It is usually stated that the L1D latency is around 4 cycles. How are those 4 cycles utilized?
1c: Calculate effective address
1c: Send request from core to cache
1c: Do cache access
1c: Send data back to core pipeline
Is there any available information on that?