[compiler-rt] [compiler-rt] Fix frame numbering for unparsable frames. (PR #148278)

via llvm-commits llvm-commits at lists.llvm.org
Fri Jul 11 12:09:30 PDT 2025


llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-compiler-rt-sanitizer

Author: Jesse Schwartzentruber (jschwartzentruber)

<details>
<summary>Changes</summary>

This can happen when JIT code is run, and we can't symbolize those frames, but they should remain numbered in the stack. An example spidermonkey trace:

```
    #<!-- -->0 0x564ac90fb80f  (/builds/worker/dist/bin/js+0x240e80f) (BuildId: 5d053c76aad4cfbd08259f8832e7ac78bbeeab58)
    #<!-- -->1 0x564ac9223a64  (/builds/worker/dist/bin/js+0x2536a64) (BuildId: 5d053c76aad4cfbd08259f8832e7ac78bbeeab58)
    #<!-- -->2 0x564ac922316f  (/builds/worker/dist/bin/js+0x253616f) (BuildId: 5d053c76aad4cfbd08259f8832e7ac78bbeeab58)
    #<!-- -->3 0x564ac9eac032  (/builds/worker/dist/bin/js+0x31bf032) (BuildId: 5d053c76aad4cfbd08259f8832e7ac78bbeeab58)
    #<!-- -->4 0x0dec477ca22e  (<unknown module>)
```

Without this change, the following symbolization is output:

```
    #<!-- -->0 0x55a6d72f980f in MOZ_CrashSequence /builds/worker/workspace/obj-build/dist/include/mozilla/Assertions.h:248:3
    #<!-- -->1 0x55a6d72f980f in Crash(JSContext*, unsigned int, JS::Value*) /builds/worker/checkouts/gecko/js/src/shell/js.cpp:4223:5
    #<!-- -->2 0x55a6d7421a64 in CallJSNative(JSContext*, bool (*)(JSContext*, unsigned int, JS::Value*), js::CallReason, JS::CallArgs const&) /builds/worker/checkouts/gecko/js/src/vm/Interpreter.cpp:501:13
    #<!-- -->3 0x55a6d742116f in js::InternalCallOrConstruct(JSContext*, JS::CallArgs const&, js::MaybeConstruct, js::CallReason) /builds/worker/checkouts/gecko/js/src/vm/Interpreter.cpp:597:12
    #<!-- -->4 0x55a6d80aa032 in js::jit::DoCallFallback(JSContext*, js::jit::BaselineFrame*, js::jit::ICFallbackStub*, unsigned int, JS::Value*, JS::MutableHandle<JS::Value>) /builds/worker/checkouts/gecko/js/src/jit/BaselineIC.cpp:1705:10
    #<!-- -->4 0x2c803bd8f22e  (<unknown module>)
```

The last frame has a duplicate number. With this change the numbering is correct:

```
    #<!-- -->0 0x5620c58ec80f in MOZ_CrashSequence /builds/worker/workspace/obj-build/dist/include/mozilla/Assertions.h:248:3
    #<!-- -->1 0x5620c58ec80f in Crash(JSContext*, unsigned int, JS::Value*) /builds/worker/checkouts/gecko/js/src/shell/js.cpp:4223:5
    #<!-- -->2 0x5620c5a14a64 in CallJSNative(JSContext*, bool (*)(JSContext*, unsigned int, JS::Value*), js::CallReason, JS::CallArgs const&) /builds/worker/checkouts/gecko/js/src/vm/Interpreter.cpp:501:13
    #<!-- -->3 0x5620c5a1416f in js::InternalCallOrConstruct(JSContext*, JS::CallArgs const&, js::MaybeConstruct, js::CallReason) /builds/worker/checkouts/gecko/js/src/vm/Interpreter.cpp:597:12
    #<!-- -->4 0x5620c669d032 in js::jit::DoCallFallback(JSContext*, js::jit::BaselineFrame*, js::jit::ICFallbackStub*, unsigned int, JS::Value*, JS::MutableHandle<JS::Value>) /builds/worker/checkouts/gecko/js/src/jit/BaselineIC.cpp:1705:10
    #<!-- -->5 0x349f24c7022e  (<unknown module>)
```

---
Full diff: https://github.com/llvm/llvm-project/pull/148278.diff


1 Files Affected:

- (modified) compiler-rt/lib/asan/scripts/asan_symbolize.py (+30-17) 


``````````diff
diff --git a/compiler-rt/lib/asan/scripts/asan_symbolize.py b/compiler-rt/lib/asan/scripts/asan_symbolize.py
index 058a1614b55e6..e70f987f03fe6 100755
--- a/compiler-rt/lib/asan/scripts/asan_symbolize.py
+++ b/compiler-rt/lib/asan/scripts/asan_symbolize.py
@@ -22,7 +22,6 @@
 import argparse
 import bisect
 import errno
-import getopt
 import logging
 import os
 import re
@@ -38,6 +37,7 @@
 allow_system_symbolizer = True
 force_system_symbolizer = False
 
+
 # FIXME: merge the code that calls fix_filename().
 def fix_filename(file_name):
     if fix_filename_patterns:
@@ -507,20 +507,29 @@ def symbolize_address(self, addr, binary, offset, arch):
         assert result
         return result
 
-    def get_symbolized_lines(self, symbolized_lines, inc_frame_counter=True):
+    def get_symbolized_lines(self, symbolized_lines):
         if not symbolized_lines:
-            if inc_frame_counter:
-                self.frame_no += 1
-            return [self.current_line]
-        else:
-            assert inc_frame_counter
-            result = []
-            for symbolized_frame in symbolized_lines:
-                result.append(
-                    "    #%s %s" % (str(self.frame_no), symbolized_frame.rstrip())
+            # If it is an unparsable frame, but contains a frame counter and address
+            # replace the frame counter so the stack is still consistent.
+            unknown_stack_frame_format = r"^( *#([0-9]+) +)(0x[0-9a-f]+) +.*"
+            match = re.match(unknown_stack_frame_format, self.current_line)
+            if match:
+                rewritten_line = (
+                    self.current_line[: match.start(2)]
+                    + str(self.frame_no)
+                    + self.current_line[match.end(2) :]
                 )
                 self.frame_no += 1
-            return result
+                return [rewritten_line]
+            # Not a frame line so don't increment the frame counter.
+            return [self.current_line]
+        result = []
+        for symbolized_frame in symbolized_lines:
+            result.append(
+                "    #%s %s" % (str(self.frame_no), symbolized_frame.rstrip())
+            )
+            self.frame_no += 1
+        return result
 
     def process_logfile(self):
         self.frame_no = 0
@@ -546,8 +555,7 @@ def process_line_posix(self, line):
         match = re.match(stack_trace_line_format, line)
         if not match:
             logging.debug('Line "{}" does not match regex'.format(line))
-            # Not a frame line so don't increment the frame counter.
-            return self.get_symbolized_lines(None, inc_frame_counter=False)
+            return self.get_symbolized_lines(None)
         logging.debug(line)
         _, frameno_str, addr, binary, offset = match.groups()
 
@@ -603,6 +611,7 @@ def _load_plugin_from_file_impl_py_gt_2(self, file_path, globals_space):
     def load_plugin_from_file(self, file_path):
         logging.info('Loading plugins from "{}"'.format(file_path))
         globals_space = dict(globals())
+
         # Provide function to register plugins
         def register_plugin(plugin):
             logging.info("Registering plugin %s", plugin.get_name())
@@ -779,9 +788,13 @@ def __str__(self):
             arch=self.arch,
             start_addr=self.start_addr,
             end_addr=self.end_addr,
-            module_path=self.module_path
-            if self.module_path == self.module_path_for_symbolization
-            else "{} ({})".format(self.module_path_for_symbolization, self.module_path),
+            module_path=(
+                self.module_path
+                if self.module_path == self.module_path_for_symbolization
+                else "{} ({})".format(
+                    self.module_path_for_symbolization, self.module_path
+                )
+            ),
             uuid=self.uuid,
         )
 

``````````

</details>


https://github.com/llvm/llvm-project/pull/148278


More information about the llvm-commits mailing list