tightly-coupled memory performance !!

The called function assembly code

05009254 <corner_turn_main>:

5009254: 2015883a mov r10,r4

5009258: 0013883a mov r9,zero

500925c: 00001506 br 50092b4 <corner_turn_main+0x60>

5009260: 40800017 ldw r2,0(r8)

5009264: 3885883a add r2,r7,r2

5009268: 11000017 ldw r4,0(r2)

500926c: 01400784 movi r5,30

5009270: 3145383a mul r2,r6,r5

5009274: 1245883a add r2,r2,r9

5009278: 1085883a add r2,r2,r2

500927c: 1085883a add r2,r2,r2

5009280: 00c14474 movhi r3,1297

5009284: 18e7c604 addi r3,r3,-24808

5009288: 10c5883a add r2,r2,r3

500928c: 11000015 stw r4,0(r2)

5009290: 00c00044 movi r3,1

5009294: 30cd883a add r6,r6,r3

5009298: 00800104 movi r2,4

500929c: 388f883a add r7,r7,r2

50092a0: 317fef1e bne r6,r5,5009260 <corner_turn_main+0xc>

50092a4: 48d3883a add r9,r9,r3

50092a8: 5095883a add r10,r10,r2

50092ac: 00800a04 movi r2,40

50092b0: 48800426 beq r9,r2,50092c4 <corner_turn_main+0x70>

50092b4: 5011883a mov r8,r10

50092b8: 000d883a mov r6,zero

50092bc: 000f883a mov r7,zero

50092c0: 003fe706 br 5009260 <corner_turn_main+0xc>

50092c4: f800283a ret

The calling function assembly code:

5008cc8: 04010034 movhi r16,1024

5008ccc: 84062804 addi r16,r16,6304

5008cd0: 84800037 ldwio r18,0(r16)

5008cd4: a009883a mov r4,r20

5008cd8: 50092540 call 5009254 <corner_turn_main>

5008cdc: 84400037 ldwio r17,0(r16)

5008ce0: 8ca3c83a sub r17,r17,r18

5008ce4: 04c000b4 movhi r19,2

5008ce8: 9cc7af04 addi r19,r19,7868

5008cec: 9809883a mov r4,r19

5008cf0: 880b883a mov r5,r17

5008cf4: 000feb00 call feb0 <printf>

5008cf8: 84800037 ldwio r18,0(r16)

5008cfc: a009883a mov r4,r20

5008d00: 50092540 call 5009254 <corner_turn_main>

5008d04: 84000037 ldwio r16,0(r16)

5008d08: 84a1c83a sub r16,r16,r18

5008d0c: 9809883a mov r4,r19

5008d10: 800b883a mov r5,r16

5008d14: 000feb00 call feb0 <printf>