utl: Adds source lines to Logger by QuantamHD · Pull Request #10231 · The-OpenROAD-Project/OpenROAD

QuantamHD · 2026-04-22T23:26:25Z

Summary

This PR adds source lines to warn+ log messages to aide in developer/agentic debugging.

This can be disabled by setting an env variable

DISABLE_SOURCE_LINES=1

ethanmoon@ethanmoon2:~/OpenROAD$ ./bazel-bin/openroad
OpenROAD bazel-nostamp
Features included (+) or not (-): -GPU -GUI -Python
This program is licensed under the BSD-3 license. See the LICENSE file for details.
Components of this program may be licensed under more restrictive licenses which must be honored.
openroad> initialize_floorplan
[ERROR IFP-0019] [bazel-out/k8-opt/bin/src/utl/swig.cc:2326] no -utilization or -die_area specified.
IFP-0019
while evaluating initialize_floorplan
openroad> [WARNING UTL-0014] [src/utl/src/Progress.cpp:52] OpenROAD will terminate after the next control+C

Type of Change

New feature

Verification

I have verified that the local build succeeds (./etc/Build.sh).
I have run the relevant tests and they pass.
My code follows the repository's formatting guidelines.
I have signed my commits (DCO).

This PR adds source lines to warn+ log messages to aide in developer/agentic debugging. This can be disabled by setting an env variable ``` DISABLE_SOURCE_LINES=1 ``` ``` ethanmoon@ethanmoon2:~/OpenROAD$ ./bazel-bin/openroad OpenROAD bazel-nostamp Features included (+) or not (-): -GPU -GUI -Python This program is licensed under the BSD-3 license. See the LICENSE file for details. Components of this program may be licensed under more restrictive licenses which must be honored. openroad> initialize_floorplan [ERROR IFP-0019] [bazel-out/k8-opt/bin/src/utl/swig.cc:2326] no -utilization or -die_area specified. IFP-0019 while evaluating initialize_floorplan openroad> [WARNING UTL-0014] [src/utl/src/Progress.cpp:52] OpenROAD will terminate after the next control+C ``` Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

gemini-code-assist

Code Review

This pull request integrates std::source_location into the logging framework to provide file and line information for warning, error, and critical log levels. It introduces a LogMessage struct to encapsulate the message and location, adds environment variable support to disable the feature, and updates the build system to handle path mapping. Review feedback identifies a significant compatibility risk with the project's minimum compiler versions, a breaking change in the SWIG-based scripting interfaces, and a recommendation to pass LogMessage by reference to avoid unnecessary string copies.

github-actions

clang-tidy made some suggestions

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

gadfort · 2026-04-23T02:43:15Z

I think this should be disabled by default and enabled as needed in a users environment

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions · 2026-04-23T06:10:35Z

clang-tidy review says "All clean, LGTM! 👍"

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions · 2026-04-23T06:26:30Z

clang-tidy review says "All clean, LGTM! 👍"

github-actions · 2026-04-23T06:27:11Z

clang-tidy review says "All clean, LGTM! 👍"

maliberty · 2026-04-23T18:31:22Z

[ERROR IFP-0019] [bazel-out/k8-opt/bin/src/utl/swig.cc:2326] no -utilization or -die_area specified.
IFP-0019

This isn't very helpful as the error is in ifp but the line in utl.

QuantamHD · 2026-04-23T23:55:01Z

The true error comes from TCL embedded in the app, but the C++ can't know that. We would need to wrap this logic in TCL proc and have TCL say what line number it came from.

I think it's a good start, and more motivation to move things into C++ rather than TCL.

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions · 2026-04-24T01:58:30Z

clang-tidy review says "All clean, LGTM! 👍"

maliberty · 2026-04-24T02:36:57Z

I think this should be off by default

QuantamHD · 2026-04-24T03:46:56Z

@gadfort @maliberty

Let me make an argument for keeping it on by default, if you still respond no I'll change it. Agents need more context to work in a loop this provides them that context to get them unstuck. I think if we're aiming for more agents to use OpenROAD in a loop this is beneficial to that use case. Keeping it off by default reduces the value of this feature significantly for the agent use case.

maliberty · 2026-04-24T14:19:42Z

Perhaps you picked a bad example as the information shown would mislead an agent afaik (i.e. nothing pointed to ifp). Do you have a positive c++ example? Thoughts on how we can suppress it when coming from tcl?

QuantamHD · 2026-04-27T19:40:37Z

ethanmoon@ethanmoon2:~/OpenROAD$ bazelisk run //src/dpl/test:pad02-tcl_test
[INFO ODB-0227] LEF file: Nangate45/Nangate45.lef, created 22 layers, 27 vias, 135 library cells
[INFO ODB-0128] Design: single_cell
[INFO ODB-0130]     Created 2 pins.
[INFO ODB-0131]     Created 1 components and 4 component-terminals.
[INFO ODB-0132]     Created 2 special nets and 2 connections.
[INFO ODB-0133]     Created 2 nets and 2 connections.
[INFO DPL-0006] Core area: 34.05 um^2, Instances area: 1.86 um^2, Utilization: 5.5%
[INFO DPL-0005] Diamond search max displacement: +/- 500 sites horizontally, +/- 100 rows vertically.
[INFO DPL-1101] Legalizing using diamond search.
[ERROR DPL-0015] [src/dpl/src/Place.cpp:376] instance _277_ does not fit inside the ROW core area.
DPL-0015

github-actions · 2026-04-27T19:45:52Z

clang-tidy review says "All clean, LGTM! 👍"

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

QuantamHD · 2026-04-27T20:29:56Z

@maliberty suppressed errors and warnings originating from TCL.

ethanmoon@ethanmoon2:~/OpenROAD$ bazel-bin/openroad
OpenROAD bazel-nostamp 
Features included (+) or not (-): -GPU -GUI -Python
This program is licensed under the BSD-3 license. See the LICENSE file for details.
Components of this program may be licensed under more restrictive licenses which must be honored.
openroad> initialize_floorplan 
[ERROR IFP-0019] no -utilization or -die_area specified.
IFP-0019
while evaluating initialize_floorplan 
openroad> [WARNING UTL-0014] [src/utl/src/Progress.cpp:52] OpenROAD will terminate after the next control+C

github-actions · 2026-04-27T20:33:13Z

clang-tidy review says "All clean, LGTM! 👍"

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions · 2026-04-27T20:45:53Z

clang-tidy review says "All clean, LGTM! 👍"

maliberty · 2026-04-27T20:50:15Z

@oharboe opinion on on/off by default? Is this useful to you or your AI?

oharboe · 2026-04-27T20:52:56Z

No opinion. My use-case have changed though, I don't read logs as much as I used to, so I'm less sensitive to log spam, I suppose. Mostly claude reads logs these days.

QuantamHD · 2026-04-27T20:57:12Z

@oharboe The goal with this change is to give Gemini/claude more context on errors, namely to allow it to explore the code to get itself unstuck.

oharboe · 2026-04-27T21:00:21Z

@oharboe The goal with this change is to give Gemini/claude more context on errors, namely to allow it to explore the code to get itself unstuck.

Yes. This sort of logging fixes help enormously in closing the feedback loop is our experience too. I think it will work.

maliberty · 2026-04-27T22:06:32Z

It is too front loaded with context, perhaps:

[ERROR IFP-0019] no -utilization or -die_area specified. [bazel-out/k8-opt/bin/src/utl/swig.cc:2326]

or

[ERROR IFP-0019] no -utilization or -die_area specified. 
     [bazel-out/k8-opt/bin/src/utl/swig.cc:2326]

maliberty · 2026-04-27T22:06:57Z

A command should be introduced to control this rather than an envar.

maliberty · 2026-04-27T22:07:37Z

debug/debugPrint too please.

QuantamHD · 2026-04-27T22:34:42Z

It is too front loaded with context, perhaps:

[ERROR IFP-0019] no -utilization or -die_area specified. [bazel-out/k8-opt/bin/src/utl/swig.cc:2326]

or

[ERROR IFP-0019] no -utilization or -die_area specified. 
     [bazel-out/k8-opt/bin/src/utl/swig.cc:2326]

This is difficult to acomodate due to the way splog interacts with std::format. You can append things to the front, but not the rear.

QuantamHD · 2026-04-27T22:36:28Z

A command should be introduced to control this rather than an

I'm amenable to a command, but the envvar is specifically for the unit tests to ensure the test harness can turn off this behavior for all unit tests at once. Are you okay if we have both?

maliberty · 2026-04-27T22:53:00Z

This is difficult to acomodate due to the way splog interacts with std::format. You can append things to the front, but not the rear.

AI suggests pre-formatting the user portion:

  auto formatted = fmt::format(FMT_RUNTIME(message), args...);         
  logger_->log(level,                                                           
               "[{} {}-{:04d}] {} [{}:{}]{}",
               level_names[level],                                              
               tool_names_[tool],                                               
               id,
               formatted,                                                       
               loc.file_name(),                                        
               loc.line(),
               spdlog::details::os::default_eol);

maliberty · 2026-04-27T22:54:29Z

I'm amenable to a command, but the envvar is specifically for the unit tests to ensure the test harness can turn off this behavior for all unit tests at once. Are you okay if we have both?

We have helpers.tcl and helpers.py in test/ where you could add the command (we already do this for supress_message).

QuantamHD · 2026-04-28T00:08:25Z

Do you want debug messages to always have source lines or be controllable similar to the other message. Since the debug messages are largely dev only I would argue the control doesn't make as much sense. It would also require duplicate code since the code paths for debug are different than those of the other log messages

QuantamHD · 2026-04-28T00:10:06Z

ethanmoon@ethanmoon2:~/OpenROAD$ bazel-bin/openroad
OpenROAD bazel-nostamp 
Features included (+) or not (-): -GPU -GUI -Python
This program is licensed under the BSD-3 license. See the LICENSE file for details.
Components of this program may be licensed under more restrictive licenses which must be honored.
openroad> [WARNING UTL-0014] OpenROAD will terminate after the next control+C [src/utl/src/Progress.cpp:52]

@gadfort @maliberty We happy with this?

gadfort · 2026-04-28T00:18:22Z

@QuantamHD much better :-) thanks

gadfort · 2026-04-28T00:19:46Z

@QuantamHD also to answer the other question, I would generally agree that on debug it's fine to how them without any gate

maliberty · 2026-04-28T03:02:37Z

Agree for debug they are helpful.

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions

clang-tidy made some suggestions

github-actions · 2026-05-15T22:00:16Z

    addMetricsSink(metrics_filename);
  }

+  const char* disable_source_lines = std::getenv("DISABLE_SOURCE_LINES");


warning: no header providing "std::getenv" is directly included [misc-include-cleaner]

src/utl/src/Logger.cpp:8:

- #include <cstring> + #include <cstdlib> + #include <cstring>

maliberty · 2026-06-01T21:51:29Z

@QuantamHD this seems to have stalled out. Is there more work from your perspective?

github-actions Bot added the size/M label Apr 22, 2026

gemini-code-assist Bot reviewed Apr 22, 2026

View reviewed changes

Comment thread CMakeLists.txt

Comment thread src/utl/src/Logger.i Outdated

Comment thread src/utl/include/utl/Logger.h Outdated

github-actions Bot reviewed Apr 22, 2026

View reviewed changes

clang-format

4ac4bb6

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

clang-tidy

716e4d1

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

QuantamHD added 2 commits April 23, 2026 06:22

fix c++ tests with source line assert

2f9f8a0

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

clang-format

cf91ddd

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

maliberty reviewed Apr 23, 2026

View reviewed changes

Comment thread CMakeLists.txt

buildifier

6690a11

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

Merge remote-tracking branch 'origin/master' into source_lines

b802f20

address feedback for suppressing source lines from TCL errors

f8120f0

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

rerun ci

61318bc

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

QuantamHD requested a review from maliberty April 27, 2026 20:42

maliberty reviewed Apr 28, 2026

View reviewed changes

Comment thread src/mpl/test/cpp/CMakeLists.txt

maliberty reviewed Apr 28, 2026

View reviewed changes

Comment thread src/mpl/test/BUILD Outdated

Add back env variable

cd7d8cf

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

QuantamHD requested review from a team as code owners May 15, 2026 21:51

QuantamHD requested review from AcKoucher, maliberty and osamahammad21 May 15, 2026 21:51

QuantamHD added 2 commits May 15, 2026 21:54

Merge remote-tracking branch 'origin/master' into source_lines

81dbc00

update sta

70fef3a

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>

github-actions Bot reviewed May 15, 2026

View reviewed changes

Uh oh!

Conversation

QuantamHD commented Apr 22, 2026

Summary

Type of Change

Verification

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gadfort commented Apr 23, 2026

Uh oh!

github-actions Bot commented Apr 23, 2026

Uh oh!

github-actions Bot commented Apr 23, 2026

Uh oh!

github-actions Bot commented Apr 23, 2026

Uh oh!

Uh oh!

maliberty commented Apr 23, 2026

Uh oh!

QuantamHD commented Apr 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Apr 24, 2026

Uh oh!

maliberty commented Apr 24, 2026

Uh oh!

QuantamHD commented Apr 24, 2026

Uh oh!

maliberty commented Apr 24, 2026

Uh oh!

QuantamHD commented Apr 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Apr 27, 2026

Uh oh!

QuantamHD commented Apr 27, 2026

Uh oh!

github-actions Bot commented Apr 27, 2026

Uh oh!

github-actions Bot commented Apr 27, 2026

Uh oh!

maliberty commented Apr 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

oharboe commented Apr 27, 2026

Uh oh!

QuantamHD commented Apr 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

oharboe commented Apr 27, 2026

Uh oh!

maliberty commented Apr 27, 2026

Uh oh!

maliberty commented Apr 27, 2026

Uh oh!

maliberty commented Apr 27, 2026

Uh oh!

QuantamHD commented Apr 27, 2026

Uh oh!

QuantamHD commented Apr 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

maliberty commented Apr 27, 2026

Uh oh!

maliberty commented Apr 27, 2026

QuantamHD commented Apr 23, 2026 •

edited

Loading

QuantamHD commented Apr 27, 2026 •

edited

Loading

maliberty commented Apr 27, 2026 •

edited

Loading

QuantamHD commented Apr 27, 2026 •

edited

Loading

QuantamHD commented Apr 27, 2026 •

edited

Loading