1650900 - Logs of "Per-test extra prefs will be set:" messages when xpcshell tests are run on a directory in non-verbose mode

Status: RESOLVED FIXED

(Testing :: XPCShell Harness, defect)

Description

[Kris Maglione [:kmag]]

When running multiple tests at once in parallel, the xpcshell harness is only meant to show TEST_START and TEST_END lines for each test it runs, except when it re-runs failed tests or is in verbose mode. Now, though, any tests with preferences specified in their manifest also generate multi-line "Per-test extra prefs will be set:" messages, which are not helpful.

Comment 1

[Geoff Brown [:gbrown]]

Probably log.debug would be better than log.info for that.

Comment 2

[Luca Greco [:rpl] [:luca] [:lgreco]]

Attachment: Bug 1650900 - Skip per-test extra prefs logs when xpcshell is not in verbose mode. r?gbrown

Comment 3

[Luca Greco [:rpl] [:luca] [:lgreco]]

(In reply to Kris Maglione [:kmag] from comment #0)

When running multiple tests at once in parallel, the xpcshell harness is only meant to show TEST_START and TEST_END lines for each test it runs, except when it re-runs failed tests or is in verbose mode. Now, though, any tests with preferences specified in their manifest also generate multi-line "Per-test extra prefs will be set:" messages, which are not helpful.

That's totally fair, sorry I missed to notice that sooner.

I've just attached a patch to avoid logging the extra prefs when mach xpcshell-test has not been executed in verbose mode.

Comment 4

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/9bb233375f3d
Skip per-test extra prefs logs when xpcshell is not in verbose mode. r=gbrown

Comment 5

[Razvan Maries]

Backed out for build bustages.

Push with failure: https://treeherder.mozilla.org/#/jobs?repo=autoland&searchStr=linux%2Cx64%2Ctsan%2Cbuild-linux64-tsan%2Fopt%2C%28bo%29&revision=6b8fa932019863bb7a192db9b6e74afc59c96e32&selectedTaskRun=Ac8MfV_9TSCtpznlszQx1A.0

Log: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=308744314&repo=autoland&lineNumber=61341

Backout: https://hg.mozilla.org/integration/autoland/rev/86a6d876a7a67b3abbd3a86e800a70877a2c2309

Comment 6

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1635227

Comment 7

[Luca Greco [:rpl] [:luca] [:lgreco]]

(In reply to Razvan Maries from comment #5)

Backed out for build bustages.

Push with failure: https://treeherder.mozilla.org/#/jobs?repo=autoland&searchStr=linux%2Cx64%2Ctsan%2Cbuild-linux64-tsan%2Fopt%2C%28bo%29&revision=6b8fa932019863bb7a192db9b6e74afc59c96e32&selectedTaskRun=Ac8MfV_9TSCtpznlszQx1A.0

Log: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=308744314&repo=autoland&lineNumber=61341

Backout: https://hg.mozilla.org/integration/autoland/rev/86a6d876a7a67b3abbd3a86e800a70877a2c2309

My bad, I forgot that I do also covered the logs emitted for the per-test prefs in the xpcshell's unit tests.

Patch updated to include the changes needed on the test side.

Comment 8

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/e9ed9b954aee
Skip per-test extra prefs logs when xpcshell is not in verbose mode. r=gbrown

Comment 9

[Atila Butkovits]

https://hg.mozilla.org/mozilla-central/rev/e9ed9b954aee

Comment 10

[BugBot [:suhaib / :marco/ :calixte]]

The patch landed in nightly and beta is affected.
:rpl, is this bug important enough to require an uplift?
If not please set status_beta to wontfix.

For more information, please visit auto_nag documentation.

Comment 11

[Luca Greco [:rpl] [:luca] [:lgreco]]

(In reply to Release mgmt bot [:sylvestre / :calixte / :marco for bugbug] from comment #10)

The patch landed in nightly and beta is affected.
:rpl, is this bug important enough to require an uplift?
If not please set status_beta to wontfix.

For more information, please visit auto_nag documentation.

It doesn't have any impact on the users, but it is also test only and so we may uplift it to beta to reduce the verbosity of the xpcshell logs on beta.

Geoff, what do you think? should we uplift the patch to beta?
(I haven't checked but it is likely that the patch could apply cleanly on beta as it is)

Comment 12

[Geoff Brown [:gbrown]]

I don't think it is that important - let's not bother.