Jon Paul Maloy [Tue, 14 Jun 2016 00:46:22 +0000 (20:46 -0400)]
tipc: add neighbor monitoring framework
TIPC based clusters are by default set up with full-mesh link
connectivity between all nodes. Those links are expected to provide
a short failure detection time, by default set to 1500 ms. Because
of this, the background load for neighbor monitoring in an N-node
cluster increases with a factor N on each node, while the overall
monitoring traffic through the network infrastructure increases at
a ~(N * (N - 1)) rate. Experience has shown that such clusters don't
scale well beyond ~100 nodes unless we significantly increase failure
discovery tolerance.
This commit introduces a framework and an algorithm that drastically
reduces this background load, while basically maintaining the original
failure detection times across the whole cluster. Using this algorithm,
background load will now grow at a rate of ~(2 * sqrt(N)) per node, and
at ~(2 * N * sqrt(N)) in traffic overhead. As an example, each node will
now have to actively monitor 38 neighbors in a 400-node cluster, instead
of as before 399.
This "Overlapping Ring Supervision Algorithm" is completely distributed
and employs no centralized or coordinated state. It goes as follows:
- Each node makes up a linearly ascending, circular list of all its N
known neighbors, based on their TIPC node identity. This algorithm
must be the same on all nodes.
- The node then selects the next M = sqrt(N) - 1 nodes downstream from
itself in the list, and chooses to actively monitor those. This is
called its "local monitoring domain".
- It creates a domain record describing the monitoring domain, and
piggy-backs this in the data area of all neighbor monitoring messages
(LINK_PROTOCOL/STATE) leaving that node. This means that all nodes in
the cluster eventually (default within 400 ms) will learn about
its monitoring domain.
- Whenever a node discovers a change in its local domain, e.g., a node
has been added or has gone down, it creates and sends out a new
version of its node record to inform all neighbors about the change.
- A node receiving a domain record from anybody outside its local domain
matches this against its own list (which may not look the same), and
chooses to not actively monitor those members of the received domain
record that are also present in its own list. Instead, it relies on
indications from the direct monitoring nodes if an indirectly
monitored node has gone up or down. If a node is indicated lost, the
receiving node temporarily activates its own direct monitoring towards
that node in order to confirm, or not, that it is actually gone.
- Since each node is actively monitoring sqrt(N) downstream neighbors,
each node is also actively monitored by the same number of upstream
neighbors. This means that all non-direct monitoring nodes normally
will receive sqrt(N) indications that a node is gone.
- A major drawback with ring monitoring is how it handles failures that
cause massive network partitionings. If both a lost node and all its
direct monitoring neighbors are inside the lost partition, the nodes in
the remaining partition will never receive indications about the loss.
To overcome this, each node also chooses to actively monitor some
nodes outside its local domain. Those nodes are called remote domain
"heads", and are selected in such a way that no node in the cluster
will be more than two direct monitoring hops away. Because of this,
each node, apart from monitoring the member of its local domain, will
also typically monitor sqrt(N) remote head nodes.
- As an optimization, local list status, domain status and domain
records are marked with a generation number. This saves senders from
unnecessarily conveying unaltered domain records, and receivers from
performing unneeded re-adaptations of their node monitoring list, such
as re-assigning domain heads.
- As a measure of caution we have added the possibility to disable the
new algorithm through configuration. We do this by keeping a threshold
value for the cluster size; a cluster that grows beyond this value
will switch from full-mesh to ring monitoring, and vice versa when
it shrinks below the value. This means that if the threshold is set to
a value larger than any anticipated cluster size (default size is 32)
the new algorithm is effectively disabled. A patch set for altering the
threshold value and for listing the table contents will follow shortly.
- This change is fully 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>
David Ahern [Tue, 14 Jun 2016 00:14:12 +0000 (17:14 -0700)]
net: vrf: Update flags and features settings
1. Default VRF devices to not having a qdisc (IFF_NO_QUEUE). Users
can add one as desired.
2. Disable adding a VLAN to a VRF device.
3. Enable offloads and hardware features similar to other logical
devices (e.g., dummy, veth)
Change provides a significant boost in TCP stream Tx performance,
from ~2,700 Mbps to ~18,100 Mbps and makes throughput close to the
performance without a VRF (18,500 Mbps). netperf TCP_STREAM benchmark
using qemu with virtio+vhost for the NICs
Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Paolo Abeni [Mon, 13 Jun 2016 22:00:04 +0000 (00:00 +0200)]
tun: fix csum generation for tap devices
The commit 34166093639b ("tuntap: use common code for virtio_net_hdr
and skb GSO conversion") replaced the tun code for header manipulation
with the generic helpers. While doing so, it implictly moved the
skb_partial_csum_set() invocation after eth_type_trans(), which
invalidate the current gso start/offset values.
Fix it by moving the helper invocation before the mac pulling.
Fixes: 34166093639 ("tuntap: use common code for virtio_net_hdr and skb GSO conversion") Signed-off-by: Paolo Abeni <pabeni@redhat.com> Acked-by: Mike Rapoport <rppt@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 15 Jun 2016 20:58:34 +0000 (13:58 -0700)]
Merge branch 'skb_array'
Michael S. Tsirkin says:
====================
skb_array: array based FIFO for skbs
This is in response to the proposal by Jason to make tun
rx packet queue lockless using a circular buffer.
My testing seems to show that at least for the common usecase
in networking, which isn't lockless, circular buffer
with indices does not perform that well, because
each index access causes a cache line to bounce between
CPUs, and index access causes stalls due to the dependency.
By comparison, an array of pointers where NULL means invalid
and !NULL means valid, can be updated without messing up barriers
at all and does not have this issue.
On the flip side, cache pressure may be caused by using large queues.
tun has a queue of 1000 entries by default and that's 8K.
At this point I'm not sure this can be solved efficiently.
The correct solution might be sizing the queues appropriately.
Here's an implementation of this idea: it can be used more
or less whenever sk_buff_head can be used, except you need
to know the queue size in advance.
As this might be useful outside of networking, I implemented
a generic array of void pointers, with a type-safe wrapper for skbs.
It remains to be seen whether resizing is required, in case it is
I included patches implementing resizing by holding both the
consumer and the producer locks.
I think this code works fine without any extra memory barriers since we
always read and write the same location, so the accesses can not be
reordered.
Multiple writes of the same value into memory would mess things up
for us, I don't think compilers would do it though.
But if people feel it's better to be safe wrt compiler optimizations,
specifying queue as volatile would probably do it in a cleaner way
than converting all accesses to READ_ONCE/WRITE_ONCE. Thoughts?
The only issue is with calls within a loop using the __ptr_ring_XXX
accessors - in theory compiler could hoist accesses out of the loop.
Following volatile-considered-harmful.txt I merely
documented that callers that busy-poll should invoke cpu_relax().
Most people will use the external skb_array_XXX APIs with a spinlock,
so this should not be an issue for them.
Eric Dumazet suggested adding an extra pointer to skb for when
we have a single outstanding packet. I could not figure out
a way to implement this without a shared consumer/producer lock
though, which would cause cache line bounces by itself.
Jesper, Jason, I know that both of you tested this,
please post Tested-by tags for whatever was tested.
changes since v7
fix typos noticed by Jesper Brouer
changes since v6
resize implemented. peek/full calls are no longer lockless
replaced _FIELD macros with _CALL which invoke a function
on the pointer rather than just returning a value
destroy now scans the array and frees all queued skbs
changes since v5
implemented a generic ptr_ring api, and
made skb_array a type-safe wrapper
apis for taking the spinlock in different contexts
following expected usecase in tun
changes since v4 (v3 was never posted)
documentation
dropped SKB_ARRAY_MIN_SIZE heuristic
unit test (in userspace, included as patch 2)
changes since v2:
fixed integer overflow pointed out by Eric.
added some comments.
changes since v1:
fixed bug pointed out by Eric.
====================
Tested-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Tested-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This adds ring resize support. Seems to be necessary as
users such as tun allow userspace control over queue size.
If resize is used, this costs us ability to peek at queue without
consumer lock - should not be a big deal as peek and consumer are
usually run on the same CPU.
If ring is made bigger, ring contents is preserved. If ring is made
smaller, extra pointers are passed to an optional destructor callback.
Cleanup function also gains destructor callback such that
all pointers in queue can be cleaned up.
This changes some APIs but we don't have any users yet,
so it won't break bisect.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
A simple array based FIFO of pointers. Intended for net stack so uses
skbs for type safety. Implemented as a set of wrappers around ptr_ring.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Tested-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
A simple array based FIFO of pointers. Intended for net stack which
commonly has a single consumer/producer.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
WANG Cong [Mon, 13 Jun 2016 20:46:28 +0000 (13:46 -0700)]
net_sched: make tcf_hash_check() boolean
Cc: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 15 Jun 2016 19:34:34 +0000 (12:34 -0700)]
Merge branch 'vrf-ipv6-mcast-link-local'
David Ahern says:
====================
net: vrf: Handle ipv6 multicast and link-local addresses
IPv6 multicast and link-local addresses require special handling by the
VRF driver. Rather than using the VRF device index and full FIB lookups,
packets to/from these addresses should use direct FIB lookups based on
the VRF device table.
Multicast routes do not make sense for the L3 master device directly.
Accordingly, do not add mcast routes for the device, and the VRF driver
should fail attempts to send packets to ipv6 mcast addresses on the
device (e.g, ping6 ff02::1%<vrf> should fail)
With this change connections into and out of a VRF enslaved device work
for multicast and link-local addresses (icmp, tcp, and udp). e.g.,
1. packets into VM with VRF config:
ping6 -c3 fe80::e0:f9ff:fe1c:b974%br1
ping6 -c3 ff02::1%br1
ssh -6 fe80::e0:f9ff:fe1c:b974%br1
2. packets going out a VRF enslaved device:
ping6 -c3 fe80::18f8:83ff:fe4b:7a2e%eth1
ping6 -c3 ff02::1%eth1
ssh -6 root@fe80::18f8:83ff:fe4b:7a2e%eth1
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David Ahern [Mon, 13 Jun 2016 20:44:19 +0000 (13:44 -0700)]
net: vrf: Handle ipv6 multicast and link-local addresses
IPv6 multicast and link-local addresses require special handling by the
VRF driver:
1. Rather than using the VRF device index and full FIB lookups,
packets to/from these addresses should use direct FIB lookups based on
the VRF device table.
2. fail sends/receives on a VRF device to/from a multicast address
(e.g, make ping6 ff02::1%<vrf> fail)
3. move the setting of the flow oif to the first dst lookup and revert
the change in icmpv6_echo_reply made in ca254490c8dfd ("net: Add VRF
support to IPv6 stack"). Linklocal/mcast addresses require use of the
skb->dev.
With this change connections into and out of a VRF enslaved device work
for multicast and link-local addresses work (icmp, tcp, and udp)
e.g.,
1. packets into VM with VRF config:
ping6 -c3 fe80::e0:f9ff:fe1c:b974%br1
ping6 -c3 ff02::1%br1
ssh -6 fe80::e0:f9ff:fe1c:b974%br1
2. packets going out a VRF enslaved device:
ping6 -c3 fe80::18f8:83ff:fe4b:7a2e%eth1
ping6 -c3 ff02::1%eth1
ssh -6 root@fe80::18f8:83ff:fe4b:7a2e%eth1
Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David Ahern [Mon, 13 Jun 2016 20:44:18 +0000 (13:44 -0700)]
net: ipv6: Do not add multicast route for l3 master devices
L3 master devices are virtual devices similar to the loopback
device. Link local and multicast routes for these devices do
not make sense. The ipv6 addrconf code already skips adding a
linklocal address; do the same for the mcast route.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eugene Crosser [Mon, 13 Jun 2016 16:46:16 +0000 (18:46 +0200)]
af_iucv: use paged SKBs for big inbound messages
When an inbound message is bigger than a page, allocate a paged SKB,
and subsequently use IUCV receive primitive with IPBUFLST flag.
This relaxes the pressure to allocate big contiguous kernel buffers.
Signed-off-by: Eugene Crosser <Eugene.Crosser@ru.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eugene Crosser [Mon, 13 Jun 2016 16:46:15 +0000 (18:46 +0200)]
af_iucv: remove fragment_skb() to use paged SKBs
Before introducing paged skbs in the receive path, get rid of the
function `iucv_fragment_skb()` that replaces one large linear skb
with several smaller linear skbs.
Signed-off-by: Eugene Crosser <Eugene.Crosser@ru.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eugene Crosser [Mon, 13 Jun 2016 16:46:14 +0000 (18:46 +0200)]
af_iucv: use paged SKBs for big outbound messages
When an outbound message is bigger than a page, allocate and fill
a paged SKB, and subsequently use IUCV send primitive with IPBUFLST
flag. This relaxes the pressure to allocate big contiguous kernel
buffers.
Signed-off-by: Eugene Crosser <Eugene.Crosser@ru.ibm.com> Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
WANG Cong [Mon, 13 Jun 2016 17:47:44 +0000 (10:47 -0700)]
act_police: rename tcf_act_police_locate() to tcf_act_police_init()
This function is just ->init(), rename it to make it obvious.
Cc: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
WANG Cong [Mon, 13 Jun 2016 17:47:43 +0000 (10:47 -0700)]
net_sched: remove internal use of TC_POLICE_*
These should be gone when we removed CONFIG_NET_CLS_POLICE.
We can not totally remove them since they are exposed
to userspace.
Cc: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 15 Jun 2016 06:50:44 +0000 (23:50 -0700)]
Merge branch 'rds-mprds-foundations'
Sowmini Varadhan says:
====================
RDS: multiple connection paths for scaling
Today RDS-over-TCP is implemented by demux-ing multiple PF_RDS sockets
between any 2 endpoints (where endpoint == [IP address, port]) over a
single TCP socket between the 2 IP addresses involved. This has the
limitation that it ends up funneling multiple RDS flows over a single
TCP flow, thus the rds/tcp connection is
(a) upper-bounded to the single-flow bandwidth,
(b) suffers from head-of-line blocking for the RDS sockets.
Better throughput (for a fixed small packet size, MTU) can be achieved
by having multiple TCP/IP flows per rds/tcp connection, i.e., multipathed
RDS (mprds). Each such TCP/IP flow constitutes a path for the rds/tcp
connection. RDS sockets will be attached to a path based on some hash
(e.g., of local address and RDS port number) and packets for that RDS
socket will be sent over the attached path using TCP to segment/reassemble
RDS datagrams on that path.
The table below, generated using a prototype that implements mprds,
shows that this is significant for scaling to 40G. Packet sizes
used were: 8K byte req, 256 byte resp. MTU: 1500. The parameters for
RDS-concurrency used below are described in the rds-stress(1) man page-
the number listed is proportional to the number of threads at which max
throughput was attained.
-------------------------------------------------------------------
RDS-concurrency Num of tx+rx K/s (iops) throughput
(-t N -d N) TCP paths
-------------------------------------------------------------------
16 1 600K - 700K 4 Gbps
28 8 5000K - 6000K 32 Gbps
-------------------------------------------------------------------
FAQ: what is the relation between mprds and mptcp?
mprds is orthogonal to mptcp. Whereas mptcp creates
sub-flows for a single TCP connection, mprds parallelizes tx/rx
at the RDS layer. MPRDS with N paths will allow N datagrams to
be sent in parallel; each path will continue to send one
datagram at a time, with sender and receiver keeping track of
the retransmit and dgram-assembly state based on the RDS header.
If desired, mptcp can additionally be used to speed up each TCP
path. That acceleration is orthogonal to the parallelization benefits
of mprds.
This patch series lays down the foundational data-structures to support
mprds in the kernel. It implements the changes to split up the
rds_connection structure into a common (to all paths) part,
and a per-path rds_conn_path. All I/O workqs are driven from
the rds_conn_path.
Note that this patchset does not (yet) actually enable multipathing
for any of the transports; all transports will continue to use a
single path with the refactored data-structures. A subsequent patchset
will add the changes to the rds-tcp module to actually use mprds
in rds-tcp.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:42 +0000 (09:44 -0700)]
RDS: Update rds_conn_destroy to be MP capable
Refactor rds_conn_destroy() so that the per-path dismantling
is done in rds_conn_path_destroy, and then iterate as needed
over rds_conn_path_destroy().
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:41 +0000 (09:44 -0700)]
RDS: Update rds_conn_shutdown to work with rds_conn_path
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>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:38 +0000 (09:44 -0700)]
RDS: update rds-info related functions to traverse multiple conn_paths
This commit updates the callbacks related to the rds-info command
so that they walk through all the rds_conn_path structures and
report the requested info.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:37 +0000 (09:44 -0700)]
RDS: Add rds_conn_path_connect_if_down() for MP-aware callers
rds_conn_path_connect_if_down() works on the rds_conn_path
that it is passed. Callers who are not t_m_capable may continue
calling rds_conn_connect_if_down, which will invoke
rds_conn_path_connect_if_down() with the default c_path[0].
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:36 +0000 (09:44 -0700)]
RDS: Make rds_send_pong() take a rds_conn_path argument
This commit allows rds_send_pong() callers to send back
the rds pong message on some path other than c_path[0] by
passing in a struct rds_conn_path * argument. It also
removes the last dependency on the #defines in rds_single.h
from send.c
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:34 +0000 (09:44 -0700)]
RDS: Pass rds_conn_path to rds_send_xmit()
Pass a struct rds_conn_path to rds_send_xmit so that MP capable
transports can transmit packets on something other than c_path[0].
The eventual goal for MP capable transports is to hash the rds
socket to a path based on the bound local address/port, and use
this path as the argument to rds_send_xmit()
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:33 +0000 (09:44 -0700)]
RDS: Make rds_send_queue_rm() rds_conn_path aware
Pass the rds_conn_path to rds_send_queue_rm, and use it to initialize
the i_conn_path field in struct rds_incoming. This commit also makes
rds_send_queue_rm() MP capable, because it now takes locks
specific to the rds_conn_path passed in, instead of defaulting to
the c_path[0] based defines from rds_single_path.h
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:32 +0000 (09:44 -0700)]
RDS: Remove stale function rds_send_get_message()
The only caller of rds_send_get_message() was
rds_iw_send_cq_comp_handler() which was removed as part of
commit dcdede0406d3 ("RDS: Drop stale iWARP RDMA transport"),
so remove rds_send_get_message() for the same reason.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sowmini Varadhan [Mon, 13 Jun 2016 16:44:26 +0000 (09:44 -0700)]
RDS: split out connection specific state from rds_connection to rds_conn_path
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>
Neal Cardwell [Mon, 13 Jun 2016 15:20:35 +0000 (11:20 -0400)]
tcp: return sizeof tcp_dctcp_info in dctcp_get_info()
Make sure that dctcp_get_info() returns only the size of the
info->dctcp struct that it zeroes out and fills in. Previously it had
been returning the size of the enclosing tcp_cc_info union,
sizeof(*info). There is no problem yet, but that union that may one
day be larger than struct tcp_dctcp_info, in which case the
TCP_CC_INFO code might accidentally copy uninitialized bytes from the
stack.
Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Wei Yongjun [Mon, 13 Jun 2016 15:08:26 +0000 (23:08 +0800)]
sctp: fix error return code in sctp_init()
Fix to return a negative error code from the error handling
case instead of 0, as done elsewhere in this function.
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Acked-by: Xin Long <lucien.xin@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Here's the next part of the AF_RXRPC rewrite. In this set I rename some of
the files in the net/rxrpc/ directory and adjust the Makefile and
ar-internal.h to reflect the changes.
The aim is twofold:
(1) Remove the "ar-" prefix on those files that have it as it's not really
useful, especially now that I'm building rxkad in.
(2) To aid splitting the local, peer, connection and call handling code
into separate files for object and event handling in future patches by
making it easier to come up with new filenames.
There are two commits:
(1) The first commit does a bunch of renames of .c files and alters the
Makefile. ar-internal.h isn't renamed at this time to avoid having to
change the contents of the files being renamed.
(2) The second commit changes the section label comments in ar-internal.h
to reflect the changed filenames and reorders the file so that the
sections are back in filename order.
Kejian Yan [Mon, 13 Jun 2016 08:56:20 +0000 (16:56 +0800)]
net: hns: update the dependency
After the patchset about adding support of ACPI (commit id is 6343488)
being applied, HNS does not depend on OF. It depends on OF or ACPI, so
the Kconfig file needs to be updated.
Signed-off-by: Kejian Yan <yankejian@huawei.com> Signed-off-by: Yisen Zhuang <Yisen.Zhuang@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
hayeswang [Mon, 13 Jun 2016 09:49:37 +0000 (17:49 +0800)]
r8152: move the setting for the default speed
Move calling set_speed() from open() to rtl_hw_phy_work_func_t().
Then, we would set the default speed only for first initialization
or after resuming.
Besides, the set_speed() could handle the flag of PHY_RESET which
would be set in rtl_ops.hw_phy_cfg().
Signed-off-by: Hayes Wang <hayeswang@realtek.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Amir Vadai [Mon, 13 Jun 2016 09:06:39 +0000 (12:06 +0300)]
net/sched: flower: Return error when hw can't offload and skip_sw is set
When skip_sw is set and hardware fails to apply filter, return error to
user. This will make error propagation logic similar to the one
currently used in u32 classifier.
Also, changed code to use tc_skip_sw() utility function.
Signed-off-by: Amir Vadai <amirva@mellanox.com> Acked-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 14 Jun 2016 23:16:18 +0000 (19:16 -0400)]
Merge branch 'bnxt_en-updates'
Michael Chan says:
====================
bnxt_en: Updates for net-next.
-Add default VLAN support for VFs.
-Add NPAR (NIC partioning) support.
-Add support for new device 5731x and 5741x. GRO logic is different.
-Support new ETHTOOL_{G|S}LINKSETTINGS.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:37 +0000 (02:25 -0400)]
bnxt_en: Don't allow autoneg on cards that don't support it.
Some cards do not support autoneg. The current code does not prevent the
user from enabling autoneg with ethtool on such cards, causing confusion.
Firmware provides the autoneg capability information and we just need to
store it in the support_auto_speeds field in bnxt_link_info struct.
The ethtool set_settings() call will check this field before proceeding
with autoneg.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:35 +0000 (02:25 -0400)]
bnxt_en: Add GRO logic for BCM5731X chips.
Add bnxt_gro_func_5731x() to handle GRO packets for this chip. The
completion structures used in the new chip have new data to help determine
the header offsets. The offsets can be off by 4 if the packet is an
internal loopback packet (e.g. from one VF to another VF). Some additional
logic is added to adjust the offsets if it is a loopback packet.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:33 +0000 (02:25 -0400)]
bnxt_en: Define the supported chip numbers.
Define all the supported chip numbers and chip categories. Store the
chip_num returned by firmware. If the call to get the version and chip
number fails, we should abort.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
NPAR type is read from bnxt_hwrm_func_qcfg. Do not allow changing link
parameters if in NPAR mode sinc ethe port is shared among multiple
partitions. The link parameters are set up by firmware.
Signed-off-by: Satish Baddipadige <sbaddipa@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:30 +0000 (02:25 -0400)]
bnxt_en: Handle VF_CFG_CHANGE event from firmware.
When the VF driver gets this event, the VF configuration has changed (such
as default VLAN). The VF driver will initiate a silent reset to pick up
the new configuration.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:29 +0000 (02:25 -0400)]
bnxt_en: Add new function bnxt_reset().
When a default VLAN is added to the VF, the VF driver needs to reset to
pick up the default VLAN ID. We can use the same tx timeout reset logic
to do that, without the debug output. This new function, with the
silent parameter to suppress debug output will now serve both purposes.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 13 Jun 2016 06:25:28 +0000 (02:25 -0400)]
bnxt_en: Add function for VF driver to query default VLAN.
The PF can setup a default VLAN for a VF. The default VLAN tag is
automatically inserted and stripped without the knowledge of the
stack running on the VF. The VF driver needs to know that default
VLAN is enabled as VLAN acceleration on the RX side is no longer
supported. Call netdev_update_features() to fix up the VLAN features
as necessary. Also, VLAN strip mode must be enabled to strip out
the default VLAN tag.
Only allow VF default VLAN to be set if the firmware spec is >= 1.2.1.
Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mike Rapoport [Tue, 14 Jun 2016 05:29:38 +0000 (08:29 +0300)]
virtio_net: fix csum generation for virtio-net devices
The commit e858fae2b0b8 ("virtio_net: use common code for virtio_net_hdr
and skb GSO conversion") replaced the tun code for header manipulation
with the generic helpers. While doing so, it implictly moved the
skb_partial_csum_set() invocation after eth_type_trans(), which
invalidate the current gso start/offset values.
Fix it by moving the helper invocation before the mac pulling.
Fixes: e858fae2b0b8 ("virtio_net: use common code for virtio_net_hdr and
skb GSO conversion")
Reported-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David Howells [Mon, 13 Jun 2016 12:30:30 +0000 (13:30 +0100)]
rxrpc: Update the comments in ar-internal.h to reflect renames
Update the section comments in ar-internal.h that indicate the locations of
the referenced items to reflect the renames done to the .c files in
net/rxrpc/.
This also involves some rearrangement to reflect keep the sections in order
of filename.
Signed-off-by: David Howells <dhowells@redhat.com>
The former file is going to have the basic connection object handling,
but there will likely be some differentiation between client
connections and service connections in additional files later. The
latter file will have all the connection-level event handling.
(*) ar-local.c -> local_object.c
This will have the local endpoint object handling code. The local
endpoint event handling code will later be split out into
local_event.c.
(*) ar-peer.c -> peer_object.c
This will have the peer endpoint object handling code. Peer event
handling code will be placed in peer_event.c (for the moment, there is
none).
(*) ar-error.c -> peer_event.c
This will become the peer event handling code, though for the moment
it's actually driven from the local endpoint's perspective.
Note that I haven't renamed ar-transport.c to transport_object.c as the
intention is to delete it when the rxrpc_transport struct is excised.
The only file that actually has its contents changed is net/rxrpc/Makefile.
net/rxrpc/ar-internal.h will need its section marker comments updating, but
I'll do that in a separate patch to make it easier for git to follow the
history across the rename. I may also want to rename ar-internal.h at some
point - but that would mean updating all the #includes and I'd rather do
that in a separate step.
Signed-off-by: David Howells <dhowells@redhat.com.
ipv6: use TOS marks from sockets for routing decision
In IPv6 the ToS values are part of the flowlabel in flowi6 and get
extracted during fib rule lookup, but we forgot to correctly initialize
the flowlabel before the routing lookup.
Reported-by: <liam.mcbirnie@boeing.com> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
HTB, CBQ and HFSC pay a very high cost updating the qdisc 'throttled'
status that nothing but CBQ seems to use.
CBQ usage is flaky anyway, since no qdisc ->enqueue() updates the
'throttled' qdisc status.
This looks like some 'optimization' that actually cost more than code
without the optimization, and might cause latency issues with CBQ.
In my tests, I could achieve a 8 % performance increase in TCP_RR
workload through HTB qdisc, in presence of throttled classes,
and 5 % without throttled classes.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Fri, 10 Jun 2016 23:41:39 +0000 (16:41 -0700)]
net_sched: remove generic throttled management
__QDISC_STATE_THROTTLED bit manipulation is rather expensive
for HTB and few others.
I already removed it for sch_fq in commit f2600cf02b5b
("net: sched: avoid costly atomic operation in fq_dequeue()")
and so far nobody complained.
When one ore more packets are stuck in one or more throttled
HTB class, a htb dequeue() performs two atomic operations
to clear/set __QDISC_STATE_THROTTLED bit, while root qdisc
lock is held.
Removing this pair of atomic operations bring me a 8 % performance
increase on 200 TCP_RR tests, in presence of throttled classes.
This patch has no side effect, since nothing actually uses
disc_is_throttled() anymore.
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Fri, 10 Jun 2016 23:41:37 +0000 (16:41 -0700)]
net_sched: cbq: remove a flaky use of qdisc_is_throttled()
So far no qdisc ever unset the throttled bit at enqueue() time,
so CBQ usage of qdisc_is_throttled() was flaky.
Since __QDISC_STATE_THROTTLED set/unset is way too expensive
considering that only CBQ was eventually caring for this status,
it would make sense to implement a Qdisc ops ->is_throttled()
if we find that this is needed.
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 11 Jun 2016 06:24:55 +0000 (23:24 -0700)]
Merge branch 'mdio-iProc-SOC'
Pramod Kumar says:
====================
Add MDIO bus multiplexer support for iProc SoCs
Broadcom iProc based SoCs use a MDIO bus multiplexer where child buses
could be internal as well external to SoCs. These buses could supports
MDIO transaction compatible to C-22/C-45.
Broadcom MDIO bus multiplexer is an integrated multiplexer where child bus
selection and mdio transaction logic lies inside multiplexer itself.
To accommodate this multiplexer in existing mux framework below changes
were required-
1. Passed MDIO parent bus via mdio_mux_init to MDIO mux framework.
This patch set includes MDIO bus multiplexer driver along with above
framework change. It includes one external bus node having Ethernet PHY
attached and two internal bus node holding PCIe PHYs.
This patch series is based on v4.7-rc1 and is available from github-
repo: https://github.com/Broadcom/arm64-linux.git
branch:mdio-mux-v5
-Changes from v4:
- disabled PCIe PHYs from dtsi and enabled in dts file.
-Changes from v3:
- Unregister and free the parent MDIO bus.
- rebased on net-next/master branch.
Reason for resend:
-Rebased on v4.7-rc1
Changes from v2:
-Addressed Rob's comments in this patch regarding typo/grammers.
-Addressed David's comments regarding local variables order.
-Removed property "mdio-integrated-mux" and used mdiobus_register()
in place of of_mdiobus_regsiter().
-removed usage of IS_ERR_OR_NULL to IS_ERR in PCIe PHY driver.
Changes from v1:
- stop using "brcm,is_c45" from bus node as suggested by Andrew. MDIO
PHY driver will logically OR MII_ADDR_C45 into the address when issues
any C45 MDIO read/write transaction.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:51 +0000 (11:03 +0530)]
phy: Add Northstar2 PCI Phy support
Add PCI Phy support for Broadcom Northstar2 SoCs. This driver uses the
interface from the iproc mdio mux driver to enable the devices
respective phys.
Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Jon Mason <jonmason@broadcom.com> Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:50 +0000 (11:03 +0530)]
binding: PHY: Binding doc for NS2 PCIe PHYs.
Binding doc for NS2 PCIe PHYs.
Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Signed-off-by: Jon Mason <jonmason@broadcom.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:49 +0000 (11:03 +0530)]
net: mdio-mux: Add MDIO mux driver for iProc SoCs
iProc based SoCs supports the integrated mdio multiplexer which
has the bus selection as well as mdio transaction generation logic
inside.
This multiplexer has child buses for PCIe, SATA, USB and ETH. These
buses could be internal or external to SOC where PHYs are attached.
These buses could use C-45 or C-22 mdio transaction.
Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:47 +0000 (11:03 +0530)]
binding: mdio-mux: Add DT binding doc for Broadcom MDIO bus multiplexer
Add DT binding doc for Broadcom MDIO bus multiplexer driver.
Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:46 +0000 (11:03 +0530)]
binding: Make "mdio-parent-bus" property from mandatory to optional
Change "mdio-parent-bus" from mandatory section to optional
as it won't be required by integrated MDIO multiplexer
which has bus selection and mdio transaction generation logic,
integrated inside.
Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Pramod Kumar [Fri, 10 Jun 2016 05:33:45 +0000 (11:03 +0530)]
mdio: mux: Enhanced MDIO mux framework for integrated multiplexers
An integrated multiplexer uses same address space for
"muxed bus selection" and "generation of mdio transaction"
hence its good to register parent bus from mux driver.
Hence added a mechanism where mux driver could register a
parent bus and pass it down to framework via mdio_mux_init api.
Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Thu, 9 Jun 2016 14:48:18 +0000 (22:48 +0800)]
sctp: sctp should change socket state when shutdown is received
Now sctp doesn't change socket state upon shutdown reception. It changes
just the assoc state, even though it's a TCP-style socket.
For some cases, if we really need to check sk->sk_state, it's necessary to
fix this issue, at least when we use ss or netstat to dump, we can get a
more exact information.
As an improvement, we will change sk->sk_state when we change asoc->state
to SHUTDOWN_RECEIVED, and also do it in sctp_shutdown to keep consistent
with sctp_close.
Signed-off-by: Xin Long <lucien.xin@gmail.com> Acked-by: Marcelo R. Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 11 Jun 2016 06:13:32 +0000 (23:13 -0700)]
Merge tag 'mac80211-next-for-davem-2016-06-09' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next
Johannes Berg says:
====================
For the next cycle, we have the following:
* the biggest change is Michał's work on integrating FQ/codel
with the mac80211 internal software queues
* cfg80211 connect result gets clarified for the
"no connection at all" case
* advertisement of per-interface type capabilities, in case
they differ (which makes a lot of sense for some capabilities)
* most of the nl80211 & hwsim unprivileged namespace operation
changes
* human-readable VHT capabilities in debugfs
* some other cleanups, like spelling
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 11 Jun 2016 06:11:50 +0000 (23:11 -0700)]
Merge branch 'arm64-bpf'
Zi Shen Lim says:
====================
arm64 BPF JIT updates
Updates for arm64 eBPF JIT.
The main addition here is implementation of bpf_tail_call.
Changes since v2:
- None. Resubmit per David Miller.
Changes since v1:
- Added patch #1 to address build error due to missing header inclusion
in linux/bpf.h. (Thanks to suggestion and ack by Daniel Borkmann)
Ordered it ahead of bpf_tail_call patch #2 so build error is not
triggered.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Zi Shen Lim [Thu, 9 Jun 2016 04:18:48 +0000 (21:18 -0700)]
arm64: bpf: implement bpf_tail_call() helper
Add support for JMP_CALL_X (tail call) introduced by commit 04fd61ab36ec
("bpf: allow bpf programs to tail-call other bpf programs").
bpf_tail_call() arguments:
ctx - context pointer passed to next program
array - pointer to map which type is BPF_MAP_TYPE_PROG_ARRAY
index - index inside array that selects specific program to run
In this implementation arm64 JIT jumps into callee program after prologue,
so callee program reuses the same stack. For tail_call_cnt, we use the
callee-saved R26 (which was already saved/restored but previously unused
by JIT).
With this patch a tail call generates the following code on arm64:
Zi Shen Lim [Thu, 9 Jun 2016 04:18:47 +0000 (21:18 -0700)]
bpf: fix missing header inclusion
Commit 0fc174dea545 ("ebpf: make internal bpf API independent of
CONFIG_BPF_SYSCALL ifdefs") introduced usage of ERR_PTR() in
bpf_prog_get(), however did not include linux/err.h.
Without this patch, when compiling arm64 BPF without CONFIG_BPF_SYSCALL:
...
In file included from arch/arm64/net/bpf_jit_comp.c:21:0:
include/linux/bpf.h: In function 'bpf_prog_get':
include/linux/bpf.h:235:9: error: implicit declaration of function 'ERR_PTR' [-Werror=implicit-function-declaration]
return ERR_PTR(-EOPNOTSUPP);
^
include/linux/bpf.h:235:9: warning: return makes pointer from integer without a cast [-Wint-conversion]
In file included from include/linux/rwsem.h:17:0,
from include/linux/mm_types.h:10,
from include/linux/sched.h:27,
from arch/arm64/include/asm/compat.h:25,
from arch/arm64/include/asm/stat.h:23,
from include/linux/stat.h:5,
from include/linux/compat.h:12,
from include/linux/filter.h:10,
from arch/arm64/net/bpf_jit_comp.c:22:
include/linux/err.h: At top level:
include/linux/err.h:23:35: error: conflicting types for 'ERR_PTR'
static inline void * __must_check ERR_PTR(long error)
^
In file included from arch/arm64/net/bpf_jit_comp.c:21:0:
include/linux/bpf.h:235:9: note: previous implicit declaration of 'ERR_PTR' was here
return ERR_PTR(-EOPNOTSUPP);
^
...
Fixes: 0fc174dea545 ("ebpf: make internal bpf API independent of CONFIG_BPF_SYSCALL ifdefs") Suggested-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Zi Shen Lim <zlim.lnx@gmail.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 11 Jun 2016 06:07:50 +0000 (23:07 -0700)]
Merge branch 'tcp_nv'
Lawrence Brakmo says:
====================
tcp: add NV congestion control
Removed most of the module parameters
Tested in a rack using between 1 and 380 active TCP-NV flows.
Consists of the following patches:
[PATCH net-next v2 1/2] tcp: add in_flight to tcp_skb_cb
[PATCH net-next v2 2/2] tcp: add NV congestion control
====================
Lawrence Brakmo [Thu, 9 Jun 2016 04:16:45 +0000 (21:16 -0700)]
tcp: add NV congestion control
TCP-NV (New Vegas) is a major update to TCP-Vegas.
An earlier version of NV was presented at 2010's LPC.
It is a delayed based congestion avoidance for the
data center. This version has been tested within a
10G rack where the HW RTTs are 20-50us and with
1 to 400 flows.
A description of TCP-NV, including implementation
details as well as experimental results, can be found at:
http://www.brakmo.org/networking/tcp-nv/TCPNV.html
Signed-off-by: Lawrence Brakmo <brakmo@fb.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Lawrence Brakmo [Thu, 9 Jun 2016 04:16:44 +0000 (21:16 -0700)]
tcp: add in_flight to tcp_skb_cb
Add in_flight (bytes in flight when packet was sent) field
to tx component of tcp_skb_cb and make it available to
congestion modules' pkts_acked() function through the
ack_sample function argument.
Signed-off-by: Lawrence Brakmo <brakmo@fb.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Since the driver is infiniband which can be used as block device and the
workqueue seems involved in regular operation of the device, so a
dedicated workqueue has been used with WQ_MEM_RECLAIM set to guarantee
forward progress under memory pressure.
Since there are only a fixed number of work items, explicit concurrency
limit is unnecessary here.
Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / karo-tx-linux.git / log