46
46
copies of tree pages are unshared. Postgres shares in-memory buffers
47
47
among backends. As a result, we do page-level read locking on btree
48
48
pages in order to guarantee that no record is modified while we are
50
50
behavior. An advantage is that when trading in a read lock for a
51
51
write lock, we need not re-read the page after getting the write lock.
52
52
Since we're also holding a pin on the shared buffer containing the