asserts that there has been no change in the code that could cause this, since 2020. Oddlly, the Android/Java tests have suddenly started failing.
Attachments
3
Activity
Show:
CB robot August 10, 2021 at 6:44 PM
Build couchbase-lite-java-3.0.0-162 contains couchbase-lite-java-ee-root commit 14bf7a8 with commit message: Re-include tests failing because of
CB robot August 10, 2021 at 6:20 PM
Build couchbase-lite-android-3.0.0-161 contains couchbase-lite-java-ee-root commit 14bf7a8 with commit message: Re-include tests failing because of
Blake Meike August 5, 2021 at 5:26 PM
It seems to work. As I said in the PR, I cherry-picked the commit onto master and tested it out. I got a clean run with no errors, built for ARM-64 and deployed on Android 30.
And idea when this will get to master?
Jim Borden August 5, 2021 at 1:17 AM
Fix in the above noted PR
Jens Alfke July 30, 2021 at 11:59 PM
figured out it's due to a change in Android 11: they've added some sort of pointer tagging that makes the high byte of a pointer nonzero.
In response I've changed the way ValueSlot tags non-pointers.
Core is crashing on this assertion:
asserts that there has been no change in the code that could cause this, since 2020. Oddlly, the Android/Java tests have suddenly started failing.