Hot Cache: use the DSO last write time/size instead of content hash

After profiling a nixglhost hot run, it turns out that we were spending more than 98% of the run time reading and sha256-hashing files. Let's give up on content hashing the files and assume that using their name, size and last write time is good enough. On a hot run, we reduce the run time from about 3s to 0.3s on a nvme-powered ryzen 7 desktop. I guess this 10x speedup probably worth the little cache corectness we lose on the way.
numtide · Dec 14, 2022 · 3ff2f01 · 3ff2f01
1 parent 61a5fcd
commit 3ff2f01
Show file tree

Hide file tree

Showing 5 changed files with 82 additions and 58 deletions.
diff --git a/src/nixglhost_wrapper.py b/src/nixglhost_wrapper.py
@@ -14,7 +14,7 @@
 from typing import List, Literal, Dict, Tuple, TypedDict, TextIO, Optional
 
 IN_NIX_STORE = False
-CACHE_VERSION = 2
+CACHE_VERSION = 3
 
 
 if IN_NIX_STORE:
@@ -30,47 +30,55 @@ class ResolvedLib:
     together with some metadata helping us to uniquely identify it."""
 
     def __init__(
-        self, name: str, dirpath: str, fullpath: str, sha256: Optional[str] = None
+        self,
+        name: str,
+        dirpath: str,
+        fullpath: str,
+        last_modification: Optional[float] = None,
+        size: Optional[int] = None,
     ):
         self.name: str = name
         self.dirpath: str = dirpath
         self.fullpath: str = fullpath
-        if sha256 is None:
-            h = hashlib.sha256()
-            with open(fullpath, "rb") as f:
-                h.update(f.read())
-            sha: str = h.hexdigest()
+        if size is None or last_modification is None:
+            stat = os.stat(fullpath)
+            self.last_modification: float = stat.st_atime
+            self.size: int = stat.st_size
         else:
-            sha = sha256
-        self.sha256: str = sha
+            self.last_modification = last_modification
+            self.size = size
 
     def __repr__(self):
-        return (
-            f"ResolvedLib<{self.name}, {self.dirpath}, {self.fullpath}, {self.sha256}>"
-        )
+        return f"ResolvedLib<{self.name}, {self.dirpath}, {self.fullpath}, {self.last_modification}, {self.size}>"
 
     def to_dict(self) -> Dict:
         return {
             "name": self.name,
             "dirpath": self.dirpath,
             "fullpath": self.fullpath,
-            "sha256": self.sha256,
+            "last_modification": self.last_modification,
+            "size": self.size,
         }
 
     def __hash__(self):
-        return hash((self.name, self.dirpath, self.fullpath, self.sha256))
+        return hash(
+            (self.name, self.dirpath, self.fullpath, self.last_modification, self.size)
+        )
 
     def __eq__(self, o):
         return (
             self.name == o.name
             and self.fullpath == o.fullpath
-            and self.sha256 == o.sha256
             and self.dirpath == o.dirpath
+            and self.last_modification == o.last_modification
+            and self.size == o.size
         )
 
     @classmethod
     def from_dict(cls, d: Dict):
-        return ResolvedLib(d["name"], d["dirpath"], d["fullpath"], d["sha256"])
+        return ResolvedLib(
+            d["name"], d["dirpath"], d["fullpath"], d["last_modification"], d["size"]
+        )
 
 
 class LibraryPath:
@@ -374,8 +382,8 @@ def is_dso_cache_up_to_date(dsos: CacheDirContent, cache_file_path: str) -> bool
 
     We keep what's in the cache through a JSON file stored at the root
     of the cache_dir. We consider a dynamically shared object to be up
-    to date if its name, its full path and its content sha256 are
-    equivalent."""
+    to date if its name, its full path, its size and last modification
+    timestamp are equivalent."""
     log_info("Checking if the cache is up to date")
     if os.path.isfile(cache_file_path):
         with open(cache_file_path, "r", encoding="utf8") as f:

diff --git a/src/nixglhost_wrapper_test.py b/src/nixglhost_wrapper_test.py
@@ -9,34 +9,26 @@ def test_hostdso_json_golden_test(self):
         lp = LibraryPath(
             glx=[
                 ResolvedLib(
-                    "dummyglx.so",
-                    "/lib",
-                    "/lib/dummyglx.so",
-                    "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406",
+                    "dummyglx.so", "/lib", "/lib/dummyglx.so", 1670260550.481498, 1612
                 )
             ],
             cuda=[
                 ResolvedLib(
-                    "dummycuda.so",
-                    "/lib",
-                    "/lib/dummycuda.so",
-                    "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407",
+                    "dummycuda.so", "/lib", "/lib/dummycuda.so", 2670260550.481498, 2612
                 )
             ],
             generic=[
                 ResolvedLib(
                     "dummygeneric.so",
                     "/lib",
                     "/lib/dummygeneric.so",
-                    "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408",
+                    3670260550.481498,
+                    3612,
                 )
             ],
             egl=[
                 ResolvedLib(
-                    "dummyegl.so",
-                    "/lib",
-                    "/lib/dummyegl.so",
-                    "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409",
+                    "dummyegl.so", "/lib", "/lib/dummyegl.so", 4670260550.481498, 4612
                 )
             ],
             path="/path/to/lib/dir",

diff --git a/tests/fixtures/json_permut/1.json b/tests/fixtures/json_permut/1.json
@@ -6,31 +6,35 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406"
+          "last_modification": 1670260550.481498,
+          "size": 1612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 2670260550.481498,
+          "size": 2612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 3670260550.481498,
+          "size": 3612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 4670260550.481498,
+          "size": 4612
         }
       ],
       "path": "/path/to/lib/dir"
@@ -41,35 +45,39 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406"
+          "last_modification": 5670260550.481498,
+          "size": 5612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 6670260550.481498,
+          "size": 6612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 7670260550.481498,
+          "size": 7612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 8670260550.481498,
+          "size": 8612
         }
       ],
       "path": "/path/to/lib/dir2"
     }
   ],
-  "version": 2
+  "version": 3
 }
diff --git a/tests/fixtures/json_permut/2.json b/tests/fixtures/json_permut/2.json
@@ -6,31 +6,35 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406"
+          "last_modification": 5670260550.481498,
+          "size": 5612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 6670260550.481498,
+          "size": 6612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 7670260550.481498,
+          "size": 7612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 8670260550.481498,
+          "size": 8612
         }
       ],
       "path": "/path/to/lib/dir2"
@@ -41,35 +45,39 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406"
+          "last_modification": 1670260550.481498,
+          "size": 1612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 2670260550.481498,
+          "size": 2612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 3670260550.481498,
+          "size": 3612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 4670260550.481498,
+          "size": 4612
         }
       ],
       "path": "/path/to/lib/dir"
     }
   ],
-  "version": 2
+  "version": 3
 }
diff --git a/tests/fixtures/json_permut/not-equal.json b/tests/fixtures/json_permut/not-equal.json
@@ -6,31 +6,35 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9406"
+          "last_modification": 2670260550.481498,
+          "size": 2612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 2670260550.481498,
+          "size": 2612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 3670260550.481498,
+          "size": 3612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "031edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 4670260550.481498,
+          "size": 4612
         }
       ],
       "path": "/path/to/lib/dir"
@@ -41,35 +45,39 @@
           "name": "dummyglx.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyglx.so",
-          "sha256": "4444444444444444444444444444444444444444444444444444444444444444"
+          "last_modification": 5670260550.481498,
+          "size": 5612
         }
       ],
       "cuda": [
         {
           "name": "dummycuda.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummycuda.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9407"
+          "last_modification": 6670260550.481498,
+          "size": 6612
         }
       ],
       "generic": [
         {
           "name": "dummygeneric.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummygeneric.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9408"
+          "last_modification": 7670260550.481498,
+          "size": 7612
         }
       ],
       "egl": [
         {
           "name": "dummyegl.so",
           "dirpath": "/lib",
           "fullpath": "/lib/dummyegl.so",
-          "sha256": "131edd7d41651593c5fe5c006fa5752b37fddff7bc4e843aa6af0c950f4b9409"
+          "last_modification": 8670260550.481498,
+          "size": 8612
         }
       ],
       "path": "/path/to/lib/dir2"
     }
   ],
-  "version": 2
+  "version": 3
 }