The existing AF_UNIX socket garbage collector destroys any socket
which may potentially be in a cycle, as indicated by its file reference
count being equal to its enqueue count. However, this can produce false
positives for in-flight sockets which aren't part of a cycle but are
part of one or more SCM_RIGHTS mssages and which have been closed
on the sending side. If the garbage collector happens to run at
exactly the wrong time, destruction of these sockets will render them
unusable on the receiving side, such that no previously-written data
may be read.
This change rewrites the garbage collector to precisely detect cycles:
- The existing check of msgcount==f_count is still used to determine whether the socket is potentially in a cycle.
- The socket is now placed on a local "dead list", which is used to reduce iteration time (and therefore contention on the global unp_link_rwlock).
- The first pass through the dead list removes each potentially-dead socket's outgoing references from the socket graph, using a gc-specific copy of the original reference count.
- The second series of passes through the dead list removes from the list any socket whose GC refcount is non-zero, as these sockets remain accessible outside of any possible cycle. Iteration is repeated until no further sockets are removed from the dead list.
- Sockets remaining in the dead list are destroyed as before.
Submitted by: firstname.lastname@example.org (prior version)