Community maintained hardware plugin for vLLM on Huawei Ascend
This report summarizes scanner evidence for vllm-project/vllm-ascend, a Python project with 2,949 GitHub stars. SynthScan examined 1,073,299 lines of code across 4030 source files, recording 12998 pattern matches distributed across 24 syntactic categories. The legacy adjusted pattern signal is 15.7 (Elevated 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? 10438 finding(s).
Are imports declared and package/API names verified? 734 finding(s).
Are wrappers, guards, and repeated structures necessary? 9800 finding(s).
Which changes need focused human review? 12998 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 12998 distinct pattern matches across 24 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 |
|---|---|---|---|---|
| LOW | collect_env.py | 110 | def run_and_parse_first_match(run_lambda, command, regex): | CODE |
| LOW | collect_env.py | 121 | def run_and_return_first_line(run_lambda, command): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 149 | def _expected_fd_compute_cores(case: FDCase, aic_num=ARCH35_MAX_AIC): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 378 | def test_fd_generalized_accuracy(case): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 386 | def test_fd_automatic_fallback_accuracy(case): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 397 | def test_zero_valid_blocks_should_produce_zero_output(): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 408 | def test_cross_base_task_partial_write_completion_is_stable(): | CODE |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 419 | def test_case_matrix_covers_fd_boundaries(): | CODE |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 25 | def _register_sparse_attention_score_op(): | CODE |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 69 | def generate_block_index_with_causal( | CODE |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 31 | def _register_sparse_attention_score_op(): | CODE |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 75 | def generate_block_index_with_causal( | CODE |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 40 | def _register_sparse_attention_score_op(): | CODE |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 84 | def generate_block_index_with_causal( | CODE |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 1031 | def broadcast_mx_scale_d_axis(self, tensor): | CODE |
| LOW | …lightning_indexer_v2/tests/pytest/collect_perf_data.py | 37 | def collect_and_save_incremental(prof_folder, test_result_path): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 27 | def cal_relative_diff_np_isclose(real_data, expect_data, type_str="fp16"): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 93 | def display_output_np_isclose(real_data, expect_data, start, end, expect_fp32_data=None): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 676 | def _gather_return_values_by_index(topk_value, result_indices, params, output_idx_offset): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 716 | def check_result_return_value( | CODE |
| LOW | …tning_indexer_v2/tests/pytest/test_qliv2_test_utils.py | 26 | def test_case_selector_preserves_requested_order(tmp_path): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/test_qliv2_test_utils.py | 40 | def test_case_selector_rejects_ambiguous_or_invalid_selection(tmp_path): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/test_qliv2_test_utils.py | 48 | def test_result_writer_uses_readable_name_and_migrates_legacy_result(tmp_path): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/test_qliv2_test_utils.py | 76 | def test_comparison_failure_raises_serializable_assertion(): | CODE |
| LOW | …tning_indexer_v2/tests/pytest/test_qliv2_test_utils.py | 88 | def test_normalize_legacy_params_adds_weight_dtype(): | CODE |
| LOW⚡ | …ightning_indexer_v2/tests/pytest/batch/replace_path.py | 16 | def replace_paths_in_test_file(test_file_path, path): | CODE |
| LOW | …htning_indexer_v2/tests/assets/impl/metadata_inputs.py | 72 | def generate_quant_lightning_indexer_metadata_inputs( | CODE |
| LOW | …ning_indexer_v2/tests/assets/impl/batch_consistency.py | 101 | def validate_disjoint_relations(self, relations): | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 461 | def normalize_compare_attributes(compare_context): | CODE |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 57 | def convert_quant_lightning_indexer_metadata( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 59 | def quant_lightning_indexer_metadata_meta( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 81 | def quant_lightning_indexer_meta( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 137 | def quant_lightning_indexer_candidate_meta( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 202 | def quant_lightning_indexer_metadata( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 256 | def quant_lightning_indexer_metadata_fallback( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 302 | def quant_lightning_indexer_candidate( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 443 | def _qli_newapi_build_metadata( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 508 | def quant_lightning_indexer_candidate_source( | CODE |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 596 | def quant_lightning_indexer_candidate_consumer( | CODE |
| LOW | csrc/cmake/scripts/parse_changed_files.py | 87 | def get_test_example_ops_test_options(self, f: Path) -> list[str]: | CODE |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 307 | def _set_all_options_to_opdescs(op_descs, soc_ver_compile_options): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_impl_build.py | 82 | def add_dtype_fmt_option_single(x, x_n, is_ref: bool = False): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_impl_build.py | 124 | def get_shortsoc_compile_option(compile_option_list: list, shortsoc:str): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_ops_config.py | 36 | def get_specified_suffix_file(root_dir, suffix): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_ops_config.py | 58 | def get_input_or_output_config(in_or_out): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_ops_config.py | 79 | def get_inputs_or_outputs_config(inputs_or_outputs): | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_ops_config.py | 253 | def generate_operator_cfg_file(json_files, binary_info_config, soc, out_dir, gen_json_status): | CODE |
| LOW | csrc/cmake/scripts/examples/get_opapi_abs_path.py | 33 | def get_environ_custom_lib_paths() -> list[Path]: | CODE |
| LOW | csrc/cmake/scripts/examples/get_opapi_abs_path.py | 53 | def get_default_custom_lib_paths() -> list[Path]: | CODE |
| LOW | csrc/cmake/scripts/examples/get_opapi_abs_path.py | 90 | def get_default_builtin_lib_paths() -> list[Path]: | CODE |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 27 | def check_if_new_tiling_file_path_existed(ori_file: Path) -> Path: | CODE |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 104 | def convert_template_tilingkey(ori_file: Path): | CODE |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 145 | def convert_to_old_tiling_struct_style(redirected_file_path): | CODE |
| LOW | csrc/cmake/scripts/utest/gen_coverage.py | 36 | def get_exclude_paths_from_yaml(yaml_path: Path): | CODE |
| LOW | csrc/cmake/scripts/utest/gen_coverage.py | 68 | def init_filter_str_from_yaml(self, source_dir: Path, yaml_path: Path): | CODE |
| LOW | csrc/scripts/build_cache.py | 699 | def _candidate_cann_metadata_files() -> list[Path]: | CODE |
| LOW | csrc/scripts/build_cache.py | 731 | def _hash_compiler_environment( | CODE |
| LOW | csrc/scripts/build_cache.py | 860 | def _expand_internal_symlink_targets( | CODE |
| LOW | csrc/scripts/build_cache.py | 1281 | def _publish_action_state_path(state_dir: Path, action_identity: str) -> Path: | CODE |
| LOW | csrc/scripts/build_cache.py | 1293 | def _publish_action_artifacts( | CODE |
| 7777 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | .gitleaks.toml | 1 | # ============================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 6 | # ============================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 17 | # ============================================================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 19 | # ============================================================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 32 | # ============================================================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 35 | # ============================================================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 47 | # ============================================================================== | COMMENT |
| MEDIUM | .gitleaks.toml | 50 | # ============================================================================== | COMMENT |
| MEDIUM⚡ | .gitleaks.toml | 79 | # ============================================================================== | COMMENT |
| MEDIUM⚡ | .gitleaks.toml | 81 | # ============================================================================== | COMMENT |
| MEDIUM⚡ | .gitleaks.toml | 88 | # ============================================================================== | COMMENT |
| MEDIUM⚡ | .gitleaks.toml | 90 | # ============================================================================== | COMMENT |
| MEDIUM | csrc/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/build.sh | 2 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/build.sh | 10 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/grouped_matmul_swiglu_quant/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/grouped_matmul_swiglu_quant/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …gmm/grouped_matmul_swiglu_quant/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …gmm/grouped_matmul_swiglu_quant/op_host/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/grouped_matmul_swiglu_quant_v2/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/grouped_matmul_swiglu_quant_v2/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/grouped_matmul_swiglu_quant_v2/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/grouped_matmul_swiglu_quant_v2/op_host/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tmul_swiglu_quant_weight_nz_tensor_list/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tmul_swiglu_quant_weight_nz_tensor_list/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/gmm_dequant_situ_quant/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/gmm/gmm_dequant_situ_quant/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …sed_scatter_copy_sparse_flash_attention/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …sed_scatter_copy_sparse_flash_attention/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/kda_layout_swap12/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/kda_layout_swap12/CMakeLists.txt | 6 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/kda_layout_swap12/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/kda_layout_swap12/op_host/CMakeLists.txt | 6 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/kv_quant_sparse_attn_sharedkv_metadata/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/kv_quant_sparse_attn_sharedkv_metadata/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t_sparse_attn_sharedkv_metadata/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …t_sparse_attn_sharedkv_metadata/op_host/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …_attn_sharedkv_metadata/op_kernel_aicpu/CMakeLists.txt | 1 | # --------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …_attn_sharedkv_metadata/op_kernel_aicpu/CMakeLists.txt | 9 | # --------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/sparse_attn_sharedkv/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/sparse_attn_sharedkv/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/sparse_attn_sharedkv/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/sparse_attn_sharedkv/op_host/CMakeLists.txt | 9 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/k2q_csr/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | csrc/attention/k2q_csr/CMakeLists.txt | 4 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tion/k2q_csr/k2q_csr_row_prefix/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tion/k2q_csr/k2q_csr_row_prefix/op_host/CMakeLists.txt | 3 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …ion/k2q_csr/k2q_csr_tile_prefix/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …ion/k2q_csr/k2q_csr_tile_prefix/op_host/CMakeLists.txt | 3 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/k2q_csr/k2q_csr_meta/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/k2q_csr/k2q_csr_meta/op_host/CMakeLists.txt | 3 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/k2q_csr/k2q_csr_hist/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …/attention/k2q_csr/k2q_csr_hist/op_host/CMakeLists.txt | 3 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tention/k2q_csr/k2q_csr_scatter/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| MEDIUM | …tention/k2q_csr/k2q_csr_scatter/op_host/CMakeLists.txt | 3 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| 1032 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | codecov.yml | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | setup.py | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | format.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/collect_user_first_contribution.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/shellcheck.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/aisbench.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/vllm_bench.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_repo.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_logger.sh | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_python_src_init.py | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/png-lint.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_symbolic_meta.py | 17 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/mooncake_installer.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/mooncake_installer.sh | 97 | # Define a function to handle the git clone operation | COMMENT |
| MEDIUM | tools/format_contributors.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/actionlint.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_boolean_context_manager.py | 17 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_forbidden_imports.py | 17 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/mypy.sh | 18 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/check_long_functions.py | 17 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/version_compat.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/coordinator.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/runner.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/config.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/auto_bisect.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/git_ops.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/worker_agent.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/good_table.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/__init__.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/build_manager.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/report.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/state.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tools/bisect/verdict.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/test_utils.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/conftest.py | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/test_batch_invariant.py | 14 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/test_envs.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/test_ascend_config.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/base.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM⚡ | tests/ut/test_logger.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/_310p/test_model_runner_310p.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/_310p/test_kv_block_zeroer_310p.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/_310p/test_block_table_310p.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/_310p/ops/test_gdn_310.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/_310p/ops/test_rotary_embedding_310.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/attention/test_sfa_o_proj_weight_switch.py | 13 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/attention/test_mla_v1.py | 2569 | # Create a mock metadata where prefill_metadata.chunked_context.seq_tot is an empty list, so that iters == 0 | COMMENT |
| MEDIUM | tests/ut/attention/utils.py | 318 | # Create a realistic slot mapping that corresponds to the block table | COMMENT |
| MEDIUM | tests/ut/core/test_short_request_first_scheduler.py | 3 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …t/core/test_recompute_scheduler_short_request_first.py | 3 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/core/test_profiling_chunk.py | 3 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …/ut/core/test_short_request_first_scheduler_install.py | 3 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …t/distributed/kv_transfer/test_kv_transfer_failures.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …ted/kv_transfer/kv_p2p/test_remote_decode_lifecycle.py | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …ed/kv_transfer/kv_p2p/test_remote_prefill_lifecycle.py | 16 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …/ut/distributed/weight_transfer/test_npu_ipc_engine.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …s/ut/distributed/weight_transfer/test_packed_tensor.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …ted/weight_transfer/test_e2e_weight_transfer_oracle.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | tests/ut/distributed/ascend_store/test_backend.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| MEDIUM | …sts/ut/distributed/ascend_store/test_pool_scheduler.py | 15 | # This file is a part of the vllm-ascend project. | COMMENT |
| 494 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | codecov.yml | 1 | # | COMMENT |
| LOW | collect_env.py | 1 | # | COMMENT |
| LOW | setup.py | 1 | # | COMMENT |
| LOW | .gitleaks.toml | 1 | # ============================================== | COMMENT |
| LOW | .gitleaks.toml | 21 | # [extend] | COMMENT |
| LOW | .gitleaks.toml | 41 | # entropy = 3.5 | COMMENT |
| LOW | .gitleaks.toml | 81 | # ============================================================================== | COMMENT |
| LOW | .readthedocs.yaml | 1 | # Read the Docs configuration file (English + Chinese) | COMMENT |
| LOW | format.sh | 1 | #!/usr/bin/env bash | COMMENT |
| LOW | csrc/utils.h | 1 | #pragma once | COMMENT |
| LOW | csrc/build.sh | 1 | #!/bin/bash | COMMENT |
| LOW | csrc/build.sh | 741 | fi | COMMENT |
| LOW | csrc/build.sh | 761 | COMMENT | |
| LOW | csrc/torch_binding.cpp | 21 | #include <torch/torch.h> | COMMENT |
| LOW | csrc/torch_binding.cpp | 41 | #include "gmm/grouped_matmul_swiglu_quant_v2/grouped_matmul_swiglu_quant_v2_torch_adpt.h" | COMMENT |
| LOW | csrc/torch_binding.cpp | 61 | #include "moe/dequant_situ_quant/dequant_situ_quant_torch_adpt.h" | COMMENT |
| LOW | csrc/gmm/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| LOW | csrc/gmm/grouped_matmul_swiglu_quant/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| LOW | …t/op_host/op_api/aclnn_grouped_matmul_swiglu_quant.cpp | 1 | /** | COMMENT |
| LOW | …t/op_host/op_api/aclnn_grouped_matmul_swiglu_quant.cpp | 541 | #endif | COMMENT |
| LOW | …lu_quant/op_kernel/grouped_matmul_swiglu_quant_utils.h | 21 | COMMENT | |
| LOW | …nt_v2/op_kernel/grouped_matmul_swiglu_quant_v2_utils.h | 21 | defined(DT_INT4) && ORIG_DTYPE_WEIGHT == DT_INT4 | COMMENT |
| LOW | …_grouped_matmul_swiglu_quant_weight_nz_tensor_list.cpp | 1 | /* | COMMENT |
| LOW | …_dequant_situ_quant/op_kernel/gmm_dequant_situ_quant.h | 501 | GmmMt mm_; | COMMENT |
| LOW | …/kv_quant_sparse_attn_sharedkv_metadata/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| LOW | …t_sparse_attn_sharedkv_metadata/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| LOW | …p_api/aclnn_kv_quant_sparse_attn_sharedkv_metadata.cpp | 21 | #include "aclnn_kernels/common/op_error_check.h" | COMMENT |
| LOW | …k2q_csr/k2q_csr_common/op_host/k2q_csr_tiling_common.h | 1 | /** | COMMENT |
| LOW | …rc/attention/kv_compress_epilog/op_host/CMakeLists.txt | 1 | # This program is free software, you can redistribute it and/or modify it. | COMMENT |
| LOW | …rc/attention/kv_compress_epilog/op_host/CMakeLists.txt | 21 | # target_sources(op_host_aclnn PRIVATE | COMMENT |
| LOW | …ttention/compressor/op_host/arch35/compressor_tiling.h | 21 | #include <vector> | COMMENT |
| LOW | …ttention/compressor/op_host/arch32/compressor_tiling.h | 21 | #include <vector> | COMMENT |
| LOW | …ssor/op_kernel/arch35/compressor_template_tiling_key.h | 21 | #define ASCENDC_TPL_1_BW 1 // 每个参数占用1个bit位 | COMMENT |
| LOW | csrc/attention/compressor/op_kernel/arch35/vf/vf_mul.h | 261 | #endif | COMMENT |
| LOW | …ssor/op_kernel/arch32/compressor_template_tiling_key.h | 21 | #define ASCENDC_TPL_1_BW 1 // 每个参数占用1个bit位 | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1781 | dual_golden_l1norm(npu_out_cpu, cpu_out_fp32, cpu_out, "NPU vs dual-golden") | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1801 | COMMENT | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1821 | # query_fp8, key_fp8, value_fp8, q_scales, k_scales, v_scales = ( | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1841 | # cpu_out = cpu_sparse_attention_score_fp8( | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1861 | # block_size=block_size, | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 1881 | # # --- Three blocks --- | COMMENT |
| LOW | …ion_score/op_kernel/sparse_attention_score_tilingkey.h | 21 | #define SASA_FP8_D128_BF16_TILING 10004 | COMMENT |
| LOW | …e_attention_score/op_kernel/attn_infra/gemm/helper.hpp | 121 | }; | COMMENT |
| LOW | …on_score/op_kernel/attn_infra/gemm/dispatch_policy.hpp | 61 | #endif // GEMM_DISPATCH_POLICY_HPP | COMMENT |
| LOW | …re/op_kernel/attn_infra/gemm/tile_common/tile_copy.hpp | 21 | #include "../../../attn_infra/gemm/tile_common/copy_l1_to_l0a_a5.hpp" | COMMENT |
| LOW | …nfra/gemm/block/block_mmad_qk_arch35_ABf16_C_to_UB.hpp | 21 | #include "../../../attn_infra/base_defs.hpp" | COMMENT |
| LOW | …n_score/op_kernel/attn_infra/gemm/block/block_mmad.hpp | 61 | #if (__CCE_AICORE__ == 310) | COMMENT |
| LOW | …nfra/gemm/block/block_mmad_pv_arch35_ABf16_C_to_UB.hpp | 21 | #include "../../../attn_infra/base_defs.hpp" | COMMENT |
| LOW | …op_kernel/attn_infra/epilogue/block/block_epilogue.hpp | 21 | > | COMMENT |
| LOW | …anage/op_kernel/fused_lightning_indexer_manage_union.h | 1121 | COMMENT | |
| LOW | …nage/fused_quant_lightning_indexer_manage_torch_adpt.h | 21 | // index_weights - bf16 [T, H] | COMMENT |
| LOW | …ernel/fused_quant_lightning_indexer_manage_workspace.h | 1 | #ifndef FUSED_QUANT_LIGHTNING_INDEXER_MANAGE_WORKSPACE_H | COMMENT |
| LOW | …el/fused_quant_lightning_indexer_manage_service_cube.h | 61 | static constexpr uint32_t M_MTE1_EVENT = EVENT_ID3; | COMMENT |
| LOW | …/fused_quant_lightning_indexer_manage_service_vector.h | 181 | __aicore__ inline void LIVectorMtpC8<LIT>::InitBuffers(TPipe *pipe) | COMMENT |
| LOW | …nel/arch35/fused_sparse_attention_overlap_kernel_mla.h | 21 | #include "kernel_tiling/kernel_tiling.h" | COMMENT |
| LOW | …35/fused_sparse_attention_overlap_service_vector_mla.h | 21 | #include "lib/matmul_intf.h" | COMMENT |
| LOW | …ch35/fused_sparse_attention_overlap_service_cube_mla.h | 21 | #include "lib/matmul_intf.h" | COMMENT |
| LOW | …ttention/rms_norm_dynamic_quant/op_host/CMakeLists.txt | 1 | # This program is free software, you can redistribute it and/or modify it. | COMMENT |
| LOW | …ttention/rms_norm_dynamic_quant/op_host/CMakeLists.txt | 21 | # ) | COMMENT |
| LOW | csrc/attention/msa_index_score/op_host/CMakeLists.txt | 1 | # ----------------------------------------------------------------------------------------------------------- | COMMENT |
| 1004 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 9 | CODE | |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 10 | CODE | |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 10 | CODE | |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 10 | CODE | |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 10 | CODE | |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 10 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 14 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 15 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 18 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 19 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 20 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 21 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 22 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 22 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 23 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 27 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 27 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 27 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 27 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 33 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 34 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 35 | CODE | |
| LOW | csrc/scripts/build_cache.py | 15 | CODE | |
| LOW | tools/test_memfabric_pd_read.py | 26 | CODE | |
| LOW | tools/test_memfabric_pd_read.py | 125 | CODE | |
| LOW | tools/upload_to_openlibing.py | 3 | CODE | |
| LOW | tools/check_symbolic_meta.py | 20 | CODE | |
| LOW | tools/send_mm_request.py | 1 | CODE | |
| LOW | tools/check_long_functions.py | 20 | CODE | |
| LOW | tools/docs_codegen/scanner.py | 1 | CODE | |
| LOW | tools/docs_codegen/sphinx_extension.py | 1 | CODE | |
| LOW | tools/docs_codegen/generator.py | 1 | CODE | |
| LOW | tools/docs_codegen/cli.py | 1 | CODE | |
| LOW | tools/docs_codegen/utils.py | 1 | CODE | |
| LOW | tools/docs_codegen/yaml_loader.py | 1 | CODE | |
| LOW | tools/docs_codegen/converters.py | 1 | CODE | |
| LOW | tools/docs_codegen/errors.py | 1 | CODE | |
| LOW | tests/ut/test_platform.py | 116 | CODE | |
| LOW | tests/ut/test_platform.py | 176 | CODE | |
| LOW | tests/ut/test_platform.py | 2098 | CODE | |
| LOW | tests/ut/_310p/test_mamba_align_fallback_310p_source.py | 9 | CODE | |
| LOW | tests/ut/core/test_batch_job_aware_scheduler.py | 14 | CODE | |
| LOW | tests/ut/distributed/test_yuanrong_backend.py | 22 | CODE | |
| LOW | …s/ut/distributed/weight_transfer/test_packed_tensor.py | 26 | CODE | |
| LOW | tests/ut/distributed/ascend_store/test_backend.py | 27 | CODE | |
| LOW | …sts/ut/distributed/ascend_store/test_pool_scheduler.py | 35 | CODE | |
| 631 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | setup.py | 202 | print(f"Error running build_aclnn.sh: {e}") | CODE |
| LOW | …/tests/pytest/test_quant_lightning_indexer_v2_batch.py | 122 | except Exception as error: | CODE |
| LOW | …/tests/pytest/test_quant_lightning_indexer_v2_batch.py | 170 | except Exception as e: | CODE |
| LOW | …/tests/pytest/test_quant_lightning_indexer_v2_batch.py | 181 | except Exception as e: | CODE |
| MEDIUM | …/tests/pytest/test_quant_lightning_indexer_v2_batch.py | 71 | def qliv2(testcase_file): | CODE |
| LOW | …lightning_indexer_v2/tests/pytest/collect_perf_data.py | 159 | except Exception as e: | CODE |
| LOW | …lightning_indexer_v2/tests/pytest/collect_perf_data.py | 181 | except Exception as e: | CODE |
| LOW | …ightning_indexer_v2/tests/pytest/batch/replace_path.py | 31 | except Exception as e: | CODE |
| LOW | …sts/pytest/batch/quant_lightning_indexer_v2_pt_save.py | 144 | except Exception as e: | CODE |
| LOW | …sts/pytest/batch/quant_lightning_indexer_v2_pt_save.py | 166 | except Exception as e: | CODE |
| LOW | …ention/quant_lightning_indexer_v2/tests/assets/spec.py | 33 | except Exception as exc: | CODE |
| LOW | …ghtning_indexer_v2/tests/assets/impl/npu_preprocess.py | 64 | except Exception: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/golden.py | 71 | except Exception as exc: | CODE |
| LOW | …htning_indexer_v2/tests/assets/impl/metadata_inputs.py | 40 | except Exception: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 83 | except Exception as exc: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 100 | except Exception as exc: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 118 | except Exception as exc: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 144 | except Exception as exc: | CODE |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 166 | except Exception as exc: | CODE |
| LOW | …uant_lightning_indexer_v2/tests/assets/impl/compare.py | 53 | except Exception as exc: | CODE |
| LOW | csrc/cmake/scripts/convert_yaml.py | 30 | except Exception as e: | CODE |
| LOW | csrc/cmake/scripts/convert_yaml.py | 81 | except Exception as e: | CODE |
| LOW | csrc/cmake/scripts/check_version_compatible.py | 92 | except Exception as e: | CODE |
| LOW | csrc/cmake/scripts/util/ascendc_gen_options.py | 23 | except Exception as err: | CODE |
| LOW | csrc/cmake/scripts/examples/get_soc_info.py | 52 | except Exception as e: | CODE |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 430 | except Exception as e: | CODE |
| LOW | csrc/scripts/opgen/opgen_standalone.py | 157 | except Exception as e: | CODE |
| LOW | csrc/scripts/package/package.py | 86 | except Exception as exception: | CODE |
| LOW | csrc/scripts/package/package.py | 166 | except Exception as ex: | CODE |
| LOW | csrc/scripts/package/package.py | 819 | except Exception as e: | CODE |
| LOW | csrc/scripts/package/common/py/pkg_parser.py | 989 | except Exception: | CODE |
| LOW | tools/aisbench.py | 360 | except Exception as e: | CODE |
| LOW | tools/send_request.py | 131 | except Exception: | CODE |
| LOW | tools/vllm_bench.py | 152 | except Exception as e: | CODE |
| MEDIUM | tools/vllm_bench.py | 146 | def run_vllm_bench_case(model_name, port, config, baseline, threshold=0.97, mode | CODE |
| LOW | tools/test_memfabric_pd_read.py | 167 | except Exception as error: | CODE |
| LOW⚡ | tools/test_memfabric_pd_read.py | 365 | except Exception as error: | CODE |
| LOW⚡ | tools/test_memfabric_pd_read.py | 372 | except Exception as error: | CODE |
| LOW⚡ | tools/test_memfabric_pd_read.py | 379 | except Exception as error: | CODE |
| LOW | tools/test_memfabric_pd_read.py | 412 | except Exception as error: | CODE |
| MEDIUM | tools/test_memfabric_pd_read.py | 122 | def _load_torch(): | CODE |
| LOW | tools/upload_to_openlibing.py | 52 | except Exception as e: | CODE |
| LOW | tools/upload_to_openlibing.py | 156 | except Exception: | CODE |
| LOW | tools/upload_to_openlibing.py | 247 | except Exception as e: | CODE |
| LOW | tools/ai_qos.py | 137 | except Exception as e: | CODE |
| MEDIUM | tools/ai_qos.py | 145 | print("Error: no visible NPU devices found.", file=sys.stderr) | CODE |
| MEDIUM | tools/ai_qos.py | 173 | print(f"Error: {VISIBLE_DEVICE_ENV} must list at least one device.", file=sys.stderr) | CODE |
| LOW | tools/format_contributors.py | 41 | except Exception: | CODE |
| LOW | tools/format_contributors.py | 45 | except Exception: | CODE |
| LOW | tools/bisect/version_compat.py | 166 | except Exception as exc: # noqa: BLE001 - broken metadata must not abort discovery | CODE |
| LOW | tools/bisect/version_compat.py | 285 | except Exception as exc: # noqa: BLE001 - add context for bisect logs | CODE |
| LOW | tools/bisect/build_manager.py | 75 | except Exception: # noqa: BLE001 - best effort; fall back to rebuild | CODE |
| LOW | tests/ut/proxy/test_load_balance_proxy_server.py | 81 | except Exception: | CODE |
| LOW | tests/ut/proxy/test_load_balance_proxy_server.py | 96 | except Exception: | CODE |
| LOW | tests/ut/distributed/ascend_store/test_glm53_store.py | 374 | except Exception as error: | CODE |
| MEDIUM | tests/ut/distributed/ascend_store/test_glm53_store.py | 371 | def save(): | CODE |
| LOW | …/ut/distributed/ascend_store/test_mooncake_pipeline.py | 106 | except Exception as exc: | CODE |
| LOW | …/ut/distributed/ascend_store/test_mooncake_pipeline.py | 420 | except Exception as exc: | CODE |
| MEDIUM | …/ut/distributed/ascend_store/test_mooncake_pipeline.py | 102 | def run(): | CODE |
| MEDIUM | …/ut/distributed/ascend_store/test_mooncake_pipeline.py | 407 | def run(): | CODE |
| 391 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | collect_env.py | 198 | CODE | |
| LOW | setup.py | 74 | CODE | |
| LOW | setup.py | 328 | CODE | |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 176 | CODE | |
| LOW | …tention_score/tests/test_flash_decoding_generalized.py | 269 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 25 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 69 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 133 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 269 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 31 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 75 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 155 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 40 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 84 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 209 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 345 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 467 | CODE | |
| LOW | …msa_index_score/tests/golden/msa_index_score_golden.py | 63 | CODE | |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 1327 | CODE | |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 2680 | CODE | |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 371 | CODE | |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 670 | CODE | |
| LOW | …r_v2/tests/pytest/quant_lightning_indexer_v2_golden.py | 982 | CODE | |
| LOW | …/tests/pytest/test_quant_lightning_indexer_v2_batch.py | 160 | CODE | |
| LOW | …lightning_indexer_v2/tests/pytest/collect_perf_data.py | 37 | CODE | |
| LOW | …lightning_indexer_v2/tests/pytest/collect_perf_data.py | 92 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 46 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 244 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 403 | CODE | |
| LOW | …2/tests/pytest/quant_lightning_indexer_v2_acl_graph.py | 74 | CODE | |
| LOW | …est/batch/quant_lightning_indexer_v2_pt_loadprocess.py | 20 | CODE | |
| LOW | …ghtning_indexer_v2/tests/assets/impl/npu_preprocess.py | 124 | CODE | |
| LOW | …ning_indexer_v2/tests/assets/impl/metadata_protocol.py | 181 | CODE | |
| LOW | …ning_indexer_v2/tests/assets/impl/batch_consistency.py | 284 | CODE | |
| LOW | …ning_indexer_v2/tests/assets/impl/batch_consistency.py | 468 | CODE | |
| LOW | …ning_indexer_v2/tests/assets/impl/batch_consistency.py | 546 | CODE | |
| LOW | csrc/cmake/scripts/parse_changed_files.py | 186 | CODE | |
| LOW | csrc/cmake/scripts/parse_changed_files.py | 209 | CODE | |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 326 | CODE | |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 361 | CODE | |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 182 | CODE | |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 212 | CODE | |
| LOW | csrc/cmake/scripts/util/opdesc_parser.py | 283 | CODE | |
| LOW | csrc/cmake/scripts/util/parse_ini_to_json.py | 158 | CODE | |
| LOW | csrc/cmake/scripts/util/parse_ini_to_json.py | 239 | CODE | |
| LOW | csrc/cmake/scripts/util/parse_ini_to_json.py | 314 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_impl_build.py | 345 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_impl_build.py | 419 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_impl_build.py | 501 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_bin_param_build.py | 437 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_bin_param_build.py | 215 | CODE | |
| LOW | csrc/cmake/scripts/util/ascendc_gen_options.py | 28 | CODE | |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 83 | CODE | |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 104 | CODE | |
| LOW | csrc/cmake/scripts/utest/gen_tiling_data_stub.py | 200 | CODE | |
| LOW | csrc/scripts/build_cache.py | 360 | CODE | |
| LOW | csrc/scripts/build_cache.py | 1216 | CODE | |
| LOW | csrc/scripts/build_cache.py | 1293 | CODE | |
| LOW | csrc/scripts/build_cache.py | 1421 | CODE | |
| LOW | csrc/scripts/package/package.py | 245 | CODE | |
| 504 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | csrc/scripts/build_cache.py | 142 | except OSError: | CODE |
| HIGH | csrc/scripts/build_cache.py | 460 | except UnicodeDecodeError: | CODE |
| HIGH | csrc/scripts/package/common/py/version_info.py | 84 | except ValueError: | CODE |
| HIGH | tools/bisect/runner.py | 114 | except psutil.Error: | CODE |
| HIGH | tools/docs_codegen/converters.py | 137 | except json.JSONDecodeError: | CODE |
| HIGH | tests/ut/conftest.py | 137 | except RuntimeError: | CODE |
| HIGH | tests/ut/proxy/test_load_balance_proxy_server.py | 96 | except Exception: | CODE |
| HIGH | tests/ut/proxy/test_load_balance_proxy_server.py | 81 | except Exception: | CODE |
| HIGH | …ts/ut/distributed/ascend_store/test_mooncake_hybrid.py | 307 | except RuntimeError: | CODE |
| HIGH | tests/e2e/conftest.py | 263 | except ProcessLookupError: | CODE |
| HIGH | …sts/e2e/pull_request/one_card/test_sfa_v1_precision.py | 47 | except Exception: | CODE |
| HIGH | tests/e2e/pull_request/one_card/rlhf/conftest.py | 168 | except Exception: | CODE |
| HIGH | tests/e2e/models/test_rm_eval_correctness.py | 68 | except ValueError: | CODE |
| HIGH | tests/e2e/common/multi_node/benchmark_results.py | 91 | except (ValueError, AttributeError): | CODE |
| HIGH | tests/e2e/common/multi_node/external_dp/runtime.py | 123 | except OSError: | CODE |
| HIGH | tests/e2e/common/single_node/kv_pool_runtime.py | 117 | except ProcessLookupError: | CODE |
| HIGH | tests/e2e/common/single_node/kv_pool_runtime.py | 83 | except OSError: | CODE |
| HIGH | tests/e2e/common/single_node/kv_pool_runtime.py | 115 | except ProcessLookupError: | CODE |
| HIGH | tests/e2e/common/single_node/test_single_node.py | 498 | except (ValueError, AttributeError): | CODE |
| HIGH | …kills/vllm-ascend-release/scripts/scan_release_bugs.py | 92 | except Exception: | CODE |
| HIGH | …ills/vllm-ascend-release/scripts/scan_test_coverage.py | 158 | except Exception: | CODE |
| HIGH | vllm_ascend/__init__.py | 46 | except Exception: | CODE |
| HIGH | vllm_ascend/logger.py | 165 | except Exception: | CODE |
| HIGH | vllm_ascend/logger.py | 120 | except OSError: | CODE |
| HIGH | vllm_ascend/core/scheduler_profiling_chunk.py | 242 | except AttributeError: | CODE |
| HIGH | vllm_ascend/eplb/core/eplb_utils.py | 107 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 50 | except ImportError: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 511 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 525 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 534 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 516 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/eplb_communicator.py | 521 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/eplb/hixl_compat.py | 469 | except Exception: | CODE |
| HIGH | vllm_ascend/distributed/kv_transfer/__init__.py | 114 | except ImportError: | CODE |
| HIGH | …tributed/kv_transfer/kv_p2p/sfa_pd_rd2h/send_thread.py | 187 | except queue.Empty: | CODE |
| HIGH | …tributed/kv_transfer/kv_p2p/sfa_pd_rd2h/send_thread.py | 198 | except queue.Empty: | CODE |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/coordinator.py | 460 | except (TypeError, ValueError): | CODE |
| HIGH | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 797 | except ValueError: | CODE |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/kv_transfer.py | 78 | except Exception: | CODE |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/kv_transfer.py | 94 | except Exception: | CODE |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/kv_transfer.py | 752 | except Exception: | CODE |
| HIGH | …nsfer/kv_pool/ascend_store/backend/mooncake_backend.py | 56 | except Exception: | CODE |
| HIGH | …del_executor/warmup/rejection_sampler_triton_warmup.py | 315 | except ImportError: | CODE |
| HIGH | vllm_ascend/model_executor/warmup/rms_triton_warmup.py | 73 | except ImportError: | CODE |
| HIGH | vllm_ascend/models/minimax_m3/ops/msa_m3_triton.py | 147 | except AssertionError: | CODE |
| HIGH | vllm_ascend/models/minimax_m3/ops/msa_m3_triton.py | 167 | except AssertionError: | CODE |
| HIGH | vllm_ascend/patch/platform/patch_balance_schedule.py | 92 | except Exception: | CODE |
| HIGH | vllm_ascend/patch/platform/patch_dyntra_lb_core.py | 60 | except Exception: | CODE |
| HIGH | vllm_ascend/patch/platform/patch_profiling_chunk.py | 66 | except RuntimeError: | CODE |
| HIGH | vllm_ascend/patch/worker/patch_triton.py | 86 | except ImportError: | CODE |
| HIGH | vllm_ascend/patch/worker/patch_triton.py | 120 | except ImportError: | CODE |
| HIGH | vllm_ascend/patch/worker/patch_triton.py | 318 | except ImportError: | CODE |
| HIGH | vllm_ascend/patch/worker/patch_kv_cache_dtype.py | 119 | except ImportError: | CODE |
| HIGH | …_ascend/patch/worker/patch_v2/patch_spec_decode_310.py | 62 | except Exception: | CODE |
| HIGH | vllm_ascend/model_loader/rfork/transfer_backend.py | 171 | except Exception: | CODE |
| HIGH | vllm_ascend/model_loader/rfork/transfer_backend.py | 177 | except Exception: | CODE |
| HIGH | vllm_ascend/model_loader/rfork/rfork_loader.py | 578 | except (TypeError, ValueError, RuntimeError): | CODE |
| HIGH | …fill_v1/load_balance_proxy_layerwise_server_example.py | 115 | except ImportError: | CODE |
| HIGH | …fill_v1/load_balance_proxy_layerwise_server_example.py | 295 | except ValueError: | CODE |
| HIGH | …fill_v1/load_balance_proxy_layerwise_server_example.py | 129 | except Exception: | CODE |
| 14 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | tests/ut/attention/utils.py | 0 | specification for a batch configuration (workload shape only). | STRING |
| HIGH | tests/ut/spec_decode/test_eagle_proposer.py | 0 | specification for a batch configuration (workload shape only). | STRING |
| HIGH | tests/e2e/pull_request/one_card/attention_utils.py | 0 | specification for a batch configuration (workload shape only). | STRING |
| HIGH | tests/ut/patch/platform/test_prefix_cache_cp_patches.py | 0 | build a kvcachetensor; vllm #51718 renamed shared_by -> layers on main. | STRING |
| HIGH | tests/ut/worker/test_model_runner_v2_mamba.py | 0 | build a kvcachetensor; vllm #51718 renamed shared_by -> layers on main. | STRING |
| HIGH | tests/ut/worker/test_attn_utils_v2.py | 0 | build a kvcachetensor; vllm #51718 renamed shared_by -> layers on main. | STRING |
| HIGH | …l_request/one_card/test_model_runner_v1_with_device.py | 0 | build a kvcachetensor; vllm #51718 renamed shared_by -> layers on main. | STRING |
| HIGH | …l_request/four_card/spec_decode/test_mtp_qwen3_next.py | 0 | compare the short outputs of hf and vllm when using greedy sampling. run `pytest tests/e2e/pull_request/one_card/test_vl | STRING |
| HIGH | …pull_request/four_card/spec_decode/test_mtp_step3p5.py | 0 | compare the short outputs of hf and vllm when using greedy sampling. run `pytest tests/e2e/pull_request/one_card/test_vl | STRING |
| HIGH | …equest/four_card/spec_decode/test_dspark_deepseekv4.py | 0 | compare the short outputs of hf and vllm when using greedy sampling. run `pytest tests/e2e/pull_request/one_card/test_vl | STRING |
| HIGH | …/e2e/pull_request/two_card/test_gpt_oss_distributed.py | 0 | compare the short outputs of hf and vllm when using greedy sampling. run `pytest tests/e2e/pull_request/one_card/test_vl | STRING |
| HIGH | tests/e2e/pull_request/one_card/test_vlm.py | 0 | compare the short outputs of hf and vllm when using greedy sampling. run `pytest tests/e2e/pull_request/one_card/test_vl | STRING |
| HIGH | …s/e2e/pull_request/two_card/lora/test_qwen3moe_lora.py | 0 | i want you to act as a sql terminal in front of an example database, you need only to return the sql command to me. do n | STRING |
| HIGH | …ts/e2e/pull_request/one_card/lora/test_llama32_lora.py | 0 | i want you to act as a sql terminal in front of an example database, you need only to return the sql command to me. do n | STRING |
| HIGH | tests/e2e/pull_request/one_card/lora/test_olmoe_lora.py | 0 | i want you to act as a sql terminal in front of an example database, you need only to return the sql command to me. do n | STRING |
| HIGH | vllm_ascend/_310p/quantization/methods/w8a8_dynamic.py | 0 | return (w1, w1_scale, w2, w2_scale) in the standard mlp layout. | STRING |
| HIGH | vllm_ascend/quantization/methods/w4a8/w4a8.py | 0 | return (w1, w1_scale, w2, w2_scale) in the standard mlp layout. | STRING |
| HIGH | vllm_ascend/quantization/methods/w8a8/w8a8_dynamic.py | 0 | return (w1, w1_scale, w2, w2_scale) in the standard mlp layout. | STRING |
| HIGH | vllm_ascend/attention/dsa_v1.py | 0 | metadata for mlacommon. note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/sfa_v1.py | 0 | metadata for mlacommon. note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/mla_v1.py | 0 | metadata for mlacommon. note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/context_parallel/dsa_cp.py | 0 | metadata for mlacommon. note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/dsa_v1.py | 0 | note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/sfa_v1.py | 0 | note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/mla_v1.py | 0 | note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/context_parallel/dsa_cp.py | 0 | note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | vllm_ascend/attention/context_parallel/mla_cp.py | 0 | note: please read the comment at the top of the file before trying to understand this class | STRING |
| HIGH | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 0 | get and clear the requests that have been completed. returns: a set of request ids that have been completed. | STRING |
| HIGH | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 0 | get and clear the requests that have been completed. returns: a set of request ids that have been completed. | STRING |
| HIGH | …tributed/kv_transfer/kv_p2p/mooncake_d2rh_connector.py | 0 | get and clear the requests that have been completed. returns: a set of request ids that have been completed. | STRING |
| HIGH | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 0 | get and clear the requests that have been completed. returns: a set of request ids that have been completed. | STRING |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/kv_transfer.py | 0 | get and clear the requests that have been completed. returns: a set of request ids that have been completed. | STRING |
| HIGH | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 0 | run the thread to handle kv cache transfer requests. | STRING |
| HIGH | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 0 | run the thread to handle kv cache transfer requests. | STRING |
| HIGH | …tributed/kv_transfer/kv_p2p/mooncake_d2rh_connector.py | 0 | run the thread to handle kv cache transfer requests. | STRING |
| HIGH | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 0 | run the thread to handle kv cache transfer requests. | STRING |
| HIGH | …ibuted/kv_transfer/kv_pool/ascend_store/kv_transfer.py | 0 | run the thread to handle kv cache transfer requests. | STRING |
| HIGH | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 0 | for remote prefill, pull all prompt blocks from remote asynchronously relative to engine execution. args: request (reque | STRING |
| HIGH | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 0 | for remote prefill, pull all prompt blocks from remote asynchronously relative to engine execution. args: request (reque | STRING |
| HIGH | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 0 | for remote prefill, pull all prompt blocks from remote asynchronously relative to engine execution. args: request (reque | STRING |
| HIGH | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 0 | once a request is finished, determine whether request blocks should be freed now or will be sent asynchronously and free | STRING |
| HIGH | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 0 | once a request is finished, determine whether request blocks should be freed now or will be sent asynchronously and free | STRING |
| HIGH | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 0 | once a request is finished, determine whether request blocks should be freed now or will be sent asynchronously and free | STRING |
| HIGH | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 0 | hash the string using sha-256 and convert it into an int64 integer. | STRING |
| HIGH | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 0 | hash the string using sha-256 and convert it into an int64 integer. | STRING |
| HIGH | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 0 | hash the string using sha-256 and convert it into an int64 integer. | STRING |
| HIGH | …scend/distributed/kv_transfer/kv_p2p/mooncake/utils.py | 0 | hash the string using sha-256 and convert it into an int64 integer. | STRING |
| HIGH | vllm_ascend/models/qwen3_dspark.py | 0 | per-position acceptance probability for each drafted token. | STRING |
| HIGH | vllm_ascend/models/deepseek_v4/dspark.py | 0 | per-position acceptance probability for each drafted token. | STRING |
| HIGH | vllm_ascend/models/deepseek_v41/dspark.py | 0 | per-position acceptance probability for each drafted token. | STRING |
| HIGH | vllm_ascend/quantization/methods/w4a8/w4a8.py | 0 | normalized weight payload for the fused_mc2 comm path. | STRING |
| HIGH | vllm_ascend/quantization/methods/w8a8/w8a8_dynamic.py | 0 | normalized weight payload for the fused_mc2 comm path. | STRING |
| HIGH | vllm_ascend/ops/fused_moe/routed_experts.py | 0 | normalized weight payload for the fused_mc2 comm path. | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 302 | // Step 1: Read grouplist reduceSum to calculate total data count | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 310 | // Step 2: Calculate core allocation | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 333 | // Step 3: Calculate total data volume | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 339 | // Step 4: Allocate space | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 396 | // Step 1: Read grouplist reduceSum to calculate total data count | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 400 | // Step 2: Calculate core allocation | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 423 | // Step 3: Calculate total data volume | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 429 | // Step 4: Allocate space | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 446 | # Step 3: PV: [groupSize, valid_len] x [valid_len, D] -> [groupSize, D] | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 449 | # Step 4: rescale O | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 457 | # Step 5: final divide + cast | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 414 | # Step 1: QK: [groupSize, D] x [D, valid_len] -> [groupSize, valid_len] | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 236 | // Step 1: Cast int8 -> half -> fp32 | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 244 | // Step 2: Mul dequant_scale | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 252 | // Step 3: Add dequant_bias (if exists) | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 289 | // Step 1: tanh(gate / beta) * beta | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 299 | // Step 2: sigmoid(gate) = 1 / (1 + exp(-gate)) | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 334 | // Step 3: situ_a = tanhResult * sigmoidResult | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 338 | // Step 4: if linear_beta > 0: up = linear_beta * tanh(up / linear_beta) | COMMENT |
| LOW | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 349 | // Step 5: output = situ_a * up = tanhResult * up | COMMENT |
| LOW⚡ | …moe/situ_mx_quant/op_host/situ_mx_quant_infershape.cpp | 89 | // Step 1: Compute y shape (Situ output) | COMMENT |
| LOW⚡ | …moe/situ_mx_quant/op_host/situ_mx_quant_infershape.cpp | 96 | // Step 2: Compute mxscale shape | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 206 | // Step 1: Situ activation | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 211 | // Step 2: MxQuant - extract max exponent per 32-element block | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 216 | // Step 3: MxQuant - compute E8M0 scale and reciprocal scale | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 224 | // Step 4: MxQuant - quantize to FP8 | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 397 | # Step 1: Create weights | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 412 | # Step 2: Simulate weight loading | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 440 | # Step 3: Process after loading | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 75 | #### Step 1: Software Installation | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 89 | #### Step 2: Configure mooncake.json and Start mooncake_master | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 200 | #### Step 4: Start vLLM inference service | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 448 | #### Step 5: MooncakeStore SSD Offload (Embedded Real Client Mode) | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 607 | #### Step 1: Prerequisite Checks | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 636 | # Step 1: Disable HDK signature verification (only needs to be executed once per machine) | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 681 | #### Step 2: Software Installation | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 688 | #### Step 3: Configure Memcache Config File | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 761 | #### Step 4: Run MetaService | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 771 | #### Step 5: Start vLLM inference service | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 990 | #### Step 6: Memcache and vLLM Separated Deployment {: #Memcache-and-vLLM-Separated-Deployment } | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1016 | #### Step 7: Enable Memcache SSD Cache | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1161 | #### Step 1: Install Yuanrong Datasystem | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1176 | #### Step 2: Choose a Service Discovery Backend | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1275 | #### Step 3: Start Datasystem Worker | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1335 | #### Step 4: Configure Environment Variables and `yuanrong.json` | COMMENT |
| LOW | …user_guide/feature_guide/kv_cache_pool_ascend_store.md | 1474 | #### Step 5: Run AscendStoreConnector (Yuanrong Backend) | COMMENT |
| LOW | docs/source/user_guide/feature_guide/ucm_deployment.md | 782 | # Step 1: Pre-seed KV cache (25600 = 32000 * 0.8) | COMMENT |
| LOW | docs/source/user_guide/feature_guide/ucm_deployment.md | 796 | # Step 2: Run performance test | COMMENT |
| LOW | …r_guide/Design_Documents/ModelRunner_prepare_inputs.md | 99 | ### Step 1: All requests in the prefill phase | COMMENT |
| LOW | …r_guide/Design_Documents/ModelRunner_prepare_inputs.md | 209 | ### Step 2: Chunked prefill | COMMENT |
| LOW | …rials/features/pd_colocated_mooncake_multi_instance.md | 276 | ### Step 1: Baseline (No Cache) | COMMENT |
| LOW | …rials/features/pd_colocated_mooncake_multi_instance.md | 288 | ### Step 2: Local DRAM Hit | COMMENT |
| LOW | …rials/features/pd_colocated_mooncake_multi_instance.md | 293 | ### Step 3: Cross-Node DRAM Hit | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 262 | # Step 1: estimate how many tokens this round will advance | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 269 | # Step 2: simulate position after this round's execution | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 272 | # Step 3: estimate the NEXT round's token demand | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 281 | # Step 4: compute new blocks needed from that position | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 46 | # Step 1: Sort the items by weight in descending order (we are sorting by weight now) | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 59 | # Step 2: Calculate the number of items per box | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 64 | # Step 3: Initialize card_num boxes with empty lists to store item IDs | COMMENT |
| 42 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | setup.py | 44 | logger = logging.getLogger(__name__) | CODE |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 24 | logger = logging.getLogger(__name__) | CODE |
| LOW | …quant_lightning_indexer_v2/torch_extension/__init__.py | 1 | __all__ = [ | CODE |
| LOW | tools/upload_to_openlibing.py | 19 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/ai_qos.py | 338 | def set_qos(self, state_path: Path) -> None: | CODE |
| LOW | tools/bisect/version_compat.py | 38 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/coordinator.py | 38 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/runner.py | 51 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/git_ops.py | 26 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/good_table.py | 40 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/build_manager.py | 44 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/state.py | 22 | logger = logging.getLogger(__name__) | CODE |
| LOW | tools/bisect/verdict.py | 30 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/ut/proxy/test_load_balance_proxy_server.py | 67 | def _set_script(port: int, items: list[dict]) -> None: | CODE |
| LOW | tests/e2e/conftest.py | 90 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/dp_mode.py | 9 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/utils.py | 11 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/benchmark_results.py | 6 | logger = logging.getLogger(__name__) | CODE |
| LOW | …s/e2e/common/multi_node/internal_dp/test_multi_node.py | 27 | logger = logging.getLogger(__name__) | CODE |
| LOW | …e2e/common/multi_node/internal_dp/multi_node_config.py | 19 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/internal_dp/utils.py | 17 | __all__ = [ | CODE |
| LOW | …2e/common/multi_node/external_dp/external_dp_config.py | 21 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/external_dp/runtime.py | 36 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/multi_node/external_dp/utils.py | 28 | logger = logging.getLogger(__name__) | CODE |
| LOW | …/e2e/common/multi_node/external_dp/test_external_dp.py | 45 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/single_node/test_single_node.py | 28 | logger = logging.getLogger(__name__) | CODE |
| LOW | tests/e2e/common/single_node/single_node_config.py | 14 | logger = logging.getLogger(__name__) | CODE |
| LOW | …2e/vllm_interface/vllm_interface_contracts/__init__.py | 34 | __all__ = [ | CODE |
| LOW | …m_interface/vllm_interface_contracts/call_contracts.py | 1479 | __all__ = [ | CODE |
| LOW | vllm_ascend/logger.py | 137 | def _setup_file_logging(log_dir: str | None = None) -> None: | CODE |
| LOW⚡ | vllm_ascend/platform.py | 1163 | def _update_compilation_modes(vllm_config: VllmConfig, ascend_config: AscendConfig) -> None: | CODE |
| LOW | vllm_ascend/platform.py | 405 | def update_block_size_for_backend(cls, vllm_config: VllmConfig) -> None: | CODE |
| LOW | vllm_ascend/platform.py | 1503 | def _set_pytorch_npu_alloc_env(vllm_config: VllmConfig) -> None: | CODE |
| LOW | vllm_ascend/utils.py | 1290 | def set_potential_max_tokens(vllm_config) -> None: | CODE |
| LOW | vllm_ascend/_310p/attention/metadata_builder.py | 39 | def set_query_lens_cpu(attn_metadata: AscendMetadata, query_lens_cpu: torch.Tensor) -> None: | CODE |
| LOW | vllm_ascend/_310p/spec_decode/__init__.py | 20 | __all__ = [ | CODE |
| LOW | vllm_ascend/_310p/quantization/__init__.py | 20 | __all__ = [ | CODE |
| LOW | vllm_ascend/_310p/sample/__init__.py | 3 | __all__ = ["AscendSampler310"] | CODE |
| LOW | vllm_ascend/_310p/ops/fla/__init__.py | 6 | __all__ = [ | CODE |
| LOW | vllm_ascend/_310p/worker/v2/model_runner.py | 674 | def update_requests(self, scheduler_output: SchedulerOutput) -> None: | CODE |
| LOW | vllm_ascend/_310p/worker/v2/spec_utils.py | 486 | def set_draft_step_host(step: int) -> None: | CODE |
| LOW | …m_ascend/_310p/worker/v2/spec_decode/mtp_speculator.py | 284 | def _set_draft_step(self, step: int) -> None: | CODE |
| LOW⚡ | vllm_ascend/attention/sfa_v1.py | 1576 | def _update_indexcache_topk_indices(self, topk_indices: torch.Tensor) -> None: | CODE |
| LOW | vllm_ascend/core/profiling_chunk_predictor.py | 214 | def set_target_latency(self, base_chunk_size: int, elapsed_time: float = 0.0) -> None: | CODE |
| LOW | vllm_ascend/xlite/utils.py | 41 | def set_dummy_tensor(tensor: torch.Tensor) -> None: | CODE |
| LOW | vllm_ascend/xlite/xlite.py | 466 | def set_xlite_attr(xlite_attr: str, layer_attr: str) -> None: | CODE |
| LOW⚡ | vllm_ascend/distributed/eplb/eplb_communicator.py | 365 | def set_stream(self, stream: torch.Stream | None) -> None: | CODE |
| LOW⚡ | vllm_ascend/distributed/eplb/eplb_communicator.py | 370 | def set_transfer_context(self, old_indices: np.ndarray, layer_idx: int) -> None: | CODE |
| LOW | vllm_ascend/distributed/eplb/eplb_communicator.py | 81 | def set_stream(self, stream: torch.Stream | None) -> None: | CODE |
| LOW | vllm_ascend/distributed/eplb/eplb_state.py | 69 | def _set_local_expert_range(self, expert_load_view: torch.Tensor) -> None: | CODE |
| LOW | …end/distributed/device_communicators/pyhccl_wrapper.py | 256 | __all__ = [ | CODE |
| LOW | …scend/distributed/ec_transfer/ec_memcache_connector.py | 196 | def update_state_after_alloc(self, request: "Request", index: int) -> None: | CODE |
| LOW⚡ | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 177 | def update_computed_tokens(self, computed_tokens: int) -> None: | CODE |
| LOW⚡ | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 181 | def update_transferred_tokens(self, transferred_tokens: int) -> None: | CODE |
| LOW | …ted/kv_transfer/kv_p2p/mooncake_layerwise_connector.py | 627 | def update_failed_task(self, req_id: str) -> None: | CODE |
| LOW⚡ | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 1299 | def set_xfer_handshake_metadata(self, metadata: dict[int, KVConnectorHandshakeMetadata]) -> None: | CODE |
| LOW⚡ | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 1309 | def set_xfer_handshake_metadata_pp_aware(self, metadata) -> None: | CODE |
| LOW | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 1632 | def set_xfer_handshake_metadata(self, metadata: dict[int, KVConnectorHandshakeMetadata]) -> None: | CODE |
| LOW | …distributed/kv_transfer/kv_p2p/mooncake/base_worker.py | 365 | __all__ = ["MooncakeBaseConnectorWorker"] | CODE |
| LOW | …nd/distributed/kv_transfer/kv_p2p/mooncake/metadata.py | 174 | __all__ = [ | CODE |
| 96 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 302 | // Step 1: Read grouplist reduceSum to calculate total data count | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 310 | // Step 2: Calculate core allocation | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 333 | // Step 3: Calculate total data volume | COMMENT |
| LOW⚡ | …el/grouped_matmul_swiglu_quant_weight_nz_tensor_list.h | 339 | // Step 4: Allocate space | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 396 | // Step 1: Read grouplist reduceSum to calculate total data count | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 400 | // Step 2: Calculate core allocation | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 423 | // Step 3: Calculate total data volume | COMMENT |
| LOW⚡ | …d_matmul_swiglu_quant_weight_nz_tensor_list_split_ws.h | 429 | // Step 4: Allocate space | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 446 | # Step 3: PV: [groupSize, valid_len] x [valid_len, D] -> [groupSize, D] | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 449 | # Step 4: rescale O | COMMENT |
| LOW⚡ | …rc/attention/sparse_attention_score/tests/test_bf16.py | 457 | # Step 5: final divide + cast | COMMENT |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 414 | # Step 1: QK: [groupSize, D] x [D, valid_len] -> [groupSize, valid_len] | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 236 | // Step 1: Cast int8 -> half -> fp32 | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 244 | // Step 2: Mul dequant_scale | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 252 | // Step 3: Add dequant_bias (if exists) | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 289 | // Step 1: tanh(gate / beta) * beta | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 299 | // Step 2: sigmoid(gate) = 1 / (1 + exp(-gate)) | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 334 | // Step 3: situ_a = tanhResult * sigmoidResult | COMMENT |
| LOW⚡ | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 338 | // Step 4: if linear_beta > 0: up = linear_beta * tanh(up / linear_beta) | COMMENT |
| LOW | …/moe/dequant_situ_quant/op_kernel/dequant_situ_quant.h | 349 | // Step 5: output = situ_a * up = tanhResult * up | COMMENT |
| LOW⚡ | …moe/situ_mx_quant/op_host/situ_mx_quant_infershape.cpp | 89 | // Step 1: Compute y shape (Situ output) | COMMENT |
| LOW⚡ | …moe/situ_mx_quant/op_host/situ_mx_quant_infershape.cpp | 96 | // Step 2: Compute mxscale shape | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 206 | // Step 1: Situ activation | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 211 | // Step 2: MxQuant - extract max exponent per 32-element block | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 216 | // Step 3: MxQuant - compute E8M0 scale and reciprocal scale | COMMENT |
| LOW⚡ | …tu_mx_quant/op_kernel/arch35/situ_mx_quant_axis_last.h | 224 | // Step 4: MxQuant - quantize to FP8 | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 397 | # Step 1: Create weights | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 412 | # Step 2: Simulate weight loading | COMMENT |
| LOW | tests/ut/quantization/methods/kv_cache/test_kv_c8.py | 440 | # Step 3: Process after loading | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 262 | # Step 1: estimate how many tokens this round will advance | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 269 | # Step 2: simulate position after this round's execution | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 272 | # Step 3: estimate the NEXT round's token demand | COMMENT |
| LOW⚡ | vllm_ascend/core/batch_job_aware_scheduler.py | 281 | # Step 4: compute new blocks needed from that position | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 46 | # Step 1: Sort the items by weight in descending order (we are sorting by weight now) | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 59 | # Step 2: Calculate the number of items per box | COMMENT |
| LOW⚡ | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 64 | # Step 3: Initialize card_num boxes with empty lists to store item IDs | COMMENT |
| LOW | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 86 | # Step 4: Distribute items into boxes based on weight | COMMENT |
| LOW | vllm_ascend/eplb/core/policy/policy_default_eplb.py | 108 | # Step 5: Output each box's contents and total weight | COMMENT |
| LOW⚡ | vllm_ascend/ops/cv_linear.py | 24 | # Step 1: Quantize (Vector) | STRING |
| LOW⚡ | vllm_ascend/ops/cv_linear.py | 27 | # Step 2: Matrix multiply (Cube) | STRING |
| LOW | vllm_ascend/ops/triton/batch_invariant/rmsnorm.py | 52 | # Step 1: Compute sum of squares in float32 to avoid overflow | COMMENT |
| LOW⚡ | vllm_ascend/ops/triton/batch_invariant/rmsnorm.py | 63 | # Step 2: Compute RMS (root mean square) in float32 | COMMENT |
| LOW⚡ | vllm_ascend/ops/triton/batch_invariant/rmsnorm.py | 68 | # Step 3: Normalize and apply weight | COMMENT |
| LOW | vllm_ascend/ops/triton/mamba/lightning_attn.py | 421 | # Step 1: Compute diagonal blocks of attention | COMMENT |
| LOW | vllm_ascend/ops/triton/mamba/lightning_attn.py | 469 | # Step 2: Compute key-value outer products for each block in parallel. | COMMENT |
| LOW | vllm_ascend/ops/triton/mamba/lightning_attn.py | 496 | # Step 3: Reduce key-value outer products across blocks and update | COMMENT |
| LOW | vllm_ascend/ops/triton/mamba/lightning_attn.py | 516 | # Step 4: Compute non-diagonal blocks of attention. | COMMENT |
| LOW⚡ | …fill_v1/load_balance_proxy_layerwise_server_example.py | 22 | # Step 1: Start Your Backend Servers | COMMENT |
| LOW⚡ | …fill_v1/load_balance_proxy_layerwise_server_example.py | 34 | # Step 2: Start the Proxy Server | COMMENT |
| LOW⚡ | …fill_v1/load_balance_proxy_layerwise_server_example.py | 48 | # Step 3: Send a Request to the Proxy | COMMENT |
| LOW⚡ | …fill_v1/load_balance_proxy_layerwise_server_example.py | 70 | # Step 4: Health Check | COMMENT |
| LOW⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 22 | # Step 1: Start Your Backend Servers | COMMENT |
| LOW⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 34 | # Step 2: Start the Proxy Server | COMMENT |
| LOW⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 48 | # Step 3: Send a Request to the Proxy | COMMENT |
| LOW⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 70 | # Step 4: Health Check | COMMENT |
| LOW⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 80 | # Step 5: Add or Remove Prefiller or Decoder Instances (Optional) | COMMENT |
| LOW⚡ | examples/eplb/eplb_deepseek.py | 108 | # Step 1: pack groups to nodes | COMMENT |
| LOW⚡ | examples/eplb/eplb_deepseek.py | 117 | # Step 2: construct redundant experts within nodes | COMMENT |
| LOW⚡ | examples/eplb/eplb_deepseek.py | 122 | # Step 3: pack physical_experts to GPUs | COMMENT |
| LOW⚡ | examples/offline_DSD_k_tuner/generate_config.py | 425 | # Step 1: get acceptance_rate_per_pos | COMMENT |
| 13 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | tools/bisect/runner.py | 297 | sources = " ; ".join(f"source {f} 2>/dev/null || true" for f in _ENV_SOURCE_FILES) | CODE |
| HIGH | tools/bisect/worker_agent.py | 53 | sources = " ; ".join(f"source {f} 2>/dev/null || true" for f in runner._ENV_SOURCE_FILES) | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 502 | (r.coverage || {}).arch && r.coverage.arch.includes(filterType) | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 507 | (r.coverage || {}).graph_mode && r.coverage.graph_mode.includes(filterGraph) | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 513 | return Object.values(cov).some(arr => arr && arr.length > 0); | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 679 | while (i < total && rows.length < pageSize && scanned < ZERO_SCAN_CAP) { | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 682 | if (!coveredSet.has(key) && !isInvalidCombo(combo, dims)) rows.push(combo); | CODE |
| HIGH⚡ | tests/e2e/generate_coverage_html.py | 687 | scanCapped: scanned >= ZERO_SCAN_CAP && rows.length < pageSize, | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 405 | let currentSort = {col: null, asc: true}; | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 421 | const c = r.card_count || 1; | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 458 | const isMarked = Object.values(cov).some(arr => arr && arr.length > 0); | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 534 | `<a href="#" onclick="window._hmFilter=null;filter();return false;" style="color:#2563eb">✕ clear</a>`; | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 592 | return (vals && vals.length) ? new Set(vals) : new Set(); | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 629 | for (const v of arr) next.push([...combo, v]); | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 707 | return (vals && vals.length) ? vals.filter(v => allowed.includes(v)) : []; | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 713 | for (const c of combos) for (const v of s) next.push([...c, v]); | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 724 | entry.tests.push(r); | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 866 | if (window._explorerState && explorerPage === 0) tbody.innerHTML = html; | CODE |
| HIGH | tests/e2e/generate_coverage_html.py | 949 | window._hmFilter = null; | CODE |
| HIGH | vllm_ascend/ascend_config.py | 425 | "draft_window_size": null, | STRING |
| HIGH | vllm_ascend/ascend_config.py | 456 | "expert_map_path": null, | STRING |
| HIGH | vllm_ascend/ascend_config.py | 459 | "expert_map_record_path": null, | STRING |
| HIGH | vllm_ascend/ascend_config.py | 500 | "need_timing": null, | STRING |
| HIGH | vllm_ascend/ascend_config.py | 523 | "method": null, | STRING |
| HIGH | …nsfer/kv_pool/ascend_store/backend/mooncake_backend.py | 578 | raise TypeError(f"tenant_id must be a string or null, got {type(value).__name__}: {value!r}") | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 379 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_fp16.py | 421 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 493 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 621 | CODE | |
| LOW | …tention_score/tests/test_sparse_attention_score_fp8.py | 771 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 566 | CODE | |
| LOW | …rc/attention/sparse_attention_score/tests/test_bf16.py | 608 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 30 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 241 | CODE | |
| LOW | …tning_indexer_v2/tests/pytest/result_compare_method.py | 244 | CODE | |
| LOW | …ghtning_indexer_v2/tests/assets/impl/npu_preprocess.py | 221 | CODE | |
| LOW | …quant_lightning_indexer_v2/tests/assets/impl/inputs.py | 790 | CODE | |
| LOW | …rch_extension/graph_convert_quant_lightning_indexer.py | 57 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 202 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 256 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 302 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 363 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 508 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 596 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 59 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 81 | CODE | |
| LOW | …_indexer_v2/torch_extension/quant_lightning_indexer.py | 137 | CODE | |
| LOW | csrc/scripts/util/build_opp_kernel_static.py | 226 | CODE | |
| LOW | tests/deepseek_v41_utils.py | 143 | CODE | |
| LOW | tests/ut/test_ascend_forward_context.py | 26 | CODE | |
| LOW | tests/ut/kv_offload/test_dspark_resident_kv.py | 49 | CODE | |
| LOW | tests/ut/kv_offload/test_dspark_prefill_config.py | 14 | CODE | |
| LOW | tests/ut/kv_offload/mooncake_v2/helpers.py | 98 | CODE | |
| LOW | tests/ut/kv_offload/mooncake_v2/helpers.py | 133 | CODE | |
| LOW | tests/ut/attention/utils.py | 139 | CODE | |
| LOW | tests/ut/distributed/ascend_store/test_pool_worker.py | 50 | CODE | |
| LOW | tests/ut/distributed/ascend_store/test_pool_worker.py | 2399 | CODE | |
| LOW | tests/ut/model_executor/warmup/helpers.py | 12 | CODE | |
| LOW | tests/ut/_tools/build_cache_test_utils.py | 15 | CODE | |
| LOW | tests/ut/_tools/test_build_cache.py | 84 | CODE | |
| LOW | tests/ut/spec_decode/test_dspark_proposer.py | 342 | CODE | |
| LOW | tests/ut/model_loader/rfork/test_identity_validation.py | 102 | CODE | |
| LOW | tests/ut/model_loader/rfork/test_identity_validation.py | 136 | CODE | |
| LOW | tests/ut/ops/test_token_dispatcher.py | 49 | CODE | |
| LOW | tests/e2e/conftest.py | 286 | CODE | |
| LOW | tests/e2e/conftest.py | 981 | CODE | |
| LOW | tests/e2e/conftest.py | 1413 | CODE | |
| LOW | …e/pull_request/one_card/test_attention_v1_precision.py | 130 | CODE | |
| LOW | tests/e2e/pull_request/one_card/attention_utils.py | 139 | CODE | |
| LOW | …/singlecard_ops/test_kimi_kda_recurrent_ascendc_npu.py | 62 | CODE | |
| LOW | …e_node/ops/singlecard_ops/test_npu_moe_gating_top_k.py | 33 | CODE | |
| LOW | …/singlecard_ops/test_recurrent_gated_delta_rule_310.py | 38 | CODE | |
| LOW | vllm_ascend/ascend_forward_context.py | 101 | CODE | |
| LOW | vllm_ascend/_310p/model_runner_310p.py | 568 | CODE | |
| LOW | vllm_ascend/_310p/npu_input_batch.py | 10 | CODE | |
| LOW | vllm_ascend/_310p/fused_moe/fused_moe.py | 155 | CODE | |
| LOW | vllm_ascend/_310p/ops/fla/gdn_310.py | 145 | CODE | |
| LOW | vllm_ascend/attention/indexer_kpool.py | 600 | CODE | |
| LOW | vllm_ascend/attention/context_parallel/common_cp.py | 315 | CODE | |
| LOW | …distributed/kv_transfer/kv_p2p/sfa_pd_rd2h/protocol.py | 197 | CODE | |
| LOW | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 535 | CODE | |
| LOW | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 848 | CODE | |
| LOW | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 968 | CODE | |
| LOW | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 1083 | CODE | |
| LOW | …stributed/kv_transfer/kv_pool/ascend_store/metadata.py | 1307 | CODE | |
| 77 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| CRITICAL | tests/ut/model_executor/test_qwen3_dspark.py | 51 | model.model.fc.weight.data.copy_(fc_weight) | CODE |
| CRITICAL⚡ | tests/ut/model_executor/test_glm_mla_dspark.py | 137 | layer.self_attn.attn.impl.exec_kv_prefill.assert_not_called() | CODE |
| CRITICAL | vllm_ascend/attention/sfa_v1.py | 1097 | weight_uq_qr = self.q_proj.weight.data.T.contiguous() | CODE |
| CRITICAL | vllm_ascend/spec_decode/llm_base_proposer.py | 519 | self.model.model.embed_tokens.weight.cpu(), | CODE |
| CRITICAL | vllm_ascend/ops/triton/triton_utils.py | 74 | device_properties: dict[str, Any] = triton.runtime.driver.active.utils.get_device_properties( | CODE |
| CRITICAL | vllm_ascend/ops/triton/rms_norm.py | 77 | device_properties = triton.runtime.driver.active.utils.get_device_properties(q.device) | CODE |
| CRITICAL | vllm_ascend/ops/triton/batch_invariant/rmsnorm.py | 115 | max_grid_size = triton.runtime.driver.active.utils.get_device_properties(torch.npu.current_device())[ | CODE |
| CRITICAL | vllm_ascend/ops/triton/batch_invariant/matmul.py | 272 | triton.runtime.driver.active.utils.get_device_properties(torch.npu.current_device())["num_vectorcore"] // 2 | CODE |
| CRITICAL | vllm_ascend/ops/triton/spec_decode/ngram.py | 214 | properties = triton.runtime.driver.active.utils.get_device_properties(device_idx) | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | setup.py | 85 | # Check if Chip Name exists (Ascend950 includes it directly) | COMMENT |
| LOW | tools/collect_user_first_contribution.sh | 135 | # Check if it's in format "ID+username" or just "username" | COMMENT |
| LOW | tools/collect_user_first_contribution.sh | 166 | # Check if we should do incremental update | COMMENT |
| LOW | tools/collect_user_first_contribution.sh | 173 | # Check if LAST_COMMIT is an ancestor of CURRENT_HEAD | COMMENT |
| LOW | tools/collect_user_first_contribution.sh | 227 | # Check if contributor already exists (deduplication) | COMMENT |
| LOW | tools/mooncake_installer.sh | 196 | # Check if thirdparties directory exists | COMMENT |
| LOW | tools/mooncake_installer.sh | 206 | # Check if yalantinglibs is already installed | COMMENT |
| LOW | tools/mooncake_installer.sh | 255 | # Check if .gitmodules exists | COMMENT |
| LOW | tools/mooncake_installer.sh | 257 | # Check if submodules are already initialized by looking for the .git directory in the first submodule | COMMENT |
| LOW | tools/mooncake_installer.sh | 307 | # Check if Go is already installed | COMMENT |
| LOW | …ts/ut/model_loader/netloader/test_netloader_elastic.py | 792 | # Check if the start() method is called. | COMMENT |
| LOW | tests/ut/ops/test_comm_utils.py | 50 | # Check if the output tensor is created properly | COMMENT |
| LOW | …four_card/rlhf/consistency/test_batch_invariant_tp4.py | 221 | # Check if tokens match first | COMMENT |
| LOW | …s/e2e/pull_request/two_card/lora/test_qwen3moe_lora.py | 46 | # Print the outputs. | COMMENT |
| LOW | …est/one_card/compile/test_graphex_norm_quant_fusion.py | 166 | # Check if fusion operator is available | COMMENT |
| LOW | …est/one_card/compile/test_graphex_norm_quant_fusion.py | 189 | # Check if the bias operator exists | COMMENT |
| LOW | tests/e2e/pull_request/one_card/lora/test_ilama_lora.py | 32 | # Print the outputs. | COMMENT |
| LOW | tests/e2e/pull_request/one_card/lora/test_olmoe_lora.py | 69 | # Print the outputs. | COMMENT |
| LOW | …ingle_node/ops/singlecard_ops/triton/test_bad_words.py | 38 | # Check if adding this bad word would exceed the token limit | COMMENT |
| LOW | …ingle_node/ops/singlecard_ops/triton/test_bad_words.py | 64 | # Set input_ids to match the first bad word | COMMENT |
| LOW | tests/e2e/common/multi_node/run.sh | 59 | # Set transformers to offline mode to avoid downloading models during tests | COMMENT |
| LOW | …lm-ascend-release/scripts/update_version_references.py | 83 | # Check if this update has a regex pattern or a template pattern | COMMENT |
| LOW | …lm-ascend-release/scripts/update_version_references.py | 180 | # Check if all updates are manual actions | COMMENT |
| LOW | …ts/skills/vllm-ascend-release/scripts/fetch_commits.py | 172 | # Check if this tag's commit is an ancestor of head | COMMENT |
| LOW | …ts/skills/vllm-ascend-release/scripts/fetch_commits.py | 1210 | # Check if message already contains the full URL pattern | COMMENT |
| LOW | …ills/vllm-ascend-release/scripts/scan_test_coverage.py | 329 | # Check if there's a linked issue and if it's resolved | COMMENT |
| LOW | vllm_ascend/profiling_config.py | 557 | # Check if the configuration file already exists | STRING |
| LOW | vllm_ascend/core/dyntra_lb_scheduler.py | 1055 | # Check if the scheduling constraints are satisfied. | COMMENT |
| LOW | vllm_ascend/core/recompute_scheduler.py | 869 | # Check if the scheduling constraints are satisfied. | COMMENT |
| LOW | vllm_ascend/core/scheduler_profiling_chunk.py | 1074 | # Check if the scheduling constraints are satisfied. | COMMENT |
| LOW | …d/distributed/kv_transfer/kv_p2p/mooncake_connector.py | 807 | # Check if we have the remote metadata cached. | COMMENT |
| LOW | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 741 | # Check if we have the remote metadata cached. | COMMENT |
| LOW | …ibuted/kv_transfer/kv_p2p/mooncake_hybrid_connector.py | 861 | # Check if we have the remote metadata cached. | COMMENT |
| LOW | vllm_ascend/quantization/utils.py | 100 | # Check if it's a local path | COMMENT |
| LOW | vllm_ascend/quantization/methods/w8a8/w8a8_mxfp8.py | 178 | # Check if already transformed to avoid double transformation | COMMENT |
| LOW | vllm_ascend/quantization/methods/w8a8/w8a8_mxfp8.py | 404 | # Check if already transformed to avoid double transformation | COMMENT |
| LOW | vllm_ascend/patch/platform/patch_balance_schedule.py | 722 | # Check if the scheduling constraints are satisfied. | COMMENT |
| LOW⚡ | …ascend/model_loader/netloader/executor/netloader_pg.py | 63 | # Check if world_size is positive | COMMENT |
| LOW⚡ | …ascend/model_loader/netloader/executor/netloader_pg.py | 66 | # Check if rank is within [0, world_size - 1] | COMMENT |
| LOW⚡ | …ascend/model_loader/netloader/executor/netloader_pg.py | 69 | # Check if HCCL is available | COMMENT |
| LOW⚡ | …ascend/model_loader/netloader/executor/netloader_pg.py | 72 | # Check if timeout is a timedelta type | COMMENT |
| LOW⚡ | …ascend/model_loader/netloader/executor/netloader_pg.py | 75 | # Check if group_name already exists | COMMENT |
| LOW | …ascend/model_loader/netloader/executor/netloader_pg.py | 121 | # Check if pg_options is None or not of type ProcessGroupHCCL.Options | COMMENT |
| LOW | …ascend/model_loader/netloader/executor/netloader_pg.py | 169 | # Check if pg is in keys of _world.pg_coalesce_state | COMMENT |
| LOW | vllm_ascend/worker/worker.py | 1220 | # Check if profiling is enabled (RFC #6954 - align with upstream vLLM) | COMMENT |
| LOW | vllm_ascend/compilation/compiler_interface.py | 128 | # Set sym_range to limit static kernel compilation to specified batch sizes. | COMMENT |
| LOW | examples/offline_data_parallel.py | 172 | # Print the outputs. | COMMENT |
| LOW | examples/eplb/eplb_strategy.py | 48 | # Check if element is numeric | COMMENT |
| LOW | .github/workflows/pr_revert_command.yml | 218 | # Check if a revert PR already exists for this PR. | COMMENT |
| LOW | .github/workflows/pr_cherry_pick_command.yml | 201 | # Check if a cherry-pick PR already exists for this target branch. | COMMENT |
| LOW | .github/workflows/scripts/test_selector.py | 1651 | # Check if this test covers any line of this function | COMMENT |
| LOW | .github/workflows/scripts/test_selector.py | 1776 | # Check if this test covers any line of this function | COMMENT |
| LOW | .github/workflows/scripts/detect_po_changes.py | 154 | # Check if this is a multi-line table block. | COMMENT |
| LOW | .github/workflows/scripts/detect_po_changes.py | 441 | # Check if this is a modification of an existing paragraph | COMMENT |
| LOW | .github/ISSUE_TEMPLATE/400-bug-report.yml | 56 | # Print the outputs. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | tools/upload_to_openlibing.py | 64 | Upload files to OpenLibing OBS bucket. Args: file_paths: List of file paths to upload openlibi | STRING |
| HIGH | vllm_ascend/_310p/attention/attention_v1.py | 327 | Main dispatch method for attention operations. Routes the execution to Decode, Prefill, or Chunked Pre | STRING |
| HIGH | vllm_ascend/_310p/attention/attention_mask.py | 123 | Retrieves the appropriate attention mask based on the model configuration. When compressed mask is sup | STRING |
| HIGH | vllm_ascend/xlite/xlite.py | 664 | Look up and initialize the appropriate xlite model adapter based on the architecture specified in vLLM config and th | STRING |
| HIGH | vllm_ascend/xlite/xlite.py | 736 | Proxy unknown attributes to the wrapped runnable model. Args: key (str): The attribute name request | STRING |
| HIGH | …nsfer/kv_pool/ascend_store/backend/mooncake_backend.py | 584 | Parse storage size strings with support for units: GB, MB, KB, B Args: value: Input value (int, str, o | STRING |
| HIGH | …nsfer/kv_pool/ascend_store/backend/mooncake_backend.py | 630 | Convert numeric string to byte count Args: number_str: Numeric portion of input multiplier: Un | STRING |
| HIGH | …cend/quantization/methods/w4a4/w4a4_mxfp4_flatquant.py | 34 | Get decomposed dimensions for Kronecker quantization. Args: n: Dimension to decompose m: Tensor para | STRING |
| HIGH | …ascend/model_loader/netloader/executor/netloader_pg.py | 41 | Initializes a stateless process group. Args: host: Hostname. port: Port number. world_size: Size o | STRING |
| HIGH | vllm_ascend/ops/triton/query_gather_prep.py | 131 | Return the fused query as a contiguous [H, T, nope+rope] tensor. Args: ql_nope: [T, H, nope_dim] partial qu | STRING |
| HIGH | vllm_ascend/ops/triton/fla/chunk.py | 276 | Args: q (torch.Tensor): queries of shape `[B, T, H, K]` if `head_first=False` else `[B, H, T, K | STRING |
| HIGH | examples/save_sharded_state_310.py | 121 | Extract quantization description from JSON configuration file. Args: json_file: Path to the JSON confi | STRING |
| HIGH | …es/dynamic_bucket_load_balancer/hybrid_proxy_server.py | 113 | Split servers into num_groups groups. Args: servers (list): the server list to group. | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | tests/ut/proxy/test_load_balance_proxy_server.py | 137 | """Create a FastAPI app that serves as a mock prefill/decode backend. | STRING |
| LOW | …s/ut/distributed/weight_transfer/test_packed_tensor.py | 84 | # already implements ``__enter__``/``__exit__`` so just return it. | COMMENT |
| LOW | tests/ut/ops/test_vocab_parallel_embedding.py | 155 | # Should just pass through without masking | COMMENT |
| MEDIUM | …rd/rlhf/state_transitions/test_hccl_weight_transfer.py | 52 | # container: the harness hands the server `ASCEND_RT_VISIBLE_DEVICES` explicitly, | COMMENT |
| MEDIUM⚡ | tests/e2e/pull_request/one_card/rlhf/conftest.py | 74 | # Server harness | COMMENT |
| MEDIUM | …rlhf/state_transitions/test_npu_ipc_weight_transfer.py | 63 | # device index inside the container: the harness hands the server | COMMENT |
| LOW | vllm_ascend/attention/attention_v1.py | 886 | # just use getattr to avoid attribute error. | COMMENT |
| MEDIUM | vllm_ascend/core/profiling_chunk_predictor.py | 139 | # Keep a robust fallback for environments where least-squares may | COMMENT |
| MEDIUM | vllm_ascend/ops/triton/v2/apply_grammar_bitmask.py | 105 | # Disable MULTIBUFFER to better utilize the UB buffer, which gives | COMMENT |
| LOW | vllm_ascend/worker/block_table.py | 213 | # NOTE(woosuk): We can't simply use `token_indices // block_size` | COMMENT |
| LOW | vllm_ascend/worker/v2/model_runner.py | 209 | # so here we just call init_speculator to reinitialize speculator. | COMMENT |
| MEDIUM | …fill_v1/load_balance_proxy_layerwise_server_example.py | 83 | # - For production, ensure your backend servers are robust and secure. | COMMENT |
| MEDIUM | …egated_prefill_v1/load_balance_proxy_server_example.py | 111 | # - For production, ensure your backend servers are robust and secure. | COMMENT |
| MEDIUM | examples/disaggregated_encoder/disagg_epd_proxy.py | 244 | # for nixl connector to facilitate kv transfer... | COMMENT |
| MEDIUM | …ted/epd_load_balance_proxy_layerwise_server_example.py | 92 | # - For production, ensure your backend servers are robust and secure. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | tests/ut/attention/test_dsa_v41_cp.py | 121 | # Keep the real prepared-query/quantization orchestration; emulate kernels. | COMMENT |
| MEDIUM | tests/ut/models/test_glm5next_kpool_model_backend.py | 21 | # The operator PR owns these kernels. Each orchestration test below replaces | COMMENT |
| MEDIUM | tests/ut/ops/test_moe_comm_method.py | 438 | # Mock the unified MoE MLP orchestration returning (tensor, event). | COMMENT |
| MEDIUM | tests/ut/ops/test_moe_comm_method.py | 486 | # Verify the unified MoE MLP orchestration was called | COMMENT |
| MEDIUM | docs/source/user_guide/feature_guide/rl.md | 312 | ## Tool calling for agentic RL | COMMENT |
| MEDIUM | docs/source/tutorials/models/GLM5.md | 7 | [GLM-5](https://huggingface.co/zai-org/GLM-5) uses a Mixture-of-Experts (MoE) architecture and targets complex systems e | CODE |
| MEDIUM | docs/source/tutorials/models/GLM5.3-Flash.md | 5 | [GLM-5.3-Flash](https://huggingface.co/zai-org/GLM-5.3-Flash) is the first natively multimodal model in the GLM-5 series | CODE |
| MEDIUM | docs/source/tutorials/models/GLM5.2.md | 5 | [GLM-5.2](https://huggingface.co/zai-org/GLM-5.2) uses a Mixture-of-Experts (MoE) architecture and targets complex syste | CODE |
| MEDIUM | vllm_ascend/attention/dsa_v1.py | 1151 | # the trailing context window plus the whole current draft block. | COMMENT |
| MEDIUM | vllm_ascend/attention/attention_v1.py | 574 | # Note: Some embedding model weights are in float32 format. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | …ransfer/sparse_kv_offload/sparse_kv_offload_manager.py | 143 | # NOTE make sure to allocate k+v together and split them after allocate. | COMMENT |
| MEDIUM⚡ | …fill_v1/load_balance_proxy_layerwise_server_example.py | 27 | # For testing, you can use the provided mock server: | COMMENT |
| MEDIUM⚡ | …egated_prefill_v1/load_balance_proxy_server_example.py | 27 | # For testing, you can use the provided mock server: | COMMENT |
| MEDIUM⚡ | …ted/epd_load_balance_proxy_layerwise_server_example.py | 27 | # For testing, you can use the provided mock server: | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW⚡ | tests/ut/ops/test_rotary_embedding.py | 247 | dummy_key = mock_rope.call_args.args[2] | CODE |
| LOW⚡ | tests/ut/ops/test_rotary_embedding.py | 248 | assert dummy_key.shape == (query.shape[0], HEAD_SIZE) | CODE |
| LOW⚡ | tests/ut/ops/test_rotary_embedding.py | 249 | assert dummy_key.dtype == query.dtype | CODE |
| LOW⚡ | tests/ut/ops/test_rotary_embedding.py | 250 | assert dummy_key.device == query.device | CODE |
| LOW | tests/e2e/pull_request/one_card/test_mla_precision.py | 206 | for layer_name in ["placeholder"]: | CODE |
| LOW | tests/e2e/pull_request/one_card/test_mla_precision.py | 211 | ["placeholder"], | CODE |
| LOW | …e/pull_request/one_card/test_attention_v1_precision.py | 291 | ["placeholder"], | CODE |
| LOW | vllm_ascend/ops/rotary_embedding.py | 292 | dummy_key = ( | CODE |
| LOW | vllm_ascend/ops/rotary_embedding.py | 304 | dummy_key, | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | tests/e2e/pull_request/two_card/test_prefix_caching.py | 17 | # A prompt containing a large markdown table. The table is randomly generated by GPT-4. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | tests/ut/kv_offload/test_d2rh_pp_lifetime.py | 16 | def handle_request(self, meta): | CODE |
| LOW | tests/ut/kv_offload/test_mooncake_hybrid_connector.py | 196 | def handle_request(req_meta: dict[str, Any]): | CODE |
| LOW | tests/ut/kv_offload/test_mooncake_hybrid_connector.py | 238 | def handle_request(req_meta: dict[str, Any]): | CODE |
| LOW | tests/ut/kv_offload/test_mooncake_hybrid_connector.py | 308 | def handle_request(req_meta: dict[str, Any]): | CODE |
| LOW | …tributed/kv_transfer/kv_p2p/test_mooncake_connector.py | 915 | def handle_request(req_meta: dict[str, Any]): | CODE |
| LOW | …tributed/kv_transfer/kv_p2p/test_mooncake_connector.py | 969 | def handle_request(req_meta): | CODE |
| LOW | …tributed/kv_transfer/kv_p2p/test_mooncake_connector.py | 1016 | def handle_request(req_meta: dict[str, Any]): | CODE |
| LOW | tests/ut/_tools/test_doctest_helper.py | 15 | def helper(): | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | vllm_ascend/worker/model_runner_v1.py | 3122 | # TODO: implement PR 28597 from vllm | COMMENT |