|
||||||||
|
This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators. For more information on JIRA, see: http://www.atlassian.com/software/jira |
||||||||
You received this message because you are subscribed to the Google Groups "Jenkins Issues" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
For more options, visit https://groups.google.com/d/optout.
Code changed in jenkins
User: James Nord
Path:
src/main/java/hudson/remoting/AtmostOneThreadExecutor.java
src/main/java/hudson/remoting/Channel.java
src/main/java/hudson/remoting/PingThread.java
src/main/java/hudson/remoting/Request.java
src/main/java/hudson/remoting/SingleLaneExecutorService.java
src/main/java/hudson/remoting/SynchronousExecutorService.java
http://jenkins-ci.org/commit/remoting/29ec965650bba57145df14f0fee73acbd8ce5248
Log:
[FIXED JENKINS-21251] Initial calculation of wait time was susseptable to
clock drift.
The calculation on how long to wait in various places on the code was
suseptable to the clock changing as multiple calls where made to
System.currentTimeMillis() for the initial calculation.
Changed this so that for the initial calculation only one call will be
made to System.currentTimeMillis().
WHilst this does not guarantee that the actuall amount of time will have
been waited for (and subsequent calculations are suseptable) at least on
the hotstop VM Object.wait() will not have spurious wake-ups.
Therefore the overhead of moving to System.nanoTime() is currently not
considered to be worth the performance overhead.
Compare: https://github.com/jenkinsci/remoting/compare/ff1a0669d363^...29ec965650bb