Filed by: rvandermeulen [at] mozilla.com
Parsed log: https://treeherder.mozilla.org/logviewer?job_id=423591499&repo=try
Full log: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/VcEYoZgDTSKi1b56wQTUug/runs/0/artifacts/public/logs/live_backing.log

I'm fairly sure what's happening is that we do not correctly use up all the allocations. Most are used but some are freed but their reuse delay has not yet expired so PHC refuses to allocate them during the initial part of the test.

Then on line 361 we continue to hammer PHC and a page's reuse delay ends and it successfully allocates a page and our test fails.

This is benign, but an annoying intermittent. To fix it we will want to be able to set PHC's parameters at runtime and decrease the reuse delay during this test.

I asked noemi_erli about this, they found that there's another crash that might suggest this test is broken anyway, or at least not always working.

https://treeherder.mozilla.org/logviewer?job_id=423880091&repo=autoland&lineNumber=2029

For now I'll disable the test so it doesn't bother anyone and fix it properly later.

Backed out for causing build bustages in TestPHC.cpp

• Backout link
• Push with failures
• Failure Log
• Failure line: /builds/worker/checkouts/gecko/memory/replace/phc/test/gtest/TestPHC.cpp(341,36): error: code will never be executed [-Werror,-Wunreachable-code]

https://hg.mozilla.org/mozilla-central/rev/5c2d20e990ab