Update CUDA architecture list in build pipeline for 12.9.1 wheels (#26592)

Signed-off-by: Will Eaton <wseaton@users.noreply.github.com>
Signed-off-by: simon-mo <simon.mo@hey.com>

.buildkite/release-pipeline.yaml

@@ -48,7 +48,7 @@ steps:
     agents:
       queue: cpu_queue_postmerge
     commands:
-      - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --build-arg torch_cuda_arch_list='7.0 7.5 8.0 8.9 9.0+PTX' --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ."
+      - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ."
       - "mkdir artifacts"
       - "docker run --rm -v $(pwd)/artifacts:/artifacts_host vllm-ci:build-image bash -c 'cp -r dist /artifacts_host && chmod -R a+rw /artifacts_host'"
       - "bash .buildkite/scripts/upload-wheels.sh"

docker/Dockerfile

@@ -13,8 +13,13 @@ ARG PYTHON_VERSION=3.12
 # private registries that use a different repository naming conventions.
 #
 # Example:
-# docker build --build-arg BUILD_BASE_IMAGE=registry.acme.org/mirror/nvidia/cuda:${CUDA_VERSION}-devel-ubuntu22.04
-ARG BUILD_BASE_IMAGE=nvidia/cuda:${CUDA_VERSION}-devel-ubuntu22.04
+# docker build --build-arg BUILD_BASE_IMAGE=registry.acme.org/mirror/nvidia/cuda:${CUDA_VERSION}-devel-ubuntu20.04
+
+# Important: We build with an old version of Ubuntu to maintain broad 
+# compatibility with other Linux OSes. The main reason for this is that the
+# glibc version is baked into the distro, and binaries built with one glibc
+# version are not backwards compatible with OSes that use an earlier version.
+ARG BUILD_BASE_IMAGE=nvidia/cuda:${CUDA_VERSION}-devel-ubuntu20.04
 # TODO: Restore to base image after FlashInfer AOT wheel fixed
 ARG FINAL_BASE_IMAGE=nvidia/cuda:${CUDA_VERSION}-devel-ubuntu22.04
 
@@ -75,34 +80,19 @@ ARG TARGETPLATFORM
 ARG INSTALL_KV_CONNECTORS=false
 ENV DEBIAN_FRONTEND=noninteractive
 
-ARG DEADSNAKES_MIRROR_URL
-ARG DEADSNAKES_GPGKEY_URL
 ARG GET_PIP_URL
 
-# Install Python and other dependencies
+# Install system dependencies and uv, then create Python virtual environment
 RUN echo 'tzdata tzdata/Areas select America' | debconf-set-selections \
     && echo 'tzdata tzdata/Zones/America select Los_Angeles' | debconf-set-selections \
     && apt-get update -y \
-    && apt-get install -y ccache software-properties-common git curl sudo \
-    && if [ ! -z ${DEADSNAKES_MIRROR_URL} ] ; then \
-        if [ ! -z "${DEADSNAKES_GPGKEY_URL}" ] ; then \
-            mkdir -p -m 0755 /etc/apt/keyrings ; \
-            curl -L ${DEADSNAKES_GPGKEY_URL} | gpg --dearmor > /etc/apt/keyrings/deadsnakes.gpg ; \
-            sudo chmod 644 /etc/apt/keyrings/deadsnakes.gpg ; \
-            echo "deb [signed-by=/etc/apt/keyrings/deadsnakes.gpg] ${DEADSNAKES_MIRROR_URL} $(lsb_release -cs) main" > /etc/apt/sources.list.d/deadsnakes.list ; \
-        fi ; \
-    else \
-        for i in 1 2 3; do \
-            add-apt-repository -y ppa:deadsnakes/ppa && break || \
-            { echo "Attempt $i failed, retrying in 5s..."; sleep 5; }; \
-        done ; \
-    fi \
-    && apt-get update -y \
-    && apt-get install -y python${PYTHON_VERSION} python${PYTHON_VERSION}-dev python${PYTHON_VERSION}-venv \
-    && update-alternatives --install /usr/bin/python3 python3 /usr/bin/python${PYTHON_VERSION} 1 \
-    && update-alternatives --set python3 /usr/bin/python${PYTHON_VERSION} \
-    && ln -sf /usr/bin/python${PYTHON_VERSION}-config /usr/bin/python3-config \
-    && curl -sS ${GET_PIP_URL} | python${PYTHON_VERSION} \
+    && apt-get install -y ccache software-properties-common git curl sudo python3-pip \
+    && curl -LsSf https://astral.sh/uv/install.sh | sh \
+    && $HOME/.local/bin/uv venv /opt/venv --python ${PYTHON_VERSION} \
+    && rm -f /usr/bin/python3 /usr/bin/python3-config /usr/bin/pip \
+    && ln -s /opt/venv/bin/python3 /usr/bin/python3 \
+    && ln -s /opt/venv/bin/python3-config /usr/bin/python3-config \
+    && ln -s /opt/venv/bin/pip /usr/bin/pip \
     && python3 --version && python3 -m pip --version
 
 ARG PIP_INDEX_URL UV_INDEX_URL
@@ -111,9 +101,9 @@ ARG PYTORCH_CUDA_INDEX_BASE_URL
 ARG PYTORCH_CUDA_NIGHTLY_INDEX_BASE_URL
 ARG PIP_KEYRING_PROVIDER UV_KEYRING_PROVIDER
 
-# Install uv for faster pip installs
-RUN --mount=type=cache,target=/root/.cache/uv \
-    python3 -m pip install uv
+# Activate virtual environment and add uv to PATH
+ENV PATH="/opt/venv/bin:/root/.local/bin:$PATH"
+ENV VIRTUAL_ENV="/opt/venv"
 
 # This timeout (in seconds) is necessary when installing some dependencies via uv since it's likely to time out
 # Reference: https://github.com/astral-sh/uv/pull/1694
@@ -142,7 +132,7 @@ WORKDIR /workspace
 COPY requirements/common.txt requirements/common.txt
 COPY requirements/cuda.txt requirements/cuda.txt
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv pip install --system -r requirements/cuda.txt \
+    uv pip install --python /opt/venv/bin/python3 -r requirements/cuda.txt \
     --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 
 # cuda arch list used by torch
@@ -172,7 +162,7 @@ ENV UV_INDEX_STRATEGY="unsafe-best-match"
 ENV UV_LINK_MODE=copy
 
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv pip install --system -r requirements/build.txt \
+    uv pip install --python /opt/venv/bin/python3 -r requirements/build.txt \
     --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 
 COPY . .
@@ -269,7 +259,7 @@ COPY requirements/lint.txt requirements/lint.txt
 COPY requirements/test.txt requirements/test.txt
 COPY requirements/dev.txt requirements/dev.txt
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv pip install --system -r requirements/dev.txt \
+    uv pip install --python /opt/venv/bin/python3 -r requirements/dev.txt \
     --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 #################### DEV IMAGE ####################
 

vllm/v1/attention/backends/flashinfer.py

@@ -29,7 +29,6 @@ from vllm.utils.flashinfer import (can_use_trtllm_attention,
                                    flashinfer_disable_q_quantization,
                                    supports_trtllm_attention,
                                    use_trtllm_attention)
-from vllm.v1.attention.backends.flash_attn import use_cascade_attention
 # yapf conflicts with isort for this block
 # yapf: disable
 from vllm.v1.attention.backends.utils import (AttentionCGSupport,
@@ -677,7 +676,9 @@ class FlashInferMetadataBuilder(AttentionMetadataBuilder[FlashInferMetadata]):
             # TODO: The cascade wrapper currently does not support setting
             # kv cache dtype to something different from query dtype.
             return False
-        return use_cascade_attention(*args, **kwargs)
+        # TODO: Cascade attention doesn't work, disable it for now
+        # return use_cascade_attention(*args, **kwargs)
+        return False
 
 
 class FlashInferImpl(AttentionImpl):