[CI] Update CI workflow to use Python 3.12 (#679)

* Update CI workflow to use Python 3.12 and enable build isolation for pip installations

- Changed the Python version in the CI configuration from 3.9 to 3.12 to ensure compatibility with the latest features and improvements.
- Updated the `PIP_NO_BUILD_ISOLATION` environment variable from `0` to `1` in the CI configuration, allowing pip to install testing requirements with build isolation enabled, which enhances the installation process during CI runs.

* Update CI workflow to trigger on pull requests instead of pull_request_target

- Changed the event trigger in the CI configuration from `pull_request_target` to `pull_request` to ensure the workflow runs on pull requests, enhancing the integration process.

* Refactor CI workflow to remove unnecessary repository and token settings

- Removed the repository and token parameters from the checkout step in the CI configuration, simplifying the workflow setup and improving security by not exposing sensitive information.

* Remove pip install command from CI workflow to streamline installation process

* Refactor reshape functions and tests for shared memory operations

- Renamed and updated `reshape_test_smem` to `reshape_test_smem_1d_2_2d` and `run_reshape_smem` to `run_reshape_smem_1d_2_2d` for clarity.
- Introduced a new reshape function `reshape_test_smem_2d_2_1d` and its corresponding runner `run_reshape_smem_2d_2_1d`.
- Updated tests to reflect the new function names and added a test for the 2D to 1D reshape functionality, enhancing test coverage and clarity.

.github/workflows/ci.yml

 name: CI
-on: [pull_request_target]
+on: [pull_request]
 
 env:
-  PYTHON_VERSION: '3.9'
+  PYTHON_VERSION: '3.12'
   VENV_DIR: tilelang_ci
 
 jobs:
@@ -17,9 +17,6 @@ jobs:
       uses: actions/checkout@v4
       with:
         fetch-depth: 0
-        repository: ${{ github.event.pull_request.head.repo.full_name }}
-        ref: ${{ github.event.pull_request.head.ref }}
-        token: ${{ secrets.PAT }}
 
     - name: Set up Python
       uses: actions/setup-python@v2
@@ -42,8 +39,7 @@ jobs:
           source "${{ runner.tool_cache }}/${{ env.VENV_DIR }}/bin/activate"
           python -m pip install --upgrade pip --no-user
           [[ -f requirements-test.txt ]] && \
-            PIP_NO_BUILD_ISOLATION=0 pip install -r requirements-test.txt --no-user
-          pip install . --no-user
+            PIP_NO_BUILD_ISOLATION=1 pip install -r requirements-test.txt --no-user
           touch "$MARKER"
         fi
 
@@ -97,7 +93,7 @@ jobs:
           source "${{ runner.tool_cache }}/${{ env.VENV_DIR }}/bin/activate"
           python -m pip install --upgrade pip --no-user
           [[ -f requirements-test.txt ]] && \
-            PIP_NO_BUILD_ISOLATION=0 pip install -r requirements-test.txt --no-user
+            PIP_NO_BUILD_ISOLATION=1 pip install -r requirements-test.txt --no-user
           pip install . --no-user
           touch "$MARKER"
         fi

testing/python/language/test_tilelang_language_reshape.py

@@ -35,7 +35,7 @@ def test_reshape_smem():
     run_reshape(2048, 64, "float16")
 
 
-def reshape_test_smem(N, M, dtype):
+def reshape_test_smem_1d_2_2d(N, M, dtype):
     import tilelang.language as T
 
     @T.prim_func
@@ -45,19 +45,17 @@ def reshape_test_smem(N, M, dtype):
     ):
         with T.Kernel(1) as _:
             A_shared = T.alloc_shared((N,), dtype)
-            for i in range(N):
+            for i in T.Parallel(N):
                 A_shared[i] = A[i]
 
             A_smem_reshaped = T.reshape(A_shared, [N // M, M])
-            for i in range(N // M):
-                for j in range(M):
-                    B[i, j] = A_smem_reshaped[i, j]
+            T.copy(A_smem_reshaped, B)
 
     return main
 
 
-def run_reshape_smem(N, M, dtype):
-    program = reshape_test_smem(N, M, dtype)
+def run_reshape_smem_1d_2_2d(N, M, dtype):
+    program = reshape_test_smem_1d_2_2d(N, M, dtype)
     jit_kernel = tl.compile(program, out_idx=-1)
     profiler = jit_kernel.get_profiler()
 
@@ -67,9 +65,43 @@ def run_reshape_smem(N, M, dtype):
     profiler.assert_allclose(ref_program, atol=1e-2, rtol=1e-2)
 
 
-def test_reshape_smem_shared():
-    run_reshape_smem(1024, 32, "float32")
-    run_reshape_smem(2048, 64, "float16")
+def test_reshape_smem_1d_2_2d():
+    run_reshape_smem_1d_2_2d(1024, 32, "float32")
+    run_reshape_smem_1d_2_2d(2048, 64, "float16")
+
+
+def reshape_test_smem_2d_2_1d(N, M, dtype):
+    import tilelang.language as T
+
+    @T.prim_func
+    def main(
+            A: T.Tensor((N // M, M), dtype),
+            B: T.Tensor((N,), dtype),
+    ):
+        with T.Kernel(1) as _:
+            A_shared = T.alloc_shared((N // M, M), dtype)
+            for i, j in T.Parallel(N // M, M):
+                A_shared[i, j] = A[i, j]
+
+            A_smem_reshaped = T.reshape(A_shared, [N])
+            T.copy(A_smem_reshaped, B)
+
+    return main
+
+def run_reshape_smem_2d_2_1d(N, M, dtype):
+    program = reshape_test_smem_2d_2_1d(N, M, dtype)
+    jit_kernel = tl.compile(program, out_idx=-1)
+    profiler = jit_kernel.get_profiler()
+
+    def ref_program(A):
+        return A.reshape(N)
+
+    profiler.assert_allclose(ref_program, atol=1e-2, rtol=1e-2)
+
+def test_reshape_smem_2d_2_1d():
+    run_reshape_smem_2d_2_1d(1024, 32, "float32")
+    run_reshape_smem_2d_2_1d(2048, 64, "float16")
+
 
 
 if __name__ == "__main__":