- * Major bug. We can't handle the case of a
- * SEG_DIFFERENCE expression in a synthetic opcode
- * variable-length instruction.
- * We don't have a frag type that is smart enough to
- * relax a SEG_DIFFERENCE, and so we just force all
- * SEG_DIFFERENCEs to behave like SEG_PASS1s.
- * Clearly, if there is a demand we can invent a new or
- * modified frag type and then coding up a frag for this
- * case will be easy. SEG_DIFFERENCE was invented for the
- * .words after a CASE opcode, and was never intended for
- * instruction operands.
- */
+ * Major bug. We can't handle the case of a operator
+ * expression in a synthetic opcode variable-length
+ * instruction. We don't have a frag type that is smart
+ * enough to relax a operator, and so we just force all
+ * operators to behave like SEG_PASS1s. Clearly, if there is
+ * a demand we can invent a new or modified frag type and
+ * then coding up a frag for this case will be easy.
+ */