Bump llama version to c22473b580807929fd9e3a3344a48e8cfbe6c88f (#355)

* feat(llama.cpp): update README and server implementation for token handling

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

---------

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

Author: ilopezluna

llamacpp/native/README.md

@@ -11,20 +11,68 @@
 
 ## Bumping llama.cpp version
 
-1. Pull and checkout the desired llama.cpp version:
+### Overview
 
-```
-pushd vendor/llama.cpp
-git fetch origin
-git checkout <desired llama.cpp sha> # usually we bump to the latest tagged commit
-popd
-```
+This project uses llama.cpp as a git submodule located at `vendor/llama.cpp`, which points to the official llama.cpp repository at https://github.com/ggml-org/llama.cpp.git.
 
-2. Apply our llama-server patch:
+The project applies custom patches to llama.cpp's server implementation (`server.cpp` and `utils.hpp`) to integrate with the Docker model-runner architecture. These patches are maintained in `src/server/server.patch`.
 
+### Prerequisites
+
+Before bumping the version, ensure the submodule is initialized. **Run this command from the project root directory** (e.g., `/path/to/model-runner`):
+
+```bash
+# From the project root directory
+git submodule update --init --recursive
 ```
-make -C src/server clean
-make -C src/server
-```
 
-3. Make sure everyting builds cleanly following the update.
+If the submodule is already initialized, this command is safe to run and will ensure it's up to date.
+
+**Note:** This command must be executed from the repository root because it needs to access the `.gitmodules` file and the submodule paths are relative to the root directory.
+
+### Step-by-Step Process
+
+1. **Find the desired llama.cpp version:**
+
+   Visit https://github.com/ggml-org/llama.cpp/releases to find the latest stable release or a specific version you want to use. We typically bump to the latest tagged commit (e.g., `b1234`, `b2345`, etc.).
+
+2. **Update the submodule to the desired version:**
+
+   ```bash
+   pushd vendor/llama.cpp
+   git fetch origin
+   git checkout <desired llama.cpp sha> # usually we bump to the latest tagged commit
+   popd
+   ```
+
+3. **Apply the custom llama-server patch:**
+
+   ```bash
+   make -C src/server clean
+   make -C src/server
+   ```
+
+   This will:
+   - Clean the previous patched files
+   - Copy the new `server.cpp` and `utils.hpp` from the updated llama.cpp
+   - Apply our custom patches from `src/server/server.patch`
+
+4. **Build and test:**
+
+   ```bash
+   # Build from the native directory   
+    cmake -B build
+    cmake --build build --parallel 8 --config Release
+   
+   # Test the build
+   ./build/bin/com.docker.llama-server --model <path to model>
+   ```
+
+   Make sure everything builds cleanly without errors.
+
+5. **Commit the submodule update:**
+
+   ```bash
+   git add vendor/llama.cpp
+   git commit -m "Bump llama.cpp to <version>"
+   ```