A133 安卓13 偶现无法开机问题(卡动画)
异常日志
1970-01-01 08:00:13.561 352-352 iorapd pid-352 W Reconnect to package manager service: 1 times
1970-01-01 08:00:13.566 352-352 ServiceManager pid-352 I Waiting for service 'package_native' on '/dev/binder'...
1970-01-01 08:00:16.139 309-468 libc++abi system_server E terminating with uncaught exception of type NSt3__112system_errorE: condition_variable timed_wait failed: Invalid argument
1970-01-01 08:00:16.139 309-468 libc system_server A Fatal signal 6 (SIGABRT), code -1 (SI_QUEUE) in tid 468 (Thread-9), pid 309 (system_server)
1970-01-01 08:00:18.614 352-352 ServiceManager iorapd W Service package_native didn't start. Returning NULL
1970-01-01 08:00:18.614 352-352 iorapd iorapd E Cannot get package manager service!
1970-01-01 08:00:18.860 320-320 PackageManager system_server I Fix for b/169414761 is applied
1970-01-01 08:00:18.860 320-320 PackageManagerTiming system_server V create package manager took to complete: 2907ms
1970-01-01 08:00:18.940 320-320 SystemServerTiming system_server V StartPackageManagerService took to complete: 2989ms
1970-01-01 08:00:19.615 352-352 iorapd iorapd W Reconnect to package manager service: 2 times
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Build fingerprint: 'WIF/BE/BE:13/TQ2A.230405.003.B2/eng.lxg.20241219.183158:userdebug/dev-keys'
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Revision: '0'
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A ABI: 'arm'
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Timestamp: 1970-01-01 08:00:15.955879757+0800
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Process uptime: 4s
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Cmdline: system_server
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A pid: 320, tid: 465, name: Thread-9 >>> system_server <<<
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A uid: 1000
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A signal 6 (SIGABRT), code -1 (SI_QUEUE), fault addr --------
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A Abort message: 'terminating with uncaught exception of type NSt3__112system_errorE: condition_variable timed_wait failed: Invalid argument'
2024-12-19 19:05:33.158 473-473 DEBUG crash_dump32 A r0 00000000 r1 000001d1 r2 00000006 r3 ba666ff8
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A r4 ba667008 r5 ba666ff0 r6 00000140 r7 0000016b
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A r8 00000000 r9 ffffffff r10 ba666ff8 r11 7fffffff
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A ip 000001d1 sp ba666fd8 lr ea7a6ddf pc ea7a6df2
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A backtrace:
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #00 pc 00062df2 /apex/com.android.runtime/lib/bionic/libc.so (abort+138) (BuildId: a2eff9b4b43a846ff70c420d25bbcd39)
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #01 pc 0001f359 /apex/com.android.os.statsd/lib/libstatspull.so (abort_message+92) (BuildId: 649822716a4e4e9612385129366d0b3d)
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #02 pc 0001f6bd /apex/com.android.os.statsd/lib/libstatspull.so (demangling_terminate_handler()+120) (BuildId: 649822716a4e4e9612385129366d0b3d)
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #03 pc 0001f57f /apex/com.android.os.statsd/lib/libstatspull.so (std::__terminate(void (*)())+2) (BuildId: 649822716a4e4e9612385129366d0b3d)
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #04 pc 0001f52f /apex/com.android.os.statsd/lib/libstatspull.so (std::terminate()+46) (BuildId: 649822716a4e4e9612385129366d0b3d)
2024-12-19 19:05:33.159 473-473 DEBUG crash_dump32 A #05 pc 0001f3c9 /apex/com.android.os.statsd/lib/libstatspull.so (__clang_call_terminate+4) (BuildId: 649822716a4e4e9612385129366d0b3d)
2024-12-19 19:05:33.923 352-352 iorapd iorapd W Reconnect to package manager service: 1 times
2024-12-19 19:05:33.924 352-352 ServiceManager iorapd I Waiting for service 'package_native' on '/dev/binder'...
1970-01-01 08:00:20.827 352-352 binder iorapd I 352:352 transaction failed 29189/-22, size 100-0 line 3188
iorapd 是 I/O 预读取守护进程，负责监测和优化应用启动速度
Google issue
terminating with uncaught exception of type NSt3__112system_errorE
Analyze：
- Through log analysis, the reason for the system error crash is： libc++abi: termination with uncaught exception of type NSt3__112system_errorE: condition_variable timed_wait failed: Invalid argument.
- Analyze the code and locate the location where the problem occurs:android/frameworks/base/services/incremental/ServiceWrappers.cpp runTimers().
- Through analysis and experiments, it is initially believed that the problem caused by the parameter “nextJobTs” passed to “mCondition.wait_until” is too large. We divide the initial value of the “static constexpr TimePoint kInfinityTs” by 2 and the problem no longer occurs.
AOSP源码中寻找解决方案
diff --git a/android/frameworks/base/services/incremental/ServiceWrappers.cpp b/android/frameworks/base/services/incremental/ServiceWrappers.cpp
index ce3d51483d..2a2fa7adf5 100644
--- a/android/frameworks/base/services/incremental/ServiceWrappers.cpp
+++ b/android/frameworks/base/services/incremental/ServiceWrappers.cpp
@@ -315,11 +315,22 @@ private:
std::unique_lock lock(mMutex);
for (;;) {
const TimePoint nextJobTs = mJobs.empty() ? kInfinityTs : mJobs.begin()->when;
- mCondition.wait_until(lock, nextJobTs, [this, oldNextJobTs = nextJobTs]() {
+ auto conditionPredicate = [this, oldNextJobTs = nextJobTs]() {
const auto now = Clock::now();
const auto newFirstJobTs = !mJobs.empty() ? mJobs.begin()->when : kInfinityTs;
return newFirstJobTs <= now || newFirstJobTs < oldNextJobTs || !mRunning;
- });
+ };
+ // libcxx's implementation of wait_until() recalculates the 'until' time into
+ // the wait duration and then goes back to the absolute timestamp when calling
+ // pthread_cond_timedwait(); this back-and-forth calculation sometimes loses
+ // the 'infinity' value because enough time passes in between, and instead
+ // passes incorrect timestamp into the syscall, causing a crash.
+ // Mitigating it by explicitly calling the non-timed wait here.
+ if (mJobs.empty()) {
+ mCondition.wait(lock, conditionPredicate);
+ } else {
+ mCondition.wait_until(lock, nextJobTs, conditionPredicate);
+ }
if (!mRunning) {
return;
}
