Trouble executing N4ITK binaries compiled under MacOS 15.1

[davidmummy]

We're having problems executing the N4ITK binary when we rebuild it under MacOS 15.1. We're using the exact same command to run it as previous, but now we get errors like the one pictured. At first I thought it we needed quotes around the input arguments or something, and that didn't work; and we've verified that all the permissions are there. Several hours of troubleshooting yielded nothing and I'm at my wits' end. It just gets stuck on that [25,0] and dies.

Seems like this is likely some basic issue where I'm doing something dumb, but we've replicated it on two separate installations now and had three people troubleshoot and we seem to be collectively blind to it. Any help would be much appreciated. Thanks!

[ntustison]

@cookpa would probably be better on this but try quotes around the brackets. This happened to me when I went from bash to zsh on the mac.

[davidmummy]

Many thanks to both of you for your quick replies -- it's really appreciated. We did some additional troubleshooting and at least have gotten to a different outcome.

We're using the bash -c command suggested by @cookpa . Instead of giving us any kind of specific execution error, it just runs for 30-45 seconds and shows this "kill" result and doesn't produce any output files in the -o folder. Does this look familiar at all? Thanks again. We're out of our depth here. FYI we're using this on 128x128x128 grayscale hyperpolarized gas images, and before the OS upgrade we were able to use N4ITK command on them with no problem whatsoever.

[cookpa]

That is usually a result of being out of memory. Not sure why the upgrade would cause that though.

Can you add --verbose and post the output?

If you can share the images I'll run it myself

[ntustison]

Also, try running with the default -t parameters (i.e., don't specify anything for -t).

[davidmummy]

Got it. We'll try both of these and let you know what happens. Thanks again

[davidmummy]

Update! We finally managed to figure this out. The "bash -c" solution did in fact work!

As @cookpa suggested, there was something insidious going on with our test data that was also causing a memory leak. Once we separated those two issues and lined up a better test case, we were able to implement this fix & it's now fully tested and in our code base. Thanks again for all your help -- no way we would we have figured this out without your suggestions. We can now all upgrade to the latest Mac OS without fear. Much appreciated!!