Synthesisable version with passed oob

.settings/com.elphel.vdt.FPGA_project.prefs

 FPGA_project_0_SimulationTopFile=tb/tb_top.v
 FPGA_project_1_SimulationTopModule=tb
 FPGA_project_2_ImplementationTopFile=dma/top.v
+FPGA_project_3_ImplementationTopModule=top
 FPGA_project_4_part=xc7z030fbg484-1
-com.elphel.store.context.FPGA_project=FPGA_project_2_ImplementationTopFile<-@\#\#@->FPGA_project_4_part<-@\#\#@->FPGA_project_0_SimulationTopFile<-@\#\#@->FPGA_project_1_SimulationTopModule<-@\#\#@->
+FPGA_project_5_part=xc7z030fbg484-1
+com.elphel.store.context.FPGA_project=FPGA_project_2_ImplementationTopFile<-@\#\#@->FPGA_project_4_part<-@\#\#@->FPGA_project_0_SimulationTopFile<-@\#\#@->FPGA_project_1_SimulationTopModule<-@\#\#@->FPGA_project_3_ImplementationTopModule<-@\#\#@->FPGA_project_5_part<-@\#\#@->
 eclipse.preferences.version=1

.settings/com.elphel.vdt.VivadoBitstream.prefs

 VivadoBitstream_103_PreBitstreamTCL=set_property BITSTREAM.STARTUP.MATCH_CYCLE NoWait [current_design]<-@\#\#@->
 VivadoBitstream_105_force=true
-com.elphel.store.context.VivadoBitstream=VivadoBitstream_105_force<-@\#\#@->VivadoBitstream_103_PreBitstreamTCL<-@\#\#@->
+VivadoBitstream_125_force=true
+com.elphel.store.context.VivadoBitstream=VivadoBitstream_105_force<-@\#\#@->VivadoBitstream_103_PreBitstreamTCL<-@\#\#@->VivadoBitstream_125_force<-@\#\#@->
 eclipse.preferences.version=1

.settings/com.elphel.vdt.VivadoSynthesis.prefs

 VivadoSynthesis_101_MaxMsg=10000
 VivadoSynthesis_102_ConstraintsFiles=x393.xdc<-@\#\#@->x393_timing.xdc<-@\#\#@->
 VivadoSynthesis_115_flatten_hierarchy=none
+VivadoSynthesis_122_ConstraintsFiles=top_timing.xdc<-@\#\#@->top.xdc<-@\#\#@->
 VivadoSynthesis_127_verbose=true
 VivadoSynthesis_81_parser_mode=1
 VivadoSynthesis_93_OtherProblems=Netlist 29-345<-@\#\#@->Board 49-26<-@\#\#@->
 VivadoSynthesis_95_ShowInfo=false
-com.elphel.store.context.VivadoSynthesis=VivadoSynthesis_102_ConstraintsFiles<-@\#\#@->VivadoSynthesis_95_ShowInfo<-@\#\#@->VivadoSynthesis_115_flatten_hierarchy<-@\#\#@->VivadoSynthesis_101_MaxMsg<-@\#\#@->VivadoSynthesis_127_verbose<-@\#\#@->VivadoSynthesis_93_OtherProblems<-@\#\#@->VivadoSynthesis_81_parser_mode<-@\#\#@->
+com.elphel.store.context.VivadoSynthesis=VivadoSynthesis_102_ConstraintsFiles<-@\#\#@->VivadoSynthesis_95_ShowInfo<-@\#\#@->VivadoSynthesis_115_flatten_hierarchy<-@\#\#@->VivadoSynthesis_101_MaxMsg<-@\#\#@->VivadoSynthesis_127_verbose<-@\#\#@->VivadoSynthesis_93_OtherProblems<-@\#\#@->VivadoSynthesis_81_parser_mode<-@\#\#@->VivadoSynthesis_122_ConstraintsFiles<-@\#\#@->
 eclipse.preferences.version=1

dma/dma_regs.v

@@ -166,7 +166,7 @@ begin
     reg04 <= rst ? 32'h0 : bram_wen & (bram_waddr[7:0] == 8'hf1) ? wdata : reg04;
     reg08 <= rst ? 32'h0 : bram_wen & (bram_waddr[7:0] == 8'hf2) ? wdata : reg08;
     reg0c <= rst ? 32'h0 : bram_wen & (bram_waddr[7:0] == 8'hf3) ? wdata : reg0c;
-    reg10 <= rst ? 32'h0 : dma_start_aclk ? 32'h0 : dma_done_aclk ? 32'hffffffff : reg10; // status reg
+    reg10 <= rst ? 32'h0 : dma_start_aclk ? 32'h0 : {31'h0, dma_done_aclk} ? 32'hffffffff : reg10; // status reg
     reg14 <= rst ? 32'h0 : dma_done_aclk ? reg00 : reg14;
 end
 

dma/sata_top.v

@@ -32,6 +32,12 @@
     output  wire    sclk,
     output  wire    sata_rst,
     input   wire    extrst,
+    
+    // reliable clock to source drp and cpll lock det circuits
+    input   wire    reliable_clk,
+    
+    input   wire    hclk,
+    
 /*
  * Commands interface
  */
@@ -228,7 +234,7 @@ wire            dma_type;
 wire            dma_start;
 wire            dma_done;
 // axi-hp clock
-wire            hclk;
+//wire            hclk;
 // dma_control <-> dma_adapter command iface
 wire            adp_busy;
 wire    [31:7]  adp_addr;
@@ -279,10 +285,9 @@ wire    [63:0]  buf_rdata;
 wire            rdata_done; // = membridge.is_last_in_page & membridge.afi_rready;
 
 //assign  rst = ARESETN;
-
-// TODO
-assign hclk = ACLK;
-
+reg hrst;
+always @ (posedge hclk)
+    hrst <= sata_rst;
 
 axi_regs axi_regs(
 // axi iface
@@ -333,7 +338,7 @@ axi_regs axi_regs(
  * Programmable sata controller registers
  */
 dma_regs dma_regs(
-    .rst            (sata_rst),
+    .rst            (ARESETN),
     .ACLK           (ACLK),
     .sclk           (sclk),
 // control iface
@@ -473,7 +478,7 @@ dma_control dma_control(
 
 dma_adapter dma_adapter(
     .clk                    (hclk),
-    .rst                    (sata_rst),
+    .rst                    (hrst),
 // command iface                            
     .cmd_type               (adp_type),
     .cmd_val                (adp_val),
@@ -526,9 +531,9 @@ V    .MEMBRIDGE_ADDR         (),
     .FRAME_HEIGHT_BITS      (),
     .FRAME_WIDTH_BITS       ()
 )*/ membridge(
-    .mrst                   (sata_rst), // input
-    .hrst                   (ARESETN), // input
-    .mclk                   (hclk), // input
+    .mrst                   (hrst), // input
+    .hrst                   (hrst), // input
+    .mclk                   (sclk), // input
     .hclk                   (hclk), // input
     .cmd_ad                 (cmd_ad),
     .cmd_stb                (cmd_stb),
@@ -605,6 +610,8 @@ sata_host sata_host(
     .rst                        (sata_rst),
     // sata clk
     .clk                        (sclk),
+    // reliable clock to source drp and cpll lock det circuits
+    .reliable_clk               (reliable_clk),
 // temporary
     .al_cmd_in                  (cmd_out), // == {cmd_type, cmd_port, cmd_val, cmd_done_bad, cmd_done_good; cmd_busy}
     .al_cmd_val_in              (cmd_val_out),

dma/top.v

@@ -139,8 +139,8 @@ BUFG bufg_axi_aclk0_i  (.O(axi_aclk0),.I(fclk[0]));
 BUFG bufg_axi_rst_i   (.O(axi_rst),.I(axi_rst_pre));
 BUFG bufg_extrst_i    (.O(extrst),.I(axi_rst_pre));
 axi_hp_clk #(
-    .CLKIN_PERIOD(6.666),
-    .CLKFBOUT_MULT_AXIHP(6),
+    .CLKIN_PERIOD(20.000),
+    .CLKFBOUT_MULT_AXIHP(18),
     .CLKFBOUT_DIV_AXIHP(6)
 ) axi_hp_clk_i (
     .rst          (axi_rst), // input
@@ -151,10 +151,13 @@ axi_hp_clk #(
 
 sata_top sata_top(
     .sclk                       (sclk),
+    // reliable clock to source drp and cpll lock det circuits
+    .reliable_clk               (axi_aclk0),
+    .hclk                       (hclk),
     .sata_rst                   (sata_rst),
     .extrst                     (extrst),
-    .ACLK                       (axi_aclk),
-    .ARESETN                    (axi_rst | sata_rst),
+    .ACLK                       (axi_aclk0),
+    .ARESETN                    (axi_rst/* | sata_rst*/),
 // AXI PS Master GP1: Read Address    
     .ARADDR                     (ARADDR),
     .ARVALID                    (ARVALID),
@@ -569,7 +572,7 @@ PS7 ps7_i (
 
 // AXI PS Master GP1    
 // AXI PS Master GP1: Clock, Reset
-    .MAXIGP1ACLK    (axi_aclk),         // AXI PS Master GP1 Clock , input
+    .MAXIGP1ACLK    (axi_aclk0),         // AXI PS Master GP1 Clock , input
     .MAXIGP1ARESETN (),          // AXI PS Master GP1 Reset, output
 // AXI PS Master GP1: Read Address    
     .MAXIGP1ARADDR  (ARADDR),           // AXI PS Master GP1 ARADDR[31:0], output  

host/command.v

@@ -25,6 +25,11 @@ module command(
     input   rst,
     input   clk,
 
+    // temporary TODO
+    input   wire    gtx_ready,
+    input   wire    phy_ready,
+    input   wire    [11:0]  debug_cnt,
+
     // tl cmd iface
     output  wire    [2:0]   cmd_type,
     output  wire            cmd_val,
@@ -217,10 +222,13 @@ assign  sh_autoact_out  = sh_autoact;
 
 // temporaty command register TODO
 reg [31:0]  cmd;
-assign  al_cmd_out = cmd;
+assign  al_cmd_out[31:12] = cmd[31:12];
+assign  al_cmd_out[11:0]  = debug_cnt;
 always @ (posedge clk)
 begin
-    cmd[31:4]   <= rst ? 28'h0 : al_cmd_val_in ? al_cmd_in[31:4] : cmd[31:4];
+    cmd[27:4]   <= rst ? 24'h0 : al_cmd_val_in ? al_cmd_in[27:4] : cmd[27:4];
+    cmd[31]     <= rst ? 1'b1 : cmd[31];
+    cmd[30:28]  <= rst ? 3'h0 : {1'b0, phy_ready, gtx_ready};
     cmd[3]      <= rst ? 1'b0 : al_cmd_val_in ? al_cmd_in[3] : cmd_val ? 1'b0 : cmd[3];
     cmd[2]      <= rst ? 1'b0 : al_cmd_val_in ? 1'b0 : cmd_done_bad ? 1'b1 : cmd[2];
     cmd[1]      <= rst ? 1'b0 : al_cmd_val_in ? 1'b0 : cmd_done_good ? 1'b1 : cmd[1];

host/gtx_elastic.v

@@ -64,7 +64,7 @@ localparam HI = DEPTH_LOG2 - 1; // hi bus index
 reg     [22:0]  ram [(1 << DEPTH_LOG2) - 1:0];
 // data to/from fifo
 wire    [22:0]  inram;
-wire    [22:0]  outram;
+reg     [22:0]  outram;
 // adresses in their natural clock domains
 reg     [HI:0]  rd_addr;
 reg     [HI:0]  wr_addr;
@@ -142,7 +142,8 @@ endgenerate
 assign  full   = wr_addr   == rd_addr_r + 1'b1;
 assign  empty  = wr_addr_r == rd_addr;
 
-assign  outram = ram[rd_addr];
+always @ (posedge rclk)
+    outram <= ram[rd_addr];
 
 always @ (posedge wclk)
     if (we)
@@ -283,7 +284,7 @@ wire    clr_skip2_align;
 wire    clr_wait_next_p;
 wire    clr_send_ack;
 
-always @ (wclk)
+always @ (posedge wclk)
     next_prim_loaded <= state_wait_next_p;
 
 assign  state_bypass_rmv = ~state_wait1_align & ~state_skip1_align & ~state_wait2_align & ~state_skip2_align & ~state_wait_next_p & ~state_send_ack;
@@ -388,11 +389,15 @@ always @ (posedge rclk)
 
 // choose 1 of 2 words of ALIGNP
 wire    [22:0]  align_word;
-assign  align_word = {outram[22], 22'h007B4A} & {23{~align_altern}} | {outram[22], 22'h014ABC} & {23{align_altern}};
+assign  align_word = {outram[22], 22'h007B4A} & {23{align_altern}} | {outram[22], 22'h014ABC} & {23{~align_altern}};
 
+// output data would be valid the next clock they are issued
+reg     pause_read_r;
+always @ (posedge rclk)
+    pause_read_r <= pause_read;
 // read when compensation is not issued and when fifo gets required fullfillment
 assign  re = ~pause_read & fifo_stable;
-assign  outdata = {23{~pause_read}} & outram | {23{pause_read}} & align_word;
+assign  outdata = {23{~pause_read_r}} & outram | {23{pause_read_r}} & align_word;
 // indicates last cycle before the next primitive
 wire    fword_strobe_correction;
 reg     fword_strobe;

host/gtx_wrap.v

@@ -33,6 +33,7 @@ module gtx_wrap #(
     parameter RXISCANRESET_TIME   = 5'h1
 )
 (
+    output  reg     debug = 0,
     output  wire    cplllock,
     input   wire    cplllockdetclk,
     input   wire    cpllreset,
@@ -75,11 +76,15 @@ module gtx_wrap #(
 
 wire    rxresetdone_gtx; 
 wire    txresetdone_gtx;
-wire    wrap_rxreset_;
-wire    wrap_txreset_;
+reg     rxresetdone_gtx_r; 
+reg     txresetdone_gtx_r;
+reg     wrap_rxreset_;
+reg     wrap_txreset_;
 // resets while PCS resets, active low
-assign  wrap_rxreset_ = rxuserrdy & rxresetdone_gtx;
-assign  wrap_txreset_ = txuserrdy & txresetdone_gtx;
+always @ (posedge rxusrclk2)
+    wrap_rxreset_ <= rxuserrdy & rxresetdone_gtx_r;
+always @ (posedge txusrclk2)
+    wrap_txreset_ <= txuserrdy & txresetdone_gtx_r;
 
 wire    [63:0]  rxdata_gtx;
 wire    [7:0]   rxcharisk_gtx;
@@ -164,9 +169,19 @@ if (DATA_BYTE_WIDTH == 4) begin
         .half_empty ()
     );
 
-    assign  txcomwake_gtx  = txdata_resync_out[36];
-    assign  txcominit_gtx  = txdata_resync_out[37];
-    assign  txelecidle_gtx = txdata_resync_out[38];
+
+    reg txcomwake_gtx_f;
+    reg txcominit_gtx_f;
+    reg txelecidle_gtx_f;
+    always @ (posedge txusrclk)
+    begin
+        txcomwake_gtx_f  <= txdata_resync_out[36];
+        txcominit_gtx_f  <= txdata_resync_out[37];
+        txelecidle_gtx_f <= txdata_resync_out[38];
+    end
+    assign  txcomwake_gtx  = txcomwake_gtx_f;
+    assign  txcominit_gtx  = txcominit_gtx_f;
+    assign  txelecidle_gtx = txelecidle_gtx_f;
 end
 else
 if (DATA_BYTE_WIDTH == 2) begin
@@ -309,6 +324,8 @@ gtx_elastic(
 */
 wire    rxcomwakedet_gtx;
 wire    rxcominitdet_gtx;
+reg     rxcomwakedet_gtx_r;
+reg     rxcominitdet_gtx_r;
 
 
 // insert resync if it's necessary
@@ -323,15 +340,15 @@ if (DATA_BYTE_WIDTH == 4) begin
     wire    rxdata_resync_strobe;
     wire    [50:0] rxdata_resync_in;
     wire    [50:0] rxdata_resync_out;
-    reg     [23:0] rxdata_resync_buf;
+    reg     [25:0] rxdata_resync_buf;
 
     assign  rxdata_resync_strobe = lword_strobe;
     assign  rxdata_resync_in = {
                                 isaligned,                                  // 1
-                                rxcomwakedet_gtx,                      // 1
-                                rxcominitdet_gtx,                      // 1
-                                rxresetdone_gtx,                       // 1 
-                                txresetdone_gtx,                       // 1
+                                rxcomwakedet_gtx_r | rxdata_resync_buf[25], // 1
+                                rxcominitdet_gtx_r | rxdata_resync_buf[24], // 1
+                                rxresetdone_gtx_r,                          // 1 
+                                txresetdone_gtx_r,                          // 1
                                 elastic_full  | rxdata_resync_buf[23], // 1 
                                 elastic_empty | rxdata_resync_buf[22], // 1
                                 rxdisperr_els_out,                     // 2
@@ -340,7 +357,7 @@ if (DATA_BYTE_WIDTH == 4) begin
                                 rxdata_els_out,                        // 16
                                 rxdata_resync_buf[21:0]};              // 22 / 51 total
     always @ (posedge rxusrclk)
-        rxdata_resync_buf    <= ~wrap_rxreset_ ? 24'h0 : ~rxdata_resync_strobe ? {elastic_full, elastic_empty, rxdisperr_els_out, rxnotintable_els_out, rxcharisk_els_out, rxdata_els_out} : rxdata_resync_buf;
+        rxdata_resync_buf    <= ~wrap_rxreset_ ? 26'h0 : ~rxdata_resync_strobe ? {rxcomwakedet_gtx_r, rxcominitdet_gtx_r, elastic_full, elastic_empty, rxdisperr_els_out, rxnotintable_els_out, rxcharisk_els_out, rxdata_els_out} : rxdata_resync_buf;
 
     always @ (posedge rxusrclk2)
         rxdata_resync_nempty_r <= rxdata_resync_nempty;
@@ -378,10 +395,10 @@ else
 if (DATA_BYTE_WIDTH == 2) begin
     // no resync is needed => straightforward assignments
     assign  rxbyteisaligned = isaligned;
-    assign  rxcomwakedet    = rxcomwakedet_gtx;
-    assign  rxcominitdet    = rxcominitdet_gtx;
-    assign  rxresetdone     = rxresetdone_gtx;
-    assign  txresetdone     = txresetdone_gtx;
+    assign  rxcomwakedet    = rxcomwakedet_gtx_r;
+    assign  rxcominitdet    = rxcominitdet_gtx_r;
+    assign  rxresetdone     = rxresetdone_gtx_r;
+    assign  txresetdone     = txresetdone_gtx_r;
     assign  rxelsfull           = elastic_full;
     assign  rxelsempty          = elastic_empty;
     assign  rxdisperr[1:0]      = rxdisperr_els_out;
@@ -398,6 +415,22 @@ else begin
 end
 endgenerate
 
+// latching gtx outputs, synchronous to RXUSRCLK2 = rxusrclk
+always @ (posedge rxusrclk)
+begin
+    rxcomwakedet_gtx_r  <= rxcomwakedet_gtx;
+    rxcominitdet_gtx_r  <= rxcominitdet_gtx;
+    rxresetdone_gtx_r   <= rxresetdone_gtx;
+    txresetdone_gtx_r   <= txresetdone_gtx;
+end
+
+wire    txoutclk_gtx;
+wire    xclk_gtx;
+wire    xclk_mr;
+BUFG bufg_txoutclk (.O(txoutclk),.I(txoutclk_gtx));
+BUFR bufr_xclk  (.O(xclk),.I(xclk_mr),.CE(1'b1),.CLR(1'b0));
+BUFMR bufmr_xclk  (.O(xclk_mr),.I(xclk_gtx));
+
 gtxe2_channel_wrapper #(
     .SIM_RECEIVER_DETECT_PASS               ("TRUE"),
     .SIM_TX_EIDLE_DRIVE_LEVEL               ("X"),
@@ -526,9 +559,9 @@ gtxe2_channel_wrapper #(
     .PD_TRANS_TIME_TO_P2                    (8'h64),
     .SAS_MAX_COM                            (64),
     .SAS_MIN_COM                            (36),
-    .SATA_BURST_SEQ_LEN                     (4'b0110),
-    .SATA_BURST_VAL                         (3'b110),
-    .SATA_EIDLE_VAL                         (3'b110),
+    .SATA_BURST_SEQ_LEN                     (4'b0101),
+    .SATA_BURST_VAL                         (3'b100),
+    .SATA_EIDLE_VAL                         (3'b100),
     .SATA_MAX_BURST                         (8),
     .SATA_MAX_INIT                          (21),
     .SATA_MAX_WAKE                          (7),
@@ -631,8 +664,8 @@ gtxe2_channel_wrapper(
     .DRPRDY                         (),
     .DRPWE                          (1'b0),
     .GTREFCLKMONITOR                (),
-    .QPLLCLK                        (gtrefclk),
-    .QPLLREFCLK                     (gtrefclk),
+    .QPLLCLK                        (1'b0/*gtrefclk*/),
+    .QPLLREFCLK                     (1'b0/*gtrefclk*/),
     .RXSYSCLKSEL                    (2'b00),
     .TXSYSCLKSEL                    (2'b00),
     .DMONITOROUT                    (),
@@ -729,7 +762,7 @@ gtxe2_channel_wrapper(
     .RXOSHOLD                       (1'b0),
     .RXOSOVRDEN                     (1'b0),
     .RXRATEDONE                     (),
-    .RXOUTCLK                       (xclk),
+    .RXOUTCLK                       (xclk_gtx),
     .RXOUTCLKFABRIC                 (),
     .RXOUTCLKPCS                    (),
     .RXOUTCLKSEL                    (3'b010),
@@ -806,7 +839,7 @@ gtxe2_channel_wrapper(
     .TXDATA                         (txdata_gtx),
     .GTXTXN                         (txn),
     .GTXTXP                         (txp),
-    .TXOUTCLK                       (txoutclk),
+    .TXOUTCLK                       (txoutclk_gtx),
     .TXOUTCLKFABRIC                 (),
     .TXOUTCLKPCS                    (),
     .TXOUTCLKSEL                    (3'b010),
@@ -832,4 +865,7 @@ gtxe2_channel_wrapper(
     .TXQPISENP                      ()
 );
 
+always @ (posedge gtrefclk)
+    debug <= ~rxelecidle | debug;
+
 endmodule

host/link.v

@@ -100,6 +100,17 @@ module link #(
     output  wire    [DATA_BYTE_WIDTH*8 - 1:0] phy_data_out,
     output  wire    [DATA_BYTE_WIDTH   - 1:0] phy_isk_out // charisk
 );
+// latching data-primitives stream from phy
+reg     [DATA_BYTE_WIDTH*8 - 1:0] phy_data_in_r;
+reg     [DATA_BYTE_WIDTH   - 1:0] phy_isk_in_r; // charisk
+reg     [DATA_BYTE_WIDTH   - 1:0] phy_err_in_r; // disperr | notintable
+always @ (posedge clk)
+begin
+    phy_data_in_r   <= phy_data_in;
+    phy_isk_in_r    <= phy_isk_in;
+    phy_err_in_r    <= phy_err_in;
+end
+
 wire    frame_done;
 // scrambled data
 wire    [DATA_BYTE_WIDTH*8 - 1:0]   scrambler_out;
@@ -494,7 +505,7 @@ scrambler scrambler(
     .val_in     (select_prim[CODE_DATA] | inc_is_data),
     .data_in    (crc_dword & {DATA_BYTE_WIDTH*8{select_prim[CODE_CRC]}} | 
                  data_in & {DATA_BYTE_WIDTH*8{select_prim[CODE_DATA]}} | 
-                 phy_data_in & {DATA_BYTE_WIDTH*8{inc_is_data}}),
+                 phy_data_in_r & {DATA_BYTE_WIDTH*8{inc_is_data}}),
     .data_out   (scrambler_out)
 );
 
@@ -557,24 +568,24 @@ assign  incom_invalidate = state_rcvr_eof & crc_bad & ~alignes_pair | state_rcvr
 // shows that incoming primitive or data is ready to be processed // TODO somehow move alignes_pair into dword_val
 assign  dword_val = |rcvd_dword & phy_ready & ~rcvd_dword[CODE_ALIGNP];
 // determine imcoming primitive type
-assign  rcvd_dword[CODE_DATA]	= ~|phy_isk_in;
+assign  rcvd_dword[CODE_DATA]	= ~|phy_isk_in_r;
 assign  rcvd_dword[CODE_CRC]	= 1'b0;
-assign  rcvd_dword[CODE_SYNCP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_SYNCP ] == phy_data_in;
-assign  rcvd_dword[CODE_ALIGNP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_ALIGNP] == phy_data_in;
-assign  rcvd_dword[CODE_XRDYP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_XRDYP ] == phy_data_in;
-assign  rcvd_dword[CODE_SOFP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_SOFP  ] == phy_data_in;
-assign  rcvd_dword[CODE_HOLDAP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_HOLDAP] == phy_data_in;
-assign  rcvd_dword[CODE_HOLDP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_HOLDP ] == phy_data_in;
-assign  rcvd_dword[CODE_EOFP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_EOFP  ] == phy_data_in;
-assign  rcvd_dword[CODE_WTRMP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_WTRMP ] == phy_data_in;
-assign  rcvd_dword[CODE_RRDYP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_RRDYP ] == phy_data_in;
-assign  rcvd_dword[CODE_IPP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_IPP   ] == phy_data_in;
-assign  rcvd_dword[CODE_DMATP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_DMATP ] == phy_data_in;
-assign  rcvd_dword[CODE_OKP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_OKP   ] == phy_data_in;
-assign  rcvd_dword[CODE_ERRP]	= phy_isk_in[0] == 1'b1 & ~|phy_isk_in[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_ERRP  ] == phy_data_in;
+assign  rcvd_dword[CODE_SYNCP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_SYNCP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_ALIGNP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_ALIGNP] == phy_data_in_r;
+assign  rcvd_dword[CODE_XRDYP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_XRDYP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_SOFP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_SOFP  ] == phy_data_in_r;
+assign  rcvd_dword[CODE_HOLDAP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_HOLDAP] == phy_data_in_r;
+assign  rcvd_dword[CODE_HOLDP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_HOLDP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_EOFP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_EOFP  ] == phy_data_in_r;
+assign  rcvd_dword[CODE_WTRMP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_WTRMP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_RRDYP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_RRDYP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_IPP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_IPP   ] == phy_data_in_r;
+assign  rcvd_dword[CODE_DMATP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_DMATP ] == phy_data_in_r;
+assign  rcvd_dword[CODE_OKP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_OKP   ] == phy_data_in_r;
+assign  rcvd_dword[CODE_ERRP]	= phy_isk_in_r[0] == 1'b1 & ~|phy_isk_in_r[DATA_BYTE_WIDTH-1:1] & prim_data[CODE_ERRP  ] == phy_data_in_r;
 
 // phy level errors handling TODO
-assign  dec_err = |phy_err_in;
+assign  dec_err = |phy_err_in_r;
 
 // form a response to transport layer
 assign  frame_done      = frame_done_good | frame_done_bad;
@@ -586,7 +597,7 @@ assign  frame_done_bad  = state_wait & dword_val & rcvd_dword[CODE_ERRP];
 always @ (posedge clk)
     if (~|rcvd_dword & phy_ready)
     begin
-        $display("%m: invalid primitive recieved : %h, conrol : %h, err : %h", phy_data_in, phy_isk_in, phy_err_in);
+        $display("%m: invalid primitive recieved : %h, conrol : %h, err : %h", phy_data_in_r, phy_isk_in_r, phy_err_in_r);
         $finish;
     end
 // States checker

host/oob.v

@@ -29,6 +29,7 @@ module oob #(
     parameter CLK_SPEED_GRADE = 1 // 1 - 75 Mhz, 2 - 150Mhz, 4 - 300Mhz
 )
 (
+    output  reg  [11:0] debug,
     // sata clk = usrclk2
     input   wire    clk,
     // reset oob
@@ -426,5 +427,20 @@ assign  eidle_timer_done = eidle_timer == 64;
 always @ (posedge clk)
     eidle_timer <= rst | rxelecidle | ~state_wait_eidle ? 8'b0 : eidle_timer + CLK_TO_TIMER_CONTRIB[7:0];
 
+always @ (posedge clk)
+    debug <= rst ? 12'h000 : { 
+                                state_idle,
+                                state_wait_cominit,
+                                state_wait_comwake,
+                                state_recal_tx,
+                                state_wait_eidle,
+                                state_wait_rxrst,
+                                state_wait_align,
+                                state_wait_synp,
+                                state_wait_linkup,
+                                state_error,
+                                oob_start,
+                                oob_error} | debug;
+                             
 
 endmodule

host/oob_ctrl.v

@@ -30,6 +30,7 @@ module oob_ctrl #(
     input   wire    rst,
     // gtx is ready = all resets are done
     input   wire    gtx_ready,
+    output  wire  [11:0] debug,
     // oob responces
     input   wire    rxcominitdet_in,
     input   wire    rxcomwakedet_in,
@@ -118,6 +119,7 @@ oob #(
 )
 oob
 (
+    .debug (debug),
 // sata clk = usrclk2
     .clk                            (clk),
 // reset oob

host/sata_host.v

@@ -29,6 +29,10 @@ module sata_host(
     output  wire    rst,
     // sata clk
     output  wire    clk,
+
+    // reliable clock to source drp and cpll lock det circuits
+    input   wire    reliable_clk,
+
 // temporary
     input   wire    [31:0]  al_cmd_in, // == {cmd_type, cmd_port, cmd_val, cmd_done_bad, cmd_done_good; cmd_busy}
     input   wire            al_cmd_val_in,
@@ -194,6 +198,12 @@ wire    [15:0]  sh_tran_cnt; // Transfer Count
 wire            sh_notif;
 wire            sh_autoact;
 
+// phy is ready - link is established
+wire    phy_ready;
+// tmp TODO
+wire    gtx_ready;
+wire    [11:0] debug_cnt;
+
 assign  sh_data_val_out = sh_data_val;
 assign  sh_data_out     = sh_data;
 assign  sh_control_out  = sh_control;
@@ -218,6 +228,10 @@ assign  sh_autoact_out  = sh_autoact;
 command command(
     .rst                                (rst),
     .clk                                (clk),
+    // temporary inputs TODO
+    .gtx_ready  (gtx_ready),
+    .phy_ready  (phy_ready),
+    .debug_cnt  (debug_cnt),
 
 // tl cmd iface
     .cmd_type                           (cl2tl_cmd_type),
@@ -560,8 +574,6 @@ transport transport(
 // oob sequence is reinitiated and link now is not established or rxelecidle
 //wire    link_reset; // use ~phy_ready instead
 
-// phy is ready - link is established
-wire    phy_ready;
 
 // data-primitives stream from phy
 wire    [DATA_BYTE_WIDTH*8 - 1:0] phy2ll_data;
@@ -657,8 +669,13 @@ sata_phy phy(
     // sata clk, generated in pll as usrclk2
     .clk                (clk),
 
+    // stable clock to source drp and cpll lock det circuits
+    .reliable_clk       (reliable_clk),
+
     // state
     .phy_ready          (phy_ready),
+    .gtx_ready          (gtx_ready),
+    .debug_cnt          (debug_cnt),
 
     // top-level ifaces
     // ref clk from an external source, shall be connected to pads

host/sata_phy.v

@@ -30,8 +30,14 @@ module sata_phy #(
     output  wire        clk,
     output  wire        rst,
 
+    // reliable clock to source drp and cpll lock det circuits
+    input   wire        reliable_clk,
+
     // state
     output  wire        phy_ready,
+    // tmp output TODO
+    output  wire        gtx_ready,
+    output  wire  [11:0] debug_cnt,
 
     // top-level ifaces
     // ref clk from an external source, shall be connected to pads
@@ -90,9 +96,9 @@ wire            rxreset_oob;
 wire            rxelsfull;
 wire            rxelsempty;
 
-wire            gtx_ready;
-
+//wire            gtx_ready;
 
+wire dummy;
 oob_ctrl oob_ctrl(
     // sata clk = usrclk2
     .clk                (clk),
@@ -100,6 +106,7 @@ oob_ctrl oob_ctrl(
     .rst                (rst),
     // gtx is ready = all resets are done
     .gtx_ready          (gtx_ready),
+    .debug ({dummy,debug_cnt[10:0]}),
     // oob responces
     .rxcominitdet_in    (rxcominitdet),
     .rxcomwakedet_in    (rxcomwakedet),
@@ -181,9 +188,26 @@ always @ (posedge gtrefclk)
  */
 wire    usrpll_locked;
 
+// make tx/rxreset synchronous to gtrefclk - gather singals from different domains: async, aclk, usrclk2, gtrefclk
+reg rxreset_f;
+reg txreset_f;
+reg rxreset_f_r;
+reg txreset_f_r;
+reg rxreset_f_rr;
+reg txreset_f_rr;
+always @ (posedge gtrefclk)
+begin
+    rxreset_f <= ~cplllock | cpllreset | rxreset_oob & gtx_configured;
+    txreset_f <= ~cplllock | cpllreset;
+    txreset_f_r <= txreset_f;
+    rxreset_f_r <= rxreset_f;
+    txreset_f_rr <= txreset_f_r;
+    rxreset_f_rr <= rxreset_f_r;
+end
+
+assign  rxreset = rxreset_f_rr;
+assign  txreset = txreset_f_rr;
 assign  cpllreset = extrst;
-assign  rxreset = ~cplllock | cpllreset | rxreset_oob & gtx_configured;
-assign  txreset = ~cplllock | cpllreset;
 assign  rxuserrdy = usrpll_locked & cplllock & ~cpllreset & ~rxreset & rxeyereset_done & sata_reset_done;
 assign  txuserrdy = usrpll_locked & cplllock & ~cpllreset & ~txreset & txpmareset_done & sata_reset_done;
 
@@ -239,9 +263,11 @@ wire    usrpll_fb_clk;
 wire    usrclk;
 wire    usrclk2;
 
-assign  txusrclk  = usrclk;
+wire usrclk_global;
+BUFG bufg_usrclk (.O(usrclk_global),.I(usrclk));
+assign  txusrclk  = usrclk_global;
 assign  txusrclk2 = usrclk2;
-assign  rxusrclk  = usrclk;
+assign  rxusrclk  = usrclk_global;
 assign  rxusrclk2 = usrclk2;
 
 PLLE2_ADV #(
@@ -330,6 +356,7 @@ gtx_wrap #(
 )
 gtx_wrap
 (
+    .debug (debug_cnt[11]),
     .cplllock           (cplllock),
     .cplllockdetclk     (cplllockdetclk),
     .cpllreset          (cpllreset),
@@ -370,8 +397,8 @@ gtx_wrap
 /*
  * Interfaces
  */
-assign  cplllockdetclk  = gtrefclk; //TODO
-assign  drpclk          = gtrefclk;
+assign  cplllockdetclk  = reliable_clk; //gtrefclk;
+assign  drpclk          = reliable_clk; //gtrefclk;
 
 //assign  clk             = usrclk2;
 BUFG bufg_sclk   (.O(clk),.I(usrclk2));

tb/tb_top.v

@@ -85,8 +85,8 @@ reg         SIMUL_AXI_FULL; // some data available
 wire        SIMUL_AXI_EMPTY;
 reg  [31:0] registered_rdata; // here read data from task
 
-//reg        CLK;
-wire    CLK;
+reg        CLK;
+//wire    CLK;
 reg        RST;
 reg        AR_SET_CMD_r;
 wire       AR_READY;
@@ -174,12 +174,12 @@ wire         #(AXI_TASK_HOLD) AW_SET_CMD = AW_SET_CMD_r;
 wire         #(AXI_TASK_HOLD) W_SET_CMD =  W_SET_CMD_r;
 
 //always #(CLKIN_PERIOD/2) CLK = ~CLK;
-assign CLK = dut.axi_aclk;
+//assign CLK = dut.axi_aclk;
 
 /*
  * connect axi ports to the dut
  */
-assign dut.ps7_i.FCLKCLK=        {4{EXTCLK_P}};
+assign dut.ps7_i.FCLKCLK=        {4{CLK}};
 assign dut.ps7_i.FCLKRESETN=     {RST,~RST,RST,~RST};
 // Read address
 assign dut.ps7_i.MAXIGP1ARADDR=  araddr;

tb/test_top.v

@@ -30,6 +30,12 @@ begin
     EXTCLK_N = ~EXTCLK_N;
 end
 
+// MAXI clock
+always #10
+begin
+    CLK = ~CLK;
+end
+
 integer i;
 integer status;
 integer id;
@@ -37,7 +43,7 @@ reg [31:0] data;
 // write registers
 initial
 begin
-//    CLK =1'b0;
+    CLK =1'b0;
     RST = 1'bx;
     AR_SET_CMD_r = 1'b0;
     AW_SET_CMD_r = 1'b0;

top.xdc

@@ -7,3 +7,8 @@ set_property PACKAGE_PIN AA5 [get_ports RXN]
 set_property PACKAGE_PIN AA6 [get_ports RXP]
 set_property PACKAGE_PIN AB3 [get_ports TXN]
 set_property PACKAGE_PIN AB4 [get_ports TXP]
+
+# manually placing usrpll in the same region where gtx is located : x0y0
+startgroup
+place_cell sata_top/sata_host/phy/usrclk_pll PLLE2_ADV_X0Y0/PLLE2_ADV
+endgroup
\ No newline at end of file

top_timing.xdc

@@ -8,6 +8,19 @@ create_clock -name gtrefclk -period 6.666 -waveform {0.000 3.333} [get_nets sata
 # after plls inside of GTX:
 create_clock -name txoutclk -period 6.666 -waveform {0.000 3.333} [get_nets sata_top/sata_host/phy/txoutclk]
 
+# recovered sata parallel clock
+create_clock -name xclk -period 6.666 -waveform {0.000 3.333} [get_nets sata_top/sata_host/phy/gtx_wrap/xclk]
+
 # txoutclk -> userpll, which gives us 2 clocks: userclk and userclk2. The second one is sata host clk
 create_generated_clock -name usrclk [get_nets sata_top/sata_host/phy/CLK]
-create_generated_clock -name sclk   [get_nets sata_top/sata_host/phy/clk]
+#create_generated_clock -name sclk   [get_nets sata_top/sata_host/phy/clk]
+create_generated_clock -name sclk   [get_nets sata_top_n_173]
+
+set_clock_groups -name async_clocks -asynchronous \
+-group {gtrefclk} \
+-group {axi_aclk0} \
+-group {xclk} \
+-group {usrclk} \
+-group {sclk} \
+-group {clk_axihp_pre} \
+-group {txoutclk}

wrapper/GTXE2_GPL.v

@@ -2685,6 +2685,9 @@ module GTXE2_GPL(
     input               QPLLCLK,
     input               QPLLREFCLK,
     
+// dunno
+    input               RXDFEVSEN,
+
 // Diffpairs
     input               GTXRXP,
     input               GTXRXN,

wrapper/gtxe2_channel_wrapper.v

@@ -986,6 +986,8 @@ gtx_unisims(
     .RXDFELFOVRDEN                                              (RXDFELFOVRDEN),
     .RXDFEUTHOLD                                                (RXDFEUTHOLD),
     .RXDFEUTOVRDEN                                              (RXDFEUTOVRDEN),
+    // this signal shall be present only in GTH, but for some reason it's included in unisims gtxe2 
+    .RXDFEVSEN                                                  (1'b0),
     .RXDFEVPHOLD                                                (RXDFEVPHOLD),
     .RXDFEVPOVRDEN                                              (RXDFEVPOVRDEN),
     .RXDFETAP2HOLD                                              (RXDFETAP2HOLD),