Details
-
Task
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
None
Description
Basically mem_heap_t should poison all not yet allocated and all freed memory. Opposite for MEM_ROOT it's not instrumented at all.
Looks like it's actually implemented in 5.5. But not in 10.0+ due to merge error. I consider it a result of macro complications and will fix in 10.0 by intruducing MY_ASAN_POISON_MEMORY_REGION() macro.