sglangv0.5.2 & support Qwen3-Next-80B-A3B-Instruct

.github/workflows/release-pypi.yml

+name: Release PyPI
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - "python/sglang/version.py"
+  workflow_dispatch:
+
+jobs:
+  publish:
+    if: github.repository == 'sgl-project/sglang'
+    runs-on: ubuntu-latest
+    environment: "prod"
+    steps:
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.10"
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Upload to pypi
+        run: |
+          cd python
+          cp ../README.md ../LICENSE .
+          pip install build
+          python3 -m build
+          pip install twine
+          python3 -m twine upload dist/* -u __token__ -p ${{ secrets.PYPI_TOKEN }}

.github/workflows/release-whl-kernel-cu118.yml

+name: Release SGLang Kernel Wheel (cu118)
+
+on:
+  workflow_dispatch:
+    inputs:
+      tag_name:
+        type: string
+  push:
+    branches:
+      - main
+    paths:
+      - sgl-kernel/python/sgl_kernel/version.py
+
+jobs:
+  build-wheels:
+    if: github.repository == 'sgl-project/sglang'
+    runs-on: sgl-kernel-release-node
+    strategy:
+      matrix:
+        python-version: ["3.9"]
+        cuda-version: ["11.8"]
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build wheels for Python ${{ matrix.python-version }} and CUDA ${{ matrix.cuda-version }}
+        run: |
+          cd sgl-kernel
+          chmod +x ./build.sh
+          ./build.sh "${{ matrix.python-version }}" "${{ matrix.cuda-version }}"
+
+      - name: Upload artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: wheel-python${{ matrix.python-version }}-cuda${{ matrix.cuda-version }}
+          path: sgl-kernel/dist/*
+
+  release:
+    needs: build-wheels
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: sgl-kernel/dist/
+          merge-multiple: true
+          pattern: wheel-*
+
+      - name: Set tag name
+        id: set_tag_name
+        run: |
+          if [ -z "${{ inputs.tag_name }}" ]; then
+            TAG_NAME="v$(cat sgl-kernel/python/sgl_kernel/version.py | cut -d'"' -f2)"
+            echo "tag_name=$TAG_NAME" >> $GITHUB_OUTPUT
+          else
+            echo "tag_name=${{ inputs.tag_name }}" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ steps.set_tag_name.outputs.tag_name }}
+          repository: sgl-project/whl
+          token: ${{ secrets.WHL_TOKEN }}
+          files: |
+            sgl-kernel/dist/*
+
+      - name: Clone wheel index
+        run: git clone https://oauth2:${WHL_TOKEN}@github.com/sgl-project/whl.git sgl-whl
+        env:
+          WHL_TOKEN: ${{ secrets.WHL_TOKEN }}
+
+      - name: Update wheel index
+        run: python3 scripts/update_kernel_whl_index.py
+
+      - name: Push wheel index
+        run: |
+          cd sgl-whl
+          git config --local user.name "github-actions[bot]"
+          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
+          git add -A
+          git commit -m "update whl index"
+          git push

.github/workflows/release-whl-kernel.yml

+name: Release SGLang Kernels
+
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - sgl-kernel/python/sgl_kernel/version.py
+  workflow_dispatch:
+    inputs:
+      tag_name:
+        type: string
+        required: false
+
+concurrency:
+  group: release-sglang-kernels-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  build-cu129:
+    if: github.repository == 'sgl-project/sglang'
+    runs-on: sgl-kernel-release-node
+    strategy:
+      matrix:
+        python-version: ["3.10"]
+        cuda-version: ["12.9"]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build wheels
+        run: |
+          cd sgl-kernel
+          chmod +x ./build.sh
+          ./build.sh "${{ matrix.python-version }}" "${{ matrix.cuda-version }}"
+
+      - name: Upload to PyPI
+        working-directory: sgl-kernel
+        run: |
+          pip install twine
+          python3 -m twine upload dist/* -u __token__ -p ${{ secrets.PYPI_TOKEN }}
+
+  build-cu124:
+    if: github.repository == 'sgl-project/sglang'
+    needs: build-cu129
+    runs-on: sgl-kernel-release-node
+    strategy:
+      matrix:
+        python-version: ["3.10"]
+        cuda-version: ["12.4"]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build wheels
+        run: |
+          cd sgl-kernel
+          chmod +x ./build.sh
+          ./build.sh "${{ matrix.python-version }}" "${{ matrix.cuda-version }}"
+
+      - name: Upload artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: wheel-python${{ matrix.python-version }}-cuda${{ matrix.cuda-version }}
+          path: sgl-kernel/dist/*
+
+  release-cu124:
+    needs: build-cu124
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: sgl-kernel/dist/
+          merge-multiple: true
+          pattern: wheel-*
+
+      - name: Set tag name
+        id: set_tag_name
+        run: |
+          if [ -z "${{ inputs.tag_name }}" ]; then
+            TAG_NAME="v$(cat sgl-kernel/python/sgl_kernel/version.py | cut -d'"' -f2)"
+            echo "tag_name=$TAG_NAME" >> $GITHUB_OUTPUT
+          else
+            echo "tag_name=${{ inputs.tag_name }}" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ steps.set_tag_name.outputs.tag_name }}
+          repository: sgl-project/whl
+          token: ${{ secrets.WHL_TOKEN }}
+          files: |
+            sgl-kernel/dist/*
+
+      - name: Clone wheel index
+        run: git clone https://oauth2:${WHL_TOKEN}@github.com/sgl-project/whl.git sgl-whl
+        env:
+          WHL_TOKEN: ${{ secrets.WHL_TOKEN }}
+
+      - name: Update wheel index
+        run: python3 scripts/update_kernel_whl_index.py --cuda 124
+
+      - name: Push wheel index
+        run: |
+          cd sgl-whl
+          git config --local user.name "github-actions[bot]"
+          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
+          git add -A
+          git commit -m "update whl index"
+          git push
+
+  build-cu128:
+    if: github.repository == 'sgl-project/sglang'
+    needs: build-cu129
+    runs-on: sgl-kernel-release-node
+    strategy:
+      matrix:
+        python-version: ["3.10"]
+        cuda-version: ["12.8"]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build wheels
+        run: |
+          cd sgl-kernel
+          chmod +x ./build.sh
+          ./build.sh "${{ matrix.python-version }}" "${{ matrix.cuda-version }}"
+
+      - name: Upload artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: wheel-python${{ matrix.python-version }}-cuda${{ matrix.cuda-version }}
+          path: sgl-kernel/dist/*
+
+  release-cu128:
+    needs: build-cu128
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: sgl-kernel/dist/
+          merge-multiple: true
+          pattern: wheel-*
+
+      - name: Set tag name
+        id: set_tag_name
+        run: |
+          if [ -z "${{ inputs.tag_name }}" ]; then
+            TAG_NAME="v$(cat sgl-kernel/python/sgl_kernel/version.py | cut -d'"' -f2)"
+            echo "tag_name=$TAG_NAME" >> $GITHUB_OUTPUT
+          else
+            echo "tag_name=${{ inputs.tag_name }}" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ steps.set_tag_name.outputs.tag_name }}
+          repository: sgl-project/whl
+          token: ${{ secrets.WHL_TOKEN }}
+          files: |
+            sgl-kernel/dist/*
+
+      - name: Clone wheel index
+        run: git clone https://oauth2:${WHL_TOKEN}@github.com/sgl-project/whl.git sgl-whl
+        env:
+          WHL_TOKEN: ${{ secrets.WHL_TOKEN }}
+
+      - name: Update wheel index
+        run: python3 scripts/update_kernel_whl_index.py --cuda 128
+
+      - name: Push wheel index
+        run: |
+          cd sgl-whl
+          git config --local user.name "github-actions[bot]"
+          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
+          git add -A
+          git commit -m "update whl index"
+          git push
+
+  build-cu129-aarch64:
+    if: github.repository == 'sgl-project/sglang'
+    runs-on: sgl-kernel-release-node-arm
+    strategy:
+      matrix:
+        python-version: ["3.10"]
+        cuda-version: ["12.9"]
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build wheels
+        run: |
+          cd sgl-kernel
+          chmod +x ./build.sh
+          ./build.sh "${{ matrix.python-version }}" "${{ matrix.cuda-version }}" aarch64
+
+      - name: Upload artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: wheel-python${{ matrix.python-version }}-cuda${{ matrix.cuda-version }}-aarch64
+          path: sgl-kernel/dist/*
+
+  release-cu129-aarch64:
+    needs: build-cu129-aarch64
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: sgl-kernel/dist/
+          merge-multiple: true
+          pattern: wheel-*
+
+      - name: Set tag name
+        id: set_tag_name
+        run: |
+          if [ -z "${{ inputs.tag_name }}" ]; then
+            TAG_NAME="v$(cat sgl-kernel/python/sgl_kernel/version.py | cut -d'"' -f2)"
+            echo "tag_name=$TAG_NAME" >> $GITHUB_OUTPUT
+          else
+            echo "tag_name=${{ inputs.tag_name }}" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ steps.set_tag_name.outputs.tag_name }}
+          repository: sgl-project/whl
+          token: ${{ secrets.WHL_TOKEN }}
+          files: |
+            sgl-kernel/dist/*
+
+      - name: Clone wheel index
+        run: git clone https://oauth2:${WHL_TOKEN}@github.com/sgl-project/whl.git sgl-whl
+        env:
+          WHL_TOKEN: ${{ secrets.WHL_TOKEN }}
+
+      - name: Update wheel index
+        run: python3 scripts/update_kernel_whl_index.py --cuda 129
+
+      - name: Push wheel index
+        run: |
+          cd sgl-whl
+          git config --local user.name "github-actions[bot]"
+          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
+          git add -A
+          git commit -m "update whl index"
+          git push

.github/workflows/vllm-dependency-test.yml

+name: VLLM Dependency Test
+
+on:
+  push:
+    branches: [ main ]
+    paths:
+      - "python/**"
+      - "scripts/ci/**"
+      - "test/**"
+  pull_request:
+    branches: [ main ]
+    paths:
+      - "python/**"
+      - "scripts/ci/**"
+      - "test/**"
+
+concurrency:
+  group: vllm-dependency-test-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  vllm-dependency-test:
+    if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') &&
+        github.event.pull_request.draft == false
+    runs-on: 1-gpu-runner
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Install dependencies
+        run: |
+          bash scripts/ci/ci_install_dependency.sh
+          pip install "bitsandbytes>=0.44.0"
+
+          pip install "sgl-kernel==0.3.7"
+
+      - name: Run vLLM dependency tests
+        timeout-minutes: 60
+        run: |
+          export SGLANG_SKIP_SGL_KERNEL_VERSION_CHECK=1
+
+          cd test/srt
+          python3 run_suite.py --suite vllm_dependency_test --timeout-per-file 3600

.gitignore

+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+
+# Tokenizer cache for tests
+.tokenizer_cache/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+.idea/
+
+# MacOS
+.DS_Store
+
+# Vim
+*.swp
+
+# Documentation
+docs/_build
+
+# SGL
+benchmark/mmlu/data
+benchmark/mmlu/data.tar
+benchmark/llava_bench/images
+benchmark/llava_bench/mme_pack
+*.jsonl
+tmp*.txt
+
+# Plots
+*.png
+*.pdf
+
+# personnal
+work_dirs/
+*.csv
+
+!logo.png
+
+# Prerequisites
+*.d
+
+# Compiled Object files
+*.slo
+*.lo
+*.o
+*.obj
+
+# Precompiled Headers
+*.gch
+*.pch
+
+# Compiled Dynamic libraries
+*.so
+*.dylib
+*.dll
+
+# Fortran module files
+*.mod
+*.smod
+
+# Compiled Static libraries
+*.lai
+*.la
+*.a
+*.lib
+
+# Executables
+*.exe
+*.out
+*.app
+
+compile_commands.json
+
+*.iml
+
+# VSCode
+.vscode
+
+1
+
+# Autoenv
+.env.leave
+
+# Rust lib
+Cargo.lock
+
+lmms-eval

.isort.cfg

+[settings]
+profile=black
+known_first_party=sglang

.pre-commit-config.yaml

+default_stages: [pre-commit, pre-push, manual]
+
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+      - id: check-symlinks
+      - id: destroyed-symlinks
+      - id: trailing-whitespace
+      - id: end-of-file-fixer
+      - id: check-yaml
+        args: [--allow-multiple-documents]
+      - id: check-toml
+      - id: check-ast
+      - id: check-added-large-files
+      - id: check-merge-conflict
+      - id: check-shebang-scripts-are-executable
+      - id: detect-private-key
+      - id: debug-statements
+      - id: no-commit-to-branch
+  - repo: https://github.com/PyCQA/isort
+    rev: 5.13.2
+    hooks:
+      - id: isort
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.11.7
+    hooks:
+      - id: ruff
+        args: [--select=F401, --fixable=F401]
+        files: ^(benchmark/|docs/|examples/)
+        exclude: \.ipynb$
+  - repo: https://github.com/psf/black
+    rev: 24.10.0
+    hooks:
+      - id: black-jupyter
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.4.1
+    hooks:
+      - id: codespell
+        additional_dependencies: ['tomli']
+        args: ['--toml', 'python/pyproject.toml', '-L', 'cann,thi,makro,wil,rouge']
+        exclude: |
+          (?x)^(
+            test/srt/test_reasoning_parser\.py|
+            docs/advanced_features/vlm_query\.ipynb
+          )$
+  - repo: https://github.com/pre-commit/mirrors-clang-format
+    rev: v18.1.8
+    hooks:
+    - id: clang-format
+      types_or: [c++, cuda]
+      args: [--style=file, --verbose]
+  - repo: https://github.com/kynan/nbstripout
+    rev: 0.8.1
+    hooks:
+      - id: nbstripout
+        args:
+          - '--keep-output'
+          - '--extra-keys=metadata.kernelspec metadata.language_info.version'

3rdparty/amd/profiling/client.sh

+#!/bin/bash
+
+# Start profiling via API
+curl http://localhost:30000/start_profile -H "Content-Type: application/json"
+
+# Benchmark serving using sglang with random dataset and tokenizer
+# Define the log file with a timestamp
+TIMESTAMP=$(date +%Y%m%d_%H%M%S)
+LOGFILE="sglang_client_log_$TIMESTAMP.json"
+
+# Run the benchmark with specified parameters and save logs
+python3 -m sglang.bench_serving \
+    --backend sglang \
+    --tokenizer Xenova/grok-1-tokenizer \
+    --dataset-name random \
+    --random-input 1024\
+    --random-output 1024 \
+    --num-prompts 240 \
+    --request-rate 8 \
+    --output-file online.jsonl 2>&1 | tee "$LOGFILE"
+
+# Stop profiling via API
+curl http://localhost:30000/stop_profile -H "Content-Type: application/json"
+
+# Convert tracing file to csv & json
+sqlite3 trace.rpd ".mode csv" ".header on" ".output trace.csv" "select * from top;" ".output stdout"
+python3 /sgl-workspace/rocmProfileData/tools/rpd2tracing.py trace.rpd trace.json

3rdparty/amd/profiling/install_rpd.sh

+# download and install RPD
+apt update && apt install -y sqlite3 libsqlite3-dev libfmt-dev
+
+# install rpd module
+git clone https://github.com/ROCmSoftwarePlatform/rocmProfileData
+cd rocmProfileData
+git apply rpd.patch
+make && make install
+cd rocpd_python && python setup.py install && cd ..
+cd rpd_tracer && make clean;make install && python setup.py install && cd ..

3rdparty/amd/profiling/loadTracer.sh

+#!/bin/bash
+################################################################################
+# Copyright (c) 2021 - 2023 Advanced Micro Devices, Inc. All rights reserved.
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+################################################################################
+OUTPUT_FILE="trace.rpd"
+
+if [ "$1" = "-o" ] ; then
+  OUTPUT_FILE=$2
+  shift
+  shift
+fi
+
+if [ -e ${OUTPUT_FILE} ] ; then
+  rm ${OUTPUT_FILE}
+fi
+
+python3 -m rocpd.schema --create ${OUTPUT_FILE}
+if [ $? != 0 ] ; then
+  echo "Error: Could not create rpd file. Please run 'python setup.py install' from the rocpd_python dir"
+  exit
+fi
+
+export RPDT_FILENAME=${OUTPUT_FILE}
+export RPDT_AUTOSTART=0
+LD_PRELOAD=librocm-smi_64:librpd_tracer.so "$@"

3rdparty/amd/profiling/rpd.patch

+diff --git a/rpd_tracer/Makefile b/rpd_tracer/Makefile
+index e9d9feb..b2e9e1a 100644
+--- a/rpd_tracer/Makefile
++++ b/rpd_tracer/Makefile
+@@ -16,7 +16,7 @@ ifneq (,$(HIP_PATH))
+         $(info Building with roctracer)
+         RPD_LIBS += -L/opt/rocm/lib -lroctracer64 -lroctx64 -lamdhip64 -lrocm_smi64
+         RPD_INCLUDES += -I/opt/rocm/include -I/opt/rocm/include/roctracer -I/opt/rocm/include/hsa
+-        RPD_SRCS += RoctracerDataSource.cpp RocmSmiDataSource.cpp
++        RPD_SRCS += RoctracerDataSource.cpp
+         RPD_INCLUDES += -D__HIP_PLATFORM_AMD__
+ endif

3rdparty/amd/profiling/rpd_profile_server_enable.patch

+diff --git a/python/sglang/srt/managers/scheduler.py b/python/sglang/srt/managers/scheduler.py
+index 62d1ff9..9021c01 100644
+--- a/python/sglang/srt/managers/scheduler.py
++++ b/python/sglang/srt/managers/scheduler.py
+@@ -71,6 +71,8 @@ from sglang.srt.utils import (
+     suppress_other_loggers,
+ )
+ from sglang.utils import get_exception_traceback
++from rpdTracerControl import rpdTracerControl
++rpdTracerControl.skipCreate()
+
+ logger = logging.getLogger(__name__)
+
+@@ -245,6 +247,7 @@ class Scheduler:
+                 ],
+                 with_stack=True,
+             )
++            self.rpd = rpdTracerControl()
+
+     @torch.inference_mode()
+     def event_loop(self):
+@@ -1027,15 +1030,24 @@ class Scheduler:
+     def start_profile(self) -> None:
+         if self.profiler is None:
+             raise RuntimeError("Profiler is not enabled.")
+-        self.profiler.start()
++        #self.profiler.start() #block pytorch profiler for rpd profiler enabling
++        if self.tp_rank == 0 or self.tp_rank == 1:
++            self.rpd.start()
++            self.rpd.rangePush("", "rpd profile range", "")
++            logger.info("rpd is enabled")
+
+     def stop_profile(self) -> None:
+         if self.profiler is None:
+             raise RuntimeError("Profiler is not enabled.")
+-        self.profiler.stop()
+-        self.profiler.export_chrome_trace(
+-            self.torch_profiler_trace_dir + "/" + str(time.time()) + ".trace.json.gz"
+-        )
++        #self.profiler.stop()
++        #self.profiler.export_chrome_trace(
++        #    self.torch_profiler_trace_dir + "/" + str(time.time()) + ".trace.json.gz"
++        #)
++        if self.tp_rank ==0 or self.tp_rank ==1:
++            self.rpd.rangePop()
++            self.rpd.stop()
++            self.rpd.flush()
++            logger.info("rpd is done")
+         logger.info("Profiler is done")

3rdparty/amd/profiling/rpd_profile_server_enable_wCPU_activities.patch

+diff --git a/python/sglang/srt/managers/scheduler.py b/python/sglang/srt/managers/scheduler.py
+index 62d1ff9..2edb427 100644
+--- a/python/sglang/srt/managers/scheduler.py
++++ b/python/sglang/srt/managers/scheduler.py
+@@ -71,6 +71,8 @@ from sglang.srt.utils import (
+     suppress_other_loggers,
+ )
+ from sglang.utils import get_exception_traceback
++from rpdTracerControl import rpdTracerControl
++rpdTracerControl.skipCreate()
+
+ logger = logging.getLogger(__name__)
+
+@@ -245,6 +247,7 @@ class Scheduler:
+                 ],
+                 with_stack=True,
+             )
++            self.rpd = rpdTracerControl()
+
+     @torch.inference_mode()
+     def event_loop(self):
+@@ -1027,15 +1030,26 @@ class Scheduler:
+     def start_profile(self) -> None:
+         if self.profiler is None:
+             raise RuntimeError("Profiler is not enabled.")
+-        self.profiler.start()
++        #self.profiler.start()
++        logger.info("torch profiler is disabled")
++        if self.tp_rank == 0 or self.tp_rank == 1:
++            self.rpd.setPythonTrace(True)
++            self.rpd.start()
++            self.rpd.rangePush("", "scheduler", "")
++        logger.info("rpd is enabled inside scheduler profiling")
+
+     def stop_profile(self) -> None:
+         if self.profiler is None:
+             raise RuntimeError("Profiler is not enabled.")
+-        self.profiler.stop()
+-        self.profiler.export_chrome_trace(
+-            self.torch_profiler_trace_dir + "/" + str(time.time()) + ".trace.json.gz"
+-        )
++        #self.profiler.stop()
++        #self.profiler.export_chrome_trace(
++        #    self.torch_profiler_trace_dir + "/" + str(time.time()) + ".trace.json.gz"
++        #)
++        if self.tp_rank ==0 or self.tp_rank ==1:
++            self.rpd.rangePop()
++            self.rpd.stop()
++            self.rpd.flush()
++            logger.info("rpd is done inside scheduler")
+         logger.info("Profiler is done")
+
+
+diff --git a/python/sglang/srt/managers/tokenizer_manager.py b/python/sglang/srt/managers/tokenizer_manager.py
+index 2621ccd..181df85 100644
+--- a/python/sglang/srt/managers/tokenizer_manager.py
++++ b/python/sglang/srt/managers/tokenizer_manager.py
+@@ -58,6 +58,10 @@ from sglang.srt.sampling.sampling_params import SamplingParams
+ from sglang.srt.server_args import PortArgs, ServerArgs
+ from sglang.srt.utils import is_generation_model, is_multimodal_model
+
++from rpdTracerControl import rpdTracerControl
++rpdTracerControl.skipCreate()
++
++
+ asyncio.set_event_loop_policy(uvloop.EventLoopPolicy())
+
+ logger = logging.getLogger(__name__)
+@@ -514,10 +518,20 @@ class TokenizerManager:
+         self.send_to_scheduler.send_pyobj(req)
+
+     def start_profile(self):
++        rpd = rpdTracerControl()
++        rpd.setPythonTrace(True)
++        rpd.start()
++        rpd.rangePush("", "tokenizer_manager", "")
++        logger.info("tokenizer_manager rpd profiling started!")
+         req = ProfileReq.START_PROFILE
+         self.send_to_scheduler.send_pyobj(req)
+
+     def stop_profile(self):
++        rpd = rpdTracerControl()
++        rpd.rangePop()
++        rpd.stop()
++        rpd.flush()
++        logger.info("rpd profiling is done inside tokenizer_manager!")
+         req = ProfileReq.STOP_PROFILE
+         self.send_to_scheduler.send_pyobj(req)
+
+diff --git a/python/sglang/srt/server.py b/python/sglang/srt/server.py
+index 7111c93..2bd722c 100644
+--- a/python/sglang/srt/server.py
++++ b/python/sglang/srt/server.py
+@@ -30,6 +30,8 @@ import threading
+ import time
+ from http import HTTPStatus
+ from typing import Dict, List, Optional, Union
++from rpdTracerControl import rpdTracerControl
++rpdTracerControl.skipCreate()
+
+ # Fix a bug of Python threading
+ setattr(threading, "_register_atexit", lambda *args, **kwargs: None)
+@@ -152,6 +154,11 @@ async def flush_cache():
+ @app.post("/start_profile")
+ async def start_profile():
+     """Start profiling."""
++    rpd = rpdTracerControl()
++    rpd.setPythonTrace(True)
++    rpd.start()
++    rpd.rangePush("", "server rpd profile range", "")
++    logger.info("rpd profiling started in server.py!")
+     tokenizer_manager.start_profile()
+     return Response(
+         content="Start profiling.\n",
+@@ -164,6 +171,11 @@ async def start_profile():
+ async def stop_profile():
+     """Stop profiling."""
+     tokenizer_manager.stop_profile()
++    rpd = rpdTracerControl()
++    rpd.rangePop()
++    rpd.stop()
++    rpd.flush()
++    logger.info("rpd profiling is done in server.py!")
+     return Response(
+         content="Stop profiling. This will take some time.\n",
+         status_code=200,

3rdparty/amd/profiling/server.sh

+#!/bin/bash
+
+# export SGLANG_TORCH_PROFILER_DIR=/data/sglang/
+export SGLANG_TORCH_PROFILER_DIR=/sgl-workspace/sglang/profile/
+
+# Get the current timestamp
+TIMESTAMP=$(date +"%Y%m%d_%H%M%S")
+
+# Define the log file with a timestamp
+LOGFILE="sglang_server_log_$TIMESTAMP.json"
+
+# Run the Python command and save the output to the log file
+loadTracer.sh python3 -m sglang.launch_server \
+    --model-path /sgl-workspace/sglang/dummy_grok1 \
+    --tokenizer-path Xenova/grok-1-tokenizer \
+    --load-format dummy \
+    --quantization fp8 \
+    --tp 8 \
+    --port 30000 \
+    --disable-radix-cache 2>&1 | tee "$LOGFILE"

3rdparty/amd/profiling/torch_profiler.patch

+diff --git a/python/sglang/srt/managers/scheduler.py b/python/sglang/srt/managers/scheduler.py
+index 62d1ff9..6ecd78c 100644
+--- a/python/sglang/srt/managers/scheduler.py
++++ b/python/sglang/srt/managers/scheduler.py
+@@ -240,7 +240,6 @@ class Scheduler:
+             )
+             self.profiler = torch.profiler.profile(
+                 activities=[
+-                    torch.profiler.ProfilerActivity.CPU,
+                     torch.profiler.ProfilerActivity.CUDA,
+                 ],
+                 with_stack=True,
+@@ -1033,9 +1032,11 @@ class Scheduler:
+         if self.profiler is None:
+             raise RuntimeError("Profiler is not enabled.")
+         self.profiler.stop()
+-        self.profiler.export_chrome_trace(
+-            self.torch_profiler_trace_dir + "/" + str(time.time()) + ".trace.json.gz"
+-        )
++        if self.tp_rank == 0:
++            with open(f"stats_repro_{int(time.time())}.txt", "w") as f:
++                print(self.profiler.key_averages(group_by_input_shape=True).table(sort_by="cuda_time_total", row_limit=-1), file=f)
++                print("Profiling stats done.")
++
+         logger.info("Profiler is done")

Makefile

+.PHONY: check-deps install-deps format update help
+
+# Show help for each target
+help:
+	@echo "Available targets:"
+	@grep -E '^[a-zA-Z0-9_-]+:.*?## .*$$' $(MAKEFILE_LIST) | sort | awk 'BEGIN {FS = ":.*?## "}; {printf "\033[36m%-20s\033[0m %s\n", $$1, $$2}'
+
+check-deps: ## Check and install required Python formatting dependencies
+	@command -v isort >/dev/null 2>&1 || (echo "Installing isort..." && pip install isort)
+	@command -v black >/dev/null 2>&1 || (echo "Installing black..." && pip install black)
+
+install-deps: ## Install Python formatting tools (isort and black)
+	pip install isort black
+
+format: check-deps ## Format modified Python files using isort and black
+	@echo "Formatting modified Python files..."
+	git diff --name-only --diff-filter=M | grep '\.py$$' | xargs -I {} sh -c 'isort {} && black {}'
+
+FILES_TO_UPDATE = docker/Dockerfile.rocm \
+                 python/pyproject.toml \
+                 python/sglang/version.py \
+                 docs/developer_guide/setup_github_runner.md \
+                 docs/get_started/install.md \
+                 docs/platforms/amd_gpu.md \
+                 docs/platforms/ascend_npu.md \
+				 benchmark/deepseek_v3/README.md
+
+update: ## Update version numbers across project files. Usage: make update <new_version>
+	@if [ -z "$(filter-out $@,$(MAKECMDGOALS))" ]; then \
+		echo "Version required. Usage: make update <new_version>"; \
+		exit 1; \
+	fi
+	@OLD_VERSION=$$(grep "version" python/sglang/version.py | cut -d '"' -f2); \
+	NEW_VERSION=$(filter-out $@,$(MAKECMDGOALS)); \
+	echo "Updating version from $$OLD_VERSION to $$NEW_VERSION"; \
+	for file in $(FILES_TO_UPDATE); do \
+		if [ "$(shell uname)" = "Darwin" ]; then \
+			sed -i '' -e "s/$$OLD_VERSION/$$NEW_VERSION/g" $$file; \
+		else \
+			sed -i -e "s/$$OLD_VERSION/$$NEW_VERSION/g" $$file; \
+		fi \
+	done; \
+	echo "Version update complete"
+
+%:
+	@: