`--stop-on-failure` no longer stops on errors

[Jean85]

Q	A
PHPUnit version	10.x-dev
PHP version	8.1
Installation Method	Composer

Summary

--stop-on-failure no longer stops on errors (as it happened on 9.x),

I don't know if this is intended (and has to be documented as a BC in the Changelog) or if it's a bug...

[garak]

Sorry to comment on a closed issue, but I see that this bug is supposed to be fixed in version 10.0.4, but I still get it:

bin/phpunit --stop-on-failure

PHPUnit 10.0.4 by Sebastian Bergmann and contributors.

Runtime:       PHP 8.1.1
Configuration: /user/myapp/phpunit.xml.dist

.....E.......................................................     61 / 61 (100%)

Time: 00:05.680, Memory: 72.50 MB

There was 1 error

[sebastianbergmann]

Quoting from the ChangeLog:

• The CLI test runner now only stops after a test errored when --stop-on-error or --stop-on-defect is used
• The CLI test runner now only stops after a test failed when --stop-on-failure or --stop-on-defect is used
• The CLI test runner now only stops after a test triggered a warning when --stop-on-warning or --stop-on-defect is used
• The CLI test runner now only stops after a test was marked as risky when --stop-on-risky or --stop-on-defect is used
• The CLI test runner now only stops after a test was skipped when --stop-on-skipped is used
• The CLI test runner now only stops after a test was marked as incomplete when --stop-on-incomplete is used

TL;DR: There is no bug here, this works as intended.

[federicotorrisi]

hello @sebastianbergmann ,
we are currently using v10.0.19 and phpunit is still not stopping using the flag --stop-on-defect when a warning is raised.

we also tried with dev-main 3a680a9 but still no luck.
any ideas?

[sebastianbergmann]

hello @sebastianbergmann , we are currently using v10.0.19 and phpunit is still not stopping using the flag --stop-on-defect when a warning is raised.

we also tried with dev-main 3a680a9 but still no luck. any ideas?

Please open a new ticket and provide a minimal, self-contained, reproducing test case that shows the problem you are reporting. Without such a minimal, self-contained, reproducing test case I will not be able to investigate this issue.

According to this test, --stop-on-defect correctly aborts test execution after the first test that triggered a warning.

[Jean85]

Maybe the failOnWarning setting is involved in the issue? @federicotorrisi what's the value that you use on that?

[federicotorrisi]

@Jean85 failOnWarning is currently set to true