Build QNN backend with ExecuTorch Failed

[luffy-yu]

• The following command raised an error:

Use of 'EXECUTORCH_BUILD_EXTENSION_MODULE' requires 'EXECUTORCH_BUILD_EXTENSION_NAMED_DATA_MAP'

cmake -DPYTHON_EXECUTABLE=python \
    -DCMAKE_INSTALL_PREFIX=cmake-out \
    -DEXECUTORCH_ENABLE_LOGGING=1 \
    -DCMAKE_BUILD_TYPE=Release \
    -DEXECUTORCH_BUILD_EXTENSION_MODULE=ON \
    -DEXECUTORCH_BUILD_EXTENSION_DATA_LOADER=ON \
    -DEXECUTORCH_BUILD_EXTENSION_TENSOR=ON \
    -DEXECUTORCH_BUILD_QNN=ON \
    -DQNN_SDK_ROOT=${QNN_SDK_ROOT} \
    -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \
    -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
    -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
    -Bcmake-out .

• After that, Setup Llama Runner also raise an error.

Fix:

cmake -DPYTHON_EXECUTABLE=python \
    -DCMAKE_INSTALL_PREFIX=cmake-out \
    -DEXECUTORCH_ENABLE_LOGGING=1 \
    -DCMAKE_BUILD_TYPE=Release \
    -DEXECUTORCH_BUILD_EXTENSION_MODULE=ON \
    -DEXECUTORCH_BUILD_EXTENSION_NAMED_DATA_MAP=ON \
    -DEXECUTORCH_BUILD_EXTENSION_DATA_LOADER=ON \
    -DEXECUTORCH_BUILD_EXTENSION_TENSOR=ON \
    -DEXECUTORCH_BUILD_EXTENSION_LLM=ON \
    -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON \
    -DEXECUTORCH_BUILD_QNN=ON \
    -DQNN_SDK_ROOT=${QNN_SDK_ROOT} \
    -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \
    -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
    -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
    -Bcmake-out .