Jongsok Choi 78ff7c86be [pallas backend] Fix scalar store shape mismatch (#171581) ...

When storing a scalar value into a buffer, use jnp.full to handle shape differences.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/171581
Approved by: https://github.com/oulgen
ghstack dependencies: #171571, #171579

2025-12-31 19:10:10 +00:00

..

cpp

…

extension_backends

Revert "[inductor] Use custom triton kernel subclass when available (#167456)"

2025-11-23 00:52:42 +00:00

pallas_expected_failures

[pallas backend] Fix scalar store shape mismatch (#171581)

2025-12-31 19:10:10 +00:00

pallas_skip_tests

[pallas backend] skip flaky test test_constant_pad_2d_strides_nonpositive_cpu_pallas (#171564)

2025-12-31 04:56:20 +00:00

__init__.py

…

CMakeLists.txt

…

custom_inductor_config.py

…

custom_ops.cpp

…

indirect_assert_helper.py

…

minifier_smoke.py

…

opinfo_harness.py

…

s429861_repro.py

…

test_alignment.py

…

test_analysis.py

[ROCm] unskip some ROCm inductor UTs (#169564)

2025-12-05 06:01:56 +00:00

test_aot_inductor_arrayref.py

[AOTI] Fix a small buffer mutation issue (#169347)

2025-12-04 05:14:39 +00:00

test_aot_inductor_custom_ops.py

…

test_aot_inductor_package.py

Skip test_compile_after_package_static on CUDA 12.6 (#169185)

2025-12-19 17:21:21 +00:00

test_aot_inductor_utils.py

[AOTI] Fix a mixed-device bug for scatter_add (#167341)

2025-11-10 16:59:44 +00:00

test_aot_inductor.py

Delete pt2e flow from pytorch/pytorch since it's migrated to torchao (#169151)

2025-12-23 20:41:04 +00:00

test_aoti_cross_compile_windows.py

…

test_async_compile.py

[ROCm][inductor] heuristic improvements for reduction kernels (#170931)

2025-12-20 18:48:49 +00:00

test_augmented_graph_helper.py

Add missing super().setUp() (#167163)

2025-11-06 17:55:23 +00:00

test_auto_chunker.py

[autochunker] override num-chunks (#171477)

2025-12-31 01:51:38 +00:00

test_auto_functionalize.py

Revert "deprecate check_is_size and guard_size_oblivious (#167198)"

2025-11-17 22:16:37 +00:00

test_autoheuristic.py

…

test_b2b_gemm.py

…

test_benchmark_fusion.py

…

test_benchmarking.py

[Inductor] Propagate benchmarking env variables for all Autotuning Types (#167867)

2025-12-12 02:15:07 +00:00

test_best_config.py

…

test_binary_folding.py

…

test_block_analysis.py

…

test_cache.py

…

test_caching.py

[inductor][caching] respect fresh cache/force disable caches (#170943)

2025-12-24 16:02:33 +00:00

test_ck_backend.py

Revert "[ROCm] Enable StaticCudaLauncher for ROCm (#166492)"

2025-12-04 20:17:50 +00:00

test_codecache.py

[xpu][static_launcher] 1/4: Add config.use_static_triton_launcher as alias of use_static_cuda_launcher. (#168951)

2025-12-29 09:10:50 +00:00

test_codegen_triton.py

…

test_collective_autotuning.py

Resolve collective autotuning test failure on arm (#168919)

2025-12-04 07:56:35 +00:00

test_combo_kernels.py

[xpu][fix] Fix UT test_fuse_mix_order_reductions_combo_kernels (#170297)

2025-12-31 01:07:25 +00:00

test_compile_subprocess.py

[inductor] layout constraint for weight-norm-bwd (#167667)

2025-11-14 23:59:24 +00:00

test_compile_worker.py

inductor: compile_worker - Fix potential race condition with quiesce waitcounters (#167025)

2025-11-07 20:49:34 +00:00

test_compile.py

[Inductor][2/2] Decouple flags for optimization and debug symbols (#167575)

2025-11-13 00:59:15 +00:00

test_compiled_autograd.py

[inductor] Fix cudagraph skip for index_put_ with boolean indices, graph partitioning logic (#170103)

2025-12-18 22:43:04 +00:00

test_compiled_optimizers.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_config.py

…

test_control_deps.py

Filter control deps while calculating lifetime (#169496)

2025-12-08 23:26:44 +00:00

test_control_flow.py

[inductor] Fix torch.cond stride mismatch when subgraph outputs are padded (#169963)

2025-12-12 19:17:42 +00:00

test_cooperative_reductions.py

Fix cooperative reduction deadlocks due to too many CTAs (#170162)

2025-12-12 17:13:01 +00:00

test_coordinate_descent_tuner.py

…

test_cpp_wrapper_hipify.py

…

test_cpu_cpp_wrapper.py

Delete pt2e flow from pytorch/pytorch since it's migrated to torchao (#169151)

2025-12-23 20:41:04 +00:00

test_cpu_repro.py

Delete pt2e flow from pytorch/pytorch since it's migrated to torchao (#169151)

2025-12-23 20:41:04 +00:00

test_cpu_select_algorithm.py

Delete pt2e flow from pytorch/pytorch since it's migrated to torchao (#169151)

2025-12-23 20:41:04 +00:00

test_cuda_repro.py

[ROCm][CI] skip on MI350 test_effn_attn_bias_padding_misaligned (#170919)

2025-12-22 17:30:00 +00:00

test_cudacodecache.py

…

test_cudagraph_trees_expandable_segments.py

…

test_cudagraph_trees.py

[ROCm] Fix and enable several Triton autotuning, CUDAGraph trees, and layout optimization tests (#169901)

2025-12-19 22:09:37 +00:00

test_custom_lowering.py

[xpu][test] reduce xpu disabled cases (1/N) (#167786)

2025-12-09 06:47:25 +00:00

test_custom_op_autotune.py

[inductor] Add range-based autotuning for custom ops (#167617)

2025-12-18 17:59:52 +00:00

test_custom_partitioner_fn.py

…

test_custom_post_grad_passes.py

…

test_cutedsl_grouped_mm.py

Add public grouped_mm (#168298)

2025-11-21 00:58:48 +00:00

test_cutedsl_template.py

…

test_cutlass_backend.py

Revert "[Inductor XPU GEMM] Step 1/N: Refactor cutlass configuration. (#160174)"

2025-12-12 00:13:49 +00:00

test_cutlass_evt.py

Revert "[Inductor XPU GEMM] Step 2/N: Move out cutlass files from torch/_inductor/codegen/cuda (#160685)"

2025-12-09 02:11:29 +00:00

test_debug_trace.py

…

test_decompose_mem_bound_mm.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_dependencies.py

…

test_deterministic.py

[inductor] skip the r2r determ test in fbcode (#169074)

2025-12-04 18:50:44 +00:00

test_device_assert.py

[ROCm] unskip some ROCm inductor UTs (#169564)

2025-12-05 06:01:56 +00:00

test_distributed_patterns.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_efficient_conv_bn_eval.py

Revert "[Inductor] Fix assertion failure in efficient_conv_bn_eval for default args (#169126)"

2025-12-23 03:10:47 +00:00

test_extension_backend.py

…

test_external_callables.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_flex_attention.py

Revert "[flex_attention] adds support for low precision K/V inputs in compiled mode with GPU (#170486)"

2025-12-29 17:44:32 +00:00

test_flex_decoding.py

Revert "[flex_attention] adds support for low precision K/V inputs in compiled mode with GPU (#170486)"

2025-12-29 17:44:32 +00:00

test_flex_flash.py

[FlexFlash] dynamic shapes (#171465)

2025-12-29 23:51:21 +00:00

test_foreach.py

Revert "[inductor] Fix unbacked floats being downcast to fp32 in Triton kernels (#170289)"

2025-12-19 00:26:35 +00:00

test_fp8.py

[Inductor][Triton][FP8] Support (BlockWise128x128, BlockWise1x128) scaling in Inductor (#170748)

2025-12-22 17:21:21 +00:00

test_fused_attention.py

[ROCm] Enable SDPA pattern rewriter tests for Pattern 16 and 18 (#170093)

2025-12-22 02:59:53 +00:00

test_fusion_regions.py

use fusion regions in overlapping (#170560)

2025-12-18 16:20:31 +00:00

test_fuzzer.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_fx_fusion.py

[12/N] Apply ruff UP035 rule (#166929)

2025-11-05 22:06:19 +00:00

test_fxir_backend.py

[Inductor] Add more tests for selective fallbacks (#170255)

2025-12-13 23:08:52 +00:00

test_gpu_cpp_wrapper.py

[Inductor] Fix combo kernels for cpu backend (#167781)

2025-12-04 23:40:47 +00:00

test_gpu_select_algorithm.py

[xpu][feature] Supports woq_int8 inductor pattern on Intel GPU (#163615)

2025-12-19 07:10:30 +00:00

test_graph_transform_observer.py

…

test_group_batch_fusion.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_halide.py

…

test_helion_kernels.py

…

test_indexing.py

…

test_inductor_annotations.py

…

test_inductor_freezing.py

[ROCm] Fix and enable several Triton autotuning, CUDAGraph trees, and layout optimization tests (#169901)

2025-12-19 22:09:37 +00:00

test_inductor_scheduler.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_inductor_utils.py

…

test_inplace_padding.py

…

test_inplacing_pass.py

reinplace effectful kernel (#170835)

2025-12-27 02:51:43 +00:00

test_kernel_benchmark.py

[inductor] fix kernel benchmark runner (#169976)

2025-12-10 23:28:47 +00:00

test_kernel_optimization.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_layout_optim.py

…

test_lookup_table.py

[ROCm] enable decompose k tests for functional coverage (#169948)

2025-12-30 09:48:53 +00:00

test_loop_ordering.py

Fix coalesce analysis to work with symints (#161819)

2025-12-15 23:30:18 +00:00

test_max_autotune_blackwell.py

[Inductor] Support appending TLX-only CUDA options in triton template (#171117)

2025-12-23 00:51:09 +00:00

test_max_autotune.py

[ROCm] enable decompose k tests for functional coverage (#169948)

2025-12-30 09:48:53 +00:00

test_mem_estimation.py

Enable ruff UP035 rule (#167307)

2025-11-09 06:40:03 +00:00

test_memory_planning.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_memory.py

[BE][1/6] fix typos in test/ (#157635)

2025-11-29 02:49:44 +00:00

test_metrics.py

…

test_minifier_isolate.py

…

test_minifier_utils.py

…

test_minifier.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_mix_order_reduction.py

[ROCm][CI] skip tests for mi200 (#170205)

2025-12-16 03:37:32 +00:00

test_mkldnn_pattern_matcher.py

Delete pt2e flow from pytorch/pytorch since it's migrated to torchao (#169151)

2025-12-23 20:41:04 +00:00

test_mmdecomp.py

…

test_move_constructors_to_gpu.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_mps_basic.py

…

test_multi_kernel.py

[xpu][test] [1/N] Enable missing Intel GPU inductor tests (#167047)

2025-11-18 05:28:35 +00:00

test_native_matmul.py

TestTritonDotReduction.test_matmul_fp16 correct tol (#169682)

2025-12-05 17:42:35 +00:00

test_needs_exact_strides.py

…

test_nv_universal_gemm.py

[Inductor] Add Nvidia Universal Gemm Backend to Inductor (#170623)

2025-12-24 00:42:12 +00:00

test_online_softmax.py

[xpu][feature] Enable triton online softmax kernels on XPU. (#163251)

2025-12-16 06:37:48 +00:00

test_op_completeness.py

…

test_op_dtype_prop.py

Revert "[inductor] Fix unbacked floats being downcast to fp32 in Triton kernels (#170289)"

2025-12-19 00:26:35 +00:00

test_ordered_set.py

Add missing super().setUp() (#167163)

2025-11-06 17:55:23 +00:00

test_pad_mm.py

Back out "Revert "[inductor][fx] clarify padding logic (#169577)"" (#170295)

2025-12-15 17:23:07 +00:00

test_padding.py

Revert "[Inductor] addmm with bias -> unfuse bias if there is a pointwise/reduction consumer (#166165)"

2025-11-06 22:34:48 +00:00

test_pallas.py

[pallas backend] skip flaky test test_constant_pad_2d_strides_nonpositive_cpu_pallas (#171564)

2025-12-31 04:56:20 +00:00

test_pattern_matcher.py

[inductor] remove redundant view/permute earlier (#170649)

2025-12-27 02:55:23 +00:00

test_perf.py

[dynamo] Rehaul the autograd.Function support (#166788)

2025-12-12 03:28:51 +00:00

test_profiler.py

[dynamo] Preserve record function in aot_eager (#167787)

2025-12-17 18:38:19 +00:00

test_provenance_tracing.py

[BE][1/6] fix typos in test/ (#157635)

2025-11-29 02:49:44 +00:00

test_quantization.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_remote_cache.py

…

test_scatter_optimization.py

…

test_segmented_tree.py

…

test_select_algorithm.py

Unit test for ExternKernelBenchmarkRequest TensorMeta construction (#170445)

2025-12-17 01:17:01 +00:00

test_selective_lowering.py

[Inductor] Add more tests for selective fallbacks (#170255)

2025-12-13 23:08:52 +00:00

test_smoke.py

…

test_snode_runtime.py

…

test_split_cat_fx_aten_passes.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_split_cat_fx_passes.py

[6/N] Use key in dict for existence checks (#168350)

2025-11-25 00:22:21 +00:00

test_static_cuda_launcher.py

[xpu][static_launcher] 1/4: Add config.use_static_triton_launcher as alias of use_static_cuda_launcher. (#168951)

2025-12-29 09:10:50 +00:00

test_subgraph_choice.py

[ROCm] unskip some ROCm inductor UTs (#169564)

2025-12-05 06:01:56 +00:00

test_template_heuristics_registry.py

…

test_torchbind.py

[hoo] Invoke subgraph + effects Inductor support (#167364)

2025-12-03 20:43:34 +00:00

test_torchinductor_codegen_config_overrides.py

…

test_torchinductor_codegen_dynamic_shapes.py

Fix unsafe_chunk decomposition for empty tensors (#170053)

2025-12-17 08:53:02 +00:00

test_torchinductor_dynamic_shapes.py

[Inductor] fix select_scatter dtype AssertionError (#171311)

2025-12-30 03:05:18 +00:00

test_torchinductor_opinfo.py

Fix FP16 tests on CPU in test_torchinductor_opinfo (#169932)

2025-12-22 00:41:22 +00:00

test_torchinductor_strided_blocks.py

[Resubmit] Fix _split_iteration_ranges handling gt handling for two-dimensional tiling (#170786)

2025-12-19 22:09:50 +00:00

test_torchinductor.py

[Inductor] fix select_scatter dtype AssertionError (#171311)

2025-12-30 03:05:18 +00:00

test_triton_cpu_backend.py

…

test_triton_extension_backend.py

Revert "[inductor] Use custom triton kernel subclass when available (#167456)"

2025-11-23 00:52:42 +00:00

test_triton_heuristics.py

[ROCm] Enable shared memory based pruning for Triton configs (#169974)

2025-12-11 16:13:09 +00:00

test_triton_kernels.py

Fix internal breakage with triton kernels (#171217)

2025-12-24 05:23:05 +00:00

test_triton_syntax.py

…

test_triton_wrapper.py

…

test_unbacked_symints.py

Revert "no need to bail out for unbacked in statically_known_multiple_of (#170058)"

2025-12-18 18:37:19 +00:00

test_utils.py

…

test_xpu_basic.py

…