Icarus hangs when running large number of tests

[GoogleCodeExporter]

I've met with no small amount of success in using Icarus to run my NUnit tests, 
but I've hit a roadblock. It seems that I can't add the 381st [Test] without 
hanging Icarus.

What steps will reproduce the problem?
1. In my case, simply add a 381st [Test] (even one that does nothing)
2.
3.

What is the expected output? What do you see instead?
I expected to be able to run all tests as I have been doing for a couple of 
weeks now. Instead, Icarus hangs, having seemingly attempted to run no tests 
(ie the 'Test Results' tab is showing '0 tests - 0 passed - 0 failed - 0 
inconclusive - 0 skipped - 00:00:00), and the status bar declares 'Running 
tests - Running the tests. (10%)

What version of the product are you using? On what operating system?
Gallio Icarus - Version 3.2 build 517

Please provide any additional information below.
If I comment out my just-added 381st test (or in fact any one of the 380 other 
tests), rebuild, reload, and rerun in Icarus -- all (380) tests run just fine 
again.
There doesn't seem to be anything special about that 381st test -- it can be 
simply an empty method or one that just does Assert.Fail()

Original issue reported on code.google.com by wayne.br...@gmail.com on 9 Sep 2010 at 4:21

[GoogleCodeExporter]

Sorry - forgot to mention OS -- it's Windows 7 x64

Original comment by wayne.br...@gmail.com on 9 Sep 2010 at 4:22

[GoogleCodeExporter]

Actually, I think I may have found the problem. I have some parameterized tests 
that look like this:

[Test, TestCaseSource("Generator")]
public void Test(object arg1, object arg2) {
  ...
}

public static IEnumerable<TestCaseData> Generator {
  get {
    foreach (...) {
      yield return new TestCaseData(...);
    }
  }
}

When I change these generators from iterators to simply building an array and 
returning it, the tests seem to work fine in Icarus.
I'm guessing that the deferred-execution of 'yield return' is causing some 
timing problem...

Original comment by wayne.br...@gmail.com on 9 Sep 2010 at 4:48

[GoogleCodeExporter]

Please try the latest nightly (http://ccnet.gallio.org/Distributables/), there 
was a problem with spin waiting for an Invoke call.

Original comment by grahamr...@gmail.com on 9 Sep 2010 at 4:55

[GoogleCodeExporter]

Graham,

Thanks, the latest nightly solves my problem (even with the yield returns), 
although the toolbars are all messed up in this build :-)

Original comment by wayne.br...@gmail.com on 9 Sep 2010 at 5:58

[GoogleCodeExporter]

... and the Execution Log is blank

Original comment by wayne.br...@gmail.com on 9 Sep 2010 at 6:06

[GoogleCodeExporter]

... and it crashed as I was navigating the test tree [I guess I should be 
filing these separately]

Original comment by wayne.br...@gmail.com on 9 Sep 2010 at 6:07

[GoogleCodeExporter]

Original comment by Yann.Tre...@gmail.com on 10 Sep 2010 at 5:48

• Changed state: Started
• Added labels: Priority-Medium, Type-Defect, Component-Icarus, Milestone-3.2

[GoogleCodeExporter]

I'm pretty sick of those toolbars :( How are they messed up? Could you attach a 
screenshot?

Can you repro the crash? There's a black box log in %APPDATA%\Gallio\Icarus, 
but it'll be gone if you've used it since.

Original comment by grahamr...@gmail.com on 11 Sep 2010 at 7:32

• Changed title: Icarus hangs when running large number of tests

[GoogleCodeExporter]

The toolbars are in odd positions (compared to where they start out in the 3.2 
release build). If I move them, they don't remember where I put them when I 
relaunch the app (not sure if this is a new bug or not). Anyway I've attached a 
screenshot. 

I didn't capture the log for the crash, but I'll try to reproduce it.

Original comment by wayne.br...@gmail.com on 13 Sep 2010 at 12:55

Attachments:

• IcarusTools.JPG

[GoogleCodeExporter]

Actually, I just noticed that a minimize and restore of the window puts to 
toolbars in their correct positions -- they only start out badly positioned 
when I first launch the app.

Original comment by wayne.br...@gmail.com on 13 Sep 2010 at 12:58

[GoogleCodeExporter]

Execution Log is fixed in 593

Original comment by wayne.br...@gmail.com on 13 Sep 2010 at 5:30

[GoogleCodeExporter]

Original comment by Yann.Tre...@gmail.com on 21 Sep 2010 at 1:46

• Added labels: Milestone-3.3

[GoogleCodeExporter]

Having looked at the toolbar issue more closely, I think it's just the crappy 
nature of the WinForms ToolStrip. I can't even get it to stay put at design 
time.

Original comment by wayne.br...@gmail.com on 2 Dec 2010 at 6:45

[GoogleCodeExporter]

I think the original problem of hanging on running tests have been resolved -- 
I haven't seen that happen in a while, so I vote to close this as resolved.

Original comment by wayne.br...@gmail.com on 2 Dec 2010 at 6:48

[GoogleCodeExporter]

Original comment by Yann.Tre...@gmail.com on 14 Jun 2011 at 5:53

• Added labels: Milestone-IdeaPool

[GoogleCodeExporter]

Facing the same issue with Icarus after I've upgraded to selenium 2.44. I'm 
using MBunit tests. 

Icarus seemed to work fine when I was using 2.37 selenium c# bindings. But now 
Icarus gets stuck at 25-26% forever.

I'm running 300 tests at a go.

Original comment by rupam08 on 26 Jan 2015 at 9:02

[GoogleCodeExporter]

sorry forgot to mention test harness details - 

OS - Win 7, x64
Test  - MbUnit
Selenium - 2.44 C# .Net bindings
IDE - MS Visual Studio 2013

Original comment by rupam08 on 26 Jan 2015 at 9:12