fix broken deepep/flashmla install in container by adding `--no-build-isolation` (#12170)

docker/Dockerfile

@@ -138,7 +138,7 @@ RUN cd /sgl-workspace/DeepEP && \
         echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 \
         ;; \
     esac && \
-    NVSHMEM_DIR=${NVSHMEM_DIR} TORCH_CUDA_ARCH_LIST="${CHOSEN_TORCH_CUDA_ARCH_LIST}" pip install .
+    NVSHMEM_DIR=${NVSHMEM_DIR} TORCH_CUDA_ARCH_LIST="${CHOSEN_TORCH_CUDA_ARCH_LIST}" pip install --no-build-isolation .
 
 # Install flashmla
 RUN git clone https://github.com/deepseek-ai/FlashMLA.git flash-mla && \
@@ -148,7 +148,7 @@ RUN git clone https://github.com/deepseek-ai/FlashMLA.git flash-mla && \
       if [ "$CUDA_VERSION" = "12.6.1" ]; then \
         export FLASH_MLA_DISABLE_SM100=1; \
       fi && \
-      pip install -v . ;
+      pip install --no-build-isolation -v . ;
 
 # Python tools
 RUN python3 -m pip install --no-cache-dir \