Memory placement in sch_dsmark is silly : Better place mask/value
in the same cache line.
Also, we can embed small arrays in the first cache line and
remove a potential cache miss.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Existing bpf_clone_redirect() helper clones skb before redirecting
it to RX or TX of destination netdev.
Introduce bpf_redirect() helper that does that without cloning.
Benchmarked with two hosts using 10G ixgbe NICs.
One host is doing line rate pktgen.
Another host is configured as:
$ tc qdisc add dev $dev ingress
$ tc filter add dev $dev root pref 10 u32 match u32 0 0 flowid 1:2 \
action bpf run object-file tcbpf1_kern.o section clone_redirect_xmit drop
so it receives the packet on $dev and immediately xmits it on $dev + 1
The section 'clone_redirect_xmit' in tcbpf1_kern.o file has the program
that does bpf_clone_redirect() and performance is 2.0 Mpps
$ tc filter add dev $dev root pref 10 u32 match u32 0 0 flowid 1:2 \
action bpf run object-file tcbpf1_kern.o section redirect_xmit drop
which is using bpf_redirect() - 2.4 Mpps
and using cls_bpf with integrated actions as:
$ tc filter add dev $dev root pref 10 \
bpf run object-file tcbpf1_kern.o section redirect_xmit integ_act classid 1
performance is 2.5 Mpps
To summarize:
u32+act_bpf using clone_redirect - 2.0 Mpps
u32+act_bpf using redirect - 2.4 Mpps
cls_bpf using redirect - 2.5 Mpps
For comparison linux bridge in this setup is doing 2.1 Mpps
and ixgbe rx + drop in ip_rcv - 7.8 Mpps
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Often cls_bpf classifier is used with single action drop attached.
Optimize this use case and let cls_bpf return both classid and action.
For backwards compatibility reasons enable this feature under
TCA_BPF_FLAG_ACT_DIRECT flag.
Then more interesting programs like the following are easier to write:
int cls_bpf_prog(struct __sk_buff *skb)
{
/* classify arp, ip, ipv6 into different traffic classes
* and drop all other packets
*/
switch (skb->protocol) {
case htons(ETH_P_ARP):
skb->tc_classid = 1;
break;
case htons(ETH_P_IP):
skb->tc_classid = 2;
break;
case htons(ETH_P_IPV6):
skb->tc_classid = 3;
break;
default:
return TC_ACT_SHOT;
}
return TC_ACT_OK;
}
Joint work with Daniel Borkmann.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The permanent protocol nodes are at the head of the list,
So only need check all these nodes.
No matter the new node is permanent or not,
insert the new node after the last permanent protocol node,
If the new node conflicts with existing permanent node,
return error.
Signed-off-by: Martin Zhang <martinbj2008@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit b73c3d0e4f ("net: Save TX flow hash in sock and set in skbuf
on xmit"), Tom provided a l4 hash to most outgoing TCP packets.
We'd like to provide one as well for SYNACK packets, so that all packets
of a given flow share same txhash, to later enable bonding driver to
also use skb->hash to perform slave selection.
Note that a SYNACK retransmit shuffles the tx hash, as Tom did
in commit 265f94ff54 ("net: Recompute sk_txhash on negative routing
advice") for established sockets.
This has nice effect making TCP flows resilient to some kind of black
holes, even at connection establish phase.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Tom Herbert <tom@herbertland.com>
Cc: Mahesh Bandewar <maheshb@google.com>
Acked-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In code review it was noticed that I had failed to add some blank lines
in places where they are customarily used. Taking a second look at the
code I have to agree blank lines would be nice so I have added them
here.
Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is immediately motivated by the bridge code that chains functions that
call into netfilter. Without passing net into the okfns the bridge code would
need to guess about the best expression for the network namespace to process
packets in.
As net is frequently one of the first things computed in continuation functions
after netfilter has done it's job passing in the desired network namespace is in
many cases a code simplification.
To support this change the function dst_output_okfn is introduced to
simplify passing dst_output as an okfn. For the moment dst_output_okfn
just silently drops the struct net.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of saying "net = dev_net(state->in?state->in:state->out)"
just say "state->net". As that information is now availabe,
much less confusing and much less error prone.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pass a network namespace parameter into the netfilter hooks. At the
call site of the netfilter hooks the path a packet is taking through
the network stack is well known which allows the network namespace to
be easily and reliabily.
This allows the replacement of magic code like
"dev_net(state->in?:state->out)" that appears at the start of most
netfilter hooks with "state->net".
In almost all cases the network namespace passed in is derived
from the first network device passed in, guaranteeing those
paths will not see any changes in practice.
The exceptions are:
xfrm/xfrm_output.c:xfrm_output_resume() xs_net(skb_dst(skb)->xfrm)
ipvs/ip_vs_xmit.c:ip_vs_nat_send_or_cont() ip_vs_conn_net(cp)
ipvs/ip_vs_xmit.c:ip_vs_send_or_cont() ip_vs_conn_net(cp)
ipv4/raw.c:raw_send_hdrinc() sock_net(sk)
ipv6/ip6_output.c:ip6_xmit() sock_net(sk)
ipv6/ndisc.c:ndisc_send_skb() dev_net(skb->dev) not dev_net(dst->dev)
ipv6/raw.c:raw6_send_hdrinc() sock_net(sk)
br_netfilter_hooks.c:br_nf_pre_routing_finish() dev_net(skb->dev) before skb->dev is set to nf_bridge->physindev
In all cases these exceptions seem to be a better expression for the
network namespace the packet is being processed in then the historic
"dev_net(in?in:out)". I am documenting them in case something odd
pops up and someone starts trying to track down what happened.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netif_receive_skb_sk is only called once in the bridge code, replace
it with a bridge specific function that calls netif_receive_skb.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is prep work for passing net to the netfilter hooks.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When struct net starts being passed through the ipv4 and ipv6 fragment
routines br_nf_push_frag_xmit will need to take a net parameter.
Prepare br_nf_push_frag_xmit before that is needed and introduce
br_nf_push_frag_xmit_sk for the call sites that still need the old
calling conventions.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is a prep work for passing struct net through ip_do_fragment and
later the netfilter okfn. Doing this independently makes the later
code changes clearer.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Keep net in a local variable so I can use it in NF_HOOK_COND
when I pass struct net to all of the netfilter hooks.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A function with weird arguments that it will never use to accomdate a
netfilter callback prototype is absolutely in the core of the
networking stack. Frankly it does not make sense and it causes a lot
of confusion as to why arguments that are never used are being passed
to the function.
As I am preparing to make a second change to arguments to the okfn even
the names stops making sense.
As I have removed the two callers of this function remove this confusion
from the networking stack.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function dev_queue_xmit_skb_sk is unncessary and very confusing.
Introduce br_send_bpdu_finish to remove the need for dev_queue_xmit_skb_sk,
and have br_send_bpdu_finish call dev_queue_xmit.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function dev_queue_xmit_skb_sk is unncessary and very confusing.
Introduce arp_xmit_finish to remove the need for dev_queue_xmit_skb_sk,
and have arp_xmit_finish call dev_queue_xmit.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Calling dev_net(dev) for is just silly.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is a prepatory patch to passing net int the netfilter hooks,
where net will be used again.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Compute struct net from the input device in ip_forward before it is
used.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a sock paramter to dst_output making dst_output_sk superfluous.
Add a skb->sk parameter to all of the callers of dst_output
Have the callers of dst_output_sk call dst_output.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Redo commit ed1acc8cd8.
Commit 822b3b2ebf ("net: Add max rate tx queue
attribute") moved get_netdev_queue_index around, but kept the old version.
Probably because of a reuse of the original patch from before Eric's change to
that function.
Remove one inline keyword, and no need for a loop to find
an index into a table.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@redhat.com>
Fixes: 822b3b2ebf ("net: Add max rate tx queue attribute")
Acked-by: Or Gerlitz <ogerlitz@mellanox.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Many commonly used functions like getifaddrs() invoke RTM_GETLINK
to dump the interface information, and do not need the
the AF_INET6 statististics that are always returned by default
from rtnl_fill_ifinfo().
Computing the statistics can be an expensive operation that impacts
scaling, so it is desirable to avoid this if the information is
not needed.
This patch adds a the RTEXT_FILTER_SKIP_STATS extended info flag that
can be passed with netlink_request() to avoid statistics computation
for the ifinfo path.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
rt_fill_info which is called for 'route get' requests hardcodes the
table id as RT_TABLE_MAIN which is not correct when multiple tables
are used. Use the newly added table id in the rtable to send back
the correct table similar to what is done for IPv6.
To maintain current ABI a new request flag, RTM_F_LOOKUP_TABLE, is
added to indicate the actual table is wanted versus the hardcoded
response.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the FIB table id to rtable to make the information available for
IPv4 as it is for IPv6.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
All callers to rt_dst_alloc have nearly the same initialization following
a successful allocation. Consolidate it into rt_dst_alloc.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking fixes from David Miller:
1) Fix out-of-bounds array access in netfilter ipset, from Jozsef
Kadlecsik.
2) Use correct free operation on netfilter conntrack templates, from
Daniel Borkmann.
3) Fix route leak in SCTP, from Marcelo Ricardo Leitner.
4) Fix sizeof(pointer) in mac80211, from Thierry Reding.
5) Fix cache pointer comparison in ip6mr leading to missed unlock of
mrt_lock. From Richard Laing.
6) rds_conn_lookup() needs to consider network namespace in key
comparison, from Sowmini Varadhan.
7) Fix deadlock in TIPC code wrt broadcast link wakeups, from Kolmakov
Dmitriy.
8) Fix fd leaks in bpf syscall, from Daniel Borkmann.
9) Fix error recovery when installing ipv6 multipath routes, we would
delete the old route before we would know if we could fully commit
to the new set of nexthops. Fix from Roopa Prabhu.
10) Fix run-time suspend problems in r8152, from Hayes Wang.
11) In fec, don't program the MAC address into the chip when the clocks
are gated off. From Fugang Duan.
12) Fix poll behavior for netlink sockets when using rx ring mmap, from
Daniel Borkmann.
13) Don't allocate memory with GFP_KERNEL from get_stats64 in r8169
driver, from Corinna Vinschen.
14) In TCP Cubic congestion control, handle idle periods better where we
are application limited, in order to keep cwnd from growing out of
control. From Eric Dumzet.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (65 commits)
tcp_cubic: better follow cubic curve after idle period
tcp: generate CA_EVENT_TX_START on data frames
xen-netfront: respect user provided max_queues
xen-netback: respect user provided max_queues
r8169: Fix sleeping function called during get_stats64, v2
ether: add IEEE 1722 ethertype - TSN
netlink, mmap: fix edge-case leakages in nf queue zero-copy
netlink, mmap: don't walk rx ring on poll if receive queue non-empty
cxgb4: changes for new firmware 1.14.4.0
net: fec: add netif status check before set mac address
r8152: fix the runtime suspend issues
r8152: split DRIVER_VERSION
ipv6: fix ifnullfree.cocci warnings
add microchip LAN88xx phy driver
stmmac: fix check for phydev being open
net: qlcnic: delete redundant memsets
net: mv643xx_eth: use kzalloc
net: jme: use kzalloc() instead of kmalloc+memset
net: cavium: liquidio: use kzalloc in setup_glist()
net: ipv6: use common fib_default_rule_pref
...
Jana Iyengar found an interesting issue on CUBIC :
The epoch is only updated/reset initially and when experiencing losses.
The delta "t" of now - epoch_start can be arbitrary large after app idle
as well as the bic_target. Consequentially the slope (inverse of
ca->cnt) would be really large, and eventually ca->cnt would be
lower-bounded in the end to 2 to have delayed-ACK slow-start behavior.
This particularly shows up when slow_start_after_idle is disabled
as a dangerous cwnd inflation (1.5 x RTT) after few seconds of idle
time.
Jana initial fix was to reset epoch_start if app limited,
but Neal pointed out it would ask the CUBIC algorithm to recalculate the
curve so that we again start growing steeply upward from where cwnd is
now (as CUBIC does just after a loss). Ideally we'd want the cwnd growth
curve to be the same shape, just shifted later in time by the amount of
the idle period.
Reported-by: Jana Iyengar <jri@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: Sangtae Ha <sangtae.ha@gmail.com>
Cc: Lawrence Brakmo <lawrence@brakmo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Issuing a CC TX_START event on control frames like pure ACK
is a waste of time, as a CC should not care.
Following patch needs this change, as we want CUBIC to properly track
idle time at a low cost, with a single TX_START being generated.
Yuchung might slightly refine the condition triggering TX_START
on a followup patch.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Cc: Jana Iyengar <jri@google.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: Sangtae Ha <sangtae.ha@gmail.com>
Cc: Lawrence Brakmo <lawrence@brakmo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
When netlink mmap on receive side is the consumer of nf queue data,
it can happen that in some edge cases, we write skb shared info into
the user space mmap buffer:
Assume a possible rx ring frame size of only 4096, and the network skb,
which is being zero-copied into the netlink skb, contains page frags
with an overall skb->len larger than the linear part of the netlink
skb.
skb_zerocopy(), which is generic and thus not aware of the fact that
shared info cannot be accessed for such skbs then tries to write and
fill frags, thus leaking kernel data/pointers and in some corner cases
possibly writing out of bounds of the mmap area (when filling the
last slot in the ring buffer this way).
I.e. the ring buffer slot is then of status NL_MMAP_STATUS_VALID, has
an advertised length larger than 4096, where the linear part is visible
at the slot beginning, and the leaked sizeof(struct skb_shared_info)
has been written to the beginning of the next slot (also corrupting
the struct nl_mmap_hdr slot header incl. status etc), since skb->end
points to skb->data + ring->frame_size - NL_MMAP_HDRLEN.
The fix adds and lets __netlink_alloc_skb() take the actual needed
linear room for the network skb + meta data into account. It's completely
irrelevant for non-mmaped netlink sockets, but in case mmap sockets
are used, it can be decided whether the available skb_tailroom() is
really large enough for the buffer, or whether it needs to internally
fallback to a normal alloc_skb().
>From nf queue side, the information whether the destination port is
an mmap RX ring is not really available without extra port-to-socket
lookup, thus it can only be determined in lower layers i.e. when
__netlink_alloc_skb() is called that checks internally for this. I
chose to add the extra ldiff parameter as mmap will then still work:
We have data_len and hlen in nfqnl_build_packet_message(), data_len
is the full length (capped at queue->copy_range) for skb_zerocopy()
and hlen some possible part of data_len that needs to be copied; the
rem_len variable indicates the needed remaining linear mmap space.
The only other workaround in nf queue internally would be after
allocation time by f.e. cap'ing the data_len to the skb_tailroom()
iff we deal with an mmap skb, but that would 1) expose the fact that
we use a mmap skb to upper layers, and 2) trim the skb where we
otherwise could just have moved the full skb into the normal receive
queue.
After the patch, in my test case the ring slot doesn't fit and therefore
shows NL_MMAP_STATUS_COPY, where a full skb carries all the data and
thus needs to be picked up via recv().
Fixes: 3ab1f683bf ("nfnetlink: add support for memory mapped netlink")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
In case of netlink mmap, there can be situations where received frames
have to be placed into the normal receive queue. The ring buffer indicates
this through NL_MMAP_STATUS_COPY, so the user is asked to pick them up
via recvmsg(2) syscall, and to put the slot back to NL_MMAP_STATUS_UNUSED.
Commit 0ef707700f ("netlink: rx mmap: fix POLLIN condition") changed
polling, so that we walk in the worst case the whole ring through the
new netlink_has_valid_frame(), for example, when the ring would have no
NL_MMAP_STATUS_VALID, but at least one NL_MMAP_STATUS_COPY frame.
Since we do a datagram_poll() already earlier to pick up a mask that could
possibly contain POLLIN | POLLRDNORM already (due to NL_MMAP_STATUS_COPY),
we can skip checking the rx ring entirely.
In case the kernel is compiled with !CONFIG_NETLINK_MMAP, then all this is
irrelevant anyway as netlink_poll() is just defined as datagram_poll().
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/route.c:2946:3-8: WARNING: NULL check before freeing functions like kfree, debugfs_remove, debugfs_remove_recursive or usb_free_urb is not needed. Maybe consider reorganizing relevant code to avoid passing NULL values.
NULL check before some freeing functions is not needed.
Based on checkpatch warning
"kfree(NULL) is safe this check is probably not required"
and kfreeaddr.cocci by Julia Lawall.
Generated by: scripts/coccinelle/free/ifnullfree.cocci
CC: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This switches IPv6 policy routing to use the shared
fib_default_rule_pref() function of IPv4 and DECnet. It is also used in
multicast routing for IPv4 as well as IPv6.
The motivation for this patch is a complaint about iproute2 behaving
inconsistent between IPv4 and IPv6 when adding policy rules: Formerly,
IPv6 rules were assigned a fixed priority of 0x3FFF whereas for IPv4 the
assigned priority value was decreased with each rule added.
Since then all users of the default_pref field have been converted to
assign the generic function fib_default_rule_pref(), fib_nl_newrule()
may just use it directly instead. Therefore get rid of the function
pointer altogether and make fib_default_rule_pref() static, as it's not
used outside fib_rules.c anymore.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
Problem:
The ecmp route replace support for ipv6 in the kernel, deletes the
existing ecmp route too early, ie when it installs the first nexthop.
If there is an error in installing the subsequent nexthops, its too late
to recover the already deleted existing route leaving the fib
in an inconsistent state.
This patch reduces the possibility of this by doing the following:
a) Changes the existing multipath route add code to a two stage process:
build rt6_infos + insert them
ip6_route_add rt6_info creation code is moved into
ip6_route_info_create.
b) This ensures that most errors are caught during building rt6_infos
and we fail early
c) Separates multipath add and del code. Because add needs the special
two stage mode in a) and delete essentially does not care.
d) In any event if the code fails during inserting a route again, a
warning is printed (This should be unlikely)
Before the patch:
$ip -6 route show
3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024
/* Try replacing the route with a duplicate nexthop */
$ip -6 route change 3000:1000:1000:1000::2/128 nexthop via
fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev
swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1
RTNETLINK answers: File exists
$ip -6 route show
/* previously added ecmp route 3000:1000:1000:1000::2 dissappears from
* kernel */
After the patch:
$ip -6 route show
3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024
/* Try replacing the route with a duplicate nexthop */
$ip -6 route change 3000:1000:1000:1000::2/128 nexthop via
fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev
swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1
RTNETLINK answers: File exists
$ip -6 route show
3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024
3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024
Fixes: 2759647247 ("ipv6: fix ECMP route replacement")
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- Create drivers/staging/rdma
- Move amso1100 driver to staging/rdma and schedule for deletion
- Move ipath driver to staging/rdma and schedule for deletion
- Add hfi1 driver to staging/rdma and set TODO for move to regular tree
- Initial support for namespaces to be used on RDMA devices
- Add RoCE GID table handling to the RDMA core caching code
- Infrastructure to support handling of devices with differing
read and write scatter gather capabilities
- Various iSER updates
- Kill off unsafe usage of global mr registrations
- Update SRP driver
- Misc. mlx4 driver updates
- Support for the mr_alloc verb
- Support for a netlink interface between kernel and user space cache
daemon to speed path record queries and route resolution
- Ininitial support for safe hot removal of verbs devices
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=EHLg
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma
Pull inifiniband/rdma updates from Doug Ledford:
"This is a fairly sizeable set of changes. I've put them through a
decent amount of testing prior to sending the pull request due to
that.
There are still a few fixups that I know are coming, but I wanted to
go ahead and get the big, sizable chunk into your hands sooner rather
than waiting for those last few fixups.
Of note is the fact that this creates what is intended to be a
temporary area in the drivers/staging tree specifically for some
cleanups and additions that are coming for the RDMA stack. We
deprecated two drivers (ipath and amso1100) and are waiting to hear
back if we can deprecate another one (ehca). We also put Intel's new
hfi1 driver into this area because it needs to be refactored and a
transfer library created out of the factored out code, and then it and
the qib driver and the soft-roce driver should all be modified to use
that library.
I expect drivers/staging/rdma to be around for three or four kernel
releases and then to go away as all of the work is completed and final
deletions of deprecated drivers are done.
Summary of changes for 4.3:
- Create drivers/staging/rdma
- Move amso1100 driver to staging/rdma and schedule for deletion
- Move ipath driver to staging/rdma and schedule for deletion
- Add hfi1 driver to staging/rdma and set TODO for move to regular
tree
- Initial support for namespaces to be used on RDMA devices
- Add RoCE GID table handling to the RDMA core caching code
- Infrastructure to support handling of devices with differing read
and write scatter gather capabilities
- Various iSER updates
- Kill off unsafe usage of global mr registrations
- Update SRP driver
- Misc mlx4 driver updates
- Support for the mr_alloc verb
- Support for a netlink interface between kernel and user space cache
daemon to speed path record queries and route resolution
- Ininitial support for safe hot removal of verbs devices"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (136 commits)
IB/ipoib: Suppress warning for send only join failures
IB/ipoib: Clean up send-only multicast joins
IB/srp: Fix possible protection fault
IB/core: Move SM class defines from ib_mad.h to ib_smi.h
IB/core: Remove unnecessary defines from ib_mad.h
IB/hfi1: Add PSM2 user space header to header_install
IB/hfi1: Add CSRs for CONFIG_SDMA_VERBOSITY
mlx5: Fix incorrect wc pkey_index assignment for GSI messages
IB/mlx5: avoid destroying a NULL mr in reg_user_mr error flow
IB/uverbs: reject invalid or unknown opcodes
IB/cxgb4: Fix if statement in pick_local_ip6adddrs
IB/sa: Fix rdma netlink message flags
IB/ucma: HW Device hot-removal support
IB/mlx4_ib: Disassociate support
IB/uverbs: Enable device removal when there are active user space applications
IB/uverbs: Explicitly pass ib_dev to uverbs commands
IB/uverbs: Fix race between ib_uverbs_open and remove_one
IB/uverbs: Fix reference counting usage of event files
IB/core: Make ib_dealloc_pd return void
IB/srp: Create an insecure all physical rkey only if needed
...
If an attempt to wake up users of broadcast link is made when there is
no enough place in send queue than it may hang up inside the
tipc_sk_rcv() function since the loop breaks only after the wake up
queue becomes empty. This can lead to complete CPU stall with the
following message generated by RCU:
INFO: rcu_sched self-detected stall on CPU { 0} (t=2101 jiffies
g=54225 c=54224 q=11465)
Task dump for CPU 0:
tpch R running task 0 39949 39948 0x0000000a
ffffffff818536c0 ffff88181fa037a0 ffffffff8106a4be 0000000000000000
ffffffff818536c0 ffff88181fa037c0 ffffffff8106d8a8 ffff88181fa03800
0000000000000001 ffff88181fa037f0 ffffffff81094a50 ffff88181fa15680
Call Trace:
<IRQ> [<ffffffff8106a4be>] sched_show_task+0xae/0x120
[<ffffffff8106d8a8>] dump_cpu_task+0x38/0x40
[<ffffffff81094a50>] rcu_dump_cpu_stacks+0x90/0xd0
[<ffffffff81097c3b>] rcu_check_callbacks+0x3eb/0x6e0
[<ffffffff8106e53f>] ? account_system_time+0x7f/0x170
[<ffffffff81099e64>] update_process_times+0x34/0x60
[<ffffffff810a84d1>] tick_sched_handle.isra.18+0x31/0x40
[<ffffffff810a851c>] tick_sched_timer+0x3c/0x70
[<ffffffff8109a43d>] __run_hrtimer.isra.34+0x3d/0xc0
[<ffffffff8109aa95>] hrtimer_interrupt+0xc5/0x1e0
[<ffffffff81030d52>] ? native_smp_send_reschedule+0x42/0x60
[<ffffffff81032f04>] local_apic_timer_interrupt+0x34/0x60
[<ffffffff810335bc>] smp_apic_timer_interrupt+0x3c/0x60
[<ffffffff8165a3fb>] apic_timer_interrupt+0x6b/0x70
[<ffffffff81659129>] ? _raw_spin_unlock_irqrestore+0x9/0x10
[<ffffffff8107eb9f>] __wake_up_sync_key+0x4f/0x60
[<ffffffffa313ddd1>] tipc_write_space+0x31/0x40 [tipc]
[<ffffffffa313dadf>] filter_rcv+0x31f/0x520 [tipc]
[<ffffffffa313d699>] ? tipc_sk_lookup+0xc9/0x110 [tipc]
[<ffffffff81659259>] ? _raw_spin_lock_bh+0x19/0x30
[<ffffffffa314122c>] tipc_sk_rcv+0x2dc/0x3e0 [tipc]
[<ffffffffa312e7ff>] tipc_bclink_wakeup_users+0x2f/0x40 [tipc]
[<ffffffffa313ce26>] tipc_node_unlock+0x186/0x190 [tipc]
[<ffffffff81597c1c>] ? kfree_skb+0x2c/0x40
[<ffffffffa313475c>] tipc_rcv+0x2ac/0x8c0 [tipc]
[<ffffffffa312ff58>] tipc_l2_rcv_msg+0x38/0x50 [tipc]
[<ffffffff815a76d3>] __netif_receive_skb_core+0x5a3/0x950
[<ffffffff815a98d3>] __netif_receive_skb+0x13/0x60
[<ffffffff815a993e>] netif_receive_skb_internal+0x1e/0x90
[<ffffffff815aa138>] napi_gro_receive+0x78/0xa0
[<ffffffffa07f93f4>] tg3_poll_work+0xc54/0xf40 [tg3]
[<ffffffff81597c8c>] ? consume_skb+0x2c/0x40
[<ffffffffa07f9721>] tg3_poll_msix+0x41/0x160 [tg3]
[<ffffffff815ab0f2>] net_rx_action+0xe2/0x290
[<ffffffff8104b92a>] __do_softirq+0xda/0x1f0
[<ffffffff8104bc26>] irq_exit+0x76/0xa0
[<ffffffff81004355>] do_IRQ+0x55/0xf0
[<ffffffff8165a12b>] common_interrupt+0x6b/0x6b
<EOI>
The issue occurs only when tipc_sk_rcv() is used to wake up postponed
senders:
tipc_bclink_wakeup_users()
// wakeupq - is a queue which consists of special
// messages with SOCK_WAKEUP type.
tipc_sk_rcv(wakeupq)
...
while (skb_queue_len(inputq)) {
filter_rcv(skb)
// Here the type of message is checked
// and if it is SOCK_WAKEUP then
// it tries to wake up a sender.
tipc_write_space(sk)
wake_up_interruptible_sync_poll()
}
After the sender thread is woke up it can gather control and perform
an attempt to send a message. But if there is no enough place in send
queue it will call link_schedule_user() function which puts a message
of type SOCK_WAKEUP to the wakeup queue and put the sender to sleep.
Thus the size of the queue actually is not changed and the while()
loop never exits.
The approach I proposed is to wake up only senders for which there is
enough place in send queue so the described issue can't occur.
Moreover the same approach is already used to wake up senders on
unicast links.
I have got into the issue on our product code but to reproduce the
issue I changed a benchmark test application (from
tipcutils/demos/benchmark) to perform the following scenario:
1. Run 64 instances of test application (nodes). It can be done
on the one physical machine.
2. Each application connects to all other using TIPC sockets in
RDM mode.
3. When setup is done all nodes start simultaneously send
broadcast messages.
4. Everything hangs up.
The issue is reproducible only when a congestion on broadcast link
occurs. For example, when there are only 8 nodes it works fine since
congestion doesn't occur. Send queue limit is 40 in my case (I use a
critical importance level) and when 64 nodes send a message at the
same moment a congestion occurs every time.
Signed-off-by: Dmitry S Kolmakov <kolmakov.dmitriy@huawei.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
