2006-01-03 02:04:38 +08:00
|
|
|
/*
|
|
|
|
* net/tipc/msg.c: TIPC message header routines
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2015-03-14 04:08:06 +08:00
|
|
|
* Copyright (c) 2000-2006, 2014-2015, Ericsson AB
|
2011-01-26 02:33:31 +08:00
|
|
|
* Copyright (c) 2005, 2010-2011, Wind River Systems
|
2006-01-03 02:04:38 +08:00
|
|
|
* All rights reserved.
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Redistribution and use in source and binary forms, with or without
|
2006-01-03 02:04:38 +08:00
|
|
|
* modification, are permitted provided that the following conditions are met:
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 3. Neither the names of the copyright holders nor the names of its
|
|
|
|
* contributors may be used to endorse or promote products derived from
|
|
|
|
* this software without specific prior written permission.
|
2006-01-03 02:04:38 +08:00
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Alternatively, this software may be distributed under the terms of the
|
|
|
|
* GNU General Public License ("GPL") version 2 as published by the Free
|
|
|
|
* Software Foundation.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
|
|
|
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
|
|
|
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
|
|
|
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
|
|
|
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
|
|
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
|
|
|
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
|
|
|
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
2006-01-03 02:04:38 +08:00
|
|
|
* POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
*/
|
|
|
|
|
2015-01-09 15:27:04 +08:00
|
|
|
#include <net/sock.h>
|
2006-01-03 02:04:38 +08:00
|
|
|
#include "core.h"
|
|
|
|
#include "msg.h"
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
#include "addr.h"
|
|
|
|
#include "name_table.h"
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2014-06-26 09:41:35 +08:00
|
|
|
#define MAX_FORWARD_SIZE 1024
|
|
|
|
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
static unsigned int align(unsigned int i)
|
2010-05-11 22:30:18 +08:00
|
|
|
{
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
return (i + 3) & ~3u;
|
2010-05-11 22:30:18 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:01 +08:00
|
|
|
/**
|
|
|
|
* tipc_buf_acquire - creates a TIPC message buffer
|
|
|
|
* @size: message size (including TIPC header)
|
|
|
|
*
|
|
|
|
* Returns a new buffer with data pointers set to the specified size.
|
|
|
|
*
|
|
|
|
* NOTE: Headroom is reserved to allow prepending of a data link header.
|
|
|
|
* There may also be unrequested tailroom present at the buffer's end.
|
|
|
|
*/
|
|
|
|
struct sk_buff *tipc_buf_acquire(u32 size)
|
|
|
|
{
|
|
|
|
struct sk_buff *skb;
|
|
|
|
unsigned int buf_size = (BUF_HEADROOM + size + 3) & ~3u;
|
|
|
|
|
|
|
|
skb = alloc_skb_fclone(buf_size, GFP_ATOMIC);
|
|
|
|
if (skb) {
|
|
|
|
skb_reserve(skb, BUF_HEADROOM);
|
|
|
|
skb_put(skb, size);
|
|
|
|
skb->next = NULL;
|
|
|
|
}
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
2015-02-05 21:36:36 +08:00
|
|
|
void tipc_msg_init(u32 own_node, struct tipc_msg *m, u32 user, u32 type,
|
|
|
|
u32 hsize, u32 dnode)
|
2010-05-11 22:30:18 +08:00
|
|
|
{
|
|
|
|
memset(m, 0, hsize);
|
|
|
|
msg_set_version(m);
|
|
|
|
msg_set_user(m, user);
|
|
|
|
msg_set_hdr_sz(m, hsize);
|
|
|
|
msg_set_size(m, hsize);
|
2015-02-05 21:36:36 +08:00
|
|
|
msg_set_prevnode(m, own_node);
|
2010-05-11 22:30:18 +08:00
|
|
|
msg_set_type(m, type);
|
2014-08-23 06:09:06 +08:00
|
|
|
if (hsize > SHORT_H_SIZE) {
|
2015-02-05 21:36:36 +08:00
|
|
|
msg_set_orignode(m, own_node);
|
|
|
|
msg_set_destnode(m, dnode);
|
2014-08-23 06:09:06 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-02-05 21:36:36 +08:00
|
|
|
struct sk_buff *tipc_msg_create(uint user, uint type,
|
2015-01-09 15:27:10 +08:00
|
|
|
uint hdr_sz, uint data_sz, u32 dnode,
|
|
|
|
u32 onode, u32 dport, u32 oport, int errcode)
|
2014-08-23 06:09:06 +08:00
|
|
|
{
|
|
|
|
struct tipc_msg *msg;
|
|
|
|
struct sk_buff *buf;
|
|
|
|
|
|
|
|
buf = tipc_buf_acquire(hdr_sz + data_sz);
|
|
|
|
if (unlikely(!buf))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
msg = buf_msg(buf);
|
2015-02-05 21:36:36 +08:00
|
|
|
tipc_msg_init(onode, msg, user, type, hdr_sz, dnode);
|
2014-08-23 06:09:06 +08:00
|
|
|
msg_set_size(msg, hdr_sz + data_sz);
|
|
|
|
msg_set_origport(msg, oport);
|
|
|
|
msg_set_destport(msg, dport);
|
|
|
|
msg_set_errcode(msg, errcode);
|
|
|
|
if (hdr_sz > SHORT_H_SIZE) {
|
|
|
|
msg_set_orignode(msg, onode);
|
|
|
|
msg_set_destnode(msg, dnode);
|
|
|
|
}
|
|
|
|
return buf;
|
2010-05-11 22:30:18 +08:00
|
|
|
}
|
|
|
|
|
2014-05-14 17:39:12 +08:00
|
|
|
/* tipc_buf_append(): Append a buffer to the fragment list of another buffer
|
2014-07-06 01:44:13 +08:00
|
|
|
* @*headbuf: in: NULL for first frag, otherwise value returned from prev call
|
|
|
|
* out: set when successful non-complete reassembly, otherwise NULL
|
|
|
|
* @*buf: in: the buffer to append. Always defined
|
2014-10-30 13:58:51 +08:00
|
|
|
* out: head buf after successful complete reassembly, otherwise NULL
|
2014-07-06 01:44:13 +08:00
|
|
|
* Returns 1 when reassembly complete, otherwise 0
|
2014-05-14 17:39:12 +08:00
|
|
|
*/
|
|
|
|
int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf)
|
|
|
|
{
|
|
|
|
struct sk_buff *head = *headbuf;
|
|
|
|
struct sk_buff *frag = *buf;
|
|
|
|
struct sk_buff *tail;
|
2014-07-26 02:48:09 +08:00
|
|
|
struct tipc_msg *msg;
|
|
|
|
u32 fragid;
|
2014-05-14 17:39:12 +08:00
|
|
|
int delta;
|
2014-07-26 02:48:09 +08:00
|
|
|
bool headstolen;
|
2014-05-14 17:39:12 +08:00
|
|
|
|
2014-07-26 02:48:09 +08:00
|
|
|
if (!frag)
|
|
|
|
goto err;
|
|
|
|
|
|
|
|
msg = buf_msg(frag);
|
|
|
|
fragid = msg_type(msg);
|
|
|
|
frag->next = NULL;
|
2014-05-14 17:39:12 +08:00
|
|
|
skb_pull(frag, msg_hdr_sz(msg));
|
|
|
|
|
|
|
|
if (fragid == FIRST_FRAGMENT) {
|
2014-07-26 02:48:09 +08:00
|
|
|
if (unlikely(head))
|
|
|
|
goto err;
|
|
|
|
if (unlikely(skb_unclone(frag, GFP_ATOMIC)))
|
|
|
|
goto err;
|
2014-05-14 17:39:12 +08:00
|
|
|
head = *headbuf = frag;
|
|
|
|
skb_frag_list_init(head);
|
2014-07-26 02:48:09 +08:00
|
|
|
TIPC_SKB_CB(head)->tail = NULL;
|
2014-07-06 01:44:13 +08:00
|
|
|
*buf = NULL;
|
2014-05-14 17:39:12 +08:00
|
|
|
return 0;
|
|
|
|
}
|
2014-07-26 02:48:09 +08:00
|
|
|
|
2014-05-14 17:39:12 +08:00
|
|
|
if (!head)
|
2014-07-26 02:48:09 +08:00
|
|
|
goto err;
|
|
|
|
|
2014-05-14 17:39:12 +08:00
|
|
|
if (skb_try_coalesce(head, frag, &headstolen, &delta)) {
|
|
|
|
kfree_skb_partial(frag, headstolen);
|
|
|
|
} else {
|
2014-07-26 02:48:09 +08:00
|
|
|
tail = TIPC_SKB_CB(head)->tail;
|
2014-05-14 17:39:12 +08:00
|
|
|
if (!skb_has_frag_list(head))
|
|
|
|
skb_shinfo(head)->frag_list = frag;
|
|
|
|
else
|
|
|
|
tail->next = frag;
|
|
|
|
head->truesize += frag->truesize;
|
|
|
|
head->data_len += frag->len;
|
|
|
|
head->len += frag->len;
|
|
|
|
TIPC_SKB_CB(head)->tail = frag;
|
|
|
|
}
|
2014-07-26 02:48:09 +08:00
|
|
|
|
2014-05-14 17:39:12 +08:00
|
|
|
if (fragid == LAST_FRAGMENT) {
|
|
|
|
*buf = head;
|
|
|
|
TIPC_SKB_CB(head)->tail = NULL;
|
|
|
|
*headbuf = NULL;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
*buf = NULL;
|
|
|
|
return 0;
|
2014-07-26 02:48:09 +08:00
|
|
|
|
|
|
|
err:
|
2014-05-14 17:39:12 +08:00
|
|
|
pr_warn_ratelimited("Unable to build fragment list\n");
|
|
|
|
kfree_skb(*buf);
|
2014-07-06 01:44:13 +08:00
|
|
|
kfree_skb(*headbuf);
|
|
|
|
*buf = *headbuf = NULL;
|
2014-05-14 17:39:12 +08:00
|
|
|
return 0;
|
|
|
|
}
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
|
2015-03-14 04:08:06 +08:00
|
|
|
/* tipc_msg_validate - validate basic format of received message
|
|
|
|
*
|
|
|
|
* This routine ensures a TIPC message has an acceptable header, and at least
|
|
|
|
* as much data as the header indicates it should. The routine also ensures
|
|
|
|
* that the entire message header is stored in the main fragment of the message
|
|
|
|
* buffer, to simplify future access to message header fields.
|
|
|
|
*
|
|
|
|
* Note: Having extra info present in the message header or data areas is OK.
|
|
|
|
* TIPC will ignore the excess, under the assumption that it is optional info
|
|
|
|
* introduced by a later release of the protocol.
|
|
|
|
*/
|
|
|
|
bool tipc_msg_validate(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct tipc_msg *msg;
|
|
|
|
int msz, hsz;
|
|
|
|
|
|
|
|
if (unlikely(TIPC_SKB_CB(skb)->validated))
|
|
|
|
return true;
|
|
|
|
if (unlikely(!pskb_may_pull(skb, MIN_H_SIZE)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
hsz = msg_hdr_sz(buf_msg(skb));
|
|
|
|
if (unlikely(hsz < MIN_H_SIZE) || (hsz > MAX_H_SIZE))
|
|
|
|
return false;
|
|
|
|
if (unlikely(!pskb_may_pull(skb, hsz)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
msg = buf_msg(skb);
|
|
|
|
if (unlikely(msg_version(msg) != TIPC_VERSION))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
msz = msg_size(msg);
|
|
|
|
if (unlikely(msz < hsz))
|
|
|
|
return false;
|
|
|
|
if (unlikely((msz - hsz) > TIPC_MAX_USER_MSG_SIZE))
|
|
|
|
return false;
|
|
|
|
if (unlikely(skb->len < msz))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
TIPC_SKB_CB(skb)->validated = true;
|
|
|
|
return true;
|
|
|
|
}
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
|
|
|
|
/**
|
2014-07-17 08:41:03 +08:00
|
|
|
* tipc_msg_build - create buffer chain containing specified header and data
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
* @mhdr: Message header, to be prepended to data
|
2014-11-15 14:16:27 +08:00
|
|
|
* @m: User message
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
* @dsz: Total length of user data
|
|
|
|
* @pktmax: Max packet size that can be used
|
2014-11-26 11:41:55 +08:00
|
|
|
* @list: Buffer or chain of buffers to be returned to caller
|
|
|
|
*
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
* Returns message data size or errno: -ENOMEM, -EFAULT
|
|
|
|
*/
|
2015-02-05 21:36:36 +08:00
|
|
|
int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
|
2015-01-09 15:27:10 +08:00
|
|
|
int offset, int dsz, int pktmax, struct sk_buff_head *list)
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
{
|
|
|
|
int mhsz = msg_hdr_sz(mhdr);
|
|
|
|
int msz = mhsz + dsz;
|
|
|
|
int pktno = 1;
|
|
|
|
int pktsz;
|
|
|
|
int pktrem = pktmax;
|
|
|
|
int drem = dsz;
|
|
|
|
struct tipc_msg pkthdr;
|
2014-11-26 11:41:55 +08:00
|
|
|
struct sk_buff *skb;
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
char *pktpos;
|
|
|
|
int rc;
|
2014-11-26 11:41:55 +08:00
|
|
|
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
msg_set_size(mhdr, msz);
|
|
|
|
|
|
|
|
/* No fragmentation needed? */
|
|
|
|
if (likely(msz <= pktmax)) {
|
2014-11-26 11:41:55 +08:00
|
|
|
skb = tipc_buf_acquire(msz);
|
|
|
|
if (unlikely(!skb))
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
return -ENOMEM;
|
2015-01-09 15:27:04 +08:00
|
|
|
skb_orphan(skb);
|
2014-11-26 11:41:55 +08:00
|
|
|
__skb_queue_tail(list, skb);
|
|
|
|
skb_copy_to_linear_data(skb, mhdr, mhsz);
|
|
|
|
pktpos = skb->data + mhsz;
|
2014-11-29 04:52:29 +08:00
|
|
|
if (copy_from_iter(pktpos, dsz, &m->msg_iter) == dsz)
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
return dsz;
|
|
|
|
rc = -EFAULT;
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Prepare reusable fragment header */
|
2015-02-05 21:36:36 +08:00
|
|
|
tipc_msg_init(msg_prevnode(mhdr), &pkthdr, MSG_FRAGMENTER,
|
|
|
|
FIRST_FRAGMENT, INT_H_SIZE, msg_destnode(mhdr));
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
msg_set_size(&pkthdr, pktmax);
|
|
|
|
msg_set_fragm_no(&pkthdr, pktno);
|
|
|
|
|
|
|
|
/* Prepare first fragment */
|
2014-11-26 11:41:55 +08:00
|
|
|
skb = tipc_buf_acquire(pktmax);
|
|
|
|
if (!skb)
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
return -ENOMEM;
|
2015-01-09 15:27:04 +08:00
|
|
|
skb_orphan(skb);
|
2014-11-26 11:41:55 +08:00
|
|
|
__skb_queue_tail(list, skb);
|
|
|
|
pktpos = skb->data;
|
|
|
|
skb_copy_to_linear_data(skb, &pkthdr, INT_H_SIZE);
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
pktpos += INT_H_SIZE;
|
|
|
|
pktrem -= INT_H_SIZE;
|
2014-11-26 11:41:55 +08:00
|
|
|
skb_copy_to_linear_data_offset(skb, INT_H_SIZE, mhdr, mhsz);
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
pktpos += mhsz;
|
|
|
|
pktrem -= mhsz;
|
|
|
|
|
|
|
|
do {
|
|
|
|
if (drem < pktrem)
|
|
|
|
pktrem = drem;
|
|
|
|
|
2014-11-29 04:52:29 +08:00
|
|
|
if (copy_from_iter(pktpos, pktrem, &m->msg_iter) != pktrem) {
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
rc = -EFAULT;
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
drem -= pktrem;
|
|
|
|
|
|
|
|
if (!drem)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* Prepare new fragment: */
|
|
|
|
if (drem < (pktmax - INT_H_SIZE))
|
|
|
|
pktsz = drem + INT_H_SIZE;
|
|
|
|
else
|
|
|
|
pktsz = pktmax;
|
2014-11-26 11:41:55 +08:00
|
|
|
skb = tipc_buf_acquire(pktsz);
|
|
|
|
if (!skb) {
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
rc = -ENOMEM;
|
|
|
|
goto error;
|
|
|
|
}
|
2015-01-09 15:27:04 +08:00
|
|
|
skb_orphan(skb);
|
2014-11-26 11:41:55 +08:00
|
|
|
__skb_queue_tail(list, skb);
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
msg_set_type(&pkthdr, FRAGMENT);
|
|
|
|
msg_set_size(&pkthdr, pktsz);
|
|
|
|
msg_set_fragm_no(&pkthdr, ++pktno);
|
2014-11-26 11:41:55 +08:00
|
|
|
skb_copy_to_linear_data(skb, &pkthdr, INT_H_SIZE);
|
|
|
|
pktpos = skb->data + INT_H_SIZE;
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
pktrem = pktsz - INT_H_SIZE;
|
|
|
|
|
|
|
|
} while (1);
|
2014-11-26 11:41:55 +08:00
|
|
|
msg_set_type(buf_msg(skb), LAST_FRAGMENT);
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
return dsz;
|
|
|
|
error:
|
2014-11-26 11:41:55 +08:00
|
|
|
__skb_queue_purge(list);
|
|
|
|
__skb_queue_head_init(list);
|
tipc: introduce direct iovec to buffer chain fragmentation function
Fragmentation at message sending is currently performed in two
places in link.c, depending on whether data to be transmitted
is delivered in the form of an iovec or as a big sk_buff. Those
functions are also tightly entangled with the send functions
that are using them.
We now introduce a re-entrant, standalone function, tipc_msg_build2(),
that builds a packet chain directly from an iovec. Each fragment is
sized according to the MTU value given by the caller, and is prepended
with a correctly built fragment header, when needed. The function is
independent from who is calling and where the chain will be delivered,
as long as the caller is able to indicate a correct MTU.
The function is tested, but not called by anybody yet. Since it is
incompatible with the existing tipc_msg_build(), and we cannot yet
remove that function, we have given it a temporary name.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:34 +08:00
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
/**
|
|
|
|
* tipc_msg_bundle(): Append contents of a buffer to tail of an existing one
|
2014-11-26 11:41:52 +08:00
|
|
|
* @list: the buffer chain of the existing buffer ("bundle")
|
|
|
|
* @skb: buffer to be appended
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
* @mtu: max allowable size for the bundle buffer
|
|
|
|
* Consumes buffer if successful
|
|
|
|
* Returns true if bundling could be performed, otherwise false
|
|
|
|
*/
|
2014-11-26 11:41:52 +08:00
|
|
|
bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu)
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
{
|
2014-11-26 11:41:52 +08:00
|
|
|
struct sk_buff *bskb = skb_peek_tail(list);
|
|
|
|
struct tipc_msg *bmsg = buf_msg(bskb);
|
|
|
|
struct tipc_msg *msg = buf_msg(skb);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
unsigned int bsz = msg_size(bmsg);
|
|
|
|
unsigned int msz = msg_size(msg);
|
|
|
|
u32 start = align(bsz);
|
|
|
|
u32 max = mtu - INT_H_SIZE;
|
|
|
|
u32 pad = start - bsz;
|
|
|
|
|
|
|
|
if (likely(msg_user(msg) == MSG_FRAGMENTER))
|
|
|
|
return false;
|
|
|
|
if (unlikely(msg_user(msg) == CHANGEOVER_PROTOCOL))
|
|
|
|
return false;
|
|
|
|
if (unlikely(msg_user(msg) == BCAST_PROTOCOL))
|
|
|
|
return false;
|
|
|
|
if (likely(msg_user(bmsg) != MSG_BUNDLER))
|
|
|
|
return false;
|
2014-11-26 11:41:52 +08:00
|
|
|
if (likely(!TIPC_SKB_CB(bskb)->bundling))
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
return false;
|
2014-11-26 11:41:52 +08:00
|
|
|
if (unlikely(skb_tailroom(bskb) < (pad + msz)))
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
return false;
|
|
|
|
if (unlikely(max < (start + msz)))
|
|
|
|
return false;
|
|
|
|
|
2014-11-26 11:41:52 +08:00
|
|
|
skb_put(bskb, pad + msz);
|
|
|
|
skb_copy_to_linear_data_offset(bskb, start, skb->data, msz);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
msg_set_size(bmsg, start + msz);
|
|
|
|
msg_set_msgcnt(bmsg, msg_msgcnt(bmsg) + 1);
|
2014-11-26 11:41:52 +08:00
|
|
|
kfree_skb(skb);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
tipc: resolve race problem at unicast message reception
TIPC handles message cardinality and sequencing at the link layer,
before passing messages upwards to the destination sockets. During the
upcall from link to socket no locks are held. It is therefore possible,
and we see it happen occasionally, that messages arriving in different
threads and delivered in sequence still bypass each other before they
reach the destination socket. This must not happen, since it violates
the sequentiality guarantee.
We solve this by adding a new input buffer queue to the link structure.
Arriving messages are added safely to the tail of that queue by the
link, while the head of the queue is consumed, also safely, by the
receiving socket. Sequentiality is secured per socket by only allowing
buffers to be dequeued inside the socket lock. Since there may be multiple
simultaneous readers of the queue, we use a 'filter' parameter to reduce
the risk that they peek the same buffer from the queue, hence also
reducing the risk of contention on the receiving socket locks.
This solves the sequentiality problem, and seems to cause no measurable
performance degradation.
A nice side effect of this change is that lock handling in the functions
tipc_rcv() and tipc_bcast_rcv() now becomes uniform, something that
will enable future simplifications of those functions.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:41 +08:00
|
|
|
/**
|
|
|
|
* tipc_msg_extract(): extract bundled inner packet from buffer
|
|
|
|
* @skb: linear outer buffer, to be extracted from.
|
|
|
|
* @iskb: extracted inner buffer, to be returned
|
|
|
|
* @pos: position of msg to be extracted. Returns with pointer of next msg
|
|
|
|
* Consumes outer buffer when last packet extracted
|
|
|
|
* Returns true when when there is an extracted buffer, otherwise false
|
|
|
|
*/
|
|
|
|
bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos)
|
|
|
|
{
|
|
|
|
struct tipc_msg *msg = buf_msg(skb);
|
|
|
|
int imsz;
|
|
|
|
struct tipc_msg *imsg = (struct tipc_msg *)(msg_data(msg) + *pos);
|
|
|
|
|
|
|
|
/* Is there space left for shortest possible message? */
|
|
|
|
if (*pos > (msg_data_sz(msg) - SHORT_H_SIZE))
|
|
|
|
goto none;
|
|
|
|
imsz = msg_size(imsg);
|
|
|
|
|
|
|
|
/* Is there space left for current message ? */
|
|
|
|
if ((*pos + imsz) > msg_data_sz(msg))
|
|
|
|
goto none;
|
|
|
|
*iskb = tipc_buf_acquire(imsz);
|
|
|
|
if (!*iskb)
|
|
|
|
goto none;
|
|
|
|
skb_copy_to_linear_data(*iskb, imsg, imsz);
|
|
|
|
*pos += align(imsz);
|
|
|
|
return true;
|
|
|
|
none:
|
|
|
|
kfree_skb(skb);
|
|
|
|
*iskb = NULL;
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
/**
|
|
|
|
* tipc_msg_make_bundle(): Create bundle buf and append message to its tail
|
2014-11-26 11:41:52 +08:00
|
|
|
* @list: the buffer chain
|
|
|
|
* @skb: buffer to be appended and replaced
|
|
|
|
* @mtu: max allowable size for the bundle buffer, inclusive header
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
* @dnode: destination node for message. (Not always present in header)
|
|
|
|
* Replaces buffer if successful
|
2014-10-30 13:58:51 +08:00
|
|
|
* Returns true if success, otherwise false
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
*/
|
2015-02-05 21:36:36 +08:00
|
|
|
bool tipc_msg_make_bundle(struct sk_buff_head *list,
|
2015-01-09 15:27:10 +08:00
|
|
|
struct sk_buff *skb, u32 mtu, u32 dnode)
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
{
|
2014-11-26 11:41:52 +08:00
|
|
|
struct sk_buff *bskb;
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
struct tipc_msg *bmsg;
|
2014-11-26 11:41:52 +08:00
|
|
|
struct tipc_msg *msg = buf_msg(skb);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
u32 msz = msg_size(msg);
|
|
|
|
u32 max = mtu - INT_H_SIZE;
|
|
|
|
|
|
|
|
if (msg_user(msg) == MSG_FRAGMENTER)
|
|
|
|
return false;
|
|
|
|
if (msg_user(msg) == CHANGEOVER_PROTOCOL)
|
|
|
|
return false;
|
|
|
|
if (msg_user(msg) == BCAST_PROTOCOL)
|
|
|
|
return false;
|
|
|
|
if (msz > (max / 2))
|
|
|
|
return false;
|
|
|
|
|
2014-11-26 11:41:52 +08:00
|
|
|
bskb = tipc_buf_acquire(max);
|
|
|
|
if (!bskb)
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
return false;
|
|
|
|
|
2014-11-26 11:41:52 +08:00
|
|
|
skb_trim(bskb, INT_H_SIZE);
|
|
|
|
bmsg = buf_msg(bskb);
|
2015-02-05 21:36:36 +08:00
|
|
|
tipc_msg_init(msg_prevnode(msg), bmsg, MSG_BUNDLER, 0,
|
|
|
|
INT_H_SIZE, dnode);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
msg_set_seqno(bmsg, msg_seqno(msg));
|
|
|
|
msg_set_ack(bmsg, msg_ack(msg));
|
|
|
|
msg_set_bcast_ack(bmsg, msg_bcast_ack(msg));
|
2014-11-26 11:41:52 +08:00
|
|
|
TIPC_SKB_CB(bskb)->bundling = true;
|
|
|
|
__skb_queue_tail(list, bskb);
|
|
|
|
return tipc_msg_bundle(list, skb, mtu);
|
tipc: introduce send functions for chained buffers in link
The current link implementation provides several different transmit
functions, depending on the characteristics of the message to be
sent: if it is an iovec or an sk_buff, if it needs fragmentation or
not, if the caller holds the node_lock or not. The permutation of
these options gives us an unwanted amount of unnecessarily complex
code.
As a first step towards simplifying the send path for all messages,
we introduce two new send functions at link level, tipc_link_xmit2()
and __tipc_link_xmit2(). The former looks up a link to the message
destination, and if one is found, it grabs the node lock and calls
the second function, which works exclusively inside the node lock
protection. If no link is found, and the destination is on the same
node, it delivers the message directly to the local destination
socket.
The new functions take a buffer chain where all packet headers are
already prepared, and the correct MTU has been used. These two
functions will later replace all other link-level transmit functions.
The functions are not backwards compatible, so we have added them
as new functions with temporary names. They are tested, but have no
users yet. Those will be added later in this series.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:32 +08:00
|
|
|
}
|
2014-06-26 09:41:35 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* tipc_msg_reverse(): swap source and destination addresses and add error code
|
|
|
|
* @buf: buffer containing message to be reversed
|
|
|
|
* @dnode: return value: node where to send message after reversal
|
|
|
|
* @err: error code to be set in message
|
|
|
|
* Consumes buffer if failure
|
|
|
|
* Returns true if success, otherwise false
|
|
|
|
*/
|
2015-02-05 21:36:36 +08:00
|
|
|
bool tipc_msg_reverse(u32 own_addr, struct sk_buff *buf, u32 *dnode,
|
2015-01-09 15:27:10 +08:00
|
|
|
int err)
|
2014-06-26 09:41:35 +08:00
|
|
|
{
|
|
|
|
struct tipc_msg *msg = buf_msg(buf);
|
|
|
|
uint imp = msg_importance(msg);
|
|
|
|
struct tipc_msg ohdr;
|
|
|
|
uint rdsz = min_t(uint, msg_data_sz(msg), MAX_FORWARD_SIZE);
|
|
|
|
|
2014-06-26 09:41:41 +08:00
|
|
|
if (skb_linearize(buf))
|
2014-06-26 09:41:35 +08:00
|
|
|
goto exit;
|
2014-06-26 09:41:41 +08:00
|
|
|
if (msg_dest_droppable(msg))
|
|
|
|
goto exit;
|
|
|
|
if (msg_errcode(msg))
|
2014-06-26 09:41:35 +08:00
|
|
|
goto exit;
|
|
|
|
|
|
|
|
memcpy(&ohdr, msg, msg_hdr_sz(msg));
|
2014-06-26 09:41:41 +08:00
|
|
|
imp = min_t(uint, imp + 1, TIPC_CRITICAL_IMPORTANCE);
|
|
|
|
if (msg_isdata(msg))
|
|
|
|
msg_set_importance(msg, imp);
|
2014-06-26 09:41:35 +08:00
|
|
|
msg_set_errcode(msg, err);
|
|
|
|
msg_set_origport(msg, msg_destport(&ohdr));
|
|
|
|
msg_set_destport(msg, msg_origport(&ohdr));
|
2015-02-05 21:36:36 +08:00
|
|
|
msg_set_prevnode(msg, own_addr);
|
2014-06-26 09:41:35 +08:00
|
|
|
if (!msg_short(msg)) {
|
|
|
|
msg_set_orignode(msg, msg_destnode(&ohdr));
|
|
|
|
msg_set_destnode(msg, msg_orignode(&ohdr));
|
|
|
|
}
|
|
|
|
msg_set_size(msg, msg_hdr_sz(msg) + rdsz);
|
|
|
|
skb_trim(buf, msg_size(msg));
|
|
|
|
skb_orphan(buf);
|
|
|
|
*dnode = msg_orignode(&ohdr);
|
|
|
|
return true;
|
|
|
|
exit:
|
|
|
|
kfree_skb(buf);
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
*dnode = 0;
|
2014-06-26 09:41:35 +08:00
|
|
|
return false;
|
|
|
|
}
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
|
|
|
|
/**
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
* tipc_msg_lookup_dest(): try to find new destination for named message
|
|
|
|
* @skb: the buffer containing the message.
|
|
|
|
* @dnode: return value: next-hop node, if destination found
|
|
|
|
* @err: return value: error code to use, if message to be rejected
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
* Does not consume buffer
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
* Returns true if a destination is found, false otherwise
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
*/
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb,
|
|
|
|
u32 *dnode, int *err)
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
{
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
struct tipc_msg *msg = buf_msg(skb);
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
u32 dport;
|
|
|
|
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
if (!msg_isdata(msg))
|
|
|
|
return false;
|
|
|
|
if (!msg_named(msg))
|
|
|
|
return false;
|
|
|
|
*err = -TIPC_ERR_NO_NAME;
|
|
|
|
if (skb_linearize(skb))
|
|
|
|
return false;
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
if (msg_reroute_cnt(msg) > 0)
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
return false;
|
2015-01-09 15:27:10 +08:00
|
|
|
*dnode = addr_domain(net, msg_lookup_scope(msg));
|
2015-01-09 15:27:09 +08:00
|
|
|
dport = tipc_nametbl_translate(net, msg_nametype(msg),
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
msg_nameinst(msg), dnode);
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
if (!dport)
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
return false;
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
msg_incr_reroute_cnt(msg);
|
|
|
|
msg_set_destnode(msg, *dnode);
|
|
|
|
msg_set_destport(msg, dport);
|
tipc: split up function tipc_msg_eval()
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-05 21:36:39 +08:00
|
|
|
*err = TIPC_OK;
|
|
|
|
return true;
|
tipc: introduce message evaluation function
When a message arrives in a node and finds no destination
socket, we may need to drop it, reject it, or forward it after
a secondary destination lookup. The latter two cases currently
results in a code path that is perceived as complex, because it
follows a deep call chain via obscure functions such as
net_route_named_msg() and net_route_msg().
We now introduce a function, tipc_msg_eval(), that takes the
decision about whether such a message should be rejected or
forwarded, but leaves it to the caller to actually perform
the indicated action.
If the decision is 'reject', it is still the task of the recently
introduced function tipc_msg_reverse() to take the final decision
about whether the message is rejectable or not. In the latter case
it drops the message.
As a result of this change, we can finally eliminate the function
net_route_named_msg(), and hence become independent of net_route_msg().
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-26 09:41:36 +08:00
|
|
|
}
|
2014-07-17 08:41:00 +08:00
|
|
|
|
|
|
|
/* tipc_msg_reassemble() - clone a buffer chain of fragments and
|
|
|
|
* reassemble the clones into one message
|
|
|
|
*/
|
2014-11-26 11:41:55 +08:00
|
|
|
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list)
|
2014-07-17 08:41:00 +08:00
|
|
|
{
|
2014-11-26 11:41:55 +08:00
|
|
|
struct sk_buff *skb;
|
|
|
|
struct sk_buff *frag = NULL;
|
2014-07-17 08:41:00 +08:00
|
|
|
struct sk_buff *head = NULL;
|
|
|
|
int hdr_sz;
|
|
|
|
|
|
|
|
/* Copy header if single buffer */
|
2014-11-26 11:41:55 +08:00
|
|
|
if (skb_queue_len(list) == 1) {
|
|
|
|
skb = skb_peek(list);
|
|
|
|
hdr_sz = skb_headroom(skb) + msg_hdr_sz(buf_msg(skb));
|
|
|
|
return __pskb_copy(skb, hdr_sz, GFP_ATOMIC);
|
2014-07-17 08:41:00 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Clone all fragments and reassemble */
|
2014-11-26 11:41:55 +08:00
|
|
|
skb_queue_walk(list, skb) {
|
|
|
|
frag = skb_clone(skb, GFP_ATOMIC);
|
2014-07-17 08:41:00 +08:00
|
|
|
if (!frag)
|
|
|
|
goto error;
|
|
|
|
frag->next = NULL;
|
|
|
|
if (tipc_buf_append(&head, &frag))
|
|
|
|
break;
|
|
|
|
if (!head)
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
return frag;
|
|
|
|
error:
|
|
|
|
pr_warn("Failed do clone local mcast rcv buffer\n");
|
|
|
|
kfree_skb(head);
|
|
|
|
return NULL;
|
|
|
|
}
|