8d94aa381d
rxrpc calls shouldn't hold refs on the sock struct. This was done so that
the socket wouldn't go away whilst the call was in progress, such that the
call could reach the socket's queues.
However, we can mark the socket as requiring an RCU release and rely on the
RCU read lock.
To make this work, we do:
(1) rxrpc_release_call() removes the call's call user ID. This is now
only called from socket operations and not from the call processor:
rxrpc_accept_call() / rxrpc_kernel_accept_call()
rxrpc_reject_call() / rxrpc_kernel_reject_call()
rxrpc_kernel_end_call()
rxrpc_release_calls_on_socket()
rxrpc_recvmsg()
Though it is also called in the cleanup path of
rxrpc_accept_incoming_call() before we assign a user ID.
(2) Pass the socket pointer into rxrpc_release_call() rather than getting
it from the call so that we can get rid of uninitialised calls.
(3) Fix call processor queueing to pass a ref to the work queue and to
release that ref at the end of the processor function (or to pass it
back to the work queue if we have to requeue).
(4) Skip out of the call processor function asap if the call is complete
and don't requeue it if the call is complete.
(5) Clean up the call immediately that the refcount reaches 0 rather than
trying to defer it. Actual deallocation is deferred to RCU, however.
(6) Don't hold socket refs for allocated calls.
(7) Use the RCU read lock when queueing a message on a socket and treat
the call's socket pointer according to RCU rules and check it for
NULL.
We also need to use the RCU read lock when viewing a call through
procfs.
(8) Transmit the final ACK/ABORT to a client call in rxrpc_release_call()
if this hasn't been done yet so that we can then disconnect the call.
Once the call is disconnected, it won't have any access to the
connection struct and the UDP socket for the call work processor to be
able to send the ACK. Terminal retransmission will be handled by the
connection processor.
(9) Release all calls immediately on the closing of a socket rather than
trying to defer this. Incomplete calls will be aborted.
The call refcount model is much simplified. Refs are held on the call by:
(1) A socket's user ID tree.
(2) A socket's incoming call secureq and acceptq.
(3) A kernel service that has a call in progress.
(4) A queued call work processor. We have to take care to put any call
that we failed to queue.
(5) sk_buffs on a socket's receive queue. A future patch will get rid of
this.
Whilst we're at it, we can do:
(1) Get rid of the RXRPC_CALL_EV_RELEASE event. Release is now done
entirely from the socket routines and never from the call's processor.
(2) Get rid of the RXRPC_CALL_DEAD state. Calls now end in the
RXRPC_CALL_COMPLETE state.
(3) Get rid of the rxrpc_call::destroyer work item. Calls are now torn
down when their refcount reaches 0 and then handed over to RCU for
final cleanup.
(4) Get rid of the rxrpc_call::deadspan timer. Calls are cleaned up
immediately they're finished with and don't hang around.
Post-completion retransmission is handled by the connection processor
once the call is disconnected.
(5) Get rid of the dead call expiry setting as there's no longer a timer
to set.
(6) rxrpc_destroy_all_calls() can just check that the call list is empty.
Signed-off-by: David Howells <dhowells@redhat.com>
209 lines
5.5 KiB
C
209 lines
5.5 KiB
C
/* ar-skbuff.c: socket buffer destruction handling
|
|
*
|
|
* Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
|
|
* Written by David Howells (dhowells@redhat.com)
|
|
*
|
|
* This program is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU General Public License
|
|
* as published by the Free Software Foundation; either version
|
|
* 2 of the License, or (at your option) any later version.
|
|
*/
|
|
|
|
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
|
|
|
|
#include <linux/module.h>
|
|
#include <linux/net.h>
|
|
#include <linux/skbuff.h>
|
|
#include <net/sock.h>
|
|
#include <net/af_rxrpc.h>
|
|
#include "ar-internal.h"
|
|
|
|
/*
|
|
* set up for the ACK at the end of the receive phase when we discard the final
|
|
* receive phase data packet
|
|
* - called with softirqs disabled
|
|
*/
|
|
static void rxrpc_request_final_ACK(struct rxrpc_call *call)
|
|
{
|
|
/* the call may be aborted before we have a chance to ACK it */
|
|
write_lock(&call->state_lock);
|
|
|
|
switch (call->state) {
|
|
case RXRPC_CALL_CLIENT_RECV_REPLY:
|
|
call->state = RXRPC_CALL_CLIENT_FINAL_ACK;
|
|
_debug("request final ACK");
|
|
|
|
set_bit(RXRPC_CALL_EV_ACK_FINAL, &call->events);
|
|
if (try_to_del_timer_sync(&call->ack_timer) >= 0)
|
|
rxrpc_queue_call(call);
|
|
break;
|
|
|
|
case RXRPC_CALL_SERVER_RECV_REQUEST:
|
|
call->state = RXRPC_CALL_SERVER_ACK_REQUEST;
|
|
default:
|
|
break;
|
|
}
|
|
|
|
write_unlock(&call->state_lock);
|
|
}
|
|
|
|
/*
|
|
* drop the bottom ACK off of the call ACK window and advance the window
|
|
*/
|
|
static void rxrpc_hard_ACK_data(struct rxrpc_call *call, struct sk_buff *skb)
|
|
{
|
|
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
|
|
int loop;
|
|
u32 seq;
|
|
|
|
spin_lock_bh(&call->lock);
|
|
|
|
_debug("hard ACK #%u", sp->hdr.seq);
|
|
|
|
for (loop = 0; loop < RXRPC_ACKR_WINDOW_ASZ; loop++) {
|
|
call->ackr_window[loop] >>= 1;
|
|
call->ackr_window[loop] |=
|
|
call->ackr_window[loop + 1] << (BITS_PER_LONG - 1);
|
|
}
|
|
|
|
seq = sp->hdr.seq;
|
|
ASSERTCMP(seq, ==, call->rx_data_eaten + 1);
|
|
call->rx_data_eaten = seq;
|
|
|
|
if (call->ackr_win_top < UINT_MAX)
|
|
call->ackr_win_top++;
|
|
|
|
ASSERTIFCMP(call->state <= RXRPC_CALL_COMPLETE,
|
|
call->rx_data_post, >=, call->rx_data_recv);
|
|
ASSERTIFCMP(call->state <= RXRPC_CALL_COMPLETE,
|
|
call->rx_data_recv, >=, call->rx_data_eaten);
|
|
|
|
if (sp->hdr.flags & RXRPC_LAST_PACKET) {
|
|
rxrpc_request_final_ACK(call);
|
|
} else if (atomic_dec_and_test(&call->ackr_not_idle) &&
|
|
test_and_clear_bit(RXRPC_CALL_TX_SOFT_ACK, &call->flags)) {
|
|
/* We previously soft-ACK'd some received packets that have now
|
|
* been consumed, so send a hard-ACK if no more packets are
|
|
* immediately forthcoming to allow the transmitter to free up
|
|
* its Tx bufferage.
|
|
*/
|
|
_debug("send Rx idle ACK");
|
|
__rxrpc_propose_ACK(call, RXRPC_ACK_IDLE,
|
|
skb->priority, sp->hdr.serial, false);
|
|
}
|
|
|
|
spin_unlock_bh(&call->lock);
|
|
}
|
|
|
|
/**
|
|
* rxrpc_kernel_data_consumed - Record consumption of data message
|
|
* @call: The call to which the message pertains.
|
|
* @skb: Message holding data
|
|
*
|
|
* Record the consumption of a data message and generate an ACK if appropriate.
|
|
* The call state is shifted if this was the final packet. The caller must be
|
|
* in process context with no spinlocks held.
|
|
*
|
|
* TODO: Actually generate the ACK here rather than punting this to the
|
|
* workqueue.
|
|
*/
|
|
void rxrpc_kernel_data_consumed(struct rxrpc_call *call, struct sk_buff *skb)
|
|
{
|
|
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
|
|
|
|
_enter("%d,%p{%u}", call->debug_id, skb, sp->hdr.seq);
|
|
|
|
ASSERTCMP(sp->call, ==, call);
|
|
ASSERTCMP(sp->hdr.type, ==, RXRPC_PACKET_TYPE_DATA);
|
|
|
|
/* TODO: Fix the sequence number tracking */
|
|
ASSERTCMP(sp->hdr.seq, >=, call->rx_data_recv);
|
|
ASSERTCMP(sp->hdr.seq, <=, call->rx_data_recv + 1);
|
|
ASSERTCMP(sp->hdr.seq, >, call->rx_data_eaten);
|
|
|
|
call->rx_data_recv = sp->hdr.seq;
|
|
rxrpc_hard_ACK_data(call, skb);
|
|
}
|
|
|
|
/*
|
|
* Destroy a packet that has an RxRPC control buffer
|
|
*/
|
|
void rxrpc_packet_destructor(struct sk_buff *skb)
|
|
{
|
|
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
|
|
struct rxrpc_call *call = sp->call;
|
|
|
|
_enter("%p{%p}", skb, call);
|
|
|
|
if (call) {
|
|
rxrpc_put_call_for_skb(call, skb);
|
|
sp->call = NULL;
|
|
}
|
|
|
|
if (skb->sk)
|
|
sock_rfree(skb);
|
|
_leave("");
|
|
}
|
|
|
|
/*
|
|
* Note the existence of a new-to-us socket buffer (allocated or dequeued).
|
|
*/
|
|
void rxrpc_new_skb(struct sk_buff *skb)
|
|
{
|
|
const void *here = __builtin_return_address(0);
|
|
int n = atomic_inc_return(&rxrpc_n_skbs);
|
|
trace_rxrpc_skb(skb, 0, atomic_read(&skb->users), n, here);
|
|
}
|
|
|
|
/*
|
|
* Note the re-emergence of a socket buffer from a queue or buffer.
|
|
*/
|
|
void rxrpc_see_skb(struct sk_buff *skb)
|
|
{
|
|
const void *here = __builtin_return_address(0);
|
|
if (skb) {
|
|
int n = atomic_read(&rxrpc_n_skbs);
|
|
trace_rxrpc_skb(skb, 1, atomic_read(&skb->users), n, here);
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Note the addition of a ref on a socket buffer.
|
|
*/
|
|
void rxrpc_get_skb(struct sk_buff *skb)
|
|
{
|
|
const void *here = __builtin_return_address(0);
|
|
int n = atomic_inc_return(&rxrpc_n_skbs);
|
|
trace_rxrpc_skb(skb, 2, atomic_read(&skb->users), n, here);
|
|
skb_get(skb);
|
|
}
|
|
|
|
/*
|
|
* Note the destruction of a socket buffer.
|
|
*/
|
|
void rxrpc_free_skb(struct sk_buff *skb)
|
|
{
|
|
const void *here = __builtin_return_address(0);
|
|
if (skb) {
|
|
int n;
|
|
CHECK_SLAB_OKAY(&skb->users);
|
|
n = atomic_dec_return(&rxrpc_n_skbs);
|
|
trace_rxrpc_skb(skb, 3, atomic_read(&skb->users), n, here);
|
|
kfree_skb(skb);
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Clear a queue of socket buffers.
|
|
*/
|
|
void rxrpc_purge_queue(struct sk_buff_head *list)
|
|
{
|
|
const void *here = __builtin_return_address(0);
|
|
struct sk_buff *skb;
|
|
while ((skb = skb_dequeue((list))) != NULL) {
|
|
int n = atomic_dec_return(&rxrpc_n_skbs);
|
|
trace_rxrpc_skb(skb, 4, atomic_read(&skb->users), n, here);
|
|
kfree_skb(skb);
|
|
}
|
|
}
|