Squash of Luca Codeluppi's thesis - Verif part #61

FrancescoConti · 2025-07-04T10:39:21Z

initial commit

multi-hwpe extension

adjustment in write check

corrected few errors

push test

QoS: bandwidth

adjusted an error in read check

few tests

handled the scenario with 0 hwpe, core, ext, or dma

waveform to check

few modifies

debug waveform

correct the previous commit: debug this waveform

PATH: corrected list of forbidden addresses, assign correct data during read transaction, create path if does't exist

modified the makefile to automatically generate the stimuli

corrected a small error in the python code

small error in the upper bound (address generation)

parametric number of HWPE_WIDTH

added latency

TEST

QoS arbitration multi-hwpe: debug

photo of waveforms for arbitration problem

Enriched the makefile: setup targets

updated python documentation

corrected an error in tb_top: PRIORITY_CHECK and RANDOM_REQ macros

adjusted QoS metrics

example config

modified hci_arbiter, added mode 1

cleaned the code

eliminated task check_hwpe_read_task: redundant

cleaned the code

changed the description of the arbiter

added another mode for the hci_arbiter

changed description of the hci_arbiter

removed unnecessary from tb_top.sv

modified a name in the makefile

removed nested generate in the arbiter

propagate arbiter mode through the hci

added TEST_RATIO: ratio between the number of transactions in the two branches

debug

corrected small error in tb_top

Improved the makefile and changed the python script accordingly

Eliminate few targets from makefile

Eliminated the following targets: setup_bandwidth, setup_data_integ, setup_arbiter_stall, setup_arbiter_no_stall. Replaced with test_1.mk, test_2.mk, test_3.mk, test_4.mk in config folder

forgot to add makefile to the staging area in the previous commit

Clean makefile

Fix MACRO_TB in the Makefile

Eliminate unnecessary in tb_top.sv

Change few messages in tb_top.sv

Change name N_TEST in config files to N_TRANSACTION and add EXACT_OR_MAX_OFFSET

Adapt python script to support EXACT_OR_MAX_OFFSET parameter

Implement hci_inteconnect_wrap.sv and fix small error in hci_arbiter.sv

Set default value for memory access pattern

Eliminate WIDTH_OF_MEMORY and improve readability of sim_time metric

Fix few comments

Add module queue_stimuli.sv

Add modules: queues_out and queues_rdata

Add comments to write transaction checker

Create verificaion_hci_package

Define parameters, structs and tasks in the package

Add arbiter_checker module

Add macro in hci_package.sv and set ecc to 0

Add progress bar

Modularize logic to compute QoS metrics

Add end_simulation, progress_bar modules and fix bugs in hci_wrapper

Add assignment modules between masters and hci

Eliminate data_for_read_transaction method from python class stimuli_generator

Fix output warning

Add comments to tb_top for clarity

Fix error in previous commit

[Update] python documentation

[Update] process_txt.py and class_stimuli_generator.py documentation

[Clean] verification_hci_package.sv

[Clean] the code

[Clean]

[FIX] small error in stimuli generation

[Documentation] Update

[FIX] README.md

[Documentation] Update

[Removed] unnecessary print

initial commit multi-hwpe extension adjustment in write check corrected few errors push test QoS: bandwidth adjusted an error in read check few tests handled the scenario with 0 hwpe, core, ext, or dma waveform to check few modifies debug waveform correct the previous commit: debug this waveform PATH: corrected list of forbidden addresses, assign correct data during read transaction, create path if does't exist modified the makefile to automatically generate the stimuli corrected a small error in the python code small error in the upper bound (address generation) parametric number of HWPE_WIDTH added latency TEST QoS arbitration multi-hwpe: debug photo of waveforms for arbitration problem Enriched the makefile: setup targets updated python documentation corrected an error in tb_top: PRIORITY_CHECK and RANDOM_REQ macros adjusted QoS metrics example config modified hci_arbiter, added mode 1 cleaned the code eliminated task check_hwpe_read_task: redundant cleaned the code changed the description of the arbiter added another mode for the hci_arbiter changed description of the hci_arbiter removed unnecessary from tb_top.sv modified a name in the makefile removed nested generate in the arbiter propagate arbiter mode through the hci added TEST_RATIO: ratio between the number of transactions in the two branches debug corrected small error in tb_top Improved the makefile and changed the python script accordingly Eliminate few targets from makefile Eliminated the following targets: setup_bandwidth, setup_data_integ, setup_arbiter_stall, setup_arbiter_no_stall. Replaced with test_1.mk, test_2.mk, test_3.mk, test_4.mk in config folder forgot to add makefile to the staging area in the previous commit Clean makefile Fix MACRO_TB in the Makefile Eliminate unnecessary in tb_top.sv Change few messages in tb_top.sv Change name N_TEST in config files to N_TRANSACTION and add EXACT_OR_MAX_OFFSET Adapt python script to support EXACT_OR_MAX_OFFSET parameter Implement hci_inteconnect_wrap.sv and fix small error in hci_arbiter.sv Set default value for memory access pattern Eliminate WIDTH_OF_MEMORY and improve readability of sim_time metric Fix few comments Add module queue_stimuli.sv Add modules: queues_out and queues_rdata Add comments to write transaction checker Create verificaion_hci_package Define parameters, structs and tasks in the package Add arbiter_checker module Add macro in hci_package.sv and set ecc to 0 Add progress bar Modularize logic to compute QoS metrics Add end_simulation, progress_bar modules and fix bugs in hci_wrapper Add assignment modules between masters and hci Eliminate data_for_read_transaction method from python class stimuli_generator Fix output warning Add comments to tb_top for clarity Fix error in previous commit [Update] python documentation [Update] process_txt.py and class_stimuli_generator.py documentation [Clean] verification_hci_package.sv [Clean] the code [Clean] [FIX] small error in stimuli generation [Documentation] Update [FIX] README.md [Documentation] Update [Removed] unnecessary print

Copilot

Pull Request Overview

This PR introduces a comprehensive verification environment for the Heterogeneous Cluster Interconnect (HCI) hardware IP, developed as part of Luca Codeluppi's thesis. The verification suite provides data integrity checking, bandwidth measurement, and QoS arbitration testing capabilities for the HCI interconnect.

Complete testbench infrastructure with clock/reset generation, stimuli generation, and result verification
Python-based stimuli generator supporting various memory access patterns (random, linear, 2D, 3D)
Configurable simulation framework with makefile-based setup and multiple test configurations

Reviewed Changes

Copilot reviewed 39 out of 41 changed files in this pull request and generated 20 comments.

Show a summary per file

File	Description
verif/sv/tb_top.sv	Main testbench module implementing HCI verification with transaction checking and QoS metrics
verif/sv/other_modules/*.sv	Supporting modules for stimuli queues, data verification, latency computation, and progress tracking
verif/stimuli_generator/*.py	Python stimuli generation framework with configurable memory access patterns
config/*.mk	Configuration files for hardware parameters, simulation settings, and test scenarios
Makefile	Build system with targets for setup, stimuli generation, compilation, and simulation
README.md	Documentation for verification environment setup and usage

Comments suppressed due to low confidence (1)

verif/sv/other_modules/end_simulation_and_final_report.sv:54

Misspelling: 'troughput' should be 'throughput' (appears twice)

    $display("PERFORMANCE RATING %f%%\n", troughput_real/troughput_theo*100);

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

verif/sv/tb_top.sv

Copilot · 2025-08-28T16:24:10Z

verif/sv/tb_top.sv

+  // REAL TROUGHPUT AND SIMULATION TIME //
+  ////////////////////////////////////////
+  real latency_per_master[N_MASTER];
+  real troughput_real;


Misspelling: 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:11Z

verif/sv/tb_top.sv

+    .END_LATENCY(END_LATENCY),
+    .rst_n(rst_n),
+    .clk(clk),
+    .troughput_real(troughput_real),


Misspelling: 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:13Z

verif/sv/other_modules/end_simulation_and_final_report.sv

+    wait(troughput_real>=0);
+    $display("\\\\BANDWIDTH\\\\");
+    $display("IDEAL APPLICATION PEAK BANDWIDTH (KERNEL DEPENDENT): %f bit per cycle",troughput_theo);
+    $display("REAL APPLICATION BANDWITH: %f bit per cycle",troughput_real);


Two misspellings: 'BANDWITH' should be 'BANDWIDTH' and 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:13Z

verif/sv/common/verification_hci_package.sv

+    end
+  endtask
+
+  task calculate_theoretical_throughput(output real troughput_theo);


Misspelling: 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:13Z

verif/sv/common/verification_hci_package.sv

+    end
+
+    tot_data = ((N_TRANSACTION_LOG * DATA_WIDTH) * (N_MASTER_REAL - N_HWPE_REAL) + (N_TRANSACTION_HWPE * HWPE_WIDTH * DATA_WIDTH) * N_HWPE_REAL); // bit
+    troughput_theo = tot_data/tot_time; // bit per cycle


Misspelling: 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:14Z

verif/sv/common/verification_hci_package.sv

+    troughput_theo = tot_data/tot_time; // bit per cycle
+    band_memory_limit = real'(N_BANKS * DATA_WIDTH);
+    if (troughput_theo >= band_memory_limit) begin
+      troughput_theo = band_memory_limit;


Misspelling: 'troughput' should be 'throughput'

Copilot · 2025-08-28T16:24:14Z

verif/stimuli_generator/stimuli_gen_main.py

+    print("ERROR: Incorrect number of parameters in --sim_and_hardware_params")
+    print("Expected: 11 parameters")
+    print("Passed: ", len(args.sim_and_hardware_params), "parameters")


The error message says '11 parameters' but the check on line 148 expects 13 parameters. This inconsistency will confuse users when the error occurs.

Suggested change

print("ERROR: Incorrect number of parameters in --sim_and_hardware_params")

print("Expected: 11 parameters")

print("Passed: ", len(args.sim_and_hardware_params), "parameters")

print("ERROR: Incorrect number of parameters in --sim_and_hardware_params")

print("Expected: 13 parameters")

print("Passed: ", len(args.sim_and_hardware_params), "parameters")

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

FrancescoConti mentioned this pull request Jul 4, 2025

Prevent spurious writes when byte enable is all zero. #58

Open

FrancescoConti mentioned this pull request Aug 28, 2025

Fix FIFO delay parameter propagation in HCI #63

Open

FrancescoConti requested a review from Copilot August 28, 2025 16:18

FrancescoConti marked this pull request as ready for review August 28, 2025 16:18

Copilot AI reviewed Aug 28, 2025

View reviewed changes

Update verif/sv/tb_top.sv

011f4a5

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Squash of Luca Codeluppi's thesis - Verif part #61

Squash of Luca Codeluppi's thesis - Verif part #61

Uh oh!

FrancescoConti commented Jul 4, 2025

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Copilot AI Aug 28, 2025

Uh oh!

Uh oh!

Squash of Luca Codeluppi's thesis - Verif part #61

Are you sure you want to change the base?

Squash of Luca Codeluppi's thesis - Verif part #61

Uh oh!

Conversation

FrancescoConti commented Jul 4, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Aug 28, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!