linux/net/rxrpc/output.c

733 lines
20 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/* RxRPC packet transmission
*
* Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/net.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/gfp.h>
#include <linux/skbuff.h>
#include <linux/export.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include <net/udp.h>
#include "ar-internal.h"
extern int udpv6_sendmsg(struct sock *sk, struct msghdr *msg, size_t len);
rxrpc: Fix oops from calling udpv6_sendmsg() on AF_INET socket If rxrpc sees an IPv6 address, it assumes it can call udpv6_sendmsg() on it - even if it got it on an IPv4 socket. Fix do_udp_sendmsg() to give an error in such a case. general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] ... RIP: 0010:ipv6_addr_v4mapped include/net/ipv6.h:749 [inline] RIP: 0010:udpv6_sendmsg+0xd0a/0x2c70 net/ipv6/udp.c:1361 ... Call Trace: do_udp_sendmsg net/rxrpc/output.c:27 [inline] do_udp_sendmsg net/rxrpc/output.c:21 [inline] rxrpc_send_abort_packet+0x73b/0x860 net/rxrpc/output.c:367 rxrpc_release_calls_on_socket+0x211/0x300 net/rxrpc/call_object.c:595 rxrpc_release_sock net/rxrpc/af_rxrpc.c:886 [inline] rxrpc_release+0x263/0x5a0 net/rxrpc/af_rxrpc.c:917 __sock_release+0xcd/0x280 net/socket.c:650 sock_close+0x18/0x20 net/socket.c:1365 __fput+0x27c/0xa90 fs/file_table.c:320 task_work_run+0x16b/0x270 kernel/task_work.c:179 exit_task_work include/linux/task_work.h:38 [inline] do_exit+0xb35/0x2a20 kernel/exit.c:820 do_group_exit+0xd0/0x2a0 kernel/exit.c:950 __do_sys_exit_group kernel/exit.c:961 [inline] __se_sys_exit_group kernel/exit.c:959 [inline] __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:959 Fixes: ed472b0c8783 ("rxrpc: Call udp_sendmsg() directly") Reported-by: Eric Dumazet <edumazet@google.com> Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-11-12 00:00:21 +08:00
static ssize_t do_udp_sendmsg(struct socket *socket, struct msghdr *msg, size_t len)
{
struct sockaddr *sa = msg->msg_name;
rxrpc: Fix oops from calling udpv6_sendmsg() on AF_INET socket If rxrpc sees an IPv6 address, it assumes it can call udpv6_sendmsg() on it - even if it got it on an IPv4 socket. Fix do_udp_sendmsg() to give an error in such a case. general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] ... RIP: 0010:ipv6_addr_v4mapped include/net/ipv6.h:749 [inline] RIP: 0010:udpv6_sendmsg+0xd0a/0x2c70 net/ipv6/udp.c:1361 ... Call Trace: do_udp_sendmsg net/rxrpc/output.c:27 [inline] do_udp_sendmsg net/rxrpc/output.c:21 [inline] rxrpc_send_abort_packet+0x73b/0x860 net/rxrpc/output.c:367 rxrpc_release_calls_on_socket+0x211/0x300 net/rxrpc/call_object.c:595 rxrpc_release_sock net/rxrpc/af_rxrpc.c:886 [inline] rxrpc_release+0x263/0x5a0 net/rxrpc/af_rxrpc.c:917 __sock_release+0xcd/0x280 net/socket.c:650 sock_close+0x18/0x20 net/socket.c:1365 __fput+0x27c/0xa90 fs/file_table.c:320 task_work_run+0x16b/0x270 kernel/task_work.c:179 exit_task_work include/linux/task_work.h:38 [inline] do_exit+0xb35/0x2a20 kernel/exit.c:820 do_group_exit+0xd0/0x2a0 kernel/exit.c:950 __do_sys_exit_group kernel/exit.c:961 [inline] __se_sys_exit_group kernel/exit.c:959 [inline] __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:959 Fixes: ed472b0c8783 ("rxrpc: Call udp_sendmsg() directly") Reported-by: Eric Dumazet <edumazet@google.com> Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-11-12 00:00:21 +08:00
struct sock *sk = socket->sk;
rxrpc: Fix oops from calling udpv6_sendmsg() on AF_INET socket If rxrpc sees an IPv6 address, it assumes it can call udpv6_sendmsg() on it - even if it got it on an IPv4 socket. Fix do_udp_sendmsg() to give an error in such a case. general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] ... RIP: 0010:ipv6_addr_v4mapped include/net/ipv6.h:749 [inline] RIP: 0010:udpv6_sendmsg+0xd0a/0x2c70 net/ipv6/udp.c:1361 ... Call Trace: do_udp_sendmsg net/rxrpc/output.c:27 [inline] do_udp_sendmsg net/rxrpc/output.c:21 [inline] rxrpc_send_abort_packet+0x73b/0x860 net/rxrpc/output.c:367 rxrpc_release_calls_on_socket+0x211/0x300 net/rxrpc/call_object.c:595 rxrpc_release_sock net/rxrpc/af_rxrpc.c:886 [inline] rxrpc_release+0x263/0x5a0 net/rxrpc/af_rxrpc.c:917 __sock_release+0xcd/0x280 net/socket.c:650 sock_close+0x18/0x20 net/socket.c:1365 __fput+0x27c/0xa90 fs/file_table.c:320 task_work_run+0x16b/0x270 kernel/task_work.c:179 exit_task_work include/linux/task_work.h:38 [inline] do_exit+0xb35/0x2a20 kernel/exit.c:820 do_group_exit+0xd0/0x2a0 kernel/exit.c:950 __do_sys_exit_group kernel/exit.c:961 [inline] __se_sys_exit_group kernel/exit.c:959 [inline] __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:959 Fixes: ed472b0c8783 ("rxrpc: Call udp_sendmsg() directly") Reported-by: Eric Dumazet <edumazet@google.com> Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-11-12 00:00:21 +08:00
if (IS_ENABLED(CONFIG_AF_RXRPC_IPV6)) {
if (sa->sa_family == AF_INET6) {
if (sk->sk_family != AF_INET6) {
pr_warn("AF_INET6 address on AF_INET socket\n");
return -ENOPROTOOPT;
}
return udpv6_sendmsg(sk, msg, len);
}
}
return udp_sendmsg(sk, msg, len);
}
struct rxrpc_abort_buffer {
struct rxrpc_wire_header whdr;
__be32 abort_code;
};
static const char rxrpc_keepalive_string[] = "";
/*
* Increase Tx backoff on transmission failure and clear it on success.
*/
static void rxrpc_tx_backoff(struct rxrpc_call *call, int ret)
{
if (ret < 0) {
if (call->tx_backoff < 1000)
call->tx_backoff += 100;
} else {
call->tx_backoff = 0;
}
}
/*
* Arrange for a keepalive ping a certain time after we last transmitted. This
* lets the far side know we're still interested in this call and helps keep
* the route through any intervening firewall open.
*
* Receiving a response to the ping will prevent the ->expect_rx_by timer from
* expiring.
*/
static void rxrpc_set_keepalive(struct rxrpc_call *call, ktime_t now)
{
ktime_t delay = ms_to_ktime(READ_ONCE(call->next_rx_timo) / 6);
call->keepalive_at = ktime_add(ktime_get_real(), delay);
trace_rxrpc_timer_set(call, delay, rxrpc_timer_trace_keepalive);
}
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
/*
* Fill out an ACK packet.
*/
static void rxrpc_fill_out_ack(struct rxrpc_call *call,
struct rxrpc_txbuf *txb,
u8 ack_reason,
rxrpc_serial_t serial)
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
{
struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base;
struct rxrpc_acktrailer *trailer = txb->kvec[2].iov_base + 3;
struct rxrpc_ackpacket *ack = (struct rxrpc_ackpacket *)(whdr + 1);
unsigned int qsize, sack, wrap, to;
rxrpc_seq_t window, wtop;
int rsize;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
u32 mtu, jmax;
u8 *filler = txb->kvec[2].iov_base;
u8 *sackp = txb->kvec[1].iov_base;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
rxrpc_inc_stat(call->rxnet, stat_tx_ack_fill);
window = call->ackr_window;
wtop = call->ackr_wtop;
sack = call->ackr_sack_base % RXRPC_SACK_SIZE;
whdr->seq = 0;
whdr->type = RXRPC_PACKET_TYPE_ACK;
txb->flags |= RXRPC_SLOW_START_OK;
ack->bufferSpace = 0;
ack->maxSkew = 0;
ack->firstPacket = htonl(window);
ack->previousPacket = htonl(call->rx_highest_seq);
ack->serial = htonl(serial);
ack->reason = ack_reason;
ack->nAcks = wtop - window;
filler[0] = 0;
filler[1] = 0;
filler[2] = 0;
if (ack_reason == RXRPC_ACK_PING)
txb->flags |= RXRPC_REQUEST_ACK;
if (after(wtop, window)) {
txb->len += ack->nAcks;
txb->kvec[1].iov_base = sackp;
txb->kvec[1].iov_len = ack->nAcks;
wrap = RXRPC_SACK_SIZE - sack;
to = min_t(unsigned int, ack->nAcks, RXRPC_SACK_SIZE);
if (sack + ack->nAcks <= RXRPC_SACK_SIZE) {
memcpy(sackp, call->ackr_sack_table + sack, ack->nAcks);
} else {
memcpy(sackp, call->ackr_sack_table + sack, wrap);
memcpy(sackp + wrap, call->ackr_sack_table, to - wrap);
}
} else if (before(wtop, window)) {
pr_warn("ack window backward %x %x", window, wtop);
} else if (ack->reason == RXRPC_ACK_DELAY) {
ack->reason = RXRPC_ACK_IDLE;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
}
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
mtu = call->peer->if_mtu;
mtu -= call->peer->hdrsize;
jmax = rxrpc_rx_jumbo_max;
qsize = (window - 1) - call->rx_consumed;
rsize = max_t(int, call->rx_winsize - qsize, 0);
txb->ack_rwind = rsize;
trailer->maxMTU = htonl(rxrpc_rx_mtu);
trailer->ifMTU = htonl(mtu);
trailer->rwind = htonl(rsize);
trailer->jumbo_max = htonl(jmax);
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
}
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
/*
* Record the beginning of an RTT probe.
*/
static void rxrpc_begin_rtt_probe(struct rxrpc_call *call, rxrpc_serial_t serial,
ktime_t now, enum rxrpc_rtt_tx_trace why)
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
{
unsigned long avail = call->rtt_avail;
int rtt_slot = 9;
if (!(avail & RXRPC_CALL_RTT_AVAIL_MASK))
goto no_slot;
rtt_slot = __ffs(avail & RXRPC_CALL_RTT_AVAIL_MASK);
if (!test_and_clear_bit(rtt_slot, &call->rtt_avail))
goto no_slot;
call->rtt_serial[rtt_slot] = serial;
call->rtt_sent_at[rtt_slot] = now;
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
smp_wmb(); /* Write data before avail bit */
set_bit(rtt_slot + RXRPC_CALL_RTT_PEND_SHIFT, &call->rtt_avail);
trace_rxrpc_rtt_tx(call, why, rtt_slot, serial);
return;
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
no_slot:
trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_no_slot, rtt_slot, serial);
}
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
/*
* Transmit an ACK packet.
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
*/
static void rxrpc_send_ack_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
{
struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base;
struct rxrpc_connection *conn;
struct rxrpc_ackpacket *ack = (struct rxrpc_ackpacket *)(whdr + 1);
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
struct msghdr msg;
ktime_t now;
int ret;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
if (test_bit(RXRPC_CALL_DISCONNECTED, &call->flags))
return;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
conn = call->conn;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
msg.msg_name = &call->peer->srx.transport;
msg.msg_namelen = call->peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = MSG_SPLICE_PAGES;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
whdr->flags = txb->flags & RXRPC_TXBUF_WIRE_FLAGS;
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
txb->serial = rxrpc_get_next_serial(conn);
whdr->serial = htonl(txb->serial);
trace_rxrpc_tx_ack(call->debug_id, txb->serial,
ntohl(ack->firstPacket),
ntohl(ack->serial), ack->reason, ack->nAcks,
txb->ack_rwind);
rxrpc_inc_stat(call->rxnet, stat_tx_ack_send);
iov_iter_kvec(&msg.msg_iter, WRITE, txb->kvec, txb->nr_kvec, txb->len);
rxrpc_local_dont_fragment(conn->local, false);
ret = do_udp_sendmsg(conn->local->socket, &msg, txb->len);
call->peer->last_tx_at = ktime_get_seconds();
if (ret < 0) {
trace_rxrpc_tx_fail(call->debug_id, txb->serial, ret,
rxrpc_tx_point_call_ack);
} else {
trace_rxrpc_tx_packet(call->debug_id, whdr,
rxrpc_tx_point_call_ack);
now = ktime_get_real();
if (ack->reason == RXRPC_ACK_PING)
rxrpc_begin_rtt_probe(call, txb->serial, now, rxrpc_rtt_tx_ping);
if (txb->flags & RXRPC_REQUEST_ACK)
call->peer->rtt_last_req = now;
rxrpc_set_keepalive(call, now);
}
rxrpc_tx_backoff(call, ret);
rxrpc: Calls shouldn't hold socket refs 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>
2016-09-07 16:19:31 +08:00
}
/*
* Queue an ACK for immediate transmission.
*/
void rxrpc_send_ACK(struct rxrpc_call *call, u8 ack_reason,
rxrpc_serial_t serial, enum rxrpc_propose_ack_trace why)
{
struct rxrpc_txbuf *txb;
if (test_bit(RXRPC_CALL_DISCONNECTED, &call->flags))
return;
rxrpc_inc_stat(call->rxnet, stat_tx_acks[ack_reason]);
txb = rxrpc_alloc_ack_txbuf(call, call->ackr_wtop - call->ackr_window);
if (!txb) {
kleave(" = -ENOMEM");
return;
}
txb->ack_why = why;
rxrpc_fill_out_ack(call, txb, ack_reason, serial);
call->ackr_nr_unacked = 0;
atomic_set(&call->ackr_nr_consumed, 0);
clear_bit(RXRPC_CALL_RX_IS_IDLE, &call->flags);
trace_rxrpc_send_ack(call, why, ack_reason, serial);
rxrpc_send_ack_packet(call, txb);
rxrpc_put_txbuf(txb, rxrpc_txbuf_put_ack_tx);
}
/*
* Send an ABORT call packet.
*/
int rxrpc_send_abort_packet(struct rxrpc_call *call)
{
struct rxrpc_connection *conn;
struct rxrpc_abort_buffer pkt;
struct msghdr msg;
struct kvec iov[1];
rxrpc_serial_t serial;
int ret;
/* Don't bother sending aborts for a client call once the server has
* hard-ACK'd all of its request data. After that point, we're not
* going to stop the operation proceeding, and whilst we might limit
* the reply, it's not worth it if we can send a new call on the same
* channel instead, thereby closing off this call.
*/
if (rxrpc_is_client_call(call) &&
rxrpc: Don't use a ring buffer for call Tx queue Change the way the Tx queueing works to make the following ends easier to achieve: (1) The filling of packets, the encryption of packets and the transmission of packets can be handled in parallel by separate threads, rather than rxrpc_sendmsg() allocating, filling, encrypting and transmitting each packet before moving onto the next one. (2) Get rid of the fixed-size ring which sets a hard limit on the number of packets that can be retained in the ring. This allows the number of packets to increase without having to allocate a very large ring or having variable-sized rings. [Note: the downside of this is that it's then less efficient to locate a packet for retransmission as we then have to step through a list and examine each buffer in the list.] (3) Allow the filler/encrypter to run ahead of the transmission window. (4) Make it easier to do zero copy UDP from the packet buffers. (5) Make it easier to do zero copy from userspace to the packet buffers - and thence to UDP (only if for unauthenticated connections). To that end, the following changes are made: (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets to be transmitted in. This allows them to be placed on multiple queues simultaneously. An sk_buff isn't really necessary as it's never passed on to lower-level networking code. (2) Keep the transmissable packets in a linked list on the call struct rather than in a ring. As a consequence, the annotation buffer isn't used either; rather a flag is set on the packet to indicate ackedness. (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be transmitted has been queued. Add RXRPC_CALL_TX_ALL_ACKED to indicate that all packets up to and including the last got hard acked. (4) Wire headers are now stored in the txbuf rather than being concocted on the stack and they're stored immediately before the data, thereby allowing zerocopy of a single span. (5) Don't bother with instant-resend on transmission failure; rather, leave it for a timer or an ACK packet to trigger. Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-04-01 06:55:08 +08:00
test_bit(RXRPC_CALL_TX_ALL_ACKED, &call->flags))
return 0;
if (test_bit(RXRPC_CALL_DISCONNECTED, &call->flags))
return -ECONNRESET;
conn = call->conn;
msg.msg_name = &call->peer->srx.transport;
msg.msg_namelen = call->peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
pkt.whdr.epoch = htonl(conn->proto.epoch);
pkt.whdr.cid = htonl(call->cid);
pkt.whdr.callNumber = htonl(call->call_id);
pkt.whdr.seq = 0;
pkt.whdr.type = RXRPC_PACKET_TYPE_ABORT;
pkt.whdr.flags = conn->out_clientflag;
pkt.whdr.userStatus = 0;
pkt.whdr.securityIndex = call->security_ix;
pkt.whdr._rsvd = 0;
pkt.whdr.serviceId = htons(call->dest_srx.srx_service);
pkt.abort_code = htonl(call->abort_code);
iov[0].iov_base = &pkt;
iov[0].iov_len = sizeof(pkt);
serial = rxrpc_get_next_serial(conn);
pkt.whdr.serial = htonl(serial);
iov_iter_kvec(&msg.msg_iter, WRITE, iov, 1, sizeof(pkt));
ret = do_udp_sendmsg(conn->local->socket, &msg, sizeof(pkt));
conn->peer->last_tx_at = ktime_get_seconds();
if (ret < 0)
trace_rxrpc_tx_fail(call->debug_id, serial, ret,
rxrpc_tx_point_call_abort);
else
trace_rxrpc_tx_packet(call->debug_id, &pkt.whdr,
rxrpc_tx_point_call_abort);
rxrpc_tx_backoff(call, ret);
return ret;
}
/*
* Prepare a (sub)packet for transmission.
*/
static void rxrpc_prepare_data_subpacket(struct rxrpc_call *call, struct rxrpc_txbuf *txb,
rxrpc_serial_t serial)
{
struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base;
enum rxrpc_req_ack_trace why;
struct rxrpc_connection *conn = call->conn;
rxrpc: Don't use a ring buffer for call Tx queue Change the way the Tx queueing works to make the following ends easier to achieve: (1) The filling of packets, the encryption of packets and the transmission of packets can be handled in parallel by separate threads, rather than rxrpc_sendmsg() allocating, filling, encrypting and transmitting each packet before moving onto the next one. (2) Get rid of the fixed-size ring which sets a hard limit on the number of packets that can be retained in the ring. This allows the number of packets to increase without having to allocate a very large ring or having variable-sized rings. [Note: the downside of this is that it's then less efficient to locate a packet for retransmission as we then have to step through a list and examine each buffer in the list.] (3) Allow the filler/encrypter to run ahead of the transmission window. (4) Make it easier to do zero copy UDP from the packet buffers. (5) Make it easier to do zero copy from userspace to the packet buffers - and thence to UDP (only if for unauthenticated connections). To that end, the following changes are made: (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets to be transmitted in. This allows them to be placed on multiple queues simultaneously. An sk_buff isn't really necessary as it's never passed on to lower-level networking code. (2) Keep the transmissable packets in a linked list on the call struct rather than in a ring. As a consequence, the annotation buffer isn't used either; rather a flag is set on the packet to indicate ackedness. (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be transmitted has been queued. Add RXRPC_CALL_TX_ALL_ACKED to indicate that all packets up to and including the last got hard acked. (4) Wire headers are now stored in the txbuf rather than being concocted on the stack and they're stored immediately before the data, thereby allowing zerocopy of a single span. (5) Don't bother with instant-resend on transmission failure; rather, leave it for a timer or an ACK packet to trigger. Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-04-01 06:55:08 +08:00
_enter("%x,{%d}", txb->seq, txb->len);
txb->serial = serial;
if (test_bit(RXRPC_CONN_PROBING_FOR_UPGRADE, &conn->flags) &&
rxrpc: Don't use a ring buffer for call Tx queue Change the way the Tx queueing works to make the following ends easier to achieve: (1) The filling of packets, the encryption of packets and the transmission of packets can be handled in parallel by separate threads, rather than rxrpc_sendmsg() allocating, filling, encrypting and transmitting each packet before moving onto the next one. (2) Get rid of the fixed-size ring which sets a hard limit on the number of packets that can be retained in the ring. This allows the number of packets to increase without having to allocate a very large ring or having variable-sized rings. [Note: the downside of this is that it's then less efficient to locate a packet for retransmission as we then have to step through a list and examine each buffer in the list.] (3) Allow the filler/encrypter to run ahead of the transmission window. (4) Make it easier to do zero copy UDP from the packet buffers. (5) Make it easier to do zero copy from userspace to the packet buffers - and thence to UDP (only if for unauthenticated connections). To that end, the following changes are made: (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets to be transmitted in. This allows them to be placed on multiple queues simultaneously. An sk_buff isn't really necessary as it's never passed on to lower-level networking code. (2) Keep the transmissable packets in a linked list on the call struct rather than in a ring. As a consequence, the annotation buffer isn't used either; rather a flag is set on the packet to indicate ackedness. (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be transmitted has been queued. Add RXRPC_CALL_TX_ALL_ACKED to indicate that all packets up to and including the last got hard acked. (4) Wire headers are now stored in the txbuf rather than being concocted on the stack and they're stored immediately before the data, thereby allowing zerocopy of a single span. (5) Don't bother with instant-resend on transmission failure; rather, leave it for a timer or an ACK packet to trigger. Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-04-01 06:55:08 +08:00
txb->seq == 1)
whdr->userStatus = RXRPC_USERSTATUS_SERVICE_UPGRADE;
/* If our RTT cache needs working on, request an ACK. Also request
* ACKs if a DATA packet appears to have been lost.
*
* However, we mustn't request an ACK on the last reply packet of a
* service call, lest OpenAFS incorrectly send us an ACK with some
* soft-ACKs in it and then never follow up with a proper hard ACK.
*/
if (txb->flags & RXRPC_REQUEST_ACK)
why = rxrpc_reqack_already_on;
else if ((txb->flags & RXRPC_LAST_PACKET) && rxrpc_sending_to_client(txb))
why = rxrpc_reqack_no_srv_last;
else if (test_and_clear_bit(RXRPC_CALL_EV_ACK_LOST, &call->events))
why = rxrpc_reqack_ack_lost;
else if (txb->flags & RXRPC_TXBUF_RESENT)
why = rxrpc_reqack_retrans;
else if (call->cong_mode == RXRPC_CALL_SLOW_START && call->cong_cwnd <= 2)
why = rxrpc_reqack_slow_start;
else if (call->tx_winsize <= 2)
why = rxrpc_reqack_small_txwin;
rxrpc: Don't use a ring buffer for call Tx queue Change the way the Tx queueing works to make the following ends easier to achieve: (1) The filling of packets, the encryption of packets and the transmission of packets can be handled in parallel by separate threads, rather than rxrpc_sendmsg() allocating, filling, encrypting and transmitting each packet before moving onto the next one. (2) Get rid of the fixed-size ring which sets a hard limit on the number of packets that can be retained in the ring. This allows the number of packets to increase without having to allocate a very large ring or having variable-sized rings. [Note: the downside of this is that it's then less efficient to locate a packet for retransmission as we then have to step through a list and examine each buffer in the list.] (3) Allow the filler/encrypter to run ahead of the transmission window. (4) Make it easier to do zero copy UDP from the packet buffers. (5) Make it easier to do zero copy from userspace to the packet buffers - and thence to UDP (only if for unauthenticated connections). To that end, the following changes are made: (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets to be transmitted in. This allows them to be placed on multiple queues simultaneously. An sk_buff isn't really necessary as it's never passed on to lower-level networking code. (2) Keep the transmissable packets in a linked list on the call struct rather than in a ring. As a consequence, the annotation buffer isn't used either; rather a flag is set on the packet to indicate ackedness. (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be transmitted has been queued. Add RXRPC_CALL_TX_ALL_ACKED to indicate that all packets up to and including the last got hard acked. (4) Wire headers are now stored in the txbuf rather than being concocted on the stack and they're stored immediately before the data, thereby allowing zerocopy of a single span. (5) Don't bother with instant-resend on transmission failure; rather, leave it for a timer or an ACK packet to trigger. Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-04-01 06:55:08 +08:00
else if (call->peer->rtt_count < 3 && txb->seq & 1)
why = rxrpc_reqack_more_rtt;
else if (ktime_before(ktime_add_ms(call->peer->rtt_last_req, 1000), ktime_get_real()))
why = rxrpc_reqack_old_rtt;
else
goto dont_set_request_ack;
rxrpc_inc_stat(call->rxnet, stat_why_req_ack[why]);
rxrpc: Don't use a ring buffer for call Tx queue Change the way the Tx queueing works to make the following ends easier to achieve: (1) The filling of packets, the encryption of packets and the transmission of packets can be handled in parallel by separate threads, rather than rxrpc_sendmsg() allocating, filling, encrypting and transmitting each packet before moving onto the next one. (2) Get rid of the fixed-size ring which sets a hard limit on the number of packets that can be retained in the ring. This allows the number of packets to increase without having to allocate a very large ring or having variable-sized rings. [Note: the downside of this is that it's then less efficient to locate a packet for retransmission as we then have to step through a list and examine each buffer in the list.] (3) Allow the filler/encrypter to run ahead of the transmission window. (4) Make it easier to do zero copy UDP from the packet buffers. (5) Make it easier to do zero copy from userspace to the packet buffers - and thence to UDP (only if for unauthenticated connections). To that end, the following changes are made: (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets to be transmitted in. This allows them to be placed on multiple queues simultaneously. An sk_buff isn't really necessary as it's never passed on to lower-level networking code. (2) Keep the transmissable packets in a linked list on the call struct rather than in a ring. As a consequence, the annotation buffer isn't used either; rather a flag is set on the packet to indicate ackedness. (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be transmitted has been queued. Add RXRPC_CALL_TX_ALL_ACKED to indicate that all packets up to and including the last got hard acked. (4) Wire headers are now stored in the txbuf rather than being concocted on the stack and they're stored immediately before the data, thereby allowing zerocopy of a single span. (5) Don't bother with instant-resend on transmission failure; rather, leave it for a timer or an ACK packet to trigger. Signed-off-by: David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org
2022-04-01 06:55:08 +08:00
trace_rxrpc_req_ack(call->debug_id, txb->seq, why);
if (why != rxrpc_reqack_no_srv_last)
txb->flags |= RXRPC_REQUEST_ACK;
dont_set_request_ack:
whdr->flags = txb->flags & RXRPC_TXBUF_WIRE_FLAGS;
whdr->serial = htonl(txb->serial);
whdr->cksum = txb->cksum;
trace_rxrpc_tx_data(call, txb->seq, txb->serial, txb->flags, false);
}
/*
* Prepare a packet for transmission.
*/
static size_t rxrpc_prepare_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
{
rxrpc_serial_t serial;
/* Each transmission of a Tx packet needs a new serial number */
serial = rxrpc_get_next_serial(call->conn);
rxrpc_prepare_data_subpacket(call, txb, serial);
return txb->len;
}
/*
* Set timeouts after transmitting a packet.
*/
static void rxrpc_tstamp_data_packets(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
{
ktime_t now = ktime_get_real();
bool ack_requested = txb->flags & RXRPC_REQUEST_ACK;
call->tx_last_sent = now;
txb->last_sent = now;
if (ack_requested) {
rxrpc_begin_rtt_probe(call, txb->serial, now, rxrpc_rtt_tx_data);
call->peer->rtt_last_req = now;
if (call->peer->rtt_count > 1) {
ktime_t delay = rxrpc_get_rto_backoff(call->peer, false);
call->ack_lost_at = ktime_add(now, delay);
trace_rxrpc_timer_set(call, delay, rxrpc_timer_trace_lost_ack);
}
}
if (!test_and_set_bit(RXRPC_CALL_BEGAN_RX_TIMER, &call->flags)) {
ktime_t delay = ms_to_ktime(READ_ONCE(call->next_rx_timo));
call->expect_rx_by = ktime_add(now, delay);
trace_rxrpc_timer_set(call, delay, rxrpc_timer_trace_expect_rx);
}
rxrpc_set_keepalive(call, now);
}
/*
* send a packet through the transport endpoint
*/
static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
{
struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base;
struct rxrpc_connection *conn = call->conn;
enum rxrpc_tx_point frag;
struct msghdr msg;
size_t len;
int ret;
_enter("%x,{%d}", txb->seq, txb->len);
len = rxrpc_prepare_data_packet(call, txb);
if (IS_ENABLED(CONFIG_AF_RXRPC_INJECT_LOSS)) {
static int lose;
if ((lose++ & 7) == 7) {
ret = 0;
trace_rxrpc_tx_data(call, txb->seq, txb->serial,
txb->flags, true);
goto done;
}
}
iov_iter_kvec(&msg.msg_iter, WRITE, txb->kvec, txb->nr_kvec, len);
msg.msg_name = &call->peer->srx.transport;
msg.msg_namelen = call->peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = MSG_SPLICE_PAGES;
/* Track what we've attempted to transmit at least once so that the
* retransmission algorithm doesn't try to resend what we haven't sent
* yet.
*/
if (txb->seq == call->tx_transmitted + 1)
call->tx_transmitted = txb->seq;
/* send the packet with the don't fragment bit set if we currently
* think it's small enough */
if (txb->len >= call->peer->maxdata) {
rxrpc_local_dont_fragment(conn->local, false);
frag = rxrpc_tx_point_call_data_frag;
} else {
rxrpc_local_dont_fragment(conn->local, true);
frag = rxrpc_tx_point_call_data_nofrag;
}
retry:
/* send the packet by UDP
* - returns -EMSGSIZE if UDP would have to fragment the packet
* to go out of the interface
* - in which case, we'll have processed the ICMP error
* message and update the peer record
*/
rxrpc_inc_stat(call->rxnet, stat_tx_data_send);
ret = do_udp_sendmsg(conn->local->socket, &msg, len);
conn->peer->last_tx_at = ktime_get_seconds();
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
if (ret < 0) {
rxrpc_inc_stat(call->rxnet, stat_tx_data_send_fail);
trace_rxrpc_tx_fail(call->debug_id, txb->serial, ret, frag);
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
} else {
trace_rxrpc_tx_packet(call->debug_id, whdr, frag);
rxrpc: Fix loss of RTT samples due to interposed ACK The Rx protocol has a mechanism to help generate RTT samples that works by a client transmitting a REQUESTED-type ACK when it receives a DATA packet that has the REQUEST_ACK flag set. The peer, however, may interpose other ACKs before transmitting the REQUESTED-ACK, as can be seen in the following trace excerpt: rxrpc_tx_data: c=00000044 DATA d0b5ece8:00000001 00000001 q=00000001 fl=07 rxrpc_rx_ack: c=00000044 00000001 PNG r=00000000 f=00000002 p=00000000 n=0 rxrpc_rx_ack: c=00000044 00000002 REQ r=00000001 f=00000002 p=00000001 n=0 ... DATA packet 1 (q=xx) has REQUEST_ACK set (bit 1 of fl=xx). The incoming ping (labelled PNG) hard-acks the request DATA packet (f=xx exceeds the sequence number of the DATA packet), causing it to be discarded from the Tx ring. The ACK that was requested (labelled REQ, r=xx references the serial of the DATA packet) comes after the ping, but the sk_buff holding the timestamp has gone and the RTT sample is lost. This is particularly noticeable on RPC calls used to probe the service offered by the peer. A lot of peers end up with an unknown RTT because we only ever sent a single RPC. This confuses the server rotation algorithm. Fix this by caching the information about the outgoing packet in RTT calculations in the rxrpc_call struct rather than looking in the Tx ring. A four-deep buffer is maintained and both REQUEST_ACK-flagged DATA and PING-ACK transmissions are recorded in there. When the appropriate response ACK is received, the buffer is checked for a match and, if found, an RTT sample is recorded. If a received ACK refers to a packet with a later serial number than an entry in the cache, that entry is presumed lost and the entry is made available to record a new transmission. ACKs types other than REQUESTED-type and PING-type cause any matching sample to be cancelled as they don't necessarily represent a useful measurement. If there's no space in the buffer on ping/data transmission, the sample base is discarded. Fixes: 50235c4b5a2f ("rxrpc: Obtain RTT data by requesting ACKs on DATA packets") Signed-off-by: David Howells <dhowells@redhat.com>
2020-08-20 06:29:16 +08:00
}
rxrpc_tx_backoff(call, ret);
if (ret == -EMSGSIZE && frag == rxrpc_tx_point_call_data_frag) {
rxrpc_local_dont_fragment(conn->local, false);
frag = rxrpc_tx_point_call_data_frag;
goto retry;
}
done:
if (ret >= 0) {
rxrpc_tstamp_data_packets(call, txb);
} else {
/* Cancel the call if the initial transmission fails,
* particularly if that's due to network routing issues that
* aren't going away anytime soon. The layer above can arrange
* the retransmission.
*/
if (!test_and_set_bit(RXRPC_CALL_BEGAN_RX_TIMER, &call->flags))
rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR,
RX_USER_ABORT, ret);
}
_leave(" = %d [%u]", ret, call->peer->maxdata);
return ret;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
/*
* Transmit a connection-level abort.
*/
void rxrpc_send_conn_abort(struct rxrpc_connection *conn)
{
struct rxrpc_wire_header whdr;
struct msghdr msg;
struct kvec iov[2];
__be32 word;
size_t len;
u32 serial;
int ret;
msg.msg_name = &conn->peer->srx.transport;
msg.msg_namelen = conn->peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
whdr.epoch = htonl(conn->proto.epoch);
whdr.cid = htonl(conn->proto.cid);
whdr.callNumber = 0;
whdr.seq = 0;
whdr.type = RXRPC_PACKET_TYPE_ABORT;
whdr.flags = conn->out_clientflag;
whdr.userStatus = 0;
whdr.securityIndex = conn->security_ix;
whdr._rsvd = 0;
whdr.serviceId = htons(conn->service_id);
word = htonl(conn->abort_code);
iov[0].iov_base = &whdr;
iov[0].iov_len = sizeof(whdr);
iov[1].iov_base = &word;
iov[1].iov_len = sizeof(word);
len = iov[0].iov_len + iov[1].iov_len;
serial = rxrpc_get_next_serial(conn);
whdr.serial = htonl(serial);
iov_iter_kvec(&msg.msg_iter, WRITE, iov, 2, len);
ret = do_udp_sendmsg(conn->local->socket, &msg, len);
if (ret < 0) {
trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
rxrpc_tx_point_conn_abort);
_debug("sendmsg failed: %d", ret);
return;
}
trace_rxrpc_tx_packet(conn->debug_id, &whdr, rxrpc_tx_point_conn_abort);
conn->peer->last_tx_at = ktime_get_seconds();
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
/*
* Reject a packet through the local endpoint.
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
*/
void rxrpc_reject_packet(struct rxrpc_local *local, struct sk_buff *skb)
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
{
struct rxrpc_wire_header whdr;
struct sockaddr_rxrpc srx;
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
struct msghdr msg;
struct kvec iov[2];
size_t size;
__be32 code;
int ret, ioc;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
rxrpc_see_skb(skb, rxrpc_skb_see_reject);
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
iov[0].iov_base = &whdr;
iov[0].iov_len = sizeof(whdr);
iov[1].iov_base = &code;
iov[1].iov_len = sizeof(code);
msg.msg_name = &srx.transport;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
memset(&whdr, 0, sizeof(whdr));
switch (skb->mark) {
case RXRPC_SKB_MARK_REJECT_BUSY:
whdr.type = RXRPC_PACKET_TYPE_BUSY;
size = sizeof(whdr);
ioc = 1;
break;
case RXRPC_SKB_MARK_REJECT_ABORT:
whdr.type = RXRPC_PACKET_TYPE_ABORT;
code = htonl(skb->priority);
size = sizeof(whdr) + sizeof(code);
ioc = 2;
break;
default:
return;
}
if (rxrpc_extract_addr_from_skb(&srx, skb) == 0) {
msg.msg_namelen = srx.transport_len;
whdr.epoch = htonl(sp->hdr.epoch);
whdr.cid = htonl(sp->hdr.cid);
whdr.callNumber = htonl(sp->hdr.callNumber);
whdr.serviceId = htons(sp->hdr.serviceId);
whdr.flags = sp->hdr.flags;
whdr.flags ^= RXRPC_CLIENT_INITIATED;
whdr.flags &= RXRPC_CLIENT_INITIATED;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
iov_iter_kvec(&msg.msg_iter, WRITE, iov, ioc, size);
ret = do_udp_sendmsg(local->socket, &msg, size);
if (ret < 0)
trace_rxrpc_tx_fail(local->debug_id, 0, ret,
rxrpc_tx_point_reject);
else
trace_rxrpc_tx_packet(local->debug_id, &whdr,
rxrpc_tx_point_reject);
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
}
}
/*
* Send a VERSION reply to a peer as a keepalive.
*/
void rxrpc_send_keepalive(struct rxrpc_peer *peer)
{
struct rxrpc_wire_header whdr;
struct msghdr msg;
struct kvec iov[2];
size_t len;
int ret;
_enter("");
msg.msg_name = &peer->srx.transport;
msg.msg_namelen = peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
whdr.epoch = htonl(peer->local->rxnet->epoch);
whdr.cid = 0;
whdr.callNumber = 0;
whdr.seq = 0;
whdr.serial = 0;
whdr.type = RXRPC_PACKET_TYPE_VERSION; /* Not client-initiated */
whdr.flags = RXRPC_LAST_PACKET;
whdr.userStatus = 0;
whdr.securityIndex = 0;
whdr._rsvd = 0;
whdr.serviceId = 0;
iov[0].iov_base = &whdr;
iov[0].iov_len = sizeof(whdr);
iov[1].iov_base = (char *)rxrpc_keepalive_string;
iov[1].iov_len = sizeof(rxrpc_keepalive_string);
len = iov[0].iov_len + iov[1].iov_len;
iov_iter_kvec(&msg.msg_iter, WRITE, iov, 2, len);
ret = do_udp_sendmsg(peer->local->socket, &msg, len);
if (ret < 0)
trace_rxrpc_tx_fail(peer->debug_id, 0, ret,
rxrpc_tx_point_version_keepalive);
else
trace_rxrpc_tx_packet(peer->debug_id, &whdr,
rxrpc_tx_point_version_keepalive);
rxrpc: Fix the keepalive generator [ver #2] AF_RXRPC has a keepalive message generator that generates a message for a peer ~20s after the last transmission to that peer to keep firewall ports open. The implementation is incorrect in the following ways: (1) It mixes up ktime_t and time64_t types. (2) It uses ktime_get_real(), the output of which may jump forward or backward due to adjustments to the time of day. (3) If the current time jumps forward too much or jumps backwards, the generator function will crank the base of the time ring round one slot at a time (ie. a 1s period) until it catches up, spewing out VERSION packets as it goes. Fix the problem by: (1) Only using time64_t. There's no need for sub-second resolution. (2) Use ktime_get_seconds() rather than ktime_get_real() so that time isn't perceived to go backwards. (3) Simplifying rxrpc_peer_keepalive_worker() by splitting it into two parts: (a) The "worker" function that manages the buckets and the timer. (b) The "dispatch" function that takes the pending peers and potentially transmits a keepalive packet before putting them back in the ring into the slot appropriate to the revised last-Tx time. (4) Taking everything that's pending out of the ring and splicing it into a temporary collector list for processing. In the case that there's been a significant jump forward, the ring gets entirely emptied and then the time base can be warped forward before the peers are processed. The warping can't happen if the ring isn't empty because the slot a peer is in is keepalive-time dependent, relative to the base time. (5) Limit the number of iterations of the bucket array when scanning it. (6) Set the timer to skip any empty slots as there's no point waking up if there's nothing to do yet. This can be triggered by an incoming call from a server after a reboot with AF_RXRPC and AFS built into the kernel causing a peer record to be set up before userspace is started. The system clock is then adjusted by userspace, thereby potentially causing the keepalive generator to have a meltdown - which leads to a message like: watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [kworker/0:1:23] ... Workqueue: krxrpcd rxrpc_peer_keepalive_worker EIP: lock_acquire+0x69/0x80 ... Call Trace: ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? _raw_spin_lock_bh+0x29/0x60 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? __lock_acquire+0x3d3/0x870 ? process_one_work+0x110/0x340 ? process_one_work+0x166/0x340 ? process_one_work+0x110/0x340 ? worker_thread+0x39/0x3c0 ? kthread+0xdb/0x110 ? cancel_delayed_work+0x90/0x90 ? kthread_stop+0x70/0x70 ? ret_from_fork+0x19/0x24 Fixes: ace45bec6d77 ("rxrpc: Fix firewall route keepalive") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-08-08 18:30:02 +08:00
peer->last_tx_at = ktime_get_seconds();
_leave("");
}
/*
* Schedule an instant Tx resend.
*/
static inline void rxrpc_instant_resend(struct rxrpc_call *call,
struct rxrpc_txbuf *txb)
{
if (!__rxrpc_call_is_complete(call))
kdebug("resend");
}
/*
* Transmit one packet.
*/
void rxrpc_transmit_one(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
{
int ret;
ret = rxrpc_send_data_packet(call, txb);
if (ret < 0) {
switch (ret) {
case -ENETUNREACH:
case -EHOSTUNREACH:
case -ECONNREFUSED:
rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR,
0, ret);
break;
default:
_debug("need instant resend %d", ret);
rxrpc_instant_resend(call, txb);
}
} else {
ktime_t delay = ns_to_ktime(call->peer->rto_us * NSEC_PER_USEC);
call->resend_at = ktime_add(ktime_get_real(), delay);
trace_rxrpc_timer_set(call, delay, rxrpc_timer_trace_resend_tx);
}
}