Open jodavies opened 5 years ago
Perhaps the call of Normalize
which doesn't check the return value at proces.c:3163
is suspicious.
Edit: No, that the term data is printed with "Called from Generator" means that a jump to GenCall
was made.
I don't have an easy way to test since I don't have a "minimal example" for this hang, but I wonder if it is related to #14 ?
Did you try to compile FORM with the patch from @tueda ? For me this actually solved all issues with broken TFORM jobs hanging on our cluster so I was about to suggest to add it to the master branch. So far I didn't observe any side effects.
No, but I haven't seen this error since I made this issue, since I changed the way the scripts were doing things to avoid it.
It is not yet obvious to me where control goes in the code, but there is a way to obtain the error
and have FORM hang indefinitely, rather than crash out. This is annoying when running large arrays of jobs, for example.
I will keep investigating the logic here, but have just made an issue already to document the problem.
Thanks, Josh.