Lines Matching refs:TODO

70 // TODO: Non-conditional direct branches take zero cycles and units.
73 // TODO: Branch and link is much different.
77 // TODO: Shift over 3 and some extensions take 2 cycles.
86 // TODO: Division blocks the divider inside C.
89 // TODO: Long multiplication take 5 cycles and also the ALU.
90 // TODO: Multiplication with accumulation can be advanced.
92 // TODO: 64-bit multiplication has a throughput of 1/2.
99 // TODO: The latency for the post or pre register is 1 cycle.
104 // TODO: Extended address requires also the ALU.
110 // TODO: Extended address requires also the ALU.
117 // TODO: FCCMP is much different.
119 // TODO: DP takes longer.
121 // TODO: MACC takes longer.
125 // TODO: Conversion between register files is much different.
128 // TODO: Copy from FPR to GPR is much different.
132 // TODO: ASIMD loads are much different.
136 // TODO: ASIMD stores are much different.
140 // TODO: Other operations are much different.
152 // TODO: Add FP register forwarding rules.
159 // TODO: The forwarding for WriteIM64 saves actually 3 cycles.