linux_dsm_epyc7002/net/tipc/discover.c

418 lines
13 KiB
C
Raw Normal View History

/*
* net/tipc/discover.c
*
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
* Copyright (c) 2003-2006, 2014-2018, Ericsson AB
* Copyright (c) 2005-2006, 2010-2011, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* 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.
*
* 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
* POSSIBILITY OF SUCH DAMAGE.
*/
#include "core.h"
#include "node.h"
#include "discover.h"
/* min delay during bearer start up */
#define TIPC_DISC_INIT msecs_to_jiffies(125)
/* max delay if bearer has no links */
#define TIPC_DISC_FAST msecs_to_jiffies(1000)
/* max delay if bearer has links */
#define TIPC_DISC_SLOW msecs_to_jiffies(60000)
/* indicates no timer in use */
#define TIPC_DISC_INACTIVE 0xffffffff
/**
* struct tipc_discoverer - information about an ongoing link setup request
* @bearer_id: identity of bearer issuing requests
* @net: network namespace instance
* @dest: destination address for request messages
* @domain: network domain to which links can be established
* @num_nodes: number of nodes currently discovered (i.e. with an active link)
* @lock: spinlock for controlling access to requests
* @skb: request message to be (repeatedly) sent
* @timer: timer governing period between requests
* @timer_intv: current interval between requests (in ms)
*/
struct tipc_discoverer {
u32 bearer_id;
struct tipc_media_addr dest;
struct net *net;
u32 domain;
int num_nodes;
spinlock_t lock;
struct sk_buff *skb;
struct timer_list timer;
unsigned long timer_intv;
};
/**
* tipc_disc_init_msg - initialize a link setup message
* @net: the applicable net namespace
* @type: message type (request or response)
* @b: ptr to bearer issuing message
*/
static void tipc_disc_init_msg(struct net *net, struct sk_buff *skb,
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
u32 mtyp, struct tipc_bearer *b)
{
struct tipc_net *tn = tipc_net(net);
u32 dest_domain = b->domain;
struct tipc_msg *hdr;
hdr = buf_msg(skb);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
tipc_msg_init(tn->trial_addr, hdr, LINK_CONFIG, mtyp,
MAX_H_SIZE, dest_domain);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
msg_set_size(hdr, MAX_H_SIZE + NODE_ID_LEN);
msg_set_non_seq(hdr, 1);
msg_set_node_sig(hdr, tn->random);
msg_set_node_capabilities(hdr, TIPC_NODE_CAPABILITIES);
msg_set_dest_domain(hdr, dest_domain);
msg_set_bc_netid(hdr, tn->net_id);
b->media->addr2msg(msg_media_addr(hdr), &b->addr);
tipc: improve throughput between nodes in netns Currently, TIPC transports intra-node user data messages directly socket to socket, hence shortcutting all the lower layers of the communication stack. This gives TIPC very good intra node performance, both regarding throughput and latency. We now introduce a similar mechanism for TIPC data traffic across network namespaces located in the same kernel. On the send path, the call chain is as always accompanied by the sending node's network name space pointer. However, once we have reliably established that the receiving node is represented by a namespace on the same host, we just replace the namespace pointer with the receiving node/namespace's ditto, and follow the regular socket receive patch though the receiving node. This technique gives us a throughput similar to the node internal throughput, several times larger than if we let the traffic go though the full network stacks. As a comparison, max throughput for 64k messages is four times larger than TCP throughput for the same type of traffic. To meet any security concerns, the following should be noted. - All nodes joining a cluster are supposed to have been be certified and authenticated by mechanisms outside TIPC. This is no different for nodes/namespaces on the same host; they have to auto discover each other using the attached interfaces, and establish links which are supervised via the regular link monitoring mechanism. Hence, a kernel local node has no other way to join a cluster than any other node, and have to obey to policies set in the IP or device layers of the stack. - Only when a sender has established with 100% certainty that the peer node is located in a kernel local namespace does it choose to let user data messages, and only those, take the crossover path to the receiving node/namespace. - If the receiving node/namespace is removed, its namespace pointer is invalidated at all peer nodes, and their neighbor link monitoring will eventually note that this node is gone. - To ensure the "100% certainty" criteria, and prevent any possible spoofing, received discovery messages must contain a proof that the sender knows a common secret. We use the hash mix of the sending node/namespace for this purpose, since it can be accessed directly by all other namespaces in the kernel. Upon reception of a discovery message, the receiver checks this proof against all the local namespaces'hash_mix:es. If it finds a match, that, along with a matching node id and cluster id, this is deemed sufficient proof that the peer node in question is in a local namespace, and a wormhole can be opened. - We should also consider that TIPC is intended to be a cluster local IPC mechanism (just like e.g. UNIX sockets) rather than a network protocol, and hence we think it can justified to allow it to shortcut the lower protocol layers. Regarding traceability, we should notice that since commit 6c9081a3915d ("tipc: add loopback device tracking") it is possible to follow the node internal packet flow by just activating tcpdump on the loopback interface. This will be true even for this mechanism; by activating tcpdump on the involved nodes' loopback interfaces their inter-name space messaging can easily be tracked. v2: - update 'net' pointer when node left/rejoined v3: - grab read/write lock when using node ref obj v4: - clone traffics between netns to loopback Suggested-by: Jon Maloy <jon.maloy@ericsson.com> Acked-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-10-29 07:51:21 +07:00
msg_set_peer_net_hash(hdr, tipc_net_hash_mixes(net, tn->random));
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
msg_set_node_id(hdr, tipc_own_id(net));
}
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
static void tipc_disc_msg_xmit(struct net *net, u32 mtyp, u32 dst,
u32 src, u32 sugg_addr,
struct tipc_media_addr *maddr,
struct tipc_bearer *b)
{
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
struct tipc_msg *hdr;
struct sk_buff *skb;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
skb = tipc_buf_acquire(MAX_H_SIZE + NODE_ID_LEN, GFP_ATOMIC);
if (!skb)
return;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
hdr = buf_msg(skb);
tipc_disc_init_msg(net, skb, mtyp, b);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
msg_set_sugg_node_addr(hdr, sugg_addr);
msg_set_dest_domain(hdr, dst);
tipc_bearer_xmit_skb(net, b->identity, skb, maddr);
}
/**
* disc_dupl_alert - issue node address duplication alert
* @b: pointer to bearer detecting duplication
* @node_addr: duplicated node address
* @media_addr: media address advertised by duplicated node
*/
static void disc_dupl_alert(struct tipc_bearer *b, u32 node_addr,
struct tipc_media_addr *media_addr)
{
char media_addr_str[64];
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 11:50:23 +07:00
tipc_media_addr_printf(media_addr_str, sizeof(media_addr_str),
media_addr);
pr_warn("Duplicate %x using %s seen on <%s>\n", node_addr,
media_addr_str, b->name);
}
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
/* tipc_disc_addr_trial(): - handle an address uniqueness trial from peer
* Returns true if message should be dropped by caller, i.e., if it is a
* trial message or we are inside trial period. Otherwise false.
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
*/
static bool tipc_disc_addr_trial_msg(struct tipc_discoverer *d,
struct tipc_media_addr *maddr,
struct tipc_bearer *b,
u32 dst, u32 src,
u32 sugg_addr,
u8 *peer_id,
int mtyp)
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
{
struct net *net = d->net;
struct tipc_net *tn = tipc_net(net);
bool trial = time_before(jiffies, tn->addr_trial_end);
u32 self = tipc_own_addr(net);
if (mtyp == DSC_TRIAL_FAIL_MSG) {
if (!trial)
return true;
/* Ignore if somebody else already gave new suggestion */
if (dst != tn->trial_addr)
return true;
/* Otherwise update trial address and restart trial period */
tn->trial_addr = sugg_addr;
msg_set_prevnode(buf_msg(d->skb), sugg_addr);
tn->addr_trial_end = jiffies + msecs_to_jiffies(1000);
return true;
}
/* Apply trial address if we just left trial period */
if (!trial && !self) {
tipc: fix lockdep warning when reinitilaizing sockets We get the following warning: [ 47.926140] 32-bit node address hash set to 2010a0a [ 47.927202] [ 47.927433] ================================ [ 47.928050] WARNING: inconsistent lock state [ 47.928661] 4.19.0+ #37 Tainted: G E [ 47.929346] -------------------------------- [ 47.929954] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. [ 47.930116] swapper/3/0 [HC0[0]:SC1[3]:HE1:SE0] takes: [ 47.930116] 00000000af8bc31e (&(&ht->lock)->rlock){+.?.}, at: rhashtable_walk_enter+0x36/0xb0 [ 47.930116] {SOFTIRQ-ON-W} state was registered at: [ 47.930116] _raw_spin_lock+0x29/0x60 [ 47.930116] rht_deferred_worker+0x556/0x810 [ 47.930116] process_one_work+0x1f5/0x540 [ 47.930116] worker_thread+0x64/0x3e0 [ 47.930116] kthread+0x112/0x150 [ 47.930116] ret_from_fork+0x3a/0x50 [ 47.930116] irq event stamp: 14044 [ 47.930116] hardirqs last enabled at (14044): [<ffffffff9a07fbba>] __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] hardirqs last disabled at (14043): [<ffffffff9a07fb81>] __local_bh_enable_ip+0x41/0xf0 [ 47.938117] softirqs last enabled at (14028): [<ffffffff9a0803ee>] irq_enter+0x5e/0x60 [ 47.938117] softirqs last disabled at (14029): [<ffffffff9a0804a5>] irq_exit+0xb5/0xc0 [ 47.938117] [ 47.938117] other info that might help us debug this: [ 47.938117] Possible unsafe locking scenario: [ 47.938117] [ 47.938117] CPU0 [ 47.938117] ---- [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] <Interrupt> [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] [ 47.938117] *** DEADLOCK *** [ 47.938117] [ 47.938117] 2 locks held by swapper/3/0: [ 47.938117] #0: 0000000062c64f90 ((&d->timer)){+.-.}, at: call_timer_fn+0x5/0x280 [ 47.938117] #1: 00000000ee39619c (&(&d->lock)->rlock){+.-.}, at: tipc_disc_timeout+0xc8/0x540 [tipc] [ 47.938117] [ 47.938117] stack backtrace: [ 47.938117] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G E 4.19.0+ #37 [ 47.938117] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 47.938117] Call Trace: [ 47.938117] <IRQ> [ 47.938117] dump_stack+0x5e/0x8b [ 47.938117] print_usage_bug+0x1ed/0x1ff [ 47.938117] mark_lock+0x5b5/0x630 [ 47.938117] __lock_acquire+0x4c0/0x18f0 [ 47.938117] ? lock_acquire+0xa6/0x180 [ 47.938117] lock_acquire+0xa6/0x180 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] _raw_spin_lock+0x29/0x60 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] rhashtable_walk_enter+0x36/0xb0 [ 47.938117] tipc_sk_reinit+0xb0/0x410 [tipc] [ 47.938117] ? mark_held_locks+0x6f/0x90 [ 47.938117] ? __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] ? lockdep_hardirqs_on+0x20/0x1a0 [ 47.938117] tipc_net_finalize+0xbf/0x180 [tipc] [ 47.938117] tipc_disc_timeout+0x509/0x540 [tipc] [ 47.938117] ? call_timer_fn+0x5/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] call_timer_fn+0xa1/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] run_timer_softirq+0x1f2/0x4d0 [ 47.938117] __do_softirq+0xfc/0x413 [ 47.938117] irq_exit+0xb5/0xc0 [ 47.938117] smp_apic_timer_interrupt+0xac/0x210 [ 47.938117] apic_timer_interrupt+0xf/0x20 [ 47.938117] </IRQ> [ 47.938117] RIP: 0010:default_idle+0x1c/0x140 [ 47.938117] Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 41 54 55 53 65 8b 2d d8 2b 74 65 0f 1f 44 00 00 e8 c6 2c 8b ff fb f4 <65> 8b 2d c5 2b 74 65 0f 1f 44 00 00 5b 5d 41 5c c3 65 8b 05 b4 2b [ 47.938117] RSP: 0018:ffffaf6ac0207ec8 EFLAGS: 00000206 ORIG_RAX: ffffffffffffff13 [ 47.938117] RAX: ffff8f5b3735e200 RBX: 0000000000000003 RCX: 0000000000000001 [ 47.938117] RDX: 0000000000000001 RSI: 0000000000000001 RDI: ffff8f5b3735e200 [ 47.938117] RBP: 0000000000000003 R08: 0000000000000001 R09: 0000000000000000 [ 47.938117] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 [ 47.938117] R13: 0000000000000000 R14: ffff8f5b3735e200 R15: ffff8f5b3735e200 [ 47.938117] ? default_idle+0x1a/0x140 [ 47.938117] do_idle+0x1bc/0x280 [ 47.938117] cpu_startup_entry+0x19/0x20 [ 47.938117] start_secondary+0x187/0x1c0 [ 47.938117] secondary_startup_64+0xa4/0xb0 The reason seems to be that tipc_net_finalize()->tipc_sk_reinit() is calling the function rhashtable_walk_enter() within a timer interrupt. We fix this by executing tipc_net_finalize() in work queue context. Acked-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>
2018-11-17 04:55:04 +07:00
tipc_sched_net_finalize(net, tn->trial_addr);
msg_set_prevnode(buf_msg(d->skb), tn->trial_addr);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
msg_set_type(buf_msg(d->skb), DSC_REQ_MSG);
}
/* Accept regular link requests/responses only after trial period */
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
if (mtyp != DSC_TRIAL_MSG)
return trial;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
sugg_addr = tipc_node_try_addr(net, peer_id, src);
if (sugg_addr)
tipc_disc_msg_xmit(net, DSC_TRIAL_FAIL_MSG, src,
self, sugg_addr, maddr, b);
return true;
}
/**
* tipc_disc_rcv - handle incoming discovery message (request or response)
* @net: applicable net namespace
* @skb: buffer containing message
* @b: bearer that message arrived on
*/
void tipc_disc_rcv(struct net *net, struct sk_buff *skb,
struct tipc_bearer *b)
{
struct tipc_net *tn = tipc_net(net);
struct tipc_msg *hdr = buf_msg(skb);
u16 caps = msg_node_capabilities(hdr);
2018-03-23 02:42:48 +07:00
bool legacy = tn->legacy_addr_format;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
u32 sugg = msg_sugg_node_addr(hdr);
u32 signature = msg_node_sig(hdr);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
u8 peer_id[NODE_ID_LEN] = {0,};
u32 dst = msg_dest_domain(hdr);
u32 net_id = msg_bc_netid(hdr);
struct tipc_media_addr maddr;
u32 src = msg_prevnode(hdr);
u32 mtyp = msg_type(hdr);
bool dupl_addr = false;
bool respond = false;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
u32 self;
int err;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
skb_linearize(skb);
hdr = buf_msg(skb);
if (caps & TIPC_NODE_ID128)
memcpy(peer_id, msg_node_id(hdr), NODE_ID_LEN);
else
sprintf(peer_id, "%x", src);
err = b->media->msg2addr(b, &maddr, msg_media_addr(hdr));
kfree_skb(skb);
if (err || maddr.broadcast) {
pr_warn_ratelimited("Rcv corrupt discovery message\n");
return;
}
/* Ignore discovery messages from own node */
if (!memcmp(&maddr, &b->addr, sizeof(maddr)))
return;
if (net_id != tn->net_id)
return;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
if (tipc_disc_addr_trial_msg(b->disc, &maddr, b, dst,
src, sugg, peer_id, mtyp))
return;
self = tipc_own_addr(net);
/* Message from somebody using this node's address */
if (in_own_node(net, src)) {
disc_dupl_alert(b, self, &maddr);
return;
}
2018-03-23 02:42:48 +07:00
if (!tipc_in_scope(legacy, dst, self))
return;
if (!tipc_in_scope(legacy, b->domain, src))
return;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
tipc_node_check_dest(net, src, peer_id, b, caps, signature,
tipc: improve throughput between nodes in netns Currently, TIPC transports intra-node user data messages directly socket to socket, hence shortcutting all the lower layers of the communication stack. This gives TIPC very good intra node performance, both regarding throughput and latency. We now introduce a similar mechanism for TIPC data traffic across network namespaces located in the same kernel. On the send path, the call chain is as always accompanied by the sending node's network name space pointer. However, once we have reliably established that the receiving node is represented by a namespace on the same host, we just replace the namespace pointer with the receiving node/namespace's ditto, and follow the regular socket receive patch though the receiving node. This technique gives us a throughput similar to the node internal throughput, several times larger than if we let the traffic go though the full network stacks. As a comparison, max throughput for 64k messages is four times larger than TCP throughput for the same type of traffic. To meet any security concerns, the following should be noted. - All nodes joining a cluster are supposed to have been be certified and authenticated by mechanisms outside TIPC. This is no different for nodes/namespaces on the same host; they have to auto discover each other using the attached interfaces, and establish links which are supervised via the regular link monitoring mechanism. Hence, a kernel local node has no other way to join a cluster than any other node, and have to obey to policies set in the IP or device layers of the stack. - Only when a sender has established with 100% certainty that the peer node is located in a kernel local namespace does it choose to let user data messages, and only those, take the crossover path to the receiving node/namespace. - If the receiving node/namespace is removed, its namespace pointer is invalidated at all peer nodes, and their neighbor link monitoring will eventually note that this node is gone. - To ensure the "100% certainty" criteria, and prevent any possible spoofing, received discovery messages must contain a proof that the sender knows a common secret. We use the hash mix of the sending node/namespace for this purpose, since it can be accessed directly by all other namespaces in the kernel. Upon reception of a discovery message, the receiver checks this proof against all the local namespaces'hash_mix:es. If it finds a match, that, along with a matching node id and cluster id, this is deemed sufficient proof that the peer node in question is in a local namespace, and a wormhole can be opened. - We should also consider that TIPC is intended to be a cluster local IPC mechanism (just like e.g. UNIX sockets) rather than a network protocol, and hence we think it can justified to allow it to shortcut the lower protocol layers. Regarding traceability, we should notice that since commit 6c9081a3915d ("tipc: add loopback device tracking") it is possible to follow the node internal packet flow by just activating tcpdump on the loopback interface. This will be true even for this mechanism; by activating tcpdump on the involved nodes' loopback interfaces their inter-name space messaging can easily be tracked. v2: - update 'net' pointer when node left/rejoined v3: - grab read/write lock when using node ref obj v4: - clone traffics between netns to loopback Suggested-by: Jon Maloy <jon.maloy@ericsson.com> Acked-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-10-29 07:51:21 +07:00
msg_peer_net_hash(hdr), &maddr, &respond,
&dupl_addr);
if (dupl_addr)
disc_dupl_alert(b, src, &maddr);
if (!respond)
return;
if (mtyp != DSC_REQ_MSG)
return;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
tipc_disc_msg_xmit(net, DSC_RESP_MSG, src, self, 0, &maddr, b);
}
/* tipc_disc_add_dest - increment set of discovered nodes
*/
void tipc_disc_add_dest(struct tipc_discoverer *d)
{
spin_lock_bh(&d->lock);
d->num_nodes++;
spin_unlock_bh(&d->lock);
}
/* tipc_disc_remove_dest - decrement set of discovered nodes
*/
void tipc_disc_remove_dest(struct tipc_discoverer *d)
{
int intv, num;
spin_lock_bh(&d->lock);
d->num_nodes--;
num = d->num_nodes;
intv = d->timer_intv;
if (!num && (intv == TIPC_DISC_INACTIVE || intv > TIPC_DISC_FAST)) {
d->timer_intv = TIPC_DISC_INIT;
mod_timer(&d->timer, jiffies + d->timer_intv);
}
spin_unlock_bh(&d->lock);
}
/* tipc_disc_timeout - send a periodic link setup request
* Called whenever a link setup request timer associated with a bearer expires.
* - Keep doubling time between sent request until limit is reached;
* - Hold at fast polling rate if we don't have any associated nodes
* - Otherwise hold at slow polling rate
*/
static void tipc_disc_timeout(struct timer_list *t)
{
struct tipc_discoverer *d = from_timer(d, t, timer);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
struct tipc_net *tn = tipc_net(d->net);
struct tipc_media_addr maddr;
struct sk_buff *skb = NULL;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
struct net *net = d->net;
u32 bearer_id;
spin_lock_bh(&d->lock);
/* Stop searching if only desired node has been found */
if (tipc_node(d->domain) && d->num_nodes) {
d->timer_intv = TIPC_DISC_INACTIVE;
goto exit;
}
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
tipc: fix lockdep warning when reinitilaizing sockets We get the following warning: [ 47.926140] 32-bit node address hash set to 2010a0a [ 47.927202] [ 47.927433] ================================ [ 47.928050] WARNING: inconsistent lock state [ 47.928661] 4.19.0+ #37 Tainted: G E [ 47.929346] -------------------------------- [ 47.929954] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. [ 47.930116] swapper/3/0 [HC0[0]:SC1[3]:HE1:SE0] takes: [ 47.930116] 00000000af8bc31e (&(&ht->lock)->rlock){+.?.}, at: rhashtable_walk_enter+0x36/0xb0 [ 47.930116] {SOFTIRQ-ON-W} state was registered at: [ 47.930116] _raw_spin_lock+0x29/0x60 [ 47.930116] rht_deferred_worker+0x556/0x810 [ 47.930116] process_one_work+0x1f5/0x540 [ 47.930116] worker_thread+0x64/0x3e0 [ 47.930116] kthread+0x112/0x150 [ 47.930116] ret_from_fork+0x3a/0x50 [ 47.930116] irq event stamp: 14044 [ 47.930116] hardirqs last enabled at (14044): [<ffffffff9a07fbba>] __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] hardirqs last disabled at (14043): [<ffffffff9a07fb81>] __local_bh_enable_ip+0x41/0xf0 [ 47.938117] softirqs last enabled at (14028): [<ffffffff9a0803ee>] irq_enter+0x5e/0x60 [ 47.938117] softirqs last disabled at (14029): [<ffffffff9a0804a5>] irq_exit+0xb5/0xc0 [ 47.938117] [ 47.938117] other info that might help us debug this: [ 47.938117] Possible unsafe locking scenario: [ 47.938117] [ 47.938117] CPU0 [ 47.938117] ---- [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] <Interrupt> [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] [ 47.938117] *** DEADLOCK *** [ 47.938117] [ 47.938117] 2 locks held by swapper/3/0: [ 47.938117] #0: 0000000062c64f90 ((&d->timer)){+.-.}, at: call_timer_fn+0x5/0x280 [ 47.938117] #1: 00000000ee39619c (&(&d->lock)->rlock){+.-.}, at: tipc_disc_timeout+0xc8/0x540 [tipc] [ 47.938117] [ 47.938117] stack backtrace: [ 47.938117] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G E 4.19.0+ #37 [ 47.938117] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 47.938117] Call Trace: [ 47.938117] <IRQ> [ 47.938117] dump_stack+0x5e/0x8b [ 47.938117] print_usage_bug+0x1ed/0x1ff [ 47.938117] mark_lock+0x5b5/0x630 [ 47.938117] __lock_acquire+0x4c0/0x18f0 [ 47.938117] ? lock_acquire+0xa6/0x180 [ 47.938117] lock_acquire+0xa6/0x180 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] _raw_spin_lock+0x29/0x60 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] rhashtable_walk_enter+0x36/0xb0 [ 47.938117] tipc_sk_reinit+0xb0/0x410 [tipc] [ 47.938117] ? mark_held_locks+0x6f/0x90 [ 47.938117] ? __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] ? lockdep_hardirqs_on+0x20/0x1a0 [ 47.938117] tipc_net_finalize+0xbf/0x180 [tipc] [ 47.938117] tipc_disc_timeout+0x509/0x540 [tipc] [ 47.938117] ? call_timer_fn+0x5/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] call_timer_fn+0xa1/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] run_timer_softirq+0x1f2/0x4d0 [ 47.938117] __do_softirq+0xfc/0x413 [ 47.938117] irq_exit+0xb5/0xc0 [ 47.938117] smp_apic_timer_interrupt+0xac/0x210 [ 47.938117] apic_timer_interrupt+0xf/0x20 [ 47.938117] </IRQ> [ 47.938117] RIP: 0010:default_idle+0x1c/0x140 [ 47.938117] Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 41 54 55 53 65 8b 2d d8 2b 74 65 0f 1f 44 00 00 e8 c6 2c 8b ff fb f4 <65> 8b 2d c5 2b 74 65 0f 1f 44 00 00 5b 5d 41 5c c3 65 8b 05 b4 2b [ 47.938117] RSP: 0018:ffffaf6ac0207ec8 EFLAGS: 00000206 ORIG_RAX: ffffffffffffff13 [ 47.938117] RAX: ffff8f5b3735e200 RBX: 0000000000000003 RCX: 0000000000000001 [ 47.938117] RDX: 0000000000000001 RSI: 0000000000000001 RDI: ffff8f5b3735e200 [ 47.938117] RBP: 0000000000000003 R08: 0000000000000001 R09: 0000000000000000 [ 47.938117] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 [ 47.938117] R13: 0000000000000000 R14: ffff8f5b3735e200 R15: ffff8f5b3735e200 [ 47.938117] ? default_idle+0x1a/0x140 [ 47.938117] do_idle+0x1bc/0x280 [ 47.938117] cpu_startup_entry+0x19/0x20 [ 47.938117] start_secondary+0x187/0x1c0 [ 47.938117] secondary_startup_64+0xa4/0xb0 The reason seems to be that tipc_net_finalize()->tipc_sk_reinit() is calling the function rhashtable_walk_enter() within a timer interrupt. We fix this by executing tipc_net_finalize() in work queue context. Acked-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>
2018-11-17 04:55:04 +07:00
/* Did we just leave trial period ? */
if (!time_before(jiffies, tn->addr_trial_end) && !tipc_own_addr(net)) {
mod_timer(&d->timer, jiffies + TIPC_DISC_INIT);
spin_unlock_bh(&d->lock);
tipc_sched_net_finalize(net, tn->trial_addr);
return;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
}
/* Adjust timeout interval according to discovery phase */
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
if (time_before(jiffies, tn->addr_trial_end)) {
d->timer_intv = TIPC_DISC_INIT;
} else {
d->timer_intv *= 2;
if (d->num_nodes && d->timer_intv > TIPC_DISC_SLOW)
d->timer_intv = TIPC_DISC_SLOW;
else if (!d->num_nodes && d->timer_intv > TIPC_DISC_FAST)
d->timer_intv = TIPC_DISC_FAST;
tipc: fix lockdep warning when reinitilaizing sockets We get the following warning: [ 47.926140] 32-bit node address hash set to 2010a0a [ 47.927202] [ 47.927433] ================================ [ 47.928050] WARNING: inconsistent lock state [ 47.928661] 4.19.0+ #37 Tainted: G E [ 47.929346] -------------------------------- [ 47.929954] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. [ 47.930116] swapper/3/0 [HC0[0]:SC1[3]:HE1:SE0] takes: [ 47.930116] 00000000af8bc31e (&(&ht->lock)->rlock){+.?.}, at: rhashtable_walk_enter+0x36/0xb0 [ 47.930116] {SOFTIRQ-ON-W} state was registered at: [ 47.930116] _raw_spin_lock+0x29/0x60 [ 47.930116] rht_deferred_worker+0x556/0x810 [ 47.930116] process_one_work+0x1f5/0x540 [ 47.930116] worker_thread+0x64/0x3e0 [ 47.930116] kthread+0x112/0x150 [ 47.930116] ret_from_fork+0x3a/0x50 [ 47.930116] irq event stamp: 14044 [ 47.930116] hardirqs last enabled at (14044): [<ffffffff9a07fbba>] __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] hardirqs last disabled at (14043): [<ffffffff9a07fb81>] __local_bh_enable_ip+0x41/0xf0 [ 47.938117] softirqs last enabled at (14028): [<ffffffff9a0803ee>] irq_enter+0x5e/0x60 [ 47.938117] softirqs last disabled at (14029): [<ffffffff9a0804a5>] irq_exit+0xb5/0xc0 [ 47.938117] [ 47.938117] other info that might help us debug this: [ 47.938117] Possible unsafe locking scenario: [ 47.938117] [ 47.938117] CPU0 [ 47.938117] ---- [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] <Interrupt> [ 47.938117] lock(&(&ht->lock)->rlock); [ 47.938117] [ 47.938117] *** DEADLOCK *** [ 47.938117] [ 47.938117] 2 locks held by swapper/3/0: [ 47.938117] #0: 0000000062c64f90 ((&d->timer)){+.-.}, at: call_timer_fn+0x5/0x280 [ 47.938117] #1: 00000000ee39619c (&(&d->lock)->rlock){+.-.}, at: tipc_disc_timeout+0xc8/0x540 [tipc] [ 47.938117] [ 47.938117] stack backtrace: [ 47.938117] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G E 4.19.0+ #37 [ 47.938117] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 47.938117] Call Trace: [ 47.938117] <IRQ> [ 47.938117] dump_stack+0x5e/0x8b [ 47.938117] print_usage_bug+0x1ed/0x1ff [ 47.938117] mark_lock+0x5b5/0x630 [ 47.938117] __lock_acquire+0x4c0/0x18f0 [ 47.938117] ? lock_acquire+0xa6/0x180 [ 47.938117] lock_acquire+0xa6/0x180 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] _raw_spin_lock+0x29/0x60 [ 47.938117] ? rhashtable_walk_enter+0x36/0xb0 [ 47.938117] rhashtable_walk_enter+0x36/0xb0 [ 47.938117] tipc_sk_reinit+0xb0/0x410 [tipc] [ 47.938117] ? mark_held_locks+0x6f/0x90 [ 47.938117] ? __local_bh_enable_ip+0x7a/0xf0 [ 47.938117] ? lockdep_hardirqs_on+0x20/0x1a0 [ 47.938117] tipc_net_finalize+0xbf/0x180 [tipc] [ 47.938117] tipc_disc_timeout+0x509/0x540 [tipc] [ 47.938117] ? call_timer_fn+0x5/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] call_timer_fn+0xa1/0x280 [ 47.938117] ? tipc_disc_msg_xmit.isra.19+0xa0/0xa0 [tipc] [ 47.938117] run_timer_softirq+0x1f2/0x4d0 [ 47.938117] __do_softirq+0xfc/0x413 [ 47.938117] irq_exit+0xb5/0xc0 [ 47.938117] smp_apic_timer_interrupt+0xac/0x210 [ 47.938117] apic_timer_interrupt+0xf/0x20 [ 47.938117] </IRQ> [ 47.938117] RIP: 0010:default_idle+0x1c/0x140 [ 47.938117] Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 41 54 55 53 65 8b 2d d8 2b 74 65 0f 1f 44 00 00 e8 c6 2c 8b ff fb f4 <65> 8b 2d c5 2b 74 65 0f 1f 44 00 00 5b 5d 41 5c c3 65 8b 05 b4 2b [ 47.938117] RSP: 0018:ffffaf6ac0207ec8 EFLAGS: 00000206 ORIG_RAX: ffffffffffffff13 [ 47.938117] RAX: ffff8f5b3735e200 RBX: 0000000000000003 RCX: 0000000000000001 [ 47.938117] RDX: 0000000000000001 RSI: 0000000000000001 RDI: ffff8f5b3735e200 [ 47.938117] RBP: 0000000000000003 R08: 0000000000000001 R09: 0000000000000000 [ 47.938117] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 [ 47.938117] R13: 0000000000000000 R14: ffff8f5b3735e200 R15: ffff8f5b3735e200 [ 47.938117] ? default_idle+0x1a/0x140 [ 47.938117] do_idle+0x1bc/0x280 [ 47.938117] cpu_startup_entry+0x19/0x20 [ 47.938117] start_secondary+0x187/0x1c0 [ 47.938117] secondary_startup_64+0xa4/0xb0 The reason seems to be that tipc_net_finalize()->tipc_sk_reinit() is calling the function rhashtable_walk_enter() within a timer interrupt. We fix this by executing tipc_net_finalize() in work queue context. Acked-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>
2018-11-17 04:55:04 +07:00
msg_set_type(buf_msg(d->skb), DSC_REQ_MSG);
msg_set_prevnode(buf_msg(d->skb), tn->trial_addr);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
}
mod_timer(&d->timer, jiffies + d->timer_intv);
memcpy(&maddr, &d->dest, sizeof(maddr));
skb = skb_clone(d->skb, GFP_ATOMIC);
bearer_id = d->bearer_id;
exit:
spin_unlock_bh(&d->lock);
if (skb)
tipc_bearer_xmit_skb(net, bearer_id, skb, &maddr);
}
/**
* tipc_disc_create - create object to send periodic link setup requests
* @net: the applicable net namespace
* @b: ptr to bearer issuing requests
* @dest: destination address for request messages
* @dest_domain: network domain to which links can be established
*
* Returns 0 if successful, otherwise -errno.
*/
int tipc_disc_create(struct net *net, struct tipc_bearer *b,
struct tipc_media_addr *dest, struct sk_buff **skb)
{
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
struct tipc_net *tn = tipc_net(net);
struct tipc_discoverer *d;
d = kmalloc(sizeof(*d), GFP_ATOMIC);
if (!d)
return -ENOMEM;
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
d->skb = tipc_buf_acquire(MAX_H_SIZE + NODE_ID_LEN, GFP_ATOMIC);
if (!d->skb) {
kfree(d);
return -ENOMEM;
}
tipc_disc_init_msg(net, d->skb, DSC_REQ_MSG, b);
tipc: handle collisions of 32-bit node address hash values When a 32-bit node address is generated from a 128-bit identifier, there is a risk of collisions which must be discovered and handled. We do this as follows: - We don't apply the generated address immediately to the node, but do instead initiate a 1 sec trial period to allow other cluster members to discover and handle such collisions. - During the trial period the node periodically sends out a new type of message, DSC_TRIAL_MSG, using broadcast or emulated broadcast, to all the other nodes in the cluster. - When a node is receiving such a message, it must check that the presented 32-bit identifier either is unused, or was used by the very same peer in a previous session. In both cases it accepts the request by not responding to it. - If it finds that the same node has been up before using a different address, it responds with a DSC_TRIAL_FAIL_MSG containing that address. - If it finds that the address has already been taken by some other node, it generates a new, unused address and returns it to the requester. - During the trial period the requesting node must always be prepared to accept a failure message, i.e., a message where a peer suggests a different (or equal) address to the one tried. In those cases it must apply the suggested value as trial address and restart the trial period. This algorithm ensures that in the vast majority of cases a node will have the same address before and after a reboot. If a legacy user configures the address explicitly, there will be no trial period and messages, so this protocol addition is completely backwards compatible. Acked-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>
2018-03-23 02:42:51 +07:00
/* Do we need an address trial period first ? */
if (!tipc_own_addr(net)) {
tn->addr_trial_end = jiffies + msecs_to_jiffies(1000);
msg_set_type(buf_msg(d->skb), DSC_TRIAL_MSG);
}
memcpy(&d->dest, dest, sizeof(*dest));
d->net = net;
d->bearer_id = b->identity;
d->domain = b->domain;
d->num_nodes = 0;
d->timer_intv = TIPC_DISC_INIT;
spin_lock_init(&d->lock);
timer_setup(&d->timer, tipc_disc_timeout, 0);
mod_timer(&d->timer, jiffies + d->timer_intv);
b->disc = d;
*skb = skb_clone(d->skb, GFP_ATOMIC);
return 0;
}
/**
* tipc_disc_delete - destroy object sending periodic link setup requests
* @d: ptr to link duest structure
*/
void tipc_disc_delete(struct tipc_discoverer *d)
{
del_timer_sync(&d->timer);
kfree_skb(d->skb);
kfree(d);
}
/**
* tipc_disc_reset - reset object to send periodic link setup requests
* @net: the applicable net namespace
* @b: ptr to bearer issuing requests
* @dest_domain: network domain to which links can be established
*/
void tipc_disc_reset(struct net *net, struct tipc_bearer *b)
{
struct tipc_discoverer *d = b->disc;
struct tipc_media_addr maddr;
struct sk_buff *skb;
spin_lock_bh(&d->lock);
tipc_disc_init_msg(net, d->skb, DSC_REQ_MSG, b);
d->net = net;
d->bearer_id = b->identity;
d->domain = b->domain;
d->num_nodes = 0;
d->timer_intv = TIPC_DISC_INIT;
memcpy(&maddr, &d->dest, sizeof(maddr));
mod_timer(&d->timer, jiffies + d->timer_intv);
skb = skb_clone(d->skb, GFP_ATOMIC);
spin_unlock_bh(&d->lock);
if (skb)
tipc_bearer_xmit_skb(net, b->identity, skb, &maddr);
}