[Bug] Fix num_splits in parallel_struct_for (#7121)

Issue: fix #7112

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

taichi/codegen/llvm/codegen_llvm.cpp

@@ -2261,7 +2261,8 @@ void TaskCodeGenLLVM::create_offload_struct_for(OffloadedStmt *stmt,
 
   int list_element_size = std::min(leaf_block->max_num_elements(),
                                    (int64)taichi_listgen_max_element_size);
-  int num_splits = std::max(1, list_element_size / stmt->block_dim);
+  int num_splits = std::max(1, list_element_size / stmt->block_dim +
+                                   (list_element_size % stmt->block_dim != 0));
 
   auto struct_for_func = get_runtime_function("parallel_struct_for");
 

tests/python/test_struct_for_non_pot.py

@@ -66,3 +66,40 @@ def test_2d():
 @test_utils.test(packed=True)
 def test_2d_packed():
     _test_2d()
+
+
+def _test_2d_pointer():
+    block_size, leaf_size = 3, 8
+    x = ti.field(ti.i32)
+    block = ti.root.pointer(ti.ij, (block_size, block_size))
+    block.dense(ti.ij, (leaf_size, leaf_size)).place(x)
+
+    @ti.kernel
+    def activate():
+        x[7, 7] = 1
+
+    activate()
+
+    @ti.kernel
+    def test() -> ti.i32:
+        res = 0
+        for I in ti.grouped(x):
+            res += I[0] + I[1] * 2
+        return res
+
+    ans = 0
+    for i in range(leaf_size):
+        for j in range(leaf_size):
+            ans += i + j * 2
+
+    assert ans == test()
+
+
+@test_utils.test(require=ti.extension.sparse, packed=False)
+def test_2d_pointer():
+    _test_2d_pointer()
+
+
+@test_utils.test(require=ti.extension.sparse, packed=True)
+def test_2d_pointer_packed():
+    _test_2d_pointer()