Re:22-Year-Old PostgreSQL Bug Turned into Remote Code Execution
by amluto on Friday November 13, 2026 @09:27AM
(#50483995)
Unless the OP is completely wrong, it’s not. The author was able to fairly straightforwardly prepare a statement (thus selecting a type id), execute another statement (changing the meaning of that type id), and then execute the prepared statement (uses the old type id with the new semantics, which had a different idea of whether a particular field was a pointer).
[ Parent
| on HN ]
There are no comments on this comment yet.