InternalError exceptions are assertions within the compiler. I would expect them to be deterministic unlike assertions in the GC which depend a lot on timing of threads. It should be possible to get to the bottom of this if I'm able to get the ML code that is causing the problem. I appreciate this may be easier said than done.
Just to check: is this with the release, with the current fixes branch or with SVN trunk? I seem to remember fixing something a bit like this.
David
On 27/02/2014 03:44, Michael Norrish wrote:
I seem to be able to pretty reliably produce the following:
Exception- InternalError: indirect - invalid constant address raised while compiling
Exception- InternalError: indirect - invalid constant address raised while compiling
Unfortunately, it takes hours to get there...
Switching to --gcthreads=1 may have fixed another error we were having, relating to unwarranted pointer-equality tests.
Michael