#!/bin/bash
set -eux

python_executable=python$1
cuda_home=/usr/local/cuda-$2

# Update paths
PATH=${cuda_home}/bin:$PATH
LD_LIBRARY_PATH=${cuda_home}/lib64:$LD_LIBRARY_PATH

# Install requirements
if [ "$(echo "$2" | cut -d. -f1)" = "12" ]; then
    sed -i 's/^nvidia-cutlass-dsl\[cu13\]>=/nvidia-cutlass-dsl>=/' requirements/cuda.txt
fi
$python_executable -m pip install -r requirements/build/cuda.txt -r requirements/cuda.txt

# Limit the number of parallel jobs to avoid OOM
export MAX_JOBS=1
# Make sure release wheels are built for the following architectures
# Do not add +PTX here: vLLM filters torch's top-level PTX flag when it
# converts global gencode flags into per-kernel arch lists. If a specific
# kernel needs PTX, add +PTX to that kernel's CMake arch list instead.
export TORCH_CUDA_ARCH_LIST="7.5 8.0 8.6 8.9 9.0 10.0 12.0"

bash tools/check_repo.sh

# Build
$python_executable setup.py bdist_wheel --dist-dir=dist