Add AMD CI test

Author: Wenlei Bao

.github/workflows/ci.yml

@@ -0,0 +1,62 @@
+name: Integration Tests
+on:
+  push:
+    branches: [main]
+
+jobs:
+  # TODO: Add pre commit checks
+  # pre-commit:
+  #   uses: ./.github/workflows/pre-commit.yml
+
+  # TODO: Add Nvidia integration tests
+  # integration-tests-nvidia:
+
+  integration-tests-amd:
+    runs-on: ${{ matrix.runner }}
+    timeout-minutes: 60
+    strategy:
+      matrix:
+        runner: "amd-gfx942"
+        include:
+          # - image: "rocm/rocm-6.4-py310-torch25-241101-3"
+          - image: "rocm/pytorch:rocm6.1_ubuntu22.04_py3.10_pytorch_2.4"
+            runner: ["amd-gfx942"]
+    env:
+      PYTHON: "python3"
+
+    container:
+      - name: "AMD Integration Tests"
+        image: ${{ matrix.image }}
+        options: --user root
+
+    steps:
+      - name: Checkout
+        # uses: actions/checkout@v3
+        # with:
+        #   submodules: 'recursive'
+        run: |
+          git clone https://github.com/ByteDance-Seed/Triton-distributed.git
+          cd Triton-distributed/
+          git submodule update --init --recursive
+
+      - name: Install dependencies
+        run: |
+          echo "Install dependencies..."
+          sudo apt-get update -y
+          sudo apt install -y libopenmpi-dev
+          pip3 install --pre torch --index-url https://download.pytorch.org/whl/nightly/rocm6.3 --no-deps
+          bash ./shmem/rocshmem_bind/build.sh
+          python3 -m pip install -i https://test.pypi.org/simple hip-python>=6.3.0 # (or whatever Rocm version you have)
+          pip3 install pybind11
+
+      - name: Build
+        runs: |
+          echo "Building Triton-distributed..."
+          pip3 install -e python --verbose --no-build-isolation --use-pep517
+
+
+      - name: Run AMD Integration Tests
+        run: |
+          echo "Running AMD integration tests..."
+          bash ./scripts/launch_amd.sh ./python/triton_dist/test/amd/test_ag_gemm_intra_node.py 8192 8192 29568
+