eth_type_trans() & get_rps_cpus() currently need two 64bytes cache
lines in packet to compute rxhash.
Increasing NET_SKB_PAD from 32 to 64 reduces the need to one cache
line only, and makes RPS faster.
NET_IP_ALIGN(2) + ethernet_header(14) + IP_header(20/40) + ports(8)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This whole patchset is for adding netpoll support to bridge and bonding
devices. I already tested it for bridge, bonding, bridge over bonding,
and bonding over bridge. It looks fine now.
To make bridge and bonding support netpoll, we need to adjust
some netpoll generic code. This patch does the following things:
1) introduce two new priv_flags for struct net_device:
IFF_IN_NETPOLL which identifies we are processing a netpoll;
IFF_DISABLE_NETPOLL is used to disable netpoll support for a device
at run-time;
2) introduce one new method for netdev_ops:
->ndo_netpoll_cleanup() is used to clean up netpoll when a device is
removed.
3) introduce netpoll_poll_dev() which takes a struct net_device * parameter;
export netpoll_send_skb() and netpoll_poll_dev() which will be used later;
4) hide a pointer to struct netpoll in struct netpoll_info, ditto.
5) introduce ->real_dev for struct netpoll.
6) introduce a new status NETDEV_BONDING_DESLAE, which is used to disable
netconsole before releasing a slave, to avoid deadlocks.
Cc: David Miller <davem@davemloft.net>
Cc: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: WANG Cong <amwang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With following patch I can reach maximum rate of my pktgen+udpsink
simulator :
- 'old' machine : dual quad core E5450 @3.00GHz
- 64 UDP rx flows (only differ by destination port)
- RPS enabled, NIC interrupts serviced on cpu0
- rps dispatched on 7 other cores. (~130.000 IPI per second)
- SLAB allocator (faster than SLUB in this workload)
- tg3 NIC
- 1.080.000 pps without a single drop at NIC level.
Idea is to add two prefetchw() calls in __alloc_skb(), one to prefetch
first sk_buff cache line, the second to prefetch the shinfo part.
Also using one memset() to initialize all skb_shared_info fields instead
of one by one to reduce number of instructions, using long word moves.
All skb_shared_info fields before 'dataref' are cleared in
__alloc_skb().
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add hlist_for_each_entry_rcu_bh() and
hlist_for_each_entry_continue_rcu_bh() macros, and use them in
ipv6_get_ifaddr(), if6_get_first() and if6_get_next() to fix lockdeps
warnings.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reviewed-by: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
virtio added mergeable buffers mode where 2 bytes of extra info is put
after vnet header but before actual data (tun does not need this data).
In hindsight, it would have been better to add the new info *before* the
packet: as it is, users need a lot of tricky code to skip the extra 2
bytes in the middle of the iovec, and in fact applications seem to get
it wrong, and only work with specific iovec layout. The fact we might
need to split iovec also means we might in theory overflow iovec max
size.
This patch adds a simpler way for applications to handle this,
and future proofs the interface against further extensions,
by making the size of the virtio net header configurable
from userspace. As a result, tun driver will simply
skip the extra 2 bytes on both input and output.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: David S. Miller <davem@davemloft.net>
Per cpu variable softnet_data.total was shared between IRQ and SoftIRQ context
without any protection. And enqueue_to_backlog should update the netdev_rx_stat
of the target CPU.
This patch renames softnet_data.total to softnet_data.processed: the number of
packets processed in uppper levels(IP stacks).
softnet_stat data is moved into softnet_data.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
include/linux/netdevice.h | 17 +++++++----------
net/core/dev.c | 26 ++++++++++++--------------
net/sched/sch_generic.c | 2 +-
3 files changed, 20 insertions(+), 25 deletions(-)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In commit 6be8ac2f ("[NET]: uninline skb_pull, de-bloats a lot")
we uninlined skb_pull.
But in some critical paths it makes sense to inline this thing
and it helps performance significantly.
Create an skb_pull_inline() so that we can do this in a way that
serves also as annotation.
Based upon a patch by Eric Dumazet.
Signed-off-by: David S. Miller <davem@davemloft.net>
sk_callback_lock rwlock actually protects sk->sk_sleep pointer, so we
need two atomic operations (and associated dirtying) per incoming
packet.
RCU conversion is pretty much needed :
1) Add a new structure, called "struct socket_wq" to hold all fields
that will need rcu_read_lock() protection (currently: a
wait_queue_head_t and a struct fasync_struct pointer).
[Future patch will add a list anchor for wakeup coalescing]
2) Attach one of such structure to each "struct socket" created in
sock_alloc_inode().
3) Respect RCU grace period when freeing a "struct socket_wq"
4) Change sk_sleep pointer in "struct sock" by sk_wq, pointer to "struct
socket_wq"
5) Change sk_sleep() function to use new sk->sk_wq instead of
sk->sk_sleep
6) Change sk_has_sleeper() to wq_has_sleeper() that must be used inside
a rcu_read_lock() section.
7) Change all sk_has_sleeper() callers to :
- Use rcu_read_lock() instead of read_lock(&sk->sk_callback_lock)
- Use wq_has_sleeper() to eventually wakeup tasks.
- Use rcu_read_unlock() instead of read_unlock(&sk->sk_callback_lock)
8) sock_wake_async() is modified to use rcu protection as well.
9) Exceptions :
macvtap, drivers/net/tun.c, af_unix use integrated "struct socket_wq"
instead of dynamically allocated ones. They dont need rcu freeing.
Some cleanups or followups are probably needed, (possible
sk_callback_lock conversion to a spinlock for example...).
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Changes:
This is a complete re-write of the socket layer. Making the socket
implementation more aligned with the other socket layers and using more
of the support functions available in sock.c. Lots of code is copied
from af_unix (and some from af_irda).
Non-blocking mode should be working as well.
Signed-off-by: Sjur Braendeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
batch skb dequeueing from softnet input_pkt_queue to reduce potential lock
contention when RPS is enabled.
Note: in the worst case, the number of packets in a softnet_data may
be double of netdev_max_backlog.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
reimplement softnet_data.output_queue as a FIFO queue to keep the
fairness among the qdiscs rescheduled.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
----
include/linux/netdevice.h | 1 +
net/core/dev.c | 22 ++++++++++++----------
2 files changed, 13 insertions(+), 10 deletions(-)
Signed-off-by: David S. Miller <davem@davemloft.net>
This is used to configure APs to not bridge traffic between connected stations.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This change corrects the fact that we were not reporting Gen2 link speeds
when we were in fact connected at Gen2 rates.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Our offset handling becomes even a little more hackish now. For some reason I
do not understand all offsets as inrelative. It assumes base offset is 0x1000
but it will work for now as we make offsets relative anyway by removing base
0x1000. Should be cleaner however.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Attempting to read registers that don't exist on the SSB bus can cause
hangs on some boxes. At least some b43 devices are 'in the wild' that
don't have SPROMs at all. When the SSB bus support loads, it attempts
to read these (non-existant) SPROMs and causes hard hangs on the box --
no console output, etc.
This patch adds some intelligence to determine whether or not the SPROM
is present before attempting to read it. This avoids those hard hangs
on those devices with no SPROM attached to their SSB bus. The
SSB-attached devices (e.g. b43, et al.) won't work, but at least the box
will survive to test further patches. :-)
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Cc: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Michael Buesch <mb@bu3sch.de>
Decouple rtnetlink address families from real address families in socket.h to
be able to add rtnetlink interfaces to code that is not a real address family
without increasing AF_MAX/NPROTO.
This will be used to add support for multicast route dumping from all tables
as the proc interface can't be extended to support anything but the main table
without breaking compatibility.
This partialy undoes the patch to introduce independant families for routing
rules and converts ipmr routing rules to a new rtnetlink family. Similar to
that patch, values up to 127 are reserved for real address families, values
above that may be used arbitrarily.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Finally add support to detect a local IPV6_DONTFRAG event
and return the relevant data to the user if they've enabled
IPV6_RECVPATHMTU on the socket. The next recvmsg() will
return no data, but have an IPV6_PATHMTU as ancillary data.
Signed-off-by: Brian Haley <brian.haley@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add underlying data structure changes and basic setsockopt()
and getsockopt() support for IPV6_RECVPATHMTU, IPV6_PATHMTU,
and IPV6_DONTFRAG. IPV6_PATHMTU is actually fully functional
at this point.
Signed-off-by: Brian Haley <brian.haley@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
remove DCB_PROTO_VERSION as we don't do netlink versioning
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
V2 Feedback from John Hughes.
- Add header for userspace implementations such as xot/xoe to use
- Use explicit values for interface stability
- No changes to driver patches
V1
- Use identifiers instead of magic numbers for X25 layer 3 to device interface.
- Also fixed checkpatch notes on updated code.
[ Add new user header to include/linux/Kbuild -DaveM ]
Signed-off-by: Andrew Hendry <andrew.hendry@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add an SKF_AD_HATYPE field to the packet ancilliary data area, giving
access to skb->dev->type, as reported in the sll_hatype field.
When capturing packets on a PF_PACKET/SOCK_RAW socket bound to all
interfaces, there doesn't appear to be a way for the filter program to
actually find out the underlying hardware type the packet was captured
on. This patch adds such ability.
This patch also handles the case where skb->dev can be NULL, such as on
netlink sockets.
Signed-off-by: Paul Evans <leonerd@leonerd.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds IPv6 support for RFC5082 Generalized TTL Security Mechanism.
Not to users of mapped address; the IPV6 and IPV4 socket options are seperate.
The server does have to deal with both IPv4 and IPv6 socket options
and the client has to handle the different for each family.
On client:
int ttl = 255;
getaddrinfo(argv[1], argv[2], &hint, &result);
for (rp = result; rp != NULL; rp = rp->ai_next) {
s = socket(rp->ai_family, rp->ai_socktype, rp->ai_protocol);
if (s < 0) continue;
if (rp->ai_family == AF_INET) {
setsockopt(s, IPPROTO_IP, IP_TTL, &ttl, sizeof(ttl));
} else if (rp->ai_family == AF_INET6) {
setsockopt(s, IPPROTO_IPV6, IPV6_UNICAST_HOPS,
&ttl, sizeof(ttl)))
}
if (connect(s, rp->ai_addr, rp->ai_addrlen) == 0) {
...
On server:
int minttl = 255 - maxhops;
getaddrinfo(NULL, port, &hints, &result);
for (rp = result; rp != NULL; rp = rp->ai_next) {
s = socket(rp->ai_family, rp->ai_socktype, rp->ai_protocol);
if (s < 0) continue;
if (rp->ai_family == AF_INET6)
setsockopt(s, IPPROTO_IPV6, IPV6_MINHOPCOUNT,
&minttl, sizeof(minttl));
setsockopt(s, IPPROTO_IP, IP_MINTTL, &minttl, sizeof(minttl));
if (bind(s, rp->ai_addr, rp->ai_addrlen) == 0)
break
...
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds platform data to the ks8842 driver.
Via the platform data a MAC address, to be used by the controller,
can be passed.
To ensure this MAC address is used, the MAC address is written
after each hardware reset.
Signed-off-by: Richard Röjfors <richard.rojfors@pelagicore.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
kill_fasync() uses a central rwlock, candidate for RCU conversion, to
avoid cache line ping pongs on SMP.
fasync_remove_entry() and fasync_add_entry() can disable IRQS on a short
section instead during whole list scan.
Use a spinlock per fasync_struct to synchronize kill_fasync_rcu() and
fasync_{remove|add}_entry(). This spinlock is IRQ safe, so sock_fasync()
doesnt need its own implementation and can use fasync_helper(), to
reduce code size and complexity.
We can remove __kill_fasync() direct use in net/socket.c, and rename it
to kill_fasync_rcu().
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is no need to export skb_under_panic() and skb_over_panic() in
skbuff.c, since these methods are used only in skbuff.c ; this patch
removes these two exports. It also marks these functions as 'static'
and removeS the extern declarations of them from
include/linux/skbuff.h
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct softnet_data holds many queues, so consistent use "sd" name
instead of "queue" is better.
Adds a rps_ipi_queued() helper to cleanup enqueue_to_backlog()
Adds a _and_irq_disable suffix to net_rps_action() name, as David
suggested.
incr_input_queue_head() becomes input_queue_head_incr()
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net_rps_action() is a bit expensive on NR_CPUS=64..4096 kernels, even if
RPS is not active.
Tom Herbert used two bitmasks to hold information needed to send IPI,
but a single LIFO list seems more appropriate.
Move all RPS logic into net_rps_action() to cleanup net_rx_action() code
(remove two ifdefs)
Move rps_remote_softirq_cpus into softnet_data to share its first cache
line, filling an existing hole.
In a future patch, we could call net_rps_action() from process_backlog()
to make sure we send IPI before handling this cpu backlog.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
rcu: Make RCU lockdep check the lockdep_recursion variable
rcu: Update docs for rcu_access_pointer and rcu_dereference_protected
rcu: Better explain the condition parameter of rcu_dereference_check()
rcu: Add rcu_access_pointer and rcu_dereference_protected
The lockdep facility temporarily disables lockdep checking by
incrementing the current->lockdep_recursion variable. Such
disabling happens in NMIs and in other situations where lockdep
might expect to recurse on itself.
This patch therefore checks current->lockdep_recursion, disabling RCU
lockdep splats when this variable is non-zero. In addition, this patch
removes the "likely()", as suggested by Lai Jiangshan.
Reported-by: Frederic Weisbecker <fweisbec@gmail.com>
Reported-by: David Miller <davem@davemloft.net>
Tested-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
Cc: eric.dumazet@gmail.com
LKML-Reference: <20100415195039.GA22623@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
wl1251 has WLAN_IRQ pin for generating interrupts to host processor,
which is mandatory in SPI mode and optional in SDIO mode (which can
use SDIO interrupts instead). However TI recommends using deditated
IRQ line for SDIO too.
Add support for using dedicated interrupt line with SDIO, but also leave
ability to switch to SDIO interrupts in case it's needed.
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Reviewed-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: wacom - switch mode upon system resume
Revert "Input: wacom - merge out and in prox events"
Input: matrix_keypad - allow platform to disable key autorepeat
Input: ALPS - add signature for HP Pavilion dm3 laptops
Input: i8042 - spelling fix
Input: sparse-keymap - implement safer freeing of the keymap
Input: update the status of the Multitouch X driver project
Input: clarify the no-finger event in multitouch protocol
Input: bcm5974 - retract efi-broken suspend_resume
Input: sparse-keymap - free the right keymap on error
Among else, this allows projects like libdc1394 to carry copies of the
ABI related header files without them or distributors having to worry
about effects on the project's overall license terms. Switch to MIT
license as suggested by Kristian. Also update the year in the
copyright statement according to source history.
Cc: Jay Fenlason <fenlason@redhat.com>
Acked-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
The new enh_desc is used for selecting the enhanced descriptors
structure. There are several scenarios; some chips (mac10/100
or gmac) want to use the enhanced descriptors; others want the normal
ones.
For example, on ST platforms: MAC10/100 uses the normal desc structure
and the GMAC uses the enhanced one.
It can be useful to get this information from the platform.
This could also be decided at run-time looking at the chip's ID number;
but it could happen that chips with the same ID want to use different
descriptor structure.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet