linux/net/sctp/ulpqueue.c

1152 lines
29 KiB
C
Raw Normal View History

/* SCTP kernel implementation
* (C) Copyright IBM Corp. 2001, 2004
* Copyright (c) 1999-2000 Cisco, Inc.
* Copyright (c) 1999-2001 Motorola, Inc.
* Copyright (c) 2001 Intel Corp.
* Copyright (c) 2001 Nokia, Inc.
* Copyright (c) 2001 La Monte H.P. Yarroll
*
* This abstraction carries sctp events to the ULP (sockets).
*
* This SCTP implementation is free software;
* you can redistribute it and/or modify it under the terms of
* the GNU General Public License as published by
* the Free Software Foundation; either version 2, or (at your option)
* any later version.
*
* This SCTP implementation is distributed in the hope that it
* will be useful, but WITHOUT ANY WARRANTY; without even the implied
* ************************
* warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with GNU CC; see the file COPYING. If not, see
* <http://www.gnu.org/licenses/>.
*
* Please send any bug reports or fixes you make to the
* email address(es):
* lksctp developers <linux-sctp@vger.kernel.org>
*
* Written or modified by:
* Jon Grimm <jgrimm@us.ibm.com>
* La Monte H.P. Yarroll <piggy@acm.org>
* Sridhar Samudrala <sri@us.ibm.com>
*/
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 08:04:11 +00:00
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/busy_poll.h>
#include <net/sctp/structs.h>
#include <net/sctp/sctp.h>
#include <net/sctp/sm.h>
/* Forward declarations for internal helpers. */
static struct sctp_ulpevent *sctp_ulpq_reasm(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *);
static struct sctp_ulpevent *sctp_ulpq_order(struct sctp_ulpq *,
struct sctp_ulpevent *);
static void sctp_ulpq_reasm_drain(struct sctp_ulpq *ulpq);
/* 1st Level Abstractions */
/* Initialize a ULP queue from a block of memory. */
struct sctp_ulpq *sctp_ulpq_init(struct sctp_ulpq *ulpq,
struct sctp_association *asoc)
{
memset(ulpq, 0, sizeof(struct sctp_ulpq));
ulpq->asoc = asoc;
skb_queue_head_init(&ulpq->reasm);
skb_queue_head_init(&ulpq->lobby);
ulpq->pd_mode = 0;
return ulpq;
}
/* Flush the reassembly and ordering queues. */
void sctp_ulpq_flush(struct sctp_ulpq *ulpq)
{
struct sk_buff *skb;
struct sctp_ulpevent *event;
while ((skb = __skb_dequeue(&ulpq->lobby)) != NULL) {
event = sctp_skb2event(skb);
sctp_ulpevent_free(event);
}
while ((skb = __skb_dequeue(&ulpq->reasm)) != NULL) {
event = sctp_skb2event(skb);
sctp_ulpevent_free(event);
}
}
/* Dispose of a ulpqueue. */
void sctp_ulpq_free(struct sctp_ulpq *ulpq)
{
sctp_ulpq_flush(ulpq);
}
/* Process an incoming DATA chunk. */
int sctp_ulpq_tail_data(struct sctp_ulpq *ulpq, struct sctp_chunk *chunk,
gfp_t gfp)
{
struct sk_buff_head temp;
struct sctp_ulpevent *event;
int event_eor = 0;
/* Create an event from the incoming chunk. */
event = sctp_ulpevent_make_rcvmsg(chunk->asoc, chunk, gfp);
if (!event)
return -ENOMEM;
/* Do reassembly if needed. */
event = sctp_ulpq_reasm(ulpq, event);
/* Do ordering if needed. */
if ((event) && (event->msg_flags & MSG_EOR)) {
/* Create a temporary list to collect chunks on. */
skb_queue_head_init(&temp);
__skb_queue_tail(&temp, sctp_event2skb(event));
event = sctp_ulpq_order(ulpq, event);
}
/* Send event to the ULP. 'event' is the sctp_ulpevent for
* very first SKB on the 'temp' list.
*/
if (event) {
event_eor = (event->msg_flags & MSG_EOR) ? 1 : 0;
sctp_ulpq_tail_event(ulpq, event);
}
return event_eor;
}
/* Add a new event for propagation to the ULP. */
/* Clear the partial delivery mode for this socket. Note: This
* assumes that no association is currently in partial delivery mode.
*/
int sctp_clear_pd(struct sock *sk, struct sctp_association *asoc)
{
struct sctp_sock *sp = sctp_sk(sk);
if (atomic_dec_and_test(&sp->pd_mode)) {
/* This means there are no other associations in PD, so
* we can go ahead and clear out the lobby in one shot
*/
if (!skb_queue_empty(&sp->pd_lobby)) {
sctp: simplify sk_receive_queue locking SCTP already serializes access to rcvbuf through its sock lock: sctp_recvmsg takes it right in the start and release at the end, while rx path will also take the lock before doing any socket processing. On sctp_rcv() it will check if there is an user using the socket and, if there is, it will queue incoming packets to the backlog. The backlog processing will do the same. Even timers will do such check and re-schedule if an user is using the socket. Simplifying this will allow us to remove sctp_skb_list_tail and get ride of some expensive lockings. The lists that it is used on are also mangled with functions like __skb_queue_tail and __skb_unlink in the same context, like on sctp_ulpq_tail_event() and sctp_clear_pd(). sctp_close() will also purge those while using only the sock lock. Therefore the lockings performed by sctp_skb_list_tail() are not necessary. This patch removes this function and replaces its calls with just skb_queue_splice_tail_init() instead. The biggest gain is at sctp_ulpq_tail_event(), because the events always contain a list, even if it's queueing a single skb and this was triggering expensive calls to spin_lock_irqsave/_irqrestore for every data chunk received. As SCTP will deliver each data chunk on a corresponding recvmsg, the more effective the change will be. Before this patch, with chunks with 30 bytes: netperf -t SCTP_STREAM -H 192.168.1.2 -cC -l 60 -- -m 30 -S 400000 400000 -s 400000 400000 on a 10Gbit link with 1500 MTU: SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.1.1 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 425984 425984 30 60.00 137.45 7.34 7.36 52.504 52.608 With it: SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.1.1 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 425984 425984 30 60.00 179.10 7.97 6.70 43.740 36.788 Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-13 22:12:29 +00:00
skb_queue_splice_tail_init(&sp->pd_lobby,
&sk->sk_receive_queue);
return 1;
}
} else {
/* There are other associations in PD, so we only need to
* pull stuff out of the lobby that belongs to the
* associations that is exiting PD (all of its notifications
* are posted here).
*/
if (!skb_queue_empty(&sp->pd_lobby) && asoc) {
struct sk_buff *skb, *tmp;
struct sctp_ulpevent *event;
sctp_skb_for_each(skb, &sp->pd_lobby, tmp) {
event = sctp_skb2event(skb);
if (event->asoc == asoc) {
__skb_unlink(skb, &sp->pd_lobby);
__skb_queue_tail(&sk->sk_receive_queue,
skb);
}
}
}
}
return 0;
}
/* Set the pd_mode on the socket and ulpq */
static void sctp_ulpq_set_pd(struct sctp_ulpq *ulpq)
{
struct sctp_sock *sp = sctp_sk(ulpq->asoc->base.sk);
atomic_inc(&sp->pd_mode);
ulpq->pd_mode = 1;
}
/* Clear the pd_mode and restart any pending messages waiting for delivery. */
static int sctp_ulpq_clear_pd(struct sctp_ulpq *ulpq)
{
ulpq->pd_mode = 0;
sctp_ulpq_reasm_drain(ulpq);
return sctp_clear_pd(ulpq->asoc->base.sk, ulpq->asoc);
}
/* If the SKB of 'event' is on a list, it is the first such member
* of that list.
*/
int sctp_ulpq_tail_event(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event)
{
struct sock *sk = ulpq->asoc->base.sk;
struct sctp_sock *sp = sctp_sk(sk);
struct sk_buff_head *queue, *skb_list;
struct sk_buff *skb = sctp_event2skb(event);
int clear_pd = 0;
skb_list = (struct sk_buff_head *) skb->prev;
/* If the socket is just going to throw this away, do not
* even try to deliver it.
*/
if (sk->sk_shutdown & RCV_SHUTDOWN &&
(sk->sk_shutdown & SEND_SHUTDOWN ||
!sctp_ulpevent_is_notification(event)))
goto out_free;
if (!sctp_ulpevent_is_notification(event)) {
sk_mark_napi_id(sk, skb);
sk_incoming_cpu_update(sk);
}
/* Check if the user wishes to receive this event. */
if (!sctp_ulpevent_is_enabled(event, &sp->subscribe))
goto out_free;
/* If we are in partial delivery mode, post to the lobby until
* partial delivery is cleared, unless, of course _this_ is
* the association the cause of the partial delivery.
*/
if (atomic_read(&sp->pd_mode) == 0) {
queue = &sk->sk_receive_queue;
} else {
if (ulpq->pd_mode) {
/* If the association is in partial delivery, we
* need to finish delivering the partially processed
* packet before passing any other data. This is
* because we don't truly support stream interleaving.
*/
if ((event->msg_flags & MSG_NOTIFICATION) ||
(SCTP_DATA_NOT_FRAG ==
(event->msg_flags & SCTP_DATA_FRAG_MASK)))
queue = &sp->pd_lobby;
else {
clear_pd = event->msg_flags & MSG_EOR;
queue = &sk->sk_receive_queue;
}
} else {
/*
* If fragment interleave is enabled, we
* can queue this to the receive queue instead
* of the lobby.
*/
if (sp->frag_interleave)
queue = &sk->sk_receive_queue;
else
queue = &sp->pd_lobby;
}
}
/* If we are harvesting multiple skbs they will be
* collected on a list.
*/
if (skb_list)
sctp: simplify sk_receive_queue locking SCTP already serializes access to rcvbuf through its sock lock: sctp_recvmsg takes it right in the start and release at the end, while rx path will also take the lock before doing any socket processing. On sctp_rcv() it will check if there is an user using the socket and, if there is, it will queue incoming packets to the backlog. The backlog processing will do the same. Even timers will do such check and re-schedule if an user is using the socket. Simplifying this will allow us to remove sctp_skb_list_tail and get ride of some expensive lockings. The lists that it is used on are also mangled with functions like __skb_queue_tail and __skb_unlink in the same context, like on sctp_ulpq_tail_event() and sctp_clear_pd(). sctp_close() will also purge those while using only the sock lock. Therefore the lockings performed by sctp_skb_list_tail() are not necessary. This patch removes this function and replaces its calls with just skb_queue_splice_tail_init() instead. The biggest gain is at sctp_ulpq_tail_event(), because the events always contain a list, even if it's queueing a single skb and this was triggering expensive calls to spin_lock_irqsave/_irqrestore for every data chunk received. As SCTP will deliver each data chunk on a corresponding recvmsg, the more effective the change will be. Before this patch, with chunks with 30 bytes: netperf -t SCTP_STREAM -H 192.168.1.2 -cC -l 60 -- -m 30 -S 400000 400000 -s 400000 400000 on a 10Gbit link with 1500 MTU: SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.1.1 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 425984 425984 30 60.00 137.45 7.34 7.36 52.504 52.608 With it: SCTP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.1.1 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 425984 425984 30 60.00 179.10 7.97 6.70 43.740 36.788 Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-13 22:12:29 +00:00
skb_queue_splice_tail_init(skb_list, queue);
else
__skb_queue_tail(queue, skb);
/* Did we just complete partial delivery and need to get
* rolling again? Move pending data to the receive
* queue.
*/
if (clear_pd)
sctp_ulpq_clear_pd(ulpq);
if (queue == &sk->sk_receive_queue && !sp->data_ready_signalled) {
sp->data_ready_signalled = 1;
sk->sk_data_ready(sk);
}
return 1;
out_free:
if (skb_list)
sctp_queue_purge_ulpevents(skb_list);
else
sctp_ulpevent_free(event);
return 0;
}
/* 2nd Level Abstractions */
/* Helper function to store chunks that need to be reassembled. */
static void sctp_ulpq_store_reasm(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *event)
{
struct sk_buff *pos;
struct sctp_ulpevent *cevent;
__u32 tsn, ctsn;
tsn = event->tsn;
/* See if it belongs at the end. */
pos = skb_peek_tail(&ulpq->reasm);
if (!pos) {
__skb_queue_tail(&ulpq->reasm, sctp_event2skb(event));
return;
}
/* Short circuit just dropping it at the end. */
cevent = sctp_skb2event(pos);
ctsn = cevent->tsn;
if (TSN_lt(ctsn, tsn)) {
__skb_queue_tail(&ulpq->reasm, sctp_event2skb(event));
return;
}
/* Find the right place in this list. We store them by TSN. */
skb_queue_walk(&ulpq->reasm, pos) {
cevent = sctp_skb2event(pos);
ctsn = cevent->tsn;
if (TSN_lt(tsn, ctsn))
break;
}
/* Insert before pos. */
__skb_queue_before(&ulpq->reasm, pos, sctp_event2skb(event));
}
/* Helper function to return an event corresponding to the reassembled
* datagram.
* This routine creates a re-assembled skb given the first and last skb's
* as stored in the reassembly queue. The skb's may be non-linear if the sctp
* payload was fragmented on the way and ip had to reassemble them.
* We add the rest of skb's to the first skb's fraglist.
*/
static struct sctp_ulpevent *sctp_make_reassembled_event(struct net *net,
struct sk_buff_head *queue, struct sk_buff *f_frag,
struct sk_buff *l_frag)
{
struct sk_buff *pos;
struct sk_buff *new = NULL;
struct sctp_ulpevent *event;
struct sk_buff *pnext, *last;
struct sk_buff *list = skb_shinfo(f_frag)->frag_list;
/* Store the pointer to the 2nd skb */
if (f_frag == l_frag)
pos = NULL;
else
pos = f_frag->next;
/* Get the last skb in the f_frag's frag_list if present. */
for (last = list; list; last = list, list = list->next)
;
/* Add the list of remaining fragments to the first fragments
* frag_list.
*/
if (last)
last->next = pos;
else {
if (skb_cloned(f_frag)) {
/* This is a cloned skb, we can't just modify
* the frag_list. We need a new skb to do that.
* Instead of calling skb_unshare(), we'll do it
* ourselves since we need to delay the free.
*/
new = skb_copy(f_frag, GFP_ATOMIC);
if (!new)
return NULL; /* try again later */
sctp_skb_set_owner_r(new, f_frag->sk);
skb_shinfo(new)->frag_list = pos;
} else
skb_shinfo(f_frag)->frag_list = pos;
}
/* Remove the first fragment from the reassembly queue. */
__skb_unlink(f_frag, queue);
/* if we did unshare, then free the old skb and re-assign */
if (new) {
kfree_skb(f_frag);
f_frag = new;
}
while (pos) {
pnext = pos->next;
/* Update the len and data_len fields of the first fragment. */
f_frag->len += pos->len;
f_frag->data_len += pos->len;
/* Remove the fragment from the reassembly queue. */
__skb_unlink(pos, queue);
/* Break if we have reached the last fragment. */
if (pos == l_frag)
break;
pos->next = pnext;
pos = pnext;
}
event = sctp_skb2event(f_frag);
SCTP_INC_STATS(net, SCTP_MIB_REASMUSRMSGS);
return event;
}
/* Helper function to check if an incoming chunk has filled up the last
* missing fragment in a SCTP datagram and return the corresponding event.
*/
static struct sctp_ulpevent *sctp_ulpq_retrieve_reassembled(struct sctp_ulpq *ulpq)
{
struct sk_buff *pos;
struct sctp_ulpevent *cevent;
struct sk_buff *first_frag = NULL;
__u32 ctsn, next_tsn;
struct sctp_ulpevent *retval = NULL;
struct sk_buff *pd_first = NULL;
struct sk_buff *pd_last = NULL;
size_t pd_len = 0;
struct sctp_association *asoc;
u32 pd_point;
/* Initialized to 0 just to avoid compiler warning message. Will
* never be used with this value. It is referenced only after it
* is set when we find the first fragment of a message.
*/
next_tsn = 0;
/* The chunks are held in the reasm queue sorted by TSN.
* Walk through the queue sequentially and look for a sequence of
* fragmented chunks that complete a datagram.
* 'first_frag' and next_tsn are reset when we find a chunk which
* is the first fragment of a datagram. Once these 2 fields are set
* we expect to find the remaining middle fragments and the last
* fragment in order. If not, first_frag is reset to NULL and we
* start the next pass when we find another first fragment.
*
* There is a potential to do partial delivery if user sets
* SCTP_PARTIAL_DELIVERY_POINT option. Lets count some things here
* to see if can do PD.
*/
skb_queue_walk(&ulpq->reasm, pos) {
cevent = sctp_skb2event(pos);
ctsn = cevent->tsn;
switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
case SCTP_DATA_FIRST_FRAG:
/* If this "FIRST_FRAG" is the first
* element in the queue, then count it towards
* possible PD.
*/
if (pos == ulpq->reasm.next) {
pd_first = pos;
pd_last = pos;
pd_len = pos->len;
} else {
pd_first = NULL;
pd_last = NULL;
pd_len = 0;
}
first_frag = pos;
next_tsn = ctsn + 1;
break;
case SCTP_DATA_MIDDLE_FRAG:
if ((first_frag) && (ctsn == next_tsn)) {
next_tsn++;
if (pd_first) {
pd_last = pos;
pd_len += pos->len;
}
} else
first_frag = NULL;
break;
case SCTP_DATA_LAST_FRAG:
if (first_frag && (ctsn == next_tsn))
goto found;
else
first_frag = NULL;
break;
}
}
asoc = ulpq->asoc;
if (pd_first) {
/* Make sure we can enter partial deliver.
* We can trigger partial delivery only if framgent
* interleave is set, or the socket is not already
* in partial delivery.
*/
if (!sctp_sk(asoc->base.sk)->frag_interleave &&
atomic_read(&sctp_sk(asoc->base.sk)->pd_mode))
goto done;
cevent = sctp_skb2event(pd_first);
pd_point = sctp_sk(asoc->base.sk)->pd_point;
if (pd_point && pd_point <= pd_len) {
retval = sctp_make_reassembled_event(sock_net(asoc->base.sk),
&ulpq->reasm,
pd_first,
pd_last);
if (retval)
sctp_ulpq_set_pd(ulpq);
}
}
done:
return retval;
found:
retval = sctp_make_reassembled_event(sock_net(ulpq->asoc->base.sk),
&ulpq->reasm, first_frag, pos);
if (retval)
retval->msg_flags |= MSG_EOR;
goto done;
}
/* Retrieve the next set of fragments of a partial message. */
static struct sctp_ulpevent *sctp_ulpq_retrieve_partial(struct sctp_ulpq *ulpq)
{
struct sk_buff *pos, *last_frag, *first_frag;
struct sctp_ulpevent *cevent;
__u32 ctsn, next_tsn;
int is_last;
struct sctp_ulpevent *retval;
/* The chunks are held in the reasm queue sorted by TSN.
* Walk through the queue sequentially and look for the first
* sequence of fragmented chunks.
*/
if (skb_queue_empty(&ulpq->reasm))
return NULL;
last_frag = first_frag = NULL;
retval = NULL;
next_tsn = 0;
is_last = 0;
skb_queue_walk(&ulpq->reasm, pos) {
cevent = sctp_skb2event(pos);
ctsn = cevent->tsn;
switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
case SCTP_DATA_FIRST_FRAG:
if (!first_frag)
return NULL;
goto done;
case SCTP_DATA_MIDDLE_FRAG:
if (!first_frag) {
first_frag = pos;
next_tsn = ctsn + 1;
last_frag = pos;
} else if (next_tsn == ctsn) {
next_tsn++;
last_frag = pos;
} else
goto done;
break;
case SCTP_DATA_LAST_FRAG:
if (!first_frag)
first_frag = pos;
else if (ctsn != next_tsn)
goto done;
last_frag = pos;
is_last = 1;
goto done;
default:
return NULL;
}
}
/* We have the reassembled event. There is no need to look
* further.
*/
done:
retval = sctp_make_reassembled_event(sock_net(ulpq->asoc->base.sk),
&ulpq->reasm, first_frag, last_frag);
if (retval && is_last)
retval->msg_flags |= MSG_EOR;
return retval;
}
/* Helper function to reassemble chunks. Hold chunks on the reasm queue that
* need reassembling.
*/
static struct sctp_ulpevent *sctp_ulpq_reasm(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *event)
{
struct sctp_ulpevent *retval = NULL;
/* Check if this is part of a fragmented message. */
if (SCTP_DATA_NOT_FRAG == (event->msg_flags & SCTP_DATA_FRAG_MASK)) {
event->msg_flags |= MSG_EOR;
return event;
}
sctp_ulpq_store_reasm(ulpq, event);
if (!ulpq->pd_mode)
retval = sctp_ulpq_retrieve_reassembled(ulpq);
else {
__u32 ctsn, ctsnap;
/* Do not even bother unless this is the next tsn to
* be delivered.
*/
ctsn = event->tsn;
ctsnap = sctp_tsnmap_get_ctsn(&ulpq->asoc->peer.tsn_map);
if (TSN_lte(ctsn, ctsnap))
retval = sctp_ulpq_retrieve_partial(ulpq);
}
return retval;
}
/* Retrieve the first part (sequential fragments) for partial delivery. */
static struct sctp_ulpevent *sctp_ulpq_retrieve_first(struct sctp_ulpq *ulpq)
{
struct sk_buff *pos, *last_frag, *first_frag;
struct sctp_ulpevent *cevent;
__u32 ctsn, next_tsn;
struct sctp_ulpevent *retval;
/* The chunks are held in the reasm queue sorted by TSN.
* Walk through the queue sequentially and look for a sequence of
* fragmented chunks that start a datagram.
*/
if (skb_queue_empty(&ulpq->reasm))
return NULL;
last_frag = first_frag = NULL;
retval = NULL;
next_tsn = 0;
skb_queue_walk(&ulpq->reasm, pos) {
cevent = sctp_skb2event(pos);
ctsn = cevent->tsn;
switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
case SCTP_DATA_FIRST_FRAG:
if (!first_frag) {
first_frag = pos;
next_tsn = ctsn + 1;
last_frag = pos;
} else
goto done;
break;
case SCTP_DATA_MIDDLE_FRAG:
if (!first_frag)
return NULL;
if (ctsn == next_tsn) {
next_tsn++;
last_frag = pos;
} else
goto done;
break;
case SCTP_DATA_LAST_FRAG:
if (!first_frag)
return NULL;
else
goto done;
break;
default:
return NULL;
}
}
/* We have the reassembled event. There is no need to look
* further.
*/
done:
retval = sctp_make_reassembled_event(sock_net(ulpq->asoc->base.sk),
&ulpq->reasm, first_frag, last_frag);
return retval;
}
/*
* Flush out stale fragments from the reassembly queue when processing
* a Forward TSN.
*
* RFC 3758, Section 3.6
*
* After receiving and processing a FORWARD TSN, the data receiver MUST
* take cautions in updating its re-assembly queue. The receiver MUST
* remove any partially reassembled message, which is still missing one
* or more TSNs earlier than or equal to the new cumulative TSN point.
* In the event that the receiver has invoked the partial delivery API,
* a notification SHOULD also be generated to inform the upper layer API
* that the message being partially delivered will NOT be completed.
*/
void sctp_ulpq_reasm_flushtsn(struct sctp_ulpq *ulpq, __u32 fwd_tsn)
{
struct sk_buff *pos, *tmp;
struct sctp_ulpevent *event;
__u32 tsn;
if (skb_queue_empty(&ulpq->reasm))
return;
skb_queue_walk_safe(&ulpq->reasm, pos, tmp) {
event = sctp_skb2event(pos);
tsn = event->tsn;
/* Since the entire message must be abandoned by the
* sender (item A3 in Section 3.5, RFC 3758), we can
* free all fragments on the list that are less then
* or equal to ctsn_point
*/
if (TSN_lte(tsn, fwd_tsn)) {
__skb_unlink(pos, &ulpq->reasm);
sctp_ulpevent_free(event);
} else
break;
}
}
/*
* Drain the reassembly queue. If we just cleared parted delivery, it
* is possible that the reassembly queue will contain already reassembled
* messages. Retrieve any such messages and give them to the user.
*/
static void sctp_ulpq_reasm_drain(struct sctp_ulpq *ulpq)
{
struct sctp_ulpevent *event = NULL;
struct sk_buff_head temp;
if (skb_queue_empty(&ulpq->reasm))
return;
while ((event = sctp_ulpq_retrieve_reassembled(ulpq)) != NULL) {
/* Do ordering if needed. */
if ((event) && (event->msg_flags & MSG_EOR)) {
skb_queue_head_init(&temp);
__skb_queue_tail(&temp, sctp_event2skb(event));
event = sctp_ulpq_order(ulpq, event);
}
/* Send event to the ULP. 'event' is the
* sctp_ulpevent for very first SKB on the temp' list.
*/
if (event)
sctp_ulpq_tail_event(ulpq, event);
}
}
/* Helper function to gather skbs that have possibly become
* ordered by an an incoming chunk.
*/
static void sctp_ulpq_retrieve_ordered(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *event)
{
struct sk_buff_head *event_list;
struct sk_buff *pos, *tmp;
struct sctp_ulpevent *cevent;
struct sctp_stream *in;
__u16 sid, csid, cssn;
sid = event->stream;
in = &ulpq->asoc->ssnmap->in;
event_list = (struct sk_buff_head *) sctp_event2skb(event)->prev;
/* We are holding the chunks by stream, by SSN. */
sctp_skb_for_each(pos, &ulpq->lobby, tmp) {
cevent = (struct sctp_ulpevent *) pos->cb;
csid = cevent->stream;
cssn = cevent->ssn;
/* Have we gone too far? */
if (csid > sid)
break;
/* Have we not gone far enough? */
if (csid < sid)
continue;
if (cssn != sctp_ssn_peek(in, sid))
break;
/* Found it, so mark in the ssnmap. */
sctp_ssn_next(in, sid);
__skb_unlink(pos, &ulpq->lobby);
/* Attach all gathered skbs to the event. */
__skb_queue_tail(event_list, pos);
}
}
/* Helper function to store chunks needing ordering. */
static void sctp_ulpq_store_ordered(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *event)
{
struct sk_buff *pos;
struct sctp_ulpevent *cevent;
__u16 sid, csid;
__u16 ssn, cssn;
pos = skb_peek_tail(&ulpq->lobby);
if (!pos) {
__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
return;
}
sid = event->stream;
ssn = event->ssn;
cevent = (struct sctp_ulpevent *) pos->cb;
csid = cevent->stream;
cssn = cevent->ssn;
if (sid > csid) {
__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
return;
}
if ((sid == csid) && SSN_lt(cssn, ssn)) {
__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
return;
}
/* Find the right place in this list. We store them by
* stream ID and then by SSN.
*/
skb_queue_walk(&ulpq->lobby, pos) {
cevent = (struct sctp_ulpevent *) pos->cb;
csid = cevent->stream;
cssn = cevent->ssn;
if (csid > sid)
break;
if (csid == sid && SSN_lt(ssn, cssn))
break;
}
/* Insert before pos. */
__skb_queue_before(&ulpq->lobby, pos, sctp_event2skb(event));
}
static struct sctp_ulpevent *sctp_ulpq_order(struct sctp_ulpq *ulpq,
struct sctp_ulpevent *event)
{
__u16 sid, ssn;
struct sctp_stream *in;
/* Check if this message needs ordering. */
if (SCTP_DATA_UNORDERED & event->msg_flags)
return event;
/* Note: The stream ID must be verified before this routine. */
sid = event->stream;
ssn = event->ssn;
in = &ulpq->asoc->ssnmap->in;
/* Is this the expected SSN for this stream ID? */
if (ssn != sctp_ssn_peek(in, sid)) {
/* We've received something out of order, so find where it
* needs to be placed. We order by stream and then by SSN.
*/
sctp_ulpq_store_ordered(ulpq, event);
return NULL;
}
/* Mark that the next chunk has been found. */
sctp_ssn_next(in, sid);
/* Go find any other chunks that were waiting for
* ordering.
*/
sctp_ulpq_retrieve_ordered(ulpq, event);
return event;
}
/* Helper function to gather skbs that have possibly become
* ordered by forward tsn skipping their dependencies.
*/
static void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq, __u16 sid)
{
struct sk_buff *pos, *tmp;
struct sctp_ulpevent *cevent;
struct sctp_ulpevent *event;
struct sctp_stream *in;
struct sk_buff_head temp;
struct sk_buff_head *lobby = &ulpq->lobby;
__u16 csid, cssn;
in = &ulpq->asoc->ssnmap->in;
/* We are holding the chunks by stream, by SSN. */
skb_queue_head_init(&temp);
event = NULL;
sctp_skb_for_each(pos, lobby, tmp) {
cevent = (struct sctp_ulpevent *) pos->cb;
csid = cevent->stream;
cssn = cevent->ssn;
/* Have we gone too far? */
if (csid > sid)
break;
/* Have we not gone far enough? */
if (csid < sid)
continue;
/* see if this ssn has been marked by skipping */
if (!SSN_lt(cssn, sctp_ssn_peek(in, csid)))
break;
__skb_unlink(pos, lobby);
if (!event)
/* Create a temporary list to collect chunks on. */
event = sctp_skb2event(pos);
/* Attach all gathered skbs to the event. */
__skb_queue_tail(&temp, pos);
}
/* If we didn't reap any data, see if the next expected SSN
* is next on the queue and if so, use that.
*/
if (event == NULL && pos != (struct sk_buff *)lobby) {
cevent = (struct sctp_ulpevent *) pos->cb;
csid = cevent->stream;
cssn = cevent->ssn;
if (csid == sid && cssn == sctp_ssn_peek(in, csid)) {
sctp_ssn_next(in, csid);
__skb_unlink(pos, lobby);
__skb_queue_tail(&temp, pos);
event = sctp_skb2event(pos);
}
}
/* Send event to the ULP. 'event' is the sctp_ulpevent for
* very first SKB on the 'temp' list.
*/
if (event) {
/* see if we have more ordered that we can deliver */
sctp_ulpq_retrieve_ordered(ulpq, event);
sctp_ulpq_tail_event(ulpq, event);
}
}
/* Skip over an SSN. This is used during the processing of
* Forwared TSN chunk to skip over the abandoned ordered data
*/
void sctp_ulpq_skip(struct sctp_ulpq *ulpq, __u16 sid, __u16 ssn)
{
struct sctp_stream *in;
/* Note: The stream ID must be verified before this routine. */
in = &ulpq->asoc->ssnmap->in;
/* Is this an old SSN? If so ignore. */
if (SSN_lt(ssn, sctp_ssn_peek(in, sid)))
return;
/* Mark that we are no longer expecting this SSN or lower. */
sctp_ssn_skip(in, sid, ssn);
/* Go find any other chunks that were waiting for
* ordering and deliver them if needed.
*/
sctp_ulpq_reap_ordered(ulpq, sid);
}
static __u16 sctp_ulpq_renege_list(struct sctp_ulpq *ulpq,
struct sk_buff_head *list, __u16 needed)
{
__u16 freed = 0;
__u32 tsn, last_tsn;
struct sk_buff *skb, *flist, *last;
struct sctp_ulpevent *event;
struct sctp_tsnmap *tsnmap;
tsnmap = &ulpq->asoc->peer.tsn_map;
while ((skb = skb_peek_tail(list)) != NULL) {
event = sctp_skb2event(skb);
tsn = event->tsn;
/* Don't renege below the Cumulative TSN ACK Point. */
if (TSN_lte(tsn, sctp_tsnmap_get_ctsn(tsnmap)))
break;
/* Events in ordering queue may have multiple fragments
* corresponding to additional TSNs. Sum the total
* freed space; find the last TSN.
*/
freed += skb_headlen(skb);
flist = skb_shinfo(skb)->frag_list;
for (last = flist; flist; flist = flist->next) {
last = flist;
freed += skb_headlen(last);
}
if (last)
last_tsn = sctp_skb2event(last)->tsn;
else
last_tsn = tsn;
/* Unlink the event, then renege all applicable TSNs. */
__skb_unlink(skb, list);
sctp_ulpevent_free(event);
while (TSN_lte(tsn, last_tsn)) {
sctp_tsnmap_renege(tsnmap, tsn);
tsn++;
}
if (freed >= needed)
return freed;
}
return freed;
}
/* Renege 'needed' bytes from the ordering queue. */
static __u16 sctp_ulpq_renege_order(struct sctp_ulpq *ulpq, __u16 needed)
{
return sctp_ulpq_renege_list(ulpq, &ulpq->lobby, needed);
}
/* Renege 'needed' bytes from the reassembly queue. */
static __u16 sctp_ulpq_renege_frags(struct sctp_ulpq *ulpq, __u16 needed)
{
return sctp_ulpq_renege_list(ulpq, &ulpq->reasm, needed);
}
/* Partial deliver the first message as there is pressure on rwnd. */
void sctp_ulpq_partial_delivery(struct sctp_ulpq *ulpq,
gfp_t gfp)
{
struct sctp_ulpevent *event;
struct sctp_association *asoc;
struct sctp_sock *sp;
__u32 ctsn;
struct sk_buff *skb;
asoc = ulpq->asoc;
sp = sctp_sk(asoc->base.sk);
/* If the association is already in Partial Delivery mode
* we have nothing to do.
*/
if (ulpq->pd_mode)
return;
/* Data must be at or below the Cumulative TSN ACK Point to
* start partial delivery.
*/
skb = skb_peek(&asoc->ulpq.reasm);
if (skb != NULL) {
ctsn = sctp_skb2event(skb)->tsn;
if (!TSN_lte(ctsn, sctp_tsnmap_get_ctsn(&asoc->peer.tsn_map)))
return;
}
/* If the user enabled fragment interleave socket option,
* multiple associations can enter partial delivery.
* Otherwise, we can only enter partial delivery if the
* socket is not in partial deliver mode.
*/
if (sp->frag_interleave || atomic_read(&sp->pd_mode) == 0) {
/* Is partial delivery possible? */
event = sctp_ulpq_retrieve_first(ulpq);
/* Send event to the ULP. */
if (event) {
sctp_ulpq_tail_event(ulpq, event);
sctp_ulpq_set_pd(ulpq);
return;
}
}
}
/* Renege some packets to make room for an incoming chunk. */
void sctp_ulpq_renege(struct sctp_ulpq *ulpq, struct sctp_chunk *chunk,
gfp_t gfp)
{
struct sctp_association *asoc;
__u16 needed, freed;
asoc = ulpq->asoc;
if (chunk) {
needed = ntohs(chunk->chunk_hdr->length);
needed -= sizeof(sctp_data_chunk_t);
} else
needed = SCTP_DEFAULT_MAXWINDOW;
freed = 0;
if (skb_queue_empty(&asoc->base.sk->sk_receive_queue)) {
freed = sctp_ulpq_renege_order(ulpq, needed);
if (freed < needed) {
freed += sctp_ulpq_renege_frags(ulpq, needed - freed);
}
}
/* If able to free enough room, accept this chunk. */
if (chunk && (freed >= needed)) {
int retval;
retval = sctp_ulpq_tail_data(ulpq, chunk, gfp);
/*
* Enter partial delivery if chunk has not been
* delivered; otherwise, drain the reassembly queue.
*/
if (retval <= 0)
sctp_ulpq_partial_delivery(ulpq, gfp);
else if (retval == 1)
sctp_ulpq_reasm_drain(ulpq);
}
[NET] CORE: Introducing new memory accounting interface. This patch introduces new memory accounting functions for each network protocol. Most of them are renamed from memory accounting functions for stream protocols. At the same time, some stream memory accounting functions are removed since other functions do same thing. Renaming: sk_stream_free_skb() -> sk_wmem_free_skb() __sk_stream_mem_reclaim() -> __sk_mem_reclaim() sk_stream_mem_reclaim() -> sk_mem_reclaim() sk_stream_mem_schedule -> __sk_mem_schedule() sk_stream_pages() -> sk_mem_pages() sk_stream_rmem_schedule() -> sk_rmem_schedule() sk_stream_wmem_schedule() -> sk_wmem_schedule() sk_charge_skb() -> sk_mem_charge() Removeing sk_stream_rfree(): consolidates into sock_rfree() sk_stream_set_owner_r(): consolidates into skb_set_owner_r() sk_stream_mem_schedule() The following functions are added. sk_has_account(): check if the protocol supports accounting sk_mem_uncharge(): do the opposite of sk_mem_charge() In addition, to achieve consolidation, updating sk_wmem_queued is removed from sk_mem_charge(). Next, to consolidate memory accounting functions, this patch adds memory accounting calls to network core functions. Moreover, present memory accounting call is renamed to new accounting call. Finally we replace present memory accounting calls with new interface in TCP and SCTP. Signed-off-by: Takahiro Yasui <tyasui@redhat.com> Signed-off-by: Hideo Aoki <haoki@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-12-31 08:11:19 +00:00
sk_mem_reclaim(asoc->base.sk);
}
/* Notify the application if an association is aborted and in
* partial delivery mode. Send up any pending received messages.
*/
void sctp_ulpq_abort_pd(struct sctp_ulpq *ulpq, gfp_t gfp)
{
struct sctp_ulpevent *ev = NULL;
struct sock *sk;
struct sctp_sock *sp;
if (!ulpq->pd_mode)
return;
sk = ulpq->asoc->base.sk;
sp = sctp_sk(sk);
if (sctp_ulpevent_type_enabled(SCTP_PARTIAL_DELIVERY_EVENT,
&sctp_sk(sk)->subscribe))
ev = sctp_ulpevent_make_pdapi(ulpq->asoc,
SCTP_PARTIAL_DELIVERY_ABORTED,
gfp);
if (ev)
__skb_queue_tail(&sk->sk_receive_queue, sctp_event2skb(ev));
/* If there is data waiting, send it up the socket now. */
if ((sctp_ulpq_clear_pd(ulpq) || ev) && !sp->data_ready_signalled) {
sp->data_ready_signalled = 1;
sk->sk_data_ready(sk);
}
}