feat: support build tabby on windows (#948)

* feat: update config to support build on windows

* resolve comment

* update release.yml

* resolve comment

.github/workflows/release.yml

@@ -26,7 +26,7 @@ jobs:
     container: ${{ matrix.container }}
     strategy:
       matrix:
-        binary: [aarch64-apple-darwin, x86_64-manylinux2014, x86_64-manylinux2014-cuda117]
+        binary: [aarch64-apple-darwin, x86_64-manylinux2014, x86_64-manylinux2014-cuda117, x86_64-pc-windows-msvc]
         include:
           - os: macos-latest
             target: aarch64-apple-darwin
@@ -40,6 +40,11 @@ jobs:
             binary: x86_64-manylinux2014-cuda117
             container: sameli/manylinux2014_x86_64_cuda_11.7
             build_args: --features cuda
+          - os: windows-latest
+            target: x86_64-pc-windows-msvc
+            binary: x86_64-pc-windows-msvc
+            ext: .exe
+            build_args: --features cuda
 
     env:
       SCCACHE_GHA_ENABLED: true
@@ -77,20 +82,34 @@ jobs:
             ~/.cargo/registry
             ~/.cargo/git
 
-      - run: bash ./ci/prepare_build_environment.sh
+      - name: Prepare build environment for macOS & Linux
+        run: bash ./ci/prepare_build_environment.sh
+        if: runner.os != 'Windows'
+
+      - name: Prepare build environment for Windows
+        run: ./ci/prepare_build_environment.ps1
+        if: runner.os == 'Windows'
+
+      - name: Install CUDA toolkit for Windows
+        uses: Jimver/cuda-toolkit@v0.2.11
+        with:
+          cuda: '11.7.1'
+          method: 'network'
+          sub-packages: '["nvcc", "cudart", "cublas", "cublas_dev", "thrust", "visual_studio_integration"]'
+        if: runner.os == 'Windows'
 
       - name: Bulid release binary
         run: cargo build ${{ matrix.build_args }} --release --target ${{ matrix.target }} --package tabby
 
       - name: Rename release binary
-        run: mv target/${{ matrix.target }}/release/tabby tabby_${{ matrix.binary }}
+        run: mv target/${{ matrix.target }}/release/tabby${{ matrix.ext }} tabby_${{ matrix.binary }}${{ matrix.ext }}
 
       - name: Upload artifacts
         uses: actions/upload-artifact@v3
         with:
           retention-days: 3
-          name: tabby_${{ matrix.binary }}
-          path: tabby_${{ matrix.binary }}
+          name: tabby_${{ matrix.binary }}${{ matrix.ext }}
+          path: tabby_${{ matrix.binary }}${{ matrix.ext }}
 
   pre-release:
     if: github.event_name == 'push'

ci/prepare_build_environment.ps1

@@ -0,0 +1 @@
+choco install --yes protoc

crates/llama-cpp-bindings/build.rs

@@ -10,11 +10,16 @@ fn main() {
         "Please init submodules with `git submodule update --init --recursive` and try again"
     );
 
-    println!("cargo:rerun-if-changed=cc/*.h");
-    println!("cargo:rerun-if-changed=cc/*.cc");
+    println!("cargo:rerun-if-changed=include/engine.h");
+    println!("cargo:rerun-if-changed=src/engine.cc");
     println!("cargo:rustc-link-lib=llama");
     println!("cargo:rustc-link-lib=ggml_static");
 
+    build_llama_cpp();
+    build_cxx_binding();
+}
+
+fn build_llama_cpp() {
     let mut config = Config::new("llama.cpp");
     if cfg!(target_os = "macos") {
         config.define("LLAMA_METAL", "ON");
@@ -26,9 +31,16 @@ fn main() {
     if cfg!(feature = "cuda") {
         config.define("LLAMA_CUBLAS", "ON");
         config.define("CMAKE_POSITION_INDEPENDENT_CODE", "ON");
-        println!("cargo:rustc-link-search=native=/usr/local/cuda/lib64");
+        if cfg!(target_os = "windows") {
+            let Ok(cuda_path) = env::var("CUDA_PATH") else {
+                panic!("CUDA_PATH is not set");
+            };
+            println!(r"cargo:rustc-link-search=native={}\lib\x64", cuda_path);
+        } else {
+            println!("cargo:rustc-link-search=native=/usr/local/cuda/lib64");
+            println!("cargo:rustc-link-lib=culibos");
+        }
         println!("cargo:rustc-link-lib=cudart");
-        println!("cargo:rustc-link-lib=culibos");
         println!("cargo:rustc-link-lib=cublas");
         println!("cargo:rustc-link-lib=cublasLt");
     }
@@ -69,9 +81,26 @@ fn main() {
         println!("cargo:rustc-link-lib=hipblas");
     }
 
+    // By default, this value is automatically inferred from Rust’s compilation profile.
+    // For Windows platform, we always build llama.cpp in release mode.
+    // See https://github.com/TabbyML/tabby/pull/948 for more details.
+    if cfg!(target_os = "windows") {
+        config.profile("Release");
+    }
+
     let dst = config.build();
-    println!("cargo:rustc-link-search=native={}/build", dst.display());
+    if cfg!(target_os = "windows") {
+        println!(
+            r"cargo:rustc-link-search=native={}\build\{}",
+            dst.display(),
+            config.get_profile()
+        );
+    } else {
+        println!("cargo:rustc-link-search=native={}/build", dst.display());
+    }
+}
 
+fn build_cxx_binding() {
     cxx_build::bridge("src/lib.rs")
         .file("src/engine.cc")
         .flag_if_supported("-Iinclude")

crates/llama-cpp-bindings/src/engine.cc

@@ -1,5 +1,3 @@
-#include "engine.h"
-
 #include <functional>
 #include <vector>
 #include <deque>
@@ -9,6 +7,7 @@
 #include <ggml.h>
 #include <llama.h>
 
+#include "llama-cpp-bindings/include/engine.h"
 #include "llama-cpp-bindings/src/lib.rs.h"
 
 namespace llama {

crates/tabby-common/Cargo.toml

@@ -5,6 +5,7 @@ edition = "2021"
 
 [dependencies]
 filenamify = "0.1.0"
+home = "0.5.5"
 lazy_static = { workspace = true }
 serde = { workspace = true }
 serdeconv = { workspace = true }

crates/tabby-common/src/path.rs

@@ -6,7 +6,7 @@ lazy_static! {
     static ref TABBY_ROOT: Mutex<Cell<PathBuf>> = {
         Mutex::new(Cell::new(match env::var("TABBY_ROOT") {
             Ok(x) => PathBuf::from(x),
-            Err(_) => PathBuf::from(env::var("HOME").unwrap()).join(".tabby"),
+            Err(_) => home::home_dir().unwrap().join(".tabby"),
         }))
     };
 }