Add constraints for cadence idma ops. (#12597)

Summary:

Add memory planning constraints for idma ops:
1. idma load: output needs to be in DTCM
2. idma store: input needs to be in DTCM
3. idma wait: output aliases the input

Reviewed By: zonglinpeng

Differential Revision: D77232760

Author: hsharma35

backends/cadence/aot/memory_constraints.py

@@ -654,6 +654,37 @@ def compute_slice_and_select_loc_constraints(
 ]
 
 
+@register_cadence_pass(CadencePassAttribute(opt_level=0))
+class GenerateIdmaConstraints(PassBase):
+    """Generate constraints for idma ops."""
+
+    def __init__(self, constraint: MemConstraints) -> None:
+        self.constraint = constraint
+
+    def call(self, graph_module: torch.fx.GraphModule) -> Optional[PassResult]:
+        for node in graph_module.graph.find_nodes(
+            op="call_function", target=torch.ops.cadence.idma_wait.out
+        ):
+            # This is just an alias op.
+            self.constraint.add_relative_placement_constraint(node.args[0], node)
+
+        for node in graph_module.graph.find_nodes(
+            op="call_function", target=torch.ops.cadence.idma_load.out
+        ):
+            # TODO: set correct dtcm bank here.
+            mem_id = 1
+            self.constraint.add_absolute_placement_constraint(node, mem_id, None)
+
+        for node in graph_module.graph.find_nodes(
+            op="call_function", target=torch.ops.cadence.idma_store.out
+        ):
+            # TODO: set correct dtcm bank here.
+            mem_id = 1
+            self.constraint.add_absolute_placement_constraint(
+                node.args[0], mem_id, None
+            )
+
+
 # The class to generate all the constraints that will be passed on to the memory
 # planning algorithm.
 class GenerateMemConstraints:
@@ -671,6 +702,7 @@ def __call__(self, graph_module: torch.fx.GraphModule) -> PassResult:
         constraint_gen_passes: Sequence[ConstraintsGenPass] = cast(
             list[ConstraintsGenPass],
             [
+                GenerateIdmaConstraints,
                 GenerateMemoryViewConstraints,
                 GenerateSliceAndSelectNopConstraints,
                 GenerateCatNopConstraints,

backends/cadence/aot/memory_planning.py

@@ -420,7 +420,9 @@ def run(
         # True.
         mem_planning = MemoryPlanningPass(
             self.algo,
-            allow_lifetime_and_storage_overlap=(self.opt_level >= 2),
+            # Always allow lifetime and storage overlap.
+            # At opt level 0, we need overlap for idma wait.
+            allow_lifetime_and_storage_overlap=True,
             alloc_graph_input=self.alloc_graph_input,
             alloc_graph_output=self.alloc_graph_output,
         )

backends/cadence/aot/ops_registrations.py

@@ -178,7 +178,13 @@
 # Post memory planning, we check that outputs/inputs for the load/store are in
 # DTCM and replace idma_load/idma_store with idma_copy.
 lib.define("idma_load(Tensor src, int task_num=0, int channel=0) -> Tensor")
+lib.define(
+    "idma_load.out(Tensor src, int task_num=0, int channel=0, *, Tensor(a!) out) -> Tensor(a!)"
+)
 lib.define("idma_store(Tensor src, int task_num=0, int channel=0) -> Tensor")
+lib.define(
+    "idma_store.out(Tensor src, int task_num=0, int channel=0, *, Tensor(a!) out) -> Tensor(a!)"
+)
 
 # Non-blocking iDMA copy.
 lib.define("idma_copy(Tensor src, int task_num=0, int channel=0) -> Tensor")