RFR: 8279124: VM does not handle SIGQUIT during initialization [v5]

[David Holmes]

On Thu, 20 Jan 2022 06:05:22 GMT, Xin Liu <xliu at openjdk.org> wrote:

>> In early stage of initialization, HotSpot doesn't handle SIGQUIT. The default signal preposition on Linux is to quit the process and generate coredump.
>> 
>> There are 2 applications for this signal.
>> 1. There's a handshake protocol between sun.tools.attach and HotSpot.  VirtualMachineImpl sends SIGQUIT(3) if the AttachListener has not been initialized. It expects "Signal Dispatcher" to handle SIGQUIT and create AttachListener.
>> 2. POSIX systems use SIGQUIT as SIGBREAK. After AttachListener is up, HotSpot will reinterpret the signal for thread dump.
>> 
>> It is possible that HotSpot is still initializing in Threads::create_vm() when SIGQUIT arrives. We should change JVM_HANDLE_XXX_SIGNAL to catch SIGQUIT and ignore it. It is installed os::init_2() and should cover the early stage of initialization. Later on, os::initialize_jdk_signal_support() still overwrites the signal handler of SIGQUIT if ReduceSignalUsage is false(default). 
>> 
>> Testing 
>> 
>> Before, this patch, once initialization takes long time, jcmd may quit the java process. 
>> 
>> $java -Xms64g -XX:+AlwaysPreTouch -Xlog:gc+heap=debug:stderr -XX:ParallelGCThreads=1 &
>> [1] 9589
>> [0.028s][debug][gc,heap] Minimum heap 68719476736  Initial heap 68719476736  Maximum heap 68719476736
>> [0.034s][debug][gc,heap] Running G1 PreTouch with 1 workers for 16384 work units pre-touching 68719476736B.
>> $jcmd 9589 VM.flags
>> 9589:
>> [1]  + 9589 quit       java -Xms64g -XX:+AlwaysPreTouch -Xlog:gc+heap=debug:stderr
>> java.io.IOException: No such process
>>         at jdk.attach/sun.tools.attach.VirtualMachineImpl.sendQuitTo(Native Method)
>>         at jdk.attach/sun.tools.attach.VirtualMachineImpl.<init>(VirtualMachineImpl.java:100)
>>         at jdk.attach/sun.tools.attach.AttachProviderImpl.attachVirtualMachine(AttachProviderImpl.java:58)
>>         at jdk.attach/com.sun.tools.attach.VirtualMachine.attach(VirtualMachine.java:207)
>>         at jdk.jcmd/sun.tools.jcmd.JCmd.executeCommandForPid(JCmd.java:113)
>>         at jdk.jcmd/sun.tools.jcmd.JCmd.main(JCmd.java:97)
>> 
>> 
>> With this patch, neither jcmd nor kill -3 will disrupt java process 45850. 
>> 
>> $java -Xms64g -XX:+AlwaysPreTouch -XX:ParallelGCThreads=1 -Xlog:os+init=debug:stderr -version &
>> [1] 45850
>> $ kill -3 45850
>> [6.902s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.                                                                                                                                                                  
>> $jcmd 45850 VM.flags
>> 45850:
>> [19.920s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> [25.422s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> [26.522s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> [27.723s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> [29.023s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> [30.423s][info][os,init] ignore BREAK_SIGNAL in the initialization phase.
>> com.sun.tools.attach.AttachNotSupportedException: Unable to open socket file /proc/45850/root/tmp/.java_pid45850: target process 45850 doesn't respond within 10500ms or HotSpot VM not loaded
>>         at jdk.attach/sun.tools.attach.VirtualMachineImpl.<init>(VirtualMachineImpl.java:105)
>>         at jdk.attach/sun.tools.attach.AttachProviderImpl.attachVirtualMachine(AttachProviderImpl.java:58)
>>         at jdk.attach/com.sun.tools.attach.VirtualMachine.attach(VirtualMachine.java:207)
>>         at jdk.jcmd/sun.tools.jcmd.JCmd.executeCommandForPid(JCmd.java:113)
>>         at jdk.jcmd/sun.tools.jcmd.JCmd.main(JCmd.java:97)
>> $openjdk version "19-internal" 2022-09-20
>> OpenJDK Runtime Environment (fastdebug build 19-internal+0-adhoc.xxinliu.jdk)
>> OpenJDK 64-Bit Server VM (fastdebug build 19-internal+0-adhoc.xxinliu.jdk, mixed mode, sharing)
>> [1]  + 45850 done       java -Xms64g -XX:+AlwaysPreTouch -XX:ParallelGCThreads=1  -version
>
> Xin Liu has updated the pull request incrementally with one additional commit since the last revision:
> 
>   remove log_info from the signal handler.

src/hotspot/os/posix/signals_posix.cpp line 600:

> 598:     assert(!ReduceSignalUsage, "Should not happen with -Xrs/-XX:+ReduceSignalUsage");
> 599:     signal_was_handled = true;
> 600:   }

Can't we just do this as the very first thing at line 564:

// If we get BREAK_SIGNAL it means we are very early in VM initialization and 
// only temporarily "handling" it to prevent the VM process getting terminated.
if (sig == BREAK_SIGNAL) {
  assert(!ReduceSignalUsage, "Should not happen with -Xrs/-XX:+ReduceSignalUsage");
  return true;
}

-------------

PR: https://git.openjdk.java.net/jdk/pull/7003