From c83823d1eb387f79511895a640c6186c9d9a85dc Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Mon, 22 Dec 2025 16:01:44 +0800
Subject: [PATCH 01/11] [CI] try test for flagtree

---
 .github/workflows/aipu-build-and-test.yml     |  9 +--
 .github/workflows/ascend-build-and-test.yml   |  9 +--
 .github/workflows/coordinator.yml             | 57 +++++++++++++++++++
 .github/workflows/hcu-build-and-test.yml      |  9 +--
 .github/workflows/iluvatar-build-and-test.yml |  9 +--
 .github/workflows/metax-build-and-test.yml    |  9 +--
 .github/workflows/mthreads-build-and-test.yml |  9 +--
 .github/workflows/nv-build-and-test.yml       |  9 +--
 .../workflows/tsingmicro-build-and-test.yml   |  9 +--
 .github/workflows/xpu-build-and-test.yml      |  9 +--
 10 files changed, 102 insertions(+), 36 deletions(-)
 create mode 100644 .github/workflows/coordinator.yml

diff --git a/.github/workflows/aipu-build-and-test.yml b/.github/workflows/aipu-build-and-test.yml
index b4e5599a5..f0c6cdbe4 100644
--- a/.github/workflows/aipu-build-and-test.yml
+++ b/.github/workflows/aipu-build-and-test.yml
@@ -1,10 +1,11 @@
 name: AIPU-Build-And-Test
 
 on:
-  push:
-    branches: [ "triton_v3.3.x" ]
-  pull_request:
-    branches: [ "triton_v3.3.x" ]
+  # push:
+  #   branches: [ "triton_v3.3.x" ]
+  # pull_request:
+  #   branches: [ "triton_v3.3.x" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/ascend-build-and-test.yml b/.github/workflows/ascend-build-and-test.yml
index d2b91baa0..bc8e95f33 100644
--- a/.github/workflows/ascend-build-and-test.yml
+++ b/.github/workflows/ascend-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Ascend-Build-And-Test
 
 on:
-  push:
-    branches: [ "triton_v3.2.x" ]
-  pull_request:
-    branches: [ "triton_v3.2.x" ]
+  # push:
+  #   branches: [ "triton_v3.2.x" ]
+  # pull_request:
+  #   branches: [ "triton_v3.2.x" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/coordinator.yml b/.github/workflows/coordinator.yml
new file mode 100644
index 000000000..54239f8aa
--- /dev/null
+++ b/.github/workflows/coordinator.yml
@@ -0,0 +1,57 @@
+name: Test Coordinator
+
+on:
+  push:
+    branches:
+      - main
+      - triton_v3.2.x
+      - triton_v3.3.x
+  pull_request:
+    branches:
+      - main
+      - triton_v3.2.x
+      - triton_v3.3.x
+      
+jobs:
+  ubuntu-native-aipu:
+    if: github.ref_name == 'triton_v3.3.x'
+    uses: ./.github/workflows/aipu-build-and-test.yml
+
+  ubuntu-native-ascend:
+    if: github.ref_name == 'triton_v3.2.x'
+    uses: ./.github/workflows/ascend-build-and-test.yml
+
+  ubuntu-native-cambricon:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/cambricon-build-and-test.yml
+
+  ubuntu-native-hcu:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/hcu-build-and-test.yml
+
+  ubuntu-native-iluvatar:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/iluvatar-build-and-test.yml
+
+  ubuntu-native-metax:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/metax-build-and-test.yml
+
+  ubuntu-native-mthreads:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/mthreads-build-and-test.yml
+
+  ubuntu-native-nv:
+    if: |
+      github.ref_name == 'main' || 
+      github.ref_name == 'triton_v3.2.x' || 
+      github.ref_name == 'triton_v3.3.x'
+    uses: ./.github/workflows/nv-build-and-test.yml
+
+  ubuntu-native-tsingmicro:
+    if: github.ref_name == 'triton_v3.3.x'
+    uses: ./.github/workflows/tsingmicro-build-and-test.yml
+
+  ubuntu-native-xpu:
+    if: github.ref_name == 'main'
+    uses: ./.github/workflows/xpu-build-and-test.yml
diff --git a/.github/workflows/hcu-build-and-test.yml b/.github/workflows/hcu-build-and-test.yml
index 0fc910cec..e0b5098ef 100644
--- a/.github/workflows/hcu-build-and-test.yml
+++ b/.github/workflows/hcu-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Hcu-Build-And-Test
 
 on:
-  push:
-    branches: [ "main" ]
-  pull_request:
-    branches: [ "main" ]
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/iluvatar-build-and-test.yml b/.github/workflows/iluvatar-build-and-test.yml
index d2ec03068..c9532b950 100644
--- a/.github/workflows/iluvatar-build-and-test.yml
+++ b/.github/workflows/iluvatar-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Iluvatar-Build-And-Test
 
 on:
-  push:
-    branches: [ "main" ]
-  pull_request:
-    branches: [ "main" ]
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/metax-build-and-test.yml b/.github/workflows/metax-build-and-test.yml
index cb9ec550d..1e409b66e 100644
--- a/.github/workflows/metax-build-and-test.yml
+++ b/.github/workflows/metax-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Metax-Build-And-Test
 
 on:
-  push:
-    branches: [ "main" ]
-  pull_request:
-    branches: [ "main" ]
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/mthreads-build-and-test.yml b/.github/workflows/mthreads-build-and-test.yml
index d86619fa8..0e87550a8 100644
--- a/.github/workflows/mthreads-build-and-test.yml
+++ b/.github/workflows/mthreads-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Mthreads-Build-And-Test
 
 on:
-  push:
-    branches: [ "main" ]
-  pull_request:
-    branches: [ "main" ]
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/nv-build-and-test.yml b/.github/workflows/nv-build-and-test.yml
index 952169159..dd4ec6155 100644
--- a/.github/workflows/nv-build-and-test.yml
+++ b/.github/workflows/nv-build-and-test.yml
@@ -3,10 +3,11 @@ name: NV-Build-And-Test
 on:
   schedule:
     - cron: '0 21 * * *'
-  push:
-    branches: [ "main", "triton_v3.2.x", "triton_v3.3.x" ]
-  pull_request:
-    branches: [ "main", "triton_v3.2.x", "triton_v3.3.x" ]
+  # push:
+  #   branches: [ "main", "triton_v3.2.x", "triton_v3.3.x" ]
+  # pull_request:
+  #   branches: [ "main", "triton_v3.2.x", "triton_v3.3.x" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/tsingmicro-build-and-test.yml b/.github/workflows/tsingmicro-build-and-test.yml
index 1cf2adf94..143bef229 100644
--- a/.github/workflows/tsingmicro-build-and-test.yml
+++ b/.github/workflows/tsingmicro-build-and-test.yml
@@ -1,10 +1,11 @@
 name: Tsingmicro-Build-And-Test
 
 on:
-  push:
-    branches: [ "triton_v3.3.x" ]
-  pull_request:
-    branches: [ "triton_v3.3.x" ]
+  # push:
+  #   branches: [ "triton_v3.3.x" ]
+  # pull_request:
+  #   branches: [ "triton_v3.3.x" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
diff --git a/.github/workflows/xpu-build-and-test.yml b/.github/workflows/xpu-build-and-test.yml
index e256e8a65..12dba45cd 100644
--- a/.github/workflows/xpu-build-and-test.yml
+++ b/.github/workflows/xpu-build-and-test.yml
@@ -2,10 +2,11 @@
 name: Xpu-Build-And-Test
 
 on:
-  push:
-    branches: [ "main" ]
-  pull_request:
-    branches: [ "main" ]
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_call:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}

From ac914d05e49602c66f5734266f49c4d2e6b99039 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 14:38:54 +0800
Subject: [PATCH 02/11] fix up

---
 .../test_ubuntuNativeContainerNV.yml          | 128 ++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 .github/workflows/test_ubuntuNativeContainerNV.yml

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
new file mode 100644
index 000000000..b868555a0
--- /dev/null
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -0,0 +1,128 @@
+name: Run Container ubuntu-native
+
+on:
+  push:
+   branches:
+     - main
+  pull_request:
+   branches:
+     - main
+#   schedule:
+#     - cron: '0 21 * * *'
+#   workflow_call:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  test-in-ubuntu-container:
+    runs-on: ubuntu-latest
+    container:
+      image:  ubuntu:22.04
+
+    steps:
+      - name: Set up Python and Install Dependencies
+        run: |
+          DEBIAN_FRONTEND=noninteractive apt-get update -y
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3-dev pybind11-dev python3-distutils
+          pip3 install pybind11 -q
+
+          wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
+          dpkg -i cuda-keyring_1.1-1_all.deb
+          DEBIAN_FRONTEND=noninteractive apt-get update -y
+          DEBIAN_FRONTEND=noninteractive apt-get install -y cuda-drivers cuda-toolkit
+          DEBIAN_FRONTEND=noninteractive apt-get install -y libnccl2 libnccl-dev
+
+      - name: Checkout Code
+        uses: actions/checkout@v6
+        with:
+          repository: ${{ github.event.pull_request.head.repo.full_name }}
+          ref: ${{ github.event.pull_request.head.ref }}
+          ssh-strict: true
+          ssh-user: git
+          persist-credentials: true
+          clean: true
+          sparse-checkout-cone-mode: true
+          fetch-tags: false
+          show-progress: true
+          lfs: false
+          submodules: true
+          set-safe-directory: true
+
+      - name: Detect Target Branch
+        shell: bash
+        run: |
+          set -x
+          if [ "${{ github.event_name }}" = "pull_request" ]; then
+            TARGET_BRANCH="${{ github.base_ref }}"
+          else
+            TARGET_BRANCH="${{ github.ref_name }}"
+          fi
+          echo "TARGET_BRANCH=$TARGET_BRANCH" >> $GITHUB_ENV
+          echo "TARGET_BRANCH=$TARGET_BRANCH"
+
+      - name: Prepare LLVM for NVidia
+        shell: bash
+        run: |
+          case "${{ env.TARGET_BRANCH }}" in
+            "main")          LLVM_HASH="10dc3a8e" ;;
+            "triton_v3.2.x") LLVM_HASH="86b69c31" ;;
+            "triton_v3.3.x") LLVM_HASH="a66376b0" ;;
+            "triton_v3.4.x") LLVM_HASH="8957e64a" ;;
+            "triton_v3.5.x") LLVM_HASH="7d5de303" ;;
+            *)               LLVM_HASH="10dc3a8e" ;; 
+          esac
+          
+          LLVM_DIR="$HOME/.flagtree/llvm_builds"
+          LLVM_FILE="llvm-${LLVM_HASH}-ubuntu-x64"
+          mkdir -p "$LLVM_DIR"
+          cd "$LLVM_DIR"
+
+          if [ ! -d "${LLVM_FILE}" ]; then
+            echo "Downloading LLVM ${LLVM_HASH}..."
+            wget -q "https://oaitriton.blob.core.windows.net/public/llvm-builds/${LLVM_FILE}.tar.gz"
+            tar -zxf "${LLVM_FILE}.tar.gz"
+            rm "${LLVM_FILE}.tar.gz"
+          else
+            echo "LLVM ${LLVM_HASH} already exists, skipping download."
+          fi          
+          echo "LLVM_SYSPATH=$LLVM_DIR/${LLVM_FILE}" >> $GITHUB_ENV
+          echo "LLVM_LIBRARY_DIR=$LLVM_DIR/${LLVM_FILE}/lib" >> $GITHUB_ENV
+
+      - name: FlagTree Build on NVidia (Main branch)
+        if: ${{ env.TARGET_BRANCH == 'main' }}
+        shell: bash
+        run: |
+          set -x
+          pip uninstall -y triton
+          cd /__w/flagtree/flagtree/python
+          MAX_JOBS=32 python3 -m pip install . --no-build-isolation
+
+      - name: FlagTree Build on NVidia (triton_v3.2.x branch)
+        if: ${{ env.TARGET_BRANCH == 'triton_v3.2.x' }}
+        shell: bash
+        run: |
+          set -x
+          pip uninstall -y triton
+          cd /__w/flagtree/flagtree/python
+          MAX_JOBS=32 python3 -m pip install . --no-build-isolation
+
+      - name: FlagTree Build on NVidia (triton_v3.3.x branch)
+        if: ${{ env.TARGET_BRANCH == 'triton_v3.3.x' }}
+        shell: bash
+        run: |
+          set -x
+          pip uninstall -y triton
+          cd /__w/flagtree/flagtree/python
+          MAX_JOBS=32 python3 -m pip install . --no-build-isolation
+
+      - name: FlagTree Test on NVidia
+        shell: bash
+        run: |
+          set -x
+          python3 -m pytest -s python/test/unit
+          if [ -d "python/test/operators" ]; then
+            python3 -m pytest -s python/test/operators
+          fi
\ No newline at end of file

From 08e94398e15a184769bfb2acf744cb5368d4ac8f Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 15:10:14 +0800
Subject: [PATCH 03/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index b868555a0..da66a4693 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -25,7 +25,7 @@ jobs:
       - name: Set up Python and Install Dependencies
         run: |
           DEBIAN_FRONTEND=noninteractive apt-get update -y
-          DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
           DEBIAN_FRONTEND=noninteractive apt-get install -y python3-dev pybind11-dev python3-distutils
           pip3 install pybind11 -q
 
@@ -97,6 +97,8 @@ jobs:
         run: |
           set -x
           pip uninstall -y triton
+          git config --global --add safe.directory /__w/flagtree/flagtree
+          export CMAKE_MAKE_PROGRAM=$(which make)
           cd /__w/flagtree/flagtree/python
           MAX_JOBS=32 python3 -m pip install . --no-build-isolation
 

From 083da2309141cddaa72cdbe733a1f2639b981a2d Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 15:25:18 +0800
Subject: [PATCH 04/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index da66a4693..0ca9c901a 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -27,7 +27,8 @@ jobs:
           DEBIAN_FRONTEND=noninteractive apt-get update -y
           DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
           DEBIAN_FRONTEND=noninteractive apt-get install -y python3-dev pybind11-dev python3-distutils
-          pip3 install pybind11 -q
+          pip3 install --upgrade pip setuptools wheel
+          pip3 install pybind11
 
           wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
           dpkg -i cuda-keyring_1.1-1_all.deb
@@ -98,7 +99,10 @@ jobs:
           set -x
           pip uninstall -y triton
           git config --global --add safe.directory /__w/flagtree/flagtree
-          export CMAKE_MAKE_PROGRAM=$(which make)
+          export Python3_ROOT_DIR=/usr/bin/python3
+          export Python3_INCLUDE_DIR=/usr/include/python3.10
+          export CMAKE_MAKE_PROGRAM=$(which ninja)
+          export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python
           MAX_JOBS=32 python3 -m pip install . --no-build-isolation
 

From 38868ae4efb295cad1c97d6f970cd66ce7bb2a2f Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 15:32:11 +0800
Subject: [PATCH 05/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 0ca9c901a..005593da7 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -99,8 +99,9 @@ jobs:
           set -x
           pip uninstall -y triton
           git config --global --add safe.directory /__w/flagtree/flagtree
-          export Python3_ROOT_DIR=/usr/bin/python3
-          export Python3_INCLUDE_DIR=/usr/include/python3.10
+          export Python3_INCLUDE_DIRS=/usr/include/python3.10
+          export Python3_LIBRARIES=/usr/lib/x86_64-linux-gnu/libpython3.10.so
+          export pybind11_DIR=$(python3 -m pybind11 --cmakedir)
           export CMAKE_MAKE_PROGRAM=$(which ninja)
           export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python

From d4c475bf1e61300e8391294aee64829f62156a95 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 15:40:48 +0800
Subject: [PATCH 06/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 005593da7..81e8acf1f 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -99,13 +99,14 @@ jobs:
           set -x
           pip uninstall -y triton
           git config --global --add safe.directory /__w/flagtree/flagtree
-          export Python3_INCLUDE_DIRS=/usr/include/python3.10
-          export Python3_LIBRARIES=/usr/lib/x86_64-linux-gnu/libpython3.10.so
-          export pybind11_DIR=$(python3 -m pybind11 --cmakedir)
           export CMAKE_MAKE_PROGRAM=$(which ninja)
           export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python
-          MAX_JOBS=32 python3 -m pip install . --no-build-isolation
+          MAX_JOBS=4 python3 -m pip install . --no-build-isolation \
+            -C="--cmake-args=-DPython3_EXECUTABLE=/usr/bin/python3" \
+            -C="--cmake-args=-DPython3_INCLUDE_DIR=/usr/include/python3.10" \
+            -C="--cmake-args=-Dpybind11_DIR=$(python3 -m pybind11 --cmakedir)" \
+            -v
 
       - name: FlagTree Build on NVidia (triton_v3.2.x branch)
         if: ${{ env.TARGET_BRANCH == 'triton_v3.2.x' }}

From 45db0f6936c53f16eddc50e62c7ef6ec4ee28ef3 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 15:55:42 +0800
Subject: [PATCH 07/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 81e8acf1f..8af071d5b 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -26,7 +26,7 @@ jobs:
         run: |
           DEBIAN_FRONTEND=noninteractive apt-get update -y
           DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
-          DEBIAN_FRONTEND=noninteractive apt-get install -y python3-dev pybind11-dev python3-distutils
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3.10-dev pybind11-dev python3-distutils
           pip3 install --upgrade pip setuptools wheel
           pip3 install pybind11
 
@@ -105,6 +105,7 @@ jobs:
           MAX_JOBS=4 python3 -m pip install . --no-build-isolation \
             -C="--cmake-args=-DPython3_EXECUTABLE=/usr/bin/python3" \
             -C="--cmake-args=-DPython3_INCLUDE_DIR=/usr/include/python3.10" \
+            -C="--cmake-args=-DPython3_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.10.so" \
             -C="--cmake-args=-Dpybind11_DIR=$(python3 -m pybind11 --cmakedir)" \
             -v
 

From 7f3ca24f10d37bf15298683b33b0a3be52ccc085 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 16:19:20 +0800
Subject: [PATCH 08/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 8af071d5b..75c5d38a7 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -24,9 +24,17 @@ jobs:
     steps:
       - name: Set up Python and Install Dependencies
         run: |
+          # DEBIAN_FRONTEND=noninteractive apt-get update -y
+          # DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
+          # DEBIAN_FRONTEND=noninteractive apt-get install -y python3.10-dev pybind11-dev python3-distutils
+          # pip3 install --upgrade pip setuptools wheel
+          # pip3 install pybind11
+
           DEBIAN_FRONTEND=noninteractive apt-get update -y
-          DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
-          DEBIAN_FRONTEND=noninteractive apt-get install -y python3.10-dev pybind11-dev python3-distutils
+          DEBIAN_FRONTEND=noninteractive apt-get install -y \
+            python3 python3-pip python3-venv git \
+            build-essential cmake ninja-build \
+            python3.10-dev python3-distutils  
           pip3 install --upgrade pip setuptools wheel
           pip3 install pybind11
 
@@ -102,7 +110,7 @@ jobs:
           export CMAKE_MAKE_PROGRAM=$(which ninja)
           export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python
-          MAX_JOBS=4 python3 -m pip install . --no-build-isolation \
+          MAX_JOBS=32 python3 -m pip install . --no-build-isolation \
             -C="--cmake-args=-DPython3_EXECUTABLE=/usr/bin/python3" \
             -C="--cmake-args=-DPython3_INCLUDE_DIR=/usr/include/python3.10" \
             -C="--cmake-args=-DPython3_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.10.so" \

From c471d58d82a48e66d0fde17712b9bec0b6760d98 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 16:23:31 +0800
Subject: [PATCH 09/11] fix up

---
 .../test_ubuntuNativeContainerNV.yml          | 20 +++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 75c5d38a7..622e8acc8 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -24,20 +24,20 @@ jobs:
     steps:
       - name: Set up Python and Install Dependencies
         run: |
-          # DEBIAN_FRONTEND=noninteractive apt-get update -y
-          # DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
-          # DEBIAN_FRONTEND=noninteractive apt-get install -y python3.10-dev pybind11-dev python3-distutils
-          # pip3 install --upgrade pip setuptools wheel
-          # pip3 install pybind11
-
           DEBIAN_FRONTEND=noninteractive apt-get update -y
-          DEBIAN_FRONTEND=noninteractive apt-get install -y \
-            python3 python3-pip python3-venv git \
-            build-essential cmake ninja-build \
-            python3.10-dev python3-distutils  
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip python3-venv git autotools-dev wget gcc g++ build-essential cmake ninja-build
+          DEBIAN_FRONTEND=noninteractive apt-get install -y python3.10-dev python3-distutils
           pip3 install --upgrade pip setuptools wheel
           pip3 install pybind11
 
+          # DEBIAN_FRONTEND=noninteractive apt-get update -y
+          # DEBIAN_FRONTEND=noninteractive apt-get install -y \
+          #   python3 python3-pip python3-venv git \
+          #   build-essential cmake ninja-build \
+          #   python3.10-dev python3-distutils  
+          # pip3 install --upgrade pip setuptools wheel
+          # pip3 install pybind11
+
           wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
           dpkg -i cuda-keyring_1.1-1_all.deb
           DEBIAN_FRONTEND=noninteractive apt-get update -y

From 00fcba73de7640a9ffb8d17dc5f64841d9014639 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 16:40:40 +0800
Subject: [PATCH 10/11] fix up

---
 .../workflows/test_ubuntuNativeContainerNV.yml  | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index 622e8acc8..a2031aacc 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -43,6 +43,19 @@ jobs:
           DEBIAN_FRONTEND=noninteractive apt-get update -y
           DEBIAN_FRONTEND=noninteractive apt-get install -y cuda-drivers cuda-toolkit
           DEBIAN_FRONTEND=noninteractive apt-get install -y libnccl2 libnccl-dev
+      
+
+      - name: Verify Python and pybind11 paths
+        shell: bash
+        run: |
+          echo "Python3 version: $(python3 --version)"
+          echo "Python3 include: $(python3 -c 'import distutils.sysconfig; print(distutils.sysconfig.get_python_inc())')"
+          PYBIND11_DIR=$(python3 -m pybind11 --cmakedir)
+          echo "pybind11 CMake dir: $PYBIND11_DIR"
+          if [ ! -d "$PYBIND11_DIR" ]; then
+            echo "Error: pybind11 CMake directory not found at $PYBIND11_DIR"
+            exit 1
+          fi
 
       - name: Checkout Code
         uses: actions/checkout@v6
@@ -110,11 +123,13 @@ jobs:
           export CMAKE_MAKE_PROGRAM=$(which ninja)
           export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python
+
+          PYBIND11_DIR=$(python3 -m pybind11 --cmakedir)
           MAX_JOBS=32 python3 -m pip install . --no-build-isolation \
             -C="--cmake-args=-DPython3_EXECUTABLE=/usr/bin/python3" \
             -C="--cmake-args=-DPython3_INCLUDE_DIR=/usr/include/python3.10" \
             -C="--cmake-args=-DPython3_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.10.so" \
-            -C="--cmake-args=-Dpybind11_DIR=$(python3 -m pybind11 --cmakedir)" \
+            -C="--cmake-args=-Dpybind11_DIR=$PYBIND11_DIR" \
             -v
 
       - name: FlagTree Build on NVidia (triton_v3.2.x branch)

From e59295fc07e2f1cb3aeb8e85e021f13918eeea88 Mon Sep 17 00:00:00 2001
From: weareyoung18 <3370789544@qq.com>
Date: Tue, 23 Dec 2025 16:51:56 +0800
Subject: [PATCH 11/11] fix up

---
 .github/workflows/test_ubuntuNativeContainerNV.yml | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/test_ubuntuNativeContainerNV.yml b/.github/workflows/test_ubuntuNativeContainerNV.yml
index a2031aacc..1200dcfe2 100644
--- a/.github/workflows/test_ubuntuNativeContainerNV.yml
+++ b/.github/workflows/test_ubuntuNativeContainerNV.yml
@@ -48,14 +48,17 @@ jobs:
       - name: Verify Python and pybind11 paths
         shell: bash
         run: |
-          echo "Python3 version: $(python3 --version)"
-          echo "Python3 include: $(python3 -c 'import distutils.sysconfig; print(distutils.sysconfig.get_python_inc())')"
           PYBIND11_DIR=$(python3 -m pybind11 --cmakedir)
           echo "pybind11 CMake dir: $PYBIND11_DIR"
           if [ ! -d "$PYBIND11_DIR" ]; then
             echo "Error: pybind11 CMake directory not found at $PYBIND11_DIR"
             exit 1
           fi
+          if [ ! -f "$PYBIND11_DIR/pybind11-config.cmake" ]; then
+            echo "Error: pybind11-config.cmake not found at $PYBIND11_DIR"
+            exit 1
+          fi
+          echo "pybind11-config.cmake found at $PYBIND11_DIR"
 
       - name: Checkout Code
         uses: actions/checkout@v6
@@ -124,14 +127,13 @@ jobs:
           export PIP_NO_CACHE_DIR=1
           cd /__w/flagtree/flagtree/python
 
-          PYBIND11_DIR=$(python3 -m pybind11 --cmakedir)
           MAX_JOBS=32 python3 -m pip install . --no-build-isolation \
             -C="--cmake-args=-DPython3_EXECUTABLE=/usr/bin/python3" \
             -C="--cmake-args=-DPython3_INCLUDE_DIR=/usr/include/python3.10" \
             -C="--cmake-args=-DPython3_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.10.so" \
             -C="--cmake-args=-Dpybind11_DIR=$PYBIND11_DIR" \
             -v
-
+            
       - name: FlagTree Build on NVidia (triton_v3.2.x branch)
         if: ${{ env.TARGET_BRANCH == 'triton_v3.2.x' }}
         shell: bash