New sw folders

Resolve "Network Processor"

Closes #93

See merge request bslathi19/super6502!74

.gitignore

@@ -8,6 +8,7 @@
 *.map
 *.list
 *.bin
+*.hex
 *.o
 
 

.gitlab-ci.yml

@@ -12,7 +12,8 @@ build:
     - efinity
     - linux
   script:
-    - source init_env.sh
+    # - source init_env.sh
+    - module load efinity/2024.1
     - make
   after_script:
     - cat hw/super6502_fpga/outflow/super6502_fpga.err.log

Makefile

@@ -1,4 +1,4 @@
-ROM_TARGET=test_code/loop_test
+ROM_TARGET=bootloader
 
 INIT_HEX=hw/super6502_fpga/init_hex.mem
 HEX=sw/$(ROM_TARGET)/$(notdir $(ROM_TARGET)).bin
@@ -26,7 +26,7 @@ $(CC65):
 	$(MAKE) -C sw/toolchain/cc65 -j $(shell nproc)
 
 $(INIT_HEX): $(CC65) script/generate_rom_image.py $(HEX)
-	python script/generate_rom_image.py -i $(HEX) -o $@
+	python3 script/generate_rom_image.py -i $(HEX) -o $@
 
 # This should get dependencies of rom, not be phony
 .PHONY: $(HEX)

hw/super6502_fpga/src/rtl/super_6502_fpga.sv

@@ -91,7 +91,9 @@ assign pre_resetn = button_resetn & vio0_resetn;
 
 assign sdram_ready = |w_sdr_state;
 
-assign master_resetn = pre_resetn & sdram_ready;
+always_ff @(posedge i_sysclk) begin
+    master_resetn <= pre_resetn & sdram_ready;
+end
 
 assign o_sd_cs = '1;
 
@@ -214,11 +216,14 @@ logic  [1:0]                sd_controller_dma_RRESP;
 axil_intf ntw_reg();
 axil_intf ntw_dma();
 
+logic cpu_wrapper_reset;
+always_ff @(posedge i_sysclk) cpu_wrapper_reset <= ~master_resetn;
+
 
 cpu_wrapper u_cpu_wrapper_0(
     .i_clk_cpu  (clk_cpu),
     .i_clk_100  (i_sysclk),
-    .i_rst      (~master_resetn),
+    .i_rst      (cpu_wrapper_reset),
 
     .o_cpu_rst  (o_cpu0_reset),
     .o_cpu_rdy  (o_cpu0_rdy),
@@ -258,6 +263,8 @@ cpu_wrapper u_cpu_wrapper_0(
     .i_nmi('0)
 );
 
+logic crossbar_resetn;
+always_ff @(posedge i_sysclk) crossbar_resetn <= master_resetn;
 
 axilxbar #(
     .NM(3),
@@ -271,7 +278,7 @@ axilxbar #(
     })
 ) u_crossbar (
     .S_AXI_ACLK         (i_sysclk),
-    .S_AXI_ARESETN      (master_resetn),
+    .S_AXI_ARESETN      (crossbar_resetn),
 
     .S_AXI_ARADDR       ({cpu0_ARADDR,  sd_controller_dma_ARADDR,       ntw_dma.araddr  }),
     .S_AXI_ARVALID      ({cpu0_ARVALID, sd_controller_dma_ARVALID,      ntw_dma.arvalid }),
@@ -310,13 +317,16 @@ axilxbar #(
 
 );
 
+logic rom_reset;
+always_ff @(posedge i_sysclk) rom_reset <= ~master_resetn;
+
 axi4_lite_rom #(
     .ROM_SIZE(12),
     .BASE_ADDRESS(32'h0000f000),
     .ROM_INIT_FILE("init_hex.mem")
 ) u_rom (
     .i_clk(i_sysclk),
-    .i_rst(~master_resetn),
+    .i_rst(rom_reset),
 
     .o_AWREADY(rom_awready),
     .o_WREADY(rom_wready),
@@ -344,12 +354,15 @@ axi4_lite_rom #(
     .i_WSTRB(rom_wstrb)
 );
 
+logic ram_reset;
+always_ff @(posedge i_sysclk) ram_reset <= ~master_resetn;
+
 axi4_lite_ram #(
     .RAM_SIZE(9),
     .ZERO_INIT(1)
 ) u_ram(
     .i_clk(i_sysclk),
-    .i_rst(~master_resetn),
+    .i_rst(ram_reset),
 
     .o_AWREADY(ram_awready),
     .o_WREADY(ram_wready),
@@ -452,12 +465,15 @@ sdram_controller u_sdram_controller(
 
 logic sd_irq;
 
+logic sd_controller_wrapper_reset;
+always_ff @(posedge i_sysclk) sd_controller_wrapper_reset <= ~master_resetn;
+
 sd_controller_wrapper #(
     .NUMIO              (1),   // board as it stands is in 1 bit mode
     .BASE_ADDRESS       (32'h0000E000)
 ) u_sdio_top (
     .i_clk              (i_sysclk),
-    .i_reset            (~master_resetn),
+    .i_reset            (sd_controller_wrapper_reset),
 
     .S_AXIL_AWVALID     (sd_controller_ctrl_AWVALID),
     .S_AXIL_AWREADY     (sd_controller_ctrl_AWREADY),
@@ -506,11 +522,14 @@ sd_controller_wrapper #(
     .o_int              (sd_irq)
 );
 
+logic network_processor_reset;
+always_ff @(posedge i_sysclk) network_processor_reset <= ~master_resetn;
+
 network_processor #(
-    .NUM_TCP(4)
+    .NUM_TCP(2)
 ) u_network_processor (
     .i_clk              (i_sysclk),
-    .i_rst              (~master_resetn),
+    .i_rst              (network_processor_reset),
 
     .s_reg_axil         (ntw_reg),
     .m_dma_axil         (ntw_dma),

hw/super6502_fpga/src/sim/Makefile

@@ -23,7 +23,7 @@ all: waves
 
 waves: $(TB_NAME)
 #	./$(TB_NAME) -fst
-	./obj_dir/Vsim_top
+	./obj_dir/Vsim_top | tee run.log
 
 $(TB_NAME): $(SUPER6502_FPGA_SOURCES) $(SIM_SOURCES) $(COPY_FILES) $(SD_IMAGE)
 # $(IVERILOG) -g2005-sv $(FLAGS) -s $@ -o $@ $(INCLUDE) $(SUPER6502_FPGA_SOURCES) $(SIM_SOURCES) -I ../../

hw/super6502_fpga/src/sim/hvl/sim_top.sv

@@ -197,8 +197,12 @@ initial begin
     button_resetn <= '0;
     repeat(10) @(clk_cpu);
     button_resetn <= '1;
-    repeat(20000) @(posedge clk_cpu);
+    repeat(2000) @(posedge clk_cpu);
     $finish();
 end
 
+always @(posedge w_cpu0_sync) begin
+    $display("[%t] CPU Executed %x:%x", $time(), w_cpu0_addr, w_cpu0_data_from_dut);
+end
+
 endmodule

hw/super6502_fpga/src/sub/network_processor/sim/cocotb/tests/scapy_irl_test.py

@@ -1,7 +1,3 @@
-from http import server
-from turtle import xcor
-from scapy.layers.inet import Ether, IP, TCP
-from scapy.layers.l2 import ARP
 from scapy.data import IP_PROTOS
 
 from scapy import sendrecv
@@ -18,8 +14,9 @@ from cocotbext.axi import AxiLiteBus, AxiLiteMaster, AxiLiteRam
 from cocotbext.eth import MiiPhy, GmiiFrame
 import struct
 
-from scapy.layers.inet import Ether, IP, TCP
+from scapy.layers.inet import IP, TCP
-from scapy.layers.l2 import ARP
+from scapy.layers.l2 import ARP, Ether
+from scapy.packet import Packet
 from scapy.utils import PcapWriter
 
 from scapy.layers.tuntap import TunTapInterface
@@ -27,6 +24,8 @@ import logging
 
 from decimal import Decimal
 
+import random
+
 CLK_PERIOD_NS = 10
 
 MII_CLK_PERIOD_NS = 40
@@ -35,7 +34,7 @@ MII_CLK_PERIOD_NS = 40
 import socket
 
 # In order for this to work, you need to run these commands:
-# sudo ip tuntap add name tun0 mode tun user $USER
+# sudo ip tuntap add name tun0 mode tun group netdev
 # sudo ip a add 172.0.0.1 peer 172.0.0.2 dev tun0
 # sudo ip link set tun0 up
 
@@ -99,6 +98,33 @@ def ip_to_hex(ip: str) -> int:
 async def test_irl(dut):
     tb = TB(dut)
 
+    async def read_tcp_from_dut():
+        resp = await tb.mii_phy.tx.recv() # type: GmiiFrame
+        packet = Ether(resp.get_payload())
+        tb.log.info(f"Packet Type: {packet.type:x}")
+
+        ip_packet = packet.payload
+        assert isinstance(ip_packet, IP)
+
+        tcp_packet = ip_packet.payload
+        assert isinstance(tcp_packet, TCP)
+
+        tb.log.info(f"Source Port: {tcp_packet.sport}")
+        tb.log.info(f"Dest Port: {tcp_packet.dport}")
+        tb.log.info(f"Seq: {tcp_packet.seq}")
+        tb.log.info(f"Ack: {tcp_packet.ack}")
+        tb.log.info(f"Data Offs: {tcp_packet.dataofs}")
+        tb.log.info(f"flags: {tcp_packet.flags}")
+        tb.log.info(f"window: {tcp_packet.window}")
+        tb.log.info(f"Checksum: {tcp_packet.chksum}")
+
+        return ip_packet
+
+    #############################
+    # Reset DUT                 #
+    #############################
+
+
     await tb.cycle_reset()
 
     dut_ip = "172.0.0.2"
@@ -106,14 +132,18 @@ async def test_irl(dut):
 
     tb_mac = "02:00:00:11:22:33"
 
+    dut_port = random.randint(1024, 65535)
+    tb_port = random.randint(1024, 65535)
+
     serversocket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-    serversocket.bind((tb_ip, 5678))
+    serversocket.bind((tb_ip, tb_port))
     serversocket.listen(1)
     t = TunTapInterface('tun0')
 
 
-    dut_port = 1234
+    ###############################
-    tb_port = 5678
+    # Configure DUT Network block #
+    ###############################
 
     await tb.axil_master.write_dword(0x0, 0x1807)
 
@@ -150,35 +180,25 @@ async def test_irl(dut):
     assert arp_request.psrc == dut_ip, "ARP psrc does not match expected"
     assert arp_request.pdst == tb_ip, "ARP pdst does not match expected"
 
+    # hardcode the ARP response for now
     arp_response = Ether(dst=dut_mac, src=tb_mac)
     arp_response /= ARP(op="is-at", hwsrc=tb_mac, hwdst=dut_mac, psrc=tb_ip, pdst=dut_ip)
     arp_response = arp_response.build()
 
     await tb.mii_phy.rx.send(GmiiFrame.from_payload(arp_response))
 
-    resp = await tb.mii_phy.tx.recv() # type: GmiiFrame
-    packet = Ether(resp.get_payload())
-    tb.log.info(f"Packet Type: {packet.type:x}")
 
-    ip_packet = packet.payload
+    ###############################
-    assert isinstance(ip_packet, IP)
+    # Start TCP handshake         #
+    ###############################
 
-    tcp_packet = ip_packet.payload
+    ip_packet = await read_tcp_from_dut()
-    assert isinstance(tcp_packet, TCP)
-
-    tb.log.info(f"Source Port: {tcp_packet.sport}")
-    tb.log.info(f"Dest Port: {tcp_packet.dport}")
-    tb.log.info(f"Seq: {tcp_packet.seq}")
-    tb.log.info(f"Ack: {tcp_packet.ack}")
-    tb.log.info(f"Data Offs: {tcp_packet.dataofs}")
-    tb.log.info(f"flags: {tcp_packet.flags}")
-    tb.log.info(f"window: {tcp_packet.window}")
-    tb.log.info(f"Checksum: {tcp_packet.chksum}")
 
     t.send(ip_packet)
 
     while True:
         pkt = t.recv()
+        assert isinstance(pkt, Packet)
         if (pkt.proto == IP_PROTOS.tcp):
             break
     print(pkt)
@@ -187,29 +207,16 @@ async def test_irl(dut):
 
     await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_synack.build()))
 
-    resp = await tb.mii_phy.tx.recv() # type: GmiiFrame
+    ip_packet = await read_tcp_from_dut()
-    packet = Ether(resp.get_payload())
-    tb.log.info(f"Packet Type: {packet.type:x}")
-
-    ip_packet = packet.payload
-    assert isinstance(ip_packet, IP)
-
-    tcp_packet = ip_packet.payload
-    assert isinstance(tcp_packet, TCP)
-
-    tb.log.info(f"Source Port: {tcp_packet.sport}")
-    tb.log.info(f"Dest Port: {tcp_packet.dport}")
-    tb.log.info(f"Seq: {tcp_packet.seq}")
-    tb.log.info(f"Ack: {tcp_packet.ack}")
-    tb.log.info(f"Data Offs: {tcp_packet.dataofs}")
-    tb.log.info(f"flags: {tcp_packet.flags}")
-    tb.log.info(f"window: {tcp_packet.window}")
-    tb.log.info(f"Checksum: {tcp_packet.chksum}")
 
     t.send(ip_packet)
 
     con, addr = serversocket.accept()
 
+    ###############################
+    # Send data from DUT to host  #
+    ###############################
+
     # Construct a descriptor in memry
     tb.axil_ram.write_dword(0x00000000, 0x00001000)
     tb.axil_ram.write_dword(0x00000004, 64)
@@ -232,11 +239,16 @@ async def test_irl(dut):
     con.recv(64)
     tb.log.info("Received 64 packets")
 
+    ###############################
+    # Close connection from host  #
+    ###############################
+
     con.close()
     serversocket.close()
 
     while True:
         pkt = t.recv()
+        assert isinstance(pkt, Packet)
         if (pkt.proto == IP_PROTOS.tcp):
             break
     print(pkt)
@@ -249,6 +261,7 @@ async def test_irl(dut):
 
     while True:
         pkt = t.recv()
+        assert isinstance(pkt, Packet)
         if (pkt.proto == IP_PROTOS.tcp):
             break
     print(pkt)
@@ -257,6 +270,39 @@ async def test_irl(dut):
 
     await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_fin.build()))
 
+    tb.log.info("Expecting to get ACK from DUT")
+
+    ip_packet = await read_tcp_from_dut()
+
+    t.send(ip_packet)
+
+    tb.log.info("Expecting to get FINACK from DUT")
+
+    ip_packet = await read_tcp_from_dut()
+
+    t.send(ip_packet)
+
+    tb.log.info("Expecting to get ACK from host")
+
+    while True:
+        pkt = t.recv()
+        assert isinstance(pkt, Packet)
+        if (pkt.proto == IP_PROTOS.tcp):
+            break
+    print(pkt)
+
+    tcp_fin = Ether(dst=dut_mac, src=tb_mac) / pkt
+
+    await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_fin.build()))
+
+    await Timer(Decimal(CLK_PERIOD_NS * 1000), units='ns')
+
+
+@cocotb.test()
+async def test_close(dut):
+    tb = TB(dut)
+
+    async def read_tcp_from_dut():
         resp = await tb.mii_phy.tx.recv() # type: GmiiFrame
         packet = Ether(resp.get_payload())
         tb.log.info(f"Packet Type: {packet.type:x}")
@@ -276,18 +322,145 @@ async def test_irl(dut):
         tb.log.info(f"window: {tcp_packet.window}")
         tb.log.info(f"Checksum: {tcp_packet.chksum}")
 
-    t.send(ip_packet)
+        return ip_packet
-
-    tb.log.info("Expecting to send last ACK here")
 
+    def get_pkt_from_host():
         while True:
             pkt = t.recv()
+            assert isinstance(pkt, Packet)
             if (pkt.proto == IP_PROTOS.tcp):
                 break
         print(pkt)
+        return pkt
 
-    tcp_fin = Ether(dst=dut_mac, src=tb_mac) / pkt
+    #############################
+    # Reset DUT                 #
+    #############################
 
-    await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_fin.build()))
 
-    await Timer(Decimal(CLK_PERIOD_NS * 1000), units='ns')
+    await tb.cycle_reset()
+
+    dut_ip = "172.0.0.2"
+    tb_ip = "172.0.0.1"
+
+    tb_mac = "02:00:00:11:22:33"
+
+    dut_port = random.randint(1024, 65535)
+    tb_port = random.randint(1024, 65535)
+
+    serversocket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    serversocket.bind((tb_ip, tb_port))
+    serversocket.listen(1)
+    t = TunTapInterface('tun0')
+
+
+    ###############################
+    # Configure DUT Network block #
+    ###############################
+
+    await tb.axil_master.write_dword(0x0, 0x1807)
+
+    await tb.axil_master.write_dword(0x200, dut_port)
+    await tb.axil_master.write_dword(0x204, ip_to_hex(dut_ip))
+    await tb.axil_master.write_dword(0x208, tb_port)
+    await tb.axil_master.write_dword(0x20c, ip_to_hex(tb_ip))
+    await tb.axil_master.write_dword(0x210, 0x3)
+
+    resp = await tb.mii_phy.tx.recv() # type: GmiiFrame
+
+    packet = Ether(resp.get_payload())
+
+    tb.log.info(f"Packet Type: {packet.type:x}")
+
+    assert packet.type == 0x806, "Packet type is not ARP!"
+
+
+    arp_request = packet.payload
+    assert isinstance(arp_request, ARP)
+
+    tb.log.info(f"Arp OP: {arp_request.op}")
+    tb.log.info(f"Arp hwsrc: {arp_request.hwsrc}")
+    tb.log.info(f"Arp hwdst: {arp_request.hwdst}")
+    tb.log.info(f"Arp psrc: {arp_request.psrc}")
+    tb.log.info(f"Arp pdst: {arp_request.pdst}")
+
+    dut_mac = arp_request.hwsrc
+    dut_ip = arp_request.psrc
+
+    assert arp_request.op == 1, "ARP type is not request!"
+    assert arp_request.hwsrc == "02:00:00:aa:bb:cc", "ARP hwsrc does not match expected"
+    assert arp_request.hwdst == "00:00:00:00:00:00", "ARP hwdst does not match expected"
+    assert arp_request.psrc == dut_ip, "ARP psrc does not match expected"
+    assert arp_request.pdst == tb_ip, "ARP pdst does not match expected"
+
+    # hardcode the ARP response for now
+    arp_response = Ether(dst=dut_mac, src=tb_mac)
+    arp_response /= ARP(op="is-at", hwsrc=tb_mac, hwdst=dut_mac, psrc=tb_ip, pdst=dut_ip)
+    arp_response = arp_response.build()
+
+    await tb.mii_phy.rx.send(GmiiFrame.from_payload(arp_response))
+
+
+    ###############################
+    # Start TCP handshake         #
+    ###############################
+
+    ip_packet = await read_tcp_from_dut()
+
+    t.send(ip_packet)
+
+
+    pkt = get_pkt_from_host()
+    tcp_synack = Ether(dst=dut_mac, src=tb_mac)  / pkt
+
+    await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_synack.build()))
+
+    ip_packet = await read_tcp_from_dut()
+
+    t.send(ip_packet)
+
+    con, addr = serversocket.accept()
+
+    tb.log.info(f"con_timeout: {con.timeout}")
+
+    ###############################
+    # Close connection from DUT   #
+    ###############################
+
+    tb.log.info("Closing connection from the DUT side")
+    await tb.axil_master.write_dword(0x210, 5)
+
+    ip_packet = await read_tcp_from_dut()
+
+    tb.log.info("Sending packet to host")
+    t.send(ip_packet)
+
+    pkt = get_pkt_from_host()
+    tcp_synack = Ether(dst=dut_mac, src=tb_mac)  / pkt
+
+    tb.log.info("Sending reply to DUT, this should be an ACK?")
+    await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_synack.build()))
+
+    tb.log.info(tcp_synack.flags)
+
+    # Host will send an ack first, then a finack?
+
+    tb.log.info("Closing server socket")
+    con.close()
+    serversocket.close()
+
+    pkt = get_pkt_from_host()
+    tcp_synack = Ether(dst=dut_mac, src=tb_mac)  / pkt
+
+    tb.log.info("Sending packet to DUT, this should be a FINACK?")
+    await tb.mii_phy.rx.send(GmiiFrame.from_payload(tcp_synack.build()))
+
+    pkt = get_pkt_from_host()
+    tcp_synack = Ether(dst=dut_mac, src=tb_mac)  / pkt
+
+    ip_packet = await read_tcp_from_dut()
+
+    tb.log.info("Sending packet to host")
+    t.send(ip_packet)
+
+    await Timer(Decimal(CLK_PERIOD_NS * 20000), units='ns')

hw/super6502_fpga/src/sub/network_processor/sources.list

@@ -20,3 +20,5 @@ src/ip_demux_wrapper.sv
 src/tcp_dest_decap.sv
 src/tcp_parser.sv
 src/checksum_calc.sv
+src/ip_pipeline_register_wrapper.sv
+src/axil_reg_slice.sv

hw/super6502_fpga/src/sub/network_processor/src/axil_reg_slice.sv

@@ -0,0 +1,76 @@
+module axil_reg_slice(
+    input clk,
+    input rst,
+
+    axil_intf.SLAVE s_axil,
+    axil_intf.MASTER m_axil
+);
+
+skidbuffer #(
+    .DW(s_axil.AXIL_ADDR_WIDTH)
+) awskid(
+    .i_clk(clk),
+    .i_reset(rst),
+
+    .i_valid(s_axil.awvalid),
+    .o_ready(s_axil.awready),
+    .i_data(s_axil.awaddr),
+    .o_valid(m_axil.awvalid),
+    .i_ready(m_axil.awready),
+    .o_data(m_axil.awaddr)
+);
+
+skidbuffer #(
+    .DW(s_axil.AXIL_DATA_WIDTH + s_axil.AXIL_STRB_WIDTH)
+) wskid(
+    .i_clk(clk),
+    .i_reset(rst),
+
+    .i_valid(s_axil.wvalid),
+    .o_ready(s_axil.wready),
+    .i_data({s_axil.wdata, s_axil.wstrb}),
+    .o_valid(m_axil.wvalid),
+    .i_ready(m_axil.wready),
+    .o_data({m_axil.wdata, m_axil.wstrb})
+);
+
+skidbuffer #(
+    .DW(s_axil.AXIL_ADDR_WIDTH)
+) arskid(
+    .i_clk(clk),
+    .i_reset(rst),
+
+    .i_valid(s_axil.arvalid),
+    .o_ready(s_axil.arready),
+    .i_data(s_axil.araddr),
+    .o_valid(m_axil.arvalid),
+    .i_ready(m_axil.arready),
+    .o_data(m_axil.araddr)
+);
+
+skidbuffer #(
+    .DW(s_axil.AXIL_DATA_WIDTH + 2)
+) rskid(
+    .i_clk(clk),
+    .i_reset(rst),
+
+    .i_valid(m_axil.rvalid),
+    .o_ready(m_axil.rready),
+    .i_data({m_axil.rdata, m_axil.rresp}),
+    .o_valid(s_axil.rvalid),
+    .i_ready(s_axil.rready),
+    .o_data({s_axil.rdata, s_axil.rresp})
+);
+skidbuffer #(.DW(2)) bskid(
+    .i_clk(clk),
+    .i_reset(rst),
+
+    .i_valid(m_axil.bvalid),
+    .o_ready(m_axil.bready),
+    .i_data(m_axil.bresp),
+    .o_valid(s_axil.bvalid),
+    .i_ready(s_axil.bready),
+    .o_data(s_axil.bresp)
+);
+
+endmodule

hw/super6502_fpga/src/sub/network_processor/src/checksum_calc.sv

@@ -11,6 +11,8 @@ module checksum_calc (
 );
 
 logic [31:0] sum;
+logic [31:0] pre_sum;
+logic [31:0] sum_next;
 logic [15:0] sum_wrapped;
 
 assign sum_wrapped = sum[15:0] + sum [31:16];
@@ -21,9 +23,14 @@ always @(posedge i_clk) begin
         sum <= '0;
     end else begin
         if (i_enable) begin
-            sum <= sum + i_data[31:16] + i_data[15:0];
+            sum <= sum_next;
         end
     end
 end
 
+always_comb begin
+    pre_sum = i_data[31:16] + i_data[15:0];
+    sum_next = sum + pre_sum;
+end
+
 endmodule

hw/super6502_fpga/src/sub/network_processor/src/ip_pipeline_register_wrapper.sv

@@ -0,0 +1,59 @@
+module ip_pipeline_register_wrapper(
+    input logic clk,
+    input logic rst,
+
+    ip_intf.SLAVE s_ip,
+    ip_intf.MASTER m_ip
+);
+
+assign m_ip.ip_hdr_valid        = s_ip.ip_hdr_valid;
+assign s_ip.ip_hdr_ready        = m_ip.ip_hdr_ready;
+assign m_ip.eth_dest_mac        = s_ip.eth_dest_mac;
+assign m_ip.eth_src_mac         = s_ip.eth_src_mac;
+assign m_ip.eth_type            = s_ip.eth_type;
+assign m_ip.ip_version          = s_ip.ip_version;
+assign m_ip.ip_ihl              = s_ip.ip_ihl;
+assign m_ip.ip_dscp             = s_ip.ip_dscp;
+assign m_ip.ip_ecn              = s_ip.ip_ecn;
+assign m_ip.ip_length           = s_ip.ip_length;
+assign m_ip.ip_identification   = s_ip.ip_identification;
+assign m_ip.ip_flags            = s_ip.ip_flags;
+assign m_ip.ip_fragment_offset  = s_ip.ip_fragment_offset;
+assign m_ip.ip_ttl              = s_ip.ip_ttl;
+assign m_ip.ip_protocol         = s_ip.ip_protocol;
+assign m_ip.ip_header_checksum  = s_ip.ip_header_checksum;
+assign m_ip.ip_source_ip        = s_ip.ip_source_ip;
+assign m_ip.ip_dest_ip          = s_ip.ip_dest_ip;
+
+axis_pipeline_register #(
+    .DATA_WIDTH(s_ip.DATA_WIDTH),
+    .KEEP_WIDTH(s_ip.KEEP_WIDTH),
+    .ID_ENABLE(1),
+    .ID_WIDTH(s_ip.ID_WIDTH),
+    .DEST_ENABLE(1),
+    .DEST_WIDTH(s_ip.DEST_WIDTH),
+    .USER_WIDTH(s_ip.USER_WIDTH)
+) u_reg (
+    .clk(clk),
+    .rst(rst),
+
+    .s_axis_tdata   (s_ip.ip_payload_axis_tdata),
+    .s_axis_tkeep   (s_ip.ip_payload_axis_tkeep),
+    .s_axis_tvalid  (s_ip.ip_payload_axis_tvalid),
+    .s_axis_tready  (s_ip.ip_payload_axis_tready),
+    .s_axis_tlast   (s_ip.ip_payload_axis_tlast),
+    .s_axis_tid     (s_ip.ip_payload_axis_tid),
+    .s_axis_tdest   (s_ip.ip_payload_axis_tdest),
+    .s_axis_tuser   (s_ip.ip_payload_axis_tuser),
+
+    .m_axis_tdata   (m_ip.ip_payload_axis_tdata),
+    .m_axis_tkeep   (m_ip.ip_payload_axis_tkeep),
+    .m_axis_tvalid  (m_ip.ip_payload_axis_tvalid),
+    .m_axis_tready  (m_ip.ip_payload_axis_tready),
+    .m_axis_tlast   (m_ip.ip_payload_axis_tlast),
+    .m_axis_tid     (m_ip.ip_payload_axis_tid),
+    .m_axis_tdest   (m_ip.ip_payload_axis_tdest),
+    .m_axis_tuser   (m_ip.ip_payload_axis_tuser)
+);
+
+endmodule

hw/super6502_fpga/src/sub/network_processor/src/network_processor.sv

@@ -39,6 +39,8 @@ localparam MAC_DATA_WIDTH = 8;
 localparam AXIS_DATA_WIDTH = 8;
 localparam AXIS_KEEP_WIDTH = ((AXIS_DATA_WIDTH+7)/8);
 
+axil_intf reg_axil_post_reg();
+
 axis_intf #(.DATA_WIDTH(MAC_DATA_WIDTH)) mac_tx_axis();
 axis_intf #(.DATA_WIDTH(MAC_DATA_WIDTH)) mac_rx_axis();
 
@@ -54,29 +56,37 @@ ip_intf #(.DATA_WIDTH(MAC_DATA_WIDTH)) proto_tx_ip[3]();
 ntw_top_regfile_pkg::ntw_top_regfile__in_t hwif_in;
 ntw_top_regfile_pkg::ntw_top_regfile__out_t hwif_out;
 
+axil_reg_slice u_reg_axil_reg_slice(
+    .clk(i_clk),
+    .rst(i_rst),
+
+    .s_axil(s_reg_axil),
+    .m_axil(reg_axil_post_reg)
+);
+
 ntw_top_regfile u_ntw_top_regfile (
     .clk                                (i_clk),
     .rst                                (i_rst),
 
-    .s_axil_awready                     (s_reg_axil.awready),
+    .s_axil_awready                     (reg_axil_post_reg.awready),
-    .s_axil_awvalid                     (s_reg_axil.awvalid),
+    .s_axil_awvalid                     (reg_axil_post_reg.awvalid),
-    .s_axil_awaddr                      (s_reg_axil.awaddr),
+    .s_axil_awaddr                      (reg_axil_post_reg.awaddr),
-    .s_axil_awprot                      (s_reg_axil.awprot),
+    .s_axil_awprot                      (reg_axil_post_reg.awprot),
-    .s_axil_wready                      (s_reg_axil.wready),
+    .s_axil_wready                      (reg_axil_post_reg.wready),
-    .s_axil_wvalid                      (s_reg_axil.wvalid),
+    .s_axil_wvalid                      (reg_axil_post_reg.wvalid),
-    .s_axil_wdata                       (s_reg_axil.wdata),
+    .s_axil_wdata                       (reg_axil_post_reg.wdata),
-    .s_axil_wstrb                       (s_reg_axil.wstrb),
+    .s_axil_wstrb                       (reg_axil_post_reg.wstrb),
-    .s_axil_bready                      (s_reg_axil.bready),
+    .s_axil_bready                      (reg_axil_post_reg.bready),
-    .s_axil_bvalid                      (s_reg_axil.bvalid),
+    .s_axil_bvalid                      (reg_axil_post_reg.bvalid),
-    .s_axil_bresp                       (s_reg_axil.bresp),
+    .s_axil_bresp                       (reg_axil_post_reg.bresp),
-    .s_axil_arready                     (s_reg_axil.arready),
+    .s_axil_arready                     (reg_axil_post_reg.arready),
-    .s_axil_arvalid                     (s_reg_axil.arvalid),
+    .s_axil_arvalid                     (reg_axil_post_reg.arvalid),
-    .s_axil_araddr                      (s_reg_axil.araddr),
+    .s_axil_araddr                      (reg_axil_post_reg.araddr),
-    .s_axil_arprot                      (s_reg_axil.arprot),
+    .s_axil_arprot                      (reg_axil_post_reg.arprot),
-    .s_axil_rready                      (s_reg_axil.rready),
+    .s_axil_rready                      (reg_axil_post_reg.rready),
-    .s_axil_rvalid                      (s_reg_axil.rvalid),
+    .s_axil_rvalid                      (reg_axil_post_reg.rvalid),
-    .s_axil_rdata                       (s_reg_axil.rdata),
+    .s_axil_rdata                       (reg_axil_post_reg.rdata),
-    .s_axil_rresp                       (s_reg_axil.rresp),
+    .s_axil_rresp                       (reg_axil_post_reg.rresp),
 
     .hwif_in                            (hwif_in),
     .hwif_out                           (hwif_out)

hw/super6502_fpga/src/sub/network_processor/src/tcp.sv

@@ -205,6 +205,8 @@ logic [$clog2(NUM_TCP)-1:0] tcp_demux_sel;
 logic [15:0] tcp_dests [NUM_TCP];
 
 always_comb begin : TCP_DEST_SEL
+    tcp_demux_sel = '0;
+
     for (int i = 0; i < NUM_TCP; i++) begin
         if (tcp_dest == tcp_dests[i]) begin
             tcp_demux_sel = i;

hw/super6502_fpga/src/sub/network_processor/src/tcp_dest_decap.sv

@@ -100,6 +100,15 @@ always_comb begin
 
     s_ip.ip_payload_axis_tready = '0;
 
+    valid = '0;
+
+    m_ip.ip_payload_axis_tdata = '0;
+    m_ip.ip_payload_axis_tvalid = '0;
+    m_ip.ip_payload_axis_tlast = '0;
+    m_ip.ip_payload_axis_tuser = '0;
+    m_ip.ip_payload_axis_tid = '0;
+    m_ip.ip_payload_axis_tdest = '0;
+
     case (state)
         PORTS: begin
             s_ip.ip_payload_axis_tready = 1;

hw/super6502_fpga/src/sub/network_processor/src/tcp_packet_generator.sv

@@ -37,7 +37,7 @@ assign pre_checksum_data.tuser = s_axis_data.tuser;
 
 axis_saf_fifo #(
     .DATA_DEPTH_L2(11),
-    .CTRL_DEPTH_L2(1)
+    .CTRL_DEPTH_L2(2)
 ) u_checksum_fifo (
     .sclk(i_clk),
     .srst(i_rst),
@@ -67,6 +67,9 @@ logic data_checksum_clear;
 logic [31:0] data_checksum_data;
 logic [15:0] data_checksum_final;
 
+logic [31:0] src_ip, dst_ip_next;
+logic [31:0] dst_ip, src_ip_next;
+
 checksum_calc u_header_checksum_calc(
     .i_rst      (i_rst),
     .i_clk      (i_clk),
@@ -83,18 +86,44 @@ always_ff @(posedge i_clk) begin
         checksum_counter <= '0;
         state <= IDLE;
         data_expand <= '0;
+        src_ip <= '0;
+        dst_ip <= '0;
     end else begin
         counter <= counter_next;
         checksum_counter <= checksum_counter_next;
         state <= state_next;
         data_expand <= data_expand_next;
+        src_ip <= src_ip_next;
+        dst_ip <= dst_ip_next;
     end
 end
 
 always_comb begin
+    state_next = state;
+
     m_ip.ip_hdr_valid = '0;
+    m_ip.ip_dscp        = '0;
+    m_ip.ip_ecn         = '0;
+    m_ip.ip_length      = '0;
+    m_ip.ip_ttl         = '0;
+    m_ip.ip_protocol    = '0;
+    m_ip.ip_source_ip   = '0;
+    m_ip.ip_dest_ip     = '0;
+
+    m_ip.ip_payload_axis_tdata  = '0;
     m_ip.ip_payload_axis_tvalid = '0;
     m_ip.ip_payload_axis_tlast  = '0;
+    m_ip.ip_payload_axis_tuser  = '0;
+    m_ip.ip_payload_axis_tid    = '0;
+    m_ip.ip_payload_axis_tdest  = '0;
+
+    post_checksum_data.tready = '0;
+
+    checksum_counter_next = checksum_counter;
+    checksum_data = '0;
+
+    counter_next = counter;
+
     o_packet_done = '0;
     checksum_clear = '0;
     checksum_enable = '0;
@@ -103,6 +132,8 @@ always_comb begin
 
     data_expand_next = data_expand;
 
+    src_ip_next = src_ip;
+    dst_ip_next = dst_ip;
     case (state)
 
         IDLE: begin
@@ -120,6 +151,9 @@ always_comb begin
                 m_ip.ip_source_ip   = i_src_ip;
                 m_ip.ip_dest_ip     = i_dst_ip;
 
+                src_ip_next = i_src_ip;
+                dst_ip_next = i_dst_ip;
+
                 if (m_ip.ip_hdr_ready) begin
                     if (i_no_data) begin
                         state_next = HEADER;
@@ -159,9 +193,9 @@ always_comb begin
             end
 
             case (checksum_counter)
-                0: checksum_data = m_ip.ip_source_ip;
+                0: checksum_data = src_ip;
-                1: checksum_data = m_ip.ip_dest_ip;
+                1: checksum_data = dst_ip;
-                2: checksum_data = {8'b0, m_ip.ip_protocol, (i_ip_len - 16'd20)}; // tcp length, not IP length
+                2: checksum_data = {8'b0, 8'h6, (i_ip_len - 16'd20)}; // tcp length, not IP length
                 3: checksum_data = {i_source_port, i_dest_port};
                 4: checksum_data = i_seq_number;
                 5: checksum_data = i_ack_number;

hw/super6502_fpga/src/sub/network_processor/src/tcp_parser.sv

@@ -66,6 +66,13 @@ always_comb begin
     checksum_next = checksum;
     hdr_valid = '0;
 
+    counter_next = counter;
+
+    state_next = state;
+
+    s_ip.ip_hdr_ready = '0;
+    s_ip.ip_payload_axis_tready = '0;
+
     case (state)
         HEADER: begin
             s_ip.ip_hdr_ready = '1;
@@ -103,6 +110,7 @@ always_comb begin
         end
 
         PAYLOAD: begin
+            s_ip.ip_payload_axis_tready = '1;
             if (s_ip.ip_payload_axis_tlast) begin
                 counter_next = '0;
                 state_next = HEADER;

hw/super6502_fpga/src/sub/network_processor/src/tcp_rx_ctrl.sv

@@ -19,34 +19,47 @@ module tcp_rx_ctrl (
     output logic [31:0]       o_ack_number
 );
 
+tcp_pkg::rx_msg_t rx_msg_next;
+logic rx_msg_valid_next;
+
 logic [31:0] ack_num, ack_num_next;
 assign o_ack_number = ack_num;
 
 always_ff @(posedge i_clk) begin
     if (i_rst) begin
         ack_num <= '0;
+        o_rx_msg <= RX_MSG_NOP;
+        o_rx_msg_valid <= '0;
     end else begin
         ack_num <= ack_num_next;
+        o_rx_msg <= rx_msg_next;
+        o_rx_msg_valid <= rx_msg_valid_next;
     end
 end
 
 always_comb begin
+    rx_msg_next = RX_MSG_NOP;
+    rx_msg_valid_next = '0;
+
+    ack_num_next = ack_num;
+
     if (i_hdr_valid) begin
         if (i_flags == 8'h12) begin
-            o_rx_msg = RX_MSG_RECV_SYNACK;
+            rx_msg_next = RX_MSG_RECV_SYNACK;
-            o_rx_msg_valid = '1;
+            rx_msg_valid_next = '1;
 
             ack_num_next = i_seq_number + 1;
         end
 
         if (i_flags == 8'h11) begin
-            o_rx_msg = RX_MSG_RECV_FIN;
+            rx_msg_next = RX_MSG_RECV_FIN;
-            o_rx_msg_valid = '1;
+            rx_msg_valid_next = '1;
+            ack_num_next = i_seq_number + 1;
         end
 
         if (i_flags == 8'h10) begin
-            o_rx_msg = RX_MSG_RECV_ACK;
+            rx_msg_next = RX_MSG_RECV_ACK;
-            o_rx_msg_valid = '1;
+            rx_msg_valid_next = '1;
         end
 
     end

hw/super6502_fpga/src/sub/network_processor/src/tcp_state_manager.sv

@@ -87,8 +87,46 @@ always_comb begin
 
         ESTABLISHED: begin
             if (i_rx_msg_valid && i_rx_msg == RX_MSG_RECV_FIN) begin
+                o_tx_ctrl = TX_CTRL_SEND_ACK;
+                o_tx_ctrl_valid = '1;
+                tcp_state_next = WAIT_CLOSE;
+            end
+
+            if (i_close) begin
                 o_tx_ctrl = TX_CTRL_SEND_FIN;
                 o_tx_ctrl_valid = '1;
+                tcp_state_next = FIN_WAIT_1;
+            end
+        end
+
+        FIN_WAIT_1: begin
+            if (i_rx_msg_valid) begin
+                if (i_rx_msg == RX_MSG_RECV_ACK) begin
+                    tcp_state_next = FIN_WAIT_2;
+                end else if (i_rx_msg == RX_MSG_RECV_FIN) begin
+                    tcp_state_next = TIME_WAIT;
+                    o_tx_ctrl_valid = '1;
+                    o_tx_ctrl = TX_CTRL_SEND_ACK;
+                end
+            end
+        end
+
+        FIN_WAIT_2: begin
+            if (i_rx_msg == RX_MSG_RECV_FIN) begin
+                tcp_state_next = TIME_WAIT;
+                o_tx_ctrl = TX_CTRL_SEND_ACK;
+                o_tx_ctrl_valid = '1;
+            end
+        end
+
+        TIME_WAIT: begin
+            tcp_state_next = IDLE;
+        end
+
+        WAIT_CLOSE: begin
+            o_tx_ctrl = TX_CTRL_SEND_FIN;
+            o_tx_ctrl_valid = '1;
+            if (i_tx_ctrl_ack) begin        // we should be doing this other places too...
                 tcp_state_next = LAST_ACK;
             end
         end

hw/super6502_fpga/src/sub/network_processor/src/tcp_stream.sv

@@ -30,8 +30,11 @@ module tcp_stream #(
 );
 
 axis_intf m2s_axis();
+axis_intf m2s_axis_pre_reg();
 axis_intf s2m_axis();
 
+ip_intf m_ip_tx_pre_reg();
+
 axis_intf m2s_post_saf_axis();
 axis_intf s2m_pre_saf_axis();
 
@@ -111,6 +114,14 @@ m2s_dma #(
     .s_cpuif_wr_err             (),
 
     .m_axil                     (m_m2s_axil),
+    .m_axis                     (m2s_axis_pre_reg)
+);
+
+axis_pipeline_register_wrapper u_m2s_reg (
+    .clk(clk),
+    .rst(rst),
+
+    .s_axis(m2s_axis_pre_reg),
     .m_axis(m2s_axis)
 );
 
@@ -204,6 +215,14 @@ tcp_packet_generator u_tcp_packet_generator (
 
     .o_packet_done              (w_tx_packet_done),
 
+    .m_ip                       (m_ip_tx_pre_reg)
+);
+
+ip_pipeline_register_wrapper u_tx_ip_reg (
+    .clk(clk),
+    .rst(rst),
+
+    .s_ip(m_ip_tx_pre_reg),
     .m_ip(m_ip_tx)
 );
 

hw/super6502_fpga/src/sub/network_processor/src/tcp_tx_ctrl.sv

@@ -24,14 +24,13 @@ module tcp_tx_ctrl(
     input  wire                 i_packet_done
 );
 
-assign m_axis.tdata     = s_axis.tdata;
+axis_pipeline_register_wrapper u_m2s_reg (
-assign m_axis.tkeep     = s_axis.tkeep;
+    .clk(i_clk),
-assign m_axis.tvalid    = s_axis.tvalid;
+    .rst(i_rst),
-assign s_axis.tready    = m_axis.tready;
+
-assign m_axis.tlast     = s_axis.tlast;
+    .s_axis(s_axis),
-assign m_axis.tid       = s_axis.tid;
+    .m_axis(m_axis)
-assign m_axis.tdest     = s_axis.tdest;
+);
-assign m_axis.tuser     = s_axis.tuser;
 
 localparam FLAG_FIN = (1 << 0);
 localparam FLAG_SYN = (1 << 1);
@@ -61,6 +60,8 @@ always_comb begin
     state_next = state;
     o_no_data = '0;
 
+    o_tx_ctrl_ack = '0;
+
     o_ack_number    = '0;
     o_flags         = '0;
     o_window_size   = 16'h100;
@@ -129,7 +130,7 @@ always_comb begin
 
             if (i_packet_done) begin
                 state_next = IDLE;
-                seq_num_next = seq_num + s_axis_len;
+                seq_num_next = seq_num + 1;
             end
         end
     endcase

hw/super6502_fpga/super6502_fpga.xml

@@ -1,4 +1,5 @@
-<efx:project xmlns:efx="http://www.efinixinc.com/enf_proj" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" name="super6502_fpga" description="" last_change_date="Mon Sep 09 2024 11:15:46 PM" location="/cluster/projects/super6502/hw/super6502_fpga" sw_version="2023.1.150" last_run_state="pass" last_run_tool="efx_pgm" last_run_flow="bitstream" config_result_in_sync="sync" design_ood="sync" place_ood="sync" route_ood="sync" xsi:schemaLocation="http://www.efinixinc.com/enf_proj enf_proj.xsd">
+<?xml version="1.0" encoding="UTF-8"?>
+<efx:project name="super6502_fpga" description="" last_change="1728874681" sw_version="2024.1.163" last_run_state="pass" last_run_flow="bitstream" config_result_in_sync="sync" design_ood="sync" place_ood="sync" route_ood="sync" xmlns:efx="http://www.efinixinc.com/enf_proj" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.efinixinc.com/enf_proj enf_proj.xsd">
     <efx:device_info>
         <efx:family name="Trion"/>
         <efx:device name="T20F256"/>
@@ -209,6 +210,9 @@
         <efx:design_file name="src/sub/my-fifos/src/fifo_fwft.sv" version="default" library="default"/>
         <efx:design_file name="src/sub/my-fifos/src/fifo.sv" version="default" library="default"/>
         <efx:design_file name="src/sub/my-fifos/src/fwft_adapter.sv" version="default" library="default"/>
+        <efx:design_file name="src/sub/network_processor/src/ip_pipeline_register_wrapper.sv" version="default" library="default"/>
+        <efx:design_file name="src/sub/verilog-ethernet/lib/axis/rtl/axis_pipeline_register_wrapper.sv" version="default" library="default"/>
+        <efx:design_file name="src/sub/network_processor/src/axil_reg_slice.sv" version="default" library="default"/>
         <efx:top_vhdl_arch name=""/>
     </efx:design_info>
     <efx:constraint_info>
@@ -249,6 +253,10 @@
         <efx:param name="mult_input_regs_packing" value="1" value_type="e_option"/>
         <efx:param name="mult_output_regs_packing" value="1" value_type="e_option"/>
         <efx:param name="include" value="ip/sdram_controller" value_type="e_string"/>
+        <efx:param name="bram-push-tco-outreg" value="0" value_type="e_option"/>
+        <efx:param name="mult-auto-pipeline" value="0" value_type="e_integer"/>
+        <efx:param name="use-logic-for-small-mem" value="64" value_type="e_integer"/>
+        <efx:param name="use-logic-for-small-rom" value="64" value_type="e_integer"/>
         <efx:defmacro name="SDIO_AXI" value="1"/>
         <efx:defmacro name="EFINIX" value="1"/>
     </efx:synthesis>
@@ -260,6 +268,7 @@
         <efx:param name="seed" value="1" value_type="e_integer"/>
         <efx:param name="placer_effort_level" value="5" value_type="e_option"/>
         <efx:param name="max_threads" value="-1" value_type="e_integer"/>
+        <efx:param name="print_critical_path" value="10" value_type="e_integer"/>
     </efx:place_and_route>
     <efx:bitstream_generation tool_name="efx_pgm">
         <efx:param name="mode" value="active" value_type="e_option"/>

sw/Makefile

@@ -0,0 +1,19 @@
+SD_IMAGE = sd_image.bin
+BOOTLOADER = bootloader/bootloader.bin
+
+all: $(SD_IMAGE) $(BOOTLOADER)
+
+$(BOOTLOADER):
+	$(MAKE) -C bootloader bootloader.bin
+
+$(SD_IMAGE):
+	dd if=/dev/zero of=$@ bs=1M count=256
+	/sbin/mkfs.vfat $@
+	udisksctl loop-setup --file $@
+	udisksctl mount -b /dev/loop1
+	udisksctl unmount -b /dev/loop1
+	udisksctl loop-delete -b /dev/loop1
+
+clean:
+	rm -rf $(SD_IMAGE)
+	rm -rf $(BOOTLOADER)

sw/bootloader/Makefile

@@ -0,0 +1,38 @@
+CC=../toolchain/cc65/bin/cl65
+LD=../toolchain/cc65/bin/cl65
+CFLAGS=-T -t none -I. --cpu "65C02"
+LDFLAGS=-C link.ld -m $(NAME).map
+
+NAME=bootloader
+
+BIN=$(NAME).bin
+HEX=$(NAME).hex
+
+LISTS=lists
+
+SRCS=$(wildcard *.s) $(wildcard *.c)
+SRCS+=$(wildcard **/*.s) $(wildcard **/*.c)
+OBJS+=$(patsubst %.s,%.o,$(filter %s,$(SRCS)))
+OBJS+=$(patsubst %.c,%.o,$(filter %c,$(SRCS)))
+
+all: $(HEX)
+
+$(HEX): $(BIN)
+	objcopy --input-target=binary --output-target=verilog $(BIN) $(HEX)
+
+$(BIN): $(OBJS)
+	$(CC) $(CFLAGS) $(LDFLAGS) $(OBJS) -o $@
+
+%.o: %.c $(LISTS)
+	$(CC) $(CFLAGS) -l $(LISTS)/$<.list -c $< -o $@
+
+%.o: %.s $(LISTS)
+	$(CC) $(CFLAGS) -l $(LISTS)/$<.list -c $< -o $@
+
+$(LISTS):
+	mkdir -p $(addprefix $(LISTS)/,$(sort $(dir $(SRCS))))
+
+.PHONY: clean
+clean:
+	rm -rf $(OBJS) $(BIN) $(HEX) $(LISTS) $(NAME).map
+

sw/bootloader/bootloader.s

@@ -0,0 +1,25 @@
+.export _init, _nmi_int, _irq_int
+
+.segment  "VECTORS"
+
+.addr      _nmi_int    ; NMI vector
+.addr      _init       ; Reset vector
+.addr      _irq_int    ; IRQ/BRK vector
+
+SDRAM= $200
+
+.code
+
+_nmi_int:
+_irq_int:
+
+_init:
+        lda #$00
+@start:
+        sta SDRAM
+        cmp SDRAM
+        bne @end
+        ina
+        bra @start
+
+@end:   bra @end

sw/bootloader/link.ld

@@ -0,0 +1,30 @@
+MEMORY
+{
+    RAM: start = $0000, size = $200;
+    ROM: start = $F000, size = $1000, file = %O;
+}
+
+SEGMENTS {
+    ZEROPAGE: load = RAM, type = zp,  define   = yes;
+    DATA:     load = ROM, type = rw,  define   = yes;
+    CODE:     load = ROM, type = ro;
+    RODATA:   load = ROM, type = ro;
+    VECTORS:  load = ROM, type = ro,  start    = $FFFA;
+}
+
+FEATURES {
+    CONDES:    segment = STARTUP,
+               type    = constructor,
+               label   = __CONSTRUCTOR_TABLE__,
+               count   = __CONSTRUCTOR_COUNT__;
+    CONDES:    segment = STARTUP,
+               type    = destructor,
+               label   = __DESTRUCTOR_TABLE__,
+               count   = __DESTRUCTOR_COUNT__;
+}
+
+SYMBOLS {
+    # Define the stack size for the application
+    __STACKSIZE__:  value = $0200, type = weak;
+    __STACKSTART__: type = weak, value = $0800; # 2k stack
+}

sw/kernel/Makefile

@@ -0,0 +1,39 @@
+CC=../toolchain/cc65/bin/cl65
+LD=../toolchain/cc65/bin/cl65
+CFLAGS=-T -t none -I. --cpu "65C02"
+LDFLAGS=-C link.ld -m $(NAME).map
+
+NAME=bootloader
+
+BIN=$(NAME).bin
+HEX=$(NAME).hex
+
+LISTS=lists
+
+SRCS=$(wildcard *.s) $(wildcard *.c)
+SRCS+=$(wildcard **/*.s) $(wildcard **/*.c)
+OBJS+=$(patsubst %.s,%.o,$(filter %s,$(SRCS)))
+OBJS+=$(patsubst %.c,%.o,$(filter %c,$(SRCS)))
+
+# Make sure the kernel linked to correct address, no relocation!
+all: $(HEX)
+
+$(HEX): $(BIN)
+	objcopy --input-target=binary --output-target=verilog $(BIN) $(HEX)
+
+$(BIN): $(OBJS)
+	$(CC) $(CFLAGS) $(LDFLAGS) $(OBJS) -o $@
+
+%.o: %.c $(LISTS)
+	$(CC) $(CFLAGS) -l $(LISTS)/$<.list -c $< -o $@
+
+%.o: %.s $(LISTS)
+	$(CC) $(CFLAGS) -l $(LISTS)/$<.list -c $< -o $@
+
+$(LISTS):
+	mkdir -p $(addprefix $(LISTS)/,$(sort $(dir $(SRCS))))
+
+.PHONY: clean
+clean:
+	rm -rf $(OBJS) $(BIN) $(HEX) $(LISTS) $(NAME).map
+

sw/kernel/kernel.s

@@ -0,0 +1,9 @@
+.segment "ENTRY"
+
+.addr main
+
+
+.code
+
+main:
+    bra main

sw/kernel/link.ld

@@ -0,0 +1,30 @@
+MEMORY
+{
+    RAM: start = $0000, size = $200;
+    KERNEL: start = $8000, size = $7000, file = %O;
+}
+
+SEGMENTS {
+    ZEROPAGE: load = RAM, type = zp,  define   = yes;
+    ENTRY:    load = KERNEL, type = ro, start = $8000;
+    DATA:     load = KERNEL, type = rw,  define   = yes;
+    CODE:     load = KERNEL, type = ro;
+    RODATA:   load = KERNEL, type = ro;
+}
+
+FEATURES {
+    CONDES:    segment = STARTUP,
+               type    = constructor,
+               label   = __CONSTRUCTOR_TABLE__,
+               count   = __CONSTRUCTOR_COUNT__;
+    CONDES:    segment = STARTUP,
+               type    = destructor,
+               label   = __DESTRUCTOR_TABLE__,
+               count   = __DESTRUCTOR_COUNT__;
+}
+
+SYMBOLS {
+    # Define the stack size for the application
+    __STACKSIZE__:  value = $0200, type = weak;
+    __STACKSTART__: type = weak, value = $0800; # 2k stack
+}