100% cpu on arm32 in Service Thread
david.holmes at oracle.com
Mon Sep 30 00:43:47 UTC 2019
On 28/09/2019 2:01 am, Craig.Condit wrote:
> It appears there is some sort of regression in serviceThread.cpp which consumes all available CPU on jdk13 arm32 builds (tested on multiple vendor builds including AdoptOpenJDK and BellSoft). This is easily reproducible -- I used multiple Raspberry Pis (2 and 3).
> All java processes consume 100% of a core in "Service Thread". To reproduce, launch any long-running java process (such as jshell), and in another terminal, run "top -H" to view CPU usage by thread.
> This does not occur on jdk 11 or jdk 12.
Doesn't occur on x86-32 either so seems to be arm32 specific. Which is a
little odd as I can't see anything the serviceThread does that is
obviously CPU dependent. Unfortunately there is no logging in the
serviceThread itself so this would need some active debugging by the
Bellsoft folk (as maintainers of the arm32 port). I've added Boris to
the cc list.
Thanks for the report.
More information about the hotspot-dev