Synchronous pernet_operations are not allowed anymore.
All are asynchronous. So, drop the structure member.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The netns deletion path does not need to wait for all net_devices
to be unregistered before dismantling rds_tcp state for the netns
(we are able to dismantle this state on module unload even when
all net_devices are active so there is no dependency here).
This patch removes code related to netdevice notifiers and
refactors all the code needed to dismantle rds_tcp state
into a ->exit callback for the pernet_operations used with
register_pernet_device().
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Reviewed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
rds_tcp_connection allocation/free management has the potential to be
called from __rds_conn_create after IRQs have been disabled, so
spin_[un]lock_bh cannot be used with rds_tcp_conn_lock.
Bottom-halves that need to synchronize for critical sections protected
by rds_tcp_conn_lock should instead use rds_destroy_pending() correctly.
Reported-by: syzbot+c68e51bb5e699d3f8d91@syzkaller.appspotmail.com
Fixes: ebeeb1ad9b ("rds: tcp: use rds_destroy_pending() to synchronize
netns/module teardown and rds connection/workq management")
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These pernet_operations create and destroy sysctl table
and listen socket. Also, exit method flushes global
workqueue and work. Everything looks per-net safe,
so we can mark them async.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Changes since v1:
Added changes in these files:
drivers/infiniband/hw/usnic/usnic_transport.c
drivers/staging/lustre/lnet/lnet/lib-socket.c
drivers/target/iscsi/iscsi_target_login.c
drivers/vhost/net.c
fs/dlm/lowcomms.c
fs/ocfs2/cluster/tcp.c
security/tomoyo/network.c
Before:
All these functions either return a negative error indicator,
or store length of sockaddr into "int *socklen" parameter
and return zero on success.
"int *socklen" parameter is awkward. For example, if caller does not
care, it still needs to provide on-stack storage for the value
it does not need.
None of the many FOO_getname() functions of various protocols
ever used old value of *socklen. They always just overwrite it.
This change drops this parameter, and makes all these functions, on success,
return length of sockaddr. It's always >= 0 and can be differentiated
from an error.
Tests in callers are changed from "if (err)" to "if (err < 0)", where needed.
rpc_sockname() lost "int buflen" parameter, since its only use was
to be passed to kernel_getsockname() as &buflen and subsequently
not used in any way.
Userspace API is not changed.
text data bss dec hex filename
30108430 2633624 873672 33615726 200ef6e vmlinux.before.o
30108109 2633612 873672 33615393 200ee21 vmlinux.o
Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
CC: David S. Miller <davem@davemloft.net>
CC: linux-kernel@vger.kernel.org
CC: netdev@vger.kernel.org
CC: linux-bluetooth@vger.kernel.org
CC: linux-decnet-user@lists.sourceforge.net
CC: linux-wireless@vger.kernel.org
CC: linux-rdma@vger.kernel.org
CC: linux-sctp@vger.kernel.org
CC: linux-nfs@vger.kernel.org
CC: linux-x25@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
An rds_connection can get added during netns deletion between lines 528
and 529 of
506 static void rds_tcp_kill_sock(struct net *net)
:
/* code to pull out all the rds_connections that should be destroyed */
:
528 spin_unlock_irq(&rds_tcp_conn_lock);
529 list_for_each_entry_safe(tc, _tc, &tmp_list, t_tcp_node)
530 rds_conn_destroy(tc->t_cpath->cp_conn);
Such an rds_connection would miss out the rds_conn_destroy()
loop (that cancels all pending work) and (if it was scheduled
after netns deletion) could trigger the use-after-free.
A similar race-window exists for the module unload path
in rds_tcp_exit -> rds_tcp_destroy_conns
Concurrency with netns deletion (rds_tcp_kill_sock()) must be handled
by checking check_net() before enqueuing new work or adding new
connections.
Concurrency with module-unload is handled by maintaining a module
specific flag that is set at the start of the module exit function,
and must be checked before enqueuing new work or adding new connections.
This commit refactors existing RDS_DESTROY_PENDING checks added by
commit 3db6e0d172 ("rds: use RCU to synchronize work-enqueue with
connection teardown") and consolidates all the concurrency checks
listed above into the function rds_destroy_pending().
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
en_rx_am.c was deleted in 'net-next' but had a bug fixed in it in
'net'.
The esp{4,6}_offload.c conflicts were overlapping changes.
The 'out' label is removed so we just return ERR_PTR(-EINVAL)
directly.
Signed-off-by: David S. Miller <davem@davemloft.net>
rds-tcp uses m_ack_seq to track the tcp ack# that indicates
that the peer has received a rds_message. The m_ack_seq is
used in rds_tcp_is_acked() to figure out when it is safe to
drop the rds_message from the RDS retransmit queue.
The m_ack_seq must be calculated as an offset from the right
edge of the in-flight tcp buffer, i.e., it should be based on
the ->write_seq, not the ->snd_nxt.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If kmem_cache_alloc() fails in the middle of the for() loop,
cleanup anything that might have been allocated so far.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit f10b4cff98 ("rds: tcp: atomically purge entries from
rds_tcp_conn_list during netns delete") adds the field t_tcp_detached,
but this needs to be initialized explicitly to false.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The rds_tcp_kill_sock() function parses the rds_tcp_conn_list
to find the rds_connection entries marked for deletion as part
of the netns deletion under the protection of the rds_tcp_conn_lock.
Since the rds_tcp_conn_list tracks rds_tcp_connections (which
have a 1:1 mapping with rds_conn_path), multiple tc entries in
the rds_tcp_conn_list will map to a single rds_connection, and will
be deleted as part of the rds_conn_destroy() operation that is
done outside the rds_tcp_conn_lock.
The rds_tcp_conn_list traversal done under the protection of
rds_tcp_conn_lock should not leave any doomed tc entries in
the list after the rds_tcp_conn_lock is released, else another
concurrently executiong netns delete (for a differnt netns) thread
may trip on these entries.
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 8edc3affc0 ("rds: tcp: Take explicit refcounts on struct net")
introduces a regression in rds-tcp netns cleanup. The cleanup_net(),
(and thus rds_tcp_dev_event notification) is only called from put_net()
when all netns refcounts go to 0, but this cannot happen if the
rds_connection itself is holding a c_net ref that it expects to
release in rds_tcp_kill_sock.
Instead, the rds_tcp_kill_sock callback should make sure to
tear down state carefully, ensuring that the socket teardown
is only done after all data-structures and workqs that depend
on it are quiesced.
The original motivation for commit 8edc3affc0 ("rds: tcp: Take explicit
refcounts on struct net") was to resolve a race condition reported by
syzkaller where workqs for tx/rx/connect were triggered after the
namespace was deleted. Those worker threads should have been
cancelled/flushed before socket tear-down and indeed,
rds_conn_path_destroy() does try to sequence this by doing
/* cancel cp_send_w */
/* cancel cp_recv_w */
/* flush cp_down_w */
/* free data structures */
Here the "flush cp_down_w" will trigger rds_conn_shutdown and thus
invoke rds_tcp_conn_path_shutdown() to close the tcp socket, so that
we ought to have satisfied the requirement that "socket-close is
done after all other dependent state is quiesced". However,
rds_conn_shutdown has a bug in that it *always* triggers the reconnect
workq (and if connection is successful, we always restart tx/rx
workqs so with the right timing, we risk the race conditions reported
by syzkaller).
Netns deletion is like module teardown- no need to restart a
reconnect in this case. We can use the c_destroy_in_prog bit
to avoid restarting the reconnect.
Fixes: 8edc3affc0 ("rds: tcp: Take explicit refcounts on struct net")
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A side-effect of Commit c14b036681 ("rds: tcp: set linger to 1
when unloading a rds-tcp") is that we always send a RST on the tcp
connection for rds_conn_destroy(), so rds_tcp_conn_paths_destroy()
is not needed any more and is removed in this patch.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We could end up executing rds_conn_shutdown before the rds_recv_worker
thread, then rds_conn_shutdown -> rds_tcp_conn_shutdown can do a
sock_release and set sock->sk to null, which may interleave in bad
ways with rds_recv_worker, e.g., it could result in:
"BUG: unable to handle kernel NULL pointer dereference at 0000000000000078"
[ffff881769f6fd70] release_sock at ffffffff815f337b
[ffff881769f6fd90] rds_tcp_recv at ffffffffa043c888 [rds_tcp]
[ffff881769f6fdb0] rds_recv_worker at ffffffffa04a4810 [rds]
[ffff881769f6fde0] process_one_work at ffffffff810a14c1
[ffff881769f6fe40] worker_thread at ffffffff810a1940
[ffff881769f6fec0] kthread at ffffffff810a6b1e
Also, do not enqueue any new shutdown workq items when the connection is
shutting down (this may happen for rds-tcp in softirq mode, if a FIN
or CLOSE is received while the modules is in the middle of an unload)
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit a93d01f577 ("RDS: TCP: avoid bad page reference in
rds_tcp_listen_data_ready") added the function
rds_tcp_listen_sock_def_readable() to handle the case when a
partially set-up acceptor socket drops into rds_tcp_listen_data_ready().
However, if the listen socket (rtn->rds_tcp_listen_sock) is itself going
through a tear-down via rds_tcp_listen_stop(), the (*ready)() will be
null and we would hit a panic of the form
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: (null)
:
? rds_tcp_listen_data_ready+0x59/0xb0 [rds_tcp]
tcp_data_queue+0x39d/0x5b0
tcp_rcv_established+0x2e5/0x660
tcp_v4_do_rcv+0x122/0x220
tcp_v4_rcv+0x8b7/0x980
:
In the above case, it is not fatal to encounter a NULL value for
ready- we should just drop the packet and let the flush of the
acceptor thread finish gracefully.
In general, the tear-down sequence for listen() and accept() socket
that is ensured by this commit is:
rtn->rds_tcp_listen_sock = NULL; /* prevent any new accepts */
In rds_tcp_listen_stop():
serialize with, and prevent, further callbacks using lock_sock()
flush rds_wq
flush acceptor workq
sock_release(listen socket)
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Order of initialization in rds_tcp_init needs to be done so
that resources are set up and destroyed in the correct synchronization
sequence with both the data path, as well as netns create/destroy
path. Specifically,
- we must call register_pernet_subsys and get the rds_tcp_netid
before calling register_netdevice_notifier, otherwise we risk
the sequence
1. register_netdevice_notifier sets up netdev notifier callback
2. rds_tcp_dev_event -> rds_tcp_kill_sock uses netid 0, and finds
the wrong rtn, resulting in a panic with string that is of the form:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000d
IP: rds_tcp_kill_sock+0x3a/0x1d0 [rds_tcp]
:
- the rds_tcp_incoming_slab kmem_cache must be initialized before the
datapath starts up. The latter can happen any time after the
pernet_subsys registration of rds_tcp_net_ops, whose -> init
function sets up the listen socket. If the rds_tcp_incoming_slab has
not been set up at that time, a panic of the form below may be
encountered
BUG: unable to handle kernel NULL pointer dereference at 0000000000000014
IP: kmem_cache_alloc+0x90/0x1c0
:
rds_tcp_data_recv+0x1e7/0x370 [rds_tcp]
tcp_read_sock+0x96/0x1c0
rds_tcp_recv_path+0x65/0x80 [rds_tcp]
:
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It is incorrect for the rds_connection to piggyback on the
sock_net() refcount for the netns because this gives rise to
a chicken-and-egg problem during rds_conn_destroy. Instead explicitly
take a ref on the net, and hold the netns down till the connection
tear-down is complete.
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function rds_trans_register always returns 0. As such, it is not
necessary to check the returned value.
Cc: Joe Jin <joe.jin@oracle.com>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the function register_netdevice_notifier fails, the memory
allocated by kmem_cache_create should be freed by the function
kmem_cache_destroy.
Cc: Joe Jin <joe.jin@oracle.com>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Acked-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Couple conflicts resolved here:
1) In the MACB driver, a bug fix to properly initialize the
RX tail pointer properly overlapped with some changes
to support variable sized rings.
2) In XGBE we had a "CONFIG_PM" --> "CONFIG_PM_SLEEP" fix
overlapping with a reorganization of the driver to support
ACPI, OF, as well as PCI variants of the chip.
3) In 'net' we had several probe error path bug fixes to the
stmmac driver, meanwhile a lot of this code was cleaned up
and reorganized in 'net-next'.
4) The cls_flower classifier obtained a helper function in
'net-next' called __fl_delete() and this overlapped with
Daniel Borkamann's bug fix to use RCU for object destruction
in 'net'. It also overlapped with Jiri's change to guard
the rhashtable_remove_fast() call with a check against
tc_skip_sw().
5) In mlx4, a revert bug fix in 'net' overlapped with some
unrelated changes in 'net-next'.
6) In geneve, a stale header pointer after pskb_expand_head()
bug fix in 'net' overlapped with a large reorganization of
the same code in 'net-next'. Since the 'net-next' code no
longer had the bug in question, there was nothing to do
other than to simply take the 'net-next' hunks.
Signed-off-by: David S. Miller <davem@davemloft.net>
If some error is encountered in rds_tcp_init_net, make sure to
unregister_netdevice_notifier(), else we could trigger a panic
later on, when the modprobe from a netns fails.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make struct pernet_operations::id unsigned.
There are 2 reasons to do so:
1)
This field is really an index into an zero based array and
thus is unsigned entity. Using negative value is out-of-bound
access by definition.
2)
On x86_64 unsigned 32-bit data which are mixed with pointers
via array indexing or offsets added or subtracted to pointers
are preffered to signed 32-bit data.
"int" being used as an array index needs to be sign-extended
to 64-bit before being used.
void f(long *p, int i)
{
g(p[i]);
}
roughly translates to
movsx rsi, esi
mov rdi, [rsi+...]
call g
MOVSX is 3 byte instruction which isn't necessary if the variable is
unsigned because x86_64 is zero extending by default.
Now, there is net_generic() function which, you guessed it right, uses
"int" as an array index:
static inline void *net_generic(const struct net *net, int id)
{
...
ptr = ng->ptr[id - 1];
...
}
And this function is used a lot, so those sign extensions add up.
Patch snipes ~1730 bytes on allyesconfig kernel (without all junk
messing with code generation):
add/remove: 0/0 grow/shrink: 70/598 up/down: 396/-2126 (-1730)
Unfortunately some functions actually grow bigger.
This is a semmingly random artefact of code generation with register
allocator being used differently. gcc decides that some variable
needs to live in new r8+ registers and every access now requires REX
prefix. Or it is shifted into r12, so [r12+0] addressing mode has to be
used which is longer than [r8]
However, overall balance is in negative direction:
add/remove: 0/0 grow/shrink: 70/598 up/down: 396/-2126 (-1730)
function old new delta
nfsd4_lock 3886 3959 +73
tipc_link_build_proto_msg 1096 1140 +44
mac80211_hwsim_new_radio 2776 2808 +32
tipc_mon_rcv 1032 1058 +26
svcauth_gss_legacy_init 1413 1429 +16
tipc_bcbase_select_primary 379 392 +13
nfsd4_exchange_id 1247 1260 +13
nfsd4_setclientid_confirm 782 793 +11
...
put_client_renew_locked 494 480 -14
ip_set_sockfn_get 730 716 -14
geneve_sock_add 829 813 -16
nfsd4_sequence_done 721 703 -18
nlmclnt_lookup_host 708 686 -22
nfsd4_lockt 1085 1063 -22
nfs_get_client 1077 1050 -27
tcf_bpf_init 1106 1076 -30
nfsd4_encode_fattr 5997 5930 -67
Total: Before=154856051, After=154854321, chg -0.00%
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The socket argument passed to rds_tcp_tc_info() is a PF_RDS socket,
so it is incorrect to report the address port info based on
rds_getname() as part of TCP state report.
Invoke inet_getname() for the t_sock associated with the
rds_tcp_connection instead.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use RDS probe-ping to compute how many paths may be used with
the peer, and to synchronously start the multiple paths. If mprds is
supported, hash outgoing traffic to one of multiple paths in rds_sendmsg()
when multipath RDS is supported by the transport.
CC: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some code duplication in rds_tcp_reset_callbacks() can be avoided
by having the function call rds_tcp_restore_callbacks() and
rds_tcp_set_callbacks().
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As the existing comments in rds_tcp_listen_data_ready() indicate,
it is possible under some race-windows to get to this function with the
accept() socket. If that happens, we could run into a sequence whereby
thread 1 thread 2
rds_tcp_accept_one() thread
sets up new_sock via ->accept().
The sk_user_data is now
sock_def_readable
data comes in for new_sock,
->sk_data_ready is called, and
we land in rds_tcp_listen_data_ready
rds_tcp_set_callbacks()
takes the sk_callback_lock and
sets up sk_user_data to be the cp
read_lock sk_callback_lock
ready = cp
unlock sk_callback_lock
page fault on ready
In the above sequence, we end up with a panic on a bad page reference
when trying to execute (*ready)(). Instead we need to call
sock_def_readable() safely, which is what this patch achieves.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/mellanox/mlx5/core/en.h
drivers/net/ethernet/mellanox/mlx5/core/en_main.c
drivers/net/usb/r8152.c
All three conflicts were overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
If register_pernet_subsys() fails, we shouldn't try to call
unregister_pernet_subsys().
Fixes: 467fa15356 ("RDS-TCP: Support multiple RDS-TCP listen endpoints, one per netns.")
Cc: stable@vger.kernel.org
Cc: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Acked-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds ->conn_path_connect callbacks in the rds_transport
that are used to set up a single connection path.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ->sk_user_data contains a pointer to the rds_conn_path
for the socket. Use this consistently in the rds_tcp_data_ready
callbacks to get the rds_conn_path for rds_recv_incoming.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The socket callbacks should all operate on a struct rds_conn_path,
in preparation for a MP capable RDS-TCP.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A single rds_connection may have multiple rds_conn_paths that have
to be carefully and correctly destroyed, for both rmmod and
netns-delete cases.
For both cases, we extract a single rds_tcp_connection for
each conn into a temporary list, and then invoke rds_conn_destroy()
which iteratively dismantles every path in the rds_connection.
For the netns deletion case, we additionally have to make sure
that we do not leave a socket in TIME_WAIT state, as this will
hold up the netns deletion. Thus we call rds_tcp_conn_paths_destroy()
to reset state quickly.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The struct rds_tcp_connection is the transport-specific private
data structure that tracks TCP information per rds_conn_path.
Modify this structure to have a back-pointer to the rds_conn_path
for which it is the ->cp_transport_data.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The c_passive bit is only intended for the IB transport and will
never be encountered in rds-tcp, so remove the dead logic that
predicates on this bit.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor code to avoid separate indirections for single-path
and multipath transports. All transports (both single and mp-capable)
will get a pointer to the rds_conn_path, and can trivially derive
the rds_connection from the ->cp_conn.
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes the following sparse warnings:
net/rds/tcp.c:59:5: warning:
symbol 'rds_tcp_min_sndbuf' was not declared. Should it be static?
net/rds/tcp.c:60:5: warning:
symbol 'rds_tcp_min_rcvbuf' was not declared. Should it be static?
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Acked-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit changes rds_conn_shutdown to take a rds_conn_path *
argument, allowing it to shutdown paths other than c_path[0] for
MP-capable transports.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation for multipath RDS, split the rds_connection
structure into a base structure, and a per-path struct rds_conn_path.
The base structure tracks information and locks common to all
paths. The workqs for send/recv/shutdown etc are tracked per
rds_conn_path. Thus the workq callbacks now work with rds_conn_path.
This commit allows for one rds_conn_path per rds_connection, and will
be extended into multiple conn_paths in subsequent commits.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The send path needs to be quiesced before resetting callbacks from
rds_tcp_accept_one(), and commit eb19284026 ("RDS:TCP: Synchronize
rds_tcp_accept_one with rds_send_xmit when resetting t_sock") achieves
this using the c_state and RDS_IN_XMIT bit following the pattern
used by rds_conn_shutdown(). However this leaves the possibility
of a race window as shown in the sequence below
take t_conn_lock in rds_tcp_conn_connect
send outgoing syn to peer
drop t_conn_lock in rds_tcp_conn_connect
incoming from peer triggers rds_tcp_accept_one, conn is
marked CONNECTING
wait for RDS_IN_XMIT to quiesce any rds_send_xmit threads
call rds_tcp_reset_callbacks
[.. race-window where incoming syn-ack can cause the conn
to be marked UP from rds_tcp_state_change ..]
lock_sock called from rds_tcp_reset_callbacks, and we set
t_sock to null
As soon as the conn is marked UP in the race-window above, rds_send_xmit()
threads will proceed to rds_tcp_xmit and may encounter a null-pointer
deref on the t_sock.
Given that rds_tcp_state_change() is invoked in softirq context, whereas
rds_tcp_reset_callbacks() is in workq context, and testing for RDS_IN_XMIT
after lock_sock could result in a deadlock with tcp_sendmsg, this
commit fixes the race by using a new c_state, RDS_TCP_RESETTING, which
will prevent a transition to RDS_CONN_UP from rds_tcp_state_change().
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When we switch a connection's sockets in rds_tcp_rest_callbacks,
any partially sent datagram must be retransmitted on the new
socket so that the receiver can correctly reassmble the RDS
datagram. Use rds_send_reset() which is designed for this purpose.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When rds_tcp_accept_one() has to replace the existing tcp socket
with a newer tcp socket (duelling-syn resolution), it must lock_sock()
to suppress the rds_tcp_data_recv() path while callbacks are being
changed. Also, existing RDS datagram reassembly state must be reset,
so that the next datagram on the new socket does not have corrupted
state. Similarly when resetting the newly accepted socket, appropriate
locks and synchronization is needed.
This commit ensures correct synchronization by invoking
kernel_sock_shutdown to reset a newly accepted sock, and by taking
appropriate lock_sock()s (for old and new sockets) when resetting
existing callbacks.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
An arbitration scheme for duelling SYNs is implemented as part of
commit 241b271952 ("RDS-TCP: Reset tcp callbacks if re-using an
outgoing socket in rds_tcp_accept_one()") which ensures that both nodes
involved will arrive at the same arbitration decision. However, this
needs to be synchronized with an outgoing SYN to be generated by
rds_tcp_conn_connect(). This commit achieves the synchronization
through the t_conn_lock mutex in struct rds_tcp_connection.
The rds_conn_state is checked in rds_tcp_conn_connect() after acquiring
the t_conn_lock mutex. A SYN is sent out only if the RDS connection is
not already UP (an UP would indicate that rds_tcp_accept_one() has
completed 3WH, so no SYN needs to be generated).
Similarly, the rds_conn_state is checked in rds_tcp_accept_one() after
acquiring the t_conn_lock mutex. The only acceptable states (to
allow continuation of the arbitration logic) are UP (i.e., outgoing SYN
was SYN-ACKed by peer after it sent us the SYN) or CONNECTING (we sent
outgoing SYN before we saw incoming SYN).
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is a race condition between rds_send_xmit -> rds_tcp_xmit
and the code that deals with resolution of duelling syns added
by commit 241b271952 ("RDS-TCP: Reset tcp callbacks if re-using an
outgoing socket in rds_tcp_accept_one()").
Specifically, we may end up derefencing a null pointer in rds_send_xmit
if we have the interleaving sequence:
rds_tcp_accept_one rds_send_xmit
conn is RDS_CONN_UP, so
invoke rds_tcp_xmit
tc = conn->c_transport_data
rds_tcp_restore_callbacks
/* reset t_sock */
null ptr deref from tc->t_sock
The race condition can be avoided without adding the overhead of
additional locking in the xmit path: have rds_tcp_accept_one wait
for rds_tcp_xmit threads to complete before resetting callbacks.
The synchronization can be done in the same manner as rds_conn_shutdown().
First set the rds_conn_state to something other than RDS_CONN_UP
(so that new threads cannot get into rds_tcp_xmit()), then wait for
RDS_IN_XMIT to be cleared in the conn->c_flags indicating that any
threads in rds_tcp_xmit are done.
Fixes: 241b271952 ("RDS-TCP: Reset tcp callbacks if re-using an
outgoing socket in rds_tcp_accept_one()")
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RDS_TCP_DEFAULT_BUFSIZE has been unused since commit 1edd6a14d2
("RDS-TCP: Do not bloat sndbuf/rcvbuf in rds_tcp_tune").
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add per-net sysctl tunables to set the size of sndbuf and
rcvbuf on the kernel tcp socket.
The tunables are added at /proc/sys/net/rds/tcp/rds_tcp_sndbuf
and /proc/sys/net/rds/tcp/rds_tcp_rcvbuf.
These values must be set before accept() or connect(),
and there may be an arbitrary number of existing rds-tcp
sockets when the tunable is modified. To make sure that all
connections in the netns pick up the same value for the tunable,
we reset existing rds-tcp connections in the netns, so that
they can reconnect with the new parameters.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using the value of RDS_TCP_DEFAULT_BUFSIZE (128K)
clobbers efficient use of TSO because it inflates the size_goal
that is computed in tcp_sendmsg/tcp_sendpage and skews packet
latency, and the default values for these parameters actually
results in significantly better performance.
In request-response tests using rds-stress with a packet size of
100K with 16 threads (test parameters -q 100000 -a 256 -t16 -d16)
between a single pair of IP addresses achieves a throughput of
6-8 Gbps. Without this patch, throughput maxes at 2-3 Gbps under
equivalent conditions on these platforms.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Register pernet subsys init/stop functions that will set up
and tear down per-net RDS-TCP listen endpoints. Unregister
pernet subusys functions on 'modprobe -r' to clean up these
end points.
Enable keepalive on both accept and connect socket endpoints.
The keepalive timer expiration will ensure that client socket
endpoints will be removed as appropriate from the netns when
an interface is removed from a namespace.
Register a device notifier callback that will clean up all
sockets (and thus avoid the need to wait for keepalive timeout)
when the loopback device is unregistered from the netns indicating
that the netns is getting deleted.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
