CUDA Templates and Python DSLs for High-Performance Linear Algebra
This report summarizes scanner evidence for NVIDIA/cutlass, a C++ project with 10,558 GitHub stars. SynthScan examined 1,897,910 lines of code across 4854 source files, recording 10577 pattern matches distributed across 21 syntactic categories. The legacy adjusted pattern signal is 9.5 (Moderate pattern signal); it is not an authorship estimate.
Use the findings to ask whether maintainers can explain intent, verify dependencies, simplify unnecessary scaffolding, and focus review effort. Risk Payload v1 adds ownership, supply-chain, over-engineering, and review-effort vectors when available. These are prompts for human verification, not probabilities, security findings, or measured maintenance costs.
Use each vector to guide manual verification of the underlying findings.
Can maintainers explain intent, assumptions, and trade-offs? 5702 finding(s).
Are imports declared and package/API names verified? 1433 finding(s).
Are wrappers, guards, and repeated structures necessary? 5648 finding(s).
Which changes need focused human review? 10577 finding(s).
This chart tracks the legacy adjusted pattern score across scan runs. Changes can reflect source changes, scan coverage, or detector/scoring versions; they do not measure AI adoption or prove that a team refactored code. Compare runs only when their scoring versions are compatible. The dashed line tracks the raw pattern-hit count.
Groups findings by configured severity weight. Severity is not a confidence estimate, security rating, or proof that a package is missing. Verify imports against the project's manifests and inspect each finding in its local design context.
This horizontal bar chart shows where configured pattern scores accumulate across directories. Treat the bars as locations to inspect, not as evidence of authorship or code quality. Use the file-level findings to decide whether a directory needs a focused ownership, dependency, complexity, or review check.
The scanner identified 10577 distinct pattern matches across 21 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.
Reading the findings table: The Severity column indicates the configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM⚡ | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 60 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM⚡ | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 62 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM⚡ | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 69 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM⚡ | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 71 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 161 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 163 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 181 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 183 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 205 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 207 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 224 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 226 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 243 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 245 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 264 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 266 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 289 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 291 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 309 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 311 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 329 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 331 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 349 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 351 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 368 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 371 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 65 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 67 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 194 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 196 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 215 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 217 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 241 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 243 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 293 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 295 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 332 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 334 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 365 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 367 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 392 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 394 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 410 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 412 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 434 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 436 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 457 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 459 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 531 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 533 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 567 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 569 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 592 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 594 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 90 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 92 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 182 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 184 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 226 | # --------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 228 | # --------------------------------------------------------------------------- | COMMENT |
| 1908 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | test/python/cutlass/conv2d/conv2d_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_mixed_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f16_sm90.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f64_sm90.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f64_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f16_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_s8_sm90.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f32_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_s8_sm80.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | test/python/cutlass/gemm/gemm_f8_sm90.py | 0 | wrapper class to which tests will be added dynamically in __main__ | STRING |
| HIGH | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 0 | this example provides an implementation of the sm107 batched dense blockscaled gemm kernel, please note that the apis an | STRING |
| HIGH | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 0 | this example provides an implementation of the sm107 batched dense blockscaled gemm kernel, please note that the apis an | STRING |
| HIGH | …/blockscaled_gemm/dense_blockscaled_gemm_persistent.py | 0 | this example provides an implementation of the sm107 batched dense blockscaled gemm kernel, please note that the apis an | STRING |
| HIGH | …mm/dense_blockscaled_gemm_persistent_mixed_clusters.py | 0 | this example provides an implementation of the sm107 batched dense blockscaled gemm kernel, please note that the apis an | STRING |
| HIGH | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …ions/sm100_dense_blockscaled_static_persistent_impl.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …mm/dense_blockscaled_gemm_persistent_mixed_clusters.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …/blockscaled_gemm/dense_blockscaled_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …led_gemm/dense_blockscaled_gemm_persistent_prefetch.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …/blockscaled_gemm/dense_blockscaled_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …kscaled_gemm/dense_blockscaled_gemm_persistent_amax.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …bin/kernel/grouped_gemm/grouped_gemm_mixed_clusters.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …ations/sm100_contiguous_offset_2d3d_dense_gemm_impl.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …/cutedsl/gemm/implementations/sm100_persistent_impl.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …/cute/rubin/kernel/dense_gemm/dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …ell/kernel/dense_gemm/dense_gemm_persistent_dynamic.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …well/kernel/dense_gemm/dense_gemm_software_pipeline.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …e/blackwell/kernel/dense_gemm/dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …CuTeDSL/cute/blackwell/kernel/dense_gemm/dense_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …ll/kernel/dense_gemm/dense_gemm_persistent_prefetch.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …buted/distributed_gemm_all_reduce_ldxstmc_blackwell.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …l/distributed/distributed_all_gather_gemm_blackwell.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …stributed/distributed_gemm_reduce_scatter_blackwell.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …buted/distributed_gemm_all_reduce_lamport_blackwell.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …cute/blackwell/kernel/blockwise_gemm/blockwise_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …blackwell/kernel/blockwise_gemm/masked_grouped_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …kwell/kernel/blockwise_gemm/contiguous_grouped_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …CuTeDSL/cute_ext/blackwell/dense_gemm_cute_pipeline.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …on/CuTeDSL/cute/blackwell/efc/common_dense_gemm_efc.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …/kernel/dense_gemm/dense_gemm_alpha_beta_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …sl/gemm/implementations/sm90_static_persistent_impl.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …er/kernel/dense_gemm/dense_gemm_fp8_gelu_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …cute/hopper/kernel/dense_gemm/dense_gemm_persistent.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …on/CuTeDSL/cute/hopper/kernel/dense_gemm/dense_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …uTeDSL/cute/hopper/kernel/grouped_gemm/grouped_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …DSL/cute/rubin/kernel/blockwise_gemm/blockwise_gemm.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …rouped_gemm/grouped_blockscaled_gemm_mixed_clusters.py | 0 | set up configurations that are dependent on gemm inputs. this method configures various attributes based on the input te | STRING |
| HIGH | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …mm/dense_blockscaled_gemm_persistent_mixed_clusters.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …ions/sm100_dense_blockscaled_static_persistent_impl.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …/blockscaled_gemm/dense_blockscaled_gemm_persistent.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| HIGH | …led_gemm/dense_blockscaled_gemm_persistent_prefetch.py | 0 | execute the gemm operation in steps: - setup static attributes before smem/grid/tma computation - setup tma load/store a | STRING |
| 667 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | CMakeLists.txt | 421 | list(APPEND CUTLASS_CUDA_NVCC_FLAGS -DCUTLASS_SM100_FAMILY_ARCHS_ENABLED) | COMMENT |
| LOW | CMakeLists.txt | 881 | # DEPENDS: A list of targets or files on which this test is dependent. | COMMENT |
| LOW | tools/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | tools/util/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | tools/util/include/cutlass/util/print_error.hpp | 41 | #include <cute/tensor.hpp> | COMMENT |
| LOW | tools/util/include/cutlass/util/cublas_wrappers.hpp | 41 | # define BLAM_DEBUG_OUT(msg) std::cerr << "BLAM: " << msg << std::endl | COMMENT |
| LOW | …l/include/cutlass/util/reference/host/rank_k_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …lude/cutlass/util/reference/host/gemm_planar_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …util/include/cutlass/util/reference/host/convolution.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …til/include/cutlass/util/reference/host/symm_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …til/include/cutlass/util/reference/host/gemm_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …/include/cutlass/util/reference/host/tensor_reduce.hpp | 41 | COMMENT | |
| LOW | …il/include/cutlass/util/reference/host/tensor_fill.hpp | 41 | COMMENT | |
| LOW | …include/cutlass/util/reference/host/tensor_compare.hpp | 41 | COMMENT | |
| LOW | …util/include/cutlass/util/reference/host/tensor_fill.h | 41 | #include <random> | COMMENT |
| LOW | …/include/cutlass/util/reference/host/rank_2k_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | tools/util/include/cutlass/util/reference/device/gemm.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …il/include/cutlass/util/reference/device/convolution.h | 41 | #include "cutlass/matrix_shape.h" | COMMENT |
| LOW | …l/include/cutlass/util/reference/device/gemm_complex.h | 41 | #include "cutlass/numeric_conversion.h" | COMMENT |
| LOW | …il/include/cutlass/util/reference/device/tensor_fill.h | 41 | COMMENT | |
| LOW | tools/util/scripts/split_test_cmake.py | 1 | ################################################################################################# | COMMENT |
| LOW | tools/util/scripts/split_test_cmake.py | 21 | # AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | COMMENT |
| LOW | tools/library/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | tools/library/include/cutlass/library/library.h | 41 | COMMENT | |
| LOW | tools/library/include/cutlass/library/library.h | 61 | #include "cutlass/tensor_coord.h" | COMMENT |
| LOW | tools/library/src/conv_operation_3x.hpp | 501 | // Convolutions split the metadata (in Conv2dConfiguration) from | COMMENT |
| LOW | tools/library/src/conv_operation_3x.hpp | 741 | else { | COMMENT |
| LOW | tools/library/src/conv_operation_3x.hpp | 841 | if (num_groups != 1) { | COMMENT |
| LOW | tools/library/src/conv2d_operation.h | 41 | #include "cutlass/conv/kernel/default_conv2d_dgrad.h" | COMMENT |
| LOW | tools/library/src/sparse_gemm_operation_3x.hpp | 41 | #include "cutlass/library/library.h" | COMMENT |
| LOW | tools/library/src/gemm_operation_3x.hpp | 41 | #include "cutlass/library/library.h" | COMMENT |
| LOW | …ary/src/reference/blockwise_gemm_reference_operation.h | 41 | #include <cstring> | COMMENT |
| LOW | tools/profiler/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | tools/profiler/include/cutlass/profiler/debug.h | 41 | COMMENT | |
| LOW | …r/include/cutlass/profiler/conv2d_operation_profiler.h | 41 | #include <algorithm> | COMMENT |
| LOW | tools/profiler/include/cutlass/profiler/options.h | 221 | COMMENT | |
| LOW | …lude/cutlass/profiler/sparse_gemm_operation_profiler.h | 41 | #include <algorithm> | COMMENT |
| LOW | …ler/include/cutlass/profiler/symm_operation_profiler.h | 41 | #include <memory> | COMMENT |
| LOW | …e/cutlass/profiler/blockwise_gemm_operation_profiler.h | 41 | #include <memory> | COMMENT |
| LOW | …ler/include/cutlass/profiler/trmm_operation_profiler.h | 41 | #include <memory> | COMMENT |
| LOW | …nclude/cutlass/profiler/reduction_operation_profiler.h | 41 | #include <algorithm> | COMMENT |
| LOW | …/include/cutlass/profiler/rank_2k_operation_profiler.h | 41 | #include <memory> | COMMENT |
| LOW | …utlass/profiler/block_scaled_gemm_operation_profiler.h | 41 | #include <string> | COMMENT |
| LOW | …profiler/include/cutlass/profiler/operation_profiler.h | 41 | COMMENT | |
| LOW | …ler/include/cutlass/profiler/gemm_operation_profiler.h | 41 | #include <algorithm> | COMMENT |
| LOW | …r/include/cutlass/profiler/conv3d_operation_profiler.h | 41 | #include <algorithm> | COMMENT |
| LOW | …r/include/cutlass/profiler/rank_k_operation_profiler.h | 41 | #include <memory> | COMMENT |
| LOW | …ude/cutlass/profiler/grouped_gemm_operation_profiler.h | 41 | #include <unordered_map> | COMMENT |
| LOW | tools/profiler/src/performance_report.cpp | 41 | #include "cutlass/library/util.h" | COMMENT |
| LOW | tools/profiler/src/cudnn_helpers.cpp | 241 | if(a_tensor.element != b_tensor.element) { | COMMENT |
| LOW | tools/profiler/src/cudnn_helpers.cpp | 381 | // where; div_up(a, b) : (a - 1)/b + 1 // | COMMENT |
| LOW | test/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | test/self_contained_includes/CMakeLists.txt | 1 | # Copyright (c) 2024 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | test/self_contained_includes/CMakeLists.txt | 21 | # DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE | COMMENT |
| LOW | test/self_contained_includes/CMakeLists.txt | 221 | # cutlass/arch | COMMENT |
| LOW | test/unit/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | test/unit/pipeline/CMakeLists.txt | 1 | # Copyright (c) 2017 - 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | test/unit/pipeline/testbed_cluster_launch_control.h | 41 | #define CUTLASS_UNIT_TEST_PIPELINE true | COMMENT |
| LOW | test/unit/pipeline/testbed.h | 41 | #include "cutlass/util/command_line.h" | COMMENT |
| 2200 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | test/python/pycute/test_right_inverse.py | 46 | def helper_test_right_inverse(self, layout): | CODE |
| LOW | test/python/CuTeDSL/test_no_source_diagnostics.py | 27 | def test_exec_defined_function_raises_user_error(self): | CODE |
| LOW | test/python/CuTeDSL/test_no_source_diagnostics.py | 48 | def test_uncaught_dsl_error_does_not_kill_interactive_session(self): | CODE |
| LOW | test/python/CuTeDSL/test_no_source_diagnostics.py | 74 | def test_uncaught_dsl_error_in_script_still_exits_nonzero(self): | CODE |
| LOW | test/python/CuTeDSL/test_link_libraries_order.py | 26 | def test_preserves_attribute_order_and_deduplicates(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_export_c_header_shape_type.py | 115 | def test_static_shape_has_no_shape_slots(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_export_c_header_shape_type.py | 120 | def test_dlpack_tensor_shape_slots(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_export_c_header_shape_type.py | 126 | def test_mixed_width_shape_dims_rejected(self): | CODE |
| LOW | test/python/CuTeDSL/test_export_c_header_shape_type.py | 138 | def test_tensor_list_shape_slots(self): | CODE |
| LOW | test/python/CuTeDSL/test_export_c_header_shape_type.py | 152 | def test_tensor_list_mismatched_widths_rejected(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 125 | def test_in_range_values_are_preserved(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 135 | def test_out_of_range_values_still_wrap(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 145 | def test_floats_truncate_toward_zero(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 152 | def test_bools_fold_to_zero_and_one(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 159 | def test_boolean_folds_nonzero_to_one(self): | CODE |
| LOW⚡ | test/python/CuTeDSL/test_numeric_construction.py | 168 | def test_nan_and_infinity_are_rejected(self): | CODE |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 182 | def test_boolean_folds_nan_and_infinity_to_true(self): | CODE |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 188 | def test_widths_without_a_numpy_dtype_are_rejected(self): | CODE |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 202 | def test_exact_range_tracks_the_numpy_dtype(self): | CODE |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 211 | def test_range_follows_the_dtype_not_the_width(self): | CODE |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 241 | def test_from_other_integer_type(self): | CODE |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 48 | def test_get_tensor_inside_dynamic_if(self): | CODE |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 70 | def test_struct_use_outside_if_still_works(self): | CODE |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 91 | def test_nested_struct_inside_dynamic_if(self): | CODE |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 137 | def test_if_elif_else_warp_spec_pattern(self): | CODE |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 166 | def test_struct_captured_in_while_body(self): | CODE |
| LOW | test/python/cutlass/installation.py | 45 | def test_cutlass_source_paths(self): | CODE |
| LOW | test/python/cutlass/interface/evt_interface.py | 92 | def test_too_much_shared_memory(self): | CODE |
| LOW | test/python/cutlass/interface/evt_interface.py | 96 | def evt_too_much_shared_memory(accum, C1, C2, C3, C4, C5, C6, C7, C8): | CODE |
| LOW | test/python/cutlass/interface/evt_interface.py | 175 | def test_missing_example_tensor(self): | CODE |
| LOW | test/python/cutlass/interface/evt_interface.py | 179 | def evt_missing_example_tensor(accum, C): | CODE |
| LOW | test/python/cutlass/interface/gemm_interface.py | 237 | def test_tensorop_availability(self): | CODE |
| LOW | test/python/cutlass/interface/gemm_interface.py | 272 | def test_invalid_tile_description(self): | CODE |
| LOW | test/python/cutlass/interface/conv2d_interface.py | 268 | def test_invalid_tile_description(self): | CODE |
| LOW | test/python/cutlass/evt/evt_mixed_sm80_90.py | 53 | def test_same_variable_used_multiple_times(self): | CODE |
| LOW | test/python/cutlass/evt/evt_mixed_sm80_90.py | 193 | def test_mixed_dag_partition_k(self): | CODE |
| LOW | test/utils/test_sharding.py | 245 | def pytest_collection_modifyitems(config, items): | CODE |
| LOW | test/utils/test_sharding.py | 465 | def pytest_runtest_makereport(item, call): | CODE |
| LOW | test/utils/test_sharding.py | 498 | def pytest_report_collectionfinish(config, start_path, items): | CODE |
| LOW | test/examples/CuTeDSL/test_for_control_flow.py | 41 | def _for_loop_carried_reused_target_kernel(out: cute.Tensor, n: cutlass.Int32): | CODE |
| LOW | test/examples/CuTeDSL/test_for_control_flow.py | 53 | def _for_loop_carried_reused_target_host(out: cute.Tensor, n: cutlass.Int32): | CODE |
| LOW | test/examples/CuTeDSL/test_for_control_flow.py | 60 | def test_for_loop_carried_var_with_reused_loop_target(): | CODE |
| LOW | test/examples/CuTeDSL/test_dataclasses.py | 50 | def _test_empty_dataclass_kernel(out: cute.Tensor, tag: A | B): | CODE |
| LOW | test/examples/CuTeDSL/test_dataclasses.py | 61 | def _test_empty_dataclass_host(out: cute.Tensor, tag: A | B): | CODE |
| LOW | test/examples/CuTeDSL/test_dataclasses.py | 66 | def test_empty_dataclass_union(tag, expected): | CODE |
| LOW | test/examples/CuTeDSL/test_tvm_ffi_env_stream.py | 53 | def test_env_stream_detected_from_nested_tuple(): | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 129 | def test_grouped_blockscaled_gemm( | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 166 | def test_grouped_blockscaled_gemm_large_persistent_repro(): | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 199 | def test_invalid_dtypes_and_sf_vec_size( | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 283 | def test_invalid_mma_tiler_and_cluster_shape( | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 319 | def test_invalid_tensor_alignment(problem_sizes_mnkl: List[Tuple[int, int, int, int]]): | CODE |
| LOW | …amples/CuTeDSL/sm_103/test_grouped_blockscaled_gemm.py | 364 | def test_batch_dimension_not_one(): | CODE |
| LOW | test/examples/CuTeDSL/sm_100a/test_rmsnorm.py | 89 | def test_rmsnorm_without_weight(self, N): | CODE |
| LOW | test/examples/CuTeDSL/sm_100a/test_rmsnorm.py | 110 | def test_cluster_path_correctness(self, N): | CODE |
| LOW | …CuTeDSL/sm_100a/test_dense_gemm_persistent_prefetch.py | 222 | def test_prefetch_dist_configurations(prefetch_dist: Optional[int]): | CODE |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 169 | def test_dense_blockscaled_gemm_prefetch( | CODE |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 290 | def test_prefetch_dist_configurations(prefetch_dist: Optional[int]): | CODE |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 352 | def test_invalid_dtypes_and_scale_factor_vec_size( | CODE |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 427 | def test_invalid_mma_tiler_and_cluster_shape( | CODE |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 481 | def test_invalid_tensor_alignment( | CODE |
| 1969 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | test/python/pycute/test_coalesce.py | 40 | CODE | |
| LOW | test/python/pycute/test_left_inverse.py | 40 | CODE | |
| LOW | test/python/pycute/test_complement.py | 40 | CODE | |
| LOW | test/python/pycute/test_int_tuple.py | 39 | CODE | |
| LOW | test/python/pycute/test_typing.py | 39 | CODE | |
| LOW | test/python/pycute/test_composition.py | 40 | CODE | |
| LOW | test/python/pycute/test_right_inverse.py | 40 | CODE | |
| LOW | test/python/CuTeDSL/test_export_c_header_shape_type.py | 30 | CODE | |
| LOW | test/python/CuTeDSL/test_struct_in_if.py | 24 | CODE | |
| LOW | test/python/cutlass/interface/gemm_interface.py | 37 | CODE | |
| LOW | test/python/cutlass/interface/conv2d_interface.py | 37 | CODE | |
| LOW | test/python/cutlass/evt/evt_layout_sm80_90.py | 41 | CODE | |
| LOW | test/python/cutlass/evt/evt_layout_sm80_90.py | 42 | CODE | |
| LOW | test/python/cutlass/evt/evt_mixed_sm80_90.py | 41 | CODE | |
| LOW | test/python/cutlass/evt/evt_mixed_sm80_90.py | 42 | CODE | |
| LOW | test/python/cutlass/evt/evt_compute_sm80_90.py | 40 | CODE | |
| LOW | test/python/cutlass/evt/evt_compute_sm80_90.py | 41 | CODE | |
| LOW | test/python/cutlass/evt/evt_compute_sm80_90.py | 42 | CODE | |
| LOW | test/python/cutlass/evt/evt_load_sm80_90.py | 41 | CODE | |
| LOW | test/python/cutlass/evt/evt_load_sm80_90.py | 42 | CODE | |
| LOW | test/python/cutlass/evt/evt_store_sm80_90.py | 41 | CODE | |
| LOW | test/python/cutlass/evt/evt_store_sm80_90.py | 42 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_sm80.py | 43 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_problem_sizes.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 41 | CODE | |
| LOW | test/python/cutlass/conv2d/conv2d_test_utils.py | 55 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_batched.py | 37 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_batched.py | 43 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_batched.py | 46 | CODE | |
| LOW | test/python/cutlass/gemm/utils.py | 36 | CODE | |
| LOW | test/python/cutlass/gemm/utils.py | 45 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 34 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 35 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 36 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 40 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 51 | CODE | |
| LOW | test/examples/CuTeDSL/conftest.py | 53 | CODE | |
| LOW | test/examples/CuTeDSL/test_dataclasses.py | 34 | CODE | |
| LOW | test/examples/CuTeDSL/test_math.py | 29 | CODE | |
| LOW | test/examples/CuTeDSL/test_math.py | 32 | CODE | |
| LOW | operators/test/conftest.py | 32 | CODE | |
| LOW | operators/test/unit/test_manifest.py | 33 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 29 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 29 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 30 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 30 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 30 | CODE | |
| LOW | operators/test/test_utils/__init__.py | 31 | CODE | |
| LOW | operators/test/test_utils/common.py | 29 | CODE | |
| LOW | operators/test/test_utils/reference_check.py | 29 | CODE | |
| LOW | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 31 | CODE | |
| LOW | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 29 | CODE | |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 31 | CODE | |
| LOW | operators/cutlass/operators/artifact.py | 29 | CODE | |
| LOW | operators/cutlass/operators/manifest.py | 29 | CODE | |
| LOW | operators/cutlass/operators/arch.py | 29 | CODE | |
| 1368 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 122 | except Exception: | CODE |
| HIGH | python/CuTeDSL/_mlir_helpers/__init__.py | 28 | except ImportError: | CODE |
| HIGH | python/CuTeDSL/_mlir_helpers/op.py | 460 | except RuntimeError: | CODE |
| HIGH | python/CuTeDSL/cutlass/testing.py | 557 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/testing.py | 591 | except Exception: | CODE |
| HIGH | …eDSL/cutlass/experimental/task_scheduling/resources.py | 1766 | except TypeError: | CODE |
| HIGH | …eDSL/cutlass/experimental/task_scheduling/resources.py | 1774 | except TypeError: | CODE |
| HIGH | python/CuTeDSL/cutlass/cute/nvgpu/tcgen05/mma.py | 366 | except KeyError: | CODE |
| HIGH | python/CuTeDSL/cutlass/utils/rubin_helpers.py | 321 | except ValueError: | CODE |
| HIGH | python/CuTeDSL/cutlass/utils/tensor_helpers.py | 65 | except AttributeError: | CODE |
| HIGH | python/CuTeDSL/cutlass/utils/profiling.py | 189 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 494 | except OSError: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2411 | except ValueError: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2802 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 3126 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2380 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1878 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 3205 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 2664 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 3161 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4299 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4310 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4781 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4982 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4990 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 5104 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 5117 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4668 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 5074 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 3091 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4978 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 3622 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4070 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 137 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 367 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 659 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 975 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 996 | except (AttributeError, TypeError): | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 1328 | except TypeError: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 1899 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 1979 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 2120 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 2210 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 3586 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 3645 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 3694 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 3895 | except TypeError: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 3924 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 4174 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 4365 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 4929 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 5123 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 5190 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 5299 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 5429 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 6164 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 7369 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 7673 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 8847 | except Exception: | CODE |
| HIGH | python/CuTeDSL/cutlass/base_dsl/pyir_core.py | 734 | except (TypeError, ValueError, AttributeError): | CODE |
| 82 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 104 | CODE | |
| LOW | test/python/CuTeDSL/test_numeric_construction.py | 188 | CODE | |
| LOW | test/python/cutlass/interface/gemm_interface.py | 272 | CODE | |
| LOW | test/python/cutlass/evt/utils/evt_testbed.py | 140 | CODE | |
| LOW | test/python/cutlass/gemm/gemm_testbed.py | 319 | CODE | |
| LOW | test/utils/test_sharding.py | 245 | CODE | |
| LOW | test/utils/test_sharding.py | 465 | CODE | |
| LOW | test/examples/CuTeDSL/conftest.py | 126 | CODE | |
| LOW | operators/test/test_utils/reference_check.py | 334 | CODE | |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 717 | CODE | |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 358 | CODE | |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 552 | CODE | |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 1119 | CODE | |
| LOW | operators/cutlass/operators/mma.py | 105 | CODE | |
| LOW | operators/cutlass/operators/fusion/library.py | 343 | CODE | |
| LOW | operators/cutlass/operators/fusion/library.py | 360 | CODE | |
| LOW | operators/cutlass/operators/fusion/pycute/layout.py | 135 | CODE | |
| LOW | operators/cutlass/operators/fusion/pycute/layout.py | 196 | CODE | |
| LOW | operators/cutlass/operators/fusion/ir/store_nodes.py | 129 | CODE | |
| LOW | operators/cutlass/operators/fusion/ir/c_types.py | 115 | CODE | |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 87 | CODE | |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 146 | CODE | |
| LOW | operators/cutlass/operators/fusion/ir/layout_nodes.py | 259 | CODE | |
| LOW | operators/cutlass/operators/fusion/ir/tensor.py | 58 | CODE | |
| LOW | operators/cutlass/operators/fusion/ir/node.py | 213 | CODE | |
| LOW | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 55 | CODE | |
| LOW | …rators/cutlass/operators/fusion/passes/pass_manager.py | 159 | CODE | |
| LOW | …rators/cutlass/operators/fusion/passes/graph_drawer.py | 76 | CODE | |
| LOW | …s/cutlass/operators/providers/cutedsl/evt/converter.py | 297 | CODE | |
| LOW | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 544 | CODE | |
| LOW | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 1734 | CODE | |
| LOW | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 552 | CODE | |
| LOW | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 599 | CODE | |
| LOW | …ers/cutedsl/gemm/sm100_persistent_preferred_cluster.py | 238 | CODE | |
| LOW | …ers/cutedsl/gemm/sm100_persistent_preferred_cluster.py | 374 | CODE | |
| LOW | …ers/cutedsl/gemm/sm100_persistent_preferred_cluster.py | 544 | CODE | |
| LOW | …/providers/cutedsl/gemm/sm100_static_persistent_efc.py | 303 | CODE | |
| LOW | …ators/providers/cutedsl/gemm/sm90_static_persistent.py | 151 | CODE | |
| LOW | …ators/providers/cutedsl/gemm/sm90_static_persistent.py | 264 | CODE | |
| LOW | …s/operators/providers/cutedsl/gemm/sm100_persistent.py | 169 | CODE | |
| LOW | …s/operators/providers/cutedsl/gemm/sm100_persistent.py | 299 | CODE | |
| LOW | …operators/providers/cutedsl/gemm/sm80_tensorop_gemm.py | 201 | CODE | |
| LOW | …/operators/providers/cutedsl/gemm/sm100_mixed_input.py | 691 | CODE | |
| LOW | …rs/cutedsl/gemm/implementations/sm100_tgv_gemm_impl.py | 276 | CODE | |
| LOW | …utedsl/gemm/implementations/sm80_tensorop_gemm_impl.py | 254 | CODE | |
| LOW | …ations/sm100_contiguous_offset_2d3d_dense_gemm_impl.py | 655 | CODE | |
| LOW | …ations/sm100_contiguous_offset_2d3d_dense_gemm_impl.py | 1828 | CODE | |
| LOW | …ions/sm100_dense_blockscaled_static_persistent_impl.py | 720 | CODE | |
| LOW | …sl/gemm/implementations/sm90_static_persistent_impl.py | 357 | CODE | |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 1053 | CODE | |
| LOW | …ementations/sm100_persistent_preferred_cluster_impl.py | 437 | CODE | |
| LOW | …/cutedsl/gemm/implementations/sm100_persistent_impl.py | 557 | CODE | |
| LOW | operators/cutlass/operators/utils/tensor.py | 347 | CODE | |
| LOW | operators/cutlass/operators/utils/tensor.py | 528 | CODE | |
| LOW | operators/cutlass/operators/arguments/base.py | 99 | CODE | |
| LOW | python/cutlass_cppgen/library_defaults.py | 109 | CODE | |
| LOW | python/cutlass_cppgen/library_defaults.py | 236 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 479 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 506 | CODE | |
| LOW | python/cutlass_cppgen/utils/datatypes.py | 84 | CODE | |
| 584 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | test/python/cutlass/evt/utils/evt_testbed.py | 108 | # Create the Gemm plan | COMMENT |
| MEDIUM | test/python/cutlass/conv2d/conv2d_test_utils.py | 347 | # Create the plan | COMMENT |
| MEDIUM | operators/test/integration/test_cuda_graph.py | 91 | # Create a CUDA Graph to run our compiled Operator N times | COMMENT |
| MEDIUM | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 466 | # Create a temporary kernel object to check base implementation | COMMENT |
| MEDIUM | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 125 | """This class implements batched matrix multiplication (C = A x SFA x B x SFB) with support for various data types | STRING |
| MEDIUM | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 191 | """This class implements batched matrix multiplication (C = A x B) with support for various data types | STRING |
| MEDIUM | operators/cutlass/operators/fusion/__init__.py | 69 | # Define the example tensors | STRING |
| MEDIUM | …rators/cutlass/operators/fusion/frontend/python_ast.py | 153 | # Create the assign node | COMMENT |
| MEDIUM | operators/cutlass/operators/fusion/ir/c_types.py | 159 | # Create the ctype tuple | COMMENT |
| MEDIUM | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 136 | # Create the subgraph | COMMENT |
| MEDIUM | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 149 | # Create the fused node | COMMENT |
| MEDIUM | …utlass/operators/providers/cutedsl/evt/efc/_helpers.py | 164 | # Create a new signature with the custom parameter names | COMMENT |
| MEDIUM | …utlass/operators/providers/cutedsl/evt/efc/_helpers.py | 167 | # Create a wrapper function that accepts arguments according to the new signature | COMMENT |
| MEDIUM | …ors/providers/cutedsl/evt/efc/dense_gemm/_operation.py | 122 | # The following functions are executed by the specialized warps for | COMMENT |
| MEDIUM | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 1136 | # The following functions are executed by the specialized warp for the | COMMENT |
| MEDIUM | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 1950 | # Create the local_tile gX_mnl for all the EFC supplemental tensors. | COMMENT |
| MEDIUM | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 2455 | # Create the tiled tensors to be loaded in the epilogue. | COMMENT |
| MEDIUM | …s/operators/providers/cutedsl/gemm/sm100_persistent.py | 80 | """This class implements batched matrix multiplication (C = A @ B) with support for various data types | STRING |
| MEDIUM | …operators/providers/cutedsl/gemm/sm80_tensorop_gemm.py | 68 | """This class implements batched matrix multiplication (C = A @ B) | STRING |
| MEDIUM | …rs/cutedsl/gemm/implementations/sm100_tgv_gemm_impl.py | 1023 | # Create the local tile of coordC, same tiling as gC | COMMENT |
| MEDIUM | …utedsl/gemm/implementations/sm80_tensorop_gemm_impl.py | 161 | # Create a copy atom for a global to shared memory asynchronous copy | COMMENT |
| MEDIUM | …utedsl/gemm/implementations/sm80_tensorop_gemm_impl.py | 519 | # Create the copy atoms for the copy from shared memory to register | COMMENT |
| MEDIUM | …ations/sm100_contiguous_offset_2d3d_dense_gemm_impl.py | 223 | """This class implements batched matrix multiplication (C = A * B) with support for fp8 (e4m3fn, e5m2) | STRING |
| MEDIUM | …ions/sm100_dense_blockscaled_static_persistent_impl.py | 120 | """This class implements batched matrix multiplication (C = A x SFA x B x SFB) with support for various data types | STRING |
| MEDIUM | python/cutlass_cppgen/backend/c_types.py | 538 | # Create the ctype tuple | COMMENT |
| MEDIUM | python/cutlass_cppgen/backend/compiler.py | 143 | # Create the table if it does not already exist | STRING |
| MEDIUM | python/cutlass_cppgen/backend/library.py | 52 | # The following block implements enum.auto() for Python 3.5 variants that don't include it such | COMMENT |
| MEDIUM | …thon/cutlass_cppgen/backend/evt/frontend/python_ast.py | 146 | # Create the assign node | COMMENT |
| MEDIUM | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 134 | # Create the subgraph | COMMENT |
| MEDIUM | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 146 | # Create the fused node | COMMENT |
| MEDIUM | python/cutlass_cppgen/epilogue/epilogue.py | 134 | # Define the example tensors | STRING |
| MEDIUM | …L/cutlass/experimental/task_scheduling/task_manager.py | 1042 | # Create a unique list of resources from all tasks and their src and dst | COMMENT |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 6005 | # Create a dictionary to store the offsets | COMMENT |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 6219 | # Create a dictionary to store the offsets (all zeros for union) | COMMENT |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 752 | # Create a scaled basis with integer scale and mode | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 755 | # Create a scaled basis with a Ratio scale | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 758 | # Create a scaled basis with a list of modes | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 954 | # Create a basis element for the first dimension (mode 0) | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 957 | # Create a basis element for the second dimension (mode 1) | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 960 | # Create a basis element for a hierarchical dimension | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3716 | # Create a 2D compact left-most layout with shape (4,4) | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3719 | # Create a left-most layout with custom strides | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3722 | # Create a layout for a 3D tensor | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3725 | # Create a layout with custom strides | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3851 | # Create a row-major layout | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3854 | # Create a column-major layout | STRING |
| MEDIUM⚡ | python/CuTeDSL/cutlass/cute/core.py | 3857 | # Create a layout with custom dimension ordering for a 3D tensor | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 1157 | # Creating a layout with shape (4,8) and default stride (layout left / "column major") | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 1160 | # Creating a layout with explicit shape and stride | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 2881 | tensor = make_tensor(...) # Create a 2D tensor | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 3807 | # Create a 2D identity layout with shape (4,4) | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 3810 | # Create a 3D identity layout | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 3999 | # Create a basic layout | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4003 | # Create a composed layout with an offset | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4350 | # Create a layout | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4607 | # Create a layout that maps (i,j) to i*4 + j | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4609 | # Create a layout that maps (i,j) to i*3 + j | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4661 | # Create a right-major layout for a 4x4 matrix | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 4663 | # Create a left-major layout that complements the row layout | STRING |
| MEDIUM | python/CuTeDSL/cutlass/cute/core.py | 6170 | # Define a union with scalar int/float elements: | STRING |
| 112 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | …ors/cutlass/operators/fusion/frontend/frontend_base.py | 223 | except Exception as e: | CODE |
| LOW | …rators/cutlass/operators/fusion/frontend/python_ast.py | 112 | except Exception as e: | CODE |
| LOW | …rators/cutlass/operators/fusion/frontend/python_ast.py | 198 | except Exception as e: | CODE |
| LOW | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 122 | except Exception: | CODE |
| LOW | …providers/cutedsl/evt/efc/dense_gemm/sm100/_helpers.py | 95 | except Exception as exc: # noqa: BLE001 -- diagnostic, must not abort. | CODE |
| LOW | operators/cutlass/operators/utils/device.py | 165 | except Exception as e: | CODE |
| LOW | operators/cutlass/operators/utils/device.py | 244 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/prep_editable_install.py | 160 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/prep_editable_install.py | 268 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/vector.py | 159 | except Exception: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 530 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 634 | except Exception: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 672 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/__init__.py | 21 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/__init__.py | 31 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 587 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 591 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/testing.py | 273 | except Exception as exc: | CODE |
| LOW | python/CuTeDSL/cutlass/testing.py | 553 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/testing.py | 557 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/testing.py | 811 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/testing.py | 924 | except Exception as e: | CODE |
| LOW | …L/cutlass/experimental/task_scheduling/task_manager.py | 364 | except Exception: | CODE |
| LOW | …hon/CuTeDSL/cutlass/experimental/primitives/gpu_ops.py | 471 | except Exception: | CODE |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 2952 | except Exception: | CODE |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 3062 | except Exception: | CODE |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 3344 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/cute/metadata.py | 555 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/cute/launch_facts.py | 47 | except Exception as exc: | CODE |
| LOW | python/CuTeDSL/cutlass/cute/runtime.py | 186 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/cute/testing.py | 730 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/memory/tmem.py | 87 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/utils/profiling.py | 189 | except Exception: | CODE |
| MEDIUM | python/CuTeDSL/cutlass/utils/profiling.py | 186 | def __del__(self) -> None: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 452 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 481 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 490 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 653 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 674 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_class_facts.py | 758 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/version_info.py | 17 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1543 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1720 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1795 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1878 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2380 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2504 | except Exception as e: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2802 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 3126 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 3205 | except Exception: | CODE |
| MEDIUM | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 3058 | def block (if any). | CODE |
| MEDIUM | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 5067 | def traced under a folded constant ``if`` arm | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4225 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4231 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4781 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4978 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4982 | except Exception: | CODE |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 4990 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 610 | except Exception: | CODE |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_entrypoints.py | 1142 | except Exception: | CODE |
| 406 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | test/utils/test_sharding.py | 429 | # Assign markers to each subtest | COMMENT |
| LOW | …00a/test_dense_blockscaled_gemm_persistent_prefetch.py | 310 | # Check if this configuration can be implemented | COMMENT |
| LOW | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 426 | # Check if prefetch is enabled (prefetch_dist > 0) | COMMENT |
| LOW | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 1263 | # Check if configuration can be implemented | COMMENT |
| LOW | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 823 | # Check if configuration can be implemented | COMMENT |
| LOW | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 965 | # Check if configuration can be implemented | COMMENT |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 2790 | # Check if configuration can be implemented | COMMENT |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 3023 | # Check if configuration can be implemented | COMMENT |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 733 | # Check if we should use Bkeep-Breuse pattern | COMMENT |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 1320 | # Check if configuration can be implemented | COMMENT |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 1445 | # Check if configuration can be implemented | COMMENT |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 458 | # Check if input data types are compatible with MMA instruction | COMMENT |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 1532 | # Check if configuration can be implemented | COMMENT |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 1639 | # Check if configuration can be implemented | COMMENT |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 553 | # Check if prefetch is enabled (prefetch_dist > 0) | COMMENT |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 2064 | # Check if the current combination is valid | COMMENT |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 2603 | # Check if configuration can be implemented | COMMENT |
| LOW | …oviders/cutedsl/evt/efc/dense_gemm/sm100/_operation.py | 1646 | # Check if input data types are compatible with MMA instruction | COMMENT |
| LOW | …ations/sm100_contiguous_offset_2d3d_dense_gemm_impl.py | 503 | # Check if input data types are compatible with MMA instruction | COMMENT |
| LOW | …ementations/sm100_persistent_preferred_cluster_impl.py | 256 | # Check if input data types are compatible with MMA instruction | COMMENT |
| LOW | …/cutedsl/gemm/implementations/sm100_persistent_impl.py | 443 | # Check if input data types are compatible with MMA instruction | COMMENT |
| LOW | python/cutlass_cppgen/op/conv.py | 473 | # Check if the iterator algorithm is valid | COMMENT |
| LOW | python/CuTeDSL/setup.sh | 81 | # Check if CUTLASS_IR_BUILD_DIR is set | COMMENT |
| LOW | python/CuTeDSL/setup.sh | 120 | # Check if requirements file exists | COMMENT |
| LOW | python/CuTeDSL/prep_editable_install.py | 118 | # Check if version already has .dev<n> pattern | COMMENT |
| LOW | python/CuTeDSL/_mlir_helpers/vector.py | 326 | # Check if all indices are static (Python ints) | COMMENT |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 531 | # Check if it's a decorator config error first | COMMENT |
| LOW | python/CuTeDSL/cutlass/block.py | 68 | # Check if required arguments are provided | COMMENT |
| LOW | python/CuTeDSL/cutlass/testing.py | 172 | # Check if test was explicitly selected like ::test_name[param1-param2-...] | COMMENT |
| LOW | python/CuTeDSL/cutlass/pipeline/sm100.py | 714 | # Set mask to None if not using clusters (i.e. 1CTA kernels) | COMMENT |
| LOW | python/CuTeDSL/cutlass/pipeline/sm100.py | 722 | # Set mask to None if not using 2CTA instructions | COMMENT |
| LOW | …L/cutlass/experimental/task_scheduling/task_manager.py | 4798 | # Check if any of those also have ProducerWork in LOOP. | COMMENT |
| LOW | python/CuTeDSL/cutlass/cute/tuple.py | 475 | # Check if first argument is a tuple to determine behavior | COMMENT |
| LOW | python/CuTeDSL/cutlass/cute/arch/nvvm_wrappers.py | 215 | # Check if user passed an enum (should be a string literal instead) | COMMENT |
| LOW | python/CuTeDSL/cutlass/cute/arch/nvvm_wrappers.py | 3698 | # Check if it's a floating-point vector type | COMMENT |
| LOW | python/CuTeDSL/cutlass/utils/mixed_input_helpers.py | 753 | # Check if scale tensor matches the TMA load 128B alignment requirement | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 2528 | # Check if all non-default arguments are provided | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/env_manager.py | 731 | # Check if the environment variable is already set, if so, return it immediately. | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/ast_preprocessor.py | 1871 | # Check if argument is Numeric, if so, call ir_value() | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/ast_preprocessor.py | 1967 | # Check if all keywords are known DSL kwargs | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/ast_preprocessor.py | 1972 | # Check if 'preprocess' is explicitly set | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/common.py | 135 | # Check if show_stacktrace is enabled via registered env manager | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/common.py | 143 | # Check if it's a DSL operation error (by name to avoid circular import issues) | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/common.py | 801 | # Check if cause is also a DSLOperationError | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/array.py | 678 | # Check if printf is already declared | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/array.py | 2456 | # Check if last element is a slice | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/array.py | 2933 | # Check if it needs type coercion for vector store | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/runtime/cuda.py | 315 | # Check if we successfully got the compute capability | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/runtime/cuda.py | 444 | # Check if a valid CUDA context already exists (e.g., created by PyTorch or | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/utils/stacktrace.py | 31 | # Check if we are at the root directory | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/utils/leaf_utils.py | 353 | # Check if this is an assignable leaf | COMMENT |
| LOW | …TeDSL/cutlass/base_dsl/tvm_ffi_builder/mlir_builder.py | 230 | # Check if align_const is a power of two | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1324 | # Check if type is int or bool (both use v_int64, bool can be converted to int) | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1363 | # Check if type is float, int, or bool (int and bool can be converted to float) | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1464 | # Check if type is opaque pointer | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1468 | # Check if type is a nullptr | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1842 | # Check if type is int or bool (both use v_int64, bool can be converted to int) | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1920 | # Check if type is ffi.Shape or ffi.Array | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 2090 | # Check if data pointer is divisible by alignment | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 2321 | # Check if type is kTVMFFIArray | COMMENT |
| 83 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | operators/test/test_utils/allocation.py | 159 | Creates a RandomIntInitializationMode from the given datatypes. Range is determined based on the bitwi | STRING |
| HIGH | operators/test/test_utils/reference_check.py | 345 | Asserts that the result tensor is similar to the reference tensor. This function automatically handles: - C | STRING |
| HIGH | operators/cutlass/operators/__init__.py | 87 | Return Operators that match the given arguments, metadata filter, and target. Args: args (RuntimeArguments | STRING |
| HIGH | operators/cutlass/operators/heuristics/base.py | 116 | Return the registered :class:`Heuristic` subclass for ``name``. Args: name (str): The name the heuristic wa | STRING |
| HIGH | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 58 | Return the compute capability nvMMH's ``gpu`` device name represents. Only SM100 devices are mapped -- nvmatmul doe | STRING |
| HIGH | …rators/cutlass/operators/heuristics/nvmatmul/ranker.py | 178 | Order ``operators`` best-first for ``args`` using nvMatmulHeuristics. See :meth:`cutlass.operators.Heuristic.ra | STRING |
| HIGH | …tors/cutlass/operators/heuristics/nvmatmul/_mapping.py | 93 | Build a query from dense-GEMM :class:`~cutlass.operators.GemmArguments`. Only plain :class:`~cutlass.operators. | STRING |
| HIGH | …tors/cutlass/operators/heuristics/nvmatmul/_mapping.py | 160 | Map a CUTLASS ``Numeric`` type to an nvMMH / cuBLAS-style precision letter. Args: numeric (type[Numeric]): | STRING |
| HIGH | …tors/cutlass/operators/heuristics/nvmatmul/_mapping.py | 219 | Map an operand's major stride to nvMMH ``'t'`` / ``'n'``. A tensor whose last dimension is contiguous is row-major | STRING |
| HIGH | …tors/cutlass/operators/heuristics/nvmatmul/_mapping.py | 300 | Reorder and prune ``operators`` best-first according to heuristic ``configs``. nvMMH provides configs describing re | STRING |
| HIGH | operators/cutlass/operators/providers/provider.py | 124 | Register an Operator class with this provider. Intended to be used as a class decorator: .. code-block | STRING |
| HIGH | operators/cutlass/operators/utils/device.py | 61 | Check whether ``result`` contains a CUDA error and raise it, otherwise return the rest. Args: result (list) | STRING |
| HIGH | operators/cutlass/operators/utils/device.py | 114 | Return the target SM from environment variable ``env_var`` or device ``device``. The environment variable takes pre | STRING |
| HIGH | operators/cutlass/operators/utils/device.py | 191 | Convert the provided stream to a ``cuda.bindings.driver.CUstream``. Args: stream (cuda.bindings.driver.CUst | STRING |
| HIGH | operators/cutlass/operators/utils/dtype.py | 94 | Convert a torch dtype to a cutlass dtype. Args: dtype (torch.dtype): The torch dtype to convert. Retur | STRING |
| HIGH | operators/cutlass/operators/utils/dtype.py | 169 | Return the number of logical elements per PyTorch storage element. The factor is the storage dtype's bit width divi | STRING |
| HIGH | operators/cutlass/operators/utils/dtype.py | 207 | Convert a dtype to a cutlass dtype. Args: dtype (torch.dtype | jnp.dtype | np.dtype): The dtype to convert | STRING |
| HIGH | operators/cutlass/operators/utils/tensor.py | 96 | Return ``tensor`` with each contiguous 8-element K group pre-shuffled. This prepares an Int4 A payload for the mixe | STRING |
| HIGH | operators/cutlass/operators/metadata/base.py | 117 | Check if the provided ``args``/``target_sm`` satisfy the properties described by this OperatorMetadata. Args: | STRING |
| HIGH | operators/cutlass/operators/arguments/operand.py | 179 | Return the number of elements expected in :attr:`scale` for an operand of ``quantized_shape``. ``quantized_shap | STRING |
| HIGH | operators/cutlass/operators/arguments/operand.py | 374 | If operand is already a ``Operand``, return it. Otherwise, wrap it in a ``DenseTensor``. This is used for convenien | STRING |
| HIGH | python/CuTeDSL/prep_editable_install.py | 45 | Download the nvidia-cutlass-dsl wheel to a temporary directory. Args: temp_dir: Temporary directory pa | STRING |
| HIGH | python/CuTeDSL/prep_editable_install.py | 96 | Extract version from wheel filename and convert to dev version. Args: wheel_path: Path to the wheel fi | STRING |
| HIGH | …n/CuTeDSL/cutlass/experimental/task_scheduling/task.py | 304 | Normalize and validate schedule_list. - Normalizes 2-element tuples to 3-element with ScheduleStageType.Loop d | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/ast_helpers.py | 498 | Executes comparison operations with a left operand and a list of comparators. Args: left: The leftmost | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/ast_helpers.py | 790 | Decomposes a joined f-string component list into a format string and dynamic arguments. This function takes a | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/array.py | 804 | Get the printf format specifier for a given MLIR type. Args: mlir_type: The MLIR type of the value Ret | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/array.py | 3511 | Load value(s) from the array. Args: idx: Index can be: - int or Int32: Single eleme | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/utils/tree_utils.py | 647 | Flatten a nested structure into a flat list of values and a tree definition. This function recursively travers | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/utils/tree_utils.py | 743 | Internal function to flatten a tree structure. This is the core implementation of tree flattening that handles | STRING |
| HIGH | python/CuTeDSL/cutlass/cutlass_dsl/cutlass.py | 2788 | Filter items based on whether corresponding iter_args are frozen dataclasses. This function filters items (whi | STRING |
| HIGH | …CuTeDSL/cute_ext/blackwell/attention/gqa_decode_opt.py | 1960 | Run and benchmark the optimized GQA flash-decode kernel on Blackwell. Args: batches/seqlen/heads_q/heads_k/ | STRING |
| HIGH | …eDSL/cute_ext/blackwell/attention/gqa_decode_simple.py | 1272 | Run and benchmark the simple GQA flash-decode kernel on Blackwell. Args: batches/seqlen/heads_q/heads_k/hea | STRING |
| HIGH | …/cute_ext/blackwell/dense_gemm/input_transform_gemm.py | 1236 | Run and benchmark the input-transform GEMM D = transform(A, B)[0] @ B. The benchmark uses the identity transform so | STRING |
| HIGH | …thon/CuTeDSL/cute_ext/blackwell/dense_gemm/glu_gemm.py | 440 | Run and benchmark the GLU GEMM D = (A @ B0) * gate(A @ B1) on Blackwell. Args: mnkl: Problem size (M, N, K, | STRING |
| HIGH | …L/cute_ext/blackwell/dense_gemm/planar_complex_gemm.py | 496 | Run and benchmark the planar complex GEMM on Blackwell. Computes D_real = A_real @ B_real - A_imag @ B_imag and | STRING |
| HIGH | …eDSL/cute_ext/blackwell/dense_gemm/mixed_input_gemm.py | 669 | Run and benchmark the mixed-input GEMM D = ((A - Zeros) * Scale) @ B. Args: mnkl: Problem size (M, N, K, L) | STRING |
| HIGH | …/blackwell/dense_gemm/sm103_dense_block_scaled_gemm.py | 804 | Run and benchmark the SM103 block-scaled dense GEMM D = (A*SFA) @ (B*SFB). Args: mnkl: Problem size (M, N, | STRING |
| HIGH | …thon/CuTeDSL/cute_ext/blackwell/dense_gemm/b2b_gemm.py | 577 | Run and benchmark the back-to-back GEMM D = (A @ B0) @ B1 on Blackwell. Args: mnkl: (M, N, K, L); K is the | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | …/operators/providers/cutedsl/evt/efc/_configuration.py | 555 | # In the @cute.kernel context, just use the normal CuTe | COMMENT |
| MEDIUM | …utedsl/gemm/implementations/sm80_tensorop_gemm_impl.py | 199 | # if atom layout's N-mode is 1, to leverage the largest coalesced | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/dsl.py | 245 | # Here we just return the first instance of a concrete subclass. | COMMENT |
| MEDIUM | python/CuTeDSL/cutlass/base_dsl/cache_helpers.py | 227 | # use temp dir to be robust against program interruptions | COMMENT |
| LOW | …n/CuTeDSL/cutlass/base_dsl/runtime/jit_arg_adapters.py | 473 | # If no registered adapter is found, just return the original element | COMMENT |
| LOW | python/CuTeDSL/cutlass/cutlass_dsl/cutlass.py | 1754 | # If fails, just return the original arg | COMMENT |
| MEDIUM | python/cutlass_library/generator.py | 12996 | # to leverage the functionality in this file without running this script via a shell prompt. | COMMENT |
| MEDIUM | include/cutlass/gemm/kernel/tile_scheduler_params.h | 917 | // the problem can leverage a basic split-K decomposition for the stream-K tiles. | COMMENT |
| MEDIUM | …cs/regular__tile__access__iterator__tensor__op_8h.html | 137 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …cs/regular__tile__access__iterator__tensor__op_8h.html | 146 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | docs/host_2tensor__fill_8h_source.html | 101 | <a href="host_2tensor__fill_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a | CODE |
| MEDIUM | docs/annotated.html | 677 | <tr id="row_0_10_1_28_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM | docs/annotated.html | 679 | <tr id="row_0_10_1_29_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 699 | <tr id="row_0_10_1_48_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 701 | <tr id="row_0_10_1_49_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 703 | <tr id="row_0_10_1_50_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 705 | <tr id="row_0_10_1_51_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 707 | <tr id="row_0_10_1_52_0_" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> < | CODE |
| MEDIUM⚡ | docs/annotated.html | 716 | <tr id="row_0_10_7_0_" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </sp | CODE |
| MEDIUM⚡ | docs/annotated.html | 718 | <tr id="row_0_10_8_0_" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </sp | CODE |
| MEDIUM⚡ | docs/annotated.html | 720 | <tr id="row_0_10_9_0_" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </sp | CODE |
| MEDIUM | …___00_01Element__ebf4714349612673e8b6609b763eeb6f.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 260 | <tr id="row_147_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 261 | <tr id="row_148_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 262 | <tr id="row_149_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 264 | <tr id="row_151_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 265 | <tr id="row_152_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 266 | <tr id="row_153_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 267 | <tr id="row_154_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 268 | <tr id="row_155_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 269 | <tr id="row_156_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class | CODE |
| MEDIUM⚡ | docs/hierarchy.html | 270 | <tr id="row_157_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><spa | CODE |
| MEDIUM | docs/device_2kernel_2tensor__elementwise_8h_source.html | 101 | <a href="device_2kernel_2tensor__elementwise_8h.html">Go to the documentation of this file.</a><div class="fragment"><di | CODE |
| MEDIUM⚡ | docs/pitch__linear__thread__map_8h.html | 145 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/pitch__linear__thread__map_8h.html | 150 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/pitch__linear__thread__map_8h.html | 157 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …01Element___00_01efd5013a2503d6567e2bf6b40c97360c.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …s_1_1transform_1_1PitchLinearWarpStripedThreadMap.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …ass_1_1transform_1_1TransposePitchLinearThreadMap.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | docs/device_2tensor__fill_8h_source.html | 101 | <a href="device_2tensor__fill_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"> | CODE |
| MEDIUM | …ass_1_1transform_1_1PitchLinearWarpRakedThreadMap.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …01Element___00_0197fef2242a3454a7d1cebe61aee28b43.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …01Element___00_01a75d2cd74e722d6ad6a3b41aabfd432d.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/regular__tile__iterator__tensor__op__sm70_8h.html | 136 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/regular__tile__iterator__tensor__op__sm70_8h.html | 145 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM⚡ | docs/regular__tile__iterator__tensor__op__sm70_8h.html | 154 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …01Element___00_01dbd6b8468d5bd787308d2f615a24d123.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …01Element___00_01f96bbeb63e6d4ce4a2551279de3a9f0e.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | docs/regular__tile__iterator__tensor__op_8h.html | 131 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | docs/regular__tile__iterator__tensor__op_8h.html | 140 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | …___00_01Element__e9a9e0f4286f652f55eb9b863b21effe.html | 115 | <tr class="memdesc:"><td class="mdescLeft"> </td><td class="mdescRight">Internal details made public to facilitate | CODE |
| MEDIUM | cutlass_compiler/cutegen/include/cutegen/rec_var.hpp | 1823 | // As written here, the encoding is not fully robust to changes over time: | COMMENT |
| MEDIUM | examples/python/CuTeDSL/dsl_tutorials/smem_allocator.py | 91 | # Note: SMEM_SIZE bytes (specified in kernel().launch(smem=...)) can be reserved for developer to utilize | COMMENT |
| LOW | …amples/python/CuTeDSL/cute/blackwell/efc/common_efc.py | 1376 | # In the @cute.kernel context, just use the normal CuTe | COMMENT |
| MEDIUM | …l/distributed/distributed_gemm_all_reduce_blackwell.py | 471 | # For most cases, in-flight depth = 4 is best as it fully utilize overlapping ability | COMMENT |
| MEDIUM | …CuTeDSL/cute/ampere/kernel/attention/hstu_attention.py | 43 | The example showcases an implementation of HSTU attention(https://arxiv.org/abs/2402.17152) within generative recommende | STRING |
| MEDIUM | …CuTeDSL/cute/ampere/kernel/dense_gemm/tensorop_gemm.py | 276 | # if atom layout's N-mode is 1, to leverage the largest coalesced | COMMENT |
| MEDIUM | …cute/blackwell_geforce/kernel/dense_gemm/dense_gemm.py | 269 | # TODO: to leverage ldmatrix.x4, when self.atom_layout[1] is 1, mma tile is ((8x16)x2) | COMMENT |
| MEDIUM | …CuTeDSL/cute_ext/blackwell/attention/gqa_decode_opt.py | 1934 | # Perf-harness dtype plumbing: Q/K/V/O share one element type in this kernel. | COMMENT |
| MEDIUM | …eDSL/cute_ext/blackwell/attention/gqa_decode_simple.py | 1247 | # Perf-harness dtype plumbing: Q/K/V/O share one element type in this kernel. | COMMENT |
| 3 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | test/python/cutlass/gemm/utils.py | 150 | CODE | |
| LOW | …t/examples/CuTeDSL/hopper/test_dense_gemm_fp8_2xacc.py | 104 | CODE | |
| LOW | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 84 | CODE | |
| LOW | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 118 | CODE | |
| LOW | test/examples/CuTeDSL/hopper/test_grouped_gemm.py | 154 | CODE | |
| LOW | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 1161 | CODE | |
| LOW | …kernels/rubin/gemm/dense_blockscaled_gemm_universal.py | 1423 | CODE | |
| LOW | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 763 | CODE | |
| LOW | …ors/cutlass/kernels/rubin/gemm/dense_gemm_universal.py | 864 | CODE | |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 2692 | CODE | |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 2925 | CODE | |
| LOW | …mm/base/blackwell_blockscaled_dense_gemm_persistent.py | 3161 | CODE | |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 1264 | CODE | |
| LOW | …tlass/kernels/rubin/gemm/base/dense_gemm_persistent.py | 1345 | CODE | |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 1495 | CODE | |
| LOW | …els/rubin/gemm/base/blackwell_dense_gemm_persistent.py | 1549 | CODE | |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 2504 | CODE | |
| LOW | …s/rubin/gemm/base/dense_blockscaled_gemm_persistent.py | 2743 | CODE | |
| LOW | python/cutlass_cppgen/op/gemm.py | 219 | CODE | |
| LOW | python/cutlass_cppgen/op/gemm.py | 632 | CODE | |
| LOW | python/cutlass_cppgen/op/gemm_grouped.py | 116 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 205 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 545 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 637 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 737 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 917 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 931 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 946 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 960 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 974 | CODE | |
| LOW | python/cutlass_cppgen/op/conv.py | 988 | CODE | |
| LOW | python/CuTeDSL/cutlass/testing.py | 309 | CODE | |
| LOW | python/CuTeDSL/cutlass/pipeline/sm100.py | 180 | CODE | |
| LOW | python/CuTeDSL/cutlass/pipeline/sm100.py | 1032 | CODE | |
| LOW | python/CuTeDSL/cutlass/experimental/cuda/tensor_map.py | 849 | CODE | |
| LOW | python/CuTeDSL/cutlass/experimental/cuda/tensor_map.py | 1113 | CODE | |
| LOW | …ass/experimental/task_scheduling/exhaustive_checker.py | 2191 | CODE | |
| LOW | …L/cutlass/experimental/task_scheduling/task_manager.py | 912 | CODE | |
| LOW | …eDSL/cutlass/experimental/task_scheduling/resources.py | 875 | CODE | |
| LOW | …CuTeDSL/cutlass/experimental/primitives/descriptors.py | 772 | CODE | |
| LOW | …CuTeDSL/cutlass/experimental/primitives/descriptors.py | 1029 | CODE | |
| LOW | …CuTeDSL/cutlass/experimental/primitives/descriptors.py | 1183 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 5589 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 7490 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 9699 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 9936 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 9976 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 10055 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 10095 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 10234 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 10278 | CODE | |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 11500 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/core.py | 2071 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/common.py | 423 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/common.py | 506 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/tcgen05/copy.py | 935 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/cpasync/copy.py | 1634 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/cpasync/copy.py | 1725 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/cpasync/copy.py | 1825 | CODE | |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/cpasync/copy.py | 1879 | CODE | |
| 117 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | operators/cutlass/operators/fusion/ir/c_types.py | 188 | # Step 1: convert the dtype | COMMENT |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 154 | # Step 1: Split the dimensions of the old layout | COMMENT |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 262 | # Step 2: Merge the splitted dimensions according to the new shape | COMMENT |
| LOW | operators/cutlass/operators/fusion/ir/layout_nodes.py | 144 | # Step 1: infer split | COMMENT |
| LOW⚡ | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 56 | # Step 1: find the nodes that have multiple parents | COMMENT |
| LOW⚡ | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 62 | # Step 2: find the lowest common ancestor (LCA) of all its parents | COMMENT |
| LOW | …cutlass/operators/fusion/passes/pass_preprocess_red.py | 51 | # Step 1: find the compute nodes with op=red | COMMENT |
| LOW | …cutlass/operators/fusion/passes/pass_preprocess_red.py | 63 | # Step 2: for each compute, merge it with the succeeding store | COMMENT |
| LOW | …ass/operators/fusion/passes/pass_layout_elimination.py | 63 | # Step 1: get the propagation direction | COMMENT |
| LOW | …ators/cutlass/operators/fusion/backend/emitter_base.py | 78 | # Step 1: emit individual node type decl | COMMENT |
| LOW | …ators/cutlass/operators/fusion/backend/emitter_base.py | 90 | # Step 2: post-processing & get callback name | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 374 | # Step 1: Compute TMEM allocations | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 385 | # Step 2: Compute SMEM bytes per stage | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 404 | # Step 3: Compute final stage counts | COMMENT |
| LOW⚡ | …rs/providers/cutedsl/gemm/implementations/scheduler.py | 542 | # Step 1: Handle persistent scheduling - map linear_idx to work_unit_id | COMMENT |
| LOW⚡ | …rs/providers/cutedsl/gemm/implementations/scheduler.py | 547 | # Step 2: Decode work_unit_id using FastDivmod objects | COMMENT |
| LOW | python/cutlass_cppgen/backend/c_types.py | 566 | # Step 1: convert the dtype | COMMENT |
| LOW | …thon/cutlass_cppgen/backend/evt/ir/layout_algorithm.py | 133 | # Step 1: Split the dimensions of the old layout | COMMENT |
| LOW | …thon/cutlass_cppgen/backend/evt/ir/layout_algorithm.py | 229 | # Step 2: Merge the splitted dimensions according to the new shape | COMMENT |
| LOW | python/cutlass_cppgen/backend/evt/ir/layout_nodes.py | 144 | # Step 1: infer split | COMMENT |
| LOW⚡ | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 56 | # Step 1: find the nodes that have multiple parents | COMMENT |
| LOW⚡ | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 62 | # Step 2: find the lowest common ancestor (LCA) of all its parents | COMMENT |
| LOW | …tlass_cppgen/backend/evt/passes/pass_preprocess_red.py | 51 | # Step 1: find the compute nodes with op=red | COMMENT |
| LOW | …tlass_cppgen/backend/evt/passes/pass_preprocess_red.py | 63 | # Step 2: for each compute, merge it with the succeeding store | COMMENT |
| LOW | …s_cppgen/backend/evt/passes/pass_layout_elimination.py | 60 | # Step 1: get the propagation direction | COMMENT |
| LOW | …hon/cutlass_cppgen/backend/evt/backend/emitter_base.py | 78 | # Step 1: emit individual node type decl | COMMENT |
| LOW | …hon/cutlass_cppgen/backend/evt/backend/emitter_base.py | 90 | # Step 2: post-processing & get callback name | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 538 | # Step 1: Capture warmup graph | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 576 | # Step 2: Capture profiling graph | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 600 | # Step 3: Instantiate executable graphs | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 610 | # Step 4: Execute warmup graph (cache warming) | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 616 | # Step 5: Profile execution using selected profiler | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 635 | # Step 6: Cleanup - Destroy graph executables | COMMENT |
| LOW | …eDSL/cutlass/utils/static_persistent_tile_scheduler.py | 622 | # Step 1: Decode current_work_linear_idx using FastDivmod objects | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_preprocessor.py | 6087 | # Step 2: capture old values. For possible-first-def attrs, capture only inside | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/utils/stacktrace.py | 149 | # Step 1: filter internal frames | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/utils/stacktrace.py | 152 | # Step 2: consolidate duplicated frames | COMMENT |
| LOW⚡ | python/cutlass_library/sm100_utils.py | 90 | #### Step 1: generate MMA instruction shapes based on levels | COMMENT |
| LOW | python/cutlass_library/sm100_utils.py | 56 | #### Step 0: define levels | COMMENT |
| LOW⚡ | python/cutlass_library/sm90_utils.py | 106 | #### Step 1: generate MMA instruction shapes based on levels | COMMENT |
| LOW⚡ | python/cutlass_library/sm90_utils.py | 406 | #### Step 3: map tile description to valid schedules | COMMENT |
| LOW | python/cutlass_library/sm90_utils.py | 71 | #### Step 0: define levels | COMMENT |
| LOW | python/cutlass_library/sm90_utils.py | 279 | #### Step 2: generate tile descriptions from math instruction shapes | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 576 | // Step 1: reduce fragment input (Src) into tRS_rSrc | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 587 | // Step 2: copy the partial results in tRS_rSrc to sRows in shared memory | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 599 | // Step 3: copy the partial results in sRows to tSR_sRow for reduction | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 613 | // Step 4: atomically reduce the results to global memory | COMMENT |
| LOW | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 341 | /// Step 0: Step 1: | COMMENT |
| LOW⚡ | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 559 | /// Step 0: Step 1: Step 2: Step 3: | COMMENT |
| LOW⚡ | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 569 | /// Step 4: Step 5: Step 6: Step 7: | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 190 | // Step 1: layout_b((0,2)) = 0*3 + 2*1 = 2 | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 191 | // Step 2: calculated_offset = off + layout_b(coord) = 0 + 2 = 2 | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 192 | // Step 3: layout_a(calculated_offset) = layout_a(2) = 2*1 = 2 | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 107 | ### Step 1: What resources are needed, and why | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 130 | ### Step 2: How to define the work on each resource (producer vs consumer) | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 190 | ### Step 3: How a value flows between resources (captured values) | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 218 | ### Step 4: How to define the schedule | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 277 | ### Step 5: How to define the dependency graph | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 297 | ### Step 6: How to define the task | COMMENT |
| LOW | …eduling/blackwell/tutorial/01_copy_basics_ts/README.md | 323 | ### Step 7: TaskManager and the fixed lifecycle | COMMENT |
| 29 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | operators/test/test_utils/__init__.py | 33 | __all__ = [ | CODE |
| LOW⚡ | …rators/test/integration/test_preferred_cluster_gemm.py | 77 | logger = logging.getLogger(__name__) | CODE |
| LOW | operators/test/integration/test_gemm_dynamic.py | 48 | logger = logging.getLogger(__name__) | CODE |
| LOW | operators/test/integration/test_gemm.py | 47 | logger = logging.getLogger(__name__) | CODE |
| LOW | operators/test/integration/test_tgv_gemm.py | 47 | logger = logging.getLogger(__name__) | CODE |
| LOW⚡ | operators/test/integration/test_mixed_input_gemm.py | 58 | logger = logging.getLogger(__name__) | CODE |
| LOW | operators/test/integration/test_gemm_jax.py | 68 | logger = logging.getLogger(__name__) | CODE |
| LOW | operators/cutlass/operators/__init__.py | 161 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/fusion/__init__.py | 122 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/fusion/frontend/__init__.py | 38 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/fusion/ir/__init__.py | 55 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/fusion/passes/__init__.py | 48 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/fusion/backend/__init__.py | 38 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/heuristics/__init__.py | 65 | __all__ = [ | CODE |
| LOW⚡ | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 148 | def _set_backend_property(self, backend_property: Any, value: int) -> None: | CODE |
| LOW⚡ | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 158 | def set_cta_div_n(self, div_n: int) -> None: | CODE |
| LOW⚡ | …ors/cutlass/operators/heuristics/nvmatmul/_provider.py | 164 | def set_cta_div_m(self, div_m: int) -> None: | CODE |
| LOW | …tors/cutlass/operators/heuristics/nvmatmul/__init__.py | 35 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/providers/__init__.py | 106 | __all__ = [ | CODE |
| LOW | …rators/providers/cutedsl/integration_utils/__init__.py | 37 | __all__ = [ | CODE |
| LOW | …utlass/operators/providers/cutedsl/evt/efc/__init__.py | 82 | __all__ = [ | CODE |
| LOW | …utlass/operators/providers/cutedsl/evt/efc/_helpers.py | 78 | logger = logging.getLogger(__name__) | CODE |
| LOW | …providers/cutedsl/evt/efc/dense_gemm/sm100/__init__.py | 41 | __all__ = ["DenseGemmEFC"] | CODE |
| LOW | operators/cutlass/operators/utils/__init__.py | 31 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/metadata/__init__.py | 54 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/metadata/design/__init__.py | 38 | __all__ = [ | CODE |
| LOW | …rators/cutlass/operators/metadata/operands/__init__.py | 36 | __all__ = [ | CODE |
| LOW | operators/cutlass/operators/arguments/__init__.py | 54 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/prep_editable_install.py | 32 | logger = logging.getLogger(__name__) | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/__init__.py | 22 | __all__ = ["arith", "vector", "lru_cache_ir", "dsl_user_op"] | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 108 | def set_verify_trace(enabled: bool) -> None: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 137 | def _set_enable_frame_filtering(enable: bool) -> None: | CODE |
| LOW | python/CuTeDSL/_mlir_helpers/op.py | 189 | def _set_include_lib_frame(enable: bool) -> None: | CODE |
| LOW⚡ | python/CuTeDSL/_mlir_helpers/math.py | 2221 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/block.py | 249 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/pipeline/__init__.py | 58 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/experimental/__init__.py | 61 | __all__ = ["cuda", "primitives", "task_scheduling"] | CODE |
| LOW | python/CuTeDSL/cutlass/experimental/cuda/tensor_map.py | 1356 | __all__ = [ | CODE |
| LOW | …n/CuTeDSL/cutlass/experimental/task_scheduling/task.py | 1009 | def _set_max_register(self) -> None: | CODE |
| LOW | …TeDSL/cutlass/experimental/task_scheduling/__init__.py | 44 | __all__ = [ | CODE |
| LOW | …L/cutlass/experimental/task_scheduling/task_manager.py | 5056 | def _setup_resources_and_tasks_impl(self) -> None: | CODE |
| LOW | …eDSL/cutlass/experimental/task_scheduling/resources.py | 170 | def update_from(self, other: _BaseWorkTileInfo) -> None: | CODE |
| LOW | …eDSL/cutlass/experimental/task_scheduling/resources.py | 3073 | def set_consumer_var(self, name: cutlass.Constexpr[str], value: object) -> None: | CODE |
| LOW | …on/CuTeDSL/cutlass/experimental/primitives/__init__.py | 42 | __all__ = [*_descriptors_all, *_nvvm_wrapper_all, *_hybrid_ops_all] | CODE |
| LOW⚡ | …CuTeDSL/cutlass/experimental/primitives/descriptors.py | 1326 | __all__ = [ | CODE |
| LOW | …/CuTeDSL/cutlass/experimental/primitives/hybrid_ops.py | 55 | __all__ = [ | CODE |
| LOW | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 11575 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/tuple.py | 526 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/__init__.py | 271 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/core.py | 91 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/tensor.py | 100 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/math.py | 140 | __all__ = [*_base_math.__all__, "absf"] | CODE |
| LOW | python/CuTeDSL/cutlass/cute/typing.py | 970 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/viz.py | 28 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/ffi.py | 60 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/__init__.py | 25 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/common.py | 25 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/helpers.py | 47 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/tcgen05/__init__.py | 37 | __all__ = [ | CODE |
| LOW | python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/__init__.py | 36 | __all__ = [ | CODE |
| 66 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | operators/cutlass/operators/fusion/ir/c_types.py | 188 | # Step 1: convert the dtype | COMMENT |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 154 | # Step 1: Split the dimensions of the old layout | COMMENT |
| LOW | …rators/cutlass/operators/fusion/ir/layout_algorithm.py | 262 | # Step 2: Merge the splitted dimensions according to the new shape | COMMENT |
| LOW | operators/cutlass/operators/fusion/ir/layout_nodes.py | 144 | # Step 1: infer split | COMMENT |
| LOW⚡ | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 56 | # Step 1: find the nodes that have multiple parents | COMMENT |
| LOW⚡ | …ors/cutlass/operators/fusion/passes/pass_dag_2_tree.py | 62 | # Step 2: find the lowest common ancestor (LCA) of all its parents | COMMENT |
| LOW | …cutlass/operators/fusion/passes/pass_preprocess_red.py | 51 | # Step 1: find the compute nodes with op=red | COMMENT |
| LOW | …cutlass/operators/fusion/passes/pass_preprocess_red.py | 63 | # Step 2: for each compute, merge it with the succeeding store | COMMENT |
| LOW | …ass/operators/fusion/passes/pass_layout_elimination.py | 63 | # Step 1: get the propagation direction | COMMENT |
| LOW | …ators/cutlass/operators/fusion/backend/emitter_base.py | 78 | # Step 1: emit individual node type decl | COMMENT |
| LOW | …ators/cutlass/operators/fusion/backend/emitter_base.py | 90 | # Step 2: post-processing & get callback name | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 374 | # Step 1: Compute TMEM allocations | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 385 | # Step 2: Compute SMEM bytes per stage | COMMENT |
| LOW | …cutedsl/gemm/implementations/sm100_mixed_input_impl.py | 404 | # Step 3: Compute final stage counts | COMMENT |
| LOW⚡ | …rs/providers/cutedsl/gemm/implementations/scheduler.py | 542 | # Step 1: Handle persistent scheduling - map linear_idx to work_unit_id | COMMENT |
| LOW⚡ | …rs/providers/cutedsl/gemm/implementations/scheduler.py | 547 | # Step 2: Decode work_unit_id using FastDivmod objects | COMMENT |
| LOW | python/cutlass_cppgen/backend/c_types.py | 566 | # Step 1: convert the dtype | COMMENT |
| LOW | …thon/cutlass_cppgen/backend/evt/ir/layout_algorithm.py | 133 | # Step 1: Split the dimensions of the old layout | COMMENT |
| LOW | …thon/cutlass_cppgen/backend/evt/ir/layout_algorithm.py | 229 | # Step 2: Merge the splitted dimensions according to the new shape | COMMENT |
| LOW | python/cutlass_cppgen/backend/evt/ir/layout_nodes.py | 144 | # Step 1: infer split | COMMENT |
| LOW⚡ | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 56 | # Step 1: find the nodes that have multiple parents | COMMENT |
| LOW⚡ | …n/cutlass_cppgen/backend/evt/passes/pass_dag_2_tree.py | 62 | # Step 2: find the lowest common ancestor (LCA) of all its parents | COMMENT |
| LOW | …tlass_cppgen/backend/evt/passes/pass_preprocess_red.py | 51 | # Step 1: find the compute nodes with op=red | COMMENT |
| LOW | …tlass_cppgen/backend/evt/passes/pass_preprocess_red.py | 63 | # Step 2: for each compute, merge it with the succeeding store | COMMENT |
| LOW | …s_cppgen/backend/evt/passes/pass_layout_elimination.py | 60 | # Step 1: get the propagation direction | COMMENT |
| LOW | …hon/cutlass_cppgen/backend/evt/backend/emitter_base.py | 78 | # Step 1: emit individual node type decl | COMMENT |
| LOW | …hon/cutlass_cppgen/backend/evt/backend/emitter_base.py | 90 | # Step 2: post-processing & get callback name | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 538 | # Step 1: Capture warmup graph | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 576 | # Step 2: Capture profiling graph | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 600 | # Step 3: Instantiate executable graphs | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 610 | # Step 4: Execute warmup graph (cache warming) | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 616 | # Step 5: Profile execution using selected profiler | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/testing.py | 635 | # Step 6: Cleanup - Destroy graph executables | COMMENT |
| LOW | …eDSL/cutlass/utils/static_persistent_tile_scheduler.py | 622 | # Step 1: Decode current_work_linear_idx using FastDivmod objects | COMMENT |
| LOW | python/CuTeDSL/cutlass/base_dsl/pyir_preprocessor.py | 6087 | # Step 2: capture old values. For possible-first-def attrs, capture only inside | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/utils/stacktrace.py | 149 | # Step 1: filter internal frames | COMMENT |
| LOW⚡ | python/CuTeDSL/cutlass/base_dsl/utils/stacktrace.py | 152 | # Step 2: consolidate duplicated frames | COMMENT |
| LOW | …SL/cutlass/base_dsl/tvm_ffi_builder/tvm_ffi_builder.py | 1746 | # otherwise, it appears more than once, we need to check if the value matches | COMMENT |
| LOW⚡ | python/cutlass_library/sm100_utils.py | 90 | #### Step 1: generate MMA instruction shapes based on levels | COMMENT |
| LOW | python/cutlass_library/sm100_utils.py | 56 | #### Step 0: define levels | COMMENT |
| LOW⚡ | python/cutlass_library/sm90_utils.py | 106 | #### Step 1: generate MMA instruction shapes based on levels | COMMENT |
| LOW⚡ | python/cutlass_library/sm90_utils.py | 406 | #### Step 3: map tile description to valid schedules | COMMENT |
| LOW | python/cutlass_library/sm90_utils.py | 71 | #### Step 0: define levels | COMMENT |
| LOW | python/cutlass_library/sm90_utils.py | 279 | #### Step 2: generate tile descriptions from math instruction shapes | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 576 | // Step 1: reduce fragment input (Src) into tRS_rSrc | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 587 | // Step 2: copy the partial results in tRS_rSrc to sRows in shared memory | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 599 | // Step 3: copy the partial results in sRows to tSR_sRow for reduction | COMMENT |
| LOW | …/cutlass/epilogue/threadblock/fusion/visitor_store.hpp | 613 | // Step 4: atomically reduce the results to global memory | COMMENT |
| LOW | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 341 | /// Step 0: Step 1: | COMMENT |
| LOW⚡ | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 559 | /// Step 0: Step 1: Step 2: Step 3: | COMMENT |
| LOW⚡ | …/cutlass/transform/collective/sm90_wgmma_transpose.hpp | 569 | /// Step 4: Step 5: Step 6: Step 7: | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 190 | // Step 1: layout_b((0,2)) = 0*3 + 2*1 = 2 | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 191 | // Step 2: calculated_offset = off + layout_b(coord) = 0 + 2 = 2 | COMMENT |
| LOW⚡ | …lass_compiler/cutegen/test/cg_composed_layout_test.cpp | 192 | // Step 3: layout_a(calculated_offset) = layout_a(2) = 2*1 = 2 | COMMENT |
| LOW⚡ | …s/python/CuTeDSL/experimental/primitives/redux_sync.py | 526 | # Step 1: load BF16 element and widen to FP32 | COMMENT |
| LOW⚡ | …s/python/CuTeDSL/experimental/primitives/redux_sync.py | 529 | # Step 2: warp abs-max — single ``redux.sync.max.abs.f32`` | COMMENT |
| LOW⚡ | …s/python/CuTeDSL/experimental/primitives/redux_sync.py | 533 | # Step 3: guard against zero-valued rows; compute inverse scale | COMMENT |
| LOW⚡ | …s/python/CuTeDSL/experimental/primitives/redux_sync.py | 539 | # Step 4: quantize — F2FP.SATFINITE saturates to ±448 automatically | COMMENT |
| LOW⚡ | …s/python/CuTeDSL/experimental/primitives/redux_sync.py | 542 | # Step 5: lane 0 writes the per-row dequantization scale | COMMENT |
| LOW⚡ | …ython/CuTeDSL/experimental/primitives/tma/tma_store.py | 122 | # Step 1: TMA load src tile into smem_in. | COMMENT |
| 19 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | …CuTeDSL/dsl_tutorials/programmatic_dependent_launch.py | 171 | # In this example, we add a loop (10 times) for all the copy and compute operations in the following code | COMMENT |
| HIGH | …CuTeDSL/dsl_tutorials/programmatic_dependent_launch.py | 183 | # In this example, the second kernel's second operand ``gB`` has no dependencies, its loading can overlap | COMMENT |
| HIGH | …scaled_gemm/dense_blockscaled_mxfp6_gemm_persistent.py | 520 | # In this example, we have a fixed strucuture that we follow: | COMMENT |
| HIGH | …CuTeDSL/cute/blackwell/tutorial/tutorial_tma/tma_v1.py | 260 | # In this example `copy_elems = 1` and `thread_layout` has shape (T, V) = (num_trans_threads, 1), | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 632 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 637 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 642 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 644 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 649 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 651 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 720 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 721 | CODE | |
| MEDIUM | python/CuTeDSL/_mlir_helpers/op.py | 723 | CODE | |
| MEDIUM | python/cutlass_library/rank_k_operation.py | 84 | CODE | |
| MEDIUM | python/cutlass_library/rank_2k_operation.py | 86 | CODE | |
| MEDIUM | python/cutlass_library/symm_operation.py | 86 | CODE | |
| MEDIUM | python/cutlass_library/trmm_operation.py | 83 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | python/cutlass_cppgen/backend/conv2d_operation.py | 257 | if (args.tensor_c_numel >= 0 && args.tensor_c_numel == tc_C.at(3) && args.tensor_c_numel < size_C) { | CODE |
| HIGH | …uTeDSL/cutlass/experimental/primitives/nvvm_wrapper.py | 4044 | non-conforming values undefined). The wrapper exposes the full option set this | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/dsl.py | 1953 | process stderr and the lookup returns null, which the upstream wrapper | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/array.py | 3023 | Elements where the mask is False use the default value (or are undefined). | STRING |
| HIGH | python/CuTeDSL/cutlass/base_dsl/array.py | 3812 | Elements where mask is False use the default value (or are undefined). | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | python/cutlass_library/emit_kernel_listing.py | 41 | # Example usage: | COMMENT |
| LOW | cutlass_compiler/scripts/run_integration_host.sh | 41 | # Usage: | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | …buted/distributed_gemm_all_reduce_lamport_blackwell.py | 2647 | # Test scaffold: slot[0] enters as the Lamport -0.0 sentinel | COMMENT |