If jemalloc tries to free an invalid pointer but one that is NOT a PHC pointer the address calculations performed by jemalloc could cause jemalloc to dereference a PHC pointer, adding PHC stacks to a crash report that could be a red herring.

Before freeing a pointer jemalloc could record in thread local storage the pointer it's about to free, then clear it when done. Then if a crash occurs the crash reporter could ask jemalloc if it was freeing a pointer, if it says "yes this one" then that's the one we want to look up, not the crash address.

There will be trade-offs, but I haven't thought very much about them yet, so this may not be a great idea, it's "to consider".

Depends on D180195

Backed out for causing failures at TestPrintf.

Backout link: https://hg.mozilla.org/integration/autoland/rev/692aa2b9cb09a4b6848708eb27d09c919bb6823b

Push with failures: https://treeherder.mozilla.org/jobs?repo=autoland&selectedTaskRun=F2mMfY_BSKGfiPr9srQUSg.0&resultStatus=testfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel&revision=301d3e2bdac376f35bb2245b05ebe9da010f1733

Failure log:
https://treeherder.mozilla.org/logviewer?job_id=419284410&repo=autoland&lineNumber=1851
There are also these Android exceptions: https://treeherder.mozilla.org/jobs?repo=autoland&resultStatus=testfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel&revision=301d3e2bdac376f35bb2245b05ebe9da010f1733&searchStr=android&selectedTaskRun=P4FG8N6iQd6umVzB8DQ7Pw.5

• xpcshell failures log: https://treeherder.mozilla.org/logviewer?job_id=419296996&repo=autoland
• bc failures log: https://treeherder.mozilla.org/logviewer?job_id=419297084&repo=autoland

There are some r+ patches which didn't land and no activity in this bug for 2 weeks.
:pbone, could you have a look please?
If you still have some work to do, you can add an action "Plan Changes" in Phabricator.
For more information, please visit BugBot documentation.