This matches Java behavior, which is specifically called out in JLS https://docs.oracle.com/javase/specs/jls/se7/html/jls-15.html#jls-15.17.2:

“There is one special case that does not satisfy this rule: if the dividend is the negative integer of largest possible magnitude for its type, and the divisor is -1, then integer overflow occurs and the result is equal to the dividend. Despite the overflow, no exception is thrown in this case.”

This matches Java behavior, which is specifically called out in JLS https://docs.oracle.com/javase/specs/jls/se7/html/jls-15.html#jls-15.17.2:

> There is one special case that does not satisfy this rule: if the dividend is the negative integer of largest possible magnitude for its type, and the divisor is -1, then integer overflow occurs and the result is equal to the dividend. Despite the overflow, no exception is thrown in this case.

Patch clj-1253-1.txt corrects LongOps method divide for the case of args Long/MIN_VALUE and -1. It returns a BigInt in this case, not a Long, but most other pairs of values passed to this function return a Ratio exact answer, so it seems reasonable in this one case to return a BigInt exact answer when it will not fit in a Long.