[Insight-developers] ImageRegionConstIterator - thread safe?

[Nicholas Tustison]

Hi Michael,

Just for clarification, and so that nobody gets worried, I wouldn't say that
we're going to replace the current B-spline transform but, since we're 
going to have the DeformableTransform framework, it will be trivial
to add one constrained by B-splines which has practical advantages
over the current framework.

Nick


On Apr 13, 2011, at 5:15 PM, M Stauffer (V) wrote:

> Thanks Luis. This may actually end up being moot for the case of
> itkBSplineDeformaableTransform since we're planning on replacing this
> with Nick's new BSpline transform based on DeformationFieldTransform.
> 
> Yes, a more certain way to test for thread-safety would be wonderful!
> Maybe we could cook up a test option that specifies to run tests that
> are designated as testing thread-safety many times (e.g. 1000x) instead
> of the single run in a normal test run? These could be run periodically
> to avoid having each test run take more time. 
> 
> Maybe a cmake var THREAD_SAFETY_TEST_REPITITIONS that is usually 1, but
> gets set to 1000 (or 10000??) once a week or month or so, and individual
> tests can pull in the var through the command params in CMakeLists and
> use it for their testing loop as appropriate.
> 
> We could do this manually for individual tests of course, but seems it
> would give a more peace of mind to be able to do this on lots of
> different platforms periodically.
> 
> -M
> 
>> -----Original Message-----
>> From: Luis Ibanez [mailto:luis.ibanez at kitware.com] 
>> Sent: Sunday, April 10, 2011 6:38 PM
>> To: M Stauffer (V)
>> Cc: ITK-dev-list-mstauff at ver
>> Subject: Re: [Insight-developers] ImageRegionConstIterator - 
>> thread safe?
>> 
>> Hi Marius,
>> 
>> I would think that it is thread-safe,
>> 
>> at least in the sense that a "Const" iterator
>> will not modify its internal member variables.
>> 
>> A "grep" didn't return any instances of "mutable"
>> in this family of classes either,
>> 
>> So it looks safe...
>> 
>> I wish there was a less uncertain way of
>> making this determination...
>> 
>> 
>>     Luis
>> 
>> 
>> -----------------------------------------------------------
>> On Tue, Apr 5, 2011 at 5:18 PM, M Stauffer (V) 
>> <mstauff at verizon.net> wrote:
>>> Hi,
>>> 
>>> Does anyone know if ImageRegionConstIterator is thread-safe for
>>> read-only operations?
>>> 
>>> I'm trying to determine the thread-safety of
>>> itkBSplineDeformaableTransform::TransformPoint( const 
>> InputPointType &
>>> point, OutputPointType & outputPoint,  WeightsType & weights,
>>> ParameterIndexArrayType & indices, bool & inside). The
>>> ImageRegionConstIterator used here is the only unknown for me at this
>>> point, the rest of it looks thread-safe.
>>> 
>>> I can look at ImageRegionConstIterator myself in detail, but 
>> thought I'd
>>> check first if someone else already knows.
>>> 
>>> Cheers,
>>> Michael
>>> 
>>> _______________________________________________
>>> Powered by www.kitware.com
>>> 
>>> Visit other Kitware open-source projects at
>>> http://www.kitware.com/opensource/opensource.html
>>> 
>>> Kitware offers ITK Training Courses, for more information visit:
>>> http://kitware.com/products/protraining.html
>>> 
>>> Please keep messages on-topic and check the ITK FAQ at:
>>> http://www.itk.org/Wiki/ITK_FAQ
>>> 
>>> Follow this link to subscribe/unsubscribe:
>>> http://www.itk.org/mailman/listinfo/insight-developers
>>> 
> 
> _______________________________________________
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
> 
> Kitware offers ITK Training Courses, for more information visit:
> http://kitware.com/products/protraining.html
> 
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
> 
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers