The instruction queue contains multiple instruction FIFOs to handle multiple parallel instructions at the same time.
The number of FIFOs equals to INSTR_PER_FETCH, and the depth of each FIFO equals FIFO_DEPTH.
The instruction queue also contains another FIFO for addresses in case of address prediction, with the same depth, FIFO_DEPTH.
The actual depth of the instruction queue is not obvious due to this incompatibility between the address FIFO and the instruction FIFOs.
Is there an existing CVA6 bug for this?
Bug Description
The instruction queue contains multiple instruction FIFOs to handle multiple parallel instructions at the same time. The number of FIFOs equals to INSTR_PER_FETCH, and the depth of each FIFO equals FIFO_DEPTH. The instruction queue also contains another FIFO for addresses in case of address prediction, with the same depth, FIFO_DEPTH. The actual depth of the instruction queue is not obvious due to this incompatibility between the address FIFO and the instruction FIFOs.