Dockerfile.rocm

# Usage (to build SGLang ROCm docker image):
#   docker build --build-arg SGL_BRANCH=v0.4.9.post1 --build-arg GPU_ARCH=gfx942 -t v0.4.9.post1-rocm630-mi30x -f Dockerfile.rocm .
#   docker build --build-arg SGL_BRANCH=v0.4.9.post1 --build-arg GPU_ARCH=gfx950 -t v0.4.9.post1-rocm700-mi35x -f Dockerfile.rocm .

# Default base images
ARG BASE_IMAGE_950="rocm/sgl-dev:rocm7.0_preview_ubuntu_22.04_vllm_0.9.2_mi35X_prealpha"
ARG BASE_IMAGE_942="rocm/sgl-dev:vllm20250114"

# This is necessary for scope purpose
ARG GPU_ARCH=gfx950

# ===============================
# Base image 942 and args
FROM $BASE_IMAGE_942 AS gfx942
ENV BUILD_VLLM="0"
ENV BUILD_TRITON="1"
ENV BUILD_LLVM="0"
ENV BUILD_AITER_ALL="1"
ENV AITER_COMMIT="v0.1.4"
ENV NO_DEPS_FLAG=""

# ===============================
# Base image 950 and args
FROM $BASE_IMAGE_950 AS gfx950
ENV BUILD_VLLM="0"
ENV BUILD_TRITON="0"
ENV BUILD_AITER_ALL="1"
ENV BUILD_LLVM="1"
ENV AITER_COMMIT="v0.1.4"
ENV HIP_CLANG_PATH="/sgl-workspace/llvm-project/build/bin/"
ENV NO_DEPS_FLAG="--no-deps"

# ===============================
# Chosen arch and args
FROM ${GPU_ARCH}

# This is necessary for scope purpose, again
ARG GPU_ARCH=gfx950
ENV GPU_ARCH_LIST=${GPU_ARCH:-${PYTORCH_ROCM_ARCH}}

ARG SGL_REPO="https://github.com/sgl-project/sglang.git"
ARG SGL_DEFAULT="main"
ARG SGL_BRANCH=${SGL_DEFAULT}

ARG TRITON_REPO="https://github.com/ROCm/triton.git"
ARG TRITON_COMMIT="improve_fa_decode_3.0.0"

ARG AITER_REPO="https://github.com/ROCm/aiter.git"

ARG LLVM_REPO="https://github.com/jrbyrnes/llvm-project.git"
ARG LLVM_BRANCH="MainOpSelV2"
ARG LLVM_COMMIT="6520ace8227ffe2728148d5f3b9872a870b0a560"

USER root

# Install some basic utilities
RUN python -m pip install --upgrade pip && pip install setuptools_scm
RUN apt-get purge -y sccache; python -m pip uninstall -y sccache; rm -f "$(which sccache)"

WORKDIR /sgl-workspace

# -----------------------
# llvm
RUN if [ "$BUILD_LLVM" = "1" ]; then \
     git clone --single-branch ${LLVM_REPO} -b ${LLVM_BRANCH} \
     && cd llvm-project \
     && git checkout ${LLVM_COMMIT} \
     && mkdir build \
     && cd build \
     && cmake -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=1 -DLLVM_TARGETS_TO_BUILD="AMDGPU;X86" -DLLVM_ENABLE_PROJECTS="clang;lld;" -DLLVM_ENABLE_RUNTIMES="compiler-rt" ../llvm \
     && make -j$(nproc); \
    fi

# -----------------------

# -----------------------
# AITER
RUN pip uninstall -y aiter
RUN git clone ${AITER_REPO} \
 && cd aiter \
 && git checkout ${AITER_COMMIT} \
 && git submodule update --init --recursive
RUN cd aiter \
     && if [ "$BUILD_AITER_ALL" = "1" ] && [ "$BUILD_LLVM" = "1" ]; then \
          HIP_CLANG_PATH=/sgl-workspace/llvm-project/build/bin/ PREBUILD_KERNELS=1 GPU_ARCHS=$GPU_ARCH_LIST python setup.py develop; \
        elif [ "$BUILD_AITER_ALL" = "1" ]; then \
          PREBUILD_KERNELS=1 GPU_ARCHS=$GPU_ARCH_LIST python setup.py develop; \
        else \
          GPU_ARCHS=$GPU_ARCH_LIST python setup.py develop; \
        fi

# -----------------------
# Triton
RUN if [ "$BUILD_TRITON" = "1" ]; then \
        pip uninstall -y triton \
     && git clone ${TRITON_REPO} \
     && cd triton \
     && git checkout ${TRITON_COMMIT} \
     && cd python \
     && python setup.py install; \
    fi

# -----------------------
# Build vLLM
ARG VLLM_REPO="https://github.com/ROCm/vllm.git"
ARG VLLM_BRANCH="9f6b92db47c3444b7a7d67451ba0c3a2d6af4c2c"
RUN if [ "$BUILD_VLLM" = "1" ]; then \
        git clone ${VLLM_REPO} \
     && cd vllm \
     && git checkout ${VLLM_BRANCH} \
     && python -m pip install -r requirements/rocm.txt \
     && python setup.py clean --all \
     && python setup.py develop; \
    fi

# -----------------------
# Build SGLang
ARG BUILD_TYPE=all

RUN pip install IPython \
    && pip install orjson \
    && pip install python-multipart \
    && pip install torchao \
    && pip install pybind11

RUN pip uninstall -y sgl_kernel sglang
RUN git clone ${SGL_REPO} \
    && cd sglang \
    && if [ "${SGL_BRANCH}" = ${SGL_DEFAULT} ]; then \
         echo "Using ${SGL_DEFAULT}, default branch."; \
         git checkout ${SGL_DEFAULT}; \
       else \
         echo "Using ${SGL_BRANCH} branch."; \
         git checkout ${SGL_BRANCH}; \
       fi \
    && cd sgl-kernel \
    && rm -f pyproject.toml \
    && mv pyproject_rocm.toml pyproject.toml \
    && AMDGPU_TARGET=$GPU_ARCH_LIST python setup_rocm.py install \
    && cd .. \
    && if [ "$BUILD_TYPE" = "srt" ]; then \
         python -m pip --no-cache-dir install -e "python[srt_hip]" ${NO_DEPS_FLAG}; \
       else \
         python -m pip --no-cache-dir install -e "python[all_hip]" ${NO_DEPS_FLAG}; \
       fi

RUN python -m pip cache purge

# Copy config files to support MI300X in virtualized environments (MI300X_VF).  Symlinks will not be created in image build.
RUN find /sgl-workspace/sglang/python/sglang/srt/layers/quantization/configs/ \
         /sgl-workspace/sglang/python/sglang/srt/layers/moe/fused_moe_triton/configs/ \
         -type f -name '*MI300X*' | xargs -I {} sh -c 'vf_config=$(echo "$1" | sed "s/MI300X/MI300X_VF/"); cp "$1" "$vf_config"' -- {}

# Performance environment variable.
ENV HIP_FORCE_DEV_KERNARG=1
ENV HSA_NO_SCRATCH_RECLAIM=1
ENV SGLANG_SET_CPU_AFFINITY=1
ENV SGLANG_ALLOW_OVERWRITE_LONGER_CONTEXT_LEN=1
ENV NCCL_MIN_NCHANNELS=112

ENV SGLANG_USE_AITER=1
ENV SGLANG_MOE_PADDING=1
ENV VLLM_FP8_PADDING=1
ENV VLLM_FP8_ACT_PADDING=1
ENV VLLM_FP8_WEIGHT_PADDING=1
ENV VLLM_FP8_REDUCE_CONV=1
ENV TORCHINDUCTOR_MAX_AUTOTUNE=1
ENV TORCHINDUCTOR_MAX_AUTOTUNE_POINTWISE=1

CMD ["/bin/bash"]