river-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Firmstone <j...@zeus.net.au>
Subject Re: IMUnit - Improved Multithreaded Unit testing
Date Sun, 16 Oct 2011 11:34:31 GMT
Well I've had a little play, simple to learn, powerful and good at 
simulating corner case thread situations.  The only drawback is you have 
to be careful about event names in your schedule syntax, it's easy for a 
typo to ruin a test, although in this case it's obvious because the test 
doesn't run, but it doesn't throw an exception to hint where the syntax 
error is either.

Helpful advise: Run your tests often while creating them, to pick up any 
string syntax errors.

Anyway one test I've written checks correct behaviour for cleared 
Reference's being treated as absent in ReferenceConcurrentMap.putIfAbsent

In this case,  putIfAbsent is followed by reference clear, followed by 
three concurrent putIfAbsent's, the result is as expected, but it's good 
confirm it works.

Cheers,

Peter.

Dan Creswell wrote:
> No objection until there's a reason for one - evidence wins over
> speculation. Use it and see....
>
> On 15 October 2011 10:10, Peter <jini@zeus.net.au> wrote:
>   
>> http://mir.cs.illinois.edu/imunit/
>>
>> No objections to me using this for concurrency unit testing in River?
>>
>> Cheers,
>>
>> Peter.
>>     
>
>   


Mime
View raw message