The l4_rxhash flag was added to the skb structure to indicate
that the rxhash value was computed over the 4 tuple for the
packet which includes the port information in the encapsulated
transport packet. This is used by the stack to preserve the
rxhash value in __skb_rx_tunnel.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This allows us to move duplicated code in <asm/atomic.h>
(atomic_inc_not_zero() for now) to <linux/atomic.h>
Signed-off-by: Arun Sharma <asharma@fb.com>
Reviewed-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
RX rings should use GFP_KERNEL allocations if possible, add
__netdev_alloc_skb_ip_align() helper to ease this.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rearrange struct sk_buff members comments to follow their
definition order. Also, add missing comments for ooo_okay
and dropcount members.
Signed-off-by: Daniel Baluta <dbaluta@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds userspace buffers support in skb shared info. A new
struct skb_ubuf_info is needed to maintain the userspace buffers
argument and index, a callback is used to notify userspace to release
the buffers once lower device has done DMA (Last reference to that skb
has gone).
If there is any userspace apps to reference these userspace buffers,
then these userspaces buffers will be copied into kernel. This way we
can prevent userspace apps from holding these userspace buffers too long.
Use destructor_arg to point to the userspace buffer info; a new tx flags
SKBTX_DEV_ZEROCOPY is added for zero-copy buffer check.
Signed-off-by: Shirley Ma <xma@...ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The comment for the skb_tx_timestamp() function suggests calling it just
after a buffer is released to the hardware for transmission. However,
for drivers that free the buffer in an ISR, this produces a race between
the time stamp code and the ISR. This commit changes the comment to advise
placing the call just before handing the buffer over to the hardware.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@davemloft.net>
Testing of VLAN_FLAG_REORDER_HDR does not belong in vlan_untag
but rather in vlan_do_receive. Otherwise the vlan header
will not be properly put on the packet in the case of
vlan header accelleration.
As we remove the check from vlan_check_reorder_header
rename it vlan_reorder_header to keep the naming clean.
Fix up the skb->pkt_type early so we don't look at the packet
after adding the vlan tag, which guarantees we don't goof
and look at the wrong field.
Use a simple if statement instead of a complicated switch
statement to decided that we need to increment rx_stats
for a multicast packet.
Hopefully at somepoint we will just declare the case where
VLAN_FLAG_REORDER_HDR is cleared as unsupported and remove
the code. Until then this keeps it working correctly.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (27 commits)
bnx2x: allow device properly initialize after hotplug
bnx2x: fix DMAE timeout according to hw specifications
bnx2x: properly handle CFC DEL in cnic flow
bnx2x: call dev_kfree_skb_any instead of dev_kfree_skb
net: filter: move forward declarations to avoid compile warnings
pktgen: refactor pg_init() code
pktgen: use vzalloc_node() instead of vmalloc_node() + memset()
net: skb_trim explicitely check the linearity instead of data_len
ipv4: Give backtrace in ip_rt_bug().
net: avoid synchronize_rcu() in dev_deactivate_many
net: remove synchronize_net() from netdev_set_master()
rtnetlink: ignore NETDEV_RELEASE and NETDEV_JOIN event
net: rename NETDEV_BONDING_DESLAVE to NETDEV_RELEASE
bridge: call NETDEV_JOIN notifiers when add a slave
netpoll: disable netpoll when enslave a device
macvlan: Forward unicast frames in bridge mode to lowerdev
net: Remove linux/prefetch.h include from linux/skbuff.h
ipv4: Include linux/prefetch.h in fib_trie.c
netlabel: Remove prefetches from list handlers.
drivers/net: add prefetch header for prefetch users
...
Fixed up prefetch parts: removed a few duplicate prefetch.h includes,
fixed the location of the igb prefetch.h, took my version of the
skbuff.h code without the extra parentheses etc.
Commit e66eed651f ("list: remove prefetching from regular list
iterators") removed the include of prefetch.h from list.h. The skbuff
list traversal still had them.
Quoth David Miller:
"Please just remove the prefetches.
Those are modelled after list.h as I intend to eventually convert
SKB list handling to "struct list_head" but we're not there yet.
Therefore if we kill prefetches from list.h we should kill it from
these things in skbuff.h too."
Requested-by: David Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The purpose of the check on data_len is to check linearity, so use the inline
helper for this. No overhead and more explicit.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes build errors on s390 and probably other archs as well:
In file included from net/ipv4/ip_forward.c:32:0:
include/net/udp.h: In function 'udp_csum_outgoing':
include/net/udp.h:141:2: error: implicit declaration of function 'prefetch'
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
After commit a715dea3c8 ("net: Always
allocate at least 16 skb frags regardless of page size"), the value
of MAX_SKB_FRAGS can now take on either an "unsigned long" or an
"int" value.
This causes warnings like:
net/packet/af_packet.c: In function ‘tpacket_fill_skb’:
net/packet/af_packet.c:948: warning: format ‘%lu’ expects type ‘long unsigned int’, but argument 2 has type ‘int’
Fix by forcing the constant to be unsigned long, otherwise we have
a situation where the type of a system wide constant is variable.
Signed-off-by: David S. Miller <davem@davemloft.net>
When analysing performance of the cxgb3 on a ppc64 box I noticed that
we weren't doing much GRO merging. It turns out we are limited by the
number of SKB frags:
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 2)
With a 4kB page size we have 18 frags, but with a 64kB page size we
only have 3 frags.
I ran a single stream TCP bandwidth test to compare the performance of
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch allows rx_handlers to better signalize what to do next to
it's caller. That makes skb->deliver_no_wcard no longer needed.
kernel-doc for rx_handler_result is taken from Nicolas' patch.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Reviewed-by: Nicolas de Pesloüan <nicolas.2p.debian@free.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Quoting Ben Hutchings: we presumably won't be defining features that
can only be enabled on 64-bit architectures.
Occurences found by `grep -r` on net/, drivers/net, include/
[ Move features and vlan_features next to each other in
struct netdev, as per Eric Dumazet's suggestion -DaveM ]
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
The IPv6 tproxy patches split IPv6 defragmentation off of conntrack, but
failed to update the #ifdef stanzas guarding the defragmentation related
fields and code in skbuff and conntrack related code in nf_defrag_ipv6.c.
This patch adds the required #ifdefs so that IPv6 tproxy can truly be used
without connection tracking.
Original report:
http://marc.info/?l=linux-netdev&m=129010118516341&w=2
Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: KOVACS Krisztian <hidden@balabit.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce skb_checksum_start_offset() to replace repetitive calculation.
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move the calcualation of the Tx hash for a given hash range into a separate
function and define the skb_tx_hash(), which calculates a Tx hash for a
[0; dev->real_num_tx_queues - 1] hash values range, using this
function (__skb_tx_hash()).
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In dev_pick_tx, don't do work in calculating queue
index or setting
the index in the sock unless the device has more than one queue. This
allows the sock to be set only with a queue index of a multi-queue
device which is desirable if device are stacked like in a tunnel.
We also allow the mapping of a socket to queue to be changed. To
maintain in order packet transmission a flag (ooo_okay) has been
added to the sk_buff structure. If a transport layer sets this flag
on a packet, the transmit queue can be changed for the socket.
Presumably, the transport would set this if there was no possbility
of creating OOO packets (for instance, there are no packets in flight
for the socket). This patch includes the modification in TCP output
for setting this flag.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is no point using RCU for dst we allocate for a very short time
(used once).
Change dst_release() to take DST_NOCACHE into account, but also change
skb_dst_set_noref() to force a refcount increment for such dst.
This is a _huge_ gain, because we dont waste memory to store xx thousand
of dsts. Instead of queueing them to RCU, we can free them instantly.
CPU caches can stay hot, re-using same memory blocks to hold temporary
dsts.
Note : remove unneeded smp_mb__before_atomic_dec(); in dst_release(),
since atomic_dec_return() implies a full memory barrier.
Stress test, 160.000.000 udp frames sent, IP route cache disabled
(DDOS).
Before:
real 0m38.091s
user 0m13.189s
sys 7m53.018s
After:
real 0m29.946s
user 0m12.157s
sys 7m40.605s
For reference, if IP route cache was enabled :
real 0m32.030s
user 0m10.521s
sys 8m15.243s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit b30973f877 (node-aware skb allocation) spread a wrong habit of
allocating net drivers skbs on a given memory node : The one closest to
the NIC hardware. This is wrong because as soon as we try to scale
network stack, we need to use many cpus to handle traffic and hit
slub/slab management on cross-node allocations/frees when these cpus
have to alloc/free skbs bound to a central node.
skb allocated in RX path are ephemeral, they have a very short
lifetime : Extra cost to maintain NUMA affinity is too expensive. What
appeared as a nice idea four years ago is in fact a bad one.
In 2010, NIC hardwares are multiqueue, or we use RPS to spread the load,
and two 10Gb NIC might deliver more than 28 million packets per second,
needing all the available cpus.
Cost of cross-node handling in network and vm stacks outperforms the
small benefit hardware had when doing its DMA transfert in its 'local'
memory node at RX time. Even trying to differentiate the two allocations
done for one skb (the sk_buff on local node, the data part on NIC
hardware node) is not enough to bring good performance.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On 32bit arches, if PAGE_SIZE is smaller than 65536, we can use 16bit
offset and size fields. This patch saves 72 bytes per skb on i386, or
128 bytes after rounding.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change "return (EXPR);" to "return EXPR;"
return is not a function, parentheses are not required.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fresh skbs have ip_summed set to CHECKSUM_NONE (0)
We can avoid setting again skb->ip_summed to CHECKSUM_NONE in drivers.
Introduce skb_checksum_none_assert() helper so that we keep this
assertion documented in driver sources.
Change most occurrences of :
skb->ip_summed = CHECKSUM_NONE;
by :
skb_checksum_none_assert(skb);
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
SKBs can be "fragmented" in two ways, via a page array (called
skb_shinfo(skb)->frags[]) and via a list of SKBs (called
skb_shinfo(skb)->frag_list).
Since skb_has_frags() tests the latter, it's name is confusing
since it sounds more like it's testing the former.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes the abstraction introduced by the union skb_shared_tx in
the shared skb data.
The access of the different union elements at several places led to some
confusion about accessing the shared tx_flags e.g. in skb_orphan_try().
http://marc.info/?l=linux-netdev&m=128084897415886&w=2
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Factor out flow calculation code from get_rps_cpu, since other
functions can use the same code.
Revisions:
v2 (Ben): Separate flow calcuation out and use in select queue.
v3 (Arnd): Don't re-implement MIN.
v4 (Changli): skb->data points to ethernet header in macvtap, and
make a fast path. Tested macvtap with this patch.
v5 (Changli):
- Cache skb->rxhash in skb_get_rxhash
- macvtap may not have pow(2) queues, so change code for
queue selection.
(Arnd):
- Use first available queue if all fails.
Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit fc6055a5ba (net: Introduce
skb_orphan_try()) allows an early orphan of the skb and takes care on
tx timestamping, which needs the sk-reference in the skb on driver level.
So does the can-raw socket, which has not been taken into account here.
The patch below adds a 'prevent_sk_orphan' bit in the skb tx shared info,
which fixes the problem discovered by Matthias Fuchs here:
http://marc.info/?t=128030411900003&r=1&w=2
Even if it's not a primary tx timestamp topic it fits well into some skb
shared tx context. Or should be find a different place for the information to
protect the sk reference until it reaches the driver level?
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move frags[] at the end of struct skb_shared_info, and make
pskb_expand_head() copy only the used part of it instead of whole array.
This should avoid kmemcheck warnings and speedup pskb_expand_head() as
well, avoiding a lot of cache misses.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a new networking option to allow hardware time stamps
from PHY devices. When enabled, likely candidates among incoming and
outgoing network packets are offered to the PHY driver for possible
time stamping. When accepted by the PHY driver, incoming packets are
deferred for later delivery by the driver.
The patch also adds phylib driver methods for the SIOCSHWTSTAMP ioctl
and callbacks for transmit and receive time stamping. Drivers may
optionally implement these functions.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a hook for transmit time stamps. The transmit hook
allows a software fallback for transmit time stamps, for MACs
lacking time stamping hardware. Using the hook will still require
adding an inline function call to each MAC driver.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@davemloft.net>
In old kernels, NET_SKB_PAD was defined to 16.
Then commit d6301d3dd1 (net: Increase default NET_SKB_PAD to 32), and
commit 18e8c134f4 (net: Increase NET_SKB_PAD to 64 bytes) increased it
to 64.
While first patch was governed by network stack needs, second was more
driven by performance issues on current hardware. Real intent was to
align data on a cache line boundary.
So use max(32, L1_CACHE_BYTES) instead of 64, to be more generic.
Remove microblaze and powerpc own NET_SKB_PAD definitions.
Thanks to Alexander Duyck and David Miller for their comments.
Suggested-by: David Miller <davem@davemloft.net>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, the accelerated receive path for VLAN's will
drop packets if the real device is an inactive slave and
is not one of the special pkts tested for in
skb_bond_should_drop(). This behavior is different then
the non-accelerated path and for pkts over a bonded vlan.
For example,
vlanx -> bond0 -> ethx
will be dropped in the vlan path and not delivered to any
packet handlers at all. However,
bond0 -> vlanx -> ethx
and
bond0 -> ethx
will be delivered to handlers that match the exact dev,
because the VLAN path checks the real_dev which is not a
slave and netif_recv_skb() doesn't drop frames but only
delivers them to exact matches.
This patch adds a sk_buff flag which is used for tagging
skbs that would previously been dropped and allows the
skb to continue to skb_netif_recv(). Here we add
logic to check for the deliver_no_wcard flag and if it
is set only deliver to handlers that match exactly. This
makes both paths above consistent and gives pkt handlers
a way to identify skbs that come from inactive slaves.
Without this patch in some configurations skbs will be
delivered to handlers with exact matches and in others
be dropped out right in the vlan path.
I have tested the following 4 configurations in failover modes
and load balancing modes.
# bond0 -> ethx
# vlanx -> bond0 -> ethx
# bond0 -> vlanx -> ethx
# bond0 -> ethx
|
vlanx -> --
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can avoid an unecessary cache miss by checking if the skb is non-linear
before accessing gso_size/gso_type in skb_warn_if_lro, the same can also be
done to avoid a cache miss on nr_frags if data_len is 0.
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>
Use low order bit of skb->_skb_dst to tell dst is not refcounted.
Change _skb_dst to _skb_refdst to make sure all uses are catched.
skb_dst() returns the dst, regardless of noref bit set or not, but
with a lockdep check to make sure a noref dst is not given if current
user is not rcu protected.
New skb_dst_set_noref() helper to set an notrefcounted dst on a skb.
(with lockdep check)
skb_dst_drop() drops a reference only if skb dst was refcounted.
skb_dst_force() helper is used to force a refcount on dst, when skb
is queued and not anymore RCU protected.
Use skb_dst_force() in __sk_add_backlog(), __dev_xmit_skb() if
!IFF_XMIT_DST_RELEASE or skb enqueued on qdisc queue, in
sock_queue_rcv_skb(), in __nf_queue().
Use skb_dst_force() in dev_requeue_skb().
Note: dst_use_noref() still dirties dst, we might transform it
later to do one dirtying per jiffies.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
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>
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>
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>
