Summary: There was a diagnostic assertion that no hazard pointers were held on pages discarded during an in-memory split, and that's not correct, so the assertion was removed.
===================================================
I've been running the repro case looking for the failure in WT-5043 with the new debugging. I was running 10 parallel jobs on ocelot-aws. I started this several days ago, so the tree is running develop as of changeset 359fb2736f.
Over the weekend, after 1100 iterations, one of them hit a hazard check assertion:
[1567955476:353235][27283:0x7fc818ff9700], t, file:WiredTigerLAS.wt, WT_CURSOR.update: __wt_hazard_check_assert, 406: hazard pointer reference to discarded object: (0x7fc80c06a810: session 0x7fc844b8bc30 name WT_CURSOR.update: __wt_row_search, line 437) [1567955476:353321][27283:0x7fc818ff9700], t, file:WiredTigerLAS.wt, WT_CURSOR.update: __split_parent, 840: __wt_hazard_check_assert(session, next _ref, false) [1567955476:353328][27283:0x7fc818ff9700], t, file:WiredTigerLAS.wt, WT_CURSOR.update: __wt_abort, 28: aborting WiredTiger library