Reduces LUT needed by mul extension

bench/tests/op/divw.s

@@ -1,25 +1,60 @@
-# expected: ffffffffffffffff, ffffffffffffffff, dead0, 1, ffffffff80000000, 7
     .text
 _start:
-    li x2, 0xdead0
-    li x3, 0x0
-    li x4, 0x1
-    li x5, (-1)
-    li x6, 0xdead0
-    li x7, (-2355)
-    li x8, (-324)
-    li x9, 0x80000000
+    li x2,  0xdead0        # 912080
+    li x3,  0x1234         # 4660
+    li x4,  1              # 1
+    li x5, -123456         # -123456
+    li x6,  0xdead0        # duplicate of x2
+    li x7, -765432         # -765432
+    li x8,  123456         # 123456
+    li x9, -234567         # -234567
+    li x10, 0x7fffffff  # max positive 32b
+    li x11, 0x80000000  # min negative 32b
 
-    # Div by 0
-    divw x31, x2, x3
-    divw x31, x5, x3
-
-    # Basic divisions
+    # 1. DIVW : a / 1
     divw x31, x2, x4
+    # expected: 000000000000dead0
+
+    # 2. DIVW : a / a
     divw x31, x2, x6
+    # expected: 0000000000000001
 
-    # Overflow
-    divw x31, x9, x5
+    # 3. DIVW : negative / positive
+    divw x31, x5, x8
+    # expected: ffffffffffffffff
 
-    # Signed / signed
+    # 4. DIVW : negative / positive
     divw x31, x7, x8
+    # expected: fffffffffffffffa
+
+    # 5. DIVW : positive / negative
+    divw x31, x8, x7
+    # expected: 0000000000000000
+
+    # 6. DIVW : negative / negative
+    divw x31, x7, x9
+    # expected: 0000000000000003
+
+    # 7. DIVW : max / 2
+    li x12, 2
+    divw x31, x10, x12
+    # expected: 000000003fffffff
+
+    # 8. DIVW : min / -1
+    li x12, -1  # -1
+    divw x31, x11, x12
+    # expected: ffffffff80000000
+
+    # 9. DIVW : min / 1
+    divw x31, x11, x4
+    # expected: ffffffff80000000
+
+    # 10. DIVW : 0 / any
+    divw x31, x0, x2
+    # expected: 0000000000000000
+
+    # 11. DIVW : any / 0
+    divw x31, x2, x0
+    # expected: ffffffffffffffff
+    divw x31, x7, x0
+    # expected: ffffffffffffffff