references to HLO insts from CudnnBatchNorm Thunks.
2020-10-06 09:24:36 -07:00
cudnn_batchnorm_thunk.cc
[NFC] Eliminate references to HLO insts from CudnnBatchNorm Thunks.
2020-10-06 09:24:36 -07:00
cudnn_batchnorm_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
cudnn_fused_conv_rewriter_test.cc
…
cudnn_fused_conv_rewriter.cc
…
cudnn_fused_conv_rewriter.h
…
cudnn_pad_for_convolutions_test.cc
…
cudnn_pad_for_convolutions.cc
…
cudnn_pad_for_convolutions.h
…
cusolver_context.cc
…
cusolver_context.h
…
cusolver_rewriter.cc
…
cusolver_rewriter.h
…
custom_call_test.cc
…
custom_call_thunk.cc
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
custom_call_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
dummy_all_reduce_thunk.cc
Fix GetNcclAllReduceConfig to initialize replica count
2020-10-08 09:44:42 -07:00
elemental_ir_emitter.cc
Make LLVM code easier to read.
2020-07-27 14:22:26 -07:00
elemental_ir_emitter.h
Enable XLA:CPU fast math for min/max by default to be similar to TF's behavior.
2020-06-17 01:27:08 -07:00
fft_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
fft_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
for_thunk.cc
[NFC] Eliminate references to HLO Inst from ForThunk
2020-10-06 12:56:52 -07:00
for_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
fusion_merger_test.cc
Fix bug when evaluating maximum code duplication.
2020-10-01 00:04:57 -07:00
fusion_merger.cc
[XLA:GPU] Do not fuse into bitcast ops.
2020-09-23 01:27:46 -07:00
fusion_merger.h
…
gemm_algorithm_picker.cc
[NFC] Eliminate references to HLO instr from Gemm Thunk.
2020-10-06 09:01:52 -07:00
gemm_algorithm_picker.h
…
gemm_rewriter.cc
…
gemm_rewriter.h
…
gemm_thunk.cc
[NFC] Eliminate references to HLO instr from Gemm Thunk.
2020-10-06 09:01:52 -07:00
gemm_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
gpu_autotuning.proto
Replace instances of "blacklist" with "denylist" where possible. See Google Developer guidelines at https://developers.google.com/style/word-list#blacklist for more information.
2020-07-20 16:05:25 -07:00
gpu_compiler.cc
Scatter HLO -> LHLO conversion test
2020-10-21 14:36:46 -07:00
gpu_compiler.h
Scatter HLO -> LHLO conversion test
2020-10-21 14:36:46 -07:00
gpu_constants.cc
…
gpu_constants.h
…
gpu_conv_algorithm_picker.cc
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
gpu_conv_algorithm_picker.h
…
gpu_conv_padding_legalization_test.cc
Fix spatial dimension reference in conv padding legalization.
2020-09-18 03:46:36 -07:00
gpu_conv_padding_legalization.cc
Fix spatial dimension reference in conv padding legalization.
2020-09-18 03:46:36 -07:00
gpu_conv_padding_legalization.h
…
gpu_conv_rewriter_test.cc
Merge pull request #39659 from AyanmoI:amoitra/batch_grp_convolutions
2020-06-03 11:12:00 -07:00
gpu_conv_rewriter.cc
Use cudnn for grouped backward input convolution.
2020-06-09 04:14:23 -07:00
gpu_conv_rewriter.h
…
gpu_conv_runner.cc
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
gpu_conv_runner.h
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
gpu_copy_insertion.cc
…
gpu_copy_insertion.h
…
gpu_debug_info_manager_test.cc
…
gpu_debug_info_manager.cc
This CL optimizes C++11 range-based for loops where the variable is copied in each iteration but it would suffice to obtain it by const reference. This is only applied to loop variables of types that are expensive to copy which means they are not trivially copyable or have a non-trivial copy constructor or destructor.
2020-04-06 17:55:23 -07:00
gpu_debug_info_manager.h
This CL optimizes C++11 range-based for loops where the variable is copied in each iteration but it would suffice to obtain it by const reference. This is only applied to loop variables of types that are expensive to copy which means they are not trivially copyable or have a non-trivial copy constructor or destructor.
2020-04-06 17:55:23 -07:00
gpu_device_info.h
[XLA] Speed up. Make XLA faster by making PW kernel use the right number of block and loops.
2020-09-02 12:57:32 -07:00
gpu_executable_run_options.cc
…
gpu_executable_run_options.h
…
gpu_executable.cc
[XLA] Don't pass on_host_shape to ShapedBuffer/ScopedShapedBuffer inside XLA.
2020-10-08 12:00:12 -07:00
gpu_executable.h
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
gpu_fusible_test.cc
Improve layout related multi-output fusion heuristic.
2020-03-27 05:43:42 -07:00
gpu_fusible.cc
[XLA/GPU] Address review comments.
2020-10-12 18:00:16 -07:00
gpu_fusible.h
[XLA/GPU] Address review comments.
2020-10-12 18:00:16 -07:00
gpu_hlo_schedule_test.cc
…
gpu_hlo_schedule.cc
…
gpu_hlo_schedule.h
…
gpu_layout_assignment_test.cc
…
gpu_layout_assignment.cc
…
gpu_layout_assignment.h
…
gpu_sanitize_constant_names_test.cc
…
gpu_sanitize_constant_names.cc
…
gpu_sanitize_constant_names.h
…
gpu_scatter_expander.cc
[XLA] Expand simple scatter operations into dynamic-update-slice.
2020-08-18 13:21:28 -07:00
gpu_scatter_expander.h
[XLA] Expand simple scatter operations into dynamic-update-slice.
2020-08-18 13:21:28 -07:00
gpu_transfer_manager.cc
Clarify CPU/GPU infeed error messages.
2020-05-19 04:27:32 -07:00
gpu_transfer_manager.h
…
gpu_types.h
[XLA/GPU] Make Thunk::Initialize() happen at compile-time, not run-time. This simplifies GpuExecutable for MLIR transition.
2020-05-12 15:55:36 -07:00
hlo_algorithm_denylist_test.cc
Rename hlo_algorithm_blacklist to hlo_algorithm_denylist
2020-07-23 11:13:39 -07:00
hlo_algorithm_denylist.cc
Rename hlo_algorithm_blacklist to hlo_algorithm_denylist
2020-07-23 11:13:39 -07:00
hlo_algorithm_denylist.h
Rename hlo_algorithm_blacklist to hlo_algorithm_denylist
2020-07-23 11:13:39 -07:00
hlo_execution_profiler.cc
[NFC] Eliminate reference to HLO Inst from ConditionalThunk
2020-10-06 10:54:24 -07:00
hlo_execution_profiler.h
[NFC] Eliminate reference to HLO Inst from ConditionalThunk
2020-10-06 10:54:24 -07:00
hlo_to_ir_bindings.cc
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
hlo_to_ir_bindings.h
Roll forward XLA GPU LHLO sort op migration
2020-08-13 12:00:17 -07:00
horizontal_input_fusion_test.cc
Merge pull request #43964 from trentlo:horizontal_input_fusion_again
2020-10-15 01:40:44 -07:00
horizontal_input_fusion.cc
Merge pull request #43964 from trentlo:horizontal_input_fusion_again
2020-10-15 01:40:44 -07:00
horizontal_input_fusion.h
Implement horizontal input fusion.
2020-10-12 17:59:55 -07:00
horizontal_loop_fusion_test.cc
Merge pull request #43316 from trentlo:rename_h_loop_fusion
2020-09-22 01:53:30 -07:00
horizontal_loop_fusion.cc
Implement horizontal input fusion.
2020-10-12 17:59:55 -07:00
horizontal_loop_fusion.h
Rename horizontal_fusion to horizontal_loop_fusion.
2020-09-16 17:58:22 -07:00
infeed_manager.cc
…
infeed_manager.h
…
infeed_thunk.cc
[NFC] Eliminate references to HLO instr from Infeed/Outfeed thunks.
2020-10-06 09:33:54 -07:00
infeed_thunk.h
[NFC] Eliminate references to HLO instr from Infeed/Outfeed thunks.
2020-10-06 09:33:54 -07:00
instruction_fusion_test.cc
The original motivation seems to have been that inside fusions we don't have
2020-03-24 06:02:39 -07:00
instruction_fusion.cc
Evaluate the maximum code duplication in fusion nodes instead of average.
2020-09-02 07:36:23 -07:00
instruction_fusion.h
…
ir_emission_utils.cc
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
ir_emission_utils.h
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
ir_emitter_context.h
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
ir_emitter_nested.cc
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
ir_emitter_nested.h
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
ir_emitter_unnested.cc
Change scatter emission to go through MLIR
2020-10-21 10:54:29 -07:00
ir_emitter_unnested.h
Change scatter emission to go through MLIR
2020-10-21 10:54:29 -07:00
ir_emitter.cc
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
ir_emitter.h
[XLA/GPU] Emit constants computation by computation. This helps with incrementally transitioning to MLIR.
2020-09-14 19:16:24 -07:00
kernel_mapping_scheme.h
Merge pull request #39327 from nouiz:upstream_copy_bug
2020-05-13 12:09:48 -07:00
kernel_thunk.cc
XLA Parallel reduce.
2020-08-27 15:21:33 -07:00
kernel_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
launch_dimensions.cc
Merge pull request #42683 from nouiz:upstream_master_grid_size
2020-10-06 04:04:24 -07:00
launch_dimensions.h
Merge pull request #42683 from nouiz:upstream_master_grid_size
2020-10-06 04:04:24 -07:00
memset_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
memset_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
multi_output_fusion_test.cc
[XLA:GPU] Do not create multi-output fusions that would cause exponential compile behavior (time and size).
2020-10-14 08:34:47 -07:00
multi_output_fusion.cc
[XLA:GPU] Do not create multi-output fusions that would cause exponential compile behavior (time and size).
2020-10-14 08:34:47 -07:00
multi_output_fusion.h
Remove dependence from service/gpu:multi_output_fusion to service:multi_output_fusion
2020-05-25 02:50:35 -07:00
nccl_all_reduce_thunk.cc
Fix GetNcclAllReduceConfig to initialize replica count
2020-10-08 09:44:42 -07:00
nccl_all_reduce_thunk.h
Fix GetNcclAllReduceConfig to initialize replica count
2020-10-08 09:44:42 -07:00
nvptx_compiler_registration.cc
…
nvptx_compiler.cc
When ptxas do not know about an SM, fallback to the driver.
2020-10-15 08:10:23 -07:00
nvptx_compiler.h
When ptxas do not know about an SM, fallback to the driver.
2020-10-15 08:10:23 -07:00
outfeed_manager.cc
…
outfeed_manager.h
…
outfeed_thunk.cc
[NFC] Eliminate references to HLO instr from Infeed/Outfeed thunks.
2020-10-06 09:33:54 -07:00
outfeed_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
parallel_loop_emitter.cc
Merge pull request #42683 from nouiz:upstream_master_grid_size
2020-10-06 04:04:24 -07:00
parallel_loop_emitter.h
Merge pull request #42683 from nouiz:upstream_master_grid_size
2020-10-06 04:04:24 -07:00
reduction_degenerate_dim_remover.cc
…
reduction_degenerate_dim_remover.h
…
reduction_dimension_grouper.cc
…
reduction_dimension_grouper.h
…
reduction_layout_normalizer.cc
Add op metadata to bitcasts inserted by ReductionLayoutNormalizer.
2020-06-16 10:36:57 -07:00
reduction_layout_normalizer.h
…
reduction_splitter_test.cc
[XLA:GPU] Split reduce ops with large but non-consecutive reduction dimensions.
2020-06-19 10:35:29 -07:00
reduction_splitter.cc
[XLA:GPU] Split reduce ops with large but non-consecutive reduction dimensions.
2020-06-19 10:35:29 -07:00
reduction_splitter.h
[XLA:GPU] Split reduce ops with large but non-consecutive reduction dimensions.
2020-06-19 10:35:29 -07:00
replica_id_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
replica_id_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
sequential_thunk.cc
[XLA/GPU] Remove Thunk::hlo_instruction().
2020-07-13 15:32:19 -07:00
sequential_thunk.h
[XLA/GPU] Remove Thunk::hlo_instruction().
2020-07-13 15:32:19 -07:00
stream_assignment_test.cc
…
stream_assignment.cc
[XLA:GPU] Add missing synchronization when multistreaming is enabled.
2020-03-07 11:34:17 -08:00
stream_assignment.h
[XLA:GPU] Add missing synchronization when multistreaming is enabled.
2020-03-07 11:34:17 -08:00
stream_executor_util.cc
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
stream_executor_util.h
Eliminate reference to HloCustomCallInstruction from ConvolutionThunk
2020-10-02 09:54:56 -07:00
target_constants.h
…
target_util.cc
[XLA:GPU] Use device specific tanh functions for f64 inputs to improve accuracy.
2020-05-26 03:39:44 -07:00
target_util.h
[XLA:GPU] Use device specific tanh functions for f64 inputs to improve accuracy.
2020-05-26 03:39:44 -07:00
thunk_emitter.cc
[NFC] Eliminate HLO Instruction pointer from ThunkInfo
2020-10-07 09:20:05 -07:00
thunk_emitter.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
thunk_schedule.cc
[XLA/GPU] Make Thunk::hlo_instruction() non-public.
2020-07-09 14:24:57 -07:00
thunk_schedule.h
[XLA/GPU] Remove Thunk::hlo_instruction().
2020-07-13 15:32:19 -07:00
thunk.cc
…
thunk.h
[NFC] Eliminate HLO Instruction pointer from ThunkInfo
2020-10-07 09:20:05 -07:00
tree_reduction_rewriter.cc
…
tree_reduction_rewriter.h
…
triangular_solve_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
triangular_solve_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
tuple_thunk.cc
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
tuple_thunk.h
[XLA/GPU] Remove uses of Thunk::hlo_instruction() for profiling.
2020-07-10 15:31:01 -07:00
variadic_op_splitter_test.cc
…
variadic_op_splitter.cc
…
variadic_op_splitter.h
…
while_thunk.cc
[NFC] Eliminate references to HLO Inst from WhileThunk
2020-10-06 13:55:38 -07:00
while_thunk.h
[NFC] Mark Thunk subclass members const when possible
2020-10-07 10:13:20 -07:00
while_transformer_test.cc
…
xfeed_queue.h
…