1. 07 Apr, 2013 12 commits
  2. 06 Apr, 2013 1 commit
  3. 05 Apr, 2013 10 commits
  4. 04 Apr, 2013 1 commit
    • Jesper Dangaard Brouer's avatar
      net: frag queue per hash bucket locking · 19952cc4
      Jesper Dangaard Brouer authored
      This patch implements per hash bucket locking for the frag queue
      hash.  This removes two write locks, and the only remaining write
      lock is for protecting hash rebuild.  This essentially reduce the
      readers-writer lock to a rebuild lock.
      
      This patch is part of "net: frag performance followup"
       http://thread.gmane.org/gmane.linux.network/263644
      of which two patches have already been accepted:
      
      Same test setup as previous:
       (http://thread.gmane.org/gmane.linux.network/257155)
       Two 10G interfaces, on seperate NUMA nodes, are under-test, and uses
       Ethernet flow-control.  A third interface is used for generating the
       DoS attack (with trafgen).
      
      Notice, I have changed the frag DoS generator script to be more
      efficient/deadly.  Before it would only hit one RX queue, now its
      sending packets causing multi-queue RX, due to "better" RX hashing.
      
      Test types summary (netperf UDP_STREAM):
       Test-20G64K     == 2x10G with 65K fragments
       Test-20G3F      == 2x10G with 3x fragments (3*1472 bytes)
       Test-20G64K+DoS == Same as 20G64K with frag DoS
       Test-20G3F+DoS  == Same as 20G3F  with frag DoS
       Test-20G64K+MQ  == Same as 20G64K with Multi-Queue frag DoS
       Test-20G3F+MQ   == Same as 20G3F  with Multi-Queue frag DoS
      
      When I rebased this-patch(03) (on top of net-next commit a210576c) and
      removed the _bh spinlock, I saw a performance regression.  BUT this
      was caused by some unrelated change in-between.  See tests below.
      
      Test (A) is what I reported before for patch-02, accepted in commit 1b5ab0de.
      Test (B) verifying-retest of commit 1b5ab0de corrospond to patch-02.
      Test (C) is what I reported before for this-patch
      
      Test (D) is net-next master HEAD (commit a210576c), which reveals some
      (unknown) performance regression (compared against test (B)).
      Test (D) function as a new base-test.
      
      Performance table summary (in Mbit/s):
      
      (#) Test-type:  20G64K    20G3F    20G64K+DoS  20G3F+DoS  20G64K+MQ 20G3F+MQ
          ----------  -------   -------  ----------  ---------  --------  -------
      (A) Patch-02  : 18848.7   13230.1   4103.04     5310.36     130.0    440.2
      (B) 1b5ab0de  : 18841.5   13156.8   4101.08     5314.57     129.0    424.2
      (C) Patch-03v1: 18838.0   13490.5   4405.11     6814.72     196.6    461.6
      
      (D) a210576c  : 18321.5   11250.4   3635.34     5160.13     119.1    405.2
      (E) with _bh  : 17247.3   11492.6   3994.74     6405.29     166.7    413.6
      (F) without bh: 17471.3   11298.7   3818.05     6102.11     165.7    406.3
      
      Test (E) and (F) is this-patch(03), with(V1) and without(V2) the _bh spinlocks.
      
      I cannot explain the slow down for 20G64K (but its an artificial
      "lab-test" so I'm not worried).  But the other results does show
      improvements.  And test (E) "with _bh" version is slightly better.
      Signed-off-by: default avatarJesper Dangaard Brouer <brouer@redhat.com>
      Acked-by: default avatarHannes Frederic Sowa <hannes@stressinduktion.org>
      Acked-by: default avatarEric Dumazet <edumazet@google.com>
      
      ----
      V2:
      - By analysis from Hannes Frederic Sowa and Eric Dumazet, we don't
        need the spinlock _bh versions, as Netfilter currently does a
        local_bh_disable() before entering inet_fragment.
      - Fold-in desc from cover-mail
      V3:
      - Drop the chain_len counter per hash bucket.
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      19952cc4
  5. 03 Apr, 2013 5 commits
    • David S. Miller's avatar
      Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net · d6624832
      David S. Miller authored
      Pull net into net-next to get the synchronize_net() bug fix in
      bonding.
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      d6624832
    • Linus Torvalds's avatar
      Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net · da241efc
      Linus Torvalds authored
      Pull networking fixes from David Miller:
      
       1) Fix VSOCK layer handling of context ID changes, from Reilly Grant.
      
       2) Now that we have a synchronize_net() in netdev_rx_handler_unregister(),
          we can't let any call sites hold locks.  Unfortunately bonding does,
          so we have to drop the rwlock there a little bit earlier, fix from
          Veaceslav Falico.
      
       3) MAC address setting loop exits one iteration too early in mlx4
          driver, from Yan Burman.
      
       4) Restore ipv6 routes properly upon ifdown/ifup of loopback, from
          Balakumaran Kannan.
      
      * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:
        VSOCK: Handle changes to the VMCI context ID.
        net IPv6 : Fix broken IPv6 routing table after loopback down-up
        cbq: incorrect processing of high limits
        net/mlx4_en: Fix setting initial MAC address
        bonding: get netdev_rx_handler_unregister out of locks
      da241efc
    • Linus Torvalds's avatar
      Merge tag 'regmap-v3.9-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap · 6e8517a9
      Linus Torvalds authored
      Pull regmap fixes from Mark Brown:
       "A small collection of fixes.  The most important ones are those from
        Stephen and Lars-Peter both of which fix cache issues that have been
        lurking for a while but not manifesting noticably enough for anyone to
        report them."
      
      * tag 'regmap-v3.9-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap:
        regmap: async: Add missing return
        regmap: don't corrupt work buffer in _regmap_raw_write()
        regmap: cache Fix regcache-rbtree sync
        regmap: Initialize `map->debugfs' before regcache
      6e8517a9
    • Linus Torvalds's avatar
      Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux · bd709bd0
      Linus Torvalds authored
      Pull DRM fixes from Dave Airlie:
       "Two core fixes, both regressions, along with some intel and some
        nouveau fixes for regressions and oopses"
      
      * 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
        drm: correctly restore mappings if drm_open fails
        drm/nouveau: fix NULL ptr dereference from nv50_disp_intr()
        drm/nouveau: fix handling empty channel list in ioctl's
        drm: don't unlock in the addfb error paths
        drm/i915: Fix build failure
        drm/i915: Be sure to turn hsync/vsync back on at crt enable (v2)
        drm/i915: duct-tape locking when eDP init fails
      bd709bd0
    • Linus Torvalds's avatar
      Merge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus · aea7fab8
      Linus Torvalds authored
      Pull MIPS fixes from Ralf Baechle:
       "A collection of fixes pretty much across the MIPS code.  Even the
        change to include/linux/signal.h by David Howells' 2a148698 ("Fix
        breakage in MIPS siginfo handling") should be considered MIPS-specific
        as it touches an ifdefed segment that is only relevant to MIPS and
        which unfortunately can't be made to go away entirely."
      
      * 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
        Fix breakage in MIPS siginfo handling
        Revert "MIPS: BCM63XX: Call board_register_device from device_initcall()"
        MIPS: BCM63XX: Make nvram checksum failure non fatal
        MIPS: Fix code generation for non-DSP capable CPUs
        MIPS: Fix inconsistent formatting inside /proc/cpuinfo
        MIPS: SEAD3: Enable LL/SC.
        MIPS: Get rid of CONFIG_CPU_HAS_LLSC again
        MIPS: Add dependencies for HAVE_ARCH_TRANSPARENT_HUGEPAGE
        MIPS: VR4133: Fix probe for LL/SC.
        MIPS: Fix logic errors in bitops.c
        MIPS: Use CONFIG_CPU_MIPSR2 in csum_partial.S
        MIPS: compat: Return same error ENOSYS as native for invalid operation.
      aea7fab8
  6. 02 Apr, 2013 11 commits