Integrate G‑ALP Floating‑Point Compression Scheme

.clang-format

@@ -34,3 +34,15 @@ CompactNamespaces: true
 BinPackArguments: false
 BinPackParameters: false
 
+# CUDA-specific additions
+AttributeMacros:
+  - "__global__"
+  - "__device__"
+  - "__host__"
+  - "__constant__"
+  - "__shared__"
+  - "__managed__"
+  - "__restrict__"
+  - "__launch_bounds__"
+  - "__forceinline__"
+  - "__noinline__"

.github/CODEOWNERS

@@ -0,0 +1 @@
+/.github/workflows/*    @azimafroozeh

.github/workflows/galp.yaml

@@ -0,0 +1,114 @@
+# .github/workflows/galp.yml
+
+name: GALP
+
+run-name: >-
+  ${{ github.workflow }} on ${{ github.ref_name }} (#${{ github.run_number }})
+  — ${{ github.event.pull_request.title || github.event.head_commit.message }}
+
+on:
+  push:
+
+permissions:
+  contents: read
+
+jobs:
+  president-build-test:
+    if: ${{ github.actor == 'azimafroozeh' }}
+    name: CMake CI on president
+    runs-on:
+      - self-hosted
+      - Linux
+      - X64
+      - president
+
+    steps:
+      # ────────────────────────────────────────────────
+      # 0. Checkout
+      # ────────────────────────────────────────────────
+      - name: Checkout source
+        uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      # ────────────────────────────────────────────────
+      # 0a. Generate synthetic dataset
+      # ────────────────────────────────────────────────
+      - name: Generate synthetic dataset
+        uses: ./.github/actions/generate-dataset
+
+      # ────────────────────────────────────────────────
+      # 1. Quick sanity of the runner
+      # ────────────────────────────────────────────────
+      - name: 🔎 Env dump
+        run: |
+          set -x
+          hostname
+          echo "Runner label: president"
+          echo "CUDA path: $CUDA_LIBRARY_PATH"
+          echo "NVCOMP_INC: $NVCOMP_INC"
+          echo "NVCOMP_LIB: $NVCOMP_LIB"
+          which nvidia-smi || true
+          which nvcc       || true
+          which clang++    || true
+          which g++        || true
+          which cmake      || true
+          which make       || true
+
+      # ────────────────────────────────────────────────
+      # 2. Verify required tools
+      # ────────────────────────────────────────────────
+      - name: ✅ Verify tools
+        run: |
+          for cmd in nvidia-smi nvcc clang++ g++ cmake make; do
+            command -v "$cmd" >/dev/null || { echo "::error::$cmd missing"; exit 127; }
+          done
+
+      - name: Version outputs
+        run: |
+          nvidia-smi
+          nvcc     --version
+          clang++  --version
+          g++      --version
+          cmake    --version
+          make     --version
+
+      # ────────────────────────────────────────────────
+      # 3. Configure with CMake
+      # ────────────────────────────────────────────────
+      - name: 🔄 CMake configure
+        run: |
+          cmake -S . -B build \
+            -DCMAKE_BUILD_TYPE=Release \
+            -DFLS_ENABLE_GALP_TESTING_AND_BENCHMARKING=ON
+
+      # ────────────────────────────────────────────────
+      # 4. Build (all targets)
+      # ────────────────────────────────────────────────
+      - name: 🏗️ Build
+        run: cmake --build build --parallel
+
+      # ────────────────────────────────────────────────
+      # 5. Run unit tests
+      # ────────────────────────────────────────────────
+      - name: 🧪 Run tests
+        run: |
+          cd build
+          ctest --output-on-failure
+
+      # ────────────────────────────────────────────────
+      # 6. Run benchmarks (no label filter)
+      # ────────────────────────────────────────────────
+      - name: ⚡ Run benchmarks
+        run: |
+          cd build
+          ctest --output-on-failure || true
+
+      # ────────────────────────────────────────────────
+      # 7. Smoke‑test helper binaries
+      # ────────────────────────────────────────────────
+      - name: 🚀 Smoke‑test binaries
+        run: |
+          ./build/galp/benchmark/micro-benchmarks       --help || true
+          ./build/galp/benchmark/compressors-benchmarks --help || true
+        shell: bash

CMakeLists.txt

@@ -108,7 +108,7 @@ endif ()
 option(FLS_BUILD_TESTING "Build Test" OFF)
 option(FLS_BUILD_BENCHMARKING "Enable Benchmark Build" OFF)
 option(FLS_BUILD_EXAMPLES "Build Examples" OFF)
-option(FLS_BUILD_GPU "Build GPU" OFF)
+option(FLS_BUILD_CUDA "Build FastLanes CUDA reader" OFF)
 option(FLS_BUILD_GTEST "Build GTEST" OFF)
 option(FLS_BUILD_PYTHON "Build FastLanes Python bindings" OFF)
 option(FLS_BUILD_SHARED_LIBS "Build FastLanes library as shared" OFF)
@@ -120,13 +120,13 @@ option(FLS_ENABLE_TABLE_LOG "Enable Table Log" OFF)
 option(FLS_ENABLE_VERBOSE_OUTPUT "Enable verbose output" OFF)
 option(FLS_ENABLE_DATA "Enable Data" OFF)
 option(FLS_ENABLE_FSST_TESTING_AND_BENCHMARKING "Build FSST" OFF)
+option(FLS_ENABLE_GALP_TESTING_AND_BENCHMARKING "Build GALP" OFF)
 
 # Definitions: ---------------------------------------------------------------------------------------------------------
 add_compile_definitions(FLS_CMAKE_SOURCE_DIR="${CMAKE_CURRENT_SOURCE_DIR}")
 
-
 # GTEST : --------------------------------------------------------------------------------------------------------------
-if (FLS_BUILD_TESTING OR FLS_BUILD_GPU OR FLS_ENABLE_FSST_TESTING_AND_BENCHMARKING)
+if (FLS_BUILD_TESTING OR FLS_BUILD_CUDA OR FLS_ENABLE_FSST_TESTING_AND_BENCHMARKING OR FLS_ENABLE_GALP_TESTING_AND_BENCHMARKING)
     message("---------------------------------------------------------------------------------------------------------")
     message("-- FLS: Building GTEST:")
     include(GoogleTest)
@@ -182,8 +182,21 @@ if (true)
 
 endif ()
 
+# CUDA : ---------------------------------------------------------------------------------------------------------------
+if (FLS_BUILD_CUDA)
+    message("---------------------------------------------------------------------------------------------------------")
+    message("--: Building CU-FastLanes Lib.")
+    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/cuda)
+endif ()
+
+
 # DATA : ---------------------------------------------------------------------------------------------------------------
-if (FLS_BUILD_TESTING OR FLS_BUILD_BENCHMARKING OR FLS_BUILD_EXAMPLES OR FLS_ENABLE_DATA)
+if (FLS_BUILD_TESTING OR
+        FLS_BUILD_BENCHMARKING OR
+        FLS_BUILD_EXAMPLES OR
+        FLS_ENABLE_DATA OR
+        FLS_ENABLE_GALP_TESTING_AND_BENCHMARKING OR
+        FLS_ENABLE_FSST_TESTING_AND_BENCHMARKING)
     message("---------------------------------------------------------------------------------------------------------")
     message("-- FLS: Enabling Data.")
 
@@ -276,21 +289,13 @@ if (FLS_ENABLE_FSST_TESTING_AND_BENCHMARKING)
     add_subdirectory(fsst)
 endif ()
 
-
 # Example : ------------------------------------------------------------------------------------------------------------
 if (FLS_BUILD_EXAMPLES)
     message("---------------------------------------------------------------------------------------------------------")
     message("- Examples:")
     add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/examples)
 endif ()
 
-# GPU : ----------------------------------------------------------------------------------------------------------------
-if (FLS_BUILD_GPU)
-    message("---------------------------------------------------------------------------------------------------------")
-    message("- GPU:")
-    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/gpu)
-endif ()
-
 # Playground : ---------------------------------------------------------------------------------------------------------
 if (EXISTS "${CMAKE_SOURCE_DIR}/playground/CMakeLists.txt")
     message("---------------------------------------------------------------------------------------------------------")
@@ -305,4 +310,9 @@ if (FLS_BUILD_PYTHON)
     add_subdirectory(python)   # your module’s CMakeLists.txt
 endif ()
 
-
+# FSST: ----------------------------------------------------------------------------------------------------------------
+if (FLS_ENABLE_GALP_TESTING_AND_BENCHMARKING)
+    message("---------------------------------------------------------------------------------------------------------")
+    message("-- FLS: Adding GALP Testing and Benchmarking.")
+    add_subdirectory(galp)
+endif ()

README.md

@@ -8,9 +8,10 @@
 
 FastLanes features:
 
-- **Fully SIMD-friendly with zero explicit SIMD instructions**:
-- **Zero dependencies**:
-    - no external libraries required
+* **Fully SIMD-friendly with zero explicit SIMD instructions**
+* **Zero dependencies**:
+
+    * no external libraries required
 
 ---
 
@@ -97,7 +98,7 @@ fn main() -> Result<()> {
 
 ## Documentation
 
-- [Typing System](docs/typing_system.md)
+* [Typing System](docs/typing_system.md)
 
 ---
 
@@ -121,6 +122,12 @@ fn main() -> Result<()> {
     * [Read the paper](https://doi.org/10.1145/3626717)
     * [Source code](https://github.com/cwida/ALP)
 
+* **Sven Hielke Hepkema, Azim Afroozeh, Charlotte Felius, Peter Boncz & Stefan Manegold**, “G‑ALP: Rethinking Light‑weight Encodings for GPUs,” *DaMoN ’25*, July 2025
+
+    * [Read the paper](https://dl.acm.org/doi/pdf/10.1145/3736227.3736242)
+    * [Source code](https://github.com/cwida/FastLanesGpu-Damon2025)
+
+
 ---
 
 ## How to Cite
@@ -161,6 +168,17 @@ If you use FastLanes in your research or projects, please cite:
     organization = {ACM},
     doi = {10.1145/3626717}
 }
+
+@inproceedings{hepke2025galp,
+    author       = {Hepkema, Sven Hielke and Afroozeh, Azim and Felius, Charlotte and Boncz, Peter and Manegold, Stefan},
+    title        = {G‑ALP: Rethinking Light‑weight Encodings for GPUs},
+    booktitle    = {DaMoN ’25: Proceedings of the 21st International Workshop on Data Management on New Hardware},
+    pages        = {11:1--11:10},
+    month        = jul,
+    year         = {2025},
+    organization = {ACM},
+    url          = {https://dl.acm.org/doi/pdf/10.1145/3736227.3736242}
+}
 ```
 
 ---

benchmark/CMakeLists.txt

@@ -30,3 +30,5 @@ target_link_libraries(bench_share_schema PUBLIC FastLanes::benchmarker)
 
 add_executable(bench_compression_time bench_compression_time.cpp)
 target_link_libraries(bench_compression_time PUBLIC FastLanes::benchmarker)
+
+add_subdirectory(bench_galp)

benchmark/bench_galp/CMakeLists.txt

@@ -0,0 +1,2 @@
+add_executable(bench_galp bench_data_parallelize_patch.cpp)
+target_link_libraries(bench_galp PUBLIC FastLanes::benchmarker)