Community
Participate
Working Groups
Since build I20090317-1029, I observed a small regression on the FullSourceWorkspaceModelTests#testInitJDTPlugin() test. Here are the numbers: SLED 10 Sun 1.5.0_10 (2 x 3.00GHz - 3GB RAM) M6 I20090313-0100 -0.3% N20090314-2000 0.0% I20090317-1029 -3.9% N20090319-2000 -4.6% N20090321-2000 -4.6% I20090324-0800 -7.1% I20090324-1325 -3.9% I20090325-1135 -3.6% RHEL 5.0 Sun 6.0_04 (2 x 3.00GHz - 3GB RAM) M6 I20090313-0100 -0.6% N20090314-2000 -0.9% I20090317-1029 -5.5% N20090319-2000 -5.7% N20090321-2000 -5.5% I20090324-0800 -5.5% I20090324-1325 -5.8% I20090325-1135 -5.4%
This is a consequence of the fix for bug 269476.... As there's now an additional clean-up of the DeltaProcessingState deltaProcessors local thread while restarting, more is now necessary to garbage memory during the start-up, hence explains the noticed regression. This concerns only the performance test case. In standard Eclipse usage, this simulateRestart method is never called. So, I will comment the test with this explanation...
Created attachment 130098 [details] Proposed patch
Released for 3.5M7 in HEAD stream.
Verified for 3.5M7