URL
https://opencores.org/ocsvn/openhmc/openhmc/trunk
Subversion Repositories openhmc
[/] [openhmc/] [trunk/] [openHMC/] [rtl/] [hmc_controller/] [rx/] [rx_descrambler.v] - Rev 15
Compare with Previous | Blame | View Log
/* * .--------------. .----------------. .------------. * | .------------. | .--------------. | .----------. | * | | ____ ____ | | | ____ ____ | | | ______ | | * | ||_ || _|| | ||_ \ / _|| | | .' ___ || | * ___ _ __ ___ _ __ | | | |__| | | | | | \/ | | | |/ .' \_|| | * / _ \| '_ \ / _ \ '_ \ | | | __ | | | | | |\ /| | | | || | | | * (_) | |_) | __/ | | || | _| | | |_ | | | _| |_\/_| |_ | | |\ `.___.'\| | * \___/| .__/ \___|_| |_|| ||____||____|| | ||_____||_____|| | | `._____.'| | * | | | | | | | | | | | | * |_| | '------------' | '--------------' | '----------' | * '--------------' '----------------' '------------' * * openHMC - An Open Source Hybrid Memory Cube Controller * (C) Copyright 2014 Computer Architecture Group - University of Heidelberg * www.ziti.uni-heidelberg.de * B6, 26 * 68159 Mannheim * Germany * * Contact: openhmc@ziti.uni-heidelberg.de * http://ra.ziti.uni-heidelberg.de/openhmc * * This source file is free software: you can redistribute it and/or modify * it under the terms of the GNU Lesser General Public License as published by * the Free Software Foundation, either version 3 of the License, or * (at your option) any later version. * * This source file is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public License * along with this source file. If not, see <http://www.gnu.org/licenses/>. * * * Module name: rx_descrambler * * Scrambler Logic (HMC Spec version 1.0) * This module implements a parallel scrambler based on the * polynomial 1+ x^(-14) + x^(-15). * * Such Scrambler is typically shown as a 15 bit Linear Feedback Shift Register * (LFSR) with bits shifting from register 1 on the left to register 15 on the * right, with register 14 and 15 combining to shift into register 1. * The HMC Serializer outputs data[0] first from parallel tx data[n:0], * so if data[n:0] is to be bitwise scrambled with LFSR[n:0], we need the LFSR * to shift from n -> 0, the opposite direction from the typical illustration. * This implementation shifts data from LFSR[14] on the left to LFSR[0] on the * right, with LFSR[1] and [0] combining to shift into LFSR[14]. This way * LFSR[14:0] can bitwise scramble data[14:0] and be compatible with serializ- * ation that shifts out on the data[0] side. * Put otherwise: Polynomial 1+ x^(-14) + x^(-15) is equiv to * x^15 + x^1 + x^0 * * This parallelized version calculates the next DWIDTH steps of values for * the LFSR. These bits are used to scramble the parallel input, and to * choose the next value of lfsr (lfsr_steps[DWIDTH-1]). * * This is the descrambler. It is self-seeding. When lock is asserted it has * successfully found the correct value for the LFSR. It is only implemented * for DWIDTH > 14. * * Since we know that scrambled zeros are being translated, we can calculate * what the seed will be in the next timestep. In order to simplify the * calculation, we assume that the top bit is a one. That has the happy side- * effect of letting us know that the seed didn't get stuck at all zeros. * * After the scrambler is locked, the input word may need to be aligned. The * bit_slip input allows the scrambler to shift one bit with the serializer * to keep the scrambler in sync. */ `default_nettype none module rx_descrambler #( parameter DWIDTH=16, parameter BITSLIP_SHIFT_RIGHT=1 ) ( input wire clk, input wire res_n, input wire bit_slip, input wire can_lock, output reg locked, input wire [DWIDTH-1:0] data_in, output reg [DWIDTH-1:0] data_out ); reg [14:0] lfsr; // LINEAR FEEDBACK SHIFT REGISTER wire [14:0] lfsr_slipped; // Temporary lfsr for bitslip wire [14:0] lfsr_steps [DWIDTH-1:0]; // LFSR values for serial time steps wire [14:0] calculated_seed; wire [DWIDTH-1:0] data_out_tmp; generate if(BITSLIP_SHIFT_RIGHT==1) begin assign lfsr_slipped = { (lfsr_steps[DWIDTH-1][1] ^ lfsr_steps[DWIDTH-1][0]) , lfsr_steps[DWIDTH-1][14:1] }; end else begin assign lfsr_slipped = { lfsr_steps[DWIDTH-1][13:0], (lfsr_steps[DWIDTH-1][14] ^ lfsr_steps[DWIDTH-1][0])}; end endgenerate `ifdef SIMULATION initial begin lfsr <= 15'h0; end `endif // SEQUENTIAL PROCESS `ifdef ASYNC_RES always @(posedge clk or negedge res_n) begin `else always @(posedge clk) begin `endif `ifdef RESET_ALL if(!res_n) begin data_out <= {DWIDTH {1'b0}}; lfsr <= 15'h0; end else `endif begin data_out <= data_out_tmp; if (!locked) begin lfsr <= calculated_seed; end else begin if (bit_slip) begin lfsr <= lfsr_slipped; end else begin lfsr <= lfsr_steps[DWIDTH-1]; end end end if(!res_n) begin locked <= 1'b0; end else begin if (!locked) begin if (calculated_seed == lfsr_steps[DWIDTH-1]) begin locked <= 1'b1; end end if(!can_lock) begin locked <= 1'b0; end end end // serial shift right with left input // SCRAMBLE genvar j; generate localparam OFFSET = DWIDTH-15; // It breaks here if DWIDTH < 15 assign calculated_seed[14] = 1'b1; // Guess the top bit is 1 // data_in is the past state of the LFSR, so we can figure out // the current value using a loop. for(j = 0; j < 14; j = j + 1) begin : seed_calc assign calculated_seed[j] = data_in[j+OFFSET] ^ data_in[j+OFFSET+1]; end assign data_out_tmp [0] = data_in[0] ^ lfsr[0]; // single bit scrambled assign lfsr_steps[0] = { (lfsr[1] ^ lfsr[0]) , lfsr[14:1] }; // lfsr at next bit clock for(j = 1; j < DWIDTH; j = j + 1) begin : scrambler_gen assign data_out_tmp[j] = data_in[j] ^ lfsr_steps[j-1][0]; assign lfsr_steps[j] = { (lfsr_steps[j-1][1] ^ lfsr_steps[j-1][0]) , lfsr_steps[j-1][14:1] }; end endgenerate endmodule `default_nettype wire