This used to be a linear search O(n), but is now done
as a red-black tree O(log n) instead.
These operations can be further opimized with a hash-table
which would acheive near constant time lookups.
- Added a list_remove_first() which is generally better than list_remove()
provided you want to remove the first element.
- Added a list_append_list() to append and move all nodes from one list to
another.
Previously you would have to do something like this:
for (type foo = (type) list_get_first(list); foo; foo = (type) list_get_next(list)
{
/* code */
}
Now, you can instead write this as:
LIST_FOREACH(type, foo, list,
{
/* code */
})
Basically, boilerplate stuff including the casting is gone.
uhub did not have this security bug since the hub did not advertise support for the
UCMD extension, but the message was still correctly relayed as specified in the
protocol specification.
However, this commit adds support for the UCMD extension, but only to the extent
that uhub will advertise it and uhub will also drop any such CMD message
generated by a client and will (currently) never issues a CMD message by itself.
GnuTLS sends a handshake with SSL 3.0 (0x0300) in the outer packet, but
mentions TLS 1.2 (0x0303) in the Client Hello. There's no real need for
uhub to validate these fields, as OpenSSL should do that itself already.
Just use the version mentioned in Client Hello for logging output.
It's not called anywhere yet.
Also reorder some typedefs, rename the ip check functions and add
struct {hub,plugin}_user parameter to on_check_ip_late(). Not sure where
to insert a call to that...
A struct plugin_hub_internals was falsely casted to struct
plugin_callback_data. This caused the contained commands list pointer to point to
a struct hub_info and commands->size took the value of a pointer to a struct
net_connection. Since size is increased/decreased every time an item is
added to/removed from the list, this resulted in some funny crashes.
This fix is a little dirty as it exports some internals.
