Optimize WARNING_QUIT fatal-exit output

[Growl1234]

This PR makes WARNING_QUIT output better match its behavior.

WARNING_QUIT exits with a non-zero status, so the screen banner is changed from NOTICE to ERROR. It also no longer prints TIME STATISTICS after the fatal message, keeping the actual error easier to find in failed runs.

Normal QUIT() behavior is unchanged.

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Updates error-exit handling to present failures as ERROR (instead of NOTICE) and adjusts WARNING_QUIT to skip timer/statistics output on error exits.

Changes:

• Refactors common quit logic into a new quit_impl(ret, print_timer) helper.
• Changes quit banner text from NOTICE to ERROR and updates multiple tests accordingly.
• Updates unit tests to assert error visibility and absence of timer output for WARNING_QUIT.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
source/source_basis/module_pw/test/test-other.cpp	Updates expected quit output substring to ERROR.
source/source_base/tool_quit.cpp	Introduces quit_impl, changes quit banners to !ERROR!, and skips timer output for WARNING_QUIT.
source/source_base/test/tool_quit_test.cpp	Updates expectations to look for ERROR and ensure timer stats are not printed on error exits.
source/source_base/test/tool_check_test.cpp	Updates expected quit output substring to ERROR.
source/source_base/test/math_chebyshev_test.cpp	Updates expected quit output substring to ERROR.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[mohanchen]

We recommend keeping the C++ syntax simple: just use void and avoid adding new syntax elements.

[mohanchen]

This is not always an error. We intentionally set the log level to NOTICE.

[mohanchen]

We prefer not to change NOTICE to ERROR. Psychologically speaking, ERROR is treated as a fault indicator, which may lead users to assume the problem stems from their own operations and create negative psychological implications.

[mohanchen]

We output timing data specifically to help find related errors.

[mohanchen]

This change targets core code. We have thought it through carefully. For similar questions, please raise an issue first.

[Growl1234]

Thanks for the clarification. I now understand that the NOTICE banner design choices.

However, I still have doubt whether timing data is the best way to help users locate the cause of a WARNING_QUIT. The timing table is quite bulky and may weaken the visibility of the actual notice message. If the purpose is diagnostics, source-location information and a routine calling stack would be more direct.