Page MenuHomePhabricator

Merge #17469: test: Remove fragile assert_memory_usage_stable

Authored by jasonbcox on Thu, May 21, 17:49.



fac942ca57dce6cfa5655a3ac8664d6a051bc01f test: Remove fragile assert_memory_usage_stable (MarcoFalke)

Pull request description:

This test fails on arm64 and a fuzz tests seems inappropriate for the functional test suite anyway, so remove it.

Example failures:


ACKs for top commit:


Tree-SHA512: 3577e7ce5891d221cb798454589ba796ed0c06621a26351bb919c23bc6bb46aafcd0b11cb02bbfde64b74d67cb2950da44959a7ecdc436491a34e8b045c1ccf4

Backport of Core PR17469

I've skipped backporting Core PR14794 which attempts to fix the issue using a not-so-robust approach of
looking at env variables to determine if the memory usage threshold should be modified.

This fixes sanitizer failures on CI by removing the problem entirely.

Test Plan

ninja check-functional
and run ASAN on CI

Diff Detail

rABC Bitcoin ABC
Automatic diff as part of commit; lint not applicable.
Automatic diff as part of commit; unit tests not applicable.

Event Timeline

jasonbcox created this revision.Thu, May 21, 17:49
Herald added a reviewer: Restricted Project. · View Herald TranscriptThu, May 21, 17:49
jasonbcox requested review of this revision.Thu, May 21, 17:49
teamcity edited the summary of this revision. (Show Details)Thu, May 21, 17:49

[Bot Message]
One or more PR numbers were detected in the summary.
Links to those PRs have been inserted into the summary for reference.

jasonbcox planned changes to this revision.Thu, May 21, 17:50
jasonbcox added inline comments.Thu, May 21, 17:53
86 ↗(On Diff #20475)

This timeout gets bumped over various backports, so we might as well bump it to the value in this PR.

jasonbcox requested review of this revision.Thu, May 21, 17:53
nakihito accepted this revision.Thu, May 21, 18:21
nakihito added a subscriber: nakihito.
nakihito added inline comments.
86 ↗(On Diff #20475)
190 ↗(On Diff #20475)

This is different from PR17469 and the change isn't made in PR14794. The changes are made in PR14693 and PR14812, but since its being removed, its probably fine to skip them.

This revision is now accepted and ready to land.Thu, May 21, 18:21
jasonbcox added inline comments.Thu, May 21, 18:26
86 ↗(On Diff #20475)


190 ↗(On Diff #20475)

Apparently there were more than one attempts to fix this, which makes the argument to avoid replicating that process all the more strong.