From d42201e235f476657672c3bba37f83194dcb2c01 Mon Sep 17 00:00:00 2001 From: Xiwen Yu <13230610+VALLIS-NERIA@users.noreply.github.com> Date: Mon, 8 Sep 2025 10:24:52 +0800 Subject: [PATCH] remove waivers and cleanup Signed-off-by: Xiwen Yu <13230610+VALLIS-NERIA@users.noreply.github.com> --- bringup_fix.sh | 49 ------------------- docker/Dockerfile.multi | 3 -- jenkins/L0_MergeRequest.groovy | 6 +-- .../singlegpu/test_ad_build_small_single.py | 1 - .../unit/singlegpu/test_ad_trtllm_bench.py | 1 - .../library/test_rope_transformation.py | 1 - .../_torch/modeling/test_modeling_mllama.py | 4 -- 7 files changed, 3 insertions(+), 62 deletions(-) delete mode 100644 bringup_fix.sh diff --git a/bringup_fix.sh b/bringup_fix.sh deleted file mode 100644 index 3ebd6eb441..0000000000 --- a/bringup_fix.sh +++ /dev/null @@ -1,49 +0,0 @@ -# ARCH=$(uname -m) -# if [ $ARCH == "x86_64" ]; then - -# wget https://urm.nvidia.com/artifactory/sw-gpu-cuda-installer-generic-local/packaging/r13.0/cuda_nvrtc/linux-x86_64/13.0.48/cuda-nvrtc-dev-13-0_13.0.48-1_amd64.deb && \ -# dpkg -i cuda-nvrtc-dev-13-0_13.0.48-1_amd64.deb && \ -# rm cuda-nvrtc-dev-13-0_13.0.48-1_amd64.deb - -# wget https://github.com/Kitware/CMake/releases/download/v4.0.3/cmake-4.0.3-linux-x86_64.sh && \ -# bash cmake-4.0.3-linux-x86_64.sh --skip-license --prefix=/usr/local/cmake --exclude-subdir - -# apt update -# apt install -y libstdc++-14-dev - -# elif [ $ARCH == "aarch64" ]; then - -# # to be moved to docker/common/ scripts -# wget https://urm.nvidia.com/artifactory/sw-gpu-cuda-installer-generic-local/packaging/r13.0/cuda_nvrtc/linux-sbsa/13.0.48/cuda-nvrtc-dev-13-0_13.0.48-1_arm64.deb && \ -# dpkg -i cuda-nvrtc-dev-13-0_13.0.48-1_arm64.deb && \ -# rm cuda-nvrtc-dev-13-0_13.0.48-1_arm64.deb - -# wget https://github.com/Kitware/CMake/releases/download/v4.0.3/cmake-4.0.3-linux-aarch64.sh && \ -# bash cmake-4.0.3-linux-aarch64.sh --skip-license --prefix=/usr/local/cmake --exclude-subdir - -# apt update -# # fix LLVM build -# apt install -y libstdc++-14-dev - -# # wait for https://github.com/NVIDIA/TensorRT-LLM/pull/6588 -# pip install deep_gemm@git+https://github.com/VALLIS-NERIA/DeepGEMM.git@97d97a20c2ecd53a248ab64242219d780cf822b8 --no-build-isolation - -# else -# echo "Unsupported architecture: $ARCH" -# exit 1 -# fi - -# # wait for new triton to be published -# cd /usr/local/lib/python3.12/dist-packages/ && \ -# ls -la | grep pytorch_triton && \ -# mv pytorch_triton-3.3.1+gitc8757738.dist-info triton-3.3.1+gitc8757738.dist-info && \ -# cd triton-3.3.1+gitc8757738.dist-info && \ -# echo "Current directory: $(pwd)" && \ -# echo "Files in directory:" && \ -# ls -la && \ -# sed -i 's/^Name: pytorch-triton/Name: triton/' METADATA && \ -# sed -i 's|pytorch_triton-3.3.1+gitc8757738.dist-info/|triton-3.3.1+gitc8757738.dist-info/|g' RECORD && \ -# echo "METADATA after update:" && \ -# grep "^Name:" METADATA - -# # pip install git+https://github.com/triton-lang/triton.git@main diff --git a/docker/Dockerfile.multi b/docker/Dockerfile.multi index c5179449e3..45e93071e9 100644 --- a/docker/Dockerfile.multi +++ b/docker/Dockerfile.multi @@ -75,9 +75,6 @@ RUN pip3 uninstall -y opencv && \ rm -rf /usr/local/lib/python3*/dist-packages/cv2/ && \ pip3 install opencv-python-headless --force-reinstall --no-deps --no-cache-dir -# COPY bringup_fix.sh bringup_fix.sh -# RUN bash ./bringup_fix.sh && rm bringup_fix.sh - # WARs against security issues inherited from pytorch:25.06 # * https://github.com/advisories/GHSA-8qvm-5x2c-j2w7 RUN pip3 install --upgrade --no-cache-dir \ diff --git a/jenkins/L0_MergeRequest.groovy b/jenkins/L0_MergeRequest.groovy index 2b519d572e..f7dbeffa6b 100644 --- a/jenkins/L0_MergeRequest.groovy +++ b/jenkins/L0_MergeRequest.groovy @@ -372,9 +372,9 @@ def preparation(pipeline, testFilter, globalVars) stage("Setup Environment") { setupPipelineEnvironment(pipeline, testFilter, globalVars) } - // stage("Merge Test Waive List") { - // mergeWaiveList(pipeline, globalVars) - // } + stage("Merge Test Waive List") { + mergeWaiveList(pipeline, globalVars) + } }) } diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_build_small_single.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_build_small_single.py index 273a50123d..3a226eee42 100644 --- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_build_small_single.py +++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_build_small_single.py @@ -41,7 +41,6 @@ def _check_ad_config(experiment_config: ExperimentConfig, llm_args: LlmArgs): ) -@pytest.mark.skip("https://nvbugswb.nvidia.com/NVBugs5/redir.aspx?url=/5443039") @pytest.mark.parametrize( "experiment_config", [ diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py index d22580ecc0..2d4f1959c3 100644 --- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py +++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py @@ -799,7 +799,6 @@ def trtllm_bench_unified_comparison( ) -@pytest.mark.skip("https://nvbugswb.nvidia.com/NVBugs5/redir.aspx?url=/5443039") def test_trtllm_bench(llm_root): # noqa: F811 model_path_or_name, model_name, model_path = tiny_llama_details() diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_rope_transformation.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_rope_transformation.py index 4a15eddfa5..ffa2c0ccd8 100644 --- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_rope_transformation.py +++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_rope_transformation.py @@ -184,7 +184,6 @@ class RoPEModel(torch.nn.Module): ], ) @torch.inference_mode() -@pytest.mark.skip("https://nvbugswb.nvidia.com/NVBugs5/redir.aspx?url=/5443039") def test_rope_variants( transformation, variant, diff --git a/tests/unittest/_torch/modeling/test_modeling_mllama.py b/tests/unittest/_torch/modeling/test_modeling_mllama.py index 1d044579fd..1ecfd39661 100644 --- a/tests/unittest/_torch/modeling/test_modeling_mllama.py +++ b/tests/unittest/_torch/modeling/test_modeling_mllama.py @@ -2,7 +2,6 @@ import re import unittest from copy import deepcopy -import pytest import torch from _torch.helpers import create_mock_engine from parameterized import parameterized @@ -10,7 +9,6 @@ from test_modeling_llama import Scenario, reduce_llama_config from transformers import MllamaConfig from transformers import \ MllamaForConditionalGeneration as HFMllamaForConditionalGeneration -from utils.util import getSMVersion import tensorrt_llm from tensorrt_llm._torch.attention_backend.utils import get_attention_backend @@ -291,8 +289,6 @@ class TestMLlama(unittest.TestCase): """ Compare output to HF """ - if scenario.backend == "FLASHINFER" and getSMVersion() == 120: - pytest.skip("https://nvbugspro.nvidia.com/bug/5458945") backend = scenario.backend metadata_cls = get_attention_backend(backend).Metadata