linux/fs/afs/rxrpc.c
David Howells 372ee16386 rxrpc: Fix races between skb free, ACK generation and replying
Inside the kafs filesystem it is possible to occasionally have a call
processed and terminated before we've had a chance to check whether we need
to clean up the rx queue for that call because afs_send_simple_reply() ends
the call when it is done, but this is done in a workqueue item that might
happen to run to completion before afs_deliver_to_call() completes.

Further, it is possible for rxrpc_kernel_send_data() to be called to send a
reply before the last request-phase data skb is released.  The rxrpc skb
destructor is where the ACK processing is done and the call state is
advanced upon release of the last skb.  ACK generation is also deferred to
a work item because it's possible that the skb destructor is not called in
a context where kernel_sendmsg() can be invoked.

To this end, the following changes are made:

 (1) kernel_rxrpc_data_consumed() is added.  This should be called whenever
     an skb is emptied so as to crank the ACK and call states.  This does
     not release the skb, however.  kernel_rxrpc_free_skb() must now be
     called to achieve that.  These together replace
     rxrpc_kernel_data_delivered().

 (2) kernel_rxrpc_data_consumed() is wrapped by afs_data_consumed().

     This makes afs_deliver_to_call() easier to work as the skb can simply
     be discarded unconditionally here without trying to work out what the
     return value of the ->deliver() function means.

     The ->deliver() functions can, via afs_data_complete(),
     afs_transfer_reply() and afs_extract_data() mark that an skb has been
     consumed (thereby cranking the state) without the need to
     conditionally free the skb to make sure the state is correct on an
     incoming call for when the call processor tries to send the reply.

 (3) rxrpc_recvmsg() now has to call kernel_rxrpc_data_consumed() when it
     has finished with a packet and MSG_PEEK isn't set.

 (4) rxrpc_packet_destructor() no longer calls rxrpc_hard_ACK_data().

     Because of this, we no longer need to clear the destructor and put the
     call before we free the skb in cases where we don't want the ACK/call
     state to be cranked.

 (5) The ->deliver() call-type callbacks are made to return -EAGAIN rather
     than 0 if they expect more data (afs_extract_data() returns -EAGAIN to
     the delivery function already), and the caller is now responsible for
     producing an abort if that was the last packet.

 (6) There are many bits of unmarshalling code where:

 		ret = afs_extract_data(call, skb, last, ...);
		switch (ret) {
		case 0:		break;
		case -EAGAIN:	return 0;
		default:	return ret;
		}

     is to be found.  As -EAGAIN can now be passed back to the caller, we
     now just return if ret < 0:

 		ret = afs_extract_data(call, skb, last, ...);
		if (ret < 0)
			return ret;

 (7) Checks for trailing data and empty final data packets has been
     consolidated as afs_data_complete().  So:

		if (skb->len > 0)
			return -EBADMSG;
		if (!last)
			return 0;

     becomes:

		ret = afs_data_complete(call, skb, last);
		if (ret < 0)
			return ret;

 (8) afs_transfer_reply() now checks the amount of data it has against the
     amount of data desired and the amount of data in the skb and returns
     an error to induce an abort if we don't get exactly what we want.

Without these changes, the following oops can occasionally be observed,
particularly if some printks are inserted into the delivery path:

general protection fault: 0000 [#1] SMP
Modules linked in: kafs(E) af_rxrpc(E) [last unloaded: af_rxrpc]
CPU: 0 PID: 1305 Comm: kworker/u8:3 Tainted: G            E   4.7.0-fsdevel+ #1303
Hardware name: ASUS All Series/H97-PLUS, BIOS 2306 10/09/2014
Workqueue: kafsd afs_async_workfn [kafs]
task: ffff88040be041c0 ti: ffff88040c070000 task.ti: ffff88040c070000
RIP: 0010:[<ffffffff8108fd3c>]  [<ffffffff8108fd3c>] __lock_acquire+0xcf/0x15a1
RSP: 0018:ffff88040c073bc0  EFLAGS: 00010002
RAX: 6b6b6b6b6b6b6b6b RBX: 0000000000000000 RCX: ffff88040d29a710
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88040d29a710
RBP: ffff88040c073c70 R08: 0000000000000001 R09: 0000000000000001
R10: 0000000000000001 R11: 0000000000000000 R12: 0000000000000000
R13: 0000000000000000 R14: ffff88040be041c0 R15: ffffffff814c928f
FS:  0000000000000000(0000) GS:ffff88041fa00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fa4595f4750 CR3: 0000000001c14000 CR4: 00000000001406f0
Stack:
 0000000000000006 000000000be04930 0000000000000000 ffff880400000000
 ffff880400000000 ffffffff8108f847 ffff88040be041c0 ffffffff81050446
 ffff8803fc08a920 ffff8803fc08a958 ffff88040be041c0 ffff88040c073c38
Call Trace:
 [<ffffffff8108f847>] ? mark_held_locks+0x5e/0x74
 [<ffffffff81050446>] ? __local_bh_enable_ip+0x9b/0xa1
 [<ffffffff8108f9ca>] ? trace_hardirqs_on_caller+0x16d/0x189
 [<ffffffff810915f4>] lock_acquire+0x122/0x1b6
 [<ffffffff810915f4>] ? lock_acquire+0x122/0x1b6
 [<ffffffff814c928f>] ? skb_dequeue+0x18/0x61
 [<ffffffff81609dbf>] _raw_spin_lock_irqsave+0x35/0x49
 [<ffffffff814c928f>] ? skb_dequeue+0x18/0x61
 [<ffffffff814c928f>] skb_dequeue+0x18/0x61
 [<ffffffffa009aa92>] afs_deliver_to_call+0x344/0x39d [kafs]
 [<ffffffffa009ab37>] afs_process_async_call+0x4c/0xd5 [kafs]
 [<ffffffffa0099e9c>] afs_async_workfn+0xe/0x10 [kafs]
 [<ffffffff81063a3a>] process_one_work+0x29d/0x57c
 [<ffffffff81064ac2>] worker_thread+0x24a/0x385
 [<ffffffff81064878>] ? rescuer_thread+0x2d0/0x2d0
 [<ffffffff810696f5>] kthread+0xf3/0xfb
 [<ffffffff8160a6ff>] ret_from_fork+0x1f/0x40
 [<ffffffff81069602>] ? kthread_create_on_node+0x1cf/0x1cf

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-08-06 00:08:40 -04:00

902 lines
21 KiB
C

/* Maintain an RxRPC server socket to do AFS communications through
*
* 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.
*/
#include <linux/slab.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include <rxrpc/packet.h>
#include "internal.h"
#include "afs_cm.h"
static struct socket *afs_socket; /* my RxRPC socket */
static struct workqueue_struct *afs_async_calls;
static atomic_t afs_outstanding_calls;
static atomic_t afs_outstanding_skbs;
static void afs_wake_up_call_waiter(struct afs_call *);
static int afs_wait_for_call_to_complete(struct afs_call *);
static void afs_wake_up_async_call(struct afs_call *);
static int afs_dont_wait_for_call_to_complete(struct afs_call *);
static void afs_process_async_call(struct afs_call *);
static void afs_rx_interceptor(struct sock *, unsigned long, struct sk_buff *);
static int afs_deliver_cm_op_id(struct afs_call *, struct sk_buff *, bool);
/* synchronous call management */
const struct afs_wait_mode afs_sync_call = {
.rx_wakeup = afs_wake_up_call_waiter,
.wait = afs_wait_for_call_to_complete,
};
/* asynchronous call management */
const struct afs_wait_mode afs_async_call = {
.rx_wakeup = afs_wake_up_async_call,
.wait = afs_dont_wait_for_call_to_complete,
};
/* asynchronous incoming call management */
static const struct afs_wait_mode afs_async_incoming_call = {
.rx_wakeup = afs_wake_up_async_call,
};
/* asynchronous incoming call initial processing */
static const struct afs_call_type afs_RXCMxxxx = {
.name = "CB.xxxx",
.deliver = afs_deliver_cm_op_id,
.abort_to_error = afs_abort_to_error,
};
static void afs_collect_incoming_call(struct work_struct *);
static struct sk_buff_head afs_incoming_calls;
static DECLARE_WORK(afs_collect_incoming_call_work, afs_collect_incoming_call);
static void afs_async_workfn(struct work_struct *work)
{
struct afs_call *call = container_of(work, struct afs_call, async_work);
call->async_workfn(call);
}
static int afs_wait_atomic_t(atomic_t *p)
{
schedule();
return 0;
}
/*
* open an RxRPC socket and bind it to be a server for callback notifications
* - the socket is left in blocking mode and non-blocking ops use MSG_DONTWAIT
*/
int afs_open_socket(void)
{
struct sockaddr_rxrpc srx;
struct socket *socket;
int ret;
_enter("");
skb_queue_head_init(&afs_incoming_calls);
ret = -ENOMEM;
afs_async_calls = create_singlethread_workqueue("kafsd");
if (!afs_async_calls)
goto error_0;
ret = sock_create_kern(&init_net, AF_RXRPC, SOCK_DGRAM, PF_INET, &socket);
if (ret < 0)
goto error_1;
socket->sk->sk_allocation = GFP_NOFS;
/* bind the callback manager's address to make this a server socket */
srx.srx_family = AF_RXRPC;
srx.srx_service = CM_SERVICE;
srx.transport_type = SOCK_DGRAM;
srx.transport_len = sizeof(srx.transport.sin);
srx.transport.sin.sin_family = AF_INET;
srx.transport.sin.sin_port = htons(AFS_CM_PORT);
memset(&srx.transport.sin.sin_addr, 0,
sizeof(srx.transport.sin.sin_addr));
ret = kernel_bind(socket, (struct sockaddr *) &srx, sizeof(srx));
if (ret < 0)
goto error_2;
ret = kernel_listen(socket, INT_MAX);
if (ret < 0)
goto error_2;
rxrpc_kernel_intercept_rx_messages(socket, afs_rx_interceptor);
afs_socket = socket;
_leave(" = 0");
return 0;
error_2:
sock_release(socket);
error_1:
destroy_workqueue(afs_async_calls);
error_0:
_leave(" = %d", ret);
return ret;
}
/*
* close the RxRPC socket AFS was using
*/
void afs_close_socket(void)
{
_enter("");
wait_on_atomic_t(&afs_outstanding_calls, afs_wait_atomic_t,
TASK_UNINTERRUPTIBLE);
_debug("no outstanding calls");
sock_release(afs_socket);
_debug("dework");
destroy_workqueue(afs_async_calls);
ASSERTCMP(atomic_read(&afs_outstanding_skbs), ==, 0);
_leave("");
}
/*
* Note that the data in a socket buffer is now consumed.
*/
void afs_data_consumed(struct afs_call *call, struct sk_buff *skb)
{
if (!skb) {
_debug("DLVR NULL [%d]", atomic_read(&afs_outstanding_skbs));
dump_stack();
} else {
_debug("DLVR %p{%u} [%d]",
skb, skb->mark, atomic_read(&afs_outstanding_skbs));
rxrpc_kernel_data_consumed(call->rxcall, skb);
}
}
/*
* free a socket buffer
*/
static void afs_free_skb(struct sk_buff *skb)
{
if (!skb) {
_debug("FREE NULL [%d]", atomic_read(&afs_outstanding_skbs));
dump_stack();
} else {
_debug("FREE %p{%u} [%d]",
skb, skb->mark, atomic_read(&afs_outstanding_skbs));
if (atomic_dec_return(&afs_outstanding_skbs) == -1)
BUG();
rxrpc_kernel_free_skb(skb);
}
}
/*
* free a call
*/
static void afs_free_call(struct afs_call *call)
{
_debug("DONE %p{%s} [%d]",
call, call->type->name, atomic_read(&afs_outstanding_calls));
ASSERTCMP(call->rxcall, ==, NULL);
ASSERT(!work_pending(&call->async_work));
ASSERT(skb_queue_empty(&call->rx_queue));
ASSERT(call->type->name != NULL);
kfree(call->request);
kfree(call);
if (atomic_dec_and_test(&afs_outstanding_calls))
wake_up_atomic_t(&afs_outstanding_calls);
}
/*
* End a call but do not free it
*/
static void afs_end_call_nofree(struct afs_call *call)
{
if (call->rxcall) {
rxrpc_kernel_end_call(call->rxcall);
call->rxcall = NULL;
}
if (call->type->destructor)
call->type->destructor(call);
}
/*
* End a call and free it
*/
static void afs_end_call(struct afs_call *call)
{
afs_end_call_nofree(call);
afs_free_call(call);
}
/*
* allocate a call with flat request and reply buffers
*/
struct afs_call *afs_alloc_flat_call(const struct afs_call_type *type,
size_t request_size, size_t reply_size)
{
struct afs_call *call;
call = kzalloc(sizeof(*call), GFP_NOFS);
if (!call)
goto nomem_call;
_debug("CALL %p{%s} [%d]",
call, type->name, atomic_read(&afs_outstanding_calls));
atomic_inc(&afs_outstanding_calls);
call->type = type;
call->request_size = request_size;
call->reply_max = reply_size;
if (request_size) {
call->request = kmalloc(request_size, GFP_NOFS);
if (!call->request)
goto nomem_free;
}
if (reply_size) {
call->buffer = kmalloc(reply_size, GFP_NOFS);
if (!call->buffer)
goto nomem_free;
}
init_waitqueue_head(&call->waitq);
skb_queue_head_init(&call->rx_queue);
return call;
nomem_free:
afs_free_call(call);
nomem_call:
return NULL;
}
/*
* clean up a call with flat buffer
*/
void afs_flat_call_destructor(struct afs_call *call)
{
_enter("");
kfree(call->request);
call->request = NULL;
kfree(call->buffer);
call->buffer = NULL;
}
/*
* attach the data from a bunch of pages on an inode to a call
*/
static int afs_send_pages(struct afs_call *call, struct msghdr *msg,
struct kvec *iov)
{
struct page *pages[8];
unsigned count, n, loop, offset, to;
pgoff_t first = call->first, last = call->last;
int ret;
_enter("");
offset = call->first_offset;
call->first_offset = 0;
do {
_debug("attach %lx-%lx", first, last);
count = last - first + 1;
if (count > ARRAY_SIZE(pages))
count = ARRAY_SIZE(pages);
n = find_get_pages_contig(call->mapping, first, count, pages);
ASSERTCMP(n, ==, count);
loop = 0;
do {
msg->msg_flags = 0;
to = PAGE_SIZE;
if (first + loop >= last)
to = call->last_to;
else
msg->msg_flags = MSG_MORE;
iov->iov_base = kmap(pages[loop]) + offset;
iov->iov_len = to - offset;
offset = 0;
_debug("- range %u-%u%s",
offset, to, msg->msg_flags ? " [more]" : "");
iov_iter_kvec(&msg->msg_iter, WRITE | ITER_KVEC,
iov, 1, to - offset);
/* have to change the state *before* sending the last
* packet as RxRPC might give us the reply before it
* returns from sending the request */
if (first + loop >= last)
call->state = AFS_CALL_AWAIT_REPLY;
ret = rxrpc_kernel_send_data(call->rxcall, msg,
to - offset);
kunmap(pages[loop]);
if (ret < 0)
break;
} while (++loop < count);
first += count;
for (loop = 0; loop < count; loop++)
put_page(pages[loop]);
if (ret < 0)
break;
} while (first <= last);
_leave(" = %d", ret);
return ret;
}
/*
* initiate a call
*/
int afs_make_call(struct in_addr *addr, struct afs_call *call, gfp_t gfp,
const struct afs_wait_mode *wait_mode)
{
struct sockaddr_rxrpc srx;
struct rxrpc_call *rxcall;
struct msghdr msg;
struct kvec iov[1];
int ret;
struct sk_buff *skb;
_enter("%x,{%d},", addr->s_addr, ntohs(call->port));
ASSERT(call->type != NULL);
ASSERT(call->type->name != NULL);
_debug("____MAKE %p{%s,%x} [%d]____",
call, call->type->name, key_serial(call->key),
atomic_read(&afs_outstanding_calls));
call->wait_mode = wait_mode;
call->async_workfn = afs_process_async_call;
INIT_WORK(&call->async_work, afs_async_workfn);
memset(&srx, 0, sizeof(srx));
srx.srx_family = AF_RXRPC;
srx.srx_service = call->service_id;
srx.transport_type = SOCK_DGRAM;
srx.transport_len = sizeof(srx.transport.sin);
srx.transport.sin.sin_family = AF_INET;
srx.transport.sin.sin_port = call->port;
memcpy(&srx.transport.sin.sin_addr, addr, 4);
/* create a call */
rxcall = rxrpc_kernel_begin_call(afs_socket, &srx, call->key,
(unsigned long) call, gfp);
call->key = NULL;
if (IS_ERR(rxcall)) {
ret = PTR_ERR(rxcall);
goto error_kill_call;
}
call->rxcall = rxcall;
/* send the request */
iov[0].iov_base = call->request;
iov[0].iov_len = call->request_size;
msg.msg_name = NULL;
msg.msg_namelen = 0;
iov_iter_kvec(&msg.msg_iter, WRITE | ITER_KVEC, iov, 1,
call->request_size);
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = (call->send_pages ? MSG_MORE : 0);
/* have to change the state *before* sending the last packet as RxRPC
* might give us the reply before it returns from sending the
* request */
if (!call->send_pages)
call->state = AFS_CALL_AWAIT_REPLY;
ret = rxrpc_kernel_send_data(rxcall, &msg, call->request_size);
if (ret < 0)
goto error_do_abort;
if (call->send_pages) {
ret = afs_send_pages(call, &msg, iov);
if (ret < 0)
goto error_do_abort;
}
/* at this point, an async call may no longer exist as it may have
* already completed */
return wait_mode->wait(call);
error_do_abort:
rxrpc_kernel_abort_call(rxcall, RX_USER_ABORT);
while ((skb = skb_dequeue(&call->rx_queue)))
afs_free_skb(skb);
error_kill_call:
afs_end_call(call);
_leave(" = %d", ret);
return ret;
}
/*
* Handles intercepted messages that were arriving in the socket's Rx queue.
*
* Called from the AF_RXRPC call processor in waitqueue process context. For
* each call, it is guaranteed this will be called in order of packet to be
* delivered.
*/
static void afs_rx_interceptor(struct sock *sk, unsigned long user_call_ID,
struct sk_buff *skb)
{
struct afs_call *call = (struct afs_call *) user_call_ID;
_enter("%p,,%u", call, skb->mark);
_debug("ICPT %p{%u} [%d]",
skb, skb->mark, atomic_read(&afs_outstanding_skbs));
ASSERTCMP(sk, ==, afs_socket->sk);
atomic_inc(&afs_outstanding_skbs);
if (!call) {
/* its an incoming call for our callback service */
skb_queue_tail(&afs_incoming_calls, skb);
queue_work(afs_wq, &afs_collect_incoming_call_work);
} else {
/* route the messages directly to the appropriate call */
skb_queue_tail(&call->rx_queue, skb);
call->wait_mode->rx_wakeup(call);
}
_leave("");
}
/*
* deliver messages to a call
*/
static void afs_deliver_to_call(struct afs_call *call)
{
struct sk_buff *skb;
bool last;
u32 abort_code;
int ret;
_enter("");
while ((call->state == AFS_CALL_AWAIT_REPLY ||
call->state == AFS_CALL_AWAIT_OP_ID ||
call->state == AFS_CALL_AWAIT_REQUEST ||
call->state == AFS_CALL_AWAIT_ACK) &&
(skb = skb_dequeue(&call->rx_queue))) {
switch (skb->mark) {
case RXRPC_SKB_MARK_DATA:
_debug("Rcv DATA");
last = rxrpc_kernel_is_data_last(skb);
ret = call->type->deliver(call, skb, last);
switch (ret) {
case -EAGAIN:
if (last) {
_debug("short data");
goto unmarshal_error;
}
break;
case 0:
ASSERT(last);
if (call->state == AFS_CALL_AWAIT_REPLY)
call->state = AFS_CALL_COMPLETE;
break;
case -ENOTCONN:
abort_code = RX_CALL_DEAD;
goto do_abort;
case -ENOTSUPP:
abort_code = RX_INVALID_OPERATION;
goto do_abort;
default:
unmarshal_error:
abort_code = RXGEN_CC_UNMARSHAL;
if (call->state != AFS_CALL_AWAIT_REPLY)
abort_code = RXGEN_SS_UNMARSHAL;
do_abort:
rxrpc_kernel_abort_call(call->rxcall,
abort_code);
call->error = ret;
call->state = AFS_CALL_ERROR;
break;
}
break;
case RXRPC_SKB_MARK_FINAL_ACK:
_debug("Rcv ACK");
call->state = AFS_CALL_COMPLETE;
break;
case RXRPC_SKB_MARK_BUSY:
_debug("Rcv BUSY");
call->error = -EBUSY;
call->state = AFS_CALL_BUSY;
break;
case RXRPC_SKB_MARK_REMOTE_ABORT:
abort_code = rxrpc_kernel_get_abort_code(skb);
call->error = call->type->abort_to_error(abort_code);
call->state = AFS_CALL_ABORTED;
_debug("Rcv ABORT %u -> %d", abort_code, call->error);
break;
case RXRPC_SKB_MARK_LOCAL_ABORT:
abort_code = rxrpc_kernel_get_abort_code(skb);
call->error = call->type->abort_to_error(abort_code);
call->state = AFS_CALL_ABORTED;
_debug("Loc ABORT %u -> %d", abort_code, call->error);
break;
case RXRPC_SKB_MARK_NET_ERROR:
call->error = -rxrpc_kernel_get_error_number(skb);
call->state = AFS_CALL_ERROR;
_debug("Rcv NET ERROR %d", call->error);
break;
case RXRPC_SKB_MARK_LOCAL_ERROR:
call->error = -rxrpc_kernel_get_error_number(skb);
call->state = AFS_CALL_ERROR;
_debug("Rcv LOCAL ERROR %d", call->error);
break;
default:
BUG();
break;
}
afs_free_skb(skb);
}
/* make sure the queue is empty if the call is done with (we might have
* aborted the call early because of an unmarshalling error) */
if (call->state >= AFS_CALL_COMPLETE) {
while ((skb = skb_dequeue(&call->rx_queue)))
afs_free_skb(skb);
if (call->incoming)
afs_end_call(call);
}
_leave("");
}
/*
* wait synchronously for a call to complete
*/
static int afs_wait_for_call_to_complete(struct afs_call *call)
{
struct sk_buff *skb;
int ret;
DECLARE_WAITQUEUE(myself, current);
_enter("");
add_wait_queue(&call->waitq, &myself);
for (;;) {
set_current_state(TASK_INTERRUPTIBLE);
/* deliver any messages that are in the queue */
if (!skb_queue_empty(&call->rx_queue)) {
__set_current_state(TASK_RUNNING);
afs_deliver_to_call(call);
continue;
}
ret = call->error;
if (call->state >= AFS_CALL_COMPLETE)
break;
ret = -EINTR;
if (signal_pending(current))
break;
schedule();
}
remove_wait_queue(&call->waitq, &myself);
__set_current_state(TASK_RUNNING);
/* kill the call */
if (call->state < AFS_CALL_COMPLETE) {
_debug("call incomplete");
rxrpc_kernel_abort_call(call->rxcall, RX_CALL_DEAD);
while ((skb = skb_dequeue(&call->rx_queue)))
afs_free_skb(skb);
}
_debug("call complete");
afs_end_call(call);
_leave(" = %d", ret);
return ret;
}
/*
* wake up a waiting call
*/
static void afs_wake_up_call_waiter(struct afs_call *call)
{
wake_up(&call->waitq);
}
/*
* wake up an asynchronous call
*/
static void afs_wake_up_async_call(struct afs_call *call)
{
_enter("");
queue_work(afs_async_calls, &call->async_work);
}
/*
* put a call into asynchronous mode
* - mustn't touch the call descriptor as the call my have completed by the
* time we get here
*/
static int afs_dont_wait_for_call_to_complete(struct afs_call *call)
{
_enter("");
return -EINPROGRESS;
}
/*
* delete an asynchronous call
*/
static void afs_delete_async_call(struct afs_call *call)
{
_enter("");
afs_free_call(call);
_leave("");
}
/*
* perform processing on an asynchronous call
* - on a multiple-thread workqueue this work item may try to run on several
* CPUs at the same time
*/
static void afs_process_async_call(struct afs_call *call)
{
_enter("");
if (!skb_queue_empty(&call->rx_queue))
afs_deliver_to_call(call);
if (call->state >= AFS_CALL_COMPLETE && call->wait_mode) {
if (call->wait_mode->async_complete)
call->wait_mode->async_complete(call->reply,
call->error);
call->reply = NULL;
/* kill the call */
afs_end_call_nofree(call);
/* we can't just delete the call because the work item may be
* queued */
call->async_workfn = afs_delete_async_call;
queue_work(afs_async_calls, &call->async_work);
}
_leave("");
}
/*
* Empty a socket buffer into a flat reply buffer.
*/
int afs_transfer_reply(struct afs_call *call, struct sk_buff *skb, bool last)
{
size_t len = skb->len;
if (len > call->reply_max - call->reply_size) {
_leave(" = -EBADMSG [%zu > %u]",
len, call->reply_max - call->reply_size);
return -EBADMSG;
}
if (len > 0) {
if (skb_copy_bits(skb, 0, call->buffer + call->reply_size,
len) < 0)
BUG();
call->reply_size += len;
}
afs_data_consumed(call, skb);
if (!last)
return -EAGAIN;
if (call->reply_size != call->reply_max) {
_leave(" = -EBADMSG [%u != %u]",
call->reply_size, call->reply_max);
return -EBADMSG;
}
return 0;
}
/*
* accept the backlog of incoming calls
*/
static void afs_collect_incoming_call(struct work_struct *work)
{
struct rxrpc_call *rxcall;
struct afs_call *call = NULL;
struct sk_buff *skb;
while ((skb = skb_dequeue(&afs_incoming_calls))) {
_debug("new call");
/* don't need the notification */
afs_free_skb(skb);
if (!call) {
call = kzalloc(sizeof(struct afs_call), GFP_KERNEL);
if (!call) {
rxrpc_kernel_reject_call(afs_socket);
return;
}
call->async_workfn = afs_process_async_call;
INIT_WORK(&call->async_work, afs_async_workfn);
call->wait_mode = &afs_async_incoming_call;
call->type = &afs_RXCMxxxx;
init_waitqueue_head(&call->waitq);
skb_queue_head_init(&call->rx_queue);
call->state = AFS_CALL_AWAIT_OP_ID;
_debug("CALL %p{%s} [%d]",
call, call->type->name,
atomic_read(&afs_outstanding_calls));
atomic_inc(&afs_outstanding_calls);
}
rxcall = rxrpc_kernel_accept_call(afs_socket,
(unsigned long) call);
if (!IS_ERR(rxcall)) {
call->rxcall = rxcall;
call = NULL;
}
}
if (call)
afs_free_call(call);
}
/*
* Grab the operation ID from an incoming cache manager call. The socket
* buffer is discarded on error or if we don't yet have sufficient data.
*/
static int afs_deliver_cm_op_id(struct afs_call *call, struct sk_buff *skb,
bool last)
{
size_t len = skb->len;
void *oibuf = (void *) &call->operation_ID;
_enter("{%u},{%zu},%d", call->offset, len, last);
ASSERTCMP(call->offset, <, 4);
/* the operation ID forms the first four bytes of the request data */
len = min_t(size_t, len, 4 - call->offset);
if (skb_copy_bits(skb, 0, oibuf + call->offset, len) < 0)
BUG();
if (!pskb_pull(skb, len))
BUG();
call->offset += len;
if (call->offset < 4) {
afs_data_consumed(call, skb);
_leave(" = -EAGAIN");
return -EAGAIN;
}
call->state = AFS_CALL_AWAIT_REQUEST;
/* ask the cache manager to route the call (it'll change the call type
* if successful) */
if (!afs_cm_incoming_call(call))
return -ENOTSUPP;
/* pass responsibility for the remainer of this message off to the
* cache manager op */
return call->type->deliver(call, skb, last);
}
/*
* send an empty reply
*/
void afs_send_empty_reply(struct afs_call *call)
{
struct msghdr msg;
_enter("");
msg.msg_name = NULL;
msg.msg_namelen = 0;
iov_iter_kvec(&msg.msg_iter, WRITE | ITER_KVEC, NULL, 0, 0);
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
call->state = AFS_CALL_AWAIT_ACK;
switch (rxrpc_kernel_send_data(call->rxcall, &msg, 0)) {
case 0:
_leave(" [replied]");
return;
case -ENOMEM:
_debug("oom");
rxrpc_kernel_abort_call(call->rxcall, RX_USER_ABORT);
default:
afs_end_call(call);
_leave(" [error]");
return;
}
}
/*
* send a simple reply
*/
void afs_send_simple_reply(struct afs_call *call, const void *buf, size_t len)
{
struct msghdr msg;
struct kvec iov[1];
int n;
_enter("");
iov[0].iov_base = (void *) buf;
iov[0].iov_len = len;
msg.msg_name = NULL;
msg.msg_namelen = 0;
iov_iter_kvec(&msg.msg_iter, WRITE | ITER_KVEC, iov, 1, len);
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
call->state = AFS_CALL_AWAIT_ACK;
n = rxrpc_kernel_send_data(call->rxcall, &msg, len);
if (n >= 0) {
/* Success */
_leave(" [replied]");
return;
}
if (n == -ENOMEM) {
_debug("oom");
rxrpc_kernel_abort_call(call->rxcall, RX_USER_ABORT);
}
afs_end_call(call);
_leave(" [error]");
}
/*
* Extract a piece of data from the received data socket buffers.
*/
int afs_extract_data(struct afs_call *call, struct sk_buff *skb,
bool last, void *buf, size_t count)
{
size_t len = skb->len;
_enter("{%u},{%zu},%d,,%zu", call->offset, len, last, count);
ASSERTCMP(call->offset, <, count);
len = min_t(size_t, len, count - call->offset);
if (skb_copy_bits(skb, 0, buf + call->offset, len) < 0 ||
!pskb_pull(skb, len))
BUG();
call->offset += len;
if (call->offset < count) {
afs_data_consumed(call, skb);
_leave(" = -EAGAIN");
return -EAGAIN;
}
return 0;
}