Something went wrong while setting issue due date.
setting ECL_OPT_TRAP_SIGFPE to 0 should be equivalent to (si::trap-fpe t nil)
In #317 (closed) we explained trouble we have in the embedded ECL setting (in sagemath system) with disabling ECL's FP overflow trap for overflows happening outside of ECL.
In particular setting ECL_OPT_TRAP_SIGFPE to 0 is not enough. However it turns out that (si::trap-fpe t nil).
My understanding is that the latter is equivalent to ECL's command line switch --no-trap-fpe, so this is not too strange.