Let us calculate effective CPI here as we need to find in no of cycles only.
Given that there may be either instruction cache miss or data cache miss accordingly we will have each stall..
So CPI(if it is an ALU instruction) = CPI(ideal) + Stall / instruction
= 1 + 0.1 * 50
= 6
CPI(if it is a branch instruction) = 1.5 + 0.1 * 50
= 6.5
CPI(if it is a jump instruction) = 1 + 0.1 * 50
= 6
CPI(if it is a load/store instruction) = 1.5 + 0.1 * 50 + 0.2 * 50
= 16.5
Therefore effective CPI = 0.4 * 6 + 0.3 * 16.5 + 0.2 * 6.5 + 0.1 * 6
= 2.4 + 4.95 + 1.3 + 0.6
= 9.25 cycles
Hence B) should be correct answer..