1. 20 Sep, 2017 5 commits
    • Xin Long's avatar
      ip6_gre: update mtu properly in ip6gre_err · c9a166be
      Xin Long authored
      
      [ Upstream commit 5c25f30c ]
      
      Now when probessing ICMPV6_PKT_TOOBIG, ip6gre_err only subtracts the
      offset of gre header from mtu info. The expected mtu of gre device
      should also subtract gre header. Otherwise, the next packets still
      can't be sent out.
      
      Jianlin found this issue when using the topo:
        client(ip6gre)<---->(nic1)route(nic2)<----->(ip6gre)server
      
      and reducing nic2's mtu, then both tcp and sctp's performance with
      big size data became 0.
      
      This patch is to fix it by also subtracting grehdr (tun->tun_hlen)
      from mtu info when updating gre device's mtu in ip6gre_err(). It
      also needs to subtract ETH_HLEN if gre dev'type is ARPHRD_ETHER.
      Reported-by: default avatarJianlin Shi <jishi@redhat.com>
      Signed-off-by: default avatarXin Long <lucien.xin@gmail.com>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      c9a166be
    • Jason Wang's avatar
      vhost_net: correctly check tx avail during rx busy polling · b076d251
      Jason Wang authored
      
      [ Upstream commit 8b949bef ]
      
      We check tx avail through vhost_enable_notify() in the past which is
      wrong since it only checks whether or not guest has filled more
      available buffer since last avail idx synchronization which was just
      done by vhost_vq_avail_empty() before. What we really want is checking
      pending buffers in the avail ring. Fix this by calling
      vhost_vq_avail_empty() instead.
      
      This issue could be noticed by doing netperf TCP_RR benchmark as
      client from guest (but not host). With this fix, TCP_RR from guest to
      localhost restores from 1375.91 trans per sec to 55235.28 trans per
      sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz).
      
      Fixes: 03088137 ("vhost_net: basic polling support")
      Signed-off-by: default avatarJason Wang <jasowang@redhat.com>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      b076d251
    • Claudiu Manoil's avatar
      gianfar: Fix Tx flow control deactivation · 80b25b4b
      Claudiu Manoil authored
      
      [ Upstream commit 5d621672 ]
      
      The wrong register is checked for the Tx flow control bit,
      it should have been maccfg1 not maccfg2.
      This went unnoticed for so long probably because the impact is
      hardly visible, not to mention the tangled code from adjust_link().
      First, link flow control (i.e. handling of Rx/Tx link level pause frames)
      is disabled by default (needs to be enabled via 'ethtool -A').
      Secondly, maccfg2 always returns 0 for tx_flow_oldval (except for a few
      old boards), which results in Tx flow control remaining always on
      once activated.
      
      Fixes: 45b679c9 ("gianfar: Implement PAUSE frame generation support")
      Signed-off-by: default avatarClaudiu Manoil <claudiu.manoil@nxp.com>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      80b25b4b
    • Jesper Dangaard Brouer's avatar
      Revert "net: fix percpu memory leaks" · ecb26e81
      Jesper Dangaard Brouer authored
      
      [ Upstream commit 5a63643e ]
      
      This reverts commit 1d6119ba.
      
      After reverting commit 6d7b857d ("net: use lib/percpu_counter API
      for fragmentation mem accounting") then here is no need for this
      fix-up patch.  As percpu_counter is no longer used, it cannot
      memory leak it any-longer.
      
      Fixes: 6d7b857d ("net: use lib/percpu_counter API for fragmentation mem accounting")
      Fixes: 1d6119ba ("net: fix percpu memory leaks")
      Signed-off-by: default avatarJesper Dangaard Brouer <brouer@redhat.com>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      ecb26e81
    • Jesper Dangaard Brouer's avatar
      Revert "net: use lib/percpu_counter API for fragmentation mem accounting" · 021c60ff
      Jesper Dangaard Brouer authored
      
      [ Upstream commit fb452a1a ]
      
      This reverts commit 6d7b857d.
      
      There is a bug in fragmentation codes use of the percpu_counter API,
      that can cause issues on systems with many CPUs.
      
      The frag_mem_limit() just reads the global counter (fbc->count),
      without considering other CPUs can have upto batch size (130K) that
      haven't been subtracted yet.  Due to the 3MBytes lower thresh limit,
      this become dangerous at >=24 CPUs (3*1024*1024/130000=24).
      
      The correct API usage would be to use __percpu_counter_compare() which
      does the right thing, and takes into account the number of (online)
      CPUs and batch size, to account for this and call __percpu_counter_sum()
      when needed.
      
      We choose to revert the use of the lib/percpu_counter API for frag
      memory accounting for several reasons:
      
      1) On systems with CPUs > 24, the heavier fully locked
         __percpu_counter_sum() is always invoked, which will be more
         expensive than the atomic_t that is reverted to.
      
      Given systems with more than 24 CPUs are becoming common this doesn't
      seem like a good option.  To mitigate this, the batch size could be
      decreased and thresh be increased.
      
      2) The add_frag_mem_limit+sub_frag_mem_limit pairs happen on the RX
         CPU, before SKBs are pushed into sockets on remote CPUs.  Given
         NICs can only hash on L2 part of the IP-header, the NIC-RXq's will
         likely be limited.  Thus, a fair chance that atomic add+dec happen
         on the same CPU.
      
      Revert note that commit 1d6119ba ("net: fix percpu memory leaks")
      removed init_frag_mem_limit() and instead use inet_frags_init_net().
      After this revert, inet_frags_uninit_net() becomes empty.
      
      Fixes: 6d7b857d ("net: use lib/percpu_counter API for fragmentation mem accounting")
      Fixes: 1d6119ba ("net: fix percpu memory leaks")
      Signed-off-by: default avatarJesper Dangaard Brouer <brouer@redhat.com>
      Acked-by: default avatarFlorian Westphal <fw@strlen.de>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      021c60ff
  2. 13 Sep, 2017 29 commits
  3. 10 Sep, 2017 1 commit
  4. 09 Sep, 2017 5 commits