Skip to content

R
raspberrypi-kernel

openeuler / raspberrypi-kernel

通知 13
Star 1
Fork 0
R
raspberrypi-kernel
切换分支/标签
  1. 02 3月, 2011 1 次提交
  3. 28 1月, 2011 1 次提交
  5. 25 1月, 2011 2 次提交
    • M
      net: change netdev->features to u32 · 04ed3e74
      由 Michał Mirosław 提交于 
      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: NMichał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      04ed3e74
    • M
      GRO: fix merging a paged skb after non-paged skbs · d1dc7abf
      由 Michal Schmidt 提交于 
      Suppose that several linear skbs of the same flow were received by GRO. They
were thus merged into one skb with a frag_list. Then a new skb of the same flow
arrives, but it is a paged skb with data starting in its frags[].

Before adding the skb to the frag_list skb_gro_receive() will of course adjust
the skb to throw away the headers. It correctly modifies the page_offset and
size of the frag, but it leaves incorrect information in the skb:
 ->data_len is not decreased at all.
 ->len is decreased only by headlen, as if no change were done to the frag.
Later in a receiving process this causes skb_copy_datagram_iovec() to return
-EFAULT and this is seen in userspace as the result of the recv() syscall.

In practice the bug can be reproduced with the sfc driver. By default the
driver uses an adaptive scheme when it switches between using
napi_gro_receive() (with skbs) and napi_gro_frags() (with pages). The bug is
reproduced when under rx load with enough successful GRO merging the driver
decides to switch from the former to the latter.

Manual control is also possible, so reproducing this is easy with netcat:
 - on machine1 (with sfc): nc -l 12345 > /dev/null
 - on machine2: nc machine1 12345 < /dev/zero
 - on machine1:
   echo 1 > /sys/module/sfc/parameters/rx_alloc_method  # use skbs
   echo 2 > /sys/module/sfc/parameters/rx_alloc_method  # use pages
 - See that nc has quit suddenly.

[v2: Modified by Eric Dumazet to avoid advancing skb->data past the end
     and to use a temporary variable.]
Signed-off-by: NMichal Schmidt <mschmidt@redhat.com>
Acked-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      d1dc7abf
  7. 13 1月, 2011 1 次提交
  9. 17 12月, 2010 1 次提交
  11. 16 12月, 2010 1 次提交
  13. 04 12月, 2010 1 次提交
  15. 17 10月, 2010 1 次提交
    • E
      net: allocate skbs on local node · 564824b0
      由 Eric Dumazet 提交于 
      commit b30973f8 (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: NEric Dumazet <eric.dumazet@gmail.com>
Acked-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      564824b0
  17. 09 9月, 2010 1 次提交
  19. 07 9月, 2010 2 次提交
    • K
      skb: Add tracepoints to freeing skb · 07dc22e7
      由 Koki Sanagi 提交于 
      This patch adds tracepoint to consume_skb and add trace_kfree_skb
before __kfree_skb in skb_free_datagram_locked and net_tx_action.
Combinating with tracepoint on dev_hard_start_xmit, we can check
how long it takes to free transmitted packets. And using it, we can
calculate how many packets driver had at that time. It is useful when
a drop of transmitted packet is a problem.

            sshd-6828  [000] 112689.258154: consume_skb: skbaddr=f2d99bb8
Signed-off-by: NKoki Sanagi <sanagi.koki@jp.fujitsu.com>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Acked-by: NNeil Horman <nhorman@tuxdriver.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Kaneshige Kenji <kaneshige.kenji@jp.fujitsu.com>
Cc: Izumo Taku <izumi.taku@jp.fujitsu.com>
Cc: Kosaki Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Scott Mcmillan <scott.a.mcmillan@intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
LKML-Reference: <4C724364.50903@jp.fujitsu.com>
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
      07dc22e7
    • E
      net: pskb_expand_head() optimization · 1fd63041
      由 Eric Dumazet 提交于 
      pskb_expand_head() blindly takes references on fragments before calling
skb_release_data(), potentially releasing these references.

We can add a fast path, avoiding these atomic operations, if we own the
last reference on skb->head.

Based on a previous patch from David
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      1fd63041
  21. 04 9月, 2010 1 次提交
  23. 02 9月, 2010 2 次提交
    • E
      gro: fix different skb headrooms · 3d3be433
      由 Eric Dumazet 提交于 
      Packets entering GRO might have different headrooms, even for a given
flow (because of implementation details in drivers, like copybreak).
We cant force drivers to deliver packets with a fixed headroom.

1) fix skb_segment()

skb_segment() makes the false assumption headrooms of fragments are same
than the head. When CHECKSUM_PARTIAL is used, this can give csum_start
errors, and crash later in skb_copy_and_csum_dev()

2) allocate a minimal skb for head of frag_list

skb_gro_receive() uses netdev_alloc_skb(headroom + skb_gro_offset(p)) to
allocate a fresh skb. This adds NET_SKB_PAD to a padding already
provided by netdevice, depending on various things, like copybreak.

Use alloc_skb() to allocate an exact padding, to reduce cache line
needs:
NET_SKB_PAD + NET_IP_ALIGN

bugzilla : https://bugzilla.kernel.org/show_bug.cgi?id=16626

Many thanks to Plamen Petrov, testing many debugging patches !
With help of Jarek Poplawski.
Reported-by: NPlamen Petrov <pvp-lsts@fs.uni-ruse.bg>
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
CC: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      3d3be433
    • E
      net: skbuff.c cleanup · 6602cebb
      由 Eric Dumazet 提交于 
      (skb->data - skb->head) can be changed by skb_headroom(skb)

Remove some uses of NET_SKBUFF_DATA_USES_OFFSET, using
(skb_end_pointer(skb) - skb->head) or
(skb_tail_pointer(skb) - skb->head) : compiler does the right thing,
and this is more readable for us ;)

(struct skb_shared_info *) casts in pskb_expand_head() to help memcpy()
to use aligned moves.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      6602cebb
  25. 23 8月, 2010 1 次提交
  27. 19 8月, 2010 1 次提交
  29. 25 7月, 2010 1 次提交
  31. 23 7月, 2010 2 次提交
  33. 13 7月, 2010 1 次提交
  35. 14 6月, 2010 2 次提交
  37. 01 6月, 2010 1 次提交
  39. 29 5月, 2010 2 次提交
  41. 22 5月, 2010 1 次提交
  43. 21 5月, 2010 1 次提交
  45. 18 5月, 2010 1 次提交
    • E
      net: add a noref bit on skb dst · 7fee226a
      由 Eric Dumazet 提交于 
      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: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      7fee226a
  47. 05 5月, 2010 1 次提交
    • E
      net: __alloc_skb() speedup · ec7d2f2c
      由 Eric Dumazet 提交于 
      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: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      ec7d2f2c
  49. 02 5月, 2010 1 次提交
  51. 21 4月, 2010 1 次提交
  53. 17 3月, 2010 1 次提交
    • T
      rps: Receive Packet Steering · 0a9627f2
      由 Tom Herbert 提交于 
      This patch implements software receive side packet steering (RPS).  RPS
distributes the load of received packet processing across multiple CPUs.

Problem statement: Protocol processing done in the NAPI context for received
packets is serialized per device queue and becomes a bottleneck under high
packet load.  This substantially limits pps that can be achieved on a single
queue NIC and provides no scaling with multiple cores.

This solution queues packets early on in the receive path on the backlog queues
of other CPUs.   This allows protocol processing (e.g. IP and TCP) to be
performed on packets in parallel.   For each device (or each receive queue in
a multi-queue device) a mask of CPUs is set to indicate the CPUs that can
process packets. A CPU is selected on a per packet basis by hashing contents
of the packet header (e.g. the TCP or UDP 4-tuple) and using the result to index
into the CPU mask.  The IPI mechanism is used to raise networking receive
softirqs between CPUs.  This effectively emulates in software what a multi-queue
NIC can provide, but is generic requiring no device support.

Many devices now provide a hash over the 4-tuple on a per packet basis
(e.g. the Toeplitz hash).  This patch allow drivers to set the HW reported hash
in an skb field, and that value in turn is used to index into the RPS maps.
Using the HW generated hash can avoid cache misses on the packet when
steering it to a remote CPU.

The CPU mask is set on a per device and per queue basis in the sysfs variable
/sys/class/net/<device>/queues/rx-<n>/rps_cpus.  This is a set of canonical
bit maps for receive queues in the device (numbered by <n>).  If a device
does not support multi-queue, a single variable is used for the device (rx-0).

Generally, we have found this technique increases pps capabilities of a single
queue device with good CPU utilization.  Optimal settings for the CPU mask
seem to depend on architectures and cache hierarcy.  Below are some results
running 500 instances of netperf TCP_RR test with 1 byte req. and resp.
Results show cumulative transaction rate and system CPU utilization.

e1000e on 8 core Intel
   Without RPS: 108K tps at 33% CPU
   With RPS:    311K tps at 64% CPU

forcedeth on 16 core AMD
   Without RPS: 156K tps at 15% CPU
   With RPS:    404K tps at 49% CPU

bnx2x on 16 core AMD
   Without RPS  567K tps at 61% CPU (4 HW RX queues)
   Without RPS  738K tps at 96% CPU (8 HW RX queues)
   With RPS:    854K tps at 76% CPU (4 HW RX queues)

Caveats:
- The benefits of this patch are dependent on architecture and cache hierarchy.
Tuning the masks to get best performance is probably necessary.
- This patch adds overhead in the path for processing a single packet.  In
a lightly loaded server this overhead may eliminate the advantages of
increased parallelism, and possibly cause some relative performance degradation.
We have found that masks that are cache aware (share same caches with
the interrupting CPU) mitigate much of this.
- The RPS masks can be changed dynamically, however whenever the mask is changed
this introduces the possibility of generating out of order packets.  It's
probably best not change the masks too frequently.
Signed-off-by: NTom Herbert <therbert@google.com>

 include/linux/netdevice.h |   32 ++++-
 include/linux/skbuff.h    |    3 +
 net/core/dev.c            |  335 +++++++++++++++++++++++++++++++++++++--------
 net/core/net-sysfs.c      |  225 ++++++++++++++++++++++++++++++-
 net/core/skbuff.c         |    2 +
 5 files changed, 538 insertions(+), 59 deletions(-)
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      0a9627f2
  55. 16 12月, 2009 1 次提交
  57. 21 11月, 2009 1 次提交
  59. 17 11月, 2009 1 次提交
  61. 12 11月, 2009 1 次提交
    • A
      skbuff: Do not allow skb recycling with disabled IRQs · e84af6dd
      由 Anton Vorontsov 提交于 
      NAPI drivers try to recycle SKBs in their polling routine, but we
generally don't know the context in which the polling will be called,
and the skb recycling itself may require IRQs to be enabled.

This patch adds irqs_disabled() test to the skb_recycle_check()
routine, so that we'll not let the drivers hit the skb recycling
path with IRQs disabled.

As a side effect, this patch actually disables skb recycling for some
[broken] drivers. E.g. gianfar driver grabs an irqsave spinlock during
TX ring processing, and then tries to recycle an skb, and that caused
the following badness:

nf_conntrack version 0.5.0 (1008 buckets, 4032 max)
------------[ cut here ]------------
Badness at kernel/softirq.c:143
NIP: c003e3c4 LR: c423a528 CTR: c003e344
...
NIP [c003e3c4] local_bh_enable+0x80/0xc4
LR [c423a528] destroy_conntrack+0xd4/0x13c [nf_conntrack]
Call Trace:
[c15d1b60] [c003e32c] local_bh_disable+0x1c/0x34 (unreliable)
[c15d1b70] [c423a528] destroy_conntrack+0xd4/0x13c [nf_conntrack]
[c15d1b80] [c02c6370] nf_conntrack_destroy+0x3c/0x70
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      e84af6dd
  63. 25 7月, 2009 1 次提交
  65. 18 6月, 2009 2 次提交