thread race with itkImage::SetPixel() found with thread sanitizer (TSan) about itk HOT 6 CLOSED

This solves the itkStatisticsUniqueLabelMapFilterTest1 test failure under TSan! Amazing! We'll see on cdash if it solves other similar failures too...

from itk.

itkStatisticsUniqueLabelMapFilterTest sounds familiar, because I think it is flaky. So this might be a genuine bug discovery. We shouldn't sweep it under the carpet. Does anyone have time to tackle this, now or in the future?

from itk.

The test makes reference to JIRA issue number 3370:

Which is referenced in this commit: a18b3cf

from itk.

There is this issue: #3031.

from itk.

There is this issue: #3031.

Ha! Totally forgot that I had apparently discovered this 2 years ago too!

We shouldn't sweep it under the carpet.

Agreed.

Does anyone have time to tackle this, now or in the future?

I don't actually use this part of ITK, so I'm afraid I won't tackle it, with so many other things to do.

But I would like to get Mac14.x-AppleClang-dbg-TSan on cdash green, even if it requires suppressing likely bugs, because only if it's green will anyone notice/care if future TSan issues are found.

from itk.

I am suspicious of the algorithm implemented here:
https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Filtering/LabelMap/include/itkAttributeUniqueLabelMapFilter.hxx#L79-L255

I am not sure how only the previous label line can be looked at to ensure they are not over lapping.

from itk.