[torch.compile] consider relevant code in compilation cache (#11614)

Signed-off-by: youkaichao <[email protected]>

Author: youkaichao

vllm/compilation/backends.py

@@ -145,6 +145,7 @@ def wrap_inductor(graph: fx.GraphModule,
                   example_inputs,
                   additional_inductor_config,
                   compilation_config: CompilationConfig,
+                  vllm_backend: "VllmBackend",
                   graph_index: int = 0,
                   num_graphs: int = 1,
                   runtime_shape: Optional[int] = None,
@@ -176,7 +177,7 @@ def wrap_inductor(graph: fx.GraphModule,
     # see https://github.com/pytorch/pytorch/issues/138980
     graph = copy.deepcopy(graph)
 
-    cache_data = compilation_config.inductor_hash_cache
+    cache_data = vllm_backend.inductor_hash_cache
     if (runtime_shape, graph_index) in cache_data:
         # we compiled this graph before
         # so we can directly lookup the compiled graph via hash
@@ -196,7 +197,7 @@ def wrap_inductor(graph: fx.GraphModule,
                 hash_str, example_inputs, True, False)
             assert inductor_compiled_graph is not None, (
                 "Inductor cache lookup failed. Please remove"
-                f"the cache file {compilation_config.inductor_hash_cache.cache_file_path} and try again."  # noqa
+                f"the cache file {cache_data.cache_file_path} and try again."  # noqa
             )
 
         # Inductor calling convention (function signature):
@@ -354,14 +355,15 @@ class PiecewiseCompileInterpreter(torch.fx.Interpreter):
 
     def __init__(self, module: torch.fx.GraphModule,
                  compile_submod_names: List[str], vllm_config: VllmConfig,
-                 graph_pool):
+                 graph_pool, vllm_backend: "VllmBackend"):
         super().__init__(module)
         from torch._guards import detect_fake_mode
         self.fake_mode = detect_fake_mode()
         self.compile_submod_names = compile_submod_names
         self.compilation_config = vllm_config.compilation_config
         self.graph_pool = graph_pool
         self.vllm_config = vllm_config
+        self.vllm_backend = vllm_backend
 
     def run(self, *args):
         fake_args = [
@@ -389,6 +391,7 @@ def call_module(self, target: torch.fx.node.Target,
                 args,
                 self.compilation_config.inductor_compile_config,
                 self.compilation_config,
+                self.vllm_backend,
                 graph_index=index,
                 num_graphs=len(self.compile_submod_names),
                 runtime_shape=None,
@@ -397,7 +400,7 @@ def call_module(self, target: torch.fx.node.Target,
             self.module.__dict__[target] = PiecewiseBackend(
                 submod, self.vllm_config, self.graph_pool, index,
                 len(self.compile_submod_names), sym_shape_indices,
-                compiled_graph_for_general_shape)
+                compiled_graph_for_general_shape, self.vllm_backend)
 
             compilation_counter.num_piecewise_capturable_graphs_seen += 1
 
@@ -430,6 +433,7 @@ class VllmBackend:
     post_grad_passes: Sequence[Callable]
     sym_tensor_indices: List[int]
     input_buffers: List[torch.Tensor]
+    inductor_hash_cache: InductorHashCache
 
     def __init__(
         self,
@@ -472,6 +476,53 @@ def configure_post_pass(self):
 
     def __call__(self, graph: fx.GraphModule, example_inputs) -> Callable:
 
+        if not self.compilation_config.cache_dir:
+            # no provided cache dir, generate one based on the known factors
+            # that affects the compilation. if none of the factors change,
+            # the cache dir will be the same so that we can reuse the compiled
+            # graph.
+
+            # 1. factors come from the vllm_config (it mainly summarizes how the
+            #    model is created)
+            vllm_config = self.vllm_config
+            config_hash = vllm_config.compute_hash()
+
+            # 2. factors come from the code files that are traced by Dynamo (
+            #    it mainly summarizes how the model is used in forward pass)
+            forward_code_files = list(
+                sorted(self.compilation_config.traced_files))
+            self.compilation_config.traced_files.clear()
+            logger.debug(
+                "Traced files (to be considered for compilation cache):\n%s",
+                "\n".join(forward_code_files))
+            hash_content = []
+            for filepath in forward_code_files:
+                hash_content.append(filepath)
+                with open(filepath) as f:
+                    hash_content.append(f.read())
+            import hashlib
+            code_hash = hashlib.md5(
+                "\n".join(hash_content).encode()).hexdigest()
+
+            # combine the two hashes to generate the cache dir
+            hash_key = hashlib.md5(
+                f"{config_hash}_{code_hash}".encode()).hexdigest()[:10]
+            cache_dir = os.path.join(
+                envs.VLLM_CACHE_ROOT, "torch_compile_cache", hash_key,
+                f"rank_{vllm_config.parallel_config.rank}")
+        else:
+            cache_dir = self.compilation_config.cache_dir
+        os.makedirs(cache_dir, exist_ok=True)
+
+        disabled = envs.VLLM_DISABLE_COMPILE_CACHE
+        self.inductor_hash_cache: InductorHashCache = InductorHashCache(
+            cache_dir, disabled=disabled)
+        if disabled:
+            logger.info("vLLM's torch.compile cache is disabled.")
+        else:
+            logger.info("Using cache directory: %s for vLLM's torch.compile",
+                        cache_dir)
+
         # when dynamo calls the backend, it means the bytecode
         # transform and analysis are done
         compilation_counter.num_graphs_seen += 1
@@ -507,8 +558,8 @@ def __call__(self, graph: fx.GraphModule, example_inputs) -> Callable:
         # propagate the split graph to the piecewise backend,
         # compile submodules with symbolic shapes
         PiecewiseCompileInterpreter(self.split_gm, submod_names_to_compile,
-                                    self.vllm_config,
-                                    self.graph_pool).run(*example_inputs)
+                                    self.vllm_config, self.graph_pool,
+                                    self).run(*example_inputs)
 
         self._called = True
 
@@ -577,7 +628,8 @@ class PiecewiseBackend:
     def __init__(self, graph: fx.GraphModule, vllm_config: VllmConfig,
                  graph_pool: Any, piecewise_compile_index: int,
                  total_piecewise_compiles: int, sym_shape_indices: List[int],
-                 compiled_graph_for_general_shape: Callable):
+                 compiled_graph_for_general_shape: Callable,
+                 vllm_backend: VllmBackend):
         """
         The backend for piecewise compilation.
         It mainly handles the compilation and cudagraph capturing.
@@ -597,6 +649,7 @@ def __init__(self, graph: fx.GraphModule, vllm_config: VllmConfig,
         self.graph_pool = graph_pool
         self.piecewise_compile_index = piecewise_compile_index
         self.total_piecewise_compiles = total_piecewise_compiles
+        self.vllm_backend = vllm_backend
 
         self.is_first_graph = piecewise_compile_index == 0
         self.is_last_graph = (
@@ -634,7 +687,7 @@ def check_for_ending_compilation(self):
         if self.is_last_graph and not self.to_be_compiled_sizes:
             # no specific sizes to compile
             # save the hash of the inductor graph for the next run
-            self.compilation_config.inductor_hash_cache.save_to_file()
+            self.vllm_backend.inductor_hash_cache.save_to_file()
             end_monitoring_torch_compile(self.vllm_config)
 
     def __call__(self, *args) -> Any:
@@ -662,6 +715,7 @@ def __call__(self, *args) -> Any:
                 args,
                 self.compilation_config.inductor_compile_config,
                 self.compilation_config,
+                self.vllm_backend,
                 graph_index=self.piecewise_compile_index,
                 num_graphs=self.total_piecewise_compiles,
                 runtime_shape=runtime_shape,

vllm/compilation/decorators.py

@@ -1,8 +1,10 @@
 import inspect
 from typing import Callable, Dict, List, Optional, TypeVar, Union, overload
+from unittest.mock import patch
 
 import torch
 import torch.nn as nn
+from torch._dynamo.symbolic_convert import InliningInstructionTranslator
 
 from vllm.compilation.counter import compilation_counter
 from vllm.compilation.wrapper import TorchCompileWrapperWithCustomDispatcher
@@ -196,7 +198,31 @@ def __call__(self, *args, **kwargs):
             # we need to control all the compilation of the model.
             torch._dynamo.eval_frame.remove_from_cache(
                 self.original_code_object)
-            return self.compiled_callable(*args, **kwargs)
+
+            # collect all relevant files traced by Dynamo,
+            # so that the compilation cache can trigger re-compilation
+            # properly when any of these files change.
+
+            # 1. the file containing the top-level forward function
+            self.vllm_config.compilation_config.traced_files.add(
+                self.original_code_object.co_filename)
+
+            # 2. every time Dynamo sees a function call, it will inline
+            # the function by calling InliningInstructionTranslator.inline_call
+            # we hijack this function to know all the functions called
+            # during Dynamo tracing, and their corresponding files
+            inline_call = InliningInstructionTranslator.inline_call
+
+            def patched_inline_call(parent, func, args, kwargs):
+                code = func.get_code()
+                self.vllm_config.compilation_config.traced_files.add(
+                    code.co_filename)
+                return inline_call(parent, func, args, kwargs)
+
+            with patch.object(InliningInstructionTranslator, 'inline_call',
+                              patched_inline_call):
+                output = self.compiled_callable(*args, **kwargs)
+            return output
 
         # usually, capturing the model once is enough, and then we can
         # dispatch to the compiled code directly, without going through

vllm/config.py

@@ -3,7 +3,6 @@
 import enum
 import hashlib
 import json
-import os
 import sys
 import warnings
 from contextlib import contextmanager
@@ -2778,9 +2777,8 @@ def model_post_init(self, __context: Any) -> None:
     # keep track of enabled and disabled custom ops
     enabled_custom_ops: Counter[str] = PrivateAttr
     disabled_custom_ops: Counter[str] = PrivateAttr
+    traced_files: Set[str] = PrivateAttr
     compilation_time: float = PrivateAttr
-    # should be InductorHashCache, but Pydantic does not support it
-    inductor_hash_cache: Any = PrivateAttr
 
     # Per-model forward context
     # Mainly used to store attention cls
@@ -2818,6 +2816,7 @@ def __repr__(self) -> str:
             "compilation_time",
             "bs_to_padded_graph_size",
             "pass_config",
+            "traced_files",
         }
         return self.model_dump_json(exclude=exclude, exclude_unset=True)
 
@@ -2877,6 +2876,7 @@ def model_post_init(self, __context: Any) -> None:
 
         self.enabled_custom_ops = Counter()
         self.disabled_custom_ops = Counter()
+        self.traced_files = set()
         self.static_forward_context = {}
         self.compilation_time = 0.0
 
@@ -2899,29 +2899,6 @@ def init_backend(self, vllm_config: "VllmConfig") -> Union[str, Callable]:
         # merge with the config use_inductor
         assert self.level == CompilationLevel.PIECEWISE
 
-        if not self.cache_dir:
-            # no provided cache dir, generate one based on the known factors
-            # that affects the compilation. if none of the factors change,
-            # the cache dir will be the same so that we can reuse the compiled
-            # graph.
-            hash_key = vllm_config.compute_hash()
-            cache_dir = os.path.join(
-                envs.VLLM_CACHE_ROOT, "torch_compile_cache", hash_key,
-                f"rank_{vllm_config.parallel_config.rank}")
-            os.makedirs(cache_dir, exist_ok=True)
-            self.cache_dir = cache_dir
-
-            disabled = envs.VLLM_DISABLE_COMPILE_CACHE
-            from vllm.compilation.backends import InductorHashCache
-            self.inductor_hash_cache: InductorHashCache = InductorHashCache(
-                self.cache_dir, disabled=disabled)
-            if disabled:
-                logger.info("vLLM's torch.compile cache is disabled.")
-            else:
-                logger.info(
-                    "Using cache directory: %s for vLLM's torch.compile",
-                    self.cache_dir)
-
         from vllm.compilation.backends import VllmBackend
         return VllmBackend(vllm_config)
 

vllm/sequence.py

@@ -1108,6 +1108,13 @@ class IntermediateTensors:
 
     tensors: Dict[str, torch.Tensor]
 
+    def __init__(self, tensors):
+        # manually define this function, so that
+        # Dynamo knows `IntermediateTensors()` comes from this file.
+        # Otherwise, dataclass will generate this function by evaluating
+        # a string, and we will lose the information about the source file.
+        self.tensors = tensors
+
     def __getitem__(self, key: Union[str, slice]):
         if isinstance(key, str):
             return self.tensors[key]