)]}'
{
  "log": [
    {
      "commit": "c9d26423e56ce1ab4d786f92aebecf859d419293",
      "tree": "9ad3b09055d299c220fd17fa2715143e859de7f8",
      "parents": [
        "a11c5c9ef6dc562fc7df7aaf7911569a85f4d71c",
        "af5b7e84d022fdea373038d831bb4ca2c0e82108"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 18:13:46 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 18:13:46 2014 -0600"
      },
      "message": "Merge tag \u0027pm+acpi-3.17-rc1-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull more ACPI and power management updates from Rafael Wysocki:\n \"These are a couple of regression fixes, cpuidle menu governor\n  optimizations, fixes for ACPI proccessor and battery drivers,\n  hibernation fix to avoid problems related to the e820 memory map,\n  fixes for a few cpufreq drivers and a new version of the suspend\n  profiling tool analyze_suspend.py.\n\n  Specifics:\n\n   - Fix for an ACPI-based device hotplug regression introduced in 3.14\n     that causes a kernel panic to trigger when memory hot-remove is\n     attempted with CONFIG_ACPI_HOTPLUG_MEMORY unset from Tang Chen\n\n   - Fix for a cpufreq regression introduced in 3.16 that triggers a\n     \"sleeping function called from invalid context\" bug in\n     dev_pm_opp_init_cpufreq_table() from Stephen Boyd\n\n   - ACPI battery driver fix for a warning message added in 3.16 that\n     prints silly stuff sometimes from Mariusz Ceier\n\n   - Hibernation fix for safer handling of mismatches in the 820 memory\n     map between the configurations during image creation and during the\n     subsequent restore from Chun-Yi Lee\n\n   - ACPI processor driver fix to handle CPU hotplug notifications\n     correctly during system suspend/resume from Lan Tianyu\n\n   - Series of four cpuidle menu governor cleanups that also should\n     speed it up a bit from Mel Gorman\n\n   - Fixes for the speedstep-smi, integrator, cpu0 and arm_big_little\n     cpufreq drivers from Hans Wennborg, Himangi Saraogi, Markus\n     Pargmann and Uwe Kleine-König\n\n   - Version 3.0 of the analyze_suspend.py suspend profiling tool from\n     Todd E Brandt\"\n\n* tag \u0027pm+acpi-3.17-rc1-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n  ACPI / battery: Fix warning message in acpi_battery_get_state()\n  PM / tools: analyze_suspend.py: update to v3.0\n  cpufreq: arm_big_little: fix module license spec\n  cpufreq: speedstep-smi: fix decimal printf specifiers\n  ACPI / hotplug: Check scan handlers in acpi_scan_hot_remove()\n  cpufreq: OPP: Avoid sleeping while atomic\n  cpufreq: cpu0: Do not print error message when deferring\n  cpufreq: integrator: Use set_cpus_allowed_ptr\n  PM / hibernate: avoid unsafe pages in e820 reserved regions\n  ACPI / processor: Make acpi_cpu_soft_notify() process CPU FROZEN events\n  cpuidle: menu: Lookup CPU runqueues less\n  cpuidle: menu: Call nr_iowait_cpu less times\n  cpuidle: menu: Use ktime_to_us instead of reinventing the wheel\n  cpuidle: menu: Use shifts when calculating averages where possible\n"
    },
    {
      "commit": "a11c5c9ef6dc562fc7df7aaf7911569a85f4d71c",
      "tree": "6e0a6c7a259287e201ab3e727c8d85286e93159a",
      "parents": [
        "179c0ac67b9d947d2de69e9f08a743e7c74a8dce",
        "9baa3c34ac4e27f7e062f266f50cc5dbea26a6c1"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 18:10:33 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 18:10:33 2014 -0600"
      },
      "message": "Merge tag \u0027pci-v3.17-changes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci\n\nPull DEFINE_PCI_DEVICE_TABLE removal from Bjorn Helgaas:\n \"Part two of the PCI changes for v3.17:\n\n    - Remove DEFINE_PCI_DEVICE_TABLE macro use (Benoit Taine)\n\n  It\u0027s a mechanical change that removes uses of the\n  DEFINE_PCI_DEVICE_TABLE macro.  I waited until later in the merge\n  window to reduce conflicts, but it\u0027s possible you\u0027ll still see a few\"\n\n* tag \u0027pci-v3.17-changes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:\n  PCI: Remove DEFINE_PCI_DEVICE_TABLE macro use\n"
    },
    {
      "commit": "179c0ac67b9d947d2de69e9f08a743e7c74a8dce",
      "tree": "3f3e22af02df6d8e685fb8aa89731f3746306e35",
      "parents": [
        "ad15afb8b9e80bec6c4337369f63207ff59c44ae",
        "10cf15e1d1289aa0bf1d26e9f55176b4c7c5c512"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 17:28:08 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 17:28:08 2014 -0600"
      },
      "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc\n\nPull Sparc fixes from David Miller:\n \"Hook up the memfd syscall, and properly claim all PCI resources\n  discovered when building the PCI device tree\"\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:\n  sparc: Hook up memfd_create system call.\n  sparc64: Properly claim resources as each PCI bus is probed.\n  sparc64: Skip bogus PCI bridge ranges.\n  sparc64: Expand PCI bridge probing debug logging.\n"
    },
    {
      "commit": "ad15afb8b9e80bec6c4337369f63207ff59c44ae",
      "tree": "cdd378c1a6f07cc580033730995e2c883379248d",
      "parents": [
        "899552d6e84babd24611fd36ac7051068cb1eb2d",
        "a61ebdfdb13a051f707b408d464f63b991aa21e3"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 17:25:21 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 17:25:21 2014 -0600"
      },
      "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nPull networking fixes from David Miller:\n \"I\u0027m sending this out, in particular, to get the iwlwifi fix\n  propagated:\n\n   1) Fix build due to missing include in i40e driver, from Lucas\n      Tanure.\n\n   2) Memory leak in openvswitch port allocation, from Chirstoph Jaeger.\n\n   3) Check DMA mapping errors in myri10ge, from Stanislaw Gruszka.\n\n   4) Fix various deadlock scenerios in sunvnet driver, from Sowmini\n      Varadhan.\n\n   5) Fix cxgb4i build failures with incompatible Kconfig settings of\n      the driver vs ipv6, from Anish Bhatt.\n\n   6) Fix generation of ACK packet timestamps in the presence of TSO\n      which will be split up, from Willem de Bruijn.\n\n   7) Don\u0027t enable sched scan in iwlwifi driver, it causes firmware\n      crashes in some revisions.  From Emmanuel Grumbach.\n\n   8) Revert a macvlan simplification that causes crashes.\n\n   9) Handle RTT calculations properly in the presence of repair\u0027d SKBs,\n      from Andrey Vagin.\n\n  10) SIT tunnel lookup uses wrong device index in compares, from\n      Shmulik Ladkani.\n\n  11) Handle MTU reductions in TCP properly for ipv4 mapped ipv6\n      sockets, from Neal Cardwell.\n\n  12) Add missing annotations in rhashtable code, from Thomas Graf.\n\n  13) Fix false interpretation of two RTOs as being from the same TCP\n      loss event in the FRTO code, from Neal Cardwell\"\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (42 commits)\n  netlink: Annotate RCU locking for seq_file walker\n  rhashtable: fix annotations for rht_for_each_entry_rcu()\n  rhashtable: unexport and make rht_obj() static\n  rhashtable: RCU annotations for next pointers\n  tcp: fix ssthresh and undo for consecutive short FRTO episodes\n  tcp: don\u0027t allow syn packets without timestamps to pass tcp_tw_recycle logic\n  tcp: fix tcp_release_cb() to dispatch via address family for mtu_reduced()\n  sit: Fix ipip6_tunnel_lookup device matching criteria\n  net: ethernet: ibm: ehea: Remove duplicate object from Makefile\n  net: xgene: Check negative return value of xgene_enet_get_ring_size()\n  tcp: don\u0027t use timestamp from repaired skb-s to calculate RTT (v2)\n  net: xilinx: Remove .owner field for driver\n  Revert \"macvlan: simplify the structure port\"\n  iwlwifi: mvm: disable scheduled scan to prevent firmware crash\n  xen-netback: remove loop waiting function\n  xen-netback: don\u0027t stop dealloc kthread too early\n  xen-netback: move NAPI add/remove calls\n  xen-netback: fix debugfs entry creation\n  xen-netback: fix debugfs write length check\n  net-timestamp: fix missing tcp fragmentation cases\n  ...\n"
    },
    {
      "commit": "a61ebdfdb13a051f707b408d464f63b991aa21e3",
      "tree": "4e175cc207463251e4335fdab7dff9fb056f199d",
      "parents": [
        "9ce12eb16ffb143f3a509da86283ddd0b10bcdb3",
        "77b2f2865956b7573f9b040db7a9f808b434acd1"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:17:37 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:17:37 2014 -0700"
      },
      "message": "Merge tag \u0027master-2014-08-14\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless\n\nJohn W. Linville says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\npull request: wireless 2014-08-14\n\nPlease pull this batch of fixes intended for the 3.17 stream...\n\nArend van Spriel brings two brcmfmac fixes, one which fixes a memory\nleak and one which corrects some merge damage.\n\nEmmanuel Grumbach fixes Linus\u0027s iwlwifi firmware-related log spam.\n\nRickard Strandqvist does some proper NULL termination after a call\nto strncpy.\n\nRonald Wahl corrects a carl9170 problem with sending URBs with the\nwrong endpoint type (resulting in log spam).\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "9ce12eb16ffb143f3a509da86283ddd0b10bcdb3",
      "tree": "c7a478ce8b7854149a2d2647fdc820622b5becf5",
      "parents": [
        "93f560811e80216e98f3fcec220aa0f8836b09af"
      ],
      "author": {
        "name": "Thomas Graf",
        "email": "tgraf@suug.ch",
        "time": "Wed Aug 13 16:38:32 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:13:40 2014 -0700"
      },
      "message": "netlink: Annotate RCU locking for seq_file walker\n\nSilences the following sparse warnings:\nnet/netlink/af_netlink.c:2926:21: warning: context imbalance in \u0027netlink_seq_start\u0027 - wrong count at exit\nnet/netlink/af_netlink.c:2972:13: warning: context imbalance in \u0027netlink_seq_stop\u0027 - unexpected unlock\n\nSigned-off-by: Thomas Graf \u003ctgraf@suug.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "93f560811e80216e98f3fcec220aa0f8836b09af",
      "tree": "c04384f2ba1a9d07431c1705bc0f3462066c1f25",
      "parents": [
        "c91eee56dc4f8c3d9ae834bacb835596d47a709e"
      ],
      "author": {
        "name": "Thomas Graf",
        "email": "tgraf@suug.ch",
        "time": "Wed Aug 13 16:38:31 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:13:39 2014 -0700"
      },
      "message": "rhashtable: fix annotations for rht_for_each_entry_rcu()\n\nCall rcu_deference_raw() directly from within rht_for_each_entry_rcu()\nas list_for_each_entry_rcu() does.\n\nFixes the following sparse warnings:\nnet/netlink/af_netlink.c:2906:25:    expected struct rhash_head const *__mptr\nnet/netlink/af_netlink.c:2906:25:    got struct rhash_head [noderef] \u003casn:4\u003e*\u003cnoident\u003e\n\nFixes: e341694e3eb57fc (\"netlink: Convert netlink_lookup() to use RCU protected hash table\")\nSigned-off-by: Thomas Graf \u003ctgraf@suug.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "c91eee56dc4f8c3d9ae834bacb835596d47a709e",
      "tree": "f4bdff8e42f8df9aa30cb8d058a172c33823516c",
      "parents": [
        "5300fdcb7b7e97d83033bc7196582705524d35ea"
      ],
      "author": {
        "name": "Thomas Graf",
        "email": "tgraf@suug.ch",
        "time": "Wed Aug 13 16:38:30 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:13:39 2014 -0700"
      },
      "message": "rhashtable: unexport and make rht_obj() static\n\nNo need to export rht_obj(), all inner to outer object translations\noccur internally. It was intended to be used with rht_for_each() which\nnow primarily serves as the iterator for rhashtable_remove_pprev() to\neffectively flush and free the full table.\n\nSigned-off-by: Thomas Graf \u003ctgraf@suug.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "5300fdcb7b7e97d83033bc7196582705524d35ea",
      "tree": "4aff86f9863ee5b4bf21827580c7423c515647d5",
      "parents": [
        "0c9ab09223fe9922baeb22546c9a90d774a4bde6"
      ],
      "author": {
        "name": "Thomas Graf",
        "email": "tgraf@suug.ch",
        "time": "Wed Aug 13 16:38:29 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 15:13:39 2014 -0700"
      },
      "message": "rhashtable: RCU annotations for next pointers\n\nProperly annotate next pointers as access is RCU protected in\nthe lookup path.\n\nSigned-off-by: Thomas Graf \u003ctgraf@suug.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "0c9ab09223fe9922baeb22546c9a90d774a4bde6",
      "tree": "d94a2e28f5f1c32fa56ad41655722fd06e64c953",
      "parents": [
        "a26552afe89438eefbe097512b3187f2c7e929fd"
      ],
      "author": {
        "name": "Neal Cardwell",
        "email": "ncardwell@google.com",
        "time": "Thu Aug 14 16:13:07 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:55 2014 -0700"
      },
      "message": "tcp: fix ssthresh and undo for consecutive short FRTO episodes\n\nFix TCP FRTO logic so that it always notices when snd_una advances,\nindicating that any RTO after that point will be a new and distinct\nloss episode.\n\nPreviously there was a very specific sequence that could cause FRTO to\nfail to notice a new loss episode had started:\n\n(1) RTO timer fires, enter FRTO and retransmit packet 1 in write queue\n(2) receiver ACKs packet 1\n(3) FRTO sends 2 more packets\n(4) RTO timer fires again (should start a new loss episode)\n\nThe problem was in step (3) above, where tcp_process_loss() returned\nearly (in the spot marked \"Step 2.b\"), so that it never got to the\nlogic to clear icsk_retransmits. Thus icsk_retransmits stayed\nnon-zero. Thus in step (4) tcp_enter_loss() would see the non-zero\nicsk_retransmits, decide that this RTO is not a new episode, and\ndecide not to cut ssthresh and remember the current cwnd and ssthresh\nfor undo.\n\nThere were two main consequences to the bug that we have\nobserved. First, ssthresh was not decreased in step (4). Second, when\nthere was a series of such FRTO (1-4) sequences that happened to be\nfollowed by an FRTO undo, we would restore the cwnd and ssthresh from\nbefore the entire series started (instead of the cwnd and ssthresh\nfrom before the most recent RTO). This could result in cwnd and\nssthresh being restored to values much bigger than the proper values.\n\nSigned-off-by: Neal Cardwell \u003cncardwell@google.com\u003e\nSigned-off-by: Yuchung Cheng \u003cycheng@google.com\u003e\nFixes: e33099f96d99c (\"tcp: implement RFC5682 F-RTO\")\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "a26552afe89438eefbe097512b3187f2c7e929fd",
      "tree": "714af26aeea98f9c3faa2d99f4ca979979f1eac2",
      "parents": [
        "4fab9071950c2021d846e18351e0f46a1cffd67b"
      ],
      "author": {
        "name": "Hannes Frederic Sowa",
        "email": "hannes@stressinduktion.org",
        "time": "Thu Aug 14 22:06:12 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "tcp: don\u0027t allow syn packets without timestamps to pass tcp_tw_recycle logic\n\ntcp_tw_recycle heavily relies on tcp timestamps to build a per-host\nordering of incoming connections and teardowns without the need to\nhold state on a specific quadruple for TCP_TIMEWAIT_LEN, but only for\nthe last measured RTO. To do so, we keep the last seen timestamp in a\nper-host indexed data structure and verify if the incoming timestamp\nin a connection request is strictly greater than the saved one during\nlast connection teardown. Thus we can verify later on that no old data\npackets will be accepted by the new connection.\n\nDuring moving a socket to time-wait state we already verify if timestamps\nwhere seen on a connection. Only if that was the case we let the\ntime-wait socket expire after the RTO, otherwise normal TCP_TIMEWAIT_LEN\nwill be used. But we don\u0027t verify this on incoming SYN packets. If a\nconnection teardown was less than TCP_PAWS_MSL seconds in the past we\ncannot guarantee to not accept data packets from an old connection if\nno timestamps are present. We should drop this SYN packet. This patch\ncloses this loophole.\n\nPlease note, this patch does not make tcp_tw_recycle in any way more\nusable but only adds another safety check:\nSporadic drops of SYN packets because of reordering in the network or\nin the socket backlog queues can happen. Users behing NAT trying to\nconnect to a tcp_tw_recycle enabled server can get caught in blackholes\nand their connection requests may regullary get dropped because hosts\nbehind an address translator don\u0027t have synchronized tcp timestamp clocks.\ntcp_tw_recycle cannot work if peers don\u0027t have tcp timestamps enabled.\n\nIn general, use of tcp_tw_recycle is disadvised.\n\nCc: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Hannes Frederic Sowa \u003channes@stressinduktion.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "4fab9071950c2021d846e18351e0f46a1cffd67b",
      "tree": "f7538ee4e32315a32ab45f3e7bb00e9c6a919bfa",
      "parents": [
        "bc8fc7b8f825ef17a0fb9e68c18ce94fa66ab337"
      ],
      "author": {
        "name": "Neal Cardwell",
        "email": "ncardwell@google.com",
        "time": "Thu Aug 14 12:40:05 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "tcp: fix tcp_release_cb() to dispatch via address family for mtu_reduced()\n\nMake sure we use the correct address-family-specific function for\nhandling MTU reductions from within tcp_release_cb().\n\nPreviously AF_INET6 sockets were incorrectly always using the IPv6\ncode path when sometimes they were handling IPv4 traffic and thus had\nan IPv4 dst.\n\nSigned-off-by: Neal Cardwell \u003cncardwell@google.com\u003e\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nDiagnosed-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nFixes: 563d34d057862 (\"tcp: dont drop MTU reduction indications\")\nReviewed-by: Hannes Frederic Sowa \u003channes@stressinduktion.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "bc8fc7b8f825ef17a0fb9e68c18ce94fa66ab337",
      "tree": "875de67e132fae67722ed9b478710b76f34eaddd",
      "parents": [
        "3b3e0ea8bc0f46e8291295003f70225e5a66acbc"
      ],
      "author": {
        "name": "Shmulik Ladkani",
        "email": "shmulik.ladkani@gmail.com",
        "time": "Thu Aug 14 15:27:20 2014 +0300"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "sit: Fix ipip6_tunnel_lookup device matching criteria\n\nAs of 4fddbf5d78 (\"sit: strictly restrict incoming traffic to tunnel link device\"),\nwhen looking up a tunnel, tunnel\u0027s underlying interface (t-\u003eparms.link)\nis verified to match incoming traffic\u0027s ingress device.\n\nHowever the comparison was incorrectly based on skb-\u003edev-\u003eiflink.\n\nInstead, dev-\u003eifindex should be used, which correctly represents the\ninterface from which the IP stack hands the ipip6 packets.\n\nThis allows setting up sit tunnels bound to vlan interfaces (otherwise\nincoming ipip6 traffic on the vlan interface was dropped due to\nipip6_tunnel_lookup match failure).\n\nSigned-off-by: Shmulik Ladkani \u003cshmulik.ladkani@gmail.com\u003e\nAcked-by: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "3b3e0ea8bc0f46e8291295003f70225e5a66acbc",
      "tree": "3245124fe8ad5be959a06c227182961415df9f32",
      "parents": [
        "9b9ba821b6c47b6d1a711b8f6fff931277217bab"
      ],
      "author": {
        "name": "Andreas Ruprecht",
        "email": "rupran@einserver.de",
        "time": "Thu Aug 14 14:26:23 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "net: ethernet: ibm: ehea: Remove duplicate object from Makefile\n\nIn the Makefile, ehea_phyp.o is included twice in the list of\nobject files compile into ehea.o.\n\nThis change removes one instance.\n\nSigned-off-by: Andreas Ruprecht \u003crupran@einserver.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "9b9ba821b6c47b6d1a711b8f6fff931277217bab",
      "tree": "19fe44ecda226e216ba2c2bb21f897144f897873",
      "parents": [
        "9d186cac7ffb1831e9f34cb4a3a8b22abb9dd9d4"
      ],
      "author": {
        "name": "Tobias Klauser",
        "email": "tklauser@distanz.ch",
        "time": "Thu Aug 14 13:59:42 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "net: xgene: Check negative return value of xgene_enet_get_ring_size()\n\nxgene_enet_get_ring_size() returns a negative value in case of an error,\nbut its only caller in xgene_enet_create_desc_ring() currently uses the\nreturn value directly as u32. Instead, check for a negative value first and\nerror out in case. Also move the call to xgene_enet_get_ring_size() before\ndevm_kzalloc() so we don\u0027t need to free anything in the error path.\n\nThis fixes the following issue reported by the Coverity Scanner:\n\n** CID 1231336:  Improper use of negative value  (NEGATIVE_RETURNS)\n/drivers/net/ethernet/apm/xgene/xgene_enet_main.c: 596 in xgene_enet_create_desc_ring()\n\nSigned-off-by: Tobias Klauser \u003ctklauser@distanz.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "9d186cac7ffb1831e9f34cb4a3a8b22abb9dd9d4",
      "tree": "bb689167aa49110128c81e0ca131abc3a08a1097",
      "parents": [
        "fdd42e4400fcb5fe837a20a8d837ef094d5a26fb"
      ],
      "author": {
        "name": "Andrey Vagin",
        "email": "avagin@openvz.org",
        "time": "Wed Aug 13 16:03:10 2014 +0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "tcp: don\u0027t use timestamp from repaired skb-s to calculate RTT (v2)\n\nWe don\u0027t know right timestamp for repaired skb-s. Wrong RTT estimations\nisn\u0027t good, because some congestion modules heavily depends on it.\n\nThis patch adds the TCPCB_REPAIRED flag, which is included in\nTCPCB_RETRANS.\n\nThanks to Eric for the advice how to fix this issue.\n\nThis patch fixes the warning:\n[  879.562947] WARNING: CPU: 0 PID: 2825 at net/ipv4/tcp_input.c:3078 tcp_ack+0x11f5/0x1380()\n[  879.567253] CPU: 0 PID: 2825 Comm: socket-tcpbuf-l Not tainted 3.16.0-next-20140811 #1\n[  879.567829] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011\n[  879.568177]  0000000000000000 00000000c532680c ffff880039643d00 ffffffff817aa2d2\n[  879.568776]  0000000000000000 ffff880039643d38 ffffffff8109afbd ffff880039d6ba80\n[  879.569386]  ffff88003a449800 000000002983d6bd 0000000000000000 000000002983d6bc\n[  879.569982] Call Trace:\n[  879.570264]  [\u003cffffffff817aa2d2\u003e] dump_stack+0x4d/0x66\n[  879.570599]  [\u003cffffffff8109afbd\u003e] warn_slowpath_common+0x7d/0xa0\n[  879.570935]  [\u003cffffffff8109b0ea\u003e] warn_slowpath_null+0x1a/0x20\n[  879.571292]  [\u003cffffffff816d0a05\u003e] tcp_ack+0x11f5/0x1380\n[  879.571614]  [\u003cffffffff816d10bd\u003e] tcp_rcv_established+0x1ed/0x710\n[  879.571958]  [\u003cffffffff816dc9da\u003e] tcp_v4_do_rcv+0x10a/0x370\n[  879.572315]  [\u003cffffffff81657459\u003e] release_sock+0x89/0x1d0\n[  879.572642]  [\u003cffffffff816c81a0\u003e] do_tcp_setsockopt.isra.36+0x120/0x860\n[  879.573000]  [\u003cffffffff8110a52e\u003e] ? rcu_read_lock_held+0x6e/0x80\n[  879.573352]  [\u003cffffffff816c8912\u003e] tcp_setsockopt+0x32/0x40\n[  879.573678]  [\u003cffffffff81654ac4\u003e] sock_common_setsockopt+0x14/0x20\n[  879.574031]  [\u003cffffffff816537b0\u003e] SyS_setsockopt+0x80/0xf0\n[  879.574393]  [\u003cffffffff817b40a9\u003e] system_call_fastpath+0x16/0x1b\n[  879.574730] ---[ end trace a17cbc38eb8c5c00 ]---\n\nv2: moving setting of skb-\u003ewhen for repaired skb-s in tcp_write_xmit,\n    where it\u0027s set for other skb-s.\n\nFixes: 431a91242d8d (\"tcp: timestamp SYN+DATA messages\")\nFixes: 740b0f1841f6 (\"tcp: switch rtt estimations to usec resolution\")\nCc: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Pavel Emelyanov \u003cxemul@parallels.com\u003e\nCc: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nSigned-off-by: Andrey Vagin \u003cavagin@openvz.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "fdd42e4400fcb5fe837a20a8d837ef094d5a26fb",
      "tree": "1025b7fa6046590a4f0ccd897f574b50f324a342",
      "parents": [
        "5e3c516b512c0f8f18359413b04918f6347f67e7"
      ],
      "author": {
        "name": "Michal Simek",
        "email": "michal.simek@xilinx.com",
        "time": "Wed Aug 13 13:54:22 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:38:54 2014 -0700"
      },
      "message": "net: xilinx: Remove .owner field for driver\n\nThere is no need to init .owner field.\n\nBased on the patch from Peter Griffin \u003cpeter.griffin@linaro.org\u003e\n\"mmc: remove .owner field for drivers using module_platform_driver\"\n\nThis patch removes the superflous .owner field for drivers which\nuse the module_platform_driver API, as this is overriden in\nplatform_driver_register anyway.\"\n\nSigned-off-by: Michal Simek \u003cmichal.simek@xilinx.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "5e3c516b512c0f8f18359413b04918f6347f67e7",
      "tree": "214212e6911756fbcba8c99d8793a4b1a9f58b72",
      "parents": [
        "db8d457acccdb605b82ae22e48d28a47861b0c86"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:32:49 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Thu Aug 14 14:32:49 2014 -0700"
      },
      "message": "Revert \"macvlan: simplify the structure port\"\n\nThis reverts commit a188a54d11629bef2169052297e61f3767ca8ce5.\n\nIt causes crashes\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n[   80.643286] BUG: unable to handle kernel NULL pointer dereference at 0000000000000878\n[   80.670103] IP: [\u003cffffffff810832e4\u003e] try_to_grab_pending+0x64/0x1f0\n[   80.691289] PGD 22c102067 PUD 235bf0067 PMD 0\n[   80.706611] Oops: 0002 [#1] SMP\n[   80.717836] Modules linked in: macvlan nfsd lockd nfs_acl exportfs auth_rpcgss sunrpc oid_registry ioatdma ixgbe(-) mdio igb dca\n[   80.757935] CPU: 37 PID: 6724 Comm: rmmod Not tainted 3.16.0-net-next-08-12-2014-FCoE+ #1\n[   80.785688] Hardware name: Intel Corporation S2600CO/S2600CO, BIOS SE5C600.86B.02.03.0003.041920141333 04/19/2014\n[   80.820310] task: ffff880235a9eae0 ti: ffff88022e844000 task.ti: ffff88022e844000\n[   80.845770] RIP: 0010:[\u003cffffffff810832e4\u003e]  [\u003cffffffff810832e4\u003e] try_to_grab_pending+0x64/0x1f0\n[   80.875326] RSP: 0018:ffff88022e847b28  EFLAGS: 00010046\n[   80.893251] RAX: 0000000000037a6a RBX: 0000000000000878 RCX: 0000000000000000\n[   80.917187] RDX: ffff880235a9eae0 RSI: 0000000000000001 RDI: ffffffff810832db\n[   80.941125] RBP: ffff88022e847b58 R08: 0000000000000000 R09: 0000000000000000\n[   80.965056] R10: 0000000000000001 R11: 0000000000000001 R12: ffff88022e847b70\n[   80.988994] R13: 0000000000000000 R14: ffff88022e847be8 R15: ffffffff81ebe440\n[   81.012929] FS:  00007fab90b07700(0000) GS:ffff88043f7a0000(0000) knlGS:0000000000000000\n[   81.040400] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[   81.059757] CR2: 0000000000000878 CR3: 0000000235a42000 CR4: 00000000001407e0\n[   81.083689] Stack:\n[   81.090739]  ffff880235a9eae0 0000000000000878 ffff88022e847b70 0000000000000000\n[   81.116253]  ffff88022e847be8 ffffffff81ebe440 ffff88022e847b98 ffffffff810847f1\n[   81.141766]  ffff88022e847b78 0000000000000286 ffff880234200000 0000000000000000\n[   81.167282] Call Trace:\n[   81.175768]  [\u003cffffffff810847f1\u003e] __cancel_work_timer+0x31/0x170\n[   81.195985]  [\u003cffffffff8108494b\u003e] cancel_work_sync+0xb/0x10\n[   81.214769]  [\u003cffffffffa015ae68\u003e] macvlan_port_destroy+0x28/0x60 [macvlan]\n[   81.237844]  [\u003cffffffffa015b930\u003e] macvlan_uninit+0x40/0x50 [macvlan]\n[   81.259209]  [\u003cffffffff816bf6e2\u003e] rollback_registered_many+0x1a2/0x2c0\n[   81.281140]  [\u003cffffffff816bf81a\u003e] unregister_netdevice_many+0x1a/0xb0\n[   81.302786]  [\u003cffffffffa015a4ff\u003e] macvlan_device_event+0x1ef/0x240 [macvlan]\n[   81.326439]  [\u003cffffffff8108a13d\u003e] notifier_call_chain+0x4d/0x70\n[   81.346366]  [\u003cffffffff8108a201\u003e] raw_notifier_call_chain+0x11/0x20\n[   81.367439]  [\u003cffffffff816bf25b\u003e] call_netdevice_notifiers_info+0x3b/0x70\n[   81.390228]  [\u003cffffffff816bf2a1\u003e] call_netdevice_notifiers+0x11/0x20\n[   81.411587]  [\u003cffffffff816bf6bd\u003e] rollback_registered_many+0x17d/0x2c0\n[   81.433518]  [\u003cffffffff816bf925\u003e] unregister_netdevice_queue+0x75/0x110\n[   81.455735]  [\u003cffffffff816bfb2b\u003e] unregister_netdev+0x1b/0x30\n[   81.475094]  [\u003cffffffffa0039b50\u003e] ixgbe_remove+0x170/0x1d0 [ixgbe]\n[   81.495886]  [\u003cffffffff813512a2\u003e] pci_device_remove+0x32/0x60\n[   81.515246]  [\u003cffffffff814c75c4\u003e] __device_release_driver+0x64/0xd0\n[   81.536321]  [\u003cffffffff814c76f8\u003e] driver_detach+0xc8/0xd0\n[   81.554530]  [\u003cffffffff814c656e\u003e] bus_remove_driver+0x4e/0xa0\n[   81.573888]  [\u003cffffffff814c828b\u003e] driver_unregister+0x2b/0x60\n[   81.593246]  [\u003cffffffff8135143e\u003e] pci_unregister_driver+0x1e/0xa0\n[   81.613749]  [\u003cffffffffa005db18\u003e] ixgbe_exit_module+0x1c/0x2e [ixgbe]\n[   81.635401]  [\u003cffffffff810e738b\u003e] SyS_delete_module+0x15b/0x1e0\n[   81.655334]  [\u003cffffffff8187a395\u003e] ? sysret_check+0x22/0x5d\n[   81.673833]  [\u003cffffffff810abd2d\u003e] ? trace_hardirqs_on_caller+0x11d/0x1e0\n[   81.696339]  [\u003cffffffff8132bfde\u003e] ? trace_hardirqs_on_thunk+0x3a/0x3f\n[   81.717985]  [\u003cffffffff8187a369\u003e] system_call_fastpath+0x16/0x1b\n[   81.738199] Code: 00 48 83 3d 6e bb da 00 00 48 89 c2 0f 84 67 01 00 00 fa 66 0f 1f 44 00 00 49 89 14 24 e8 b5 4b 02 00 45 84 ed 0f 85 ac 00 00 00 \u003cf0\u003e 0f ba 2b 00 72 1d 31 c0 48 8b 5d d8 4c 8b 65 e0 4c 8b 6d e8\n[   81.807026] RIP  [\u003cffffffff810832e4\u003e] try_to_grab_pending+0x64/0x1f0\n[   81.828468]  RSP \u003cffff88022e847b28\u003e\n[   81.840384] CR2: 0000000000000878\n[   81.851731] ---[ end trace 9f6c7232e3464e11 ]---\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nThis bug could be triggered by these steps:\n\nmodprobe ixgbe ; modprobe macvlan\nip link add link p96p1 address 00:1B:21:6E:06:00 macvlan0 type macvlan\nip link add link p96p1 address 00:1B:21:6E:06:01 macvlan1 type macvlan\nip link add link p96p1 address 00:1B:21:6E:06:02 macvlan2 type macvlan\nip link add link p96p1 address 00:1B:21:6E:06:03 macvlan3 type macvlan\nrmmod ixgbe\n\nReported-by: \"Keller, Jacob E\" \u003cjacob.e.keller@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "899552d6e84babd24611fd36ac7051068cb1eb2d",
      "tree": "9a02d66b74a0a9267bbf6637765517f5e8558497",
      "parents": [
        "3b7b3e6ec5f56118046594d3c62469e7d1d0aadd",
        "5a5da78b3a48120d942c8a18ecc645f6acdf7da6"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:14:29 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:14:29 2014 -0600"
      },
      "message": "Merge branch \u0027misc\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild\n\nPull misc kbuild updates from Michal Marek:\n \"This is the non-critical part of kbuild for 3.17-rc1:\n\n   - make help hint to use make -s with make kernelrelease et al.\n   - moved a kbuild document to Documentation/kbuild where it belongs\n   - four new Coccinelle scripts, one dropped and one fixed\n   - new make kselftest target to run various tests on the kernel\"\n\n* \u0027misc\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:\n  kbuild: kselftest - new make target to build and run kernel selftests\n  Coccinelle: Script to replace if and BUG with BUG_ON\n  Coccinelle: Script to detect incorrect argument to sizeof\n  Coccinelle: Script to use ARRAY_SIZE instead of division of two sizeofs\n  Coccinelle: Script to detect cast after memory allocation\n  coccinelle/null: solve parse error\n  Documentation: headers_install.txt is part of kbuild\n  kbuild: make -s should be used with kernelrelease/kernelversion/image_name\n"
    },
    {
      "commit": "3b7b3e6ec5f56118046594d3c62469e7d1d0aadd",
      "tree": "3a6eb410ee27b660b7c085c81619f9e69fb5f0fa",
      "parents": [
        "e3b1fd56f175526db42ae94c457f29c2fa810aca",
        "164f0d2efaaef835473e74b162eae750341d6a8c"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:12:46 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:12:46 2014 -0600"
      },
      "message": "Merge branch \u0027kbuild\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild\n\nPull kbuild updates from Michal Marek:\n - make clean also considers $(extra-m) and $(extra-) to be consistent\n - cleanup and fixes in scripts/Makefile.host\n - allow to override the name of the Python 2 executable with make\n   PYTHON\u003d... (only needed for ia64 in practice)\n - option to split debugingo into *.dwo files to save disk space if the\n   compiler supports it (CONFIG_DEBUG_INFO_SPLIT)\n - option to use dwarf4 debuginfo if the compiler supports it\n   (CONFIG_DEBUG_INFO_DWARF4)\n - fix for disabling certain warnings with clang\n - fix for unneeded rebuild with dash when a command contains\n   backslashes\n\n* \u0027kbuild\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:\n  kbuild: Fix handling of backslashes in *.cmd files\n  kbuild, LLVMLinux: Supress warnings unless W\u003d1-3\n  Kbuild: Add a option to enable dwarf4 v2\n  kbuild: Support split debug info v4\n  kbuild: allow to override Python command name\n  kbuild: clean-up and bug fix of scripts/Makefile.host\n  kbuild: clean up scripts/Makefile.host\n  kbuild: drop shared library support from Makefile.host\n  kbuild: fix a bug of C++ host program handling\n  kbuild: fix a typo in scripts/Makefile.host\n  scripts/Makefile.clean: clean also $(extra-m) and $(extra-)\n"
    },
    {
      "commit": "e3b1fd56f175526db42ae94c457f29c2fa810aca",
      "tree": "3e2948ca44fb7fd5348244c2a83eef864b3110b4",
      "parents": [
        "0680eb1f485ba5aac2ee02c9f0622239c9a4b16c",
        "d087f6ad724dfbcdc3df8e0191b80d9d8d839e71"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:09:05 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:09:05 2014 -0600"
      },
      "message": "Merge tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull infiniband/rdma updates from Roland Dreier:\n \"Main set of InfiniBand/RDMA updates for 3.17 merge window:\n\n   - MR reregistration support\n   - MAD support for RMPP in userspace\n   - iSER and SRP initiator updates\n   - ocrdma hardware driver updates\n   - other fixes...\"\n\n* tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (52 commits)\n  IB/srp: Fix return value check in srp_init_module()\n  RDMA/ocrdma: report asic-id in query device\n  RDMA/ocrdma: Update sli data structure for endianness\n  RDMA/ocrdma: Obtain SL from device structure\n  RDMA/uapi: Include socket.h in rdma_user_cm.h\n  IB/srpt: Handle GID change events\n  IB/mlx5: Use ARRAY_SIZE instead of sizeof/sizeof[0]\n  IB/mlx4: Use ARRAY_SIZE instead of sizeof/sizeof[0]\n  RDMA/amso1100: Check for integer overflow in c2_alloc_cq_buf()\n  IPoIB: Remove unnecessary test for NULL before debugfs_remove()\n  IB/mad: Add user space RMPP support\n  IB/mad: add new ioctl to ABI to support new registration options\n  IB/mad: Add dev_notice messages for various umad/mad registration failures\n  IB/mad: Update module to [pr|dev]_* style print messages\n  IB/ipoib: Avoid multicast join attempts with invalid P_key\n  IB/umad: Update module to [pr|dev]_* style print messages\n  IB/ipoib: Avoid flushing the workqueue from worker context\n  IB/ipoib: Use P_Key change event instead of P_Key polling mechanism\n  IB/ipath: Add P_Key change event support\n  mlx4_core: Add support for secure-host and SMP firewall\n  ...\n"
    },
    {
      "commit": "0680eb1f485ba5aac2ee02c9f0622239c9a4b16c",
      "tree": "d3878c2a5fe3a3d1709ba8fdd937ea3666093c18",
      "parents": [
        "f2937e4540836be5c4db7e0a721ce3cfc59036e3"
      ],
      "author": {
        "name": "John Stultz",
        "email": "john.stultz@linaro.org",
        "time": "Wed Aug 13 12:47:14 2014 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 11:04:11 2014 -0600"
      },
      "message": "timekeeping: Another fix to the VSYSCALL_OLD update_vsyscall\n\nBenjamin Herrenschmidt pointed out that I further missed modifying\nupdate_vsyscall after the wall_to_mono value was changed to a\ntimespec64.  This causes issues on powerpc32, which expects a 32bit\ntimespec.\n\nThis patch fixes the problem by properly converting from a timespec64 to\na timespec before passing the value on to the arch-specific vsyscall\nlogic.\n\n[ Thomas is currently on vacation, but reviewed it and wanted me to send\n  this fix on to you directly. ]\n\nCc: LKML \u003clinux-kernel@vger.kernel.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Ingo Molnar \u003cmingo@redhat.com\u003e\nCc: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\nReported-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\nReviewed-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: John Stultz \u003cjohn.stultz@linaro.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "f2937e4540836be5c4db7e0a721ce3cfc59036e3",
      "tree": "f7c68473490deabbe334279a132bc87816eaf694",
      "parents": [
        "1d508f8ace3e78e96da70f1cf4e36b80f22b5fa9",
        "a2a368d905472293d4e13d09fdd9e537edc74347"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:56:25 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:56:25 2014 -0600"
      },
      "message": "Merge branch \u0027akpm\u0027 (fixes from Andrew Morton)\n\nMerge leftovers from Andrew Morton:\n \"A few leftovers.\n\n  I have a bunch of OCFS2 patches which are still out for review and\n  which I might sneak along after -rc1.  Partly my fault - I should send\n  my review pokes out earlier\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e:\n  mm: fix CROSS_MEMORY_ATTACH help text grammar\n  drivers/mfd/rtsx_usb.c: export device table\n  mm, hugetlb_cgroup: align hugetlb cgroup limit to hugepage size\n"
    },
    {
      "commit": "a2a368d905472293d4e13d09fdd9e537edc74347",
      "tree": "1a50d83acea375052797d9c1ed2bb59da77134aa",
      "parents": [
        "1813908986e36119228c158aae1c6a0267c99e77"
      ],
      "author": {
        "name": "Geert Uytterhoeven",
        "email": "geert@linux-m68k.org",
        "time": "Tue Aug 12 13:46:11 2014 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:56:15 2014 -0600"
      },
      "message": "mm: fix CROSS_MEMORY_ATTACH help text grammar\n\nSigned-off-by: Geert Uytterhoeven \u003cgeert@linux-m68k.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "1813908986e36119228c158aae1c6a0267c99e77",
      "tree": "e41538cb82546450a898a998ebb7709c433db4c9",
      "parents": [
        "24d7cd207f8e0d6864a314d5f03b5434403ff940"
      ],
      "author": {
        "name": "Jeff Mahoney",
        "email": "jeffm@suse.com",
        "time": "Tue Aug 12 13:46:09 2014 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:56:15 2014 -0600"
      },
      "message": "drivers/mfd/rtsx_usb.c: export device table\n\nThe rtsx_usb driver contains the table for the devices it supports but\ndoesn\u0027t export it.  As a result, no alias is generated and it doesn\u0027t\nget loaded automatically.\n\nVia https://bugzilla.novell.com/show_bug.cgi?id\u003d890096\n\nSigned-off-by: Jeff Mahoney \u003cjeffm@suse.com\u003e\nReported-by: Marcel Witte \u003cwittemar@googlemail.com\u003e\nCc: Roger Tseng \u003crogerable@realtek.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "24d7cd207f8e0d6864a314d5f03b5434403ff940",
      "tree": "906bd7bcf4297a42b94ec7ade6bbee7361404afa",
      "parents": [
        "c8d6637d0497d62093dbba0694c7b3a80b79bfe1"
      ],
      "author": {
        "name": "David Rientjes",
        "email": "rientjes@google.com",
        "time": "Tue Aug 12 13:46:07 2014 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:56:15 2014 -0600"
      },
      "message": "mm, hugetlb_cgroup: align hugetlb cgroup limit to hugepage size\n\nMemcg aligns memory.limit_in_bytes to PAGE_SIZE as part of the resource\ncounter since it makes no sense to allow a partial page to be charged.\n\nAs a result of the hugetlb cgroup using the resource counter, it is also\naligned to PAGE_SIZE but makes no sense unless aligned to the size of\nthe hugepage being limited.\n\nAlign hugetlb cgroup limit to hugepage size.\n\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.cz\u003e\nCc: \"Aneesh Kumar K.V\" \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nCc: Tejun Heo \u003ctj@kernel.org\u003e\nCc: Li Zefan \u003clizefan@huawei.com\u003e\nCc: Michal Hocko \u003cmhocko@suse.cz\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "77b2f2865956b7573f9b040db7a9f808b434acd1",
      "tree": "cbbfa7b95278133202e6223857b0c3b041a16e44",
      "parents": [
        "2da2c5854ed759a64f8a9231a7f544b472c3aa6d"
      ],
      "author": {
        "name": "Emmanuel Grumbach",
        "email": "emmanuel.grumbach@intel.com",
        "time": "Thu Aug 14 19:32:16 2014 +0300"
      },
      "committer": {
        "name": "Emmanuel Grumbach",
        "email": "emmanuel.grumbach@intel.com",
        "time": "Thu Aug 14 19:47:41 2014 +0300"
      },
      "message": "iwlwifi: mvm: disable scheduled scan to prevent firmware crash\n\nThere are firmwares which don\u0027t support scheduled scan.\nDisable it for now.\nLinus\u0027s system encoutered this issue.\nThanks to David Spinadel for his help.\n\nTested-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nSigned-off-by: Emmanuel Grumbach \u003cemmanuel.grumbach@intel.com\u003e\n"
    },
    {
      "commit": "1d508f8ace3e78e96da70f1cf4e36b80f22b5fa9",
      "tree": "1374288c038234bff0bde464be72f0645ed44f33",
      "parents": [
        "2d0c05e1c5a6587a13c910845229ec5cdcd387f9",
        "9e813308a5c18c58f9ccae1ec72ed4e14eaf9025"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:14:07 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:14:07 2014 -0600"
      },
      "message": "Merge branch \u0027next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc\n\nPull more powerpc updates from Ben Herrenschmidt:\n \"Here are some more powerpc bits for 3.17, essentially fixes.\n\n  The biggest series, also aimed at -stable, is from Aneesh and is the\n  result of weeks and weeks of debugging to find out why the heck or THP\n  implementation was occasionally triggering multi-hit errors in our\n  level 1 TLB.  It ended up being a combination of issues including\n  subtleties as to how we should invalidate those special \u0027MPSS\u0027 pages\n  we use to allow the use of 16M pages inside 4K/64K \"base page size\"\n  segments (you really have to love our MMU !)\n\n  Another interesting one in the \"OMG\" category is the series from\n  Michael adding memory barriers to spin_is_locked().  That\u0027s also the\n  result of many days of debugging to figure out why the semaphore code\n  would occasionally crash in ways that made no sense.  It ended up\n  being some creative lock stacking that was defeated by the fact that\n  our locks allow a load inside the locked section to be re-ordered with\n  the load of the lock value itself (I\u0027m still of two mind about whether\n  to kill that once and for all by putting a heavier barrier back into\n  our lock implementation...).  The fixes come with a long explanation\n  in the cset comments, feel free to read it if you feel like having a\n  headache today\"\n\n* \u0027next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (25 commits)\n  powerpc/thp: Add tracepoints to track hugepage invalidate\n  powerpc/mm: Use read barrier when creating real_pte\n  powerpc/thp: Use ACCESS_ONCE when loading pmdp\n  powerpc/thp: Invalidate with vpn in loop\n  powerpc/thp: Handle combo pages in invalidate\n  powerpc/thp: Invalidate old 64K based hash page mapping before insert of 4k pte\n  powerpc/thp: Don\u0027t recompute vsid and ssize in loop on invalidate\n  powerpc/thp: Add write barrier after updating the valid bit\n  powerpc: reorder per-cpu NUMA information\u0027s initialization\n  powerpc/perf/hv-24x7: Use kmem_cache_free\n  powerpc/pseries/hvcserver: Fix endian issue in hvcs_get_partner_info\n  powerpc: Hard disable interrupts in xmon\n  powerpc: remove duplicate definition of TEXASR_FS\n  powerpc/pseries: Avoid deadlock on removing ddw\n  powerpc/pseries: Failure on removing device node\n  powerpc/boot: Use correct zlib types for comparison\n  powerpc/powernv: Interface to register/unregister opal dump region\n  printk: Add function to return log buffer address and size\n  powerpc: Add POWER8 features to CPU_FTRS_POSSIBLE/ALWAYS\n  powerpc/ppc476: Disable BTAC\n  ...\n"
    },
    {
      "commit": "2d0c05e1c5a6587a13c910845229ec5cdcd387f9",
      "tree": "7460e285a0454c3ab7fe78caacdb25e36e8854db",
      "parents": [
        "311bf6d1cbfc288ccd6e6480cbf104da2447b524",
        "ceca89e89ee21a3af050c3ea4c634b3be0e34d51"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:11:58 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:11:58 2014 -0600"
      },
      "message": "Merge tag \u0027hwspinlock-3.17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ohad/hwspinlock\n\nPull hwspinlock updates from Ohad Ben-Cohen:\n \"Two small hwspinlock changes for better OMAP support, coming from\n  Suman Anna\"\n\n* tag \u0027hwspinlock-3.17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ohad/hwspinlock:\n  hwspinlock: enable OMAP build for AM33xx, AM43xx \u0026 DRA7xx\n  hwspinlock/omap: enable module before reading SYSSTATUS register\n"
    },
    {
      "commit": "311bf6d1cbfc288ccd6e6480cbf104da2447b524",
      "tree": "1daf6dea42eee46481feef90a66242df9e768f38",
      "parents": [
        "82f05a08e27755835c948b7a45b86fac9e114e1d",
        "69f6a34bdeea4fec50bb90619bc9602973119572"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:09:48 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 10:09:48 2014 -0600"
      },
      "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security\n\nPull seccomp fix from James Morris.\n\nBUG(!spin_is_locked()) really doesn\u0027t work very well in UP\nconfigurations without any actual spinlock state.  Which is very much\nwhy we have that \"assert_spin_lock()\" function for this.\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:\n  seccomp: Replace BUG(!spin_is_locked()) with assert_spin_lock\n"
    },
    {
      "commit": "d087f6ad724dfbcdc3df8e0191b80d9d8d839e71",
      "tree": "32fb414b570f3f1bdac688931d8a4395df49a503",
      "parents": [
        "9376932d0c26d5f5f89c95d5bd45123bba96d3a9",
        "678ea9b5baab6800692b249bdba77c3c07261d61",
        "e42fa2092c1049ac9c0e38aaac39ef3c40e91a36",
        "8d4aca7f04a098086c6545df158e5a022cbbfcd1",
        "2f0304d21867476394cd51a54e97f7273d112261",
        "1471cb6ca67990a306500e69e52ffb28c93ccbbc",
        "db1044d458a287c18c4d413adc4ad12e92e253b5",
        "a57f23f675bf679b2be093571ae09fe2199c5e76",
        "a8f731ebd1b4f94cf52ff07fe524414b4fbf9793",
        "96c51abecc4379273fb751ff44eb57e4b243b5f0",
        "da05be290f5040e488ea4d13edc4da889f6f6929"
      ],
      "author": {
        "name": "Roland Dreier",
        "email": "roland@purestorage.com",
        "time": "Thu Aug 14 08:58:04 2014 -0700"
      },
      "committer": {
        "name": "Roland Dreier",
        "email": "roland@purestorage.com",
        "time": "Thu Aug 14 08:58:04 2014 -0700"
      },
      "message": "Merge branches \u0027core\u0027, \u0027cxgb4\u0027, \u0027ipoib\u0027, \u0027iser\u0027, \u0027iwcm\u0027, \u0027mad\u0027, \u0027misc\u0027, \u0027mlx4\u0027, \u0027mlx5\u0027, \u0027ocrdma\u0027 and \u0027srp\u0027 into for-next\n"
    },
    {
      "commit": "da05be290f5040e488ea4d13edc4da889f6f6929",
      "tree": "3dd1517388178e0c1660560f448bc58780f5d819",
      "parents": [
        "2aa1cf64aa576671b3593d73a6cb0a28b90ee444"
      ],
      "author": {
        "name": "Wei Yongjun",
        "email": "yongjun_wei@trendmicro.com.cn",
        "time": "Thu Aug 14 08:56:22 2014 +0800"
      },
      "committer": {
        "name": "Roland Dreier",
        "email": "roland@purestorage.com",
        "time": "Thu Aug 14 08:57:54 2014 -0700"
      },
      "message": "IB/srp: Fix return value check in srp_init_module()\n\nIn case of error, the function create_workqueue() returns NULL pointer\nnot ERR_PTR().  The IS_ERR() test in the return value check should be\nreplaced with NULL test.\n\nSigned-off-by: Wei Yongjun \u003cyongjun_wei@trendmicro.com.cn\u003e\nAcked-by: Bart Van Assche \u003cbvanassche@acm.org\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n"
    },
    {
      "commit": "82f05a08e27755835c948b7a45b86fac9e114e1d",
      "tree": "2fa6a102f3611c2aecb48e2845e83e86d560719f",
      "parents": [
        "ae36e95cf81c98b111b84317adeb358aaffa80e2",
        "6ddd855c13bcd324a2d9756134e6b57c659cfde6"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:55:13 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:55:13 2014 -0600"
      },
      "message": "Merge tag \u0027hwmon-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull hwmon fixes from Guenter Roeck:\n \"Several bug fixes in various drivers, plus a minor cleanup in the\n  tmp103 driver\"\n\n* tag \u0027hwmon-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n  hwmon: (tmp103) Remove duplicate test for I2C_FUNC_SMBUS_BYTE_DATA functionality\n  hwmon: (w83793) Fix vrm write operation\n  hwmon: (w83791d) Fix vrm write operation\n  hwmon: (w83627hf) Fix vrm write operation\n  hwmon: (vt1211) Fix vrm write operation\n  hwmon: (pc87360) Fix vrm write operation\n  hwmon: (lm87) Fix vrm write operation\n  hwmon: (asb100) Fix vrm write operation\n  hwmon: (adm1026) Fix vrm write operation\n  hwmon: (adm1025) Fix vrm write operation\n  hwmon: (hih6130) Fix missing hih6130-\u003ewrite_length setting\n  hwmon: (dme1737) Prevent overflow problem when writing large limits\n  hwmon: (emc6w201) Fix temperature limit range\n  hwmon: (ads1015) Fix out-of-bounds array access\n  hwmon: (lm92) Prevent overflow problem when writing large limits\n"
    },
    {
      "commit": "ae36e95cf81c98b111b84317adeb358aaffa80e2",
      "tree": "87ef144d3392cca199448117a737c657b7c4663a",
      "parents": [
        "cc8a44c671fd3a2c792e9e1f59ea1df52697cc8b",
        "663d3f7c2e5e1b018a4c53277ccfde40329d98ca"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:53:39 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:53:39 2014 -0600"
      },
      "message": "Merge tag \u0027devicetree-for-linus\u0027 of git://git.secretlab.ca/git/linux\n\nPull device tree updates from Grant Likely:\n \"The branch contains the following device tree changes the v3.17 merge\n  window:\n\n  Group changes to the device tree.  In preparation for adding device\n  tree overlay support, OF_DYNAMIC is reworked so that a set of device\n  tree changes can be prepared and applied to the tree all at once.\n  OF_RECONFIG notifiers see the most significant change here so that\n  users always get a consistent view of the tree.  Notifiers generation\n  is moved from before a change to after it, and notifiers for a group\n  of changes are emitted after the entire block of changes have been\n  applied\n\n  Automatic console selection from DT.  Console drivers can now use\n  of_console_check() to see if the device node is specified as a console\n  device.  If so then it gets added as a preferred console.  UART\n  devices get this support automatically when uart_add_one_port() is\n  called.\n\n  DT unit tests no longer depend on pre-loaded data in the device tree.\n  Data is loaded dynamically at the start of unit tests, and then\n  unloaded again when the tests have completed.\n\n  Also contains a few bugfixes for reserved regions and early memory\n  setup\"\n\n* tag \u0027devicetree-for-linus\u0027 of git://git.secretlab.ca/git/linux: (21 commits)\n  of: Fixing OF Selftest build error\n  drivers: of: add automated assignment of reserved regions to client devices\n  of: Use proper types for checking memory overflow\n  of: typo fix in __of_prop_dup()\n  Adding selftest testdata dynamically into live tree\n  of: Add todo tasklist for Devicetree\n  of: Transactional DT support.\n  of: Reorder device tree changes and notifiers\n  of: Move dynamic node fixups out of powerpc and into common code\n  of: Make sure attached nodes don\u0027t carry along extra children\n  of: Make devicetree sysfs update functions consistent.\n  of: Create unlocked versions of node and property add/remove functions\n  OF: Utility helper functions for dynamic nodes\n  of: Move CONFIG_OF_DYNAMIC code into a separate file\n  of: rename of_aliases_mutex to just of_mutex\n  of/platform: Fix of_platform_device_destroy iteration of devices\n  of: Migrate of_find_node_by_name() users to for_each_node_by_name()\n  tty: Update hypervisor tty drivers to use core stdout parsing code.\n  arm/versatile: Add the uart as the stdout device.\n  of: Enable console on serial ports specified by /chosen/stdout-path\n  ...\n"
    },
    {
      "commit": "cc8a44c671fd3a2c792e9e1f59ea1df52697cc8b",
      "tree": "9521e782447bb95bb04772aeb274d5387e4cdda1",
      "parents": [
        "4dacb91c7d1ab30aa9aac0d55216fc177d454254",
        "9b936c960f22954bfb89f2fefd8f96916bb42908"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:41:41 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:41:41 2014 -0600"
      },
      "message": "Merge tag \u0027vfio-v3.17-rc1\u0027 of git://github.com/awilliam/linux-vfio\n\nPull VFIO updates from Alex Williamson:\n - enable support for bus reset on device release\n - fixes for EEH support\n\n* tag \u0027vfio-v3.17-rc1\u0027 of git://github.com/awilliam/linux-vfio:\n  drivers/vfio: Enable VFIO if EEH is not supported\n  drivers/vfio: Allow EEH to be built as module\n  drivers/vfio: Fix EEH build error\n  vfio-pci: Attempt bus/slot reset on release\n  vfio-pci: Use mutex around open, release, and remove\n  vfio-pci: Release devices with BusMaster disabled\n"
    },
    {
      "commit": "4dacb91c7d1ab30aa9aac0d55216fc177d454254",
      "tree": "51ccb90ca716b2b5e03c53a4598c20573ecdd3dc",
      "parents": [
        "ba368991f63f020afe4ee9d5b647c5397cf3c7f2",
        "7d951f3ccb0308c95bf76d5eef9886dea35a7013"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:40:44 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:40:44 2014 -0600"
      },
      "message": "Merge tag \u0027stable/for-linus-3.17-b-rc0-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip\n\nPull Xen bugfixes from David Vrabel:\n - fix ARM build\n - fix boot crash with PVH guests\n - improve reliability of resume/migration\n\n* tag \u0027stable/for-linus-3.17-b-rc0-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:\n  x86/xen: use vmap() to map grant table pages in PVH guests\n  x86/xen: resume timer irqs early\n  arm/xen: remove duplicate arch_gnttab_init() function\n"
    },
    {
      "commit": "ba368991f63f020afe4ee9d5b647c5397cf3c7f2",
      "tree": "8d391c8921acab5bd70cd04edaeb3de4c38ed426",
      "parents": [
        "a8e4def604a9affa04fdd4efa0692da1385ffa3f",
        "200612ec33e555a356eebc717630b866ae2b694f"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:17:56 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:17:56 2014 -0600"
      },
      "message": "Merge tag \u0027dm-3.17-changes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm\n\nPull device mapper changes from Mike Snitzer:\n\n - Allow the thin target to paired with any size external origin; also\n   allow thin snapshots to be larger than the external origin.\n\n - Add support for quickly loading a repetitive pattern into the\n   dm-switch target.\n\n - Use per-bio data in the dm-crypt target instead of always using a\n   mempool for each allocation.  Required switching to kmalloc alignment\n   for the bio slab.\n\n - Fix DM core to properly stack the QUEUE_FLAG_NO_SG_MERGE flag\n\n - Fix the dm-cache and dm-thin targets\u0027 export of the minimum_io_size\n   to match the data block size -- this fixes an issue where mkfs.xfs\n   would improperly infer raid striping was in place on the underlying\n   storage.\n\n - Small cleanups in dm-io, dm-mpath and dm-cache\n\n* tag \u0027dm-3.17-changes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:\n  dm table: propagate QUEUE_FLAG_NO_SG_MERGE\n  dm switch: efficiently support repetitive patterns\n  dm switch: factor out switch_region_table_read\n  dm cache: set minimum_io_size to cache\u0027s data block size\n  dm thin: set minimum_io_size to pool\u0027s data block size\n  dm crypt: use per-bio data\n  block: use kmalloc alignment for bio slab\n  dm table: make dm_table_supports_discards static\n  dm cache metadata: use dm-space-map-metadata.h defined size limits\n  dm cache: fail migrations in the do_worker error path\n  dm cache: simplify deferred set reference count increments\n  dm thin: relax external origin size constraints\n  dm thin: switch to an atomic_t for tracking pending new block preparations\n  dm mpath: eliminate pg_ready() wrapper\n  dm io: simplify dec_count and sync_io\n"
    },
    {
      "commit": "a8e4def604a9affa04fdd4efa0692da1385ffa3f",
      "tree": "ab7c02cbfbd72cc4fe60674cd5da7e3d01035f00",
      "parents": [
        "d429a3639ca967ce2f35e3e8d4e70caec7149ded",
        "c83c8737e3edb33f60101d2d7692675d0cb6bdf1"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:15:53 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:15:53 2014 -0600"
      },
      "message": "Merge tag \u0027mmc-v3.17-1\u0027 of git://git.linaro.org/people/ulf.hansson/mmc\n\nPull MMC updates from Ulf Hansson:\n \"Me and Chris Ball decided to try out using my MMC tree as the primary\n  one, to simplify handling of patches.\n\n  This pull does thus contains all the MMC patches for 3.17 rc1, no pull\n  from Chris this time.\n\n  Details:\n\n  MMC core:\n   - forward compatibility for eMMC\n   - fix some blacklisted cards with broken secure discard\n\n  MMC host:\n   - mmci: Add support for Qualcomm variant\n   - mmci: Fix regression for arm_variant\n   - sdhci: Various fixes and cleanups\n   - sdhci: Improve external VDD regulator support\n   - sdhci: Support for DDR50 1.8V mode for BayTrail\n   - sdhci-st: Add driver for ST SDHCI controller\n   - sh-mmcif: DMA fixes\n   - omap_hsmmc: Add support for SDIO interrupts\n   - sdhci-pci: Add support for Intel Quark X1000\n   - dw_mmc: Update the reset sequence\n   - s3cmci: port DMA code to dmaengine API\"\n\n* tag \u0027mmc-v3.17-1\u0027 of git://git.linaro.org/people/ulf.hansson/mmc: (67 commits)\n  mmc: dw_mmc: modify the dt-binding for removing slot-node and supports-highspeed\n  mmc: dw_mmc: Slot quirk \"disable-wp\" is deprecated.\n  mmc: mmci: Reverse IRQ handling for the arm_variant\n  mmc: mmci: Move all CMD irq handling to mmci_cmd_irq()\n  mmc: mmci: Remove redundant check of status for DATA irq\n  mmc: dw_mmc: change to use recommended reset procedure\n  mmc: sdhci-pxav3: Use devm_* managed helpers\n  mmc: tmio: Configure DMA slave bus width\n  mmc: sh_mmcif: Configure DMA slave bus width\n  mmc: sh_mmcif: Fix DMA slave address configuration\n  mmc: sh_mmcif: Document DT bindings\n  mmc: sdhci-pci: remove PCI PM functions in suspend/resume callback\n  mmc: Do not advertise secure discard if it is blacklisted\n  mmc: sdhci-msm: Get COMPILE_TEST support\n  mmc: sdhci-msm: Remove unnecessary header file inclusion\n  mmc: sdhci-msm: Fix the binding example\n  mmc: sdhci: add DDR50 1.8V mode support for BayTrail eMMC Controller\n  mmc: sdhci: Preset value not supported in Baytrail eMMC\n  mmc: MMC_USDHI6ROL0 should depend on HAS_DMA\n  mmc: MMC_SH_MMCIF should depend on HAS_DMA\n  ...\n"
    },
    {
      "commit": "d429a3639ca967ce2f35e3e8d4e70caec7149ded",
      "tree": "cad1e5602551b6a744f63ef062de2c2e21cfe39a",
      "parents": [
        "4a319a490ca59a746b3d36768c0e29ee19832366",
        "99d540018caa920b7a54e2d3048f1dff530b294b"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:10:21 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:10:21 2014 -0600"
      },
      "message": "Merge branch \u0027for-3.17/drivers\u0027 of git://git.kernel.dk/linux-block\n\nPull block driver changes from Jens Axboe:\n \"Nothing out of the ordinary here, this pull request contains:\n\n   - A big round of fixes for bcache from Kent Overstreet, Slava Pestov,\n     and Surbhi Palande.  No new features, just a lot of fixes.\n\n   - The usual round of drbd updates from Andreas Gruenbacher, Lars\n     Ellenberg, and Philipp Reisner.\n\n   - virtio_blk was converted to blk-mq back in 3.13, but now Ming Lei\n     has taken it one step further and added support for actually using\n     more than one queue.\n\n   - Addition of an explicit SG_FLAG_Q_AT_HEAD for block/bsg, to\n     compliment the the default behavior of adding to the tail of the\n     queue.  From Douglas Gilbert\"\n\n* \u0027for-3.17/drivers\u0027 of git://git.kernel.dk/linux-block: (86 commits)\n  bcache: Drop unneeded blk_sync_queue() calls\n  bcache: add mutex lock for bch_is_open\n  bcache: Correct printing of btree_gc_max_duration_ms\n  bcache: try to set b-\u003eparent properly\n  bcache: fix memory corruption in init error path\n  bcache: fix crash with incomplete cache set\n  bcache: Fix more early shutdown bugs\n  bcache: fix use-after-free in btree_gc_coalesce()\n  bcache: Fix an infinite loop in journal replay\n  bcache: fix crash in bcache_btree_node_alloc_fail tracepoint\n  bcache: bcache_write tracepoint was crashing\n  bcache: fix typo in bch_bkey_equal_header\n  bcache: Allocate bounce buffers with GFP_NOWAIT\n  bcache: Make sure to pass GFP_WAIT to mempool_alloc()\n  bcache: fix uninterruptible sleep in writeback thread\n  bcache: wait for buckets when allocating new btree root\n  bcache: fix crash on shutdown in passthrough mode\n  bcache: fix lockdep warnings on shutdown\n  bcache allocator: send discards with correct size\n  bcache: Fix to remove the rcu_sched stalls.\n  ...\n"
    },
    {
      "commit": "4a319a490ca59a746b3d36768c0e29ee19832366",
      "tree": "41c34e68fdf2477aa9ae40d8c3d77cdec93f568b",
      "parents": [
        "f0094b28f3038936c1985be64dbe83f0e950b671",
        "d97a86c170b4e432f76db072a827fe30b4d6f659"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:07:02 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Thu Aug 14 09:07:02 2014 -0600"
      },
      "message": "Merge branch \u0027for-3.17/core\u0027 of git://git.kernel.dk/linux-block\n\nPull block core bits from Jens Axboe:\n \"Small round this time, after the massive blk-mq dump for 3.16.  This\n  pull request contains:\n\n   - Fixes for max_sectors overflow in ioctls from Akinoby Mita.\n\n   - Partition off-by-one bug fix in aix partitions from Dan Carpenter.\n\n   - Various small partition cleanups from Fabian Frederick.\n\n   - Fix for the block integrity code sometimes returning the wrong\n     vector count from Gu Zheng.\n\n   - Cleanup an re-org of the blk-mq queue enter/exit percpu counters\n     from Tejun.  Dependent on the percpu pull for 3.17 (which was in\n     the block tree too), that you have already pulled in.\n\n   - A blkcg oops fix, also from Tejun\"\n\n* \u0027for-3.17/core\u0027 of git://git.kernel.dk/linux-block:\n  partitions: aix.c: off by one bug\n  blkcg: don\u0027t call into policy draining if root_blkg is already gone\n  Revert \"bio: modify __bio_add_page() to accept pages that don\u0027t start a new segment\"\n  bio: modify __bio_add_page() to accept pages that don\u0027t start a new segment\n  block: fix SG_[GS]ET_RESERVED_SIZE ioctl when max_sectors is huge\n  block: fix BLKSECTGET ioctl when max_sectors is greater than USHRT_MAX\n  block/partitions/efi.c: kerneldoc fixing\n  block/partitions/msdos.c: code clean-up\n  block/partitions/amiga.c: replace nolevel printk by pr_err\n  block/partitions/aix.c: replace count*size kzalloc by kcalloc\n  bio-integrity: add \"bip_max_vcnt\" into struct bio_integrity_payload\n  blk-mq: use percpu_ref for mq usage count\n  blk-mq: collapse __blk_mq_drain_queue() into blk_mq_freeze_queue()\n  blk-mq: decouble blk-mq freezing from generic bypassing\n  block, blk-mq: draining can\u0027t be skipped even if bypass_depth was non-zero\n  blk-mq: fix a memory ordering bug in blk_mq_queue_enter()\n"
    },
    {
      "commit": "10cf15e1d1289aa0bf1d26e9f55176b4c7c5c512",
      "tree": "f5e3127632889c170b25ac6583c90a8c3d002d72",
      "parents": [
        "f1d25d37d316b8af202e51a4f82df01e12fe2661"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 22:00:09 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 22:00:09 2014 -0700"
      },
      "message": "sparc: Hook up memfd_create system call.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "f1d25d37d316b8af202e51a4f82df01e12fe2661",
      "tree": "59169536ef35250f2929cbf2bac5e430324ca619",
      "parents": [
        "4afba24e5fc2626a3c604ec990539572f6662ff9"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Tue Aug 12 23:29:09 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 21:17:49 2014 -0700"
      },
      "message": "sparc64: Properly claim resources as each PCI bus is probed.\n\nPerform a pci_claim_resource() on all valid resources discovered\nduring the OF device tree scan.\n\nBased almost entirely upon the PCI OF bus probing code which does\nthe same thing there.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "4afba24e5fc2626a3c604ec990539572f6662ff9",
      "tree": "9747dae7206d7aa396ece4c22397d81ba848eba5",
      "parents": [
        "93a6423bd84d977bd768a001c6c3868e6a20b63a"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Tue Aug 12 23:27:01 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 21:17:49 2014 -0700"
      },
      "message": "sparc64: Skip bogus PCI bridge ranges.\n\nIt seems that when a PCI Express bridge is not in use and has no devices\nbehind it, the ranges property is bogus.  Specifically the size property\nis of the form [0xffffffff:...], and if you add this size to the resource\nstart address the 64-bit calculation will overflow.\n\nJust check specifically for this size value signature and skip them.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "93a6423bd84d977bd768a001c6c3868e6a20b63a",
      "tree": "5134930d02aaa372ded8a4f78bcd1ceae53dcae6",
      "parents": [
        "f0094b28f3038936c1985be64dbe83f0e950b671"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Tue Aug 12 23:22:39 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 21:17:49 2014 -0700"
      },
      "message": "sparc64: Expand PCI bridge probing debug logging.\n\nDump the various aspects of the PCI bridge probed at boot time, most\nimportantly the bridge number ranges, and the ranges property.\n\nThis helps diagnose PCI resource issues and other problems by giving\nofpci_debug\u003d1 on the boot command line.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "db8d457acccdb605b82ae22e48d28a47861b0c86",
      "tree": "90b224b7831a7771951661f64b9aed36e5b15884",
      "parents": [
        "a4688132a1629779e4eb8f95660cbebf650819b4",
        "5e815d84180edcc6bf8e3d9c7d3b2ee0cf79b6ba"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:08:55 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:08:55 2014 -0700"
      },
      "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net\n\nJeff Kirsher says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nIntel Wired LAN Driver Updates 2014-08-12\n\nThis series contains updates to i40e and e1000e.\n\nLucas provides a fix for i40e to resolve a compile issue where a header\nwas missing in the #includes.\n\nWei Yongjun provides a fix for i40e to resolve a sparse warning, where\na non-static function should be static.\n\nJulia Lawall provides a fix for i40e which was found using Coccinelle,\nwhere there was a typo in the name of the type given to sizeof().\n\nRickard Strandqvist provides a fix for i40e to replace the use of\nstrncpy() with strlcpy() to avoid strings that lack null termination.\n\nJean Sacren provides two e1000e fixes, first is a comment fix and second\nremoves an excessive space character in a debug message.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "a4688132a1629779e4eb8f95660cbebf650819b4",
      "tree": "a97ba22dad20761ac3eddcd7cf51463a872c4d60",
      "parents": [
        "68809958710ac80e2fad387edd6383f422498f9e",
        "b1252858213f39700dac1bc3295b6e88f6cce24b"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:07:52 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:07:52 2014 -0700"
      },
      "message": "Merge branch \u0027xen-netback-synchronization\u0027\n\nWei Liu says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nxen-netback: synchronisation between core driver and netback\n\nThe zero-copy netback has far more interactions with core network driver than\nthe old copying backend. One significant thing is that netback now relies on\na callback from core driver to correctly release resources.\n\nHowever correct synchronisation between core driver and netback is missing.\nCurrently netback relies on a loop to wait for core driver to release\nresources. This is proven not enough and erroneous recently, partly due to code\nstructure, partly due to missing synchronisation. Short-live domains like\nOpenMirage unikernels can easily trigger race in backend, rendering backend\nunresponsive.\n\nThis patch series aims to slove this issue by introducing proper\nsynchronisation between core driver and netback.\n\nChagges in v4:\n* avoid using wait queue\n* remove dedicated loop for netif_napi_del\n* remove unnecessary check on callback\n\nChange in v3: improve commit message in patch 1\n\nChange in v2: fix Zoltan\u0027s email address in commit message\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "b1252858213f39700dac1bc3295b6e88f6cce24b",
      "tree": "a97ba22dad20761ac3eddcd7cf51463a872c4d60",
      "parents": [
        "a64bd934528e26e8956112e43a279fba2ee0634e"
      ],
      "author": {
        "name": "Wei Liu",
        "email": "wei.liu2@citrix.com",
        "time": "Tue Aug 12 11:48:08 2014 +0100"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:07:44 2014 -0700"
      },
      "message": "xen-netback: remove loop waiting function\n\nThe original implementation relies on a loop to check if all inflight\npackets are freed. Now we have proper reference counting, there\u0027s no\nneed to use loop anymore.\n\nSigned-off-by: Wei Liu \u003cwei.liu2@citrix.com\u003e\nCc: Ian Campbell \u003cian.campbell@citrix.com\u003e\nCc: Zoltan Kiss \u003czoltan.kiss@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "a64bd934528e26e8956112e43a279fba2ee0634e",
      "tree": "304e4fca43ec1c2907bb0db71f50445ef5fcae32",
      "parents": [
        "ea2c5e134237eadc9924ce821ded678750024549"
      ],
      "author": {
        "name": "Wei Liu",
        "email": "wei.liu2@citrix.com",
        "time": "Tue Aug 12 11:48:07 2014 +0100"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:07:44 2014 -0700"
      },
      "message": "xen-netback: don\u0027t stop dealloc kthread too early\n\nReference count the number of packets in host stack, so that we don\u0027t\nstop the deallocation thread too early. If not, we can end up with\nxenvif_free permanently waiting for deallocation thread to unmap grefs.\n\nReported-by: Thomas Leonard \u003ctalex5@gmail.com\u003e\nSigned-off-by: Wei Liu \u003cwei.liu2@citrix.com\u003e\nCc: Ian Campbell \u003cian.campbell@citrix.com\u003e\nCc: Zoltan Kiss \u003czoltan.kiss@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "ea2c5e134237eadc9924ce821ded678750024549",
      "tree": "f54d00e659f58fbe53ba5683faa1f7c301b9c755",
      "parents": [
        "68809958710ac80e2fad387edd6383f422498f9e"
      ],
      "author": {
        "name": "Wei Liu",
        "email": "wei.liu2@citrix.com",
        "time": "Tue Aug 12 11:48:06 2014 +0100"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:07:44 2014 -0700"
      },
      "message": "xen-netback: move NAPI add/remove calls\n\nOriginally netif_napi_add was in xenvif_init_queue and netif_napi_del\nwas in xenvif_deinit_queue, while kthreads were handled in\nxenvif_connect and xenvif_disconnect. Move netif_napi_add and\nnetif_napi_del to xenvif_connect and xenvif_disconnect so that they\nreside together with kthread operations.\n\nSigned-off-by: Wei Liu \u003cwei.liu2@citrix.com\u003e\nCc: Ian Campbell \u003cian.campbell@citrix.com\u003e\nCc: Zoltan Kiss \u003czoltan.kiss@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "68809958710ac80e2fad387edd6383f422498f9e",
      "tree": "1c738b4350f79c4f65f1a2441a857b88eb8ccb7b",
      "parents": [
        "490cc7d03c21871eef5949ba77a18b0c7ef70ef5",
        "628fa76b09d7b0923c142631fc25b6affbfb868d"
      ],
      "author": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:51 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:51 2014 -0700"
      },
      "message": "Merge branch \u0027xen-netback-debugfs\u0027\n\nWei Liu says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nxen-netback: fix debugfs code\n\nThis small series fixes two problems in xen-netback debugfs code.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "628fa76b09d7b0923c142631fc25b6affbfb868d",
      "tree": "1c738b4350f79c4f65f1a2441a857b88eb8ccb7b",
      "parents": [
        "5c807005fa60deef2db6616d9b7b24fc4c436be9"
      ],
      "author": {
        "name": "Wei Liu",
        "email": "wei.liu2@citrix.com",
        "time": "Tue Aug 12 11:59:30 2014 +0100"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:43 2014 -0700"
      },
      "message": "xen-netback: fix debugfs entry creation\n\nThe original code is bogus. The function gets called in a loop which\nleaks entries created in previous rounds.\n\nSigned-off-by: Wei Liu \u003cwei.liu2@citrix.com\u003e\nCc: Zoltan Kiss \u003czoltan.kiss@citrix.com\u003e\nCc: Ian Campbell \u003cian.campbell@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "5c807005fa60deef2db6616d9b7b24fc4c436be9",
      "tree": "6e1ca61dc320ede63d38b91def87edeaae31a30c",
      "parents": [
        "490cc7d03c21871eef5949ba77a18b0c7ef70ef5"
      ],
      "author": {
        "name": "Wei Liu",
        "email": "wei.liu2@citrix.com",
        "time": "Tue Aug 12 11:59:29 2014 +0100"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:43 2014 -0700"
      },
      "message": "xen-netback: fix debugfs write length check\n\nEnlarge buffer size and check input length properly, so that we don\u0027t\nmisuse -ENOSPC.\n\nNote that command like \"kickXXXX\" is still allowed, that\u0027s one patch for\nanother day if we really want to be very strict on this.\n\nReported-by: SeeChen Ng \u003cseechen81@gmail.com\u003e\nSigned-off-by: Wei Liu \u003cwei.liu2@citrix.com\u003e\nCc: Zoltan Kiss \u003czoltan.kiss@citrix.com\u003e\nCc: Ian Campbell \u003cian.campbell@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "490cc7d03c21871eef5949ba77a18b0c7ef70ef5",
      "tree": "baa7f8953e8cbd89e29515bd13b889e06b7b192a",
      "parents": [
        "712a72213fad36cc9e6ec706b5e020d7eb6e03bc"
      ],
      "author": {
        "name": "Willem de Bruijn",
        "email": "willemb@google.com",
        "time": "Tue Aug 12 15:08:12 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:06 2014 -0700"
      },
      "message": "net-timestamp: fix missing tcp fragmentation cases\n\nBytestream timestamps are correlated with a single byte in the skbuff,\nrecorded in skb_shinfo(skb)-\u003etskey. When fragmenting skbuffs, ensure\nthat the tskey is set for the fragment in which the tskey falls\n(seqno \u003c\u003d tskey \u003c end_seqno).\n\nThe original implementation did not address fragmentation in\ntcp_fragment or tso_fragment. Add code to inspect the sequence numbers\nand move both tskey and the relevant tx_flags if necessary.\n\nReported-by: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nSigned-off-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "712a72213fad36cc9e6ec706b5e020d7eb6e03bc",
      "tree": "7cc2a560a0b95d500b12c172f9e8b32b1d1321d1",
      "parents": [
        "cd094927c68718c8bda432c18e85b6ec3a9ee3d1"
      ],
      "author": {
        "name": "Willem de Bruijn",
        "email": "willemb@google.com",
        "time": "Tue Aug 12 14:53:16 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:06:06 2014 -0700"
      },
      "message": "net-timestamp: fix missing ACK timestamp\n\nACK timestamps are generated in tcp_clean_rtx_queue. The TSO datapath\ncan break out early, causing the timestamp code to be skipped. Move\nthe code up before the break.\n\nReported-by: David S. Miller \u003cdavem@davemloft.net\u003e\n\nAlso fix a boundary condition: tp-\u003esnd_una is the next unacknowledged\nbyte and between tests inclusive (a \u003c\u003d b \u003c\u003d c), so generate a an ACK\ntimestamp if (prior_snd_una \u003c\u003d tskey \u003c\u003d tp-\u003esnd_una - 1).\n\nSigned-off-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "cd094927c68718c8bda432c18e85b6ec3a9ee3d1",
      "tree": "6f420fce3b40b867cbf0738459942ae920dc6891",
      "parents": [
        "efd5029010c5b13c20bd6364fce9b3eae454419a"
      ],
      "author": {
        "name": "Libo Chen",
        "email": "clbchenlibo.chen@huawei.com",
        "time": "Tue Aug 12 13:31:54 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:05:52 2014 -0700"
      },
      "message": "drivers/net/irda/donauboe.c: convert to module_pci_driver\n\nSigned-off-by: Libo Chen \u003clibo.chen@huawei.com\u003e\nCc: Samuel Ortiz \u003csamuel@sortiz.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "efd5029010c5b13c20bd6364fce9b3eae454419a",
      "tree": "7337a4d950da13dc92876f5d4ea3bc523aad9e5d",
      "parents": [
        "8d21797df9ce6537a46ff58e20f359ef05c8faf2"
      ],
      "author": {
        "name": "Maks Naumov",
        "email": "maksqwe1@ukr.net",
        "time": "Tue Aug 12 11:28:41 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:05:52 2014 -0700"
      },
      "message": "irda: Fix rd_frame control field initialization in irlap_send_rd_frame()\n\nSigned-off-by: Maks Naumov \u003cmaksqwe1@ukr.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "8d21797df9ce6537a46ff58e20f359ef05c8faf2",
      "tree": "b084eea8f12704a69191a57cc13f4be3b0dc54de",
      "parents": [
        "7b31b4deda7679120379c005df8d3d647161eff7"
      ],
      "author": {
        "name": "Anish Bhatt",
        "email": "anish@chelsio.com",
        "time": "Tue Aug 12 22:38:26 2014 -0700"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:47 2014 -0700"
      },
      "message": "libcxgbi/cxgb4i : Fix ipv6 build failure caught with randconfig\n\nPrevious guard of IS_ENABLED(CONFIG_IPV6) is not sufficient when cxgbi drivers\nare built into kernel but ipv6 is not.\n\nv2: Use Kconfig to disable compiling cxgbi built into kernel when ipv6 is\ncompiled as a module\n\nFixes: e81fbf6cd652 (\"libcxgbi:cxgb4i Guard ipv6 code with a config check\")\nFixes: fc8d0590d914 (\"libcxgbi: Add ipv6 api to driver\")\nSigned-off-by: Anish Bhatt \u003canish@chelsio.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "7b31b4deda7679120379c005df8d3d647161eff7",
      "tree": "8596d1e8d49180a1e68eac5749e21768626a65e2",
      "parents": [
        "1d311ad2f9fe5d75cf228372174170d0f435e3ef"
      ],
      "author": {
        "name": "Govindarajulu Varadarajan",
        "email": "_govind@gmx.com",
        "time": "Wed Aug 13 13:04:56 2014 +0530"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:47 2014 -0700"
      },
      "message": "tg3: fix return value in tg3_get_stats64\n\nWhen tp-\u003ehw_stats is 0, tg3_get_stats64 should display previously\nrecorded stats. So it returns \u0026tp-\u003enet_stats_prev. But the caller,\ndev_get_stats, ignores the return value.\n\nFix this by assigning tp-\u003enet_stats_prev to stats and returning stats.\n\nSigned-off-by: Govindarajulu Varadarajan \u003c_govind@gmx.com\u003e\nAcked-by: Prashant Sreedharan \u003cprashant@broadcom.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "1d311ad2f9fe5d75cf228372174170d0f435e3ef",
      "tree": "e1ceb11d8f5ea8602b970baac8475735dd88eda9",
      "parents": [
        "adddc32d6fde82156c779997f928865100542e55"
      ],
      "author": {
        "name": "Sowmini Varadhan",
        "email": "sowmini.varadhan@oracle.com",
        "time": "Wed Aug 13 10:29:41 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "sunvnet: Schedule maybe_tx_wakeup() as a tasklet from ldc_rx path\n\nAt the tail of vnet_event(), if we hit the maybe_tx_wakeup()\ncondition, we try to take the netif_tx_lock() in the\nrecv-interrupt-context and can deadlock with dev_watchdog().\nvnet_event() should schedule maybe_tx_wakeup() as a tasklet\nto avoid this deadlock\n\nSigned-off-by: Sowmini Varadhan \u003csowmini.varadhan@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "adddc32d6fde82156c779997f928865100542e55",
      "tree": "b7605e8c600203c43768f48256c4382f30e5c390",
      "parents": [
        "1f6394e382f137f07136182c591ca8a7eeba6a06"
      ],
      "author": {
        "name": "Sowmini Varadhan",
        "email": "sowmini.varadhan@oracle.com",
        "time": "Tue Aug 12 10:33:10 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "sunvnet: Do not spin in an infinite loop when vio_ldc_send() returns EAGAIN\n\nldc_rx -\u003e vnet_rx -\u003e .. -\u003e vnet_walk_rx-\u003evnet_send_ack should not\nspin into an infinite loop waiting  EAGAIN to lift.\n\nThe sender could have sent us a burst, and gone to lunch without\ndoing any more ldc_read()\u0027s. That should not cause the receiver to\nloop infinitely till soft-lockup kicks in.\n\nSimilarly __vnet_tx_trigger should only loop on EAGAIN a finite\nnumber of times. The caller (vnet_start_xmit()) already has code\nto reset the dring state and bail on errors from __vnet_tx_trigger\n\nSigned-off-by: Sowmini Varadhan \u003csowmini.varadhan@oracle.com\u003e\nAcked-by: Raghuram Kothakota \u003craghuram.kothakota@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "1f6394e382f137f07136182c591ca8a7eeba6a06",
      "tree": "4271803a31d0e1a584709ac2c6bcd9035eaadf12",
      "parents": [
        "8356f9d564ed3da4f98e8c44cbd9f397ff9c49a5"
      ],
      "author": {
        "name": "Sowmini Varadhan",
        "email": "sowmini.varadhan@oracle.com",
        "time": "Tue Aug 12 10:27:24 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "sunvnet: Do not ask for an ACK for every dring transmit\n\nNo need to ask for an ack with every vnet_start_xmit()- the single\nACK with DRING_STOPPED is sufficient for the protocol, and we free\nthe sk_buff in vnet_start_xmit itself, so we dont need an ACK back.\n\nSigned-off-by: Sowmini Varadhan \u003csowmini.varadhan@oracle.com\u003e\nAcked-by: Raghuram Kothakota \u003craghuram.kothakota@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "8356f9d564ed3da4f98e8c44cbd9f397ff9c49a5",
      "tree": "83cbb19267443b00009201aad10beb1b669ddcc3",
      "parents": [
        "de713b57947a7a1b9cc605c0296ee14065cd86b6"
      ],
      "author": {
        "name": "chas williams - CONTRACTOR",
        "email": "chas@cmf.nrl.navy.mil",
        "time": "Tue Aug 12 09:00:36 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "lec: Fix bug introduced by b67bfe0d42cac56c512dd5da4b1b347a23f4b70a\n\nb67bfe0d42cac56c512dd5da4b1b347a23f4b70a (hlist: drop the node\nparameter from iterators) dropped the node parameter from\niterators which lec_tbl_walk() was using to iterate the list.\n\nSigned-off-by: Chas Williams \u003cchas@cmf.nrl.navy.mil\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "de713b57947a7a1b9cc605c0296ee14065cd86b6",
      "tree": "d47fa60ae55956e23ff7c1d57c32bdcafd3f65f8",
      "parents": [
        "10545937e866ccdbb7ab583031dbdcc6b14e4eb4"
      ],
      "author": {
        "name": "chas williams - CONTRACTOR",
        "email": "chas@cmf.nrl.navy.mil",
        "time": "Tue Aug 12 08:12:26 2014 -0400"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "atm/svc: Fix blocking in wait loop\n\nOne should not call blocking primitives inside a wait loop, since both\nrequire task_struct::state to sleep, so the inner will destroy the\nouter state.\n\nsigd_enq() will possibly sleep for alloc_skb().  Move sigd_enq() before\nprepare_to_wait() to avoid sleeping while waiting interruptibly.  You do\nnot actually need to call sigd_enq() after the initial prepare_to_wait()\nbecause we test the termination condition before calling schedule().\n\nBased on suggestions from Peter Zijlstra.\n\nSigned-off-by: Chas Williams \u003cchas@cmf.n4rl.navy.mil\u003e\nAcked-by: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "10545937e866ccdbb7ab583031dbdcc6b14e4eb4",
      "tree": "7b9254702b49e25ada44dc95c72688c63ce2a5ce",
      "parents": [
        "3791b3f6fb74d265c93d493d9bbf29c1e769ceae"
      ],
      "author": {
        "name": "Stanislaw Gruszka",
        "email": "sgruszka@redhat.com",
        "time": "Tue Aug 12 10:35:19 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "myri10ge: check for DMA mapping errors\n\nOn IOMMU systems DMA mapping can fail, we need to check for\nthat possibility.\n\nSigned-off-by: Stanislaw Gruszka \u003csgruszka@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "3791b3f6fb74d265c93d493d9bbf29c1e769ceae",
      "tree": "eb1eac4b5160d84014501558c0deabb92fd086c6",
      "parents": [
        "f0094b28f3038936c1985be64dbe83f0e950b671"
      ],
      "author": {
        "name": "Christoph Jaeger",
        "email": "cj@linux.com",
        "time": "Tue Aug 12 09:27:57 2014 +0200"
      },
      "committer": {
        "name": "David S. Miller",
        "email": "davem@davemloft.net",
        "time": "Wed Aug 13 20:04:46 2014 -0700"
      },
      "message": "openvswitch: Fix memory leak in ovs_vport_alloc() error path\n\novs_vport_alloc() bails out without freeing the memory \u0027vport\u0027 points to.\n\nPicked up by Coverity - CID 1230503.\n\nFixes: 5cd667b0a4 (\"openvswitch: Allow each vport to have an array of \u0027port_id\u0027s.\")\nSigned-off-by: Christoph Jaeger \u003ccj@linux.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n"
    },
    {
      "commit": "f0094b28f3038936c1985be64dbe83f0e950b671",
      "tree": "c3a61d5583cc2b39f65ef3b01e6660696b359656",
      "parents": [
        "13b102bf48d39d84d87bee31544f4588502649b6",
        "61dac43ee6bedd8507294c7af8959157b686c221"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:27:40 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:27:40 2014 -0600"
      },
      "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nPull networking fixes from David Miller:\n \"Several networking final fixes and tidies for the merge window:\n\n   1) Changes during the merge window unintentionally took away the\n      ability to build bluetooth modular, fix from Geert Uytterhoeven.\n\n   2) Several phy_node reference count bug fixes from Uwe Kleine-König.\n\n   3) Fix ucc_geth build failures, also from Uwe Kleine-König.\n\n   4) Fix klog false positivies when netlink messages go to network\n      taps, by properly resetting the network header.  Fix from Daniel\n      Borkmann.\n\n   5) Sizing estimate of VF netlink messages is too small, from Jiri\n      Benc.\n\n   6) New APM X-Gene SoC ethernet driver, from Iyappan Subramanian.\n\n   7) VLAN untagging is erroneously dependent upon whether the VLAN\n      module is loaded or not, but there are generic dependencies that\n      matter wrt what can be expected as the SKB enters the stack.\n      Make the basic untagging generic code, and do it unconditionally.\n      From Vlad Yasevich.\n\n   8) xen-netfront only has so many slots in it\u0027s transmit queue so\n      linearize packets that have too many frags.  From Zoltan Kiss.\n\n   9) Fix suspend/resume PHY handling in bcmgenet driver, from Florian\n      Fainelli\"\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (55 commits)\n  net: bcmgenet: correctly resume adapter from Wake-on-LAN\n  net: bcmgenet: update UMAC_CMD only when link is detected\n  net: bcmgenet: correctly suspend and resume PHY device\n  net: bcmgenet: request and enable main clock earlier\n  net: ethernet: myricom: myri10ge: myri10ge.c: Cleaning up missing null-terminate after strncpy call\n  xen-netfront: Fix handling packets on compound pages with skb_linearize\n  net: fec: Support phys probed from devicetree and fixed-link\n  smsc: replace WARN_ON() with WARN_ON_SMP()\n  xen-netback: Don\u0027t deschedule NAPI when carrier off\n  net: ethernet: qlogic: qlcnic: Remove duplicate object file from Makefile\n  wan: wanxl: Remove typedefs from struct names\n  m68k/atari: EtherNEC - ethernet support (ne)\n  net: ethernet: ti: cpmac.c: Cleaning up missing null-terminate after strncpy call\n  hdlc: Remove typedefs from struct names\n  airo_cs: Remove typedef local_info_t\n  atmel: Remove typedef atmel_priv_ioctl\n  com20020_cs: Remove typedef com20020_dev_t\n  ethernet: amd: Remove typedef local_info_t\n  net: Always untag vlan-tagged traffic on input.\n  drivers: net: Add APM X-Gene SoC ethernet driver support.\n  ...\n"
    },
    {
      "commit": "13b102bf48d39d84d87bee31544f4588502649b6",
      "tree": "c51846d0ca82ed2b7c01f52925c1cd56e1802ce7",
      "parents": [
        "81c02a21b293faecab725f691d709111831e1944",
        "8bccf5b313180faefce38e0d1140f76e0f327d28"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:26:50 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:26:50 2014 -0600"
      },
      "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc\n\nPull Sparc fixes from David Miller:\n \"Sparc bug fixes, one of which was preventing successful SMP boots with\n  mainline\"\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:\n  sparc64: Fix pcr_ops initialization and usage bugs.\n  sparc64: Do not disable interrupts in nmi_cpu_busy()\n  sparc: Hook up seccomp and getrandom system calls.\n  sparc: fix decimal printf format specifiers prefixed with 0x\n"
    },
    {
      "commit": "81c02a21b293faecab725f691d709111831e1944",
      "tree": "146d41d8ff3ad03dca611b8fa8ab03c46c5d8978",
      "parents": [
        "d27c0d90184a13e9e9f28c38e84f889a259f6b5f",
        "3eec595235c17a74094daa1e02d1b0af2e9a7125"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:23:32 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:23:32 2014 -0600"
      },
      "message": "Merge branch \u0027x86-apic-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86/apic updates from Thomas Gleixner:\n \"This is a major overhaul to the x86 apic subsystem consisting of the\n  following parts:\n\n   - Remove obsolete APIC driver abstractions (David Rientjes)\n\n   - Use the irqdomain facilities to dynamically allocate IRQs for\n     IOAPICs.  This is a prerequisite to enable IOAPIC hotplug support,\n     and it also frees up wasted vectors (Jiang Liu)\n\n   - Misc fixlets.\n\n  Despite the hickup in Ingos previous pull request - caused by the\n  missing fixup for the suspend/resume issue reported by Borislav - I\n  strongly recommend that this update finds its way into 3.17.  Some\n  history for you:\n\n  This is preparatory work for physical IOAPIC hotplug.  The first\n  attempt to support this was done by Yinghai and I shot it down because\n  it just added another layer of obscurity and complexity to the already\n  existing mess without tackling the underlying shortcomings of the\n  current implementation.\n\n  After quite some on- and offlist discussions, I requested that the\n  design of this functionality must use generic infrastructure, i.e.\n  irq domains, which provide all the mechanisms to dynamically map linux\n  interrupt numbers to physical interrupts.\n\n  Jiang picked up the idea and did a great job of consolidating the\n  existing interfaces to manage the x86 (IOAPIC) interrupt system by\n  utilizing irq domains.\n\n  The testing in tip, Linux-next and inside of Intel on various machines\n  did not unearth any oddities until Borislav exposed it to one of his\n  oddball machines.  The issue was resolved quickly, but unfortunately\n  the fix fell through the cracks and did not hit the tip tree before\n  Ingo sent the pull request.  Not entirely Ingos fault, I also assumed\n  that the fix was already merged when Ingo asked me whether he could\n  send it.\n\n  Nevertheless this work has a proper design, has undergone several\n  rounds of review and the final fallout after applying it to tip and\n  integrating it into Linux-next has been more than moderate.  It\u0027s the\n  ground work not only for IOAPIC hotplug, it will also allow us to move\n  the lowlevel vector allocation into the irqdomain hierarchy, which\n  will benefit other architectures as well.  Patches are posted already,\n  but they are on hold for two weeks, see below.\n\n  I really appreciate the competence and responsiveness Jiang has shown\n  in course of this endavour.  So I\u0027m sure that any fallout of this will\n  be addressed in a timely manner.\n\n  FYI, I\u0027m vanishing for 2 weeks into my annual kids summer camp kitchen\n  duty^Wvacation, while you folks are drooling at KS/LinuxCon :) But HPA\n  will have a look at the hopefully zero fallout until I\u0027m back\"\n\n* \u0027x86-apic-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (53 commits)\n  x86, irq, PCI: Keep IRQ assignment for PCI devices during suspend/hibernation\n  x86/apic/vsmp: Make is_vsmp_box() static\n  x86, apic: Remove enable_apic_mode callback\n  x86, apic: Remove setup_portio_remap callback\n  x86, apic: Remove multi_timer_check callback\n  x86, apic: Replace noop_check_apicid_used\n  x86, apic: Remove check_apicid_present callback\n  x86, apic: Remove mps_oem_check callback\n  x86, apic: Remove smp_callin_clear_local_apic callback\n  x86, apic: Replace trampoline physical addresses with defaults\n  x86, apic: Remove x86_32_numa_cpu_node callback\n  x86: intel-mid: Use the new io_apic interfaces\n  x86, vsmp: Remove is_vsmp_box() from apic_is_clustered_box()\n  x86, irq: Clean up irqdomain transition code\n  x86, irq, devicetree: Release IOAPIC pin when PCI device is disabled\n  x86, irq, SFI: Release IOAPIC pin when PCI device is disabled\n  x86, irq, mpparse: Release IOAPIC pin when PCI device is disabled\n  x86, irq, ACPI: Release IOAPIC pin when PCI device is disabled\n  x86, irq: Introduce helper functions to release IOAPIC pin\n  x86, irq: Simplify the way to handle ISA IRQ\n  ...\n"
    },
    {
      "commit": "d27c0d90184a13e9e9f28c38e84f889a259f6b5f",
      "tree": "a933c7b9c9ac96af1787fa11317842557358bf86",
      "parents": [
        "7453f33b2e07fc2835e24cda0893de83c78d8d76",
        "9e13bcf7e0981f1db0c8c8255ac17d5f898903e9"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:21:35 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:21:35 2014 -0600"
      },
      "message": "Merge branch \u0027x86-efi-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86/efix fixes from Peter Anvin:\n \"Two EFI-related Kconfig changes, which happen to touch immediately\n  adjacent lines in Kconfig and thus collapse to a single patch\"\n\n* \u0027x86-efi-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  x86/efi: Enforce CONFIG_RELOCATABLE for EFI boot stub\n  x86/efi: Fix 3DNow optimization build failure in EFI stub\n"
    },
    {
      "commit": "7453f33b2e07fc2835e24cda0893de83c78d8d76",
      "tree": "02d82193515fcc94cf39e284fd325c4491913331",
      "parents": [
        "fd1cf90580289f83f9c972bb367a74d846d281c4",
        "d0f2dd186133a0241a2ccefb188a0e49e8187859"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:20:04 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:20:04 2014 -0600"
      },
      "message": "Merge branch \u0027x86-xsave-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86/xsave changes from Peter Anvin:\n \"This is a patchset to support the XSAVES instruction required to\n  support context switch of supervisor-only features in upcoming\n  silicon.\n\n  This patchset missed the 3.16 merge window, which is why it is based\n  on 3.15-rc7\"\n\n* \u0027x86-xsave-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  x86, xsave: Add forgotten inline annotation\n  x86/xsaves: Clean up code in xstate offsets computation in xsave area\n  x86/xsave: Make it clear that the XSAVE macros use (%edi)/(%rdi)\n  Define kernel API to get address of each state in xsave area\n  x86/xsaves: Enable xsaves/xrstors\n  x86/xsaves: Call booting time xsaves and xrstors in setup_init_fpu_buf\n  x86/xsaves: Save xstate to task\u0027s xsave area in __save_fpu during booting time\n  x86/xsaves: Add xsaves and xrstors support for booting time\n  x86/xsaves: Clear reserved bits in xsave header\n  x86/xsaves: Use xsave/xrstor for saving and restoring user space context\n  x86/xsaves: Use xsaves/xrstors for context switch\n  x86/xsaves: Use xsaves/xrstors to save and restore xsave area\n  x86/xsaves: Define a macro for handling xsave/xrstor instruction fault\n  x86/xsaves: Define macros for xsave instructions\n  x86/xsaves: Change compacted format xsave area header\n  x86/alternative: Add alternative_input_2 to support alternative with two features and input\n  x86/xsaves: Add a kernel parameter noxsaves to disable xsaves/xrstors\n"
    },
    {
      "commit": "fd1cf90580289f83f9c972bb367a74d846d281c4",
      "tree": "79e067defccd0c9ef4610f2db352762da5eef216",
      "parents": [
        "06b8ab55289345ab191bf4bf0e4acc6d4bdf293d",
        "f93125ae17d1d259b5aba84016db260903822194"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:18:09 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:18:09 2014 -0600"
      },
      "message": "Merge tag \u0027metag-for-v3.17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jhogan/metag\n\nPull metag architecture updates from James Hogan:\n \"Just a couple of minor static analysis fixes, removal of a NULL check\n  that should never happen, and fix an error check where an unsigned\n  value was being checked to see if it was negative\"\n\n* tag \u0027metag-for-v3.17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jhogan/metag:\n  metag: cachepart: Fix failure check\n  metag: hugetlbpage: Remove null pointer checks that could never happen\n"
    },
    {
      "commit": "06b8ab55289345ab191bf4bf0e4acc6d4bdf293d",
      "tree": "9af9215097e26c026f30a58c6ca3092ec15d1e1e",
      "parents": [
        "dc1cc85133120e49c223f36aa77d398b8abac727",
        "71a6ec8ac587418ceb6b420def1ca44b334c1ff7"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:13:19 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 18:13:19 2014 -0600"
      },
      "message": "Merge tag \u0027nfs-for-3.17-1\u0027 of git://git.linux-nfs.org/projects/trondmy/linux-nfs\n\nPull NFS client updates from Trond Myklebust:\n \"Highlights include:\n\n   - stable fix for a bug in nfs3_list_one_acl()\n   - speed up NFS path walks by supporting LOOKUP_RCU\n   - more read/write code cleanups\n   - pNFS fixes for layout return on close\n   - fixes for the RCU handling in the rpcsec_gss code\n   - more NFS/RDMA fixes\"\n\n* tag \u0027nfs-for-3.17-1\u0027 of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (79 commits)\n  nfs: reject changes to resvport and sharecache during remount\n  NFS: Avoid infinite loop when RELEASE_LOCKOWNER getting expired error\n  SUNRPC: remove all refcounting of groupinfo from rpcauth_lookupcred\n  NFS: fix two problems in lookup_revalidate in RCU-walk\n  NFS: allow lockless access to access_cache\n  NFS: teach nfs_lookup_verify_inode to handle LOOKUP_RCU\n  NFS: teach nfs_neg_need_reval to understand LOOKUP_RCU\n  NFS: support RCU_WALK in nfs_permission()\n  sunrpc/auth: allow lockless (rcu) lookup of credential cache.\n  NFS: prepare for RCU-walk support but pushing tests later in code.\n  NFS: nfs4_lookup_revalidate: only evaluate parent if it will be used.\n  NFS: add checks for returned value of try_module_get()\n  nfs: clear_request_commit while holding i_lock\n  pnfs: add pnfs_put_lseg_async\n  pnfs: find swapped pages on pnfs commit lists too\n  nfs: fix comment and add warn_on for PG_INODE_REF\n  nfs: check wait_on_bit_lock err in page_group_lock\n  sunrpc: remove \"ec\" argument from encrypt_v2 operation\n  sunrpc: clean up sparse endianness warnings in gss_krb5_wrap.c\n  sunrpc: clean up sparse endianness warnings in gss_krb5_seal.c\n  ...\n"
    },
    {
      "commit": "dc1cc85133120e49c223f36aa77d398b8abac727",
      "tree": "3c9a823d83144867d13a6ba7b1b0e09fd3ccace8",
      "parents": [
        "cec997093bbff881c3da49084dfba4f76361e96a",
        "645f9857213476407d8ed1b59619fdff7128d3e6"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:49:53 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:49:53 2014 -0600"
      },
      "message": "Merge tag \u0027xfs-for-linus-3.17-rc1\u0027 of git://oss.sgi.com/xfs/xfs\n\nPull xfs update from Dave Chinner:\n \"This update contains:\n   - conversion of the XFS core to pass negative error numbers\n   - restructing of core XFS code that is shared with userspace to\n     fs/xfs/libxfs\n   - introduction of sysfs interface for XFS\n   - bulkstat refactoring\n   - demand driven speculative preallocation removal\n   - XFS now always requires 64 bit sectors to be configured\n   - metadata verifier changes to ensure CRCs are calculated during log\n     recovery\n   - various minor code cleanups\n   - miscellaneous bug fixes\n\n  The diffstat is kind of noisy because of the restructuring of the code\n  to make kernel/userspace code sharing simpler, along with the XFS wide\n  change to use the standard negative error return convention (at last!)\"\n\n* tag \u0027xfs-for-linus-3.17-rc1\u0027 of git://oss.sgi.com/xfs/xfs: (45 commits)\n  xfs: fix coccinelle warnings\n  xfs: flush both inodes in xfs_swap_extents\n  xfs: fix swapext ilock deadlock\n  xfs: kill xfs_vnode.h\n  xfs: kill VN_MAPPED\n  xfs: kill VN_CACHED\n  xfs: kill VN_DIRTY()\n  xfs: dquot recovery needs verifiers\n  xfs: quotacheck leaves dquot buffers without verifiers\n  xfs: ensure verifiers are attached to recovered buffers\n  xfs: catch buffers written without verifiers attached\n  xfs: avoid false quotacheck after unclean shutdown\n  xfs: fix rounding error of fiemap length parameter\n  xfs: introduce xfs_bulkstat_ag_ichunk\n  xfs: require 64-bit sector_t\n  xfs: fix uflags detection at xfs_fs_rm_xquota\n  xfs: remove XFS_IS_OQUOTA_ON macros\n  xfs: tidy up xfs_set_inode32\n  xfs: allow inode allocations in post-growfs disk space\n  xfs: mark xfs_qm_quotacheck as static\n  ...\n"
    },
    {
      "commit": "cec997093bbff881c3da49084dfba4f76361e96a",
      "tree": "7c84f8c30ceef7209a18d7cd216a3c16536008c5",
      "parents": [
        "8d2d441ac4af223eae466c3c31ff737cc31a1411",
        "01777836c87081e4f68c4a43c9abe6114805f91e"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:45:40 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:45:40 2014 -0600"
      },
      "message": "Merge branch \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs\n\nPull quota, reiserfs, UDF updates from Jan Kara:\n \"Scalability improvements for quota, a few reiserfs fixes, and couple\n  of misc cleanups (udf, ext2)\"\n\n* \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:\n  reiserfs: Fix use after free in journal teardown\n  reiserfs: fix corruption introduced by balance_leaf refactor\n  udf: avoid redundant memcpy when writing data in ICB\n  fs/udf: re-use hex_asc_upper_{hi,lo} macros\n  fs/quota: kernel-doc warning fixes\n  udf: use linux/uaccess.h\n  fs/ext2/super.c: Drop memory allocation cast\n  quota: remove dqptr_sem\n  quota: simplify remove_inode_dquot_ref()\n  quota: avoid unnecessary dqget()/dqput() calls\n  quota: protect Q_GETFMT by dqonoff_mutex\n"
    },
    {
      "commit": "8d2d441ac4af223eae466c3c31ff737cc31a1411",
      "tree": "d14b0f72e80f94c1575c281bd21d43a86de0a92d",
      "parents": [
        "89838b80bbbf9774cf010905851db7913c9331f0",
        "5f740d7e1531099b888410e6bab13f68da9b1a4d"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:43:29 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:43:29 2014 -0600"
      },
      "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client\n\nPull Ceph updates from Sage Weil:\n \"There is a lot of refactoring and hardening of the libceph and rbd\n  code here from Ilya that fix various smaller bugs, and a few more\n  important fixes with clone overlap.  The main fix is a critical change\n  to the request_fn handling to not sleep that was exposed by the recent\n  mutex changes (which will also go to the 3.16 stable series).\n\n  Yan Zheng has several fixes in here for CephFS fixing ACL handling,\n  time stamps, and request resends when the MDS restarts.\n\n  Finally, there are a few cleanups from Himangi Saraogi based on\n  Coccinelle\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (39 commits)\n  libceph: set last_piece in ceph_msg_data_pages_cursor_init() correctly\n  rbd: remove extra newlines from rbd_warn() messages\n  rbd: allocate img_request with GFP_NOIO instead GFP_ATOMIC\n  rbd: rework rbd_request_fn()\n  ceph: fix kick_requests()\n  ceph: fix append mode write\n  ceph: fix sizeof(struct tYpO *) typo\n  ceph: remove redundant memset(0)\n  rbd: take snap_id into account when reading in parent info\n  rbd: do not read in parent info before snap context\n  rbd: update mapping size only on refresh\n  rbd: harden rbd_dev_refresh() and callers a bit\n  rbd: split rbd_dev_spec_update() into two functions\n  rbd: remove unnecessary asserts in rbd_dev_image_probe()\n  rbd: introduce rbd_dev_header_info()\n  rbd: show the entire chain of parent images\n  ceph: replace comma with a semicolon\n  rbd: use rbd_segment_name_free() instead of kfree()\n  ceph: check zero length in ceph_sync_read()\n  ceph: reset r_resend_mds after receiving -ESTALE\n  ...\n"
    },
    {
      "commit": "89838b80bbbf9774cf010905851db7913c9331f0",
      "tree": "4d47f408fc90c2f575f5741093fdf34dc9f74b1e",
      "parents": [
        "f6f993328b2abcab86a3c99d7bd9f2066ab03d36",
        "25601a3c9737fed554169759582c690b98ead5d4"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:42:11 2014 -0600"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Aug 13 17:42:11 2014 -0600"
      },
      "message": "Merge tag \u0027upstream-3.17-rc1\u0027 of git://git.infradead.org/linux-ubifs\n\nPull UBI/UBIFS changes from Artem Bityutskiy:\n \"No significant changes, mostly small fixes here and there.  The more\n  important fixes are:\n\n   - UBI deleted list items while iterating the list with\n     \u0027list_for_each_entry\u0027\n   - The UBI block driver did not work properly with very large UBI\n     volumes\"\n\n* tag \u0027upstream-3.17-rc1\u0027 of git://git.infradead.org/linux-ubifs: (21 commits)\n  UBIFS: Add log overlap assertions\n  Revert \"UBIFS: add a log overlap assertion\"\n  UBI: bugfix in ubi_wl_flush()\n  UBI: block: Avoid disk size integer overflow\n  UBI: block: Set disk_capacity out of the mutex\n  UBI: block: Make ubiblock_resize return something\n  UBIFS: add a log overlap assertion\n  UBIFS: remove unnecessary check\n  UBIFS: remove mst_mutex\n  UBIFS: kernel-doc warning fix\n  UBI: init_volumes: Ignore volumes with no LEBs\n  UBIFS: replace seq_printf by seq_puts\n  UBIFS: replace count*size kzalloc by kcalloc\n  UBIFS: kernel-doc warning fix\n  UBIFS: fix error path in create_default_filesystem()\n  UBIFS: fix spelling of \"scanned\"\n  UBIFS: fix some comments\n  UBIFS: remove useless @ecc in struct ubifs_scan_leb\n  UBIFS: remove useless statements\n  UBIFS: Add missing break statements in dbg_chk_pnode()\n  ...\n"
    },
    {
      "commit": "9e813308a5c18c58f9ccae1ec72ed4e14eaf9025",
      "tree": "3e72135c0e403930b809112552e059deaa7e5c11",
      "parents": [
        "85c1fafd7262e68ad821ee1808686b1392b1167d"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:32:04 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:42 2014 +1000"
      },
      "message": "powerpc/thp: Add tracepoints to track hugepage invalidate\n\nAdd tracepoint to track hugepage invalidate. This help us\nin debugging difficult to track bugs.\n\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "85c1fafd7262e68ad821ee1808686b1392b1167d",
      "tree": "9b454eb7a172d0b583f7132527fd3d5098f2defa",
      "parents": [
        "7e467245bf5226db34c4b12d3cbacfa2f7a15a8b"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:32:03 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:41 2014 +1000"
      },
      "message": "powerpc/mm: Use read barrier when creating real_pte\n\nOn ppc64 we support 4K hash pte with 64K page size. That requires\nus to track the hash pte slot information on a per 4k basis. We do that\nby storing the slot details in the second half of pte page. The pte bit\n_PAGE_COMBO is used to indicate whether the second half need to be\nlooked while building real_pte. We need to use read memory barrier while\ndoing that so that load of hidx is not reordered w.r.t _PAGE_COMBO\ncheck. On the store side we already do a lwsync in __hash_page_4K\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "7e467245bf5226db34c4b12d3cbacfa2f7a15a8b",
      "tree": "943f6aa4fe6c82e64372cbac2f5e2f94dc015419",
      "parents": [
        "969b7b208f7408712a3526856e4ae60ad13f6928"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:32:02 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:41 2014 +1000"
      },
      "message": "powerpc/thp: Use ACCESS_ONCE when loading pmdp\n\nWe would get wrong results in compiler recomputed old_pmd. Avoid\nthat by using ACCESS_ONCE\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "969b7b208f7408712a3526856e4ae60ad13f6928",
      "tree": "2c1230292444bc1e01c926349030261b1b9e496d",
      "parents": [
        "fc0479557572375100ef16c71170b29a98e0d69a"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:32:01 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:40 2014 +1000"
      },
      "message": "powerpc/thp: Invalidate with vpn in loop\n\nAs per ISA, for 4k base page size we compare 14..65 bits of VA specified\nwith the entry_VA in tlb. That implies we need to make sure we do a\ntlbie with all the possible 4k va we used to access the 16MB hugepage.\nWith 64k base page size we compare 14..57 bits of VA. Hence we cannot\nignore the lower 24 bits of va while tlbie .We also cannot tlb\ninvalidate a 16MB entry with just one tlbie instruction because\nwe don\u0027t track which va was used to instantiate the tlb entry.\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "fc0479557572375100ef16c71170b29a98e0d69a",
      "tree": "89c0b9da580f2b47bcc8663deb02323052f5025b",
      "parents": [
        "629149fae478f0ac6bf705a535708b192e9c6b59"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:32:00 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:39 2014 +1000"
      },
      "message": "powerpc/thp: Handle combo pages in invalidate\n\nIf we changed base page size of the segment, either via sub_page_protect\nor via remap_4k_pfn, we do a demote_segment which doesn\u0027t flush the hash\ntable entries. We do a lazy hash page table flush for all mapped pages\nin the demoted segment. This happens when we handle hash page fault for\nthese pages.\n\nWe use _PAGE_COMBO bit along with _PAGE_HASHPTE to indicate whether a\npte is backed by 4K hash pte. If we find _PAGE_COMBO not set on the pte,\nthat implies that we could possibly have older 64K hash pte entries in\nthe hash page table and we need to invalidate those entries.\n\nUse _PAGE_COMBO to determine the page size with which we should\ninvalidate the hash table entries on unmap.\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "629149fae478f0ac6bf705a535708b192e9c6b59",
      "tree": "2ffdc967eac8e67df163da7c53c15799697455e1",
      "parents": [
        "fa1f8ae80f8bb996594167ff4750a0b0a5a5bb5d"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:31:59 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:39 2014 +1000"
      },
      "message": "powerpc/thp: Invalidate old 64K based hash page mapping before insert of 4k pte\n\nIf we changed base page size of the segment, either via sub_page_protect\nor via remap_4k_pfn, we do a demote_segment which doesn\u0027t flush the hash\ntable entries. We do a lazy hash page table flush for all mapped pages\nin the demoted segment. This happens when we handle hash page fault\nfor these pages.\n\nWe use _PAGE_COMBO bit along with _PAGE_HASHPTE to indicate whether a\npte is backed by 4K hash pte. If we find _PAGE_COMBO not set on the pte,\nthat implies that we could possibly have older 64K hash pte entries in\nthe hash page table and we need to invalidate those entries.\n\nHandle this correctly for 16M pages\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "fa1f8ae80f8bb996594167ff4750a0b0a5a5bb5d",
      "tree": "f2ca5e2cde1b9d9d9fbe03316231db8eda125de2",
      "parents": [
        "b0aa44a3dfae3d8f45bd1264349aa87f87b7774f"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:31:58 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:38 2014 +1000"
      },
      "message": "powerpc/thp: Don\u0027t recompute vsid and ssize in loop on invalidate\n\nThe segment identifier and segment size will remain the same in\nthe loop, So we can compute it outside. We also change the\nhugepage_invalidate interface so that we can use it the later patch\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "b0aa44a3dfae3d8f45bd1264349aa87f87b7774f",
      "tree": "92b61ca0ea8b8cbd4ec4c9934aafa39eab29fd4d",
      "parents": [
        "2fabf084b6ad6337675d700b159a6091023544f2"
      ],
      "author": {
        "name": "Aneesh Kumar K.V",
        "email": "aneesh.kumar@linux.vnet.ibm.com",
        "time": "Wed Aug 13 12:31:57 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 18:20:37 2014 +1000"
      },
      "message": "powerpc/thp: Add write barrier after updating the valid bit\n\nWith hugepages, we store the hpte valid information in the pte page\nwhose address is stored in the second half of the PMD. Use a\nwrite barrier to make sure clearing pmd busy bit and updating\nhpte valid info are ordered properly.\n\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "2fabf084b6ad6337675d700b159a6091023544f2",
      "tree": "bd3258f45d01c46af601b3a0b21e304a73a3bbb3",
      "parents": [
        "d6589722846a57a4ddf7af595a7f854ff5180950"
      ],
      "author": {
        "name": "Nishanth Aravamudan",
        "email": "nacc@linux.vnet.ibm.com",
        "time": "Thu Jul 17 16:15:12 2014 -0700"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:14:05 2014 +1000"
      },
      "message": "powerpc: reorder per-cpu NUMA information\u0027s initialization\n\nThere is an issue currently where NUMA information is used on powerpc\n(and possibly ia64) before it has been read from the device-tree, which\nleads to large slab consumption with CONFIG_SLUB and memoryless nodes.\n\nNUMA powerpc non-boot CPU\u0027s cpu_to_node/cpu_to_mem is only accurate\nafter start_secondary(), similar to ia64, which is invoked via\nsmp_init().\n\nCommit 6ee0578b4daae (\"workqueue: mark init_workqueues() as\nearly_initcall()\") made init_workqueues() be invoked via\ndo_pre_smp_initcalls(), which is obviously before the secondary\nprocessors are online.\n\nAdditionally, the following commits changed init_workqueues() to use\ncpu_to_node to determine the node to use for kthread_create_on_node:\n\nbce903809ab3f (\"workqueue: add wq_numa_tbl_len and\nwq_numa_possible_cpumask[]\")\nf3f90ad469342 (\"workqueue: determine NUMA node of workers accourding to\nthe allowed cpumask\")\n\nTherefore, when init_workqueues() runs, it sees all CPUs as being on\nNode 0. On LPARs or KVM guests where Node 0 is memoryless, this leads to\na high number of slab deactivations\n(http://www.spinics.net/lists/linux-mm/msg67489.html).\n\nFix this by initializing the powerpc-specific CPU\u003c-\u003enode/local memory\nnode mapping as early as possible, which on powerpc is\ndo_init_bootmem(). Currently that function initializes the mapping for\nthe boot CPU, but we extend it to setup the mapping for all possible\nCPUs. Then, in smp_prepare_cpus(), we can correspondingly set the\nper-cpu values for all possible CPUs. That ensures that before the\nearly_initcalls run (and really as early as possible), the per-cpu NUMA\nmapping is accurate.\n\nWhile testing memoryless nodes on PowerKVM guests with a fix to the\nworkqueue logic to use cpu_to_mem() instead of cpu_to_node(), with a\nguest topology of:\n\navailable: 2 nodes (0-1)\nnode 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49\nnode 0 size: 0 MB\nnode 0 free: 0 MB\nnode 1 cpus: 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99\nnode 1 size: 16336 MB\nnode 1 free: 15329 MB\nnode distances:\nnode   0   1\n  0:  10  40\n  1:  40  10\n\nthe slab consumption decreases from\n\nSlab:             932416 kB\nSUnreclaim:       902336 kB\n\nto\n\nSlab:             395264 kB\nSUnreclaim:       359424 kB\n\nAnd we a corresponding increase in the slab efficiency from\n\nslab                                   mem     objs    slabs\n                                      used   active   active\n------------------------------------------------------------\nkmalloc-16384                       337 MB   11.28%  100.00%\ntask_struct                         288 MB    9.93%  100.00%\n\nto\n\nslab                                   mem     objs    slabs\n                                      used   active   active\n------------------------------------------------------------\nkmalloc-16384                        37 MB  100.00%  100.00%\ntask_struct                          31 MB  100.00%  100.00%\n\nPowerpc didn\u0027t support memoryless nodes until recently (64bb80d87f01\n\"powerpc/numa: Enable CONFIG_HAVE_MEMORYLESS_NODES\" and 8c272261194d\n\"powerpc/numa: Enable USE_PERCPU_NUMA_NODE_ID\"). Those commits also\nhelped improve memory consumption with these kind of environments.\n\nSigned-off-by: Nishanth Aravamudan \u003cnacc@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "d6589722846a57a4ddf7af595a7f854ff5180950",
      "tree": "05db4975ccd1af28b3c3e63638112a41594be60f",
      "parents": [
        "587870e8650a0571e895cc879cd895c78c6391bf"
      ],
      "author": {
        "name": "Himangi Saraogi",
        "email": "himangi774@gmail.com",
        "time": "Tue Jul 22 23:40:19 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:14:04 2014 +1000"
      },
      "message": "powerpc/perf/hv-24x7: Use kmem_cache_free\n\nFree memory allocated using kmem_cache_zalloc using kmem_cache_free\nrather than kfree.\n\nThe Coccinelle semantic patch that makes this change is as follows:\n\n// \u003csmpl\u003e\n@@\nexpression x,E,c;\n@@\n\n x \u003d \\(kmem_cache_alloc\\|kmem_cache_zalloc\\|kmem_cache_alloc_node\\)(c,...)\n ... when !\u003d x \u003d E\n     when !\u003d \u0026x\n?-kfree(x)\n+kmem_cache_free(c,x)\n// \u003c/smpl\u003e\n\nSigned-off-by: Himangi Saraogi \u003chimangi774@gmail.com\u003e\nAcked-by: Julia Lawall \u003cjulia.lawall@lip6.fr\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "587870e8650a0571e895cc879cd895c78c6391bf",
      "tree": "19ab81bfde8f7d2165f19396262ae4bcf4ffe5ff",
      "parents": [
        "a71d64b4dc4067808549935583d984c7fc9ea647"
      ],
      "author": {
        "name": "Thomas Falcon",
        "email": "tlfalcon@linux.vnet.ibm.com",
        "time": "Tue Aug 05 16:42:39 2014 -0500"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:14:04 2014 +1000"
      },
      "message": "powerpc/pseries/hvcserver: Fix endian issue in hvcs_get_partner_info\n\nA buffer returned by H_VTERM_PARTNER_INFO contains device information\nin big endian format, causing problems for little endian architectures.\nThis patch ensures that they are in cpu endian.\n\nSigned-off-by: Thomas Falcon \u003ctlfalcon@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "a71d64b4dc4067808549935583d984c7fc9ea647",
      "tree": "8679cdc9688707feb94e24f19a7c848009f654c4",
      "parents": [
        "56758e3c3ca7e3d8960e0e8f881e80adf119f4fe"
      ],
      "author": {
        "name": "Anton Blanchard",
        "email": "anton@samba.org",
        "time": "Tue Aug 05 14:55:00 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:48 2014 +1000"
      },
      "message": "powerpc: Hard disable interrupts in xmon\n\nxmon only soft disables interrupts. This seems like a bad idea - we\ncertainly don\u0027t want decrementer and PMU exceptions going off when\nwe are debugging something inside xmon.\n\nThis issue was uncovered when the hard lockup detector went off\ninside xmon. To ensure we wont get a spurious hard lockup warning,\nI also call touch_nmi_watchdog() when exiting xmon.\n\nSigned-off-by: Anton Blanchard \u003canton@samba.org\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "56758e3c3ca7e3d8960e0e8f881e80adf119f4fe",
      "tree": "e9c8aa463f205058b73fdcad109514e7fe55326f",
      "parents": [
        "5efbabe09d986f25c02d19954660238fcd7f008a"
      ],
      "author": {
        "name": "Nishanth Aravamudan",
        "email": "nacc@linux.vnet.ibm.com",
        "time": "Mon Aug 11 16:43:18 2014 -0700"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:47 2014 +1000"
      },
      "message": "powerpc: remove duplicate definition of TEXASR_FS\n\nIt appears that commits 7f06f21d40a6 (\"powerpc/tm: Add checking to\ntreclaim/trechkpt\") and e4e38121507a (\"KVM: PPC: Book3S HV: Add\ntransactional memory support\") both added definitions of TEXASR_FS.\nRemove one of them. At the same time, fix the alignment of the remaining\ndefinition (should be tab-separated like the rest of the #defines).\n\nSigned-off-by: Nishanth Aravamudan \u003cnacc@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "5efbabe09d986f25c02d19954660238fcd7f008a",
      "tree": "9ed3958ab9e81e79d17362c6867e6aa35b91b259",
      "parents": [
        "f1b3929c232784580e5d8ee324b6bc634e709575"
      ],
      "author": {
        "name": "Gavin Shan",
        "email": "gwshan@linux.vnet.ibm.com",
        "time": "Mon Aug 11 19:16:20 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:47 2014 +1000"
      },
      "message": "powerpc/pseries: Avoid deadlock on removing ddw\n\nFunction remove_ddw() could be called in of_reconfig_notifier and\nwe potentially remove the dynamic DMA window property, which invokes\nof_reconfig_notifier again. Eventually, it leads to the deadlock as\nfollowing backtrace shows.\n\nThe patch fixes the above issue by deferring releasing the dynamic\nDMA window property while releasing the device node.\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n[ INFO: possible recursive locking detected ]\n3.16.0+ #428 Tainted: G        W\n---------------------------------------------\ndrmgr/2273 is trying to acquire lock:\n ((of_reconfig_chain).rwsem){.+.+..}, at: [\u003cc000000000091890\u003e] \\\n .__blocking_notifier_call_chain+0x40/0x78\n\nbut task is already holding lock:\n ((of_reconfig_chain).rwsem){.+.+..}, at: [\u003cc000000000091890\u003e] \\\n .__blocking_notifier_call_chain+0x40/0x78\n\nother info that might help us debug this:\n Possible unsafe locking scenario:\n\n       CPU0\n       ----\n  lock((of_reconfig_chain).rwsem);\n  lock((of_reconfig_chain).rwsem);\n *** DEADLOCK ***\n\n May be due to missing lock nesting notation\n\n2 locks held by drmgr/2273:\n #0:  (sb_writers#4){.+.+.+}, at: [\u003cc0000000001cbe70\u003e] \\\n      .vfs_write+0xb0/0x1f8\n #1:  ((of_reconfig_chain).rwsem){.+.+..}, at: [\u003cc000000000091890\u003e] \\\n      .__blocking_notifier_call_chain+0x40/0x78\n\nstack backtrace:\nCPU: 17 PID: 2273 Comm: drmgr Tainted: G        W     3.16.0+ #428\nCall Trace:\n[c0000000137e7000] [c000000000013d9c] .show_stack+0x88/0x148 (unreliable)\n[c0000000137e70b0] [c00000000083cd34] .dump_stack+0x7c/0x9c\n[c0000000137e7130] [c0000000000b8afc] .__lock_acquire+0x128c/0x1c68\n[c0000000137e7280] [c0000000000b9a4c] .lock_acquire+0xe8/0x104\n[c0000000137e7350] [c00000000083588c] .down_read+0x4c/0x90\n[c0000000137e73e0] [c000000000091890] .__blocking_notifier_call_chain+0x40/0x78\n[c0000000137e7490] [c000000000091900] .blocking_notifier_call_chain+0x38/0x48\n[c0000000137e7520] [c000000000682a28] .of_reconfig_notify+0x34/0x5c\n[c0000000137e75b0] [c000000000682a9c] .of_property_notify+0x4c/0x54\n[c0000000137e7650] [c000000000682bf0] .of_remove_property+0x30/0xd4\n[c0000000137e76f0] [c000000000052a44] .remove_ddw+0x144/0x168\n[c0000000137e7790] [c000000000053204] .iommu_reconfig_notifier+0x30/0xe0\n[c0000000137e7820] [c00000000009137c] .notifier_call_chain+0x6c/0xb4\n[c0000000137e78c0] [c0000000000918ac] .__blocking_notifier_call_chain+0x5c/0x78\n[c0000000137e7970] [c000000000091900] .blocking_notifier_call_chain+0x38/0x48\n[c0000000137e7a00] [c000000000682a28] .of_reconfig_notify+0x34/0x5c\n[c0000000137e7a90] [c000000000682e14] .of_detach_node+0x44/0x1fc\n[c0000000137e7b40] [c0000000000518e4] .ofdt_write+0x3ac/0x688\n[c0000000137e7c20] [c000000000238430] .proc_reg_write+0xb8/0xd4\n[c0000000137e7cd0] [c0000000001cbeac] .vfs_write+0xec/0x1f8\n[c0000000137e7d70] [c0000000001cc3b0] .SyS_write+0x58/0xa0\n[c0000000137e7e30] [c00000000000a064] syscall_exit+0x0/0x98\n\nCc: stable@vger.kernel.org\nSigned-off-by: Gavin Shan \u003cgwshan@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "f1b3929c232784580e5d8ee324b6bc634e709575",
      "tree": "b0eafa3e7a12a953b69ed5bd7ec175377bc3c1c5",
      "parents": [
        "ce8f150a174c105abddbd3cc193cbba57f9db255"
      ],
      "author": {
        "name": "Gavin Shan",
        "email": "gwshan@linux.vnet.ibm.com",
        "time": "Mon Aug 11 19:16:19 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:46 2014 +1000"
      },
      "message": "powerpc/pseries: Failure on removing device node\n\nWhile running command \"drmgr -c phb -r -s \u0027PHB 528\u0027\", following\nbacktrace jumped out because the target device node isn\u0027t marked\nwith OF_DETACHED by of_detach_node(), which caused by error\nreturned from memory hotplug related reconfig notifier when\ndisabling CONFIG_MEMORY_HOTREMOVE. The patch fixes it.\n\nERROR: Bad of_node_put() on /pci@800000020000210/ethernet@0\nCPU: 14 PID: 2252 Comm: drmgr Tainted: G        W     3.16.0+ #427\nCall Trace:\n[c000000012a776a0] [c000000000013d9c] .show_stack+0x88/0x148 (unreliable)\n[c000000012a77750] [c00000000083cd34] .dump_stack+0x7c/0x9c\n[c000000012a777d0] [c0000000006807c4] .of_node_release+0x58/0xe0\n[c000000012a77860] [c00000000038a7d0] .kobject_release+0x174/0x1b8\n[c000000012a77900] [c00000000038a884] .kobject_put+0x70/0x78\n[c000000012a77980] [c000000000681680] .of_node_put+0x28/0x34\n[c000000012a77a00] [c000000000681ea8] .__of_get_next_child+0x64/0x70\n[c000000012a77a90] [c000000000682138] .of_find_node_by_path+0x1b8/0x20c\n[c000000012a77b40] [c000000000051840] .ofdt_write+0x308/0x688\n[c000000012a77c20] [c000000000238430] .proc_reg_write+0xb8/0xd4\n[c000000012a77cd0] [c0000000001cbeac] .vfs_write+0xec/0x1f8\n[c000000012a77d70] [c0000000001cc3b0] .SyS_write+0x58/0xa0\n[c000000012a77e30] [c00000000000a064] syscall_exit+0x0/0x98\n\nCc: stable@vger.kernel.org\nSigned-off-by: Gavin Shan \u003cgwshan@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "ce8f150a174c105abddbd3cc193cbba57f9db255",
      "tree": "426705827fccf91e39c864da9cec515c9277d729",
      "parents": [
        "b09c2ec4082c63584491f35df2cb530ee8ca312d"
      ],
      "author": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Mon Aug 11 14:37:22 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:45 2014 +1000"
      },
      "message": "powerpc/boot: Use correct zlib types for comparison\n\nAvoids this warning:\n\narch/powerpc/boot/gunzip_util.c:118:9: warning: comparison of distinct pointer types lacks a cast\n\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "b09c2ec4082c63584491f35df2cb530ee8ca312d",
      "tree": "05200fb46b8c6839065a64e148a9b01813f5de54",
      "parents": [
        "14c4000a88afaaa2d0877cc86d42a74fde0f35e0"
      ],
      "author": {
        "name": "Vasant Hegde",
        "email": "hegdevasant@linux.vnet.ibm.com",
        "time": "Sat Aug 09 11:15:45 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:45 2014 +1000"
      },
      "message": "powerpc/powernv: Interface to register/unregister opal dump region\n\nPowerNV platform is capable of capturing host memory region when system\ncrashes (because of host/firmware). We have new OPAL API to register/\nunregister memory region to be captured when system crashes.\n\nThis patch adds support for new API. Also during boot time we register\nkernel log buffer and unregister before doing kexec.\n\nSigned-off-by: Vasant Hegde \u003chegdevasant@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "14c4000a88afaaa2d0877cc86d42a74fde0f35e0",
      "tree": "b4133597b6990120a60302cf6b677b463bbc1752",
      "parents": [
        "3609e09fd824c37df6f2bf13caa88f6f54a11922"
      ],
      "author": {
        "name": "Vasant Hegde",
        "email": "hegdevasant@linux.vnet.ibm.com",
        "time": "Sat Aug 09 11:15:30 2014 +0530"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:44 2014 +1000"
      },
      "message": "printk: Add function to return log buffer address and size\n\nPlatforms like IBM Power Systems supports service processor\nassisted dump. It provides interface to add memory region to\nbe captured when system is crashed.\n\nDuring initialization/running we can add kernel memory region\nto be collected.\n\nPresently we don\u0027t have a way to get the log buffer base address\nand size. This patch adds support to return log buffer address\nand size.\n\nSigned-off-by: Vasant Hegde \u003chegdevasant@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\nAcked-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n"
    },
    {
      "commit": "3609e09fd824c37df6f2bf13caa88f6f54a11922",
      "tree": "0835aa502a217e3cd30f1243f5318fbd84f56e9e",
      "parents": [
        "97b3be1e9404f4c1645f3286428abd4d2052041e"
      ],
      "author": {
        "name": "Michael Ellerman",
        "email": "mpe@ellerman.id.au",
        "time": "Wed Aug 06 15:42:17 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:43 2014 +1000"
      },
      "message": "powerpc: Add POWER8 features to CPU_FTRS_POSSIBLE/ALWAYS\n\nWe have been a bit slack about updating the CPU_FTRS_POSSIBLE and\nCPU_FTRS_ALWAYS masks. When we added POWER8, and also POWER8E we forgot\nto update the ALWAYS mask. And when we added POWER8_DD1 we forgot to\nupdate both the POSSIBLE and ALWAYS masks.\n\nLuckily this hasn\u0027t caused any actual bugs AFAICS. Failing to update the\nALWAYS mask just forgoes a potential optimisation opportunity. Failing\nto update the POSSIBLE mask for POWER8_DD1 is also OK because it only\nremoves a bit rather than adding any.\n\nRegardless they should all be in both masks so as to avoid any future\nbugs when the set of ALWAYS/POSSIBLE bits changes, or the masks\nthemselves change.\n\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\nAcked-by: Michael Neuling \u003cmikey@neuling.org\u003e\nAcked-by: Joel Stanley \u003cjoel@jms.id.au\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "97b3be1e9404f4c1645f3286428abd4d2052041e",
      "tree": "31e79eebdf8c4ec0cb539c7bcd6d75ee7f889034",
      "parents": [
        "763fe0addb8fe15ccea67c0aebddc06f4bb25439"
      ],
      "author": {
        "name": "Alistair Popple",
        "email": "alistair@popple.id.au",
        "time": "Wed Aug 06 17:03:09 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:42 2014 +1000"
      },
      "message": "powerpc/ppc476: Disable BTAC\n\nThis patch disables the branch target address CAM which under specific\ncircumstances may cause the processor to skip execution of 1-4\ninstructions. This fixes IBM Erratum #47.\n\nSigned-off-by: Alistair Popple \u003calistair@popple.id.au\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "763fe0addb8fe15ccea67c0aebddc06f4bb25439",
      "tree": "cc02c5838271e72aba44e6473ced75d5f92bae79",
      "parents": [
        "78e05b1421fa41ae8457701140933baa5e7d9479"
      ],
      "author": {
        "name": "Gavin Shan",
        "email": "gwshan@linux.vnet.ibm.com",
        "time": "Wed Aug 06 17:10:16 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:42 2014 +1000"
      },
      "message": "powerpc/powernv: Fix IOMMU group lost\n\nWhen we take full hotplug to recover from EEH errors, PCI buses\ncould be involved. For the case, the child devices of involved\nPCI buses can\u0027t be attached to IOMMU group properly, which is\ncaused by commit 3f28c5a (\"powerpc/powernv: Reduce multi-hit of\niommu_add_device()\").\n\nWhen adding the PCI devices of the newly created PCI buses to\nthe system, the IOMMU group is expected to be added in (C).\n(A) fails to bind the IOMMU group because bus-\u003eis_added is\nfalse. (B) fails because the device doesn\u0027t have binding IOMMU\ntable yet. bus-\u003eis_added is set to true at end of (C) and\npdev-\u003eis_added is set to true at (D).\n\n   pcibios_add_pci_devices()\n      pci_scan_bridge()\n         pci_scan_child_bus()\n            pci_scan_slot()\n               pci_scan_single_device()\n                  pci_scan_device()\n                  pci_device_add()\n                     pcibios_add_device()           A: Ignore\n                     device_add()                   B: Ignore\n                  pcibios_fixup_bus()\n                     pcibios_setup_bus_devices()\n                        pcibios_setup_device()      C: Hit\n      pcibios_finish_adding_to_bus()\n         pci_bus_add_devices()\n            pci_bus_add_device()                    D: Add device\n\nIf the parent PCI bus isn\u0027t involved in hotplug, the IOMMU\ngroup is expected to be bound in (B). (A) should fail as the\nsysfs entries aren\u0027t populated.\n\nThe patch fixes the issue by reverting commit 3f28c5a and remove\nWARN_ON() in iommu_add_device() to allow calling the function\neven the specified device already has associated IOMMU group.\n\nCc: \u003cstable@vger.kernel.org\u003e  # 3.16+\nReported-by: Thadeu Lima de Souza Cascardo \u003ccascardo@linux.vnet.ibm.com\u003e\nSigned-off-by: Gavin Shan \u003cgwshan@linux.vnet.ibm.com\u003e\nAcked-by: Wei Yang \u003cweiyang@linux.vnet.ibm.com\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "78e05b1421fa41ae8457701140933baa5e7d9479",
      "tree": "ddd3d45946163cdc5f5836d2e460b9b4f4dd24c2",
      "parents": [
        "51d7d5205d3389a32859f9939f1093f267409929"
      ],
      "author": {
        "name": "Michael Ellerman",
        "email": "mpe@ellerman.id.au",
        "time": "Thu Aug 07 15:36:18 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:27 2014 +1000"
      },
      "message": "powerpc: Add smp_mb()s to arch_spin_unlock_wait()\n\nSimilar to the previous commit which described why we need to add a\nbarrier to arch_spin_is_locked(), we have a similar problem with\nspin_unlock_wait().\n\nWe need a barrier on entry to ensure any spinlock we have previously\ntaken is visibly locked prior to the load of lock-\u003eslock.\n\nIt\u0027s also not clear if spin_unlock_wait() is intended to have ACQUIRE\nsemantics. For now be conservative and add a barrier on exit to give it\nACQUIRE semantics.\n\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "51d7d5205d3389a32859f9939f1093f267409929",
      "tree": "7b8804d9918c271747f685245ec480b4125b7675",
      "parents": [
        "11d549048e5bea3055053d8dcf45089b1fd711f1"
      ],
      "author": {
        "name": "Michael Ellerman",
        "email": "mpe@ellerman.id.au",
        "time": "Thu Aug 07 15:36:17 2014 +1000"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:26 2014 +1000"
      },
      "message": "powerpc: Add smp_mb() to arch_spin_is_locked()\n\nThe kernel defines the function spin_is_locked(), which can be used to\ncheck if a spinlock is currently locked.\n\nUsing spin_is_locked() on a lock you don\u0027t hold is obviously racy. That\nis, even though you may observe that the lock is unlocked, it may become\nlocked at any time.\n\nThere is (at least) one exception to that, which is if two locks are\nused as a pair, and the holder of each checks the status of the other\nbefore doing any update.\n\nAssuming *A and *B are two locks, and *COUNTER is a shared non-atomic\nvalue:\n\nThe first CPU does:\n\n\tspin_lock(*A)\n\n\tif spin_is_locked(*B)\n\t\t# nothing\n\telse\n\t\tsmp_mb()\n\t\tLOAD r \u003d *COUNTER\n\t\tr++\n\t\tSTORE *COUNTER \u003d r\n\n\tspin_unlock(*A)\n\nAnd the second CPU does:\n\n\tspin_lock(*B)\n\n\tif spin_is_locked(*A)\n\t\t# nothing\n\telse\n\t\tsmp_mb()\n\t\tLOAD r \u003d *COUNTER\n\t\tr++\n\t\tSTORE *COUNTER \u003d r\n\n\tspin_unlock(*B)\n\nAlthough this is a strange locking construct, it should work.\n\nIt seems to be understood, but not documented, that spin_is_locked() is\nnot a memory barrier, so in the examples above and below the caller\ninserts its own memory barrier before acting on the result of\nspin_is_locked().\n\nFor now we assume spin_is_locked() is implemented as below, and we break\nit out in our examples:\n\n\tbool spin_is_locked(*LOCK) {\n\t\tLOAD l \u003d *LOCK\n\t\treturn l.locked\n\t}\n\nOur intuition is that there should be no problem even if the two code\nsequences run simultaneously such as:\n\n\tCPU 0\t\t\tCPU 1\n\t\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\tspin_lock(*A)\t\tspin_lock(*B)\n\tLOAD b \u003d *B\t\tLOAD a \u003d *A\n\tif b.locked # true\tif a.locked # true\n\t# nothing\t\t# nothing\n\tspin_unlock(*A)\t\tspin_unlock(*B)\n\nIf one CPU gets the lock before the other then it will do the update and\nthe other CPU will back off:\n\n\tCPU 0\t\t\tCPU 1\n\t\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\tspin_lock(*A)\n\tLOAD b \u003d *B\n\t\t\t\tspin_lock(*B)\n\tif b.locked # false\tLOAD a \u003d *A\n\telse\t\t\tif a.locked # true\n\tsmp_mb()\t\t# nothing\n\tLOAD r1 \u003d *COUNTER\tspin_unlock(*B)\n\tr1++\n\tSTORE *COUNTER \u003d r1\n\tspin_unlock(*A)\n\nHowever in reality spin_lock() itself is not indivisible. On powerpc we\nimplement it as a load-and-reserve and store-conditional.\n\nIgnoring the retry logic for the lost reservation case, it boils down to:\n\tspin_lock(*LOCK) {\n\t\tLOAD l \u003d *LOCK\n\t\tl.locked \u003d true\n\t\tSTORE *LOCK \u003d l\n\t\tACQUIRE_BARRIER\n\t}\n\nThe ACQUIRE_BARRIER is required to give spin_lock() ACQUIRE semantics as\ndefined in memory-barriers.txt:\n\n     This acts as a one-way permeable barrier.  It guarantees that all\n     memory operations after the ACQUIRE operation will appear to happen\n     after the ACQUIRE operation with respect to the other components of\n     the system.\n\nOn modern powerpc systems we use lwsync for ACQUIRE_BARRIER. lwsync is\nalso know as \"lightweight sync\", or \"sync 1\".\n\nAs described in Power ISA v2.07 section B.2.1.1, in this scenario the\nlwsync is not the barrier itself. It instead causes the LOAD of *LOCK to\nact as the barrier, preventing any loads or stores in the locked region\nfrom occurring prior to the load of *LOCK.\n\nWhether this behaviour is in accordance with the definition of ACQUIRE\nsemantics in memory-barriers.txt is open to discussion, we may switch to\na different barrier in future.\n\nWhat this means in practice is that the following can occur:\n\n\tCPU 0\t\t\tCPU 1\n\t\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\tLOAD a \u003d *A \t\tLOAD b \u003d *B\n\ta.locked \u003d true\t\tb.locked \u003d true\n\tLOAD b \u003d *B\t\tLOAD a \u003d *A\n\tSTORE *A \u003d a\t\tSTORE *B \u003d b\n\tif b.locked # false\tif a.locked # false\n\telse\t\t\telse\n\tsmp_mb()\t\tsmp_mb()\n\tLOAD r1 \u003d *COUNTER\tLOAD r2 \u003d *COUNTER\n\tr1++\t\t\tr2++\n\tSTORE *COUNTER \u003d r1\n\t\t\t\tSTORE *COUNTER \u003d r2\t# Lost update\n\tspin_unlock(*A)\t\tspin_unlock(*B)\n\nThat is, the load of *B can occur prior to the store that makes *A\nvisibly locked. And similarly for CPU 1. The result is both CPUs hold\ntheir lock and believe the other lock is unlocked.\n\nThe easiest fix for this is to add a full memory barrier to the start of\nspin_is_locked(), so adding to our previous definition would give us:\n\n\tbool spin_is_locked(*LOCK) {\n\t\tsmp_mb()\n\t\tLOAD l \u003d *LOCK\n\t\treturn l.locked\n\t}\n\nThe new barrier orders the store to the lock we are locking vs the load\nof the other lock:\n\n\tCPU 0\t\t\tCPU 1\n\t\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\tLOAD a \u003d *A \t\tLOAD b \u003d *B\n\ta.locked \u003d true\t\tb.locked \u003d true\n\tSTORE *A \u003d a\t\tSTORE *B \u003d b\n\tsmp_mb()\t\tsmp_mb()\n\tLOAD b \u003d *B\t\tLOAD a \u003d *A\n\tif b.locked # true\tif a.locked # true\n\t# nothing\t\t# nothing\n\tspin_unlock(*A)\t\tspin_unlock(*B)\n\nAlthough the above example is theoretical, there is code similar to this\nexample in sem_lock() in ipc/sem.c. This commit in addition to the next\ncommit appears to be a fix for crashes we are seeing in that code where\nwe believe this race happens in practice.\n\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    },
    {
      "commit": "11d549048e5bea3055053d8dcf45089b1fd711f1",
      "tree": "c621f0b08610159ba41763cbb80a017b9ba19865",
      "parents": [
        "5d61a2172a0142c635ab6d7c3b1589af85a3603e"
      ],
      "author": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Aug 08 22:22:12 2014 -0700"
      },
      "committer": {
        "name": "Benjamin Herrenschmidt",
        "email": "benh@kernel.crashing.org",
        "time": "Wed Aug 13 15:13:25 2014 +1000"
      },
      "message": "powerpc: Fix \"attempt to move .org backwards\" error\n\nOnce again, we see\n\narch/powerpc/kernel/exceptions-64s.S: Assembler messages:\narch/powerpc/kernel/exceptions-64s.S:865: Error: attempt to move .org backwards\narch/powerpc/kernel/exceptions-64s.S:866: Error: attempt to move .org backwards\narch/powerpc/kernel/exceptions-64s.S:890: Error: attempt to move .org backwards\n\nwhen compiling ppc:allmodconfig.\n\nThis time the problem has been caused by to commit 0869b6fd209bda\n(\"powerpc/book3s: Add basic infrastructure to handle HMI in Linux\"),\nwhich adds functions hmi_exception_early and hmi_exception_after_realmode\ninto a critical (size-limited) code area, even though that does not appear\nto be necessary.\n\nMove those functions to a non-critical area of the file.\n\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n"
    }
  ],
  "next": "5d61a2172a0142c635ab6d7c3b1589af85a3603e"
}