feat(tensor): implement all TensorOp kernels (214/214 passing) (#40)

Author: simonCatBot

.github/workflows/conformance.yml

@@ -608,6 +608,35 @@ jobs:
           export VX_TEST_DATA_PATH=${{ github.workspace }}/OpenVX-cts/test_data/
           timeout 300 ./bin/vx_test_conformance --filter="BilateralFilter.*"
 
+  # Tensor ops — 226 tests (Tensor 12 + TensorOp 214), separated to keep
+  # each CI job focused and avoid state leakage with other categories.
+  tensor-ops:
+    name: "tensor-ops"
+    runs-on: ubuntu-22.04
+    needs: build
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: recursive
+      - name: Download build artifacts
+        uses: actions/download-artifact@v4
+        with:
+          name: build-artifacts
+      - name: Run Tensor tests
+        run: |
+          chmod +x OpenVX-cts/build/bin/vx_test_conformance
+          cd OpenVX-cts/build
+          export LD_LIBRARY_PATH=${{ github.workspace }}/target/release
+          export VX_TEST_DATA_PATH=${{ github.workspace }}/OpenVX-cts/test_data/
+          timeout 300 ./bin/vx_test_conformance --filter="Tensor.*"
+      - name: Run TensorOp tests
+        run: |
+          chmod +x OpenVX-cts/build/bin/vx_test_conformance
+          cd OpenVX-cts/build
+          export LD_LIBRARY_PATH=${{ github.workspace }}/target/release
+          export VX_TEST_DATA_PATH=${{ github.workspace }}/OpenVX-cts/test_data/
+          timeout 600 ./bin/vx_test_conformance --filter="TensorOp.*"
+
   # Performance benchmark using openvx-mark, comparing rustVX against the
   # Khronos OpenVX sample implementation on the SAME runner so the two
   # numbers come from identical hardware. This job does NOT rebuild either

openvx-core/src/c_api.rs

@@ -366,6 +366,14 @@ fn register_standard_kernels(context_id: u32) {
         ("org.khronos.openvx.weighted_average", 0x40, 4),
         // OpenVX 1.1 extensions
         ("org.khronos.openvx.sobel_5x5", 0x30, 3),
+        // Tensor operations (Enhanced Vision)
+        ("org.khronos.openvx.tensor_add", 0x41, 4),
+        ("org.khronos.openvx.tensor_subtract", 0x42, 4),
+        ("org.khronos.openvx.tensor_multiply", 0x43, 6),
+        ("org.khronos.openvx.tensor_convert_depth", 0x44, 5),
+        ("org.khronos.openvx.tensor_tablelookup", 0x45, 3),
+        ("org.khronos.openvx.tensor_transpose", 0x46, 4),
+        ("org.khronos.openvx.tensor_matrix_multiply", 0x47, 5),
     ];
 
     if let Ok(mut kernels) = KERNELS.lock() {