Redesign Docker build system to two-stage architecture with builder/runtime separation

Redesigned the Docker build system from a single-stage monolithic design to a clean
two-stage architecture that separates build environment from compilation process while
maintaining library path compatibility.

## Architecture Changes

### Builder Image (docker/builder/Dockerfile)
- Provides base environment: CUDA 11.4, GCC 10, CMake 4, Go 1.25.3
- Built once, cached for subsequent builds (~90 min first time)
- Removed config file copying (cuda-11.4.sh, gcc-10.conf, go.sh)
- Added comprehensive comments explaining each build step
- Added git installation for runtime stage source cloning

### Runtime Image (docker/runtime/Dockerfile)
- Two-stage build using ollama37-builder as base for BOTH stages
- Stage 1 (compile): Clone source from GitHub → CMake configure → Build C/C++/CUDA → Build Go
- Stage 2 (runtime): Copy artifacts from stage 1 → Setup environment → Configure server
- Both stages use identical base image to ensure library path compatibility
- Removed -buildvcs=false flag (VCS info embedded from git clone)
- Comprehensive comments documenting library paths and design rationale

### Makefile (docker/Makefile)
- Simplified from 289 to 145 lines (-50% complexity)
- Removed: run, stop, logs, shell, test targets (use docker-compose instead)
- Removed: build orchestration targets (start-builder, copy-source, run-cmake, etc.)
- Removed: artifact copying (handled internally by multi-stage build)
- Focus: Build images only (build, build-builder, build-runtime, clean, help)
- All runtime operations delegated to docker-compose.yml

### Documentation (docker/README.md)
- Completely rewritten for new two-stage architecture
- Added "Build System Components" section with file structure
- Documented why both runtime stages use builder base (library path compatibility)
- Updated build commands to use Makefile
- Updated runtime commands to use docker-compose
- Added comprehensive troubleshooting section
- Added build time and image size tables
- Reference to archived single-stage design

## Key Design Decision

**Problem**: Compiled binaries have hardcoded library paths
**Solution**: Use ollama37-builder as base for BOTH compile and runtime stages
**Trade-off**: Larger image (~18GB) vs guaranteed library compatibility

## Benefits

- ✅ Cleaner separation of concerns (builder env vs compilation vs runtime)
- ✅ Builder image cached after first build (90 min → <1 min rebuilds)
- ✅ Runtime rebuilds only take ~10 min (pulls latest code from GitHub)
- ✅ No library path mismatches (identical base images)
- ✅ No complex artifact extraction (multi-stage COPY)
- ✅ Simpler Makefile focused on image building
- ✅ Runtime management via docker-compose (industry standard)

## Files Changed

Modified:
- docker/builder/Dockerfile - Added comments, removed COPY config files
- docker/runtime/Dockerfile - Converted to two-stage build
- docker/Makefile - Simplified to focus on image building only
- docker/README.md - Comprehensive rewrite for new architecture

Deleted:
- docker/builder/README.md - No longer needed
- docker/builder/cuda-11.4.sh - Generated in Dockerfile
- docker/builder/gcc-10.conf - Generated in Dockerfile
- docker/builder/go.sh - Generated in Dockerfile

Archived:
- docker/Dockerfile → docker/Dockerfile.single-stage.archived

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

docker/Dockerfile.single-stage.archived

@@ -0,0 +1,96 @@
+FROM rockylinux/rockylinux:8
+
+# Install CUDA toolkit 11.4
+# Note: NVIDIA driver is NOT needed in builder container - only CUDA toolkit for compilation
+# The host system provides the driver at runtime via --gpus flag
+# dnf-plugins-core is required for the config-manager command
+
+RUN dnf -y install dnf-plugins-core\
+    && dnf -y config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/rhel8/x86_64/cuda-rhel8.repo\
+    && dnf -y install cuda-toolkit-11-4
+
+# Setup CUDA path
+RUN echo 'export PATH="${PATH}:/usr/local/cuda-11.4/bin"' > /etc/profile.d/cuda-11.4.sh
+ENV PATH="$PATH:/usr/local/cuda-11.4/bin"
+
+# Install gcc 10
+RUN dnf -y install wget unzip bzip2 git\
+    && dnf -y groupinstall "Development Tools"\
+    && cd /usr/local/src\
+    && wget https://github.com/gcc-mirror/gcc/archive/refs/heads/releases/gcc-10.zip\
+    && unzip gcc-10.zip\
+    && cd gcc-releases-gcc-10\
+    && contrib/download_prerequisites\
+    && mkdir /usr/local/gcc-10\
+    && cd /usr/local/gcc-10\
+    && /usr/local/src/gcc-releases-gcc-10/configure --disable-multilib\
+    && make -j $(nproc)\
+    && make install
+
+# Setup GCC 10 library path
+RUN echo '/usr/local/lib64' > /etc/ld.so.conf.d/gcc-10.conf\
+    && ldconfig\
+    && rm -f /usr/bin/cc\
+    && ln -s /usr/local/bin/gcc /usr/bin/cc
+
+# Install cmake 4
+RUN dnf -y install openssl-devel\
+    && cd /usr/local/src\
+    && wget https://github.com/Kitware/CMake/releases/download/v4.0.0/cmake-4.0.0.tar.gz\
+    && tar xvf cmake-4.0.0.tar.gz\
+    && mkdir /usr/local/cmake-4\
+    && cd /usr/local/cmake-4\
+    && /usr/local/src/cmake-4.0.0/configure\
+    && make -j $(nproc)\
+    && make install
+
+# Install go 1.25.3
+RUN cd /usr/local\
+    && wget https://go.dev/dl/go1.25.3.linux-amd64.tar.gz\
+    && tar xvf go1.25.3.linux-amd64.tar.gz
+
+# Setup Go path
+RUN echo 'export PATH="${PATH}:/usr/local/go/bin"' > /etc/profile.d/go.sh
+ENV PATH="$PATH:/usr/local/go/bin"
+
+# Clone ollama37 source code
+RUN cd /usr/local/src\
+    && git clone https://github.com/dogkeeper886/ollama37.git
+
+# Build ollama37
+WORKDIR /usr/local/src/ollama37
+
+# Configure with CMake (CUDA 11 preset for Tesla K80 compute capability 3.7)
+RUN bash -c 'LD_LIBRARY_PATH=/usr/local/lib:/usr/local/lib64:/usr/lib64:$LD_LIBRARY_PATH \
+    CC=/usr/local/bin/gcc CXX=/usr/local/bin/g++ \
+    cmake --preset "CUDA 11"'
+
+# Build C/C++/CUDA libraries
+RUN bash -c 'LD_LIBRARY_PATH=/usr/local/lib:/usr/local/lib64:/usr/lib64:$LD_LIBRARY_PATH \
+    CC=/usr/local/bin/gcc CXX=/usr/local/bin/g++ \
+    cmake --build build -j$(nproc)'
+
+# Build Go binary
+RUN go build -buildvcs=false -o /usr/local/bin/ollama .
+
+# Setup library paths for runtime
+# Include: Ollama CUDA/GGML libraries, GCC 10 runtime, CUDA 11.4 libraries
+ENV LD_LIBRARY_PATH=/usr/local/src/ollama37/build/lib/ollama:/usr/local/lib64:/usr/local/cuda-11.4/lib64:/usr/lib64
+
+# Ollama server configuration
+ENV OLLAMA_HOST=0.0.0.0:11434
+
+# Expose the Ollama API port
+EXPOSE 11434
+
+# Create data directory for models
+RUN mkdir -p /root/.ollama
+VOLUME ["/root/.ollama"]
+
+# Health check
+HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
+    CMD /usr/local/bin/ollama list || exit 1
+
+# Set entrypoint and default command
+ENTRYPOINT ["/usr/local/bin/ollama"]
+CMD ["serve"]