Comment 1 Dani Megert 2010-11-16 10:12:12 EST

Deepak, please take a look at this during our test pass when you're at the Linux machine.

Comment 2 Dani Megert 2010-11-17 03:23:56 EST

Looks good again in latest I-build.

Comment 3 Satyam Kandula 2010-11-23 06:16:40 EST

This failed again in N20101118-2000 build with the following call stack .. Reopening! 
#########
org.eclipse.core.runtime.AssertionFailedException: assertion failed:
at org.eclipse.core.runtime.Assert.isTrue(Assert.java:110)
at org.eclipse.core.runtime.Assert.isTrue(Assert.java:96)
at org.eclipse.jface.text.Position.<init>(Position.java:63)
at org.eclipse.jface.text.AbstractDocument.getPositions(AbstractDocument.java:1750)
at org.eclipse.core.internal.filebuffers.SynchronizableDocument.getPositions(SynchronizableDocument.java:277)
at org.eclipse.jface.text.source.AnnotationModel.getRegionAnnotationIterator(AnnotationModel.java:737)
at org.eclipse.jface.text.source.AnnotationModel.getAnnotationIterator(AnnotationModel.java:699)
at org.eclipse.jface.text.source.AnnotationRulerColumn.doPaint1(AnnotationRulerColumn.java:727)
at org.eclipse.jface.text.source.AnnotationRulerColumn.doubleBufferPaint(AnnotationRulerColumn.java:541)
at org.eclipse.jface.text.source.AnnotationRulerColumn.redraw(AnnotationRulerColumn.java:824)
at org.eclipse.jface.text.source.AnnotationRulerColumn$InternalListener.viewportChanged(AnnotationRulerColumn.java:79)
at org.eclipse.jface.text.TextViewer.updateViewportListeners(TextViewer.java:3112)
at org.eclipse.jface.text.TextViewer$ViewportGuard.keyReleased(TextViewer.java:299)
at org.eclipse.swt.widgets.TypedListener.handleEvent(TypedListener.java:168)
at org.eclipse.swt.widgets.EventTable.sendEvent(EventTable.java:84)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1258)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1282)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1267)
at org.eclipse.swt.widgets.Widget.sendKeyEvent(Widget.java:1294)
at org.eclipse.swt.widgets.Widget.gtk_key_release_event(Widget.java:736)
at org.eclipse.swt.widgets.Control.gtk_key_release_event(Control.java:2854)
at org.eclipse.swt.widgets.Widget.windowProc(Widget.java:1744)
at org.eclipse.swt.widgets.Control.windowProc(Control.java:4800)
at org.eclipse.swt.widgets.Display.windowProc(Display.java:4359)
at org.eclipse.swt.internal.gtk.OS._gtk_main_do_event(Native Method)
at org.eclipse.swt.internal.gtk.OS.gtk_main_do_event(OS.java:8253)
at org.eclipse.swt.widgets.Display.eventProc(Display.java:1239)
at org.eclipse.swt.internal.gtk.OS._g_main_context_iteration(Native Method)
at org.eclipse.swt.internal.gtk.OS.g_main_context_iteration(OS.java:2255)
at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3158)
at org.eclipse.jdt.text.tests.performance.EditorTestHelper.runEventQueue(EditorTestHelper.java:196)
at org.eclipse.jdt.text.tests.performance.EditorTestHelper.runEventQueue(EditorTestHelper.java:192)
at org.eclipse.jdt.text.tests.performance.EditorTestHelper.runEventQueue(EditorTestHelper.java:184)
at org.eclipse.jdt.text.tests.performance.MoveLineTest.measureMoveLine(MoveLineTest.java:95)
at org.eclipse.jdt.text.tests.performance.MoveLineTest.test(MoveLineTest.java:65)
at junit.extensions.TestDecorator.basicRun(TestDecorator.java:24)
at junit.extensions.TestSetup$1.protect(TestSetup.java:23)
at junit.extensions.TestSetup.run(TestSetup.java:27)
at junit.extensions.TestDecorator.basicRun(TestDecorator.java:24)
at junit.extensions.TestSetup$1.protect(TestSetup.java:23)
at junit.extensions.TestSetup.run(TestSetup.java:27)
at org.eclipse.test.EclipseTestRunner.run(EclipseTestRunner.java:377)
at org.eclipse.test.EclipseTestRunner.run(EclipseTestRunner.java:210)
at org.eclipse.test.UITestApplication$2.run(UITestApplication.java:197)
at org.eclipse.swt.widgets.RunnableLock.run(RunnableLock.java:35)
at org.eclipse.swt.widgets.Synchronizer.runAsyncMessages(Synchronizer.java:134)
at org.eclipse.swt.widgets.Display.runAsyncMessages(Display.java:3514)
at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3163)
at org.eclipse.ui.internal.Workbench.runEventLoop(Workbench.java:2640)
at org.eclipse.ui.internal.Workbench.runUI(Workbench.java:2604)
at org.eclipse.ui.internal.Workbench.access$4(Workbench.java:2438)
at org.eclipse.ui.internal.Workbench$7.run(Workbench.java:671)
at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
at org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:664)
at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:149)
at org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:115)
at org.eclipse.test.UITestApplication.runApplication(UITestApplication.java:140)
at org.eclipse.test.UITestApplication.run(UITestApplication.java:62)
at org.eclipse.test.UITestApplication.start(UITestApplication.java:212)
at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:196)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:110)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:79)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:369)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:179)
at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:621)
at org.eclipse.equinox.launcher.Main.basicRun(Main.java:576)
at org.eclipse.equinox.launcher.Main.run(Main.java:1409)
at org.eclipse.equinox.launcher.Main.main(Main.java:1385)
at org.eclipse.core.launcher.Main.main(Main.java:34)
##############

Comment 4 Dani Megert 2010-11-23 06:38:25 EST

:-(

Comment 5 Deepak Azad 2010-12-08 08:30:14 EST

There are four tests in question here
- TextMoveLineTest
- JavaMoveLineTest
- JavaReplaceAllTest
- JavaReplaceAllWithQuickDiffTest

I was looking at the performance data with Satyam, and found that all these tests fail together quite often, but it is not so in all cases. For example, in some cases TextMoveLineTest passes but JavaMoveLineTest passes. In some cases one of JavaReplace* passes but JavaMoveLineTest fails.

Plus the failures are intermittent, and quite difficult to reproduce.

Comment 6 Deepak Azad 2011-02-07 10:02:16 EST

(In reply to comment #5)
> There are four tests in question here
> - TextMoveLineTest
> - JavaMoveLineTest
> - JavaReplaceAllTest
> - JavaReplaceAllWithQuickDiffTest
Filed bug 336514 for for JavaReplaceAllTest. This bug will deal with *MoveLineTests.

Comment 7 Deepak Azad 2011-02-10 04:18:04 EST

(In reply to comment #6)
> Filed bug 336514 for for JavaReplaceAllTest. This bug will deal with
> *MoveLineTests.

I think both these bugs are caused by Bug 323044.

Comment 8 Deepak Azad 2011-02-14 04:54:08 EST

Created attachment 188881 [details]
proposed fix

Comment 9 Deepak Azad 2011-02-14 04:57:23 EST

Calling EditorTestHelper.runEventQueue() multiple times seems to be executing the same event multiple times on Linux (a bug ?). If I just call it just once, the test seems to be working, at least it does not fail on my Linux machine anymore.

I have tested this fix on WinXP as well.

Comment 10 Deepak Azad 2011-02-14 04:58:07 EST

Dani, ok to release this fix in HEAD ?

Comment 11 Dani Megert 2011-02-14 04:59:11 EST

(In reply to comment #9)
> Calling EditorTestHelper.runEventQueue() multiple times seems to be executing
> the same event multiple times on Linux (a bug ?).
"seems" or "does"? If "does" then it's a bug. You should be able to check that using a display filter.

Comment 12 Deepak Azad 2011-02-14 05:32:38 EST

(In reply to comment #11)
> (In reply to comment #9)
> > Calling EditorTestHelper.runEventQueue() multiple times seems to be executing
> > the same event multiple times on Linux (a bug ?).
> "seems" or "does"? If "does" then it's a bug. You should be able to check that
> using a display filter.

The display filter gets the same number of key events that we generate from our code. It could be an OS issue...

Comment 13 Dani Megert 2011-02-14 06:05:16 EST

(In reply to comment #10)
> Dani, ok to release this fix in HEAD ?
First
- Verify that the changes don't affect the performance (if it does, we need to
  backport the fix.
- Update the copyright date.
- Get rid of remaining (old) warnings in that file.
If that's done you can commit to HEAD.

Comment 14 Deepak Azad 2011-02-14 07:36:11 EST

Created attachment 188894 [details]
final fix

Fixed in HEAD.

Performance numbers are not affected.

*Without patch*
Scenario 'org.eclipse.jdt.text.tests.performance.JavaMoveLineTest#test()' (average over 5 samples):
  Used Java Heap:        -9.65M         (95% in [-33.69M, 14.4M])      Measurable effect: 34.64M (1.8 SDs) (required sample size for an effect of 5% of stdev: 6401)
  Working Set:           -1.25M         (95% in [-2.68M, 174.51K])     Measurable effect: 2.05M (1.8 SDs) (required sample size for an effect of 5% of mean: 5356)
  Committed:             -1.29M         (95% in [-2.63M, 44.48K])      Measurable effect: 1.93M (1.8 SDs) (required sample size for an effect of 5% of mean: 4435)
  Working Set Peak:          0          (95% in [0, 0])               
  Elapsed Process:        2.01s         (95% in [1.95s, 2.08s])        Measurable effect: 91ms (1.8 SDs)
  Kernel time:            1.04s         (95% in [929ms, 1.15s])        Measurable effect: 159ms (1.8 SDs) (required sample size for an effect of 5% of mean: 47)
  Page Faults:             440          (95% in [-64, 945])            Measurable effect: 727 (1.8 SDs) (required sample size for an effect of 5% of mean: 5458)
  CPU Time:               2.14s         (95% in [1.84s, 2.43s])        Measurable effect: 420ms (1.8 SDs) (required sample size for an effect of 5% of mean: 78)
  GDI Objects:               0          (95% in [0, 0])               

*With patch*
Scenario 'org.eclipse.jdt.text.tests.performance.JavaMoveLineTest#test()' (average over 5 samples):
  Used Java Heap:       -13.68M         (95% in [-15.82M, -11.54M])    Measurable effect: 3.08M (1.8 SDs) (required sample size for an effect of 5% of mean: 102)
  Working Set:            1.43M         (95% in [-2.97M, 5.84M])       Measurable effect: 6.34M (1.8 SDs) (required sample size for an effect of 5% of stdev: 6401)
  Committed:              1.46M         (95% in [-3.11M, 6.03M])       Measurable effect: 6.58M (1.8 SDs) (required sample size for an effect of 5% of stdev: 6401)
  Working Set Peak:          0          (95% in [0, 0])               
  Elapsed Process:         1.9s         (95% in [1.83s, 1.97s])        Measurable effect: 96ms (1.8 SDs) (required sample size for an effect of 5% of mean: 6)
  Kernel time:             990ms        (95% in [849ms, 1.13s])        Measurable effect: 203ms (1.8 SDs) (required sample size for an effect of 5% of mean: 85)
  Page Faults:             1.2K         (95% in [-419, 2.82K])         Measurable effect: 2.33K (1.8 SDs) (required sample size for an effect of 5% of stdev: 6401)
  CPU Time:               2.15s         (95% in [1.81s, 2.49s])        Measurable effect: 494ms (1.8 SDs) (required sample size for an effect of 5% of mean: 106)
  GDI Objects:               0          (95% in [0, 0])

Comment 15 Deepak Azad 2011-02-16 08:42:12 EST

The test completed successfully on the latest I-Build
http://download.eclipse.org/eclipse/downloads/drops/I20110215-0800/performance/eplnx2/Scenario114.html

Comment 16 Satyam Kandula 2011-03-02 04:21:29 EST

Deepak, 
Can you please release the fix on the baseline branch too? There is a huge regression on Linux.

Comment 17 Dani Megert 2011-03-02 04:23:49 EST

(In reply to comment #16)
> Deepak, 
> Can you please release the fix on the baseline branch too? There is a huge
> regression on Linux.

This contradicts to comment 14: "Performance numbers are not affected."

Details please.

Comment 18 Satyam Kandula 2011-03-02 04:30:03 EST

(In reply to comment #17)
> This contradicts to comment 14: "Performance numbers are not affected."
> 
> Details please.
I think the numbers were taken on windows machine and there is no regression there. On Linux, the baseline shows 73ms and the 3.7 shows 975ms.

Comment 19 Deepak Azad 2011-03-02 04:36:06 EST

There is no regression in the sense that the performance has NOT degraded, as can be seen on Windows. 

However the baseline for Linux isn't perfect, because the test was also not perfect. However the numbers on Linux after the fix are consistent and similar to the numbers on other platforms.

Comment 20 Dani Megert 2011-03-02 04:38:35 EST

(In reply to comment #19)
> There is no regression in the sense that the performance has NOT degraded, as
> can be seen on Windows. 
> 
> However the baseline for Linux isn't perfect, because the test was also not
> perfect. However the numbers on Linux after the fix are consistent and similar
> to the numbers on other platforms.

Sounds good then. Please backport to 'perf_36x' and release it into the 'org.eclipse.releng' project from the same branch.

Comment 21 Deepak Azad 2011-03-07 06:04:46 EST

Fix backported to 'perf_36x' and released.

Comment 22 Deepak Azad 2011-03-07 06:05:44 EST

.