add fast configuration option to fpga_verilog to speed up full testbench simulation

openfpga/src/base/openfpga_verilog.cpp

@@ -67,6 +67,7 @@ int write_verilog_testbench(OpenfpgaContext& openfpga_ctx,
   CommandOptionId opt_output_dir = cmd.option("file");
   CommandOptionId opt_reference_benchmark = cmd.option("reference_benchmark_file_path");
   CommandOptionId opt_print_top_testbench = cmd.option("print_top_testbench");
+  CommandOptionId opt_fast_configuration = cmd.option("fast_configuration");
   CommandOptionId opt_print_formal_verification_top_netlist = cmd.option("print_formal_verification_top_netlist");
   CommandOptionId opt_print_preconfig_top_testbench = cmd.option("print_preconfig_top_testbench");
   CommandOptionId opt_print_simulation_ini = cmd.option("print_simulation_ini");
@@ -81,6 +82,7 @@ int write_verilog_testbench(OpenfpgaContext& openfpga_ctx,
   options.set_reference_benchmark_file_path(cmd_context.option_value(cmd, opt_reference_benchmark));
   options.set_print_formal_verification_top_netlist(cmd_context.option_enable(cmd, opt_print_formal_verification_top_netlist));
   options.set_print_preconfig_top_testbench(cmd_context.option_enable(cmd, opt_print_preconfig_top_testbench));
+  options.set_fast_configuration(cmd_context.option_enable(cmd, opt_fast_configuration));
   options.set_print_top_testbench(cmd_context.option_enable(cmd, opt_print_top_testbench));
   options.set_print_simulation_ini(cmd_context.option_value(cmd, opt_print_simulation_ini));
   options.set_explicit_port_mapping(cmd_context.option_enable(cmd, opt_explicit_port_mapping));

openfpga/src/base/openfpga_verilog_command.cpp

@@ -79,6 +79,9 @@ ShellCommandId add_openfpga_write_verilog_testbench_command(openfpga::Shell<Open
   /* Add an option '--print_top_testbench' */
   shell_cmd.add_option("print_top_testbench", false, "Generate a full testbench for top-level fabric module with autocheck capability");
 
+  /* Add an option '--fast_configuration' */
+  shell_cmd.add_option("fast_configuration", false, "Reduce the period of configuration by skip zero data points");
+
   /* Add an option '--print_formal_verification_top_netlist' */
   shell_cmd.add_option("print_formal_verification_top_netlist", false, "Generate a top-level module which can be used in formal verification");
 

openfpga/src/fpga_verilog/verilog_api.cpp

@@ -213,6 +213,7 @@ void fpga_verilog_testbench(const ModuleManager& module_manager,
                                 netlist_name,
                                 top_testbench_file_path,
                                 simulation_setting,
+                                options.fast_configuration(),
                                 options.explicit_port_mapping());
   }
  

openfpga/src/fpga_verilog/verilog_testbench_options.cpp

@@ -46,6 +46,10 @@ bool VerilogTestbenchOption::print_top_testbench() const {
   return print_top_testbench_;
 }
 
+bool VerilogTestbenchOption::fast_configuration() const {
+  return fast_configuration_;
+}
+
 bool VerilogTestbenchOption::print_simulation_ini() const {
   return !simulation_ini_path_.empty();
 }
@@ -82,6 +86,10 @@ void VerilogTestbenchOption::set_print_formal_verification_top_netlist(const boo
   print_formal_verification_top_netlist_ = enabled;
 }
 
+void VerilogTestbenchOption::set_fast_configuration(const bool& enabled) {
+  fast_configuration_ = enabled;
+}
+
 void VerilogTestbenchOption::set_print_preconfig_top_testbench(const bool& enabled) {
   print_preconfig_top_testbench_ = enabled
                                  && (!reference_benchmark_file_path_.empty());

openfpga/src/fpga_verilog/verilog_testbench_options.h

@@ -24,6 +24,7 @@ class VerilogTestbenchOption {
   public: /* Public accessors */
     std::string output_directory() const;
     std::string reference_benchmark_file_path() const;
+    bool fast_configuration() const;
     bool print_formal_verification_top_netlist() const;
     bool print_preconfig_top_testbench() const;
     bool print_top_testbench() const;
@@ -44,6 +45,7 @@ class VerilogTestbenchOption {
     void set_print_formal_verification_top_netlist(const bool& enabled);
     /* The preconfig top testbench generation can be enabled only when formal verification top netlist is enabled */
     void set_print_preconfig_top_testbench(const bool& enabled);
+    void set_fast_configuration(const bool& enabled);
     void set_print_top_testbench(const bool& enabled);
     void set_print_simulation_ini(const std::string& simulation_ini_path);
     void set_explicit_port_mapping(const bool& enabled);
@@ -51,6 +53,7 @@ class VerilogTestbenchOption {
   private: /* Internal Data */
     std::string output_directory_;
     std::string reference_benchmark_file_path_;
+    bool fast_configuration_;
     bool print_formal_verification_top_netlist_;
     bool print_preconfig_top_testbench_;
     bool print_top_testbench_;

openfpga/src/fpga_verilog/verilog_top_testbench.cpp

@@ -481,6 +481,48 @@ void print_verilog_top_testbench_ports(std::fstream& fp,
   fp << "\tinteger " << TOP_TESTBENCH_ERROR_COUNTER << "= 0;" << std::endl;
 }
 
+/********************************************************************
+ * Estimate the number of configuration clock cycles 
+ * by traversing the linked-list and count the number of SRAM=1 or BL=1&WL=1 in it.
+ * We plus 1 additional config clock cycle here because we need to reset everything during the first clock cycle
+ * If we consider fast configuration, the number of clock cycles will be 
+ * the number of non-zero data points in the fabric bitstream
+ * Note that this will not applicable to configuration chain!!!
+ *******************************************************************/
+static 
+size_t calculate_num_config_clock_cycles(const e_config_protocol_type& sram_orgz_type,
+                                         const bool& fast_configuration,
+                                         const FabricBitstream& fabric_bitstream) {
+  size_t num_config_clock_cycles = 1 + fabric_bitstream.bits().size();
+
+  /* Branch on the type of configuration protocol */
+  switch (sram_orgz_type) {
+  case CONFIG_MEM_STANDALONE:
+  case CONFIG_MEM_SCAN_CHAIN:
+  case CONFIG_MEM_MEMORY_BANK:
+    /* TODO */
+    break;
+  case CONFIG_MEM_FRAME_BASED: {
+    /* For fast configuration, we will skip all the zero data points */
+    if (true == fast_configuration) {
+      num_config_clock_cycles = 1;
+      for (const FabricBitId& bit_id : fabric_bitstream.bits()) {
+        if (true == fabric_bitstream.bit_din(bit_id)) {
+          num_config_clock_cycles++;
+        }
+      }
+    }
+    break;
+  }
+  default: 
+    VTR_LOGF_ERROR(__FILE__, __LINE__,
+                   "Invalid SRAM organization type!\n");
+    exit(1);
+  }
+
+  return num_config_clock_cycles;
+}
+
 /********************************************************************
  * Instanciate the input benchmark module
  *******************************************************************/
@@ -865,6 +907,7 @@ void print_verilog_top_testbench_configuration_chain_bitstream(std::fstream& fp,
  *******************************************************************/
 static 
 void print_verilog_top_testbench_frame_decoder_bitstream(std::fstream& fp,
+                                                         const bool& fast_configuration,
                                                          const ModuleManager& module_manager,
                                                          const ModuleId& top_module,
                                                          const FabricBitstream& fabric_bitstream) {
@@ -904,6 +947,12 @@ void print_verilog_top_testbench_frame_decoder_bitstream(std::fstream& fp,
    * We will visit the fabric bitstream in a reverse way  
    */
   for (const FabricBitId& bit_id : fabric_bitstream.bits()) {
+    /* When fast configuration is enabled, we skip zero data_in values */
+    if ((true == fast_configuration)
+      && (false == fabric_bitstream.bit_din(bit_id))) {
+      continue;
+    }
+
     fp << "\t\t" << std::string(TOP_TESTBENCH_PROG_TASK_NAME);
     fp << "(" << addr_port.get_width() << "'b";
     VTR_ASSERT(addr_port.get_width() == fabric_bitstream.bit_address(bit_id).size());
@@ -957,6 +1006,7 @@ void print_verilog_top_testbench_frame_decoder_bitstream(std::fstream& fp,
 static 
 void print_verilog_top_testbench_bitstream(std::fstream& fp,
                                            const e_config_protocol_type& sram_orgz_type,
+                                           const bool& fast_configuration,
                                            const ModuleManager& module_manager,
                                            const ModuleId& top_module,
                                            const BitstreamManager& bitstream_manager,
@@ -973,7 +1023,7 @@ void print_verilog_top_testbench_bitstream(std::fstream& fp,
     /* TODO */
     break;
   case CONFIG_MEM_FRAME_BASED:
-    print_verilog_top_testbench_frame_decoder_bitstream(fp,
+    print_verilog_top_testbench_frame_decoder_bitstream(fp, fast_configuration,
                                                         module_manager, top_module,
                                                         fabric_bitstream);
     break;
@@ -1017,6 +1067,7 @@ void print_verilog_top_testbench(const ModuleManager& module_manager,
                                  const std::string& circuit_name,
                                  const std::string& verilog_fname,
                                  const SimulationSetting& simulation_parameters,
+                                 const bool& fast_configuration,
                                  const bool& explicit_port_mapping) {
 
   std::string timer_message = std::string("Write autocheck testbench for FPGA top-level Verilog netlist for '") + circuit_name + std::string("'");
@@ -1050,11 +1101,10 @@ void print_verilog_top_testbench(const ModuleManager& module_manager,
   /* Find the clock period */
   float prog_clock_period = (1./simulation_parameters.programming_clock_frequency());
   float op_clock_period = (1./simulation_parameters.operating_clock_frequency());
-  /* Estimate the number of configuration clock cycles 
+  /* Estimate the number of configuration clock cycles */ 
-   * by traversing the linked-list and count the number of SRAM=1 or BL=1&WL=1 in it.
+  size_t num_config_clock_cycles = calculate_num_config_clock_cycles(sram_orgz_type,
-   * We plus 1 additional config clock cycle here because we need to reset everything during the first clock cycle
+                                                                     fast_configuration,
-   */
+                                                                     fabric_bitstream);
-  size_t num_config_clock_cycles = 1 + fabric_bitstream.bits().size();
 
   /* Generate stimuli for general control signals */
   print_verilog_top_testbench_generic_stimulus(fp,
@@ -1095,6 +1145,7 @@ void print_verilog_top_testbench(const ModuleManager& module_manager,
 
   /* load bitstream to FPGA fabric in a configuration phase */
   print_verilog_top_testbench_bitstream(fp, sram_orgz_type,
+                                        fast_configuration,
                                         module_manager, top_module,
                                         bitstream_manager, fabric_bitstream);
 

openfpga/src/fpga_verilog/verilog_top_testbench.h

@@ -35,6 +35,7 @@ void print_verilog_top_testbench(const ModuleManager& module_manager,
                                  const std::string& circuit_name,
                                  const std::string& verilog_fname,
                                  const SimulationSetting& simulation_parameters,
+                                 const bool& fast_configuration,
                                  const bool& explicit_port_mapping);
 
 } /* end namespace openfpga */

openfpga_flow/OpenFPGAShellScripts/configuration_frame_example_script.openfpga

@@ -49,7 +49,7 @@ write_fabric_verilog --file ./SRC --explicit_port_mapping --include_timing --inc
 #  - Enable top-level testbench which is a full verification including programming circuit and core logic of FPGA
 #  - Enable pre-configured top-level testbench which is a fast verification skipping programming phase
 #  - Simulation ini file is optional and is needed only when you need to interface different HDL simulators using openfpga flow-run scripts
-write_verilog_testbench --file ./SRC --reference_benchmark_file_path ${REFERENCE_VERILOG_TESTBENCH} --print_top_testbench --print_preconfig_top_testbench --print_simulation_ini ./SimulationDeck/simulation_deck.ini --explicit_port_mapping
+write_verilog_testbench --file ./SRC --reference_benchmark_file_path ${REFERENCE_VERILOG_TESTBENCH} --print_top_testbench --print_preconfig_top_testbench --print_simulation_ini ./SimulationDeck/simulation_deck.ini --explicit_port_mapping --fast_configuration
 
 # Write the SDC files for PnR backend
 #  - Turn on every options here

openfpga_flow/OpenFPGAShellScripts/full_testbench_example_script.openfpga

@@ -0,0 +1,68 @@
+# Run VPR for the 'and' design
+#--write_rr_graph example_rr_graph.xml
+vpr ${VPR_ARCH_FILE} ${VPR_TESTBENCH_BLIF} --clock_modeling route
+
+# Read OpenFPGA architecture definition
+read_openfpga_arch -f ${OPENFPGA_ARCH_FILE}
+
+# Annotate the OpenFPGA architecture to VPR data base
+# to debug use --verbose options
+link_openfpga_arch --activity_file ${ACTIVITY_FILE} --sort_gsb_chan_node_in_edges
+
+# Check and correct any naming conflicts in the BLIF netlist
+check_netlist_naming_conflict --fix --report ./netlist_renaming.xml
+
+# Apply fix-up to clustering nets based on routing results
+pb_pin_fixup --verbose
+
+# Apply fix-up to Look-Up Table truth tables based on packing results
+lut_truth_table_fixup
+
+# Build the module graph
+#  - Enabled compression on routing architecture modules
+#  - Enable pin duplication on grid modules
+build_fabric --compress_routing #--verbose
+
+# Write the fabric hierarchy of module graph to a file
+# This is used by hierarchical PnR flows
+write_fabric_hierarchy --file ./fabric_hierarchy.txt
+
+# Repack the netlist to physical pbs
+# This must be done before bitstream generator and testbench generation
+# Strongly recommend it is done after all the fix-up have been applied
+repack #--verbose
+
+# Build the bitstream
+#  - Output the fabric-independent bitstream to a file
+build_architecture_bitstream --verbose --file fabric_indepenent_bitstream.xml
+
+# Build fabric-dependent bitstream
+build_fabric_bitstream --verbose
+
+# Write the Verilog netlist for FPGA fabric
+#  - Enable the use of explicit port mapping in Verilog netlist
+write_fabric_verilog --file ./SRC --explicit_port_mapping --include_timing --include_signal_init --support_icarus_simulator --print_user_defined_template --verbose
+
+# Write the Verilog testbench for FPGA fabric
+#  - We suggest the use of same output directory as fabric Verilog netlists
+#  - Must specify the reference benchmark file if you want to output any testbenches
+#  - Enable top-level testbench which is a full verification including programming circuit and core logic of FPGA
+#  - Enable pre-configured top-level testbench which is a fast verification skipping programming phase
+#  - Simulation ini file is optional and is needed only when you need to interface different HDL simulators using openfpga flow-run scripts
+write_verilog_testbench --file ./SRC --reference_benchmark_file_path ${REFERENCE_VERILOG_TESTBENCH} --print_top_testbench --print_preconfig_top_testbench --print_simulation_ini ./SimulationDeck/simulation_deck.ini --explicit_port_mapping
+
+# Write the SDC files for PnR backend
+#  - Turn on every options here
+write_pnr_sdc --file ./SDC
+
+# Write SDC to disable timing for configure ports
+write_sdc_disable_timing_configure_ports --file ./SDC/disable_configure_ports.sdc
+
+# Write the SDC to run timing analysis for a mapped FPGA fabric
+write_analysis_sdc --file ./SDC_analysis
+
+# Finish and exit OpenFPGA
+exit
+
+# Note :
+# To run verification at the end of the flow maintain source in ./SRC directory

openfpga_flow/tasks/openfpga_shell/full_testbench/configuration_frame/config/task.conf

@@ -8,7 +8,7 @@
 
 [GENERAL]
 run_engine=openfpga_shell
-openfpga_shell_template=${PATH:OPENFPGA_PATH}/openfpga_flow/OpenFPGAShellScripts/configuration_frame_example_script.openfpga
+openfpga_shell_template=${PATH:OPENFPGA_PATH}/openfpga_flow/OpenFPGAShellScripts/full_testbench_example_script.openfpga
 power_tech_file = ${PATH:OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.xml
 power_analysis = true
 spice_output=false