RFR: 8279124: VM does not handle SIGQUIT during initialization [v2]

Xin Liu xliu at openjdk.java.net
Wed Jan 19 09:26:32 UTC 2022


On Wed, 19 Jan 2022 08:36:38 GMT, Thomas Stuefe <stuefe at openjdk.org> wrote:

>> I would not print anything here, first because it is in the signal handler and secondly I don't think it is needed. No real thread dump usages are going to hit a VM that hasn't initialized yet IMO.
>
>> This is in the signal handler. Is it safe to print message or do something complex?
> 
> raw write to stdout should be fine I think.
> 
>>I would not print anything here, first because it is in the signal handler and secondly I don't think it is needed. No real thread dump usages are going to hit a VM that hasn't initialized yet IMO.
> 
> Might be nice for the user to know why his jcmd or kill -3 gets ignored though.

because JVM_HANDLE_XXX_SIGNAL is installed in os::init_2(), it's safe to use log. I added a log entry with the tag os+init.  -Xlog:os+init=info will print out a message "ignore BREAK_SIGNAL in the initialization phase." I also update the testing example in the description section.

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

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


More information about the serviceability-dev mailing list