)]}' { "log": [ { "commit": "c8d2bc9bc39ebea8437fd974fdbc21847bb897a3", "tree": "c28505808ed54d68352da0909620d0b43512c6c0", "parents": [ "f76d9c61d91343806e59335493806e87daf78947" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 16:24:33 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 16:24:33 2016 -0700" }, "message": "Linux 4.8\n" }, { "commit": "f76d9c61d91343806e59335493806e87daf78947", "tree": "e3c98a687783c5fbe9862646f455435fb6bfe3f8", "parents": [ "be67d60ba944bdd38571b79bdcd506e34c0f16c1", "117e5e9c4cfcb7628f08de074fbfefec1bb678b7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 15:23:00 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 15:23:00 2016 -0700" }, "message": "Merge branch \u0027fixes\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm\n\nPull ARM fixes from Russell King:\n \"Three relatively small fixes for ARM:\n\n - Roger noticed that dma_max_pfn() was calculating the upper limit\n wrongly, by adding the PFN offset of memory twice.\n\n - A fix from Robin to correct parsing of MPIDR values when the\n address size is larger than one BE32 unit.\n\n - A fix from Srinivas to ensure that we do not rely on the boot\n loader (or previous Linux kernel) setting the translation table\n base register a certain way in the decompressor, which can lead to\n crashes\"\n\n* \u0027fixes\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm:\n ARM: 8618/1: decompressor: reset ttbcr fields to use TTBR0 on ARMv7\n ARM: 8617/1: dma: fix dma_max_pfn()\n ARM: 8616/1: dt: Respect property size when parsing CPUs\n" }, { "commit": "117e5e9c4cfcb7628f08de074fbfefec1bb678b7", "tree": "c2fff3c27b8602ce3e9d49abb9c42daba10f20d4", "parents": [ "d248220f0465b818887baa9829e691fe662b2c5e" ], "author": { "name": "Srinivas Ramana", "email": "sramana@codeaurora.org", "time": "Fri Sep 30 15:03:31 2016 +0100" }, "committer": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Sun Oct 02 20:05:14 2016 +0100" }, "message": "ARM: 8618/1: decompressor: reset ttbcr fields to use TTBR0 on ARMv7\n\nIf the bootloader uses the long descriptor format and jumps to\nkernel decompressor code, TTBCR may not be in a right state.\nBefore enabling the MMU, it is required to clear the TTBCR.PD0\nfield to use TTBR0 for translation table walks.\n\nThe commit dbece45894d3a (\"ARM: 7501/1: decompressor:\nreset ttbcr for VMSA ARMv7 cores\") does the reset of TTBCR.N, but\ndoesn\u0027t consider all the bits for the size of TTBCR.N.\n\nClear TTBCR.PD0 field and reset all the three bits of TTBCR.N to\nindicate the use of TTBR0 and the correct base address width.\n\nFixes: dbece45894d3 (\"ARM: 7501/1: decompressor: reset ttbcr for VMSA ARMv7 cores\")\nAcked-by: Robin Murphy \u003crobin.murphy@arm.com\u003e\nSigned-off-by: Srinivas Ramana \u003csramana@codeaurora.org\u003e\nSigned-off-by: Russell King \u003crmk+kernel@arm.linux.org.uk\u003e\n" }, { "commit": "be67d60ba944bdd38571b79bdcd506e34c0f16c1", "tree": "8b829fefea32222290d0ff58f1ef9d7e90479377", "parents": [ "66188fb11a82692629e85b6cbc3ecc08c752d2dc", "2fa5f04f85730d0c4f49f984b7efeb4f8d5bd1fc" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 11:04:29 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 11:04:29 2016 -0700" }, "message": "Merge branch \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Thomas Gleixner:\n \"The last regression fixes for 4.8 final:\n\n - Two patches addressing the fallout of the CR4 optimizations which\n caused CR4-less machines to fail.\n\n - Fix the VDSO build on big endian machines\n\n - Take care of FPU initialization if no CPUID is available otherwise\n task struct size ends up being zero\n\n - Fix up context tracking in case load_gs_index fails\"\n\n* \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/entry/64: Fix context tracking state warning when load_gs_index fails\n x86/boot: Initialize FPU and X86_FEATURE_ALWAYS even if we don\u0027t have CPUID\n x86/vdso: Fix building on big endian host\n x86/boot: Fix another __read_cr4() case on 486\n x86/init: Fix cr4_init_shadow() on CR4-less machines\n" }, { "commit": "66188fb11a82692629e85b6cbc3ecc08c752d2dc", "tree": "baa7fea3bd3df098d81947c6a52fc60c923b722b", "parents": [ "0c7fc30f18220c0d1b8fb2636811a0c0d0d65714", "6605d156bdfbb2502ba301bc4fbd8db696ae4b6d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:53:38 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:53:38 2016 -0700" }, "message": "Merge branch \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus\n\nPull MIPS fixes from Ralf Baechle:\n \"Another round of fixes:\n\n - CM: Fix mips_cm_max_vp_width for non-MT kernels on MT systems\n - CPS: Avoid BUG() when offlining pre-r6 CPUs\n - DEC: Avoid gas warnings due to suspicious instruction scheduling by\n manually expanding assembler macros.\n - FTLB: Fix configuration by moving confiuguratoin after probing\n - FTLB: clear execution hazard after changing FTLB enable\n - Highmem: Fix detection of unsupported highmem with cache aliases\n - I6400: Don\u0027t touch FTLBP chicken bits\n - microMIPS: Fix BUILD_ROLLBACK_PROLOGUE\n - Malta: Fix IOCU disable switch read for MIPS64\n - Octeon: Fix probing of devices attached to GPIO lines\n - uprobes: Misc small fixes\"\n\n* \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:\n MIPS: CM: Fix mips_cm_max_vp_width for non-MT kernels on MT systems\n MIPS: Fix detection of unsupported highmem with cache aliases\n MIPS: Malta: Fix IOCU disable switch read for MIPS64\n MIPS: Fix BUILD_ROLLBACK_PROLOGUE for microMIPS\n MIPS: clear execution hazard after changing FTLB enable\n MIPS: Configure FTLB after probing TLB sizes from config4\n MIPS: Stop setting I6400 FTLBP\n MIPS: DEC: Avoid la pseudo-instruction in delay slots\n MIPS: Octeon: mark GPIO controller node not populated after IRQ init.\n MIPS: uprobes: fix use of uninitialised variable\n MIPS: uprobes: remove incorrect set_orig_insn\n MIPS: fix uretprobe implementation\n MIPS: smp-cps: Avoid BUG() when offlining pre-r6 CPUs\n" }, { "commit": "0c7fc30f18220c0d1b8fb2636811a0c0d0d65714", "tree": "0bfdc30ba7b529b3845b63bfb5c12bb9198ebbf0", "parents": [ "bb6bbc7ca2254fd885f5b85f4cc0cda7cf04f8c1", "2a0100d7beebd2fb2c436852648f75396253e9f8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:42:26 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:42:26 2016 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc\n\nPull sparc fixes from David Miller:\n\n 1) Fix section mismatches in some builds, from Paul Gortmaker.\n\n 2) Need to count huge zero page mappings when doing TSB sizing, from\n Mike Kravetz.\n\n 3) Fix handing of cpu_possible_mask when nr_cpus module option is\n specified, from Atish Patra.\n\n 4) Don\u0027t allocate irq stacks until nr_irqs has been processed, also\n from Atish Patra.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:\n sparc64: Fix non-SMP build.\n sparc64: Fix irq stack bootmem allocation.\n sparc64: Fix cpu_possible_mask if nr_cpus is set\n sparc64 mm: Fix more TSB sizing issues\n sparc64: fix section mismatch in find_numa_latencies_for_group\n" }, { "commit": "bb6bbc7ca2254fd885f5b85f4cc0cda7cf04f8c1", "tree": "00ea99c4249a670f2325b84fc177a7a182f10e86", "parents": [ "f51fdffad5b7709d0ade40736b58a2da2707fa15", "1cceda7849809a8857fd9f26efe8846506c710e1" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:36:41 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 02 10:36:41 2016 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nPull networking fixes from David Miller:\n\n 1) Fix wrong TCP checksums on MTU probing when checksum offloading is\n disabled, from Douglas Caetano dos Santos.\n\n 2) Fix qdisc backlog updates in qfq and sfb schedulers, from Cong Wang.\n\n 3) Route lookup flow key protocol value is wrong in ip6gre_xmit_other(),\n fix from Lance Richardson.\n\n 4) Scheduling while atomic in multicast routing code of ipv4 and ipv6,\n fix from Nikolay Aleksandrov.\n\n 5) Fix packet alignment in fec driver, from Eric Nelson.\n\n 6) Fix perf regression in sctp due to struct layout and cache misses,\n from Xin Long.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:\n sctp: fix the issue sctp_diag uses lock_sock in rcu_read_lock\n sctp: change to check peer prsctp_capable when using prsctp polices\n sctp: remove prsctp_param from sctp_chunk\n sctp: move sent_count to the memory hole in sctp_chunk\n tg3: Avoid NULL pointer dereference in tg3_io_error_detected()\n act_ife: Fix false encoding\n act_ife: Fix external mac header on encode\n VSOCK: Don\u0027t dec ack backlog twice for rejected connections\n Revert \"net: ethernet: bcmgenet: use phydev from struct net_device\"\n net: fec: align IP header in hardware\n net: fec: remove QUIRK_HAS_RACC from i.mx27\n net: fec: remove QUIRK_HAS_RACC from i.mx25\n ipmr, ip6mr: fix scheduling while atomic and a deadlock with ipmr_get_route\n ip6_gre: fix flowi6_proto value in ip6gre_xmit_other()\n tcp: fix a compile error in DBGUNDO()\n tcp: fix wrong checksum calculation on MTU probing\n sch_sfb: keep backlog updated with qlen\n sch_qfq: keep backlog updated with qlen\n can: dev: fix deadlock reported after bus-off\n" }, { "commit": "6605d156bdfbb2502ba301bc4fbd8db696ae4b6d", "tree": "a0348acb786b807e158e3164fa7fd50a4ab653eb", "parents": [ "058effe7fdc5776b017356f690976a857eea473f" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Sep 30 17:25:01 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Sun Oct 02 01:40:56 2016 +0200" }, "message": "MIPS: CM: Fix mips_cm_max_vp_width for non-MT kernels on MT systems\n\nWhen discovering the number of VPEs per core, smp_num_siblings will be\nincorrect for kernels built without support for the MIPS MultiThreading\n(MT) ASE running on systems which implement said ASE. This leads to\naccesses to VPEs in secondary cores being performed incorrectly since\nmips_cm_vp_id calculates the wrong ID to write to the local \"other\"\nregisters. Fix this by examining the number of VPEs in the core as\nreported by the CM.\n\nThis patch presumes that the number of VPEs will be the same in each\ncore of the system. As this path only applies to systems with CM version\n2.5 or lower, and this property is true of all such known systems, this\nis likely to be fine but is described in a comment for good measure.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14338/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "f51fdffad5b7709d0ade40736b58a2da2707fa15", "tree": "6d188ad3361bf26446f3718d2c28f609082fd924", "parents": [ "2161a2a644a6d33a29d68395518d103ed805758f", "539294b76af8922297702a7ebb8cafe68f7e5376" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 01 07:37:15 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 01 07:37:15 2016 -0700" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fix from James Bottomley:\n \"One final fix before 4.8.\n\n There was a memory leak triggered by turning scsi mq off due to the\n fact that we assume on host release that the already running hosts\n weren\u0027t mq based because that\u0027s the state of the global flag (even\n though they were).\n\n Fix it by tracking this on a per host host basis\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: Avoid that toggling use_blk_mq triggers a memory leak\n" }, { "commit": "2161a2a644a6d33a29d68395518d103ed805758f", "tree": "0e28b059ef641b27988c307ce319c1ca6a421a58", "parents": [ "dbd8805b0ad46f04cababea5d01f072142961411", "9fb6de1b0bf4ec11573b76059da3c3b39ac7f2ff" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 21:25:09 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 21:25:09 2016 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input\n\nPull input fix from Dmitry Torokhov:\n \"One small change to make joydev (which is used by older games) to bind\n to devices that export Z axis but not X or Y (such as TRC rudder)\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Input: joydev - recognize devices with Z axis as joysticks\n" }, { "commit": "dbd8805b0ad46f04cababea5d01f072142961411", "tree": "6ac45a7687ba67362d8a39e5da032625fc2e11b1", "parents": [ "9a2172a8d52cf14ce44e9cadd8c9df84cf832d75", "37aa7271d9742b574763e5ce019bde9c49aa8bfe" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 15:51:10 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 15:51:10 2016 -0700" }, "message": "Merge branch \u0027akpm\u0027 (patches from Andrew)\n\nMerge more fixes from Andrew Morton:\n \"Three fixes\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e:\n include/linux/property.h: fix typo/compile error\n ocfs2: fix deadlock on mmapped page in ocfs2_write_begin_nolock()\n mm: workingset: fix crash in shadow node shrinker caused by replace_page_cache_page()\n" }, { "commit": "37aa7271d9742b574763e5ce019bde9c49aa8bfe", "tree": "f988b42d07ef6d11efabbd374c9a069dfba2d236", "parents": [ "c33f0785bf292cf1d15f4fbe42869c63e205b21c" ], "author": { "name": "John Youn", "email": "johnyoun@synopsys.com", "time": "Fri Sep 30 15:11:35 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 15:26:52 2016 -0700" }, "message": "include/linux/property.h: fix typo/compile error\n\nThis fixes commit d76eebfa175e (\"include/linux/property.h: fix build\nissues with gcc-4.4.4\").\n\nWith that commit we get the following compile error when using the\nPROPERTY_ENTRY_INTEGER_ARRAY macro.\n\n include/linux/property.h:201:39: error: `u32_data\u0027 undeclared (first\n use in this function)\n PROPERTY_ENTRY_INTEGER_ARRAY(_name_, u32, _val_)\n ^\n include/linux/property.h:193:17: note: in definition of macro\n `PROPERTY_ENTRY_INTEGER_ARRAY\u0027\n { .pointer \u003d { _type_##_data \u003d _val_ } }, \\\n ^\n\nThis needs a \u0027.\u0027 to reference the union member. It seems this was just\noverlooked here since it is done correctly in similar constructs in\nother parts of the original commit.\n\nThis fix is in preparation of upcoming commits that will use this macro.\n\nFixes: commit d76eebfa175e (\"include/linux/property.h: fix build issues with gcc-4.4.4\")\nLink: http://lkml.kernel.org/r/2de3b929290d88a723ed829a3e3cbd02044714df.1475114627.git.johnyoun@synopsys.com\nSigned-off-by: John Youn \u003cjohnyoun@synopsys.com\u003e\nCc: \"Rafael J. Wysocki\" \u003crafael.j.wysocki@intel.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "c33f0785bf292cf1d15f4fbe42869c63e205b21c", "tree": "eb73b7096ef554702a66b8d37f1dbc5fdd419702", "parents": [ "22f2ac51b6d643666f4db093f13144f773ff3f3a" ], "author": { "name": "Eric Ren", "email": "zren@suse.com", "time": "Fri Sep 30 15:11:32 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 15:26:52 2016 -0700" }, "message": "ocfs2: fix deadlock on mmapped page in ocfs2_write_begin_nolock()\n\nThe testcase \"mmaptruncate\" of ocfs2-test deadlocks occasionally.\n\nIn this testcase, we create a 2*CLUSTER_SIZE file and mmap() on it;\nthere are 2 process repeatedly performing the following operations\nrespectively: one is doing memset(mmaped_addr + 2*CLUSTER_SIZE - 1, \u0027a\u0027,\n1), while the another is playing ftruncate(fd, 2*CLUSTER_SIZE) and then\nftruncate(fd, CLUSTER_SIZE) again and again.\n\nThis is the backtrace when the deadlock happens:\n\n __wait_on_bit_lock+0x50/0xa0\n __lock_page+0xb7/0xc0\n ocfs2_write_begin_nolock+0x163f/0x1790 [ocfs2]\n ocfs2_page_mkwrite+0x1c7/0x2a0 [ocfs2]\n do_page_mkwrite+0x66/0xc0\n handle_mm_fault+0x685/0x1350\n __do_page_fault+0x1d8/0x4d0\n trace_do_page_fault+0x37/0xf0\n do_async_page_fault+0x19/0x70\n async_page_fault+0x28/0x30\n\nIn ocfs2_write_begin_nolock(), we first grab the pages and then allocate\ndisk space for this write; ocfs2_try_to_free_truncate_log() will be\ncalled if -ENOSPC is returned; if we\u0027re lucky to get enough clusters,\nwhich is usually the case, we start over again.\n\nBut in ocfs2_free_write_ctxt() the target page isn\u0027t unlocked, so we\nwill deadlock when trying to grab the target page again.\n\nAlso, -ENOMEM might be returned in ocfs2_grab_pages_for_write().\nAnother deadlock will happen in __do_page_mkwrite() if\nocfs2_page_mkwrite() returns non-VM_FAULT_LOCKED, and along with a\nlocked target page.\n\nThese two errors fail on the same path, so fix them by unlocking the\ntarget page manually before ocfs2_free_write_ctxt().\n\nJan Kara helps me clear out the JBD2 part, and suggest the hint for root\ncause.\n\nChanges since v1:\n1. Also put ENOMEM error case into consideration.\n\nLink: http://lkml.kernel.org/r/1474173902-32075-1-git-send-email-zren@suse.com\nSigned-off-by: Eric Ren \u003czren@suse.com\u003e\nReviewed-by: He Gang \u003cghe@suse.com\u003e\nAcked-by: Joseph Qi \u003cjoseph.qi@huawei.com\u003e\nCc: Mark Fasheh \u003cmfasheh@suse.de\u003e\nCc: Joel Becker \u003cjlbec@evilplan.org\u003e\nCc: Junxiao Bi \u003cjunxiao.bi@oracle.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "22f2ac51b6d643666f4db093f13144f773ff3f3a", "tree": "d5a0830636a18b247b02a22602d57637f3bac110", "parents": [ "e3b3656ca63e23b5755183718df36fb9ff518b02" ], "author": { "name": "Johannes Weiner", "email": "hannes@cmpxchg.org", "time": "Fri Sep 30 15:11:29 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 15:26:52 2016 -0700" }, "message": "mm: workingset: fix crash in shadow node shrinker caused by replace_page_cache_page()\n\nAntonio reports the following crash when using fuse under memory pressure:\n\n kernel BUG at /build/linux-a2WvEb/linux-4.4.0/mm/workingset.c:346!\n invalid opcode: 0000 [#1] SMP\n Modules linked in: all of them\n CPU: 2 PID: 63 Comm: kswapd0 Not tainted 4.4.0-36-generic #55-Ubuntu\n Hardware name: System manufacturer System Product Name/P8H67-M PRO, BIOS 3904 04/27/2013\n task: ffff88040cae6040 ti: ffff880407488000 task.ti: ffff880407488000\n RIP: shadow_lru_isolate+0x181/0x190\n Call Trace:\n __list_lru_walk_one.isra.3+0x8f/0x130\n list_lru_walk_one+0x23/0x30\n scan_shadow_nodes+0x34/0x50\n shrink_slab.part.40+0x1ed/0x3d0\n shrink_zone+0x2ca/0x2e0\n kswapd+0x51e/0x990\n kthread+0xd8/0xf0\n ret_from_fork+0x3f/0x70\n\nwhich corresponds to the following sanity check in the shadow node\ntracking:\n\n BUG_ON(node-\u003ecount \u0026 RADIX_TREE_COUNT_MASK);\n\nThe workingset code tracks radix tree nodes that exclusively contain\nshadow entries of evicted pages in them, and this (somewhat obscure)\nline checks whether there are real pages left that would interfere with\nreclaim of the radix tree node under memory pressure.\n\nWhile discussing ways how fuse might sneak pages into the radix tree\npast the workingset code, Miklos pointed to replace_page_cache_page(),\nand indeed there is a problem there: it properly accounts for the old\npage being removed - __delete_from_page_cache() does that - but then\ndoes a raw raw radix_tree_insert(), not accounting for the replacement\npage. Eventually the page count bits in node-\u003ecount underflow while\nleaving the node incorrectly linked to the shadow node LRU.\n\nTo address this, make sure replace_page_cache_page() uses the tracked\npage insertion code, page_cache_tree_insert(). This fixes the page\naccounting and makes sure page-containing nodes are properly unlinked\nfrom the shadow node LRU again.\n\nAlso, make the sanity checks a bit less obscure by using the helpers for\nchecking the number of pages and shadows in a radix tree node.\n\nFixes: 449dd6984d0e (\"mm: keep page cache radix tree nodes in check\")\nLink: http://lkml.kernel.org/r/20160919155822.29498-1-hannes@cmpxchg.org\nSigned-off-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nReported-by: Antonio SJ Musumeci \u003ctrapexit@spawn.link\u003e\nDebugged-by: Miklos Szeredi \u003cmiklos@szeredi.hu\u003e\nCc: \u003cstable@vger.kernel.org\u003e\t[3.15+]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "9a2172a8d52cf14ce44e9cadd8c9df84cf832d75", "tree": "bbccf78193273c745021ffa328a4117287e0616f", "parents": [ "e3b3656ca63e23b5755183718df36fb9ff518b02" ], "author": { "name": "Javi Merino", "email": "javi.merino@arm.com", "time": "Fri Sep 30 13:14:28 2016 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 10:45:11 2016 -0700" }, "message": "MAINTAINERS: Switch to kernel.org email address for Javi Merino\n\nChange my email address to my kernel.org account instead of the ARM one.\n\nSigned-off-by: Javi Merino \u003cjavi.merino@arm.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "2fa5f04f85730d0c4f49f984b7efeb4f8d5bd1fc", "tree": "18963452531142ae6e6b4a8609374dd23c54d4ac", "parents": [ "05fb3c199bb09f5b85de56cc3ede194ac95c5e1f" ], "author": { "name": "Wanpeng Li", "email": "wanpeng.li@hotmail.com", "time": "Fri Sep 30 09:01:06 2016 +0800" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Sep 30 13:53:12 2016 +0200" }, "message": "x86/entry/64: Fix context tracking state warning when load_gs_index fails\n\nThis warning:\n\n WARNING: CPU: 0 PID: 3331 at arch/x86/entry/common.c:45 enter_from_user_mode+0x32/0x50\n CPU: 0 PID: 3331 Comm: ldt_gdt_64 Not tainted 4.8.0-rc7+ #13\n Call Trace:\n dump_stack+0x99/0xd0\n __warn+0xd1/0xf0\n warn_slowpath_null+0x1d/0x20\n enter_from_user_mode+0x32/0x50\n error_entry+0x6d/0xc0\n ? general_protection+0x12/0x30\n ? native_load_gs_index+0xd/0x20\n ? do_set_thread_area+0x19c/0x1f0\n SyS_set_thread_area+0x24/0x30\n do_int80_syscall_32+0x7c/0x220\n entry_INT80_compat+0x38/0x50\n\n... can be reproduced by running the GS testcase of the ldt_gdt test unit in\nthe x86 selftests.\n\ndo_int80_syscall_32() will call enter_form_user_mode() to convert context\ntracking state from user state to kernel state. The load_gs_index() call\ncan fail with user gsbase, gsbase will be fixed up and proceed if this\nhappen.\n\nHowever, enter_from_user_mode() will be called again in the fixed up path\nthough it is context tracking kernel state currently.\n\nThis patch fixes it by just fixing up gsbase and telling lockdep that IRQs\nare off once load_gs_index() failed with user gsbase.\n\nSigned-off-by: Wanpeng Li \u003cwanpeng.li@hotmail.com\u003e\nAcked-by: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Brian Gerst \u003cbrgerst@gmail.com\u003e\nCc: Denys Vlasenko \u003cdvlasenk@redhat.com\u003e\nCc: H. Peter Anvin \u003chpa@zytor.com\u003e\nCc: Josh Poimboeuf \u003cjpoimboe@redhat.com\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nLink: http://lkml.kernel.org/r/1475197266-3440-1-git-send-email-wanpeng.li@hotmail.com\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "05fb3c199bb09f5b85de56cc3ede194ac95c5e1f", "tree": "a77a4e4668b1f1458cf3afed17eeb3489fc24eb8", "parents": [ "e4aad64597d7a2455a541f904365b48d607916db" ], "author": { "name": "Andy Lutomirski", "email": "luto@kernel.org", "time": "Wed Sep 28 16:06:33 2016 -0700" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Sep 30 13:53:04 2016 +0200" }, "message": "x86/boot: Initialize FPU and X86_FEATURE_ALWAYS even if we don\u0027t have CPUID\n\nOtherwise arch_task_struct_size \u003d\u003d 0 and we die. While we\u0027re at it,\nset X86_FEATURE_ALWAYS, too.\n\nReported-by: David Saggiorato \u003cdavid@saggiorato.net\u003e\nTested-by: David Saggiorato \u003cdavid@saggiorato.net\u003e\nSigned-off-by: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Brian Gerst \u003cbrgerst@gmail.com\u003e\nCc: Dave Hansen \u003cdave@sr71.net\u003e\nCc: Denys Vlasenko \u003cdvlasenk@redhat.com\u003e\nCc: H. Peter Anvin \u003chpa@zytor.com\u003e\nCc: Josh Poimboeuf \u003cjpoimboe@redhat.com\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: stable@vger.kernel.org\nFixes: aaeb5c01c5b (\"x86/fpu, sched: Introduce CONFIG_ARCH_WANTS_DYNAMIC_TASK_STRUCT and use it on x86\")\nLink: http://lkml.kernel.org/r/8de723afbf0811071185039f9088733188b606c9.1475103911.git.luto@kernel.org\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "e4aad64597d7a2455a541f904365b48d607916db", "tree": "4fc661b8ab9279677a41ac905a7fb96ebba53504", "parents": [ "192d1dccbfc5b901b66527df9df80304693cf06e" ], "author": { "name": "Segher Boessenkool", "email": "segher@kernel.crashing.org", "time": "Thu Sep 29 11:51:00 2016 +0000" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Sep 30 12:37:40 2016 +0200" }, "message": "x86/vdso: Fix building on big endian host\n\nWe need to call GET_LE to read hdr-\u003ee_type.\n\nFixes: 57f90c3dfc75 (\"x86/vdso: Error out if the vDSO isn\u0027t a valid DSO\")\nReported-by: Paul Gortmaker \u003cpaul.gortmaker@windriver.com\u003e\nSigned-off-by: Segher Boessenkool \u003csegher@kernel.crashing.org\u003e\nAcked-by: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Stephen Rothwell \u003csfr@canb.auug.org.au\u003e\nCc: linux-next@vger.kernel.org\nLink: http://lkml.kernel.org/r/20160929193442.GA16617@gate.crashing.org\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n\n" }, { "commit": "192d1dccbfc5b901b66527df9df80304693cf06e", "tree": "8ee15ef8ca5f7bc9887dbd61f66c171013c4aae5", "parents": [ "e1bfc11c5a6f40222a698a818dc269113245820e" ], "author": { "name": "Andy Lutomirski", "email": "luto@kernel.org", "time": "Thu Sep 29 12:48:11 2016 -0700" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Sep 30 12:37:40 2016 +0200" }, "message": "x86/boot: Fix another __read_cr4() case on 486\n\nThe condition for reading CR4 was wrong: there are some CPUs with\nCPUID but not CR4. Rather than trying to make the condition exact,\nuse __read_cr4_safe().\n\nFixes: 18bc7bd523e0 (\"x86/boot: Synchronize trampoline_cr4_features and mmu_cr4_features directly\")\nReported-by: david@saggiorato.net\nSigned-off-by: Andy Lutomirski \u003cluto@kernel.org\u003e\nReviewed-by: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Brian Gerst \u003cbrgerst@gmail.com\u003e\nLink: http://lkml.kernel.org/r/8c453a61c4f44ab6ff43c29780ba04835234d2e5.1475178369.git.luto@kernel.org\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n\n" }, { "commit": "1cceda7849809a8857fd9f26efe8846506c710e1", "tree": "fbdd3dec3706dabe2ef7d83e8c78a1f4e224ceb3", "parents": [ "75b005b949d3dc93b526c3da0a750fd1fc9a703a" ], "author": { "name": "Xin Long", "email": "lucien.xin@gmail.com", "time": "Thu Sep 29 02:55:44 2016 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:08:57 2016 -0400" }, "message": "sctp: fix the issue sctp_diag uses lock_sock in rcu_read_lock\n\nWhen sctp dumps all the ep-\u003eassocs, it needs to lock_sock first,\nbut now it locks sock in rcu_read_lock, and lock_sock may sleep,\nwhich would break rcu_read_lock.\n\nThis patch is to get and hold one sock when traversing the list.\nAfter that and get out of rcu_read_lock, lock and dump it. Then\nit will traverse the list again to get the next one until all\nsctp socks are dumped.\n\nFor sctp_diag_dump_one, it fixes this issue by holding asoc and\nmoving cb() out of rcu_read_lock in sctp_transport_lookup_process.\n\nFixes: 8f840e47f190 (\"sctp: add the sctp_diag.c file\")\nSigned-off-by: Xin Long \u003clucien.xin@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "75b005b949d3dc93b526c3da0a750fd1fc9a703a", "tree": "859152992bd65e3da93767dbda6fbdf7e011e69c", "parents": [ "1b0ff89852d79354e8a091c81a88df21f5aa9f0a", "be4947bf46cb0e7a7d089e03c61bab35f1e695ce" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:07:10 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:07:10 2016 -0400" }, "message": "Merge branch \u0027sctp-fixes\u0027\n\nXin Long says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nsctp: a bunch of fixes for prsctp polices\n\nThis patchset is to fix 2 issues for prsctp polices:\n\n 1. patch 1 and 2 fix \"netperf-Throughput_Mbps -37.2% regression\" issue\n when overloading the CPU.\n\n 2. patch 3 fix \"prsctp polices should check both sides\u0027 prsctp_capable,\n instead of only local side\".\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": "be4947bf46cb0e7a7d089e03c61bab35f1e695ce", "tree": "859152992bd65e3da93767dbda6fbdf7e011e69c", "parents": [ "0605483f6ace1f6b63e397c819a115ddcd13af0d" ], "author": { "name": "Xin Long", "email": "lucien.xin@gmail.com", "time": "Thu Sep 29 02:37:28 2016 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:07:05 2016 -0400" }, "message": "sctp: change to check peer prsctp_capable when using prsctp polices\n\nNow before using prsctp polices, sctp uses asoc-\u003eprsctp_enable to\ncheck if prsctp is enabled. However asoc-\u003eprsctp_enable is set only\nmeans local host support prsctp, sctp should not abandon packet if\npeer host doesn\u0027t enable prsctp.\n\nSo this patch is to use asoc-\u003epeer.prsctp_capable to check if prsctp\nis enabled on both side, instead of asoc-\u003eprsctp_enable, as asoc\u0027s\npeer.prsctp_capable is set only when local and peer both enable prsctp.\n\nFixes: a6c2f792873a (\"sctp: implement prsctp TTL policy\")\nSigned-off-by: Xin Long \u003clucien.xin@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0605483f6ace1f6b63e397c819a115ddcd13af0d", "tree": "e71a84a43cb26e43cf85d066ff98e87a5245e0d1", "parents": [ "73dca124cdbad2d67d47d6196c08325f18447d07" ], "author": { "name": "Xin Long", "email": "lucien.xin@gmail.com", "time": "Thu Sep 29 02:37:27 2016 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:07:05 2016 -0400" }, "message": "sctp: remove prsctp_param from sctp_chunk\n\nNow sctp uses chunk-\u003eprsctp_param to save the prsctp param for all the\nprsctp polices, we didn\u0027t need to introduce prsctp_param to sctp_chunk.\nWe can just use chunk-\u003esinfo.sinfo_timetolive for RTX and BUF polices,\nand reuse msg-\u003eexpires_at for TTL policy, as the prsctp polices and old\nexpires policy are mutual exclusive.\n\nThis patch is to remove prsctp_param from sctp_chunk, and reuse msg\u0027s\nexpires_at for TTL and chunk\u0027s sinfo.sinfo_timetolive for RTX and BUF\npolices.\n\nNote that sctp can\u0027t use chunk\u0027s sinfo.sinfo_timetolive for TTL policy,\nas it needs a u64 variables to save the expires_at time.\n\nThis one also fixes the \"netperf-Throughput_Mbps -37.2% regression\"\nissue.\n\nFixes: a6c2f792873a (\"sctp: implement prsctp TTL policy\")\nSigned-off-by: Xin Long \u003clucien.xin@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "73dca124cdbad2d67d47d6196c08325f18447d07", "tree": "3c25a84f0e3d4e9446f644d4a4e0311a6cf0aeb4", "parents": [ "1b0ff89852d79354e8a091c81a88df21f5aa9f0a" ], "author": { "name": "Xin Long", "email": "lucien.xin@gmail.com", "time": "Thu Sep 29 02:37:26 2016 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 02:07:05 2016 -0400" }, "message": "sctp: move sent_count to the memory hole in sctp_chunk\n\nNow pahole sctp_chunk, it has 2 memory holes:\n struct sctp_chunk {\n\tstruct list_head list;\n\tatomic_t refcnt;\n\t/* XXX 4 bytes hole, try to pack */\n\t...\n\tlong unsigned int prsctp_param;\n\tint sent_count;\n\t/* XXX 4 bytes hole, try to pack */\n\nThis patch is to move up sent_count to fill the 1st one and eliminate\nthe 2nd one.\n\nIt\u0027s not just another struct compaction, it also fixes the \"netperf-\nThroughput_Mbps -37.2% regression\" issue when overloading the CPU.\n\nFixes: a6c2f792873a (\"sctp: implement prsctp TTL policy\")\nSigned-off-by: Xin Long \u003clucien.xin@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1b0ff89852d79354e8a091c81a88df21f5aa9f0a", "tree": "1bacfcf36e69e20bc8e77dc70d28f10ea7288f07", "parents": [ "7b8147aae741fedb7eab970361cd051d47bc8021" ], "author": { "name": "Milton Miller", "email": "miltonm@us.ibm.com", "time": "Thu Sep 29 13:24:08 2016 -0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 30 01:27:27 2016 -0400" }, "message": "tg3: Avoid NULL pointer dereference in tg3_io_error_detected()\n\nWhile the driver is probing the adapter, an error may occur before the\nnetdev structure is allocated and attached to pci_dev. In this case,\nnot only netdev isn\u0027t available, but the tg3 private structure is also\nnot available as it is just math from the NULL pointer, so dereferences\nmust be skipped.\n\nThe following trace is seen when the error is triggered:\n\n [1.402247] Unable to handle kernel paging request for data at address 0x00001a99\n [1.402410] Faulting instruction address: 0xc0000000007e33f8\n [1.402450] Oops: Kernel access of bad area, sig: 11 [#1]\n [1.402481] SMP NR_CPUS\u003d2048 NUMA PowerNV\n [1.402513] Modules linked in:\n [1.402545] CPU: 0 PID: 651 Comm: eehd Not tainted 4.4.0-36-generic #55-Ubuntu\n [1.402591] task: c000001fe4e42a20 ti: c000001fe4e88000 task.ti: c000001fe4e88000\n [1.402742] NIP: c0000000007e33f8 LR: c0000000007e3164 CTR: c000000000595ea0\n [1.402787] REGS: c000001fe4e8b790 TRAP: 0300 Not tainted (4.4.0-36-generic)\n [1.402832] MSR: 9000000100009033 \u003cSF,HV,EE,ME,IR,DR,RI,LE\u003e CR: 28000422 XER: 20000000\n [1.403058] CFAR: c000000000008468 DAR: 0000000000001a99 DSISR: 42000000 SOFTE: 1\n GPR00: c0000000007e3164 c000001fe4e8ba10 c0000000015c5e00 0000000000000000\n GPR04: 0000000000000001 0000000000000000 0000000000000039 0000000000000299\n GPR08: 0000000000000000 0000000000000001 c000001fe4e88000 0000000000000006\n GPR12: 0000000000000000 c00000000fb40000 c0000000000e6558 c000003ca1bffd00\n GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000\n GPR20: 0000000000000000 0000000000000000 0000000000000000 c000000000d52768\n GPR24: c000000000d52740 0000000000000100 c000003ca1b52000 0000000000000002\n GPR28: 0000000000000900 0000000000000000 c00000000152a0c0 c000003ca1b52000\n [1.404226] NIP [c0000000007e33f8] tg3_io_error_detected+0x308/0x340\n [1.404265] LR [c0000000007e3164] tg3_io_error_detected+0x74/0x340\n\nThis patch avoids the NULL pointer dereference by moving the access after\nthe netdev NULL pointer check on tg3_io_error_detected(). Also, we add a\ncheck for netdev being NULL on tg3_io_resume() [suggested by Michael Chan].\n\nFixes: 0486a063b1ff (\"tg3: prevent ifup/ifdown during PCI error recovery\")\nFixes: dfc8f370316b (\"net/tg3: Release IRQs on permanent error\")\nTested-by: Guilherme G. Piccoli \u003cgpiccoli@linux.vnet.ibm.com\u003e\nSigned-off-by: Milton Miller \u003cmiltonm@us.ibm.com\u003e\nSigned-off-by: Guilherme G. Piccoli \u003cgpiccoli@linux.vnet.ibm.com\u003e\nAcked-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e3b3656ca63e23b5755183718df36fb9ff518b02", "tree": "3e8d9954353530e785b31d9967720bb491b2a413", "parents": [ "c6169de7308b397824bd418c5b871b5a42de83d2", "90fd68dcf9a763f7e575c8467415bd8a66d073f4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 20:16:57 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 20:16:57 2016 -0700" }, "message": "Merge tag \u0027drm-fixes-for-v4.8-final\u0027 of git://people.freedesktop.org/~airlied/linux\n\nPull drm fixes from Dave Airlie:\n \"drm fixes for final 4.8.\n\n One big regression fix for udl, along with two amdgpu fixes and two\n nouveau fixes.\n\n All seems pretty safe and useful\"\n\n* tag \u0027drm-fixes-for-v4.8-final\u0027 of git://people.freedesktop.org/~airlied/linux:\n drm/udl: fix line iterator in damage handling\n drm/radeon/si/dpm: add workaround for for Jet parts\n drm/amdgpu: disable CRTCs before teardown\n drm/nouveau: Revert \"bus: remove cpu_coherent flag\"\n drm/nouveau/fifo/nv04: avoid ramht race against cookie insertion\n" }, { "commit": "c6169de7308b397824bd418c5b871b5a42de83d2", "tree": "d2dbc2041c8e222b70997d24eef26fe63ebf57dc", "parents": [ "53061afee43bc5041b67a45b6d793e7afdcf9ca7", "595c73071e6641e59b83911fbb4026e767471000" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 14:59:11 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 14:59:11 2016 -0700" }, "message": "Merge branch \u0027libnvdimm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm\n\nPull libnvdimm fixes from Dan Williams:\n\n - Four fixes for \"flush hint\" support.\n\n Flush hints are addresses advertised by the ACPI 6+ NFIT (NVDIMM\n Firmware Interface Table) that when written and fenced guarantee that\n writes pending in platform write buffers (outside the cpu) have been\n flushed to media. They might also be used by hypervisors as a\n trigger condition to flush guest-persistent memory ranges to storage.\n\n Fix a potential data corruption issue, a broken definition of the\n hint array, a wrong allocation size for the unit test implementation\n of the flush hint table, and missing NULL check in an error path.\n\n The unit test, while it did not prevent these bugs from being\n merged, at least triggered occasional crashes in advance of\n production usages.\n\n - Fix handling of ACPI DSM error status results. The DSM mechanism\n allows communication with platform and memory device firmware. We\n correctly parse known errors, but were silently ignoring others.\n\n Fix it to consistently fail any command with a non-zero status return\n that we otherwise do not interpret / handle.\n\n* \u0027libnvdimm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:\n libnvdimm, region: fix flush hint table thinko\n nfit: fail DSMs that return non-zero status by default\n libnvdimm: fix devm_nvdimm_memremap() error path\n tools/testing/nvdimm: fix allocation range for mock flush hint tables\n nvdimm: fix PHYS_PFN/PFN_PHYS mixup\n" }, { "commit": "e1bfc11c5a6f40222a698a818dc269113245820e", "tree": "b6a55b3ed54de70282b676db15103fec1efbc052", "parents": [ "8ab293e3a1376574e11f9059c09cc0db212546cb" ], "author": { "name": "Andy Lutomirski", "email": "luto@kernel.org", "time": "Wed Sep 28 12:34:14 2016 -0700" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Sep 29 19:08:30 2016 +0200" }, "message": "x86/init: Fix cr4_init_shadow() on CR4-less machines\n\ncr4_init_shadow() will panic on 486-like machines without CR4. Fix\nit using __read_cr4_safe().\n\nReported-by: david@saggiorato.net\nSigned-off-by: Andy Lutomirski \u003cluto@kernel.org\u003e\nReviewed-by: Borislav Petkov \u003cbp@suse.de\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Brian Gerst \u003cbrgerst@gmail.com\u003e\nCc: Denys Vlasenko \u003cdvlasenk@redhat.com\u003e\nCc: H. Peter Anvin \u003chpa@zytor.com\u003e\nCc: Josh Poimboeuf \u003cjpoimboe@redhat.com\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: stable@vger.kernel.org\nFixes: 1e02ce4cccdc (\"x86: Store a per-cpu shadow copy of CR4\")\nLink: http://lkml.kernel.org/r/43a20f81fb504013bf613913dc25574b45336a61.1475091074.git.luto@kernel.org\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "058effe7fdc5776b017356f690976a857eea473f", "tree": "fb493ec17acf848dcdcb2d6d4e6a09a615ef3035", "parents": [ "305723ab439e14debc1d339aa04e835d488b8253" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Sep 02 15:17:31 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Fix detection of unsupported highmem with cache aliases\n\nThe paging_init() function contains code which detects that highmem is\nin use but unsupported due to dcache aliasing. However this code was\nineffective because it was being run before the caches are probed,\nmeaning that cpu_has_dc_aliases would always evaluate to false (unless a\nplatform overrides it to a compile-time constant) and the detection of\nthe unsupported case is never triggered. The kernel would then go on to\nattempt to use highmem \u0026 either hit coherency issues or trigger the\nBUG_ON in flush_kernel_dcache_page().\n\nFix this by running paging_init() later than cpu_cache_init(), such that\nthe cpu_has_dc_aliases macro will evaluate correctly \u0026 the unsupported\nhighmem case will be detected successfully.\n\nThis then leads to a formerly hidden issue in that\nmem_init_free_highmem() will attempt to free all highmem pages, even\nthough we\u0027re avoiding use of them \u0026 don\u0027t have valid page structs for\nthem. This leads to an invalid pointer dereference \u0026 a TLB exception.\nAvoid this by skipping the loop in mem_init_free_highmem() if\ncpu_has_dc_aliases evaluates true.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: Rabin Vincent \u003crabinv@axis.com\u003e\nCc: Matt Redfearn \u003cmatt.redfearn@imgtec.com\u003e\nCc: Jerome Marchand \u003cjmarchan@redhat.com\u003e\nCc: Alexander Sverdlin \u003calexander.sverdlin@gmail.com\u003e\nCc: Aurelien Jarno \u003caurelien@aurel32.net\u003e\nCc: Jaedon Shin \u003cjaedon.shin@gmail.com\u003e\nCc: Toshi Kani \u003ctoshi.kani@hpe.com\u003e\nCc: James Hogan \u003cjames.hogan@imgtec.com\u003e\nCc: Sergey Ryazanov \u003cryazanov.s.a@gmail.com\u003e\nCc: Jonas Gorski \u003cjogo@openwrt.org\u003e\nCc: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: linux-mips@linux-mips.org\nCc: linux-kernel@vger.kernel.org\nPatchwork: https://patchwork.linux-mips.org/patch/14184/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "305723ab439e14debc1d339aa04e835d488b8253", "tree": "b9050300d840c04354748bdf580e057cff3629bf", "parents": [ "1eefcbc89cf3a8e252e5aeb25825594699b47360" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Sep 02 16:07:10 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Malta: Fix IOCU disable switch read for MIPS64\n\nMalta boards used with CPU emulators feature a switch to disable use of\nan IOCU. Software has to check this switch \u0026 ignore any present IOCU if\nthe switch is closed. The read used to do this was unsafe for 64 bit\nkernels, as it simply casted the address 0xbf403000 to a pointer \u0026\ndereferenced it. Whilst in a 32 bit kernel this would access kseg1, in a\n64 bit kernel this attempts to access xuseg \u0026 results in an address\nerror exception.\n\nFix by accessing a correctly formed ckseg1 address generated using the\nCKSEG1ADDR macro.\n\nWhilst modifying this code, define the name of the register and the bit\nwe care about within it, which indicates whether PCI DMA is routed to\nthe IOCU or straight to DRAM. The code previously checked that bit 0 was\nalso set, but the least significant 7 bits of the CONFIG_GEN0 register\ncontain the value of the MReqInfo signal provided to the IOCU OCP bus,\nso singling out bit 0 makes little sense \u0026 that part of the check is\ndropped.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nFixes: b6d92b4a6bdb (\"MIPS: Add option to disable software I/O coherency.\")\nCc: Matt Redfearn \u003cmatt.redfearn@imgtec.com\u003e\nCc: Masahiro Yamada \u003cyamada.masahiro@socionext.com\u003e\nCc: Kees Cook \u003ckeescook@chromium.org\u003e\nCc: linux-mips@linux-mips.org\nCc: linux-kernel@vger.kernel.org\nPatchwork: https://patchwork.linux-mips.org/patch/14187/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "1eefcbc89cf3a8e252e5aeb25825594699b47360", "tree": "05a846f3e7251dca3c41df792e69aca667e26f0d", "parents": [ "67acd8d5c606cf42e6726767d705851dec9f6a34" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Aug 19 18:15:40 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Fix BUILD_ROLLBACK_PROLOGUE for microMIPS\n\nWhen the kernel is built for microMIPS, branches targets need to be\nknown to be microMIPS code in order to result in bit 0 of the PC being\nset. The branch target in the BUILD_ROLLBACK_PROLOGUE macro was simply\nthe end of the macro, which may be pointing at padding rather than at\ncode. This results in recent enough GNU linkers complaining like so:\n\n mips-img-linux-gnu-ld: arch/mips/built-in.o: .text+0x3e3c: Unsupported branch between ISA modes.\n mips-img-linux-gnu-ld: final link failed: Bad value\n Makefile:936: recipe for target \u0027vmlinux\u0027 failed\n make: *** [vmlinux] Error 1\n\nFix this by changing the branch target to be the start of the\nappropriate handler, skipping over any padding.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14019/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "67acd8d5c606cf42e6726767d705851dec9f6a34", "tree": "f47a7004f00792f0c9c05590dda085bdfa2e1a72", "parents": [ "ebd0e0f503d0774407a63ebb5ec1a90bb54941f5" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Aug 19 18:18:28 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: clear execution hazard after changing FTLB enable\n\nOn current P-series cores from Imagination the FTLB can be enabled or\ndisabled via a bit in the Config6 register, and an execution hazard is\ncreated by changing the value of bit. The ftlb_disable function already\ncleared that hazard but that does no good for other callers. Clear the\nhazard in the set_ftlb_enable function that creates it, and only for the\ncores where it applies.\n\nThis has the effect of reverting c982c6d6c48b (\"MIPS: cpu-probe: Remove\ncp0 hazard barrier when enabling the FTLB\") which was incorrect.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nFixes: c982c6d6c48b (\"MIPS: cpu-probe: Remove cp0 hazard barrier when enabling the FTLB\")\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14023/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "ebd0e0f503d0774407a63ebb5ec1a90bb54941f5", "tree": "d16868a3dd85ee52fa64414f0c6062a29c5e28ec", "parents": [ "72c70f010dfcc9ea6cc13500602a29e33748452f" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Aug 19 18:18:27 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Configure FTLB after probing TLB sizes from config4\n\nOn some cores (proAptiv, P5600) we make use of the sizes of the TLBs\nto determine the desired FTLB:VTLB write ratio. However set_ftlb_enable\n\u0026 thus calculate_ftlb_probability is called before decode_config4. This\nresults in us calculating a probability based on zero sizes, and we end\nup setting FTLBP\u003d3 for a 3:1 FTLB:VTLB write ratio in all cases. This\nwill make abysmal use of the available FTLB resources in the affected\ncores.\n\nFix this by configuring the FTLB probability after having decoded\nconfig4. However we do need to have enabled the FTLB before that point\nsuch that fields in config4 actually reflect that an FTLB is present. So\nset_ftlb_enable is now called twice, with flags indicating that it\nshould configure the write probability only the second time.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nFixes: cf0a8aa0226d (\"MIPS: cpu-probe: Set the FTLB probability bit on supported cores\")\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14022/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "72c70f010dfcc9ea6cc13500602a29e33748452f", "tree": "c891259ef4181e3ebac3fc2bf7b9a85034cdc9ce", "parents": [ "3021773c7c3e75e20b693931a19362681e744ea9" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Fri Aug 19 18:18:26 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Stop setting I6400 FTLBP\n\nThe FTLBP field in Config7 for the I6400 is intended as chicken bits for\ndebugging rather than as a field that software actually makes use of.\nFor best performance, FTLBP should be left at its default value of 0\nwith all TLB writes hitting the FTLB by default.\n\nAdditionally, since set_ftlb_enable is called from decode_configs before\ndecode_config4 which determines the size of the TLBs, this was\npreviously always setting FTLBP\u003d3 for a 3:1 FTLB:VTLB write ratio which\nmakes abysmal use of the available FTLB resources.\n\nThis effectively reverts b0c4e1b79d8a (\"MIPS: Set up FTLB probability\nfor I6400\").\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nFixes: b0c4e1b79d8a (\"MIPS: Set up FTLB probability for I6400\")\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14021/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "3021773c7c3e75e20b693931a19362681e744ea9", "tree": "1d261df1423b1216268f7de2e4c61bc9952322ae", "parents": [ "0a900553715c39cfb6170ccc9846b194a4c13ceb" ], "author": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Tue Sep 20 14:33:01 2016 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: DEC: Avoid la pseudo-instruction in delay slots\n\nWhen expanding the la or dla pseudo-instruction in a delay slot the GNU\nassembler will complain should the pseudo-instruction expand to multiple\nactual instructions, since only the first of them will be in the delay\nslot leading to the pseudo-instruction being only partially executed if\nthe branch is taken. Use of PTR_LA in the dec int-handler.S leads to\nsuch warnings:\n\n arch/mips/dec/int-handler.S: Assembler messages:\n arch/mips/dec/int-handler.S:149: Warning: macro instruction expanded into multiple instructions in a branch delay slot\n arch/mips/dec/int-handler.S:198: Warning: macro instruction expanded into multiple instructions in a branch delay slot\n\nAvoid this by open coding the PTR_LA macros.\n\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "0a900553715c39cfb6170ccc9846b194a4c13ceb", "tree": "72c992170a89d3a290fd449f45283e40ee65536a", "parents": [ "ca86c9ef2b322ebf24772009fdea037688cbdac1" ], "author": { "name": "Steven J. Hill", "email": "steven.hill@cavium.com", "time": "Fri Aug 26 14:02:04 2016 -0500" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: Octeon: mark GPIO controller node not populated after IRQ init.\n\nWe clear the OF_POPULATED flag for the GPIO controller node on Octeon\nprocessors. Otherwise, none of the devices hanging on the GPIO lines\nare probed. The \u0027gpio-leds\u0027 driver on OCTEON failed to probe in addition\nto other devices on Cavium 71xx and 78xx development boards.\n\nFixes: 15cc2ed6dcf9 (\"of/irq: Mark initialised interrupt controllers as populated\")\nSigned-off-by: Steven J. Hill \u003csteven.hill@cavium.com\u003e\nTested-by: Aaro Koskinen \u003caaro.koskinen@iki.fi\u003e\nCc: David Daney \u003cdavid.daney@cavium.com\u003e\nCc: Rob Herring \u003crobh@kernel.org\u003e\nCc: linux-mips@linux-mips.org\nCc: devicetree@vger.kernel.org\nPatchwork: https://patchwork.linux-mips.org/patch/14091/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "ca86c9ef2b322ebf24772009fdea037688cbdac1", "tree": "4e75f9772e002ae39af5fd62745137961f8c1443", "parents": [ "ddabfa5c2e33f1b495f3e0176de7057850915c0b" ], "author": { "name": "Marcin Nowakowski", "email": "marcin.nowakowski@imgtec.com", "time": "Thu Sep 22 15:38:33 2016 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: uprobes: fix use of uninitialised variable\n\narch_uprobe_pre_xol needs to emulate a branch if a branch instruction\nhas been replaced with a breakpoint, but in fact an uninitialised local\nvariable was passed to the emulator routine instead of the original\ninstruction\n\nSigned-off-by: Marcin Nowakowski \u003cmarcin.nowakowski@imgtec.com\u003e\nFixes: 40e084a506eb (\u0027MIPS: Add uprobes support.\u0027)\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14300/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "ddabfa5c2e33f1b495f3e0176de7057850915c0b", "tree": "822ba923abc3cc07492ddc4d4465e74ae500fff9", "parents": [ "db06068a4fd44a57b642b369d2a295b8448f6b65" ], "author": { "name": "Marcin Nowakowski", "email": "marcin.nowakowski@imgtec.com", "time": "Thu Sep 22 15:38:32 2016 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: uprobes: remove incorrect set_orig_insn\n\nGeneric kernel code implements a weak version of set_orig_insn that\nmoves cached \u0027insn\u0027 from arch_uprobe to the original code location when\nthe trap is removed.\nMIPS variant used arch_uprobe-\u003eorig_inst which was never initialised\nproperly, so this code only inserted a nop instead of the original\ninstruction. With that change orig_inst can also be safely removed.\n\nSigned-off-by: Marcin Nowakowski \u003cmarcin.nowakowski@imgtec.com\u003e\nFixes: 40e084a506eb (\u0027MIPS: Add uprobes support.\u0027)\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14299/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "db06068a4fd44a57b642b369d2a295b8448f6b65", "tree": "92098a88b6b2874843f0638622ad27a8380d9ff0", "parents": [ "6ca8ac773e97e2dfa5734ae435c40e672dd19ac4" ], "author": { "name": "Marcin Nowakowski", "email": "marcin.nowakowski@imgtec.com", "time": "Thu Sep 22 15:38:31 2016 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: fix uretprobe implementation\n\narch_uretprobe_hijack_return_addr should replace the return address for\na call with a trampoline address.\n\nSigned-off-by: Marcin Nowakowski \u003cmarcin.nowakowski@imgtec.com\u003e\nFixes: 40e084a506eb (\u0027MIPS: Add uprobes support.\u0027)\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14298/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "6ca8ac773e97e2dfa5734ae435c40e672dd19ac4", "tree": "51296ee068824fda591eb74a9ccf8dcd038040ac", "parents": [ "08895a8b6b06ed2323cd97a36ee40a116b3db8ed" ], "author": { "name": "Matt Redfearn", "email": "matt.redfearn@imgtec.com", "time": "Thu Sep 22 11:59:47 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Sep 29 18:59:49 2016 +0200" }, "message": "MIPS: smp-cps: Avoid BUG() when offlining pre-r6 CPUs\n\nCommit 0d2808f338c7 (\"MIPS: smp-cps: Add support for CPU hotplug of\nMIPSr6 processors\") added a call to mips_cm_lock_other in order to lock\nthe CPC in CPUs containing a version 3 or higher Coherence Manager,\nwhich use the general CM core other register, where previous CMs had a\ndedicated core other register for the CPC.\n\nA kernel BUG() is triggered, however, if mips_cm_lock_other is called\nwith a VP other than 0 on a CPU with CM \u003c 3, a condition introduced by\n0d2808f338c7.\n\nAvoid the BUG() by always locking VP0 when locking the CPC, since the\nrequired register, cpc_stat_conf, is shared by all vps in a core.\n\nFixes: 0d2808f338c7 (\"MIPS: smp-cps: Add support for CPU hotplug...)\n\nSigned-off-by: Matt Redfearn \u003cmatt.redfearn@imgtec.com\u003e\nCc: Qais Yousef \u003cqsyousef@gmail.com\u003e\nCc: Masahiro Yamada \u003cyamada.masahiro@socionext.com\u003e\nCc: James Hogan \u003cjames.hogan@imgtec.com\u003e\nCc: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: linux-mips@linux-mips.org\nCc: linux-kernel@vger.kernel.org\nPatchwork: https://patchwork.linux-mips.org/patch/14297/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "d248220f0465b818887baa9829e691fe662b2c5e", "tree": "a7eba33d46ee26de6a5d6738b0ba4c870d0f3f04", "parents": [ "ba6dea4f7cedb4b1c17e36f4087675d817c2e24b" ], "author": { "name": "Roger Quadros", "email": "rogerq@ti.com", "time": "Thu Sep 29 08:32:55 2016 +0100" }, "committer": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Thu Sep 29 16:57:44 2016 +0100" }, "message": "ARM: 8617/1: dma: fix dma_max_pfn()\n\nSince commit 6ce0d2001692 (\"ARM: dma: Use dma_pfn_offset for dma address translation\"),\ndma_to_pfn() already returns the PFN with the physical memory start offset\nso we don\u0027t need to add it again.\n\nThis fixes USB mass storage lock-up problem on systems that can\u0027t do DMA\nover the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM\ncan only do DMA over the first 2GB. [K2E-EVM].\n\nWhat happens there is that without this patch SCSI layer sets a wrong\nbounce buffer limit in scsi_calculate_bounce_limit() for the USB mass\nstorage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit\nis set to 0x8fffff000 whereas maximum DMA\u0027ble physical memory on Keystone 2\nis 0x87fffffff. This results in non DMA\u0027ble pages being given to the\nUSB controller and hence the lock-up.\n\nNOTE: in the above case, USB-SCSI-device\u0027s dma_pfn_offset was showing as 0.\nThis should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000\nand HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be\n0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because\nUSB devices are not correctly inheriting the dma_pfn_offset from the\nUSB host controller. This will be fixed by a separate patch.\n\nFixes: 6ce0d2001692 (\"ARM: dma: Use dma_pfn_offset for dma address translation\")\nCc: stable@vger.kernel.org\nCc: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\nCc: Santosh Shilimkar \u003csantosh.shilimkar@oracle.com\u003e\nCc: Arnd Bergmann \u003carnd@arndb.de\u003e\nCc: Olof Johansson \u003colof@lixom.net\u003e\nCc: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\nCc: Linus Walleij \u003clinus.walleij@linaro.org\u003e\nReported-by: Grygorii Strashko \u003cgrygorii.strashko@ti.com\u003e\nSigned-off-by: Roger Quadros \u003crogerq@ti.com\u003e\nSigned-off-by: Russell King \u003crmk+kernel@arm.linux.org.uk\u003e\n" }, { "commit": "ba6dea4f7cedb4b1c17e36f4087675d817c2e24b", "tree": "9bd09872a666fa24a3b698765d15a0d44b495314", "parents": [ "1a57c286d8ced1e4144c6201a19bbb70827edee6" ], "author": { "name": "Robin Murphy", "email": "robin.murphy@arm.com", "time": "Mon Sep 26 16:50:55 2016 +0100" }, "committer": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Thu Sep 29 16:57:43 2016 +0100" }, "message": "ARM: 8616/1: dt: Respect property size when parsing CPUs\n\nWhilst MPIDR values themselves are less than 32 bits, it is still\nperfectly valid for a DT to have #address-cells \u003e 1 in the CPUs node,\nresulting in the \"reg\" property having leading zero cell(s). In that\nsituation, the big-endian nature of the data conspires with the current\nbehaviour of only reading the first cell to cause the kernel to think\nall CPUs have ID 0, and become resoundingly unhappy as a consequence.\n\nTake the full property length into account when parsing CPUs so as to\nbe correct under any circumstances.\n\nCc: Russell King \u003clinux@armlinux.org.uk\u003e\nSigned-off-by: Robin Murphy \u003crobin.murphy@arm.com\u003e\nSigned-off-by: Russell King \u003crmk+kernel@arm.linux.org.uk\u003e\n" }, { "commit": "2a0100d7beebd2fb2c436852648f75396253e9f8", "tree": "3dc90bf436b4e4c0f2496f5558ce66e2c9509aae", "parents": [ "ebb99a4c12e4daabe1940ae936e8e7e97ae68c6f" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 20:40:52 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 20:40:52 2016 -0400" }, "message": "sparc64: Fix non-SMP build.\n\nNeed to provide a dummy smp_fill_in_cpu_possible_map.\n\nFixes: 9b2f753ec237 (\"sparc64: Fix cpu_possible_mask if nr_cpus is set\")\nReported-by: kbuild test robot \u003cfengguang.wu@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "53061afee43bc5041b67a45b6d793e7afdcf9ca7", "tree": "ed7e918bd7323ded4bdd899aded6384488dad9a4", "parents": [ "ae6dd8d61913b5341067c7c660e8e91430d22e7c", "231e97e2b8ec9a1556ced5d8a89cda03a480b179" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:20:24 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:20:24 2016 -0700" }, "message": "Merge branch \u0027akpm\u0027 (patches from Andrew)\n\nMerge fixes from Andrew Morton:\n \"4 fixes\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e:\n mem-hotplug: use nodes that contain memory as mask in new_node_page()\n scripts/recordmcount.c: account for .softirqentry.text\n dma-mapping.h: preserve unmap info for CONFIG_DMA_API_DEBUG\n mm,ksm: fix endless looping in allocating memory when ksm enable\n" }, { "commit": "231e97e2b8ec9a1556ced5d8a89cda03a480b179", "tree": "762b3c2cb8698d59ea3a4059b509895afb2f0fe6", "parents": [ "e436fd61a8f62cb7a16310a42b95ab076ff72eff" ], "author": { "name": "Li Zhong", "email": "zhong@linux.vnet.ibm.com", "time": "Wed Sep 28 15:22:38 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:19:02 2016 -0700" }, "message": "mem-hotplug: use nodes that contain memory as mask in new_node_page()\n\n9bb627be47a5 (\"mem-hotplug: don\u0027t clear the only node in new_node_page()\")\nprevents allocating from an empty nodemask, but as David points out, it is\nstill wrong. As node_online_map may include memoryless nodes, only\nallocating from these nodes is meaningless.\n\nThis patch uses node_states[N_MEMORY] mask to prevent the above case.\n\nFixes: 9bb627be47a5 (\"mem-hotplug: don\u0027t clear the only node in new_node_page()\")\nFixes: 394e31d2ceb4 (\"mem-hotplug: alloc new page from a nearest neighbor node when mem-offline\")\nLink: http://lkml.kernel.org/r/1474447117.28370.6.camel@TP420\nSigned-off-by: Li Zhong \u003czhong@linux.vnet.ibm.com\u003e\nSuggested-by: David Rientjes \u003crientjes@google.com\u003e\nAcked-by: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nCc: Michal Hocko \u003cmhocko@suse.cz\u003e\nCc: John Allen \u003cjallen@linux.vnet.ibm.com\u003e\nCc: Xishi Qiu \u003cqiuxishi@huawei.com\u003e\nCc: Joonsoo Kim \u003ciamjoonsoo.kim@lge.com\u003e\nCc: Naoya Horiguchi \u003cn-horiguchi@ah.jp.nec.com\u003e\nCc: Tetsuo Handa \u003cpenguin-kernel@i-love.sakura.ne.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "e436fd61a8f62cb7a16310a42b95ab076ff72eff", "tree": "7b835aff31c20c8ac8321a2ce45f82dad6d8f1c6", "parents": [ "2481366afd71a0c0b7cd725e6750c04cf589673b" ], "author": { "name": "Dmitry Vyukov", "email": "dvyukov@google.com", "time": "Wed Sep 28 15:22:36 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:19:02 2016 -0700" }, "message": "scripts/recordmcount.c: account for .softirqentry.text\n\nbe7635e7287e (\"arch, ftrace: for KASAN put hard/soft IRQ entries into\nseparate sections\") added .softirqentry.text section, but it was not added\nto recordmcount. So functions in the section are untracable. Add the\nsection to scripts/recordmcount.c and scripts/recordmcount.pl.\n\nFixes: be7635e7287e (\"arch, ftrace: for KASAN put hard/soft IRQ entries into separate sections\")\nLink: http://lkml.kernel.org/r/1474902626-73468-1-git-send-email-dvyukov@google.com\nSigned-off-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nAcked-by: Steve Rostedt \u003crostedt@goodmis.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\t[4.6+]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "2481366afd71a0c0b7cd725e6750c04cf589673b", "tree": "bddefbe87e77de1c6d7d5d6e77ea85155973b45e", "parents": [ "5b398e416e880159fe55eefd93c6588fa072cd66" ], "author": { "name": "Andrey Smirnov", "email": "andrew.smirnov@gmail.com", "time": "Wed Sep 28 15:22:33 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:19:01 2016 -0700" }, "message": "dma-mapping.h: preserve unmap info for CONFIG_DMA_API_DEBUG\n\nWhen CONFIG_DMA_API_DEBUG is enabled we need to preserve unmapping address\neven if \"unmap\" is a no-op for our architecutre because we need\ndebug_dma_unmap_page() to correctly cleanup all of the debug bookkeeping.\nFailing to do so results in a false positive warnings about previously\nmapped areas never being unmapped.\n\nLink: http://lkml.kernel.org/r/1474387125-3713-1-git-send-email-andrew.smirnov@gmail.com\nSigned-off-by: Andrey Smirnov \u003candrew.smirnov@gmail.com\u003e\nReviewed-by: Robin Murphy \u003crobin.murphy@arm.com\u003e\nCc: Joerg Roedel \u003cjroedel@suse.de\u003e\nCc: Will Deacon \u003cwill.deacon@arm.com\u003e\nCc: Zhen Lei \u003cthunder.leizhen@huawei.com\u003e\nCc: \"Luis R. Rodriguez\" \u003cmcgrof@suse.com\u003e\nCc: Christian Borntraeger \u003cborntraeger@de.ibm.com\u003e\nCc: Geliang Tang \u003cgeliangtang@163.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5b398e416e880159fe55eefd93c6588fa072cd66", "tree": "9e3f7eb51383127271824518c0edca9d50c34c0d", "parents": [ "8ab293e3a1376574e11f9059c09cc0db212546cb" ], "author": { "name": "zhong jiang", "email": "zhongjiang@huawei.com", "time": "Wed Sep 28 15:22:30 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 16:19:01 2016 -0700" }, "message": "mm,ksm: fix endless looping in allocating memory when ksm enable\n\nI hit the following hung task when runing a OOM LTP test case with 4.1\nkernel.\n\nCall trace:\n[\u003cffffffc000086a88\u003e] __switch_to+0x74/0x8c\n[\u003cffffffc000a1bae0\u003e] __schedule+0x23c/0x7bc\n[\u003cffffffc000a1c09c\u003e] schedule+0x3c/0x94\n[\u003cffffffc000a1eb84\u003e] rwsem_down_write_failed+0x214/0x350\n[\u003cffffffc000a1e32c\u003e] down_write+0x64/0x80\n[\u003cffffffc00021f794\u003e] __ksm_exit+0x90/0x19c\n[\u003cffffffc0000be650\u003e] mmput+0x118/0x11c\n[\u003cffffffc0000c3ec4\u003e] do_exit+0x2dc/0xa74\n[\u003cffffffc0000c46f8\u003e] do_group_exit+0x4c/0xe4\n[\u003cffffffc0000d0f34\u003e] get_signal+0x444/0x5e0\n[\u003cffffffc000089fcc\u003e] do_signal+0x1d8/0x450\n[\u003cffffffc00008a35c\u003e] do_notify_resume+0x70/0x78\n\nThe oom victim cannot terminate because it needs to take mmap_sem for\nwrite while the lock is held by ksmd for read which loops in the page\nallocator\n\nksm_do_scan\n\tscan_get_next_rmap_item\n\t\tdown_read\n\t\tget_next_rmap_item\n\t\t\talloc_rmap_item #ksmd will loop permanently.\n\nThere is no way forward because the oom victim cannot release any memory\nin 4.1 based kernel. Since 4.6 we have the oom reaper which would solve\nthis problem because it would release the memory asynchronously.\nNevertheless we can relax alloc_rmap_item requirements and use\n__GFP_NORETRY because the allocation failure is acceptable as ksm_do_scan\nwould just retry later after the lock got dropped.\n\nSuch a patch would be also easy to backport to older stable kernels which\ndo not have oom_reaper.\n\nWhile we are at it add GFP_NOWARN so the admin doesn\u0027t have to be alarmed\nby the allocation failure.\n\nLink: http://lkml.kernel.org/r/1474165570-44398-1-git-send-email-zhongjiang@huawei.com\nSigned-off-by: zhong jiang \u003czhongjiang@huawei.com\u003e\nSuggested-by: Hugh Dickins \u003chughd@google.com\u003e\nSuggested-by: Michal Hocko \u003cmhocko@suse.cz\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.com\u003e\nAcked-by: Hugh Dickins \u003chughd@google.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": "ae6dd8d61913b5341067c7c660e8e91430d22e7c", "tree": "51636068899cf37b2cc830611ef819124dae4bd9", "parents": [ "0a966fa8914aafeb430f9e8489227e86d8468625", "f6d7c1b5598b6407c3f1da795dd54acf99c1990c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 12:53:08 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 12:53:08 2016 -0700" }, "message": "Merge tag \u0027for-linus-20160928\u0027 of git://git.infradead.org/linux-mtd\n\nPull late MTD fixes from Brian Norris:\n \"Another round of MTD fixes for v4.8\n\n My apologies for sending this so late. I\u0027ve been fairly absent as a\n maintainer this cycle, but I did queue these up weeks ago. In the\n meantime, Richard was able to handle some other fixes (thanks!) but\n didn\u0027t pick these up.\n\n On the bright side, these are very simple changes that should carry\n little risk.\n\n Summary:\n\n - Davinci NAND: fix a long-standing bug in how we clear/prep 4-bit ECC\n\n - OMAP NAND: an error-handling fix that made it into v4.8-rc1 caused\n error-handling cases in other configurations/code-paths; this fixes\n the fix\"\n\n* tag \u0027for-linus-20160928\u0027 of git://git.infradead.org/linux-mtd:\n mtd: nand: davinci: Reinitialize the HW ECC engine in 4bit hwctl\n mtd: nand: omap2: Don\u0027t call dma_release_channel() if dma_request_chan() failed\n" }, { "commit": "0a966fa8914aafeb430f9e8489227e86d8468625", "tree": "b347142146b2334e664283c7f92ddc92a381c2dd", "parents": [ "8ab293e3a1376574e11f9059c09cc0db212546cb" ], "author": { "name": "Mark Fasheh", "email": "mfasheh@suse.de", "time": "Wed Sep 28 12:51:04 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 12:52:05 2016 -0700" }, "message": "MAINTAINERS: Update my e-mail\n\nI will be starting employment at Versity next week and would like to update\nmy MAINTAINERS e-mail to reflect that change. My versity e-mail is already\nactivated so I shouldn\u0027t get any bounces on the new one. My ability to help\nwith Ocfs2 kernel maintenance won\u0027t change as a result of the new job.\n\nSigned-off-by: Mark Fasheh \u003cmfasheh@versity.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ebb99a4c12e4daabe1940ae936e8e7e97ae68c6f", "tree": "e857fbe7a91bb78022da0b9c1abee53ffa06c6df", "parents": [ "9b2f753ec23710aa32c0d837d2499db92fe9115b" ], "author": { "name": "Atish Patra", "email": "atish.patra@oracle.com", "time": "Thu Sep 15 14:54:41 2016 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 08:24:03 2016 -0700" }, "message": "sparc64: Fix irq stack bootmem allocation.\n\nCurrently, irq stack bootmem is allocated for all possible cpus\nbefore nr_cpus value changes the list of possible cpus. As a result,\nthere is unnecessary wastage of bootmemory.\n\nMove the irq stack bootmem allocation so that it happens after\npossible cpu list is modified based on nr_cpus value.\n\nSigned-off-by: Atish Patra \u003catish.patra@oracle.com\u003e\nReviewed-by: Bob Picco \u003cbob.picco@oracle.com\u003e\nReviewed-by: Vijay Kumar \u003cvijay.ac.kumar@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9b2f753ec23710aa32c0d837d2499db92fe9115b", "tree": "3fe8894d45a41028a16506e98286c974e954942a", "parents": [ "1e953d846ac015fbfcf09c857e8f893924cb629c" ], "author": { "name": "Atish Patra", "email": "atish.patra@oracle.com", "time": "Thu Sep 15 14:54:40 2016 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 08:24:02 2016 -0700" }, "message": "sparc64: Fix cpu_possible_mask if nr_cpus is set\n\nIf kernel boot parameter nr_cpus is set, it should define the number\nof CPUs that can ever be available in the system i.e.\ncpu_possible_mask. setup_nr_cpu_ids() overrides the nr_cpu_ids based\non the cpu_possible_mask during kernel initialization. If\ncpu_possible_mask is not set based on the nr_cpus value, earlier part\nof the kernel would be initialized using nr_cpus value leading to a\nkernel crash.\n\nSet cpu_possible_mask based on nr_cpus value. Thus setup_nr_cpu_ids()\nbecomes redundant and does not corrupt nr_cpu_ids value.\n\nSigned-off-by: Atish Patra \u003catish.patra@oracle.com\u003e\nReviewed-by: Bob Picco \u003cbob.picco@oracle.com\u003e\nReviewed-by: Vijay Kumar \u003cvijay.ac.kumar@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1e953d846ac015fbfcf09c857e8f893924cb629c", "tree": "d476520c50e49a4d19ec25816ec6486d1642b126", "parents": [ "bdf2f59e64eb9cd9e9dd90f990a9577640470c8a" ], "author": { "name": "Mike Kravetz", "email": "mike.kravetz@oracle.com", "time": "Wed Aug 31 13:48:19 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 08:24:02 2016 -0700" }, "message": "sparc64 mm: Fix more TSB sizing issues\n\nCommit af1b1a9b36b8 (\"sparc64 mm: Fix base TSB sizing when hugetlb\npages are used\") addressed the difference between hugetlb and THP\npages when computing TSB sizes. The following additional issues\nwere also discovered while working with the code.\n\nIn order to save memory, THP makes use of a huge zero page. This huge\nzero page does not count against a task\u0027s RSS, but it does consume TSB\nentries. This is similar to hugetlb pages. Therefore, count huge\nzero page entries in hugetlb_pte_count.\n\nAccounting of THP pages is done in the routine set_pmd_at().\nUnfortunately, this does not catch the case where a THP page is split.\nTo handle this case, decrement the count in pmdp_invalidate().\npmdp_invalidate is only called when splitting a THP. However, \u0027sanity\nchecks\u0027 are added in case it is ever called for other purposes.\n\nA more general issue exists with HPAGE_SIZE accounting.\nhugetlb_pte_count tracks the number of HPAGE_SIZE (8M) pages. This\nvalue is used to size the TSB for HPAGE_SIZE pages. However,\neach HPAGE_SIZE page consists of two REAL_HPAGE_SIZE (4M) pages.\nThe TSB contains an entry for each REAL_HPAGE_SIZE page. Therefore,\nthe number of REAL_HPAGE_SIZE pages should be used to size the huge\npage TSB. A new compile time constant REAL_HPAGE_PER_HPAGE is used\nto multiply hugetlb_pte_count before sizing the TSB.\n\nChanges from V1\n- Fixed build issue if hugetlb or THP not configured\n\nSigned-off-by: Mike Kravetz \u003cmike.kravetz@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bdf2f59e64eb9cd9e9dd90f990a9577640470c8a", "tree": "6d49d369ce65d1e184c28f9a5b172e56a88386c8", "parents": [ "8ab293e3a1376574e11f9059c09cc0db212546cb" ], "author": { "name": "Paul Gortmaker", "email": "paul.gortmaker@windriver.com", "time": "Sat Aug 06 00:31:48 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Sep 28 08:24:02 2016 -0700" }, "message": "sparc64: fix section mismatch in find_numa_latencies_for_group\n\nTo fix:\n\n WARNING: vmlinux.o(.text.unlikely+0x580): Section mismatch in\n reference from the function find_numa_latencies_for_group() to the\n function .init.text:find_mlgroup()\n\n The function find_numa_latencies_for_group() references the\n function __init find_mlgroup(). This is often because\n find_numa_latencies_for_group lacks a __init annotation or the\n annotation of find_mlgroup is wrong.\n\nIt turns out find_numa_latencies_for_group is only called from:\n static int __init numa_parse_mdesc(void)\nand hence we can tag find_numa_latencies_for_group with __init.\n\nIn doing so we see that find_best_numa_node_for_mlgroup is only\ncalled from within __init and hence can also be marked with __init.\n\nCc: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCc: Nitin Gupta \u003cnitin.m.gupta@oracle.com\u003e\nCc: Chris Hyser \u003cchris.hyser@oracle.com\u003e\nCc: Santosh Shilimkar \u003csantosh.shilimkar@oracle.com\u003e\nCc: sparclinux@vger.kernel.org\nSigned-off-by: Paul Gortmaker \u003cpaul.gortmaker@windriver.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "539294b76af8922297702a7ebb8cafe68f7e5376", "tree": "eac3ae2ecac2b358b6151ad6a62261490ec5a211", "parents": [ "14a5916e0cce6dbc9acedc99576261c6568689d2", "8d58881b995904bf8b150dae69be0829f832e7be" ], "author": { "name": "James Bottomley", "email": "James.Bottomley@HansenPartnership.com", "time": "Tue Sep 27 22:30:51 2016 -0700" }, "committer": { "name": "James Bottomley", "email": "James.Bottomley@HansenPartnership.com", "time": "Tue Sep 27 22:30:51 2016 -0700" }, "message": "Merge remote-tracking branch \u0027mkp-scsi/4.8/scsi-fixes\u0027 into fixes\n" }, { "commit": "90fd68dcf9a763f7e575c8467415bd8a66d073f4", "tree": "862590353f693cb4948bd3db8bf5569f4d4a3316", "parents": [ "aaee1d1e2d97e3cb99cf0e096d2172237d762e4e" ], "author": { "name": "David Herrmann", "email": "dh.herrmann@gmail.com", "time": "Fri Sep 23 12:36:02 2016 +0200" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 28 13:29:18 2016 +1000" }, "message": "drm/udl: fix line iterator in damage handling\n\nThe udl damage handler is supposed to render \u0027height\u0027 lines, but its\niterator has an obvious typo that makes it miss most lines if the\nrectangle does not cover 0/0.\n\nFix the damage handler to correctly render all lines.\n\nThis is a fallout from:\n\n commit e375882406d0cc24030746638592004755ed4ae0\n Author: Noralf Trønnes \u003cnoralf@tronnes.org\u003e\n Date: Thu Apr 28 17:18:37 2016 +0200\n\n drm/udl: Use drm_fb_helper deferred_io support\n\nTested-by: poma \u003cpoma@gmail.com\u003e\nCc: stable@vger.kernel.org # 4.7+\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nSigned-off-by: David Herrmann \u003cdh.herrmann@gmail.com\u003e\nReviewed-by: Eric Engestrom \u003ceric.engestrom@imgtec.com\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "aaee1d1e2d97e3cb99cf0e096d2172237d762e4e", "tree": "b157e95fcba896ec4c6b4e73322bb9e8ab32fb94", "parents": [ "b86f9faa34b4f08f4fbca333648d999bfcf42192", "bad3d80fd001d3b74ab8bb3d561bc4d1b08797d3" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 28 10:23:50 2016 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 28 10:23:50 2016 +1000" }, "message": "Merge branch \u0027linux-4.8\u0027 of git://github.com/skeggsb/linux into drm-fixes\n\nnouveau: couple of fixes.\n\n* \u0027linux-4.8\u0027 of git://github.com/skeggsb/linux:\n drm/nouveau: Revert \"bus: remove cpu_coherent flag\"\n drm/nouveau/fifo/nv04: avoid ramht race against cookie insertion\n" }, { "commit": "b86f9faa34b4f08f4fbca333648d999bfcf42192", "tree": "5048bc17e2cff8fe10dca77d8d6348c43f35d4fb", "parents": [ "08895a8b6b06ed2323cd97a36ee40a116b3db8ed", "670bb4fd21c966d0d2a59ad4a99bb4889f9a2987" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 28 10:19:35 2016 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 28 10:19:35 2016 +1000" }, "message": "Merge branch \u0027drm-fixes-4.8\u0027 of git://people.freedesktop.org/~agd5f/linux into drm-fixes\n\ntwo amd fixes.\n\n* \u0027drm-fixes-4.8\u0027 of git://people.freedesktop.org/~agd5f/linux:\n drm/radeon/si/dpm: add workaround for for Jet parts\n drm/amdgpu: disable CRTCs before teardown\n" }, { "commit": "8ab293e3a1376574e11f9059c09cc0db212546cb", "tree": "ab3868ca55ae136df4deb8d70f78faf14b961f90", "parents": [ "08895a8b6b06ed2323cd97a36ee40a116b3db8ed", "9157056da8f8c4a6305f15619e269f164b63a6de" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 16:43:11 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 16:43:11 2016 -0700" }, "message": "Merge branch \u0027for-4.8-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup\n\nPull cgroup fixes from Tejun Heo:\n \"Three late fixes for cgroup: Two cpuset ones, one trivial and the\n other pretty obscure, and a cgroup core fix for a bug which impacts\n cgroup v2 namespace users\"\n\n* \u0027for-4.8-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:\n cgroup: fix invalid controller enable rejections with cgroup namespace\n cpuset: fix non static symbol warning\n cpuset: handle race between CPU hotplug and cpuset_hotplug_work\n" }, { "commit": "670bb4fd21c966d0d2a59ad4a99bb4889f9a2987", "tree": "b648149857bf721db6d0a4dff0265af07bbdb856", "parents": [ "a951ed85abd4615e98e36b536e3b3b07b22a88ac" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Sep 26 15:32:50 2016 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 12:22:29 2016 -0400" }, "message": "drm/radeon/si/dpm: add workaround for for Jet parts\n\nAdd clock quirks for Jet parts.\n\nReviewed-by: Sonny Jiang \u003csonny.jiang@amd.com\u003e\nTested-by: Sonny Jiang \u003csonny.jiang@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "a951ed85abd4615e98e36b536e3b3b07b22a88ac", "tree": "be39fc799fb6342242dbeebacd2d0688d2874979", "parents": [ "47a66e45d7a7613322549c2475ea9d809baaf514" ], "author": { "name": "Grazvydas Ignotas", "email": "notasas@gmail.com", "time": "Sun Sep 25 23:34:48 2016 +0300" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 12:22:21 2016 -0400" }, "message": "drm/amdgpu: disable CRTCs before teardown\n\nSome code called by drm_crtc_force_disable_all() wants to wait for all\nfences, so only do fence teardown after CRTCs are disabled.\n\nFixes: 84b89bdcedf8 (\"drm/amdgpu: Turn off CRTCs on driver unload\")\nCc: stable@vger.kernel.org # v4.8+\nSigned-off-by: Grazvydas Ignotas \u003cnotasas@gmail.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "7b8147aae741fedb7eab970361cd051d47bc8021", "tree": "eb6f82ecd60bbcea3c16a16809b2272b9147db59", "parents": [ "1190cfdb1a19d89561ae51cff7d9c2ead24b3ebe", "c006da0be033b6ddcd27ee603d0ee01491236642" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 09:53:30 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 09:53:30 2016 -0400" }, "message": "Merge branch \u0027act_ife-fixes\u0027\n\nYotam Gigi says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nFix tc-ife bugs\n\nThis patch-set contains two bugfixes in the tc-ife action, one fixing some\nrandom behaviour in encode side, and one fixing the decode side packet\nparsing logic.\n\nv2-\u003ev3\n - Fix the encode side instead of the decode side\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": "c006da0be033b6ddcd27ee603d0ee01491236642", "tree": "eb6f82ecd60bbcea3c16a16809b2272b9147db59", "parents": [ "4b1d488a285a446329825d5fe91f987b7880e6e5" ], "author": { "name": "Yotam Gigi", "email": "yotam.gi@gmail.com", "time": "Mon Sep 26 13:45:26 2016 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 09:53:17 2016 -0400" }, "message": "act_ife: Fix false encoding\n\nOn ife encode side, the action stores the different tlvs inside the ife\nheader, where each tlv length field should refer to the length of the\nwhole tlv (without additional padding) and not just the data length.\n\nOn ife decode side, the action iterates over the tlvs in the ife header\nand parses them one by one, where in each iteration the current pointer is\nadvanced according to the tlv size.\n\nBefore, the encoding encoded only the data length inside the tlv, which led\nto false parsing of ife the header. In addition, due to the fact that the\nloop counter was unsigned, it could lead to infinite parsing loop.\n\nThis fix changes the loop counter to be signed and fixes the encoding to\ntake into account the tlv type and size.\n\nFixes: 28a10c426e81 (\"net sched: fix encoding to use real length\")\nAcked-by: Jamal Hadi Salim \u003cjhs@mojatatu.com\u003e\nSigned-off-by: Yotam Gigi \u003cyotamg@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4b1d488a285a446329825d5fe91f987b7880e6e5", "tree": "b2ef6d341e8eac085aa828f66ecc08a22979ed17", "parents": [ "1190cfdb1a19d89561ae51cff7d9c2ead24b3ebe" ], "author": { "name": "Yotam Gigi", "email": "yotam.gi@gmail.com", "time": "Mon Sep 26 13:45:25 2016 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 09:53:16 2016 -0400" }, "message": "act_ife: Fix external mac header on encode\n\nOn ife encode side, external mac header is copied from the original packet\nand may be overridden if the user requests. Before, the mac header copy\nwas done from memory region that might not be accessible anymore, as\nskb_cow_head might free it and copy the packet. This led to random values\nin the external mac header once the values were not set by user.\n\nThis fix takes the internal mac header from the packet, after the call to\nskb_cow_head.\n\nFixes: ef6980b6becb (\"net sched: introduce IFE action\")\nAcked-by: Jamal Hadi Salim \u003cjhs@mojatatu.com\u003e\nSigned-off-by: Yotam Gigi \u003cyotamg@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1190cfdb1a19d89561ae51cff7d9c2ead24b3ebe", "tree": "c7c3e0042bd4ab2e75eb98a275788bd59cf1ac57", "parents": [ "bf1a85a8381a0f749aa321d7881b456b36eb398d" ], "author": { "name": "Jorgen Hansen", "email": "jhansen@vmware.com", "time": "Mon Sep 26 23:59:53 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:59:25 2016 -0400" }, "message": "VSOCK: Don\u0027t dec ack backlog twice for rejected connections\n\nIf a pending socket is marked as rejected, we will decrease the\nsk_ack_backlog twice. So don\u0027t decrement it for rejected sockets\nin vsock_pending_work().\n\nTesting of the rejected socket path was done through code\nmodifications.\n\nReported-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nSigned-off-by: Jorgen Hansen \u003cjhansen@vmware.com\u003e\nReviewed-by: Adit Ranadive \u003caditr@vmware.com\u003e\nReviewed-by: Aditya Sarwade \u003casarwade@vmware.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bf1a85a8381a0f749aa321d7881b456b36eb398d", "tree": "d5bcf71835c398805464c083e2cf7a52668ed646", "parents": [ "6c1394f30bae67d5dd8648a38679876ad4afc4f1" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Sat Sep 24 12:58:30 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:41:12 2016 -0400" }, "message": "Revert \"net: ethernet: bcmgenet: use phydev from struct net_device\"\n\nThis reverts commit 62469c76007e (\"net: ethernet: bcmgenet: use phydev\nfrom struct net_device\") because it causes GENETv1/2/3 adapters to\nexpose the following behavior after an ifconfig down/up sequence:\n\nPING fainelli-linux (10.112.156.244): 56 data bytes\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.352 ms\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.472 ms (DUP!)\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.496 ms (DUP!)\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.517 ms (DUP!)\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.536 ms (DUP!)\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d1.557 ms (DUP!)\n64 bytes from 10.112.156.244: seq\u003d1 ttl\u003d61 time\u003d752.448 ms (DUP!)\n\nThis was previously fixed by commit 5dbebbb44a6a (\"net: bcmgenet:\nSoftware reset EPHY after power on\") but the commit we are reverting was\nessentially making this previous commit void, here is why.\n\nWithout commit 62469c76007e we would have the following scenario after\nan ifconfig down then up sequence:\n\n- bcmgenet_open() calls bcmgenet_power_up() to make sure the PHY is\n initialized *before* we get to initialize the UniMAC, this is\n critical to ensure the PHY is in a correct state, priv-\u003ephydev is\n valid, this code executes fine\n\n- second time from bcmgenet_mii_probe(), through the normal\n phy_init_hw() call (which arguably could be optimized out)\n\nEverything is fine in that case. With commit 62469c76007e, we would have\nthe following scenario to happen after an ifconfig down then up\nsequence:\n\n- bcmgenet_close() calls phy_disonnect() which makes dev-\u003ephydev become\n NULL\n\n- when bcmgenet_open() executes again and calls bcmgenet_mii_reset() from\n bcmgenet_power_up() to initialize the internal PHY, the NULL check\n becomes true, so we do not reset the PHY, yet we keep going on and\n initialize the UniMAC, causing MAC activity to occur\n\n- we call bcmgenet_mii_reset() from bcmgenet_mii_probe(), but this is\n too late, the PHY is botched, and causes the above bogus pings/packets\n transmission/reception to occur\n\nReported-by: Jaedon Shin \u003cjaedon.shin@gmail.com\u003e\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6c1394f30bae67d5dd8648a38679876ad4afc4f1", "tree": "daf1ca189ea3eb820b243e1752a6be19bdbdfbfb", "parents": [ "2cf750704bb6d7ed8c7d732e071dd1bc890ea5e8", "3ac72b7b63d57b231dda1e8f8d13872e0d7e8603" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:39:38 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:40:19 2016 -0400" }, "message": "Merge branch \u0027fec-align\u0027\n\nEric Nelson says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet: fec: updates to align IP header\n\nThis patch series is the outcome of investigation into very high\nnumbers of alignment faults on kernel 4.1.33 from the linux-fslc\ntree:\n https://github.com/freescale/linux-fslc/tree/4.1-1.0.x-imx\n\nThe first two patches remove support for the receive accelerator (RACC) from\nthe i.MX25 and i.MX27 SoCs which don\u0027t support the function.\n\nThe third patch enables hardware alignment of the ethernet packet payload\n(and especially the IP header) to prevent alignment faults in the IP stack.\n\nTesting on i.MX6UL on the 4.1.33 kernel showed that this patch removed\non the order of 70k alignment faults during a 100MiB transfer using\nwget.\n\nTesting on an i.MX6Q (SABRE Lite) board on net-next (4.8.0-rc7) showed\na much more modest improvement from 10\u0027s of faults, and it\u0027s not clear\nwhy that\u0027s the case.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nAcked-by: Fugang Duan \u003cfugang.duan@nxp.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3ac72b7b63d57b231dda1e8f8d13872e0d7e8603", "tree": "daf1ca189ea3eb820b243e1752a6be19bdbdfbfb", "parents": [ "97dc499c1ae396bc0d529f5b1aef3138de1ae7de" ], "author": { "name": "Eric Nelson", "email": "eric@nelint.com", "time": "Sat Sep 24 07:42:19 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:39:34 2016 -0400" }, "message": "net: fec: align IP header in hardware\n\nThe FEC receive accelerator (RACC) supports shifting the data payload of\nreceived packets by 16-bits, which aligns the payload (IP header) on a\n4-byte boundary, which is, if not required, at least strongly suggested\nby the Linux networking layer.\n\nWithout this patch, a huge number of alignment faults will be taken by the\nIP stack, as seen in /proc/cpu/alignment:\n\n\t~/$ cat /proc/cpu/alignment\n\tUser:\t\t0\n\tSystem:\t\t72645 (inet_gro_receive+0x104/0x27c)\n\tSkipped:\t0\n\tHalf:\t\t0\n\tWord:\t\t0\n\tDWord:\t\t0\n\tMulti:\t\t72645\n\tUser faults:\t3 (fixup+warn)\n\nThis patch was suggested by Andrew Lunn in this message to linux-netdev:\n\thttp://marc.info/?l\u003dlinux-arm-kernel\u0026m\u003d147465452108384\u0026w\u003d2\n\nand adapted from a patch by Russell King from 2014:\n\thttp://git.arm.linux.org.uk/cgit/linux-arm.git/commit/?id\u003d70d8a8a\n\nSigned-off-by: Eric Nelson \u003ceric@nelint.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "97dc499c1ae396bc0d529f5b1aef3138de1ae7de", "tree": "c365067e8d52ca506ef841ad8e303cd45eadbf8d", "parents": [ "653d37d8bca0be2c7c53c8578980b06840482427" ], "author": { "name": "Eric Nelson", "email": "eric@nelint.com", "time": "Sat Sep 24 07:42:18 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:39:34 2016 -0400" }, "message": "net: fec: remove QUIRK_HAS_RACC from i.mx27\n\nAccording to the i.MX27 reference manual, this SoC does not have support\nfor the receive accelerator (RACC) register at offset 0x1C4.\n\n\thttp://cache.nxp.com/files/32bit/doc/ref_manual/MCIMX27RM.pdf\n\nSigned-off-by: Eric Nelson \u003ceric@nelint.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "653d37d8bca0be2c7c53c8578980b06840482427", "tree": "d47180bc6b4709da6e7fe36271725d50f8e5b24b", "parents": [ "2cf750704bb6d7ed8c7d732e071dd1bc890ea5e8" ], "author": { "name": "Eric Nelson", "email": "eric@nelint.com", "time": "Sat Sep 24 07:42:17 2016 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Sep 27 07:39:33 2016 -0400" }, "message": "net: fec: remove QUIRK_HAS_RACC from i.mx25\n\nAccording to the i.MX25 reference manual, this SoC does not have support\nfor the receive accelerator (RACC) register at offset 0x1C4.\n\nhttp://www.nxp.com/files/dsp/doc/ref_manual/IMX25RM.pdf\n\nSigned-off-by: Eric Nelson \u003ceric@nelint.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9fb6de1b0bf4ec11573b76059da3c3b39ac7f2ff", "tree": "ef62417c2f3210a0724d31700b3adc5f222dc1a2", "parents": [ "4af2ff91ec3f42b538a65cf12df5f9faf6aaa914" ], "author": { "name": "Ville Ranki", "email": "ville.ranki@iki.fi", "time": "Thu Sep 15 15:31:27 2016 -0700" }, "committer": { "name": "Dmitry Torokhov", "email": "dmitry.torokhov@gmail.com", "time": "Mon Sep 26 20:03:06 2016 -0700" }, "message": "Input: joydev - recognize devices with Z axis as joysticks\n\nCurrent implementation of joydev\u0027s input_device_id table recognizes only\ndevices with ABS_X, ABS_WHEEL or ABS_THROTTLE axes as joysticks.\n\nThere are joystick devices that do not have those axes, for example TRC\nRudder device. The device in question has ABS_Z, ABS_RX and ABS_RY axes\ncausing it not being detected as joystick.\n\nThis patch adds ABS_Z to the input_device_id list allowing devices with\nABS_Z axis to be detected correctly.\n\nSigned-off-by: Ville Ranki \u003cville.ranki@iki.fi\u003e\nSigned-off-by: Dmitry Torokhov \u003cdmitry.torokhov@gmail.com\u003e" }, { "commit": "8d58881b995904bf8b150dae69be0829f832e7be", "tree": "38bb289ed3d1fa2912d9b7b47b22330affc66b18", "parents": [ "a0f81dbeef13aa0aeaa8b955b38735dbf09de392" ], "author": { "name": "Bart Van Assche", "email": "bart.vanassche@sandisk.com", "time": "Thu Sep 22 14:20:54 2016 -0700" }, "committer": { "name": "Martin K. Petersen", "email": "martin.petersen@oracle.com", "time": "Mon Sep 26 20:58:42 2016 -0400" }, "message": "scsi: Avoid that toggling use_blk_mq triggers a memory leak\n\nThis patch avoids that the following memory leak is triggered if\nuse_blk_mq is disabled after a SCSI host has been allocated by the\nib_srp driver and before the same SCSI host is freed:\n\nunreferenced object 0xffff8803a168c568 (size 256):\n backtrace:\n [\u003cffffffff81620c95\u003e] kmemleak_alloc+0x45/0xa0\n [\u003cffffffff811bb104\u003e] __kmalloc_node+0x1e4/0x400\n [\u003cffffffff81309fe4\u003e] blk_mq_alloc_tag_set+0xb4/0x230\n [\u003cffffffff814731b7\u003e] scsi_mq_setup_tags+0xc7/0xd0\n [\u003cffffffff81469c26\u003e] scsi_add_host_with_dma+0x216/0x2d0\n [\u003cffffffffa064bef5\u003e] srp_create_target+0xe55/0x13d0 [ib_srp]\n [\u003cffffffff8143ce23\u003e] dev_attr_store+0x13/0x20\n [\u003cffffffff8125f030\u003e] sysfs_kf_write+0x40/0x50\n [\u003cffffffff8125e397\u003e] kernfs_fop_write+0x137/0x1c0\n [\u003cffffffff811d8c13\u003e] __vfs_write+0x23/0x140\n [\u003cffffffff811d92e0\u003e] vfs_write+0xb0/0x190\n [\u003cffffffff811da5b4\u003e] SyS_write+0x44/0xa0\n [\u003cffffffff8162c8a5\u003e] entry_SYSCALL_64_fastpath+0x18/0xa8\n\nFixes: 9aa9cc4221f5 (\"scsi: remove the disable_blk_mq host flag\")\nSigned-off-by: Bart Van Assche \u003cbart.vanassche@sandisk.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Martin K. Petersen \u003cmartin.petersen@oracle.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Martin K. Petersen \u003cmartin.petersen@oracle.com\u003e\n" }, { "commit": "2cf750704bb6d7ed8c7d732e071dd1bc890ea5e8", "tree": "3af7d034bba1f98d1e5178568f958f31bc606244", "parents": [ "db32e4e49ce2b0e5fcc17803d011a401c0a637f6" ], "author": { "name": "Nikolay Aleksandrov", "email": "nikolay@cumulusnetworks.com", "time": "Sun Sep 25 23:08:31 2016 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 25 23:41:39 2016 -0400" }, "message": "ipmr, ip6mr: fix scheduling while atomic and a deadlock with ipmr_get_route\n\nSince the commit below the ipmr/ip6mr rtnl_unicast() code uses the portid\ninstead of the previous dst_pid which was copied from in_skb\u0027s portid.\nSince the skb is new the portid is 0 at that point so the packets are sent\nto the kernel and we get scheduling while atomic or a deadlock (depending\non where it happens) by trying to acquire rtnl two times.\nAlso since this is RTM_GETROUTE, it can be triggered by a normal user.\n\nHere\u0027s the sleeping while atomic trace:\n[ 7858.212557] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:620\n[ 7858.212748] in_atomic(): 1, irqs_disabled(): 0, pid: 0, name: swapper/0\n[ 7858.212881] 2 locks held by swapper/0/0:\n[ 7858.213013] #0: (((\u0026mrt-\u003eipmr_expire_timer))){+.-...}, at: [\u003cffffffff810fbbf5\u003e] call_timer_fn+0x5/0x350\n[ 7858.213422] #1: (mfc_unres_lock){+.....}, at: [\u003cffffffff8161e005\u003e] ipmr_expire_process+0x25/0x130\n[ 7858.213807] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.8.0-rc7+ #179\n[ 7858.213934] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014\n[ 7858.214108] 0000000000000000 ffff88005b403c50 ffffffff813a7804 0000000000000000\n[ 7858.214412] ffffffff81a1338e ffff88005b403c78 ffffffff810a4a72 ffffffff81a1338e\n[ 7858.214716] 000000000000026c 0000000000000000 ffff88005b403ca8 ffffffff810a4b9f\n[ 7858.215251] Call Trace:\n[ 7858.215412] \u003cIRQ\u003e [\u003cffffffff813a7804\u003e] dump_stack+0x85/0xc1\n[ 7858.215662] [\u003cffffffff810a4a72\u003e] ___might_sleep+0x192/0x250\n[ 7858.215868] [\u003cffffffff810a4b9f\u003e] __might_sleep+0x6f/0x100\n[ 7858.216072] [\u003cffffffff8165bea3\u003e] mutex_lock_nested+0x33/0x4d0\n[ 7858.216279] [\u003cffffffff815a7a5f\u003e] ? netlink_lookup+0x25f/0x460\n[ 7858.216487] [\u003cffffffff8157474b\u003e] rtnetlink_rcv+0x1b/0x40\n[ 7858.216687] [\u003cffffffff815a9a0c\u003e] netlink_unicast+0x19c/0x260\n[ 7858.216900] [\u003cffffffff81573c70\u003e] rtnl_unicast+0x20/0x30\n[ 7858.217128] [\u003cffffffff8161cd39\u003e] ipmr_destroy_unres+0xa9/0xf0\n[ 7858.217351] [\u003cffffffff8161e06f\u003e] ipmr_expire_process+0x8f/0x130\n[ 7858.217581] [\u003cffffffff8161dfe0\u003e] ? ipmr_net_init+0x180/0x180\n[ 7858.217785] [\u003cffffffff8161dfe0\u003e] ? ipmr_net_init+0x180/0x180\n[ 7858.217990] [\u003cffffffff810fbc95\u003e] call_timer_fn+0xa5/0x350\n[ 7858.218192] [\u003cffffffff810fbbf5\u003e] ? call_timer_fn+0x5/0x350\n[ 7858.218415] [\u003cffffffff8161dfe0\u003e] ? ipmr_net_init+0x180/0x180\n[ 7858.218656] [\u003cffffffff810fde10\u003e] run_timer_softirq+0x260/0x640\n[ 7858.218865] [\u003cffffffff8166379b\u003e] ? __do_softirq+0xbb/0x54f\n[ 7858.219068] [\u003cffffffff816637c8\u003e] __do_softirq+0xe8/0x54f\n[ 7858.219269] [\u003cffffffff8107a948\u003e] irq_exit+0xb8/0xc0\n[ 7858.219463] [\u003cffffffff81663452\u003e] smp_apic_timer_interrupt+0x42/0x50\n[ 7858.219678] [\u003cffffffff816625bc\u003e] apic_timer_interrupt+0x8c/0xa0\n[ 7858.219897] \u003cEOI\u003e [\u003cffffffff81055f16\u003e] ? native_safe_halt+0x6/0x10\n[ 7858.220165] [\u003cffffffff810d64dd\u003e] ? trace_hardirqs_on+0xd/0x10\n[ 7858.220373] [\u003cffffffff810298e3\u003e] default_idle+0x23/0x190\n[ 7858.220574] [\u003cffffffff8102a20f\u003e] arch_cpu_idle+0xf/0x20\n[ 7858.220790] [\u003cffffffff810c9f8c\u003e] default_idle_call+0x4c/0x60\n[ 7858.221016] [\u003cffffffff810ca33b\u003e] cpu_startup_entry+0x39b/0x4d0\n[ 7858.221257] [\u003cffffffff8164f995\u003e] rest_init+0x135/0x140\n[ 7858.221469] [\u003cffffffff81f83014\u003e] start_kernel+0x50e/0x51b\n[ 7858.221670] [\u003cffffffff81f82120\u003e] ? early_idt_handler_array+0x120/0x120\n[ 7858.221894] [\u003cffffffff81f8243f\u003e] x86_64_start_reservations+0x2a/0x2c\n[ 7858.222113] [\u003cffffffff81f8257c\u003e] x86_64_start_kernel+0x13b/0x14a\n\nFixes: 2942e9005056 (\"[RTNETLINK]: Use rtnl_unicast() for rtnetlink unicasts\")\nSigned-off-by: Nikolay Aleksandrov \u003cnikolay@cumulusnetworks.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "08895a8b6b06ed2323cd97a36ee40a116b3db8ed", "tree": "b2d12f6f5eda6efab1e55e1dd74ba63b68460925", "parents": [ "4c04b4b534cbe8c0cc0661e232bbb9708e212bc2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 18:47:13 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 18:47:13 2016 -0700" }, "message": "Linux 4.8-rc8\n" }, { "commit": "4c04b4b534cbe8c0cc0661e232bbb9708e212bc2", "tree": "1e103a03475fbf447dfae4ec874e05778adb5032", "parents": [ "90b75db6498a19da96dac4b55c909ff3721f3045", "1ae2293dd6d2f5c823cf97e60b70d03631cd622f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 18:40:13 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 18:40:13 2016 -0700" }, "message": "Merge tag \u0027trace-v4.8-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull tracefs fixes from Steven Rostedt:\n \"Al Viro has been looking at the tracefs code, and has pointed out some\n issues. This contains one fix by me and one by Al. I\u0027m sure that\n he\u0027ll come up with more but for now I tested these patches and they\n don\u0027t appear to have any negative impact on tracing\"\n\n* tag \u0027trace-v4.8-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n fix memory leaks in tracing_buffers_splice_read()\n tracing: Move mutex to protect against resetting of seq data\n" }, { "commit": "90b75db6498a19da96dac4b55c909ff3721f3045", "tree": "fa40265db710d5f6b599eeeeac14f5df7ec37842", "parents": [ "38e088546522e1e86d2b8f401a1354ad3a9b3303" ], "author": { "name": "Dave Chinner", "email": "dchinner@redhat.com", "time": "Mon Sep 26 09:57:33 2016 +1000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 18:16:44 2016 -0700" }, "message": "fault_in_multipages_readable() throws set-but-unused error\n\nWhen building XFS with -Werror, it now fails with:\n\n include/linux/pagemap.h: In function \u0027fault_in_multipages_readable\u0027:\n include/linux/pagemap.h:602:16: error: variable \u0027c\u0027 set but not used [-Werror\u003dunused-but-set-variable]\n volatile char c;\n ^\n\nThis is a regression caused by commit e23d4159b109 (\"fix\nfault_in_multipages_...() on architectures with no-op access_ok()\").\nFix it by re-adding the \"(void)c\" trick taht was previously used to make\nthe compiler think the variable is used.\n\nSigned-off-by: Dave Chinner \u003cdchinner@redhat.com\u003e\nCc: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "38e088546522e1e86d2b8f401a1354ad3a9b3303", "tree": "3e7144eb3ecd99edd02d0c2bb44e962d1982fd42", "parents": [ "831e45d84a971495c882bc186d98bbb825b2ee59" ], "author": { "name": "Lorenzo Stoakes", "email": "lstoakes@gmail.com", "time": "Sun Sep 11 23:54:25 2016 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 15:43:42 2016 -0700" }, "message": "mm: check VMA flags to avoid invalid PROT_NONE NUMA balancing\n\nThe NUMA balancing logic uses an arch-specific PROT_NONE page table flag\ndefined by pte_protnone() or pmd_protnone() to mark PTEs or huge page\nPMDs respectively as requiring balancing upon a subsequent page fault.\nUser-defined PROT_NONE memory regions which also have this flag set will\nnot normally invoke the NUMA balancing code as do_page_fault() will send\na segfault to the process before handle_mm_fault() is even called.\n\nHowever if access_remote_vm() is invoked to access a PROT_NONE region of\nmemory, handle_mm_fault() is called via faultin_page() and\n__get_user_pages() without any access checks being performed, meaning\nthe NUMA balancing logic is incorrectly invoked on a non-NUMA memory\nregion.\n\nA simple means of triggering this problem is to access PROT_NONE mmap\u0027d\nmemory using /proc/self/mem which reliably results in the NUMA handling\nfunctions being invoked when CONFIG_NUMA_BALANCING is set.\n\nThis issue was reported in bugzilla (issue 99101) which includes some\nsimple repro code.\n\nThere are BUG_ON() checks in do_numa_page() and do_huge_pmd_numa_page()\nadded at commit c0e7cad to avoid accidentally provoking strange\nbehaviour by attempting to apply NUMA balancing to pages that are in\nfact PROT_NONE. The BUG_ON()\u0027s are consistently triggered by the repro.\n\nThis patch moves the PROT_NONE check into mm/memory.c rather than\ninvoking BUG_ON() as faulting in these pages via faultin_page() is a\nvalid reason for reaching the NUMA check with the PROT_NONE page table\nflag set and is therefore not always a bug.\n\nLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d99101\nReported-by: Trevor Saunders \u003ctbsaunde@tbsaunde.org\u003e\nSigned-off-by: Lorenzo Stoakes \u003clstoakes@gmail.com\u003e\nAcked-by: Rik van Riel \u003criel@redhat.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Mel Gorman \u003cmgorman@techsingularity.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "831e45d84a971495c882bc186d98bbb825b2ee59", "tree": "51d9e306fe09d44e9cd90b976e37bcafdb9a65e0", "parents": [ "751b9a5d16cf318d715b0e6c7f3631d8185143ee", "116e7111c8e3cc65ceef9664741bd593483e9517" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:59:52 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:59:52 2016 -0700" }, "message": "Merge branch \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus\n\nPull MIPS fixes from Ralf Baechle:\n \"A round of 4.8 fixes:\n\n MIPS generic code:\n - Add a missing \".set pop\" in an early commit\n - Fix memory regions reaching top of physical\n - MAAR: Fix address alignment\n - vDSO: Fix Malta EVA mapping to vDSO page structs\n - uprobes: fix incorrect uprobe brk handling\n - uprobes: select HAVE_REGS_AND_STACK_ACCESS_API\n - Avoid a BUG warning during PR_SET_FP_MODE prctl\n - SMP: Fix possibility of deadlock when bringing CPUs online\n - R6: Remove compact branch policy Kconfig entries\n - Fix size calc when avoiding IPIs for small icache flushes\n - Fix pre-r6 emulation FPU initialisation\n - Fix delay slot emulation count in debugfs\n\n ATH79:\n - Fix test for error return of clk_register_fixed_factor.\n\n Octeon:\n - Fix kernel header to work for VDSO build.\n - Fix initialization of platform device probing.\n\n paravirt:\n - Fix undefined reference to smp_bootstrap\"\n\n* \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:\n MIPS: Fix delay slot emulation count in debugfs\n MIPS: SMP: Fix possibility of deadlock when bringing CPUs online\n MIPS: Fix pre-r6 emulation FPU initialisation\n MIPS: vDSO: Fix Malta EVA mapping to vDSO page structs\n MIPS: Select HAVE_REGS_AND_STACK_ACCESS_API\n MIPS: Octeon: Fix platform bus probing\n MIPS: Octeon: mangle-port: fix build failure with VDSO code\n MIPS: Avoid a BUG warning during prctl(PR_SET_FP_MODE, ...)\n MIPS: c-r4k: Fix size calc when avoiding IPIs for small icache flushes\n MIPS: Add a missing \".set pop\" in an early commit\n MIPS: paravirt: Fix undefined reference to smp_bootstrap\n MIPS: Remove compact branch policy Kconfig entries\n MIPS: MAAR: Fix address alignment\n MIPS: Fix memory regions reaching top of physical\n MIPS: uprobes: fix incorrect uprobe brk handling\n MIPS: ath79: Fix test for error return of clk_register_fixed_factor().\n" }, { "commit": "751b9a5d16cf318d715b0e6c7f3631d8185143ee", "tree": "fe9ba09c8d1241a772974ccf1a8364038b901de4", "parents": [ "8d2c0d36d6826ddc3114801c599619d3f2932f0a", "b79331a5eb9f96e4dfd216974581168ec4c8a4d4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:52:59 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:52:59 2016 -0700" }, "message": "Merge tag \u0027powerpc-4.8-7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull one more powerpc fix from Michael Ellerman:\n \"powernv/pci: Fix m64 checks for SR-IOV and window alignment from\n Russell Currey\"\n\n* tag \u0027powerpc-4.8-7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/powernv/pci: Fix m64 checks for SR-IOV and window alignment\n" }, { "commit": "8d2c0d36d6826ddc3114801c599619d3f2932f0a", "tree": "826496294c0a46b6f66a65584e01ec6fe70278bf", "parents": [ "62fd5258ebe3ea240371234955a6e2cc99e0b6c3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:32:46 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 13:32:46 2016 -0700" }, "message": "radix tree: fix sibling entry handling in radix_tree_descend()\n\nThe fixes to the radix tree test suite show that the multi-order case is\nbroken. The basic reason is that the radix tree code uses tagged\npointers with the \"internal\" bit in the low bits, and calculating the\npointer indices was supposed to mask off those bits. But gcc will\nnotice that we then use the index to re-create the pointer, and will\navoid doing the arithmetic and use the tagged pointer directly.\n\nThis cleans the code up, using the existing is_sibling_entry() helper to\nvalidate the sibling pointer range (instead of open-coding it), and\nusing entry_to_node() to mask off the low tag bit from the pointer. And\nonce you do that, you might as well just use the now cleaned-up pointer\ndirectly.\n\n[ Side note: the multi-order code isn\u0027t actually ever used in the kernel\n right now, and the only reason I didn\u0027t just delete all that code is\n that Kirill Shutemov piped up and said:\n\n \"Well, my ext4-with-huge-pages patchset[1] uses multi-order entries.\n It also converts shmem-with-huge-pages and hugetlb to them.\n\n I\u0027m okay with converting it to other mechanism, but I need\n something. (I looked into Konstantin\u0027s RFC patchset[2]. It looks\n okay, but I don\u0027t feel myself qualified to review it as I don\u0027t\n know much about radix-tree internals.)\"\n\n [1] http://lkml.kernel.org/r/20160915115523.29737-1-kirill.shutemov@linux.intel.com\n [2] http://lkml.kernel.org/r/147230727479.9957.1087787722571077339.stgit@zurg ]\n\nReported-by: Matthew Wilcox \u003cmawilcox@microsoft.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Ross Zwisler \u003cross.zwisler@linux.intel.com\u003e\nCc: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Konstantin Khlebnikov \u003ckoct9i@gmail.com\u003e\nCc: Cedric Blancher \u003ccedric.blancher@gmail.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "62fd5258ebe3ea240371234955a6e2cc99e0b6c3", "tree": "c138367e16b619c13a21fdb562b541936d4a4136", "parents": [ "9c0e28a7be656d737fb18998e2dcb0b8ce595643" ], "author": { "name": "Matthew Wilcox", "email": "mawilcox@microsoft.com", "time": "Thu Sep 22 11:53:34 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 25 11:49:16 2016 -0700" }, "message": "radix tree test suite: Test radix_tree_replace_slot() for multiorder entries\n\nWhen we replace a multiorder entry, check that all indices reflect the\nnew value.\n\nAlso, compile the test suite with -O2, which shows other problems with\nthe code due to some dodgy pointer operations in the radix tree code.\n\nSigned-off-by: Matthew Wilcox \u003cmawilcox@microsoft.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "1ae2293dd6d2f5c823cf97e60b70d03631cd622f", "tree": "1cd2b21bd5f6d29ba96ce66daa111cd38fd3aaed", "parents": [ "1245800c0f96eb6ebb368593e251d66c01e61022" ], "author": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Sat Sep 17 18:31:46 2016 -0400" }, "committer": { "name": "Steven Rostedt", "email": "rostedt@goodmis.org", "time": "Sun Sep 25 13:30:13 2016 -0400" }, "message": "fix memory leaks in tracing_buffers_splice_read()\n\nCc: stable@vger.kernel.org\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "1245800c0f96eb6ebb368593e251d66c01e61022", "tree": "c945fb377258b480d806786d33fbd0f6ce118022", "parents": [ "c87edb36118664f1fa275107c1138b6f47793240" ], "author": { "name": "Steven Rostedt (Red Hat)", "email": "rostedt@goodmis.org", "time": "Fri Sep 23 22:57:13 2016 -0400" }, "committer": { "name": "Steven Rostedt", "email": "rostedt@goodmis.org", "time": "Sun Sep 25 10:27:08 2016 -0400" }, "message": "tracing: Move mutex to protect against resetting of seq data\n\nThe iter-\u003eseq can be reset outside the protection of the mutex. So can\nreading of user data. Move the mutex up to the beginning of the function.\n\nFixes: d7350c3f45694 (\"tracing/core: make the read callbacks reentrants\")\nCc: stable@vger.kernel.org # 2.6.30+\nReported-by: Al Viro \u003cviro@ZenIV.linux.org.uk\u003e\nSigned-off-by: Steven Rostedt \u003crostedt@goodmis.org\u003e\n" }, { "commit": "116e7111c8e3cc65ceef9664741bd593483e9517", "tree": "1ae6cda95eddf35d3424b27d7f567b316804bf66", "parents": [ "8f46cca1e6c06a058374816887059bcc017b382f" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Thu Sep 22 15:47:40 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Sun Sep 25 01:59:16 2016 +0200" }, "message": "MIPS: Fix delay slot emulation count in debugfs\n\nCommit 432c6bacbd0c (\"MIPS: Use per-mm page to execute branch delay slot\ninstructions\") accidentally removed use of the MIPS_FPU_EMU_INC_STATS\nmacro from do_dsemulret, leading to the ds_emul file in debugfs always\nreturning zero even though we perform delay slot emulations.\n\nFix this by re-adding the use of the MIPS_FPU_EMU_INC_STATS macro.\n\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nFixes: 432c6bacbd0c (\"MIPS: Use per-mm page to execute branch delay slot instructions\")\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14301/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "8f46cca1e6c06a058374816887059bcc017b382f", "tree": "0208125c0c12aac63194f8864e79cd26a39c4e2f", "parents": [ "7e956304eb8a285304a78582e4537e72c6365f20" ], "author": { "name": "Matt Redfearn", "email": "matt.redfearn@imgtec.com", "time": "Thu Sep 22 17:15:47 2016 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Sun Sep 25 01:43:52 2016 +0200" }, "message": "MIPS: SMP: Fix possibility of deadlock when bringing CPUs online\n\nThis patch fixes the possibility of a deadlock when bringing up\nsecondary CPUs.\nThe deadlock occurs because the set_cpu_online() is called before\nsynchronise_count_slave(). This can cause a deadlock if the boot CPU,\nhaving scheduled another thread, attempts to send an IPI to the\nsecondary CPU, which it sees has been marked online. The secondary is\nblocked in synchronise_count_slave() waiting for the boot CPU to enter\nsynchronise_count_master(), but the boot cpu is blocked in\nsmp_call_function_many() waiting for the secondary to respond to it\u0027s\nIPI request.\n\nFix this by marking the CPU online in cpu_callin_map and synchronising\ncounters before declaring the CPU online and calculating the maps for\nIPIs.\n\nSigned-off-by: Matt Redfearn \u003cmatt.redfearn@imgtec.com\u003e\nReported-by: Justin Chen \u003cjustinpopo6@gmail.com\u003e\nTested-by: Justin Chen \u003cjustinpopo6@gmail.com\u003e\nCc: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nCc: stable@vger.kernel.org # v4.1+\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/14302/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "9c0e28a7be656d737fb18998e2dcb0b8ce595643", "tree": "10f54f40da96c519ba2d781cae9398dfc73a3a2c", "parents": [ "2507c856620cc7474e6101b0a05f82ac0ae5bf69", "3bf6215a1b30db7df6083c708caab3fe1a8e8abe" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:44:28 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:44:28 2016 -0700" }, "message": "Merge branch \u0027perf-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull perf fixes from Thomas Gleixner:\n \"Three fixlets for perf:\n\n - add a missing NULL pointer check in the intel BTS driver\n\n - make BTS an exclusive PMU because BTS can only handle one event at\n a time\n\n - ensure that exclusive events are limited to one PMU so that several\n exclusive events can be scheduled on different PMU instances\"\n\n* \u0027perf-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n perf/core: Limit matching exclusive events to one PMU\n perf/x86/intel/bts: Make it an exclusive PMU\n perf/x86/intel/bts: Make sure debug store is valid\n" }, { "commit": "2507c856620cc7474e6101b0a05f82ac0ae5bf69", "tree": "a7bdba4229890416e1e5fe50fa9052a7bdb122d1", "parents": [ "709b8f67d72078cbbd166aaecd2d0f48d1d1560c", "96b03ab86d843524ec4aed7fe0ceef412c684c68" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:41:19 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:41:19 2016 -0700" }, "message": "Merge branch \u0027locking-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull locking fixes from Thomas Gleixner:\n \"Two smallish fixes:\n\n - use the proper asm constraint in the Super-H atomic_fetch_ops\n\n - a trivial typo fix in the Kconfig help text\"\n\n* \u0027locking-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n locking/hung_task: Fix typo in CONFIG_DETECT_HUNG_TASK help text\n locking/atomic, arch/sh: Fix ATOMIC_FETCH_OP()\n" }, { "commit": "709b8f67d72078cbbd166aaecd2d0f48d1d1560c", "tree": "9d86b5110b0086f43099fe1813b637dcdd3fda6d", "parents": [ "4b8b0ff60f1d6553914a5fc17f16f9aa38a2036e", "7597cdc066313bfd211cca2f9252dfeb41271391" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:35:26 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:35:26 2016 -0700" }, "message": "Merge branch \u0027efi-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull EFI fixes from Thomas Gleixner:\n \"Two fixes for EFI/PAT:\n\n - a 32bit overflow bug in the PAT code which was unearthed by the\n large EFI mappings\n\n - prevent a boot hang on large systems when EFI mixed mode is enabled\n but not used\"\n\n* \u0027efi-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/efi: Only map RAM into EFI page tables if in mixed-mode\n x86/mm/pat: Prevent hang during boot when mapping pages\n" }, { "commit": "4b8b0ff60f1d6553914a5fc17f16f9aa38a2036e", "tree": "4c841fbb7ae2baf0a8a9926f717079408eb51cbd", "parents": [ "0f26574178f6c698e5d76e66ca68a95cc35eef9f", "e875bd66dfb68f4e898e9a43ef42858c504a7f23" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:30:12 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 12:30:12 2016 -0700" }, "message": "Merge branch \u0027irq-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull irq fixes from Thomas Gleixner:\n \"Three fixes for irq core and irq chip drivers:\n\n - Do not set the irq type if type is NONE. Fixes a boot regression\n on various SoCs\n\n - Use the proper cpu for setting up the GIC target list. Discovered\n by the cpumask debugging code.\n\n - A rather large fix for the MIPS-GIC so per cpu local interrupts\n work again. This was discovered late because the code falls back\n to slower timers which use normal device interrupts\"\n\n* \u0027irq-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n irqchip/mips-gic: Fix local interrupts\n irqchip/gicv3: Silence noisy DEBUG_PER_CPU_MAPS warning\n genirq: Skip chained interrupt trigger setup if type is IRQ_TYPE_NONE\n" }, { "commit": "595c73071e6641e59b83911fbb4026e767471000", "tree": "52feca3df3886d10dfc2fd1750658b212846cbb2", "parents": [ "11294d63ac915230a36b0603c62134ef7b173d0a" ], "author": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Fri Sep 23 17:53:52 2016 -0700" }, "committer": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Sat Sep 24 11:45:38 2016 -0700" }, "message": "libnvdimm, region: fix flush hint table thinko\n\nThe definition of the flush hint table as:\n\n\tvoid __iomem *flush_wpq[0][0];\n\n...passed the unit test, but is broken as flush_wpq[0][1] and\nflush_wpq[1][0] refer to the same entry. Fix this to use a helper that\ncalculates a slot in the table based on the geometry of flush hints in\nthe region. This is important to get right since virtualization\nsolutions use this mechanism to trigger hypervisor flushes to platform\npersistence.\n\nReported-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nTested-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nSigned-off-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\n" }, { "commit": "0f26574178f6c698e5d76e66ca68a95cc35eef9f", "tree": "af9591d14bd7f2d941cc9baabedaa2c802a0c20c", "parents": [ "bd5dbcb4be37ce97b115bf9e8d48ae75f0d8a9a2", "b385d21f27d86426472f6ae92a231095f7de2a8d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 11:31:45 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 11:31:45 2016 -0700" }, "message": "Merge branch \u0027hughd-fixes\u0027 (patches from Hugh Dickins)\n\nMerge VM fixes from High Dickins:\n \"I get the impression that Andrew is away or busy at the moment, so I\u0027m\n going to send you three independent uncontroversial little mm fixes\n directly - though none is strictly a 4.8 regression fix.\n\n - shmem: fix tmpfs to handle the huge\u003d option properly from Toshi\n Kani is a one-liner to fix a major embarrassment in 4.8\u0027s hugepages\n on tmpfs feature: although Hillf pointed it out in June, somehow\n both Kirill and I repeatedly dropped the ball on this one. You\n might wonder if the feature got tested at all with that bug in:\n yes, it did, but for wider testing coverage, Kirill and I had each\n relied too much on an override which bypasses that condition.\n\n - huge tmpfs: fix Committed_AS leak just a run-of-the-mill accounting\n fix in the same feature.\n\n - mm: delete unnecessary and unsafe init_tlb_ubc() is an unrelated\n fix to 4.3\u0027s TLB flush batching in reclaim: the bug would be rare,\n and none of us will be shamed if this one misses 4.8; but it got\n such a quick ack from Mel today that I\u0027m inclined to offer it along\n with the first two\"\n\n* emailed patches from Hugh Dickins \u003chughd@google.com\u003e:\n mm: delete unnecessary and unsafe init_tlb_ubc()\n huge tmpfs: fix Committed_AS leak\n shmem: fix tmpfs to handle the huge\u003d option properly\n" }, { "commit": "b385d21f27d86426472f6ae92a231095f7de2a8d", "tree": "a8e2a20246c8beaf70a429a1e66f07ad437320af", "parents": [ "71664665c3e3ca5ff61ef5fc65480f82cd575eb2" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Fri Sep 23 20:27:04 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 11:20:01 2016 -0700" }, "message": "mm: delete unnecessary and unsafe init_tlb_ubc()\n\ninit_tlb_ubc() looked unnecessary to me: tlb_ubc is statically\ninitialized with zeroes in the init_task, and copied from parent to\nchild while it is quiescent in arch_dup_task_struct(); so I went to\ndelete it.\n\nBut inserted temporary debug WARN_ONs in place of init_tlb_ubc() to\ncheck that it was always empty at that point, and found them firing:\nbecause memcg reclaim can recurse into global reclaim (when allocating\nbiosets for swapout in my case), and arrive back at the init_tlb_ubc()\nin shrink_node_memcg().\n\nResetting tlb_ubc.flush_required at that point is wrong: if the upper\nlevel needs a deferred TLB flush, but the lower level turns out not to,\nwe miss a TLB flush. But fortunately, that\u0027s the only part of the\nprotocol that does not nest: with the initialization removed, cpumask\ncollects bits from upper and lower levels, and flushes TLB when needed.\n\nFixes: 72b252aed506 (\"mm: send one IPI per CPU to TLB flush all entries after unmapping pages\")\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Mel Gorman \u003cmgorman@techsingularity.net\u003e\nCc: stable@vger.kernel.org # 4.3+\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "71664665c3e3ca5ff61ef5fc65480f82cd575eb2", "tree": "96572603989e027b559a3e117a773bb62ec434e3", "parents": [ "3089bf614c7e2fd441ee001e3ff3d18326f6f091" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Fri Sep 23 20:24:23 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 11:20:01 2016 -0700" }, "message": "huge tmpfs: fix Committed_AS leak\n\nUnder swapping load on huge tmpfs, /proc/meminfo\u0027s Committed_AS grows\nbigger and bigger: just a cosmetic issue for most users, but disabling\nfor those who run without overcommit (/proc/sys/vm/overcommit_memory 2).\n\nshmem_uncharge() was forgetting to unaccount __vm_enough_memory\u0027s\ncharge, and shmem_charge() was forgetting it on the filesystem-full\nerror path.\n\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3089bf614c7e2fd441ee001e3ff3d18326f6f091", "tree": "b16cd2c7b0213c55dfe2d8b867e54745b403a861", "parents": [ "3be7988674ab33565700a37b210f502563d932e6" ], "author": { "name": "Toshi Kani", "email": "toshi.kani@hpe.com", "time": "Fri Sep 23 20:21:56 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 24 11:20:01 2016 -0700" }, "message": "shmem: fix tmpfs to handle the huge\u003d option properly\n\nshmem_get_unmapped_area() checks SHMEM_SB(sb)-\u003ehuge incorrectly, which\nleads to a reversed effect of \"huge\u003d\" mount option.\n\nFix the check in shmem_get_unmapped_area().\n\nNote, the default value of SHMEM_SB(sb)-\u003ehuge remains as\nSHMEM_HUGE_NEVER. User will need to specify \"huge\u003d\" option to enable\nhuge page mappings.\n\nReported-by: Hillf Danton \u003chillf.zj@alibaba-inc.com\u003e\nSigned-off-by: Toshi Kani \u003ctoshi.kani@hpe.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nReviewed-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "db32e4e49ce2b0e5fcc17803d011a401c0a637f6", "tree": "dd23eb8d057170e143636366351f17c59626499b", "parents": [ "019b1c9fe32a2a32c1153e31375f87ec3e591273" ], "author": { "name": "Lance Richardson", "email": "lrichard@redhat.com", "time": "Fri Sep 23 15:50:29 2016 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 24 09:44:31 2016 -0400" }, "message": "ip6_gre: fix flowi6_proto value in ip6gre_xmit_other()\n\nSimilar to commit 3be07244b733 (\"ip6_gre: fix flowi6_proto value in\nxmit path\"), set flowi6_proto to IPPROTO_GRE for output route lookup.\n\nUp until now, ip6gre_xmit_other() has set flowi6_proto to a bogus value.\nThis affected output route lookup for packets sent on an ip6gretap device\nin cases where routing was dependent on the value of flowi6_proto.\n\nSince the correct proto is already set in the tunnel flowi6 template via\ncommit 252f3f5a1189 (\"ip6_gre: Set flowi6_proto as IPPROTO_GRE in xmit\npath.\"), simply delete the line setting the incorrect flowi6_proto value.\n\nSuggested-by: Jiri Benc \u003cjbenc@redhat.com\u003e\nFixes: c12b395a4664 (\"gre: Support GRE over IPv6\")\nReviewed-by: Shmulik Ladkani \u003cshmulik.ladkani@gmail.com\u003e\nSigned-off-by: Lance Richardson \u003clrichard@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bd5dbcb4be37ce97b115bf9e8d48ae75f0d8a9a2", "tree": "5c8f44fc553089aa48de80ea9e267fcd65bdc8f8", "parents": [ "d0c1d15f5ed7f028d2c284b7fed7e73575d9c24d", "331dcf421c34d227784d07943eb01e4023a42b0a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:44:12 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:44:12 2016 -0700" }, "message": "Merge branch \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull i2c fixes from Wolfram Sang:\n \"Three driver bugfixes: fixing uninitialized memory pointers (eg20t),\n pm/clock imbalance (qup), and a wrongly set cached variable (pc954x)\"\n\n* \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n i2c: qup: skip qup_i2c_suspend if the device is already runtime suspended\n i2c: mux: pca954x: retry updating the mux selection on failure\n i2c-eg20t: fix race between i2c init and interrupt enable\n" }, { "commit": "d0c1d15f5ed7f028d2c284b7fed7e73575d9c24d", "tree": "95668ae49d855b243ab5f31f36eb1d12652373ef", "parents": [ "4ee69866258620b105d06ae07da3bbe9fdb1699a", "4af2ff91ec3f42b538a65cf12df5f9faf6aaa914" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:34:24 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:34:24 2016 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input\n\nPull input updates from Dmitry Torokhov:\n \"Just a fix up for the firmware handling to the Silead driver (which is\n a new driver in this release)\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Input: silead_gsl1680 - use \"silead/\" prefix for firmware loading\n Input: silead_gsl1680 - document firmware-name, fix implementation\n" }, { "commit": "4ee69866258620b105d06ae07da3bbe9fdb1699a", "tree": "e17e4aea5675b4abcf548f1d87f9586e1cd19541", "parents": [ "b22734a55067adbc10216e459762dbd7dcef29d5", "c8712c6a674e3382fe4d26d108251ccfa55d08e0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:24:36 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 16:24:36 2016 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"Three fixes, two regressions and one that poses a problem in blk-mq\n with the new nvmef code\"\n\n* \u0027for-linus\u0027 of git://git.kernel.dk/linux-block:\n blk-mq: skip unmapped queues in blk_mq_alloc_request_hctx\n nvme-rdma: only clear queue flags after successful connect\n blk-throttle: Extend slice if throttle group is not empty\n" }, { "commit": "9157056da8f8c4a6305f15619e269f164b63a6de", "tree": "3b2cea5209fae8b62b46ab19c3ce4d4a897b9b40", "parents": [ "8a15b81741879fa89601b03c0e50b0d780d65bc0" ], "author": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Fri Sep 23 16:55:49 2016 -0400" }, "committer": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Fri Sep 23 16:55:49 2016 -0400" }, "message": "cgroup: fix invalid controller enable rejections with cgroup namespace\n\nOn the v2 hierarchy, \"cgroup.subtree_control\" rejects controller\nenables if the cgroup has processes in it. The enforcement of this\nlogic assumes that the cgroup wouldn\u0027t have any css_sets associated\nwith it if there are no tasks in the cgroup, which is no longer true\nsince a79a908fd2b0 (\"cgroup: introduce cgroup namespaces\").\n\nWhen a cgroup namespace is created, it pins the css_set of the\ncreating task to use it as the root css_set of the namespace. This\nextra reference stays as long as the namespace is around and makes\n\"cgroup.subtree_control\" think that the namespace root cgroup is not\nempty even when it is and thus reject controller enables.\n\nFix it by making cgroup_subtree_control() walk and test emptiness of\neach css_set instead of testing whether the list_head is empty.\n\nWhile at it, update the comment of cgroup_task_count() to indicate\nthat the returned value may be higher than the number of tasks, which\nhas always been true due to temporary references and doesn\u0027t break\nanything.\n\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\nReported-by: Evgeny Vereshchagin \u003cevvers@ya.ru\u003e\nCc: Serge E. Hallyn \u003cserge.hallyn@ubuntu.com\u003e\nCc: Aditya Kali \u003cadityakali@google.com\u003e\nCc: Eric W. Biederman \u003cebiederm@xmission.com\u003e\nCc: stable@vger.kernel.org # v4.6+\nFixes: a79a908fd2b0 (\"cgroup: introduce cgroup namespaces\")\nLink: https://github.com/systemd/systemd/pull/3589#issuecomment-249089541\n" }, { "commit": "b22734a55067adbc10216e459762dbd7dcef29d5", "tree": "b672082875eea08035930888ffda3177599c0c43", "parents": [ "78bbf153fa96e5e40234b7b05567602535645460", "325c50e3cebb9208009083e841550f98a863bfa0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 13:39:37 2016 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 23 13:39:37 2016 -0700" }, "message": "Merge branch \u0027for-linus-4.8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs\n\nPull btrfs fixes from Chris Mason:\n \"Josef fixed a problem when quotas are enabled with his latest ENOSPC\n rework, and Jeff added more checks into the subvol ioctls to avoid\n tripping up lookup_one_len\"\n\n* \u0027for-linus-4.8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:\n btrfs: ensure that file descriptor used with subvol ioctls is a dir\n Btrfs: handle quota reserve failure properly\n" } ], "next": "78bbf153fa96e5e40234b7b05567602535645460" }