src-test/sys/net/raw_cb.c, branch stable/7

Merge r180391 from head to stable/7:

2008-08-31T16:14:41Z

Rather than checking for a NULL so_pcb in raw_attach(), assert that it's non-NULL, as all callers can and should already do the required checking. Update comments a bit more to talk about rawcb allocation for consumers. Reviewed by: bz

Merge r180390 from head to stable/7:

2008-08-31T16:02:54Z

Add sysctl subtree net.raw for generic raw socket infrastructure; expose default send and receive socket buffer sizes using sysctls so that they can be administered centrally. Reviewed by: bz

Merge r180385 from head to stable/7:

2008-08-31T15:55:59Z

Remove unused support for local and foreign addresses in generic raw socket support. These utility routines are used only for routing and pfkey sockets, neither of which have a notion of address, so were required to mock up fake socket addresses to avoid connection requirements for applications that did not specify their own fake addresses (most of them). Quite a bit of the removed code is #ifdef notdef, since raw sockets don't support bind() or connect() in practice. Removing this simplifies the raw socket implementation, and removes two (commented out) uses of dtom(9). Fake addresses passed to sendto(2) by applications are ignored for compatibility reasons, but this is now done in a more consistent way (and with a comment). Possibly, EINVAL could be returned here in the future if it is determined that no applications depend on the semantic inconsistency of specifying a destination address for a protocol without address support, but this will require some amount of careful surveying. NB: This does not affect netinet, netinet6, or other wire protocol raw sockets, which provide their own independent infrastructure with control block address support specific to the protocol. Reviewed by: bz

Merge r180305 from head to stable/7:

2008-08-31T15:39:06Z

Improve approximation of style(9) in raw socket code.

raw_disconnect() now disconnects but does not detach the raw pcb. As a

2006-06-02T08:27:15Z

result, raw_uabort() now needs to call raw_detach() directly. As raw_uabort() is never called, and raw_disconnect() is probably not ever actually called in practice, this is likely not a functional change, but improves congruence between protocols, and avoids a NULL raw cb pointer after disconnect, which could result in a panic. MFC after: 1 month

Chance protocol switch method pru_detach() so that it returns void

2006-04-01T15:42:02Z

rather than an error. Detaches do not "fail", they other occur or the protocol flags SS_PROTOREF to take ownership of the socket. soclose() no longer looks at so_pcb to see if it's NULL, relying entirely on the protocol to decide whether it's time to free the socket or not using SS_PROTOREF. so_pcb is now entirely owned and managed by the protocol code. Likewise, no longer test so_pcb in other socket functions, such as soreceive(), which have no business digging into protocol internals. Protocol detach routines no longer try to free the socket on detach, this is performed in the socket code if the protocol permits it. In rts_detach(), no longer test for rp != NULL in detach, and likewise in other protocols that don't permit a NULL so_pcb, reduce the incidence of testing for it during detach. netinet and netinet6 are not fully updated to this change, which will be in an upcoming commit. In their current state they may leak memory or panic. MFC after: 3 months

Acquire the raw_cb mutex around LIST_REMOVE() of a raw socket control

2005-01-24T22:56:09Z

block from the global raw socket list. Submitted by: Roselyn Lee MFC after: 1 week

/* -> /*- for license, minor formatting changes

2005-01-07T01:45:51Z

Push acquisition of the accept mutex out of sofree() into the caller

2004-10-18T22:19:43Z

(sorele()/sotryfree()): - This permits the caller to acquire the accept mutex before the socket mutex, avoiding sofree() having to drop the socket mutex and re-order, which could lead to races permitting more than one thread to enter sofree() after a socket is ready to be free'd. - This also covers clearing of the so_pcb weak socket reference from the protocol to the socket, preventing races in clearing and evaluation of the reference such that sofree() might be called more than once on the same socket. This appears to close a race I was able to easily trigger by repeatedly opening and resetting TCP connections to a host, in which the tcp_close() code called as a result of the RST raced with the close() of the accepted socket in the user process resulting in simultaneous attempts to de-allocate the same socket. The new locking increases the overhead for operations that may potentially free the socket, so we will want to revise the synchronization strategy here as we normalize the reference counting model for sockets. The use of the accept mutex in freeing of sockets that are not listen sockets is primarily motivated by the potential need to remove the socket from the incomplete connection queue on its parent (listen) socket, so cleaning up the reference model here may allow us to substantially weaken the synchronization requirements. RELENG_5_3 candidate. MFC after: 3 days Reviewed by: dwhite Discussed with: gnn, dwhite, green Reported by: Marc UBM Bocklet Reported by: Vlad

Lock down rawcb_list, a global list of control blocks for raw sockets,

2004-06-15T04:13:59Z

using rawcb_mtx. Hold this mutex while modifying or iterating over the control list; this means that the mutex is held over calls into socket delivery code, which no longer causes a lock order reversal as the routing socket code uses a netisr to avoid recursing socket -> routing -> socket. Note: Locking of IPsec consumers of rawcb_list is not included in this commit.