)]}' { "log": [ { "commit": "9d99c7123c9af5f47a2037933a3ffaec3e37efad", "tree": "73617b729d320323cc3f9362fc1ce1b5acb9d847", "parents": [ "049e6dde7e57f0054fdc49102e7ef4830c698b46" ], "author": { "name": "Christian Melki", "email": "christian.melki@t2data.com", "time": "Mon Oct 05 17:31:33 2015 +0200" }, "committer": { "name": "Konrad Rzeszutek Wilk", "email": "konrad.wilk@oracle.com", "time": "Wed Oct 07 15:31:35 2015 -0400" }, "message": "swiotlb: Enable it under x86 PAE\n\nMost distributions end up enabling SWIOTLB already with 32-bit\nkernels due to the combination of CONFIG_HYPERVISOR_GUEST|CONFIG_XEN\u003dy\nas those end up requiring the SWIOTLB.\n\nHowever for those that are not interested in virtualization and\nrun in 32-bit they will discover that: \"32-bit PAE 4.2.0 kernel\n(no IOMMU code) would hang when writing to my USB disk. The kernel\nspews million(-ish messages per sec) to syslog, effectively\n\"hanging\" userspace with my kernel.\n\nOct 2 14:33:06 voodoochild kernel: [ 223.287447] nommu_map_sg:\noverflow 25dcac000+1024 of device mask ffffffff\nOct 2 14:33:06 voodoochild kernel: [ 223.287448] nommu_map_sg:\noverflow 25dcac000+1024 of device mask ffffffff\nOct 2 14:33:06 voodoochild kernel: [ 223.287449] nommu_map_sg:\noverflow 25dcac000+1024 of device mask ffffffff\n... etc ...\"\n\nEnabling it makes the problem go away.\n\nN.B. With a6dfa128ce5c414ab46b1d690f7a1b8decb8526d\n\"config: Enable NEED_DMA_MAP_STATE by default when SWIOTLB is selected\"\nwe also have the important part of the SG macros enabled to make this\nwork properly - in case anybody wants to backport this patch.\n\nReported-and-Tested-by: Christian Melki \u003cchristian.melki@t2data.com\u003e\nSigned-off-by: Christian Melki \u003cchristian.melki@t2data.com\u003e\nSigned-off-by: Konrad Rzeszutek Wilk \u003ckonrad.wilk@oracle.com\u003e\n" }, { "commit": "049e6dde7e57f0054fdc49102e7ef4830c698b46", "tree": "ec9c177c64815ee50957bd972f8be3a27a9261e4", "parents": [ "30c44659f4a3e7e1f9f47e895591b4b40bf62671" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 16:57:17 2015 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 16:57:17 2015 +0100" }, "message": "Linux 4.3-rc4\n" }, { "commit": "30c44659f4a3e7e1f9f47e895591b4b40bf62671", "tree": "0bc2af55dd7f7e7fa0a2d1ff11b5929f5ed1fc9e", "parents": [ "15ecf9a986e2678f5de36ead23b89235612fc03f", "30059d494a72603d066baf55c748803df968aa08" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 16:31:13 2015 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 16:31:13 2015 +0100" }, "message": "Merge branch \u0027strscpy\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile\n\nPull strscpy string copy function implementation from Chris Metcalf.\n\nChris sent this during the merge window, but I waffled back and forth on\nthe pull request, which is why it\u0027s going in only now.\n\nThe new \"strscpy()\" function is definitely easier to use and more secure\nthan either strncpy() or strlcpy(), both of which are horrible nasty\ninterfaces that have serious and irredeemable problems.\n\nstrncpy() has a useless return value, and doesn\u0027t NUL-terminate an\noverlong result. To make matters worse, it pads a short result with\nzeroes, which is a performance disaster if you have big buffers.\n\nstrlcpy(), by contrast, is a mis-designed \"fix\" for strlcpy(), lacking\nthe insane NUL padding, but having a differently broken return value\nwhich returns the original length of the source string. Which means\nthat it will read characters past the count from the source buffer, and\nyou have to trust the source to be properly terminated. It also makes\nerror handling fragile, since the test for overflow is unnecessarily\nsubtle.\n\nstrscpy() avoids both these problems, guaranteeing the NUL termination\n(but not excessive padding) if the destination size wasn\u0027t zero, and\nmaking the overflow condition very obvious by returning -E2BIG. It also\ndoesn\u0027t read past the size of the source, and can thus be used for\nuntrusted source data too.\n\nSo why did I waffle about this for so long?\n\nEvery time we introduce a new-and-improved interface, people start doing\nthese interminable series of trivial conversion patches.\n\nAnd every time that happens, somebody does some silly mistake, and the\nconversion patch to the improved interface actually makes things worse.\nBecause the patch is mindnumbing and trivial, nobody has the attention\nspan to look at it carefully, and it\u0027s usually done over large swatches\nof source code which means that not every conversion gets tested.\n\nSo I\u0027m pulling the strscpy() support because it *is* a better interface.\nBut I will refuse to pull mindless conversion patches. Use this in\nplaces where it makes sense, but don\u0027t do trivial patches to fix things\nthat aren\u0027t actually known to be broken.\n\n* \u0027strscpy\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile:\n tile: use global strscpy() rather than private copy\n string: provide strscpy()\n Make asm/word-at-a-time.h available on all architectures\n" }, { "commit": "15ecf9a986e2678f5de36ead23b89235612fc03f", "tree": "2eb463a995f42091a147a5093e1d3b818b78206e", "parents": [ "0d8770815f70cf41b69a82ede272b026dbb2df7d", "da6fb7a9e5bd6f04f7e15070f630bdf1ea502841" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:47:28 2015 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:47:28 2015 +0100" }, "message": "Merge tag \u0027md/4.3-fixes\u0027 of git://neil.brown.name/md\n\nPull md fixes from Neil Brown:\n \"Assorted fixes for md in 4.3-rc.\n\n Two tagged for -stable, and one is really a cleanup to match and\n improve kmemcache interface.\n\n* tag \u0027md/4.3-fixes\u0027 of git://neil.brown.name/md:\n md/bitmap: don\u0027t pass -1 to bitmap_storage_alloc.\n md/raid1: Avoid raid1 resync getting stuck\n md: drop null test before destroy functions\n md: clear CHANGE_PENDING in readonly array\n md/raid0: apply base queue limits *before* disk_stack_limits\n md/raid5: don\u0027t index beyond end of array in need_this_block().\n raid5: update analysis state for failed stripe\n md: wait for pending superblock updates before switching to read-only\n" }, { "commit": "0d8770815f70cf41b69a82ede272b026dbb2df7d", "tree": "3f088af1c614fd3b73e3aeae64223088797ef873", "parents": [ "3e519dde1e8a73e59b0eab52575a14a2986283b6", "d218af78492a36a4ae607c08fedfb59258440314" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:41:58 2015 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:41:58 2015 +0100" }, "message": "Merge branch \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus\n\nPull MIPS updates from Ralf Baechle:\n \"This week\u0027s round of MIPS fixes:\n - Fix JZ4740 build\n - Fix fallback to GFP_DMA\n - FP seccomp in case of ENOSYS\n - Fix bootmem panic\n - A number of FP and CPS fixes\n - Wire up new syscalls\n - Make sure BPF assembler objects can properly be disassembled\n - Fix BPF assembler code for MIPS I\"\n\n* \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:\n MIPS: scall: Always run the seccomp syscall filters\n MIPS: Octeon: Fix kernel panic on startup from memory corruption\n MIPS: Fix R2300 FP context switch handling\n MIPS: Fix octeon FP context switch handling\n MIPS: BPF: Fix load delay slots.\n MIPS: BPF: Do all exports of symbols with FEXPORT().\n MIPS: Fix the build on jz4740 after removing the custom gpio.h\n MIPS: CPS: #ifdef on CONFIG_MIPS_MT_SMP rather than CONFIG_MIPS_MT\n MIPS: CPS: Don\u0027t include MT code in non-MT kernels.\n MIPS: CPS: Stop dangling delay slot from has_mt.\n MIPS: dma-default: Fix 32-bit fall back to GFP_DMA\n MIPS: Wire up userfaultfd and membarrier syscalls.\n" }, { "commit": "3e519dde1e8a73e59b0eab52575a14a2986283b6", "tree": "16e19515f19a239aaa155e927a37d8e59304b202", "parents": [ "2cf30826bbc6c940568be07e8fab0aee02165cf8", "791c76d58465a248cbd1ee422c8075cb90fa615f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:40:09 2015 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Oct 04 11:40:09 2015 +0100" }, "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 \"This update contains:\n\n - Fix for a long standing race affecting /proc/irq/NNN\n\n - One line fix for ARM GICV3-ITS counting the wrong data\n\n - Warning silencing in ARM GICV3-ITS. Another GCC trying to be\n overly clever issue\"\n\n* \u0027irq-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n irqchip/gic-v3-its: Count additional LPIs for the aliased devices\n irqchip/gic-v3-its: Silence warning when its_lpi_alloc_chunks gets inlined\n genirq: Fix race in register_irq_proc()\n" }, { "commit": "d218af78492a36a4ae607c08fedfb59258440314", "tree": "fb9a1d1c08161c39c6f8b8db04b0cc116959718b", "parents": [ "66803dd9198cb57a4b7ed4a6846a63ab1d59a2e0" ], "author": { "name": "Markos Chandras", "email": "markos.chandras@imgtec.com", "time": "Fri Sep 25 08:17:42 2015 +0100" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Sun Oct 04 12:10:56 2015 +0200" }, "message": "MIPS: scall: Always run the seccomp syscall filters\n\nThe MIPS syscall handler code used to return -ENOSYS on invalid\nsyscalls. Whilst this is expected, it caused problems for seccomp\nfilters because the said filters never had the change to run since\nthe code returned -ENOSYS before triggering them. This caused\nproblems on the chromium testsuite for filters looking for invalid\nsyscalls. This has now changed and the seccomp filters are always\nrun even if the syscall is invalid. We return -ENOSYS once we\nreturn from the seccomp filters. Moreover, similar codepaths have\nbeen merged in the process which simplifies somewhat the overall\nsyscall code.\n\nSigned-off-by: Markos Chandras \u003cmarkos.chandras@imgtec.com\u003e\nCc: linux-mips@linux-mips.org\nPatchwork: https://patchwork.linux-mips.org/patch/11236/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "2cf30826bbc6c940568be07e8fab0aee02165cf8", "tree": "69a3606e8e965407aee9022ed7ae3f3441f33595", "parents": [ "37cc7ab1d2c317cc989b8aa0224cfc5f0478ccbd", "f4b4aae1828855db761bf998ce37d3062b1d6446" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:53:05 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:53:05 2015 -0400" }, "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 Ingo Molnar:\n \"Fixes all around the map: W+X kernel mapping fix, WCHAN fixes, two\n build failure fixes for corner case configs, x32 header fix and a\n speling fix\"\n\n* \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/headers/uapi: Fix __BITS_PER_LONG value for x32 builds\n x86/mm: Set NX on gap between __ex_table and rodata\n x86/kexec: Fix kexec crash in syscall kexec_file_load()\n x86/process: Unify 32bit and 64bit implementations of get_wchan()\n x86/process: Add proper bound checks in 64bit get_wchan()\n x86, efi, kasan: Fix build failure on !KASAN \u0026\u0026 KMEMCHECK\u003dy kernels\n x86/hyperv: Fix the build in the !CONFIG_KEXEC_CORE case\n x86/cpufeatures: Correct spelling of the HWP_NOTIFY flag\n" }, { "commit": "37cc7ab1d2c317cc989b8aa0224cfc5f0478ccbd", "tree": "3db756ca12b5ab24fab1b7b5cb99ef7635cd370a", "parents": [ "a758379b031f50b9def094aad071ef547a5cb335", "67dfae0cd72fec5cd158b6e5fb1647b7dbe0834c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:51:41 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:51:41 2015 -0400" }, "message": "Merge branch \u0027timers-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull timer fixes from Ingo Molnar:\n \"An abs64() fix in the watchdog driver, and two clocksource driver\n NO_IRQ assumption fixes\"\n\n* \u0027timers-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n clocksource: Fix abs() usage w/ 64bit values\n clocksource/drivers/keystone: Fix bad NO_IRQ usage\n clocksource/drivers/rockchip: Fix bad NO_IRQ usage\n" }, { "commit": "a758379b031f50b9def094aad071ef547a5cb335", "tree": "3a29fedfd77b578a568916ce4c3b6cf2750b8a77", "parents": [ "14f97d9713283adfadf2193e287e21d079f72ee7", "0ce3cc008ec04258b6a6314b09f1a6012810881a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:46:41 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:46:41 2015 -0400" }, "message": "Merge branch \u0027core-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull EFI fixes from Ingo Molnar:\n \"Two EFI fixes: one for x86, one for ARM, fixing a boot crash bug that\n can trigger under newer EFI firmware\"\n\n* \u0027core-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n arm64/efi: Fix boot crash by not padding between EFI_MEMORY_RUNTIME regions\n x86/efi: Fix boot crash by mapping EFI memmap entries bottom-up at runtime, instead of top-down\n" }, { "commit": "14f97d9713283adfadf2193e287e21d079f72ee7", "tree": "ac0e6497aacd976bf49b404a517a671fcb883324", "parents": [ "978ab6a009e6691375a0b6f8a44ebbb471a64b6d", "ccf03d6995fa4b784f5b987726ba98f4859bf326" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:39:31 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 03 10:39:31 2015 -0400" }, "message": "Merge branch \u0027drm-fixes\u0027 of git://people.freedesktop.org/~airlied/linux\n\nPull drm fixes from Dave Airlie:\n \"Bunch of fixes all over the place, all pretty small: amdgpu, i915,\n exynos, one qxl and one vmwgfx.\n\n There is also a bunch of mst fixes, I left some cleanups in the series\n as I didn\u0027t think it was worth splitting up the tested series\"\n\n* \u0027drm-fixes\u0027 of git://people.freedesktop.org/~airlied/linux: (37 commits)\n drm/dp/mst: add some defines for logical/physical ports\n drm/dp/mst: drop cancel work sync in the mstb destroy path (v2)\n drm/dp/mst: split connector registration into two parts (v2)\n drm/dp/mst: update the link_address_sent before sending the link address (v3)\n drm/dp/mst: fixup handling hotplug on port removal.\n drm/dp/mst: don\u0027t pass port into the path builder function\n drm/radeon: drop radeon_fb_helper_set_par\n drm: handle cursor_set2 in restore_fbdev_mode\n drm/exynos: Staticize local function in exynos_drm_gem.c\n drm/exynos: fimd: actually disable dp clock\n drm/exynos: dp: remove suspend/resume functions\n drm/qxl: recreate the primary surface when the bo is not primary\n drm/amdgpu: only print meaningful VM faults\n drm/amdgpu/cgs: remove import_gpu_mem\n drm/i915: Call non-locking version of drm_kms_helper_poll_enable(), v2\n drm: Add a non-locking version of drm_kms_helper_poll_enable(), v2\n drm/vmwgfx: Fix a command submission hang regression\n drm/exynos: remove unused mode_fixup() code\n drm/exynos: remove decon_mode_fixup()\n drm/exynos: remove fimd_mode_fixup()\n ...\n" }, { "commit": "978ab6a009e6691375a0b6f8a44ebbb471a64b6d", "tree": "eec7ee99a0502a38ed82a67ab880da66d29a7f08", "parents": [ "5634347dee31373a8faf084f4cdbf6d5ea0b03a4", "62d78461447198b49383f20301aaa15fe97dfa4f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 17:53:25 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 17:53:25 2015 -0400" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input\n\nPull input layer fixes from Dmitry Torokhov:\n \"Fixes for two recent regressions (in Synaptics PS/2 and uinput\n drivers) and some more driver fixups\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Revert \"Input: synaptics - fix handling of disabling gesture mode\"\n Input: psmouse - fix data race in __ps2_command\n Input: elan_i2c - add all valid ic type for i2c/smbus\n Input: zhenhua - ensure we have BITREVERSE\n Input: omap4-keypad - fix memory leak\n Input: serio - fix blocking of parport\n Input: uinput - fix crash when using ABS events\n Input: elan_i2c - expand maximum product_id form 0xFF to 0xFFFF\n Input: elan_i2c - add ic type 0x03\n Input: elan_i2c - don\u0027t require known iap version\n Input: imx6ul_tsc - fix controller name\n Input: imx6ul_tsc - use the preferred method for kzalloc()\n Input: imx6ul_tsc - check for negative return value\n Input: imx6ul_tsc - propagate the errors\n Input: walkera0701 - fix abs() calculations on 64 bit values\n Input: mms114 - remove unneded semicolons\n Input: pm8941-pwrkey - remove unneded semicolon\n Input: fix typo in MT documentation\n Input: cyapa - fix address of Gen3 devices in device tree documentation\n" }, { "commit": "67dfae0cd72fec5cd158b6e5fb1647b7dbe0834c", "tree": "6b4740d09261098ec7731412b8537418df5ec255", "parents": [ "bdf7344e14d826d0df438a55fc51146d179e198d" ], "author": { "name": "John Stultz", "email": "john.stultz@linaro.org", "time": "Mon Sep 14 18:05:20 2015 -0700" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Oct 02 22:53:01 2015 +0200" }, "message": "clocksource: Fix abs() usage w/ 64bit values\n\nThis patch fixes one cases where abs() was being used with 64-bit\nnanosecond values, where the result may be capped at 32-bits.\n\nThis potentially could cause watchdog false negatives on 32-bit\nsystems, so this patch addresses the issue by using abs64().\n\nSigned-off-by: John Stultz \u003cjohn.stultz@linaro.org\u003e\nCc: Prarit Bhargava \u003cprarit@redhat.com\u003e\nCc: Richard Cochran \u003crichardcochran@gmail.com\u003e\nCc: Ingo Molnar \u003cmingo@kernel.org\u003e\nCc: stable@vger.kernel.org\nLink: http://lkml.kernel.org/r/1442279124-7309-2-git-send-email-john.stultz@linaro.org\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "5634347dee31373a8faf084f4cdbf6d5ea0b03a4", "tree": "1ff99fa3cbdd23eb25dcb5a4f83e27389f61786b", "parents": [ "b55a97e759c9e9efdce0470f520026383c514a13", "ee556d00cf20012e889344a0adbbf809ab5015a3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:54:16 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:54:16 2015 -0400" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fixes from Catalin Marinas:\n\n - Fix for transparent huge page change_protection() logic which was\n inadvertently changing a huge pmd page into a pmd table entry.\n\n - Function graph tracer panic fix caused by the return_to_handler code\n corrupting the multi-regs function return value (composite types).\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: ftrace: fix function_graph tracer panic\n arm64: Fix THP protection change logic\n" }, { "commit": "b55a97e759c9e9efdce0470f520026383c514a13", "tree": "502a67e8598807803164358cc684f7590ede73db", "parents": [ "83dc311ce0b430ee8e37e62976e6753bfd022004", "95bc06ef049b808a067327bd8490b608b47e3870" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:51:46 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:51:46 2015 -0400" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k\n\nPull m68k updates from Geert Uytterhoeven:\n \"Summary:\n - Fix for accidental modification of arguments of syscall functions\n - Wire up new syscalls\n - Update defconfigs\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:\n m68k/defconfig: Update defconfigs for v4.3-rc1\n m68k: Define asmlinkage_protect\n m68k: Wire up membarrier\n m68k: Wire up userfaultfd\n m68k: Wire up direct socket calls\n" }, { "commit": "791c76d58465a248cbd1ee422c8075cb90fa615f", "tree": "0587c21bf759c782127cb0dfb70840acbe4e80e6", "parents": [ "c8415b9470727f70afce8607d4fe521789aa6c1c" ], "author": { "name": "Marc Zyngier", "email": "marc.zyngier@arm.com", "time": "Fri Oct 02 16:44:06 2015 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Oct 02 20:51:41 2015 +0200" }, "message": "irqchip/gic-v3-its: Count additional LPIs for the aliased devices\n\nWhen configuring the interrupt mapping for a new device, we\niterate over all the possible aliases to account for their\nmaximum MSI allocation. This was introduced by e8137f4f5088\n(\"irqchip: gicv3-its: Iterate over PCI aliases to generate ITS configuration\").\n\nTurns out that the code doing that is a bit braindead, and repeatedly\naccounts for the same device over and over.\n\nFix this by counting the actual alias that is passed to us by the\ncore code.\n\nSigned-off-by: Marc Zyngier \u003cmarc.zyngier@arm.com\u003e\nCc: linux-arm-kernel@lists.infradead.org\nCc: Alex Shi \u003calex.shi@linaro.org\u003e\nCc: Ard Biesheuvel \u003card.biesheuvel@linaro.org\u003e\nCc: David Daney \u003cddaney.cavm@gmail.com\u003e\nCc: Jason Cooper \u003cjason@lakedaemon.net\u003e\nLink: http://lkml.kernel.org/r/1443800646-8074-3-git-send-email-marc.zyngier@arm.com\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "c8415b9470727f70afce8607d4fe521789aa6c1c", "tree": "31778c67d2bc90ddcc8892ebc3f4797c8c4da324", "parents": [ "95c2b17534654829db428f11bcf4297c059a2a7e" ], "author": { "name": "Marc Zyngier", "email": "marc.zyngier@arm.com", "time": "Fri Oct 02 16:44:05 2015 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Oct 02 20:51:41 2015 +0200" }, "message": "irqchip/gic-v3-its: Silence warning when its_lpi_alloc_chunks gets inlined\n\nMore agressive inlining in recent versions of GCC have uncovered\na new set of warnings:\n\n drivers/irqchip/irq-gic-v3-its.c: In function its_msi_prepare:\n drivers/irqchip/irq-gic-v3-its.c:1148:26: warning: lpi_base may be used\n uninitialized in this function [-Wmaybe-uninitialized]\n dev-\u003eevent_map.lpi_base \u003d lpi_base;\n ^\n drivers/irqchip/irq-gic-v3-its.c:1116:6: note: lpi_base was declared here\n int lpi_base;\n\t ^\n drivers/irqchip/irq-gic-v3-its.c:1149:25: warning: nr_lpis may be used\n uninitialized in this function [-Wmaybe-uninitialized]\n dev-\u003eevent_map.nr_lpis \u003d nr_lpis;\n\t ^\n drivers/irqchip/irq-gic-v3-its.c:1117:6: note: nr_lpis was declared here\n int nr_lpis;\n\t ^\nThe warning is fairly benign (there is no code path that could\nactually use uninitialized variables), but let\u0027s silence it anyway\nby zeroing the variables on the error path.\n\nReported-by: Alex Shi \u003calex.shi@linaro.org\u003e\nTested-by: Ard Biesheuvel \u003card.biesheuvel@linaro.org\u003e\nSigned-off-by: Marc Zyngier \u003cmarc.zyngier@arm.com\u003e\nCc: linux-arm-kernel@lists.infradead.org\nCc: David Daney \u003cddaney.cavm@gmail.com\u003e\nCc: Jason Cooper \u003cjason@lakedaemon.net\u003e\nLink: http://lkml.kernel.org/r/1443800646-8074-2-git-send-email-marc.zyngier@arm.com\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "83dc311ce0b430ee8e37e62976e6753bfd022004", "tree": "7241969107dd15d92e0f14ad2f3eca5b26b466d1", "parents": [ "27728bf04b99a0abc5c27343c06e18379f57c726", "7b09a1bba4091a9d208481d7831682a1f3061ab9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:46:15 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:46:15 2015 -0400" }, "message": "Merge tag \u0027dmaengine-fix-4.3-rc4\u0027 of git://git.infradead.org/users/vkoul/slave-dma\n\nPull dmaengine fixes from Vinod Koul:\n \"This contains fixes spread throughout the drivers, and also fixes one\n more instance of privatecnt in dmaengine.\n\n Driver fixes summary:\n - bunch of pxa_dma fixes for reuse of descriptor issue, residue and\n no-requestor\n - odd fixes in xgene, idma, sun4i and zxdma\n - at_xdmac fixes for cleaning descriptor and block addr mode\"\n\n* tag \u0027dmaengine-fix-4.3-rc4\u0027 of git://git.infradead.org/users/vkoul/slave-dma:\n dmaengine: pxa_dma: fix residue corner case\n dmaengine: pxa_dma: fix the no-requestor case\n dmaengine: zxdma: Fix off-by-one for testing valid pchan request\n dmaengine: at_xdmac: clean used descriptor\n dmaengine: at_xdmac: change block increment addressing mode\n dmaengine: dw: properly read DWC_PARAMS register\n dmaengine: xgene-dma: Fix overwritting DMA tx ring\n dmaengine: fix balance of privatecnt\n dmaengine: sun4i: fix unsafe list iteration\n dmaengine: idma64: improve residue estimation\n dmaengine: xgene-dma: fix handling xgene_dma_get_ring_size result\n dmaengine: pxa_dma: fix initial list move\n" }, { "commit": "27728bf04b99a0abc5c27343c06e18379f57c726", "tree": "80b1341742676a2ab7e4dcd81b5ce12eec9898c0", "parents": [ "36f8dafe52b0d0726edd5102900e37192f1d3c65", "0bf6cd5b9531bcc29c0a5e504b6ce2984c6fd8d8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:40:57 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 14:40:57 2015 -0400" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"Another week, another round of fixes.\n\n These have been brewing for a bit and in various iterations, but I\n feel pretty comfortable about the quality of them. They fix real\n issues. The pull request is mostly blk-mq related, and the only one\n not fixing a real bug, is the tag iterator abstraction from Christoph.\n But it\u0027s pretty trivial, and we\u0027ll need it for another fix soon.\n\n Apart from the blk-mq fixes, there\u0027s an NVMe affinity fix from Keith,\n and a single fix for xen-blkback from Roger fixing failure to free\n requests on disconnect\"\n\n* \u0027for-linus\u0027 of git://git.kernel.dk/linux-block:\n blk-mq: factor out a helper to iterate all tags for a request_queue\n blk-mq: fix racy updates of rq-\u003eerrors\n blk-mq: fix deadlock when reading cpu_list\n blk-mq: avoid inserting requests before establishing new mapping\n blk-mq: fix q-\u003emq_usage_counter access race\n blk-mq: Fix use after of free q-\u003emq_map\n blk-mq: fix sysfs registration/unregistration race\n blk-mq: avoid setting hctx-\u003etags-\u003ecpumask before allocation\n NVMe: Set affinity after allocating request queues\n xen/blkback: free requests on disconnection\n" }, { "commit": "62d78461447198b49383f20301aaa15fe97dfa4f", "tree": "2eef5c1e037dcd589531cafe0cb350b7bd5593a3", "parents": [ "218c1f76b8b25d6dc9d01443f071cb618e206b0c" ], "author": { "name": "Dmitry Torokhov", "email": "dmitry.torokhov@gmail.com", "time": "Fri Oct 02 10:31:32 2015 -0700" }, "committer": { "name": "Dmitry Torokhov", "email": "dmitry.torokhov@gmail.com", "time": "Fri Oct 02 10:31:32 2015 -0700" }, "message": "Revert \"Input: synaptics - fix handling of disabling gesture mode\"\n\nThis reverts commit e51e38494a8ecc18650efb0c840600637891de2c: we\nactually do want the device to work in extended W mode, as this is the\nmode that allows us receiving multiple contact information.\n\nCc: stable@vger.kernel.org\n" }, { "commit": "66803dd9198cb57a4b7ed4a6846a63ab1d59a2e0", "tree": "fac03809b0103c8c4eb9e9bd34999c1955ac2b1c", "parents": [ "085c2f25d36ef4a69bb1dab933daee0692426f15" ], "author": { "name": "Matt Bennett", "email": "matt.bennett@alliedtelesis.co.nz", "time": "Wed Sep 30 17:40:42 2015 +1300" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Oct 02 19:19:55 2015 +0200" }, "message": "MIPS: Octeon: Fix kernel panic on startup from memory corruption\n\nDuring development it was found that a number of builds would panic\nduring the kernel init process, more specifically in \u0027delayed_fput()\u0027.\nThe panic showed the kernel trying to access a memory address of\n\u00270xb7fdc00\u0027 while traversing the \u0027delayed_fput_list\u0027 structure.\nComparing this memory address to the value of the pointer used on\nbuilds that did not panic confirmed that the pointer on crashing\nbuilds must have been corrupted at some stage earlier in the init\nprocess.\n\nBy traversing the list earlier and earlier in the code it was found\nthat \u0027plat_mem_setup()\u0027 was responsible for corrupting the list.\nSpecifically the line:\n\n memory \u003d cvmx_bootmem_phy_alloc(mem_alloc_size,\n\t\t\t__pa_symbol(\u0026__init_end), -1,\n\t\t\t0x100000,\n\t\t\tCVMX_BOOTMEM_FLAG_NO_LOCKING);\n\nWhich would eventually call:\n\n cvmx_bootmem_phy_set_size(new_ent_addr,\n\t\tcvmx_bootmem_phy_get_size\n\t\t(ent_addr) -\n\t\t(desired_min_addr -\n\t\t\tent_addr));\n\nWhere \u0027new_ent_addr\u0027\u003d0x4800000 (the address of \u0027delayed_fput_list\u0027)\nand the second argument (size)\u003d0xb7fdc00 (the address causing the\nkernel panic). The job of this part of \u0027plat_mem_setup()\u0027 is to\nallocate chunks of memory for the kernel to use. At the start of\neach chunk of memory the size of the chunk is written, hence the\nvalue 0xb7fdc00 is written onto memory at 0x4800000, therefore the\nkernel panics when it goes back to access \u0027delayed_fput_list\u0027 later\non in the initialisation process.\n\nOn builds that were not crashing it was found that the compiler had\nplaced \u0027delayed_fput_list\u0027 at 0x4800008, meaning it wasn\u0027t corrupted\n(but something else in memory was overwritten).\n\nAs can be seen in the first function call above the code begins to\nallocate chunks of memory beginning from the symbol \u0027__init_end\u0027.\nThe MIPS linker script (vmlinux.lds.S) however defines the .bss\nsection to begin after \u0027__init_end\u0027. Therefore memory within the\n.bss section is allocated to the kernel to use (System.map shows\n\u0027delayed_fput_list\u0027 and other kernel structures to be in .bss).\n\nTo stop the kernel panic (and the .bss section being corrupted)\nmemory should begin being allocated from the symbol \u0027_end\u0027.\n\nSigned-off-by: Matt Bennett \u003cmatt.bennett@alliedtelesis.co.nz\u003e\nAcked-by: David Daney \u003cdavid.daney@cavium.com\u003e\nCc: linux-mips@linux-mips.org\nCc: aleksey.makarov@auriga.com\nPatchwork: https://patchwork.linux-mips.org/patch/11251/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "085c2f25d36ef4a69bb1dab933daee0692426f15", "tree": "7fc03ac0ba5a66889daa06524abb8f5159c1ec09", "parents": [ "0fa24340f7c88d2814547d8d24d5e3a1803009cc" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Mon Sep 21 10:07:42 2015 -0700" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Oct 02 19:16:46 2015 +0200" }, "message": "MIPS: Fix R2300 FP context switch handling\n\nCommit 1a3d59579b9f (\"MIPS: Tidy up FPU context switching\") removed FP\ncontext saving from the asm-written resume function in favour of reusing\nexisting code to perform the same task. However it only removed the FP\ncontext saving code from the r4k_switch.S implementation of resume.\nRemove it from the r2300_switch.S implementation too in order to prevent\nattempting to save the FP context twice, which would likely lead to an\nexception from the second save because the FPU had already been disabled\nby the first save.\n\nThis patch has only been build tested, using rbtx49xx_defconfig.\n\nFixes: 1a3d59579b9f (\"MIPS: Tidy up FPU context switching\")\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: linux-mips@linux-mips.org\nCc: Maciej W. Rozycki \u003cmacro@linux-mips.org\u003e\nCc: linux-kernel@vger.kernel.org\nCc: Manuel Lauss \u003cmanuel.lauss@gmail.com\u003e\nPatchwork: https://patchwork.linux-mips.org/patch/11167/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "0fa24340f7c88d2814547d8d24d5e3a1803009cc", "tree": "cb21b22c6f00e5beeb966bd2796667229fe9376b", "parents": [ "0c5d187828588dd1b36cb93b4481a8db467ef3e8" ], "author": { "name": "Paul Burton", "email": "paul.burton@imgtec.com", "time": "Mon Sep 21 10:07:41 2015 -0700" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Oct 02 19:16:06 2015 +0200" }, "message": "MIPS: Fix octeon FP context switch handling\n\nCommit 1a3d59579b9f (\"MIPS: Tidy up FPU context switching\") removed FP\ncontext saving from the asm-written resume function in favour of reusing\nexisting code to perform the same task. However it only removed the FP\ncontext saving code from the r4k_switch.S implementation of resume.\nOcteon uses its own implementation in octeon_switch.S, so remove FP\ncontext saving there too in order to prevent attempting to save context\ntwice. That formerly led to an exception from the second save as follows\nbecause the FPU had already been disabled by the first save:\n\n do_cpu invoked from kernel context![#1]:\n CPU: 0 PID: 2 Comm: kthreadd Not tainted 4.3.0-rc2-dirty #2\n task: 800000041f84a008 ti: 800000041f864000 task.ti: 800000041f864000\n $ 0 : 0000000000000000 0000000010008ce1 0000000000100000 ffffffffbfffffff\n $ 4 : 800000041f84a008 800000041f84ac08 800000041f84c000 0000000000000004\n $ 8 : 0000000000000001 0000000000000000 0000000000000000 0000000000000001\n $12 : 0000000010008ce3 0000000000119c60 0000000000000036 800000041f864000\n $16 : 800000041f84ac08 800000000792ce80 800000041f84a008 ffffffff81758b00\n $20 : 0000000000000000 ffffffff8175ae50 0000000000000000 ffffffff8176c740\n $24 : 0000000000000006 ffffffff81170300\n $28 : 800000041f864000 800000041f867d90 0000000000000000 ffffffff815f3fa0\n Hi : 0000000000fa8257\n Lo : ffffffffe15cfc00\n epc : ffffffff8112821c resume+0x9c/0x200\n ra : ffffffff815f3fa0 __schedule+0x3f0/0x7d8\n Status: 10008ce2 KX SX UX KERNEL EXL\n Cause : 1080002c (ExcCode 0b)\n PrId : 000d0601 (Cavium Octeon+)\n Modules linked in:\n Process kthreadd (pid: 2, threadinfo\u003d800000041f864000, task\u003d800000041f84a008, tls\u003d0000000000000000)\n Stack : ffffffff81604218 ffffffff815f7e08 800000041f84a008 ffffffff811681b0\n 800000041f84a008 ffffffff817e9878 0000000000000000 ffffffff81770000\n ffffffff81768340 ffffffff81161398 0000000000000001 0000000000000000\n 0000000000000000 ffffffff815f4424 0000000000000000 ffffffff81161d68\n ffffffff81161be8 0000000000000000 0000000000000000 0000000000000000\n 0000000000000000 0000000000000000 0000000000000000 ffffffff8111e16c\n 0000000000000000 0000000000000000 0000000000000000 0000000000000000\n 0000000000000000 0000000000000000 0000000000000000 0000000000000000\n 0000000000000000 0000000000000000 0000000000000000 0000000000000000\n 0000000000000000 0000000000000000 0000000000000000 0000000000000000\n ...\n Call Trace:\n [\u003cffffffff8112821c\u003e] resume+0x9c/0x200\n [\u003cffffffff815f3fa0\u003e] __schedule+0x3f0/0x7d8\n [\u003cffffffff815f4424\u003e] schedule+0x34/0x98\n [\u003cffffffff81161d68\u003e] kthreadd+0x180/0x198\n [\u003cffffffff8111e16c\u003e] ret_from_kernel_thread+0x14/0x1c\n\nTested using cavium_octeon_defconfig on an EdgeRouter Lite.\n\nFixes: 1a3d59579b9f (\"MIPS: Tidy up FPU context switching\")\nReported-by: Aaro Koskinen \u003caaro.koskinen@nokia.com\u003e\nSigned-off-by: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: linux-mips@linux-mips.org\nCc: Aleksey Makarov \u003caleksey.makarov@auriga.com\u003e\nCc: linux-kernel@vger.kernel.org\nCc: Chandrakala Chavva \u003ccchavva@caviumnetworks.com\u003e\nCc: David Daney \u003cdavid.daney@cavium.com\u003e\nCc: Leonid Rosenboim \u003clrosenboim@caviumnetworks.com\u003e\nPatchwork: https://patchwork.linux-mips.org/patch/11166/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "36f8dafe52b0d0726edd5102900e37192f1d3c65", "tree": "275620178300956525d5f01f70f6dd5cf2ce91ee", "parents": [ "8c25ab8b5a04a7c559aa8fd4cabe5fc4463b8ada", "031277d4d33d33f0174fbb569ca8f68238175617" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 08:03:04 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 08:03:04 2015 -0400" }, "message": "Merge tag \u0027mmc-v4.3-rc3\u0027 of git://git.linaro.org/people/ulf.hansson/mmc\n\nPull MMC fixes from Ulf Hansson:\n \"Here are some mmc fixes intended for v4.3 rc4:\n\n MMC core:\n - Allow users of mmc_of_parse() to succeed when CONFIG_GPIOLIB is\n unset\n - Prevent infinite loop of re-tuning for CRC-errors for CMD19 and\n CMD21\n\n MMC host:\n - pxamci: Fix issues with card detect\n - sunxi: Fix clk-delay settings\"\n\n* tag \u0027mmc-v4.3-rc3\u0027 of git://git.linaro.org/people/ulf.hansson/mmc:\n mmc: core: fix dead loop of mmc_retune\n mmc: pxamci: fix card detect with slot-gpio API\n mmc: sunxi: Fix clk-delay settings\n mmc: core: Don\u0027t return an error for CD/WP GPIOs when GPIOLIB is unset\n" }, { "commit": "8c25ab8b5a04a7c559aa8fd4cabe5fc4463b8ada", "tree": "9cdc7d9cb6576fa757277cd2e3ad8231191442b9", "parents": [ "bde17b90dd9712cb61a7ab0c1ccd0f7f6aa57957", "15bbdec3931e617231c12b0920e497e87ec8c2c6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 07:59:29 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 02 07:59:29 2015 -0400" }, "message": "Merge git://git.infradead.org/intel-iommu\n\nPull IOVA fixes from David Woodhouse:\n \"The main fix here is the first one, fixing the over-allocation of\n size-aligned requests. The other patches simply make the existing\n IOVA code available to users other than the Intel VT-d driver, with no\n functional change.\n\n I concede the latter really *should* have been submitted during the\n merge window, but since it\u0027s basically risk-free and people are\n waiting to build on top of it and it\u0027s my fault I didn\u0027t get it in, I\n (and they) would be grateful if you\u0027d take it\"\n\n* git://git.infradead.org/intel-iommu:\n iommu: Make the iova library a module\n iommu: iova: Export symbols\n iommu: iova: Move iova cache management to the iova library\n iommu/iova: Avoid over-allocating when size-aligned\n" }, { "commit": "ee556d00cf20012e889344a0adbbf809ab5015a3", "tree": "21880eb6317793473dea4af2612e292479918f02", "parents": [ "1a541b4e3cd6f5795022514114854b3e1345f24e" ], "author": { "name": "Li Bin", "email": "huawei.libin@huawei.com", "time": "Wed Sep 30 10:49:55 2015 +0800" }, "committer": { "name": "Catalin Marinas", "email": "catalin.marinas@arm.com", "time": "Fri Oct 02 11:12:56 2015 +0100" }, "message": "arm64: ftrace: fix function_graph tracer panic\n\nWhen function graph tracer is enabled, the following operation\nwill trigger panic:\n\nmount -t debugfs nodev /sys/kernel\necho next_tgid \u003e /sys/kernel/tracing/set_ftrace_filter\necho function_graph \u003e /sys/kernel/tracing/current_tracer\nls /proc/\n\n------------[ cut here ]------------\n[ 198.501417] Unable to handle kernel paging request at virtual address cb88537fdc8ba316\n[ 198.506126] pgd \u003d ffffffc008f79000\n[ 198.509363] [cb88537fdc8ba316] *pgd\u003d00000000488c6003, *pud\u003d00000000488c6003, *pmd\u003d0000000000000000\n[ 198.517726] Internal error: Oops: 94000005 [#1] SMP\n[ 198.518798] Modules linked in:\n[ 198.520582] CPU: 1 PID: 1388 Comm: ls Tainted: G\n[ 198.521800] Hardware name: linux,dummy-virt (DT)\n[ 198.522852] task: ffffffc0fa9e8000 ti: ffffffc0f9ab0000 task.ti: ffffffc0f9ab0000\n[ 198.524306] PC is at next_tgid+0x30/0x100\n[ 198.525205] LR is at return_to_handler+0x0/0x20\n[ 198.526090] pc : [\u003cffffffc0002a1070\u003e] lr : [\u003cffffffc0000907c0\u003e] pstate: 60000145\n[ 198.527392] sp : ffffffc0f9ab3d40\n[ 198.528084] x29: ffffffc0f9ab3d40 x28: ffffffc0f9ab0000\n[ 198.529406] x27: ffffffc000d6a000 x26: ffffffc000b786e8\n[ 198.530659] x25: ffffffc0002a1900 x24: ffffffc0faf16c00\n[ 198.531942] x23: ffffffc0f9ab3ea0 x22: 0000000000000002\n[ 198.533202] x21: ffffffc000d85050 x20: 0000000000000002\n[ 198.534446] x19: 0000000000000002 x18: 0000000000000000\n[ 198.535719] x17: 000000000049fa08 x16: ffffffc000242efc\n[ 198.537030] x15: 0000007fa472b54c x14: ffffffffff000000\n[ 198.538347] x13: ffffffc0fada84a0 x12: 0000000000000001\n[ 198.539634] x11: ffffffc0f9ab3d70 x10: ffffffc0f9ab3d70\n[ 198.540915] x9 : ffffffc0000907c0 x8 : ffffffc0f9ab3d40\n[ 198.542215] x7 : 0000002e330f08f0 x6 : 0000000000000015\n[ 198.543508] x5 : 0000000000000f08 x4 : ffffffc0f9835ec0\n[ 198.544792] x3 : cb88537fdc8ba316 x2 : cb88537fdc8ba306\n[ 198.546108] x1 : 0000000000000002 x0 : ffffffc000d85050\n[ 198.547432]\n[ 198.547920] Process ls (pid: 1388, stack limit \u003d 0xffffffc0f9ab0020)\n[ 198.549170] Stack: (0xffffffc0f9ab3d40 to 0xffffffc0f9ab4000)\n[ 198.582568] Call trace:\n[ 198.583313] [\u003cffffffc0002a1070\u003e] next_tgid+0x30/0x100\n[ 198.584359] [\u003cffffffc0000907bc\u003e] ftrace_graph_caller+0x6c/0x70\n[ 198.585503] [\u003cffffffc0000907bc\u003e] ftrace_graph_caller+0x6c/0x70\n[ 198.586574] [\u003cffffffc0000907bc\u003e] ftrace_graph_caller+0x6c/0x70\n[ 198.587660] [\u003cffffffc0000907bc\u003e] ftrace_graph_caller+0x6c/0x70\n[ 198.588896] Code: aa0003f5 2a0103f4 b4000102 91004043 (885f7c60)\n[ 198.591092] ---[ end trace 6a346f8f20949ac8 ]---\n\nThis is because when using function graph tracer, if the traced\nfunction return value is in multi regs ([x0-x7]), return_to_handler\nmay corrupt them. So in return_to_handler, the parameter regs should\nbe protected properly.\n\nCc: \u003cstable@vger.kernel.org\u003e # 3.18+\nSigned-off-by: Li Bin \u003chuawei.libin@huawei.com\u003e\nAcked-by: AKASHI Takahiro \u003ctakahiro.akashi@linaro.org\u003e\nSigned-off-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\n" }, { "commit": "0c5d187828588dd1b36cb93b4481a8db467ef3e8", "tree": "899cdb3b4b6e2b99718d7916296bcfbffc853846", "parents": [ "1e16a8f11669c98a0adf5fb5f8522aebc1f69f71" ], "author": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Oct 02 09:48:57 2015 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Oct 02 09:48:57 2015 +0200" }, "message": "MIPS: BPF: Fix load delay slots.\n\nThe entire bpf_jit_asm.S is written in noreorder mode because \"we know\nbetter\" according to a comment. This also prevented the assembler from\nthrowing in the required NOPs for MIPS I processors which have no\nload-use interlock, thus the load\u0027s consumer might end up using the\nold value of the register from prior to the load.\n\nFixed by putting the assembler in reorder mode for just the affected\nload instructions. This is not enough for gas to actually try to be\nclever by looking at the next instruction and inserting a nop only\nwhen needed but as the comment said \"we know better\", so getting gas\nto unconditionally emit a NOP is just right in this case and prevents\nadding further ifdefery.\n\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "f4b4aae1828855db761bf998ce37d3062b1d6446", "tree": "e2bf23ffeb061a9949415114a7538c00f4f7605a", "parents": [ "ab76f7b4ab2397ffdd2f1eb07c55697d19991d10" ], "author": { "name": "Ben Hutchings", "email": "ben@decadent.org.uk", "time": "Thu Oct 01 01:40:43 2015 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Oct 02 09:43:21 2015 +0200" }, "message": "x86/headers/uapi: Fix __BITS_PER_LONG value for x32 builds\n\nOn x32, gcc predefines __x86_64__ but long is only 32-bit. Use\n__ILP32__ to distinguish x32.\n\nFixes this compiler error in perf:\n\n\ttools/include/asm-generic/bitops/__ffs.h: In function \u0027__ffs\u0027:\n\ttools/include/asm-generic/bitops/__ffs.h:19:8: error: right shift count \u003e\u003d width of type [-Werror\u003dshift-count-overflow]\n\t word \u003e\u003e\u003d 32;\n\t ^\n\nThis isn\u0027t sufficient to build perf for x32, though.\n\nSigned-off-by: Ben Hutchings \u003cben@decadent.org.uk\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: linux-kernel@vger.kernel.org\nLink: http://lkml.kernel.org/r/1443660043.2730.15.camel@decadent.org.uk\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "da6fb7a9e5bd6f04f7e15070f630bdf1ea502841", "tree": "cc7f1bf1d5b2e213bc3ff759d8dcf3c318907da7", "parents": [ "e8ff8bf09ff49733534ff3cee91bde030186055f" ], "author": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Thu Oct 01 16:03:38 2015 +1000" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:24:13 2015 +1000" }, "message": "md/bitmap: don\u0027t pass -1 to bitmap_storage_alloc.\n\nPassing -1 to bitmap_storage_alloc() causes page-\u003eindex to be set to\n-1, which is quite problematic.\n\nSo only pass -\u003ecluster_slot if mddev_is_clustered().\n\nFixes: b97e92574c0b (\"Use separate bitmaps for each nodes in the cluster\")\nCc: stable@vger.kernel.org (v4.1+)\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "e8ff8bf09ff49733534ff3cee91bde030186055f", "tree": "a88b4a7f0d402a2aefeecc97b752c13e931874f6", "parents": [ "644df1a85fc4b0c7a16800f55717261546f4e651" ], "author": { "name": "Jes Sorensen", "email": "Jes.Sorensen@redhat.com", "time": "Wed Sep 16 10:20:05 2015 -0400" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:44 2015 +1000" }, "message": "md/raid1: Avoid raid1 resync getting stuck\n\nclose_sync() needs to set conf-\u003enext_resync to a large, but safe value\nbelow MaxSector and use it to determine whether or not to set\nstart_next_window in wait_barrier()\n\nSolution suggested by Neil Brown.\n\nReported-by: Nate Dailey \u003cnate.dailey@stratus.com\u003e\nTested-by: Xiao Ni \u003cxni@redhat.com\u003e\nSigned-off-by: Jes Sorensen \u003cJes.Sorensen@redhat.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "644df1a85fc4b0c7a16800f55717261546f4e651", "tree": "30ed717350163c534c5923d0557f5dafa4a43053", "parents": [ "d4929add83ad4660b1824a9282ab5dd4d60140fa" ], "author": { "name": "Julia Lawall", "email": "Julia.Lawall@lip6.fr", "time": "Sun Sep 13 14:15:10 2015 +0200" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:44 2015 +1000" }, "message": "md: drop null test before destroy functions\n\nRemove unneeded NULL test.\n\nThe semantic patch that makes this change is as follows:\n(http://coccinelle.lip6.fr/)\n\n// \u003csmpl\u003e\n@@ expression x; @@\n-if (x !\u003d NULL)\n \\(kmem_cache_destroy\\|mempool_destroy\\|dma_pool_destroy\\)(x);\n// \u003c/smpl\u003e\n\nSigned-off-by: Julia Lawall \u003cJulia.Lawall@lip6.fr\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "d4929add83ad4660b1824a9282ab5dd4d60140fa", "tree": "c087245ba01366fb5c568915fce7d353897cc7a1", "parents": [ "66eefe5de11db1e0d8f2edc3880d50e7c36a9d43" ], "author": { "name": "Shaohua Li", "email": "shli@fb.com", "time": "Fri Sep 18 10:20:12 2015 -0700" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:44 2015 +1000" }, "message": "md: clear CHANGE_PENDING in readonly array\n\nIf faulty disks of an array are more than allowed degraded number, the\narray enters error handling. It will be marked as read-only with\nMD_CHANGE_PENDING/RECOVERY_NEEDED set. But currently recovery doesn\u0027t\nclear CHANGE_PENDING bit for read-only array. If MD_CHANGE_PENDING is\nset for a raid5 array, all returned IO will be hold on a list till the\nbit is clear. But recovery nevery clears this bit, the IO is always in\npending state and nevery finish. This has bad effects like upper layer\ncan\u0027t get an IO error and the array can\u0027t be stopped.\n\nFixes: c3cce6cda162 (\"md/raid5: ensure device failure recorded before write request returns.\")\nSigned-off-by: Shaohua Li \u003cshli@fb.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "66eefe5de11db1e0d8f2edc3880d50e7c36a9d43", "tree": "f387b71a4c0eeaf506fb1cba79e9d6b7413dbbd2", "parents": [ "36707bb2e7c6730d79d6cdc6d1475d3d7e94c518" ], "author": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Thu Sep 24 15:47:47 2015 +1000" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:44 2015 +1000" }, "message": "md/raid0: apply base queue limits *before* disk_stack_limits\n\nCalling e.g. blk_queue_max_hw_sectors() after calls to\ndisk_stack_limits() discards the settings determined by\ndisk_stack_limits().\nSo we need to make those calls first.\n\nFixes: 199dc6ed5179 (\"md/raid0: update queue parameter in a safer location.\")\nCc: stable@vger.kernel.org (v2.6.35+ - please apply with 199dc6ed5179).\nReported-by: Jes Sorensen \u003cJes.Sorensen@redhat.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "36707bb2e7c6730d79d6cdc6d1475d3d7e94c518", "tree": "37ea9a4b0e716212bfd411f4c2d9e971ff8dc5f3", "parents": [ "ebda780bce8d58ec0abab157397c9e099c41a05f" ], "author": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Thu Sep 24 15:25:36 2015 +1000" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:43 2015 +1000" }, "message": "md/raid5: don\u0027t index beyond end of array in need_this_block().\n\nWhen need_this_block probably shouldn\u0027t be called when there\nare more than 2 failed devices, we really don\u0027t want it to try\nindexing beyond the end of the failed_num[] of fdev[] arrays.\n\nSo limit the loops to at most 2 iterations.\n\nReported-by: Shaohua Li \u003cshli@fb.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.de\u003e\n" }, { "commit": "ebda780bce8d58ec0abab157397c9e099c41a05f", "tree": "bb3995ca492ded73273329f69791cf814b5dfdbe", "parents": [ "88724bfa68be792c1487d759e87568c36ac1a1cc" ], "author": { "name": "Shaohua Li", "email": "shli@fb.com", "time": "Fri Sep 18 10:20:13 2015 -0700" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:43 2015 +1000" }, "message": "raid5: update analysis state for failed stripe\n\nhandle_failed_stripe() makes the stripe fail, eg, all IO will return\nwith a failure, but it doesn\u0027t update stripe_head_state. Later\nhandle_stripe() has special handling for raid6 for handle_stripe_fill().\nThat check before handle_stripe_fill() doesn\u0027t skip the failed stripe\nand we get a kernel crash in need_this_block. This patch clear the\nanalysis state to make sure no functions wrongly called after\nhandle_failed_stripe()\n\nSigned-off-by: Shaohua Li \u003cshli@fb.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "88724bfa68be792c1487d759e87568c36ac1a1cc", "tree": "12cf787a5922610396529d90ae82ccaf223a57d9", "parents": [ "bcee19f424a0d8c26ecf2607b73c690802658b29" ], "author": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Thu Sep 24 14:00:51 2015 +1000" }, "committer": { "name": "NeilBrown", "email": "neilb@suse.com", "time": "Fri Oct 02 17:23:43 2015 +1000" }, "message": "md: wait for pending superblock updates before switching to read-only\n\nIf a superblock update is pending, wait for it to complete before\nletting md_set_readonly() switch to readonly.\nOtherwise we might lose important information about a device having\nfailed.\n\nFor external arrays, waiting for superblock updates can wait on\nuser-space, so in that case, just return an error.\n\nReported-and-tested-by: Shaohua Li \u003cshli@fb.com\u003e\nSigned-off-by: NeilBrown \u003cneilb@suse.com\u003e\n" }, { "commit": "ab76f7b4ab2397ffdd2f1eb07c55697d19991d10", "tree": "47a592eec724e5bf0d89e8824fac1c9ad4bb8e6f", "parents": [ "e3c41e37b0f4b18cbd4dac76cbeece5a7558b909" ], "author": { "name": "Stephen Smalley", "email": "sds@tycho.nsa.gov", "time": "Thu Oct 01 09:04:22 2015 -0400" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Oct 02 09:21:06 2015 +0200" }, "message": "x86/mm: Set NX on gap between __ex_table and rodata\n\nUnused space between the end of __ex_table and the start of\nrodata can be left W+x in the kernel page tables. Extend the\nsetting of the NX bit to cover this gap by starting from\ntext_end rather than rodata_start.\n\n Before:\n ---[ High Kernel Mapping ]---\n 0xffffffff80000000-0xffffffff81000000 16M pmd\n 0xffffffff81000000-0xffffffff81600000 6M ro PSE GLB x pmd\n 0xffffffff81600000-0xffffffff81754000 1360K ro GLB x pte\n 0xffffffff81754000-0xffffffff81800000 688K RW GLB x pte\n 0xffffffff81800000-0xffffffff81a00000 2M ro PSE GLB NX pmd\n 0xffffffff81a00000-0xffffffff81b3b000 1260K ro GLB NX pte\n 0xffffffff81b3b000-0xffffffff82000000 4884K RW GLB NX pte\n 0xffffffff82000000-0xffffffff82200000 2M RW PSE GLB NX pmd\n 0xffffffff82200000-0xffffffffa0000000 478M pmd\n\n After:\n ---[ High Kernel Mapping ]---\n 0xffffffff80000000-0xffffffff81000000 16M pmd\n 0xffffffff81000000-0xffffffff81600000 6M ro PSE GLB x pmd\n 0xffffffff81600000-0xffffffff81754000 1360K ro GLB x pte\n 0xffffffff81754000-0xffffffff81800000 688K RW GLB NX pte\n 0xffffffff81800000-0xffffffff81a00000 2M ro PSE GLB NX pmd\n 0xffffffff81a00000-0xffffffff81b3b000 1260K ro GLB NX pte\n 0xffffffff81b3b000-0xffffffff82000000 4884K RW GLB NX pte\n 0xffffffff82000000-0xffffffff82200000 2M RW PSE GLB NX pmd\n 0xffffffff82200000-0xffffffffa0000000 478M pmd\n\nSigned-off-by: Stephen Smalley \u003csds@tycho.nsa.gov\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: linux-kernel@vger.kernel.org\nLink: http://lkml.kernel.org/r/1443704662-3138-1-git-send-email-sds@tycho.nsa.gov\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "e3c41e37b0f4b18cbd4dac76cbeece5a7558b909", "tree": "304c80cb661ee68130d3fb5dc6c5ddcc661701f0", "parents": [ "95c632f4e4e6365a20397f4fd04dcf27aab02958" ], "author": { "name": "Lee, Chun-Yi", "email": "joeyli.kernel@gmail.com", "time": "Tue Sep 29 20:58:57 2015 +0800" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Oct 02 09:13:06 2015 +0200" }, "message": "x86/kexec: Fix kexec crash in syscall kexec_file_load()\n\nThe original bug is a page fault crash that sometimes happens\non big machines when preparing ELF headers:\n\n BUG: unable to handle kernel paging request at ffffc90613fc9000\n IP: [\u003cffffffff8103d645\u003e] prepare_elf64_ram_headers_callback+0x165/0x260\n\nThe bug is caused by us under-counting the number of memory ranges\nand subsequently not allocating enough ELF header space for them.\nThe bug is typically masked on smaller systems, because the ELF header\nallocation is rounded up to the next page.\n\nThis patch modifies the code in fill_up_crash_elf_data() by using\nwalk_system_ram_res() instead of walk_system_ram_range() to correctly\ncount the max number of crash memory ranges. That\u0027s because the\nwalk_system_ram_range() filters out small memory regions that\nreside in the same page, but walk_system_ram_res() does not.\n\nHere\u0027s how I found the bug:\n\nAfter tracing prepare_elf64_headers() and prepare_elf64_ram_headers_callback(),\nthe code uses walk_system_ram_res() to fill-in crash memory regions information\nto the program header, so it counts those small memory regions that\nreside in a page area.\n\nBut, when the kernel was using walk_system_ram_range() in\nfill_up_crash_elf_data() to count the number of crash memory regions,\nit filters out small regions.\n\nI printed those small memory regions, for example:\n\n kexec: Get nr_ram ranges. vaddr\u003d0xffff880077592258 paddr\u003d0x77592258, sz\u003d0xdc0\n\nBased on the code in walk_system_ram_range(), this memory region\nwill be filtered out:\n\n pfn \u003d (0x77592258 + 0x1000 - 1) \u003e\u003e 12 \u003d 0x77593\n end_pfn \u003d (0x77592258 + 0xfc0 -1 + 1) \u003e\u003e 12 \u003d 0x77593\n end_pfn - pfn \u003d 0x77593 - 0x77593 \u003d 0 \u003c\u003d\u003d\u003d if (end_pfn \u003e pfn) is FALSE\n\nSo, the max_nr_ranges that\u0027s counted by the kernel doesn\u0027t include\nsmall memory regions - causing us to under-allocate the required space.\nThat causes the page fault crash that happens in a later code path\nwhen preparing ELF headers.\n\nThis bug is not easy to reproduce on small machines that have few\nCPUs, because the allocated page aligned ELF buffer has more free\nspace to cover those small memory regions\u0027 PT_LOAD headers.\n\nSigned-off-by: Lee, Chun-Yi \u003cjlee@suse.com\u003e\nCc: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Baoquan He \u003cbhe@redhat.com\u003e\nCc: Jiang Liu \u003cjiang.liu@linux.intel.com\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Stephen Rothwell \u003csfr@canb.auug.org.au\u003e\nCc: Takashi Iwai \u003ctiwai@suse.de\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Viresh Kumar \u003cviresh.kumar@linaro.org\u003e\nCc: Vivek Goyal \u003cvgoyal@redhat.com\u003e\nCc: kexec@lists.infradead.org\nCc: linux-kernel@vger.kernel.org\nCc: \u003cstable@vger.kernel.org\u003e\nLink: http://lkml.kernel.org/r/1443531537-29436-1-git-send-email-jlee@suse.com\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "ccf03d6995fa4b784f5b987726ba98f4859bf326", "tree": "6fdc0aabe330fac0ec6ddc9d1cbf54b0a000ae3b", "parents": [ "274d83524895fe41ca8debae4eec60ede7252bb5" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Oct 01 16:28:25 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:42 2015 +1000" }, "message": "drm/dp/mst: add some defines for logical/physical ports\n\nThis just removes the magic number.\n\nAcked-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "274d83524895fe41ca8debae4eec60ede7252bb5", "tree": "343a2a9e9cba7f74ac5446047d09c05c5055da0e", "parents": [ "d9515c5ec1a20c77d83471e634ad9bb12deb0eac" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 30 10:39:42 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:42 2015 +1000" }, "message": "drm/dp/mst: drop cancel work sync in the mstb destroy path (v2)\n\nSince 9eb1e57f564d4e6e10991402726cc83fe0b9172f\ndrm/dp/mst: make sure mst_primary mstb is valid in work function\n\nwe validate the mstb structs in the work function, and doing\nthat takes a reference. So we should never get here with the\nwork function running using the mstb device, only if the work\nfunction hasn\u0027t run yet or is running for another mstb.\n\nSo we don\u0027t need to sync the work here, this was causing\nlockdep spew as below.\n\n[ +0.000160] \u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n[ +0.000001] [ INFO: possible recursive locking detected ]\n[ +0.000002] 3.10.0-320.el7.rhel72.stable.backport.3.x86_64.debug #1 Tainted: G W ------------\n[ +0.000001] ---------------------------------------------\n[ +0.000001] kworker/4:2/1262 is trying to acquire lock:\n[ +0.000001] ((\u0026mgr-\u003ework)){+.+.+.}, at: [\u003cffffffff810b29a5\u003e] flush_work+0x5/0x2e0\n[ +0.000007]\nbut task is already holding lock:\n[ +0.000001] ((\u0026mgr-\u003ework)){+.+.+.}, at: [\u003cffffffff810b57e4\u003e] process_one_work+0x1b4/0x710\n[ +0.000004]\nother info that might help us debug this:\n[ +0.000001] Possible unsafe locking scenario:\n\n[ +0.000002] CPU0\n[ +0.000000] ----\n[ +0.000001] lock((\u0026mgr-\u003ework));\n[ +0.000002] lock((\u0026mgr-\u003ework));\n[ +0.000001]\n *** DEADLOCK ***\n\n[ +0.000001] May be due to missing lock nesting notation\n\n[ +0.000002] 2 locks held by kworker/4:2/1262:\n[ +0.000001] #0: (events_long){.+.+.+}, at: [\u003cffffffff810b57e4\u003e] process_one_work+0x1b4/0x710\n[ +0.000004] #1: ((\u0026mgr-\u003ework)){+.+.+.}, at: [\u003cffffffff810b57e4\u003e] process_one_work+0x1b4/0x710\n[ +0.000003]\nstack backtrace:\n[ +0.000003] CPU: 4 PID: 1262 Comm: kworker/4:2 Tainted: G W ------------ 3.10.0-320.el7.rhel72.stable.backport.3.x86_64.debug #1\n[ +0.000001] Hardware name: LENOVO 20EGS0R600/20EGS0R600, BIOS GNET71WW (2.19 ) 02/05/2015\n[ +0.000008] Workqueue: events_long drm_dp_mst_link_probe_work [drm_kms_helper]\n[ +0.000001] ffffffff82c26c90 00000000a527b914 ffff88046399bae8 ffffffff816fe04d\n[ +0.000004] ffff88046399bb58 ffffffff8110f47f ffff880461438000 0001009b840fc003\n[ +0.000002] ffff880461438a98 0000000000000000 0000000804dc26e1 ffffffff824a2c00\n[ +0.000003] Call Trace:\n[ +0.000004] [\u003cffffffff816fe04d\u003e] dump_stack+0x19/0x1b\n[ +0.000004] [\u003cffffffff8110f47f\u003e] __lock_acquire+0x115f/0x1250\n[ +0.000002] [\u003cffffffff8110fd49\u003e] lock_acquire+0x99/0x1e0\n[ +0.000002] [\u003cffffffff810b29a5\u003e] ? flush_work+0x5/0x2e0\n[ +0.000002] [\u003cffffffff810b29ee\u003e] flush_work+0x4e/0x2e0\n[ +0.000002] [\u003cffffffff810b29a5\u003e] ? flush_work+0x5/0x2e0\n[ +0.000004] [\u003cffffffff81025905\u003e] ? native_sched_clock+0x35/0x80\n[ +0.000002] [\u003cffffffff81025959\u003e] ? sched_clock+0x9/0x10\n[ +0.000002] [\u003cffffffff810da1f5\u003e] ? local_clock+0x25/0x30\n[ +0.000002] [\u003cffffffff8110dca9\u003e] ? mark_held_locks+0xb9/0x140\n[ +0.000003] [\u003cffffffff810b4ed5\u003e] ? __cancel_work_timer+0x95/0x160\n[ +0.000002] [\u003cffffffff810b4ee8\u003e] __cancel_work_timer+0xa8/0x160\n[ +0.000002] [\u003cffffffff810b4fb0\u003e] cancel_work_sync+0x10/0x20\n[ +0.000007] [\u003cffffffffa0160d17\u003e] drm_dp_destroy_mst_branch_device+0x27/0x120 [drm_kms_helper]\n[ +0.000006] [\u003cffffffffa0163968\u003e] drm_dp_mst_link_probe_work+0x78/0xa0 [drm_kms_helper]\n[ +0.000002] [\u003cffffffff810b5850\u003e] process_one_work+0x220/0x710\n[ +0.000002] [\u003cffffffff810b57e4\u003e] ? process_one_work+0x1b4/0x710\n[ +0.000005] [\u003cffffffff810b5e5b\u003e] worker_thread+0x11b/0x3a0\n[ +0.000003] [\u003cffffffff810b5d40\u003e] ? process_one_work+0x710/0x710\n[ +0.000002] [\u003cffffffff810beced\u003e] kthread+0xed/0x100\n[ +0.000003] [\u003cffffffff810bec00\u003e] ? insert_kthread_work+0x80/0x80\n[ +0.000003] [\u003cffffffff817121d8\u003e] ret_from_fork+0x58/0x90\n\nv2: add flush_work.\n\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "d9515c5ec1a20c77d83471e634ad9bb12deb0eac", "tree": "fc0f8cb6098aa942f4dfe6749fb8c55b34161a34", "parents": [ "68d8c9fc91a0f63a2a10ccf7adf56f69125c11c1" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 16 17:55:23 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:41 2015 +1000" }, "message": "drm/dp/mst: split connector registration into two parts (v2)\n\nIn order to cache the EDID properly for tiled displays, we\nneed to retrieve it before we register the connector with\nuserspace, otherwise userspace can call get resources\nand try and get the edid before we\u0027ve even cached it.\n\nThis fixes some problems when hotplugging mst monitors,\nwith X/mutter running. As mutter seems to get 0 modes\nfor one of the monitors in the tile.\n\nv2: fix warning in radeon\nhandle tile setting in cached path rather than\nget edid path.\n\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "68d8c9fc91a0f63a2a10ccf7adf56f69125c11c1", "tree": "800a523c72e47843484f4d91a088562b0713da21", "parents": [ "df4839fdc9b3c922586b945f062f38cbbda022bb" ], "author": { "name": "Dave Airlie", "email": "airlied@gmail.com", "time": "Sun Sep 06 18:53:00 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:41 2015 +1000" }, "message": "drm/dp/mst: update the link_address_sent before sending the link address (v3)\n\nUpdate the state before sending the msg to close it.\n\nv2: reset value if return indicates we haven\u0027t send the msg.\nv3: just clean the code up.\nPointed out by Adam J Richter on\n\nBugzilla: https://bugs.freedesktop.org/show_bug.cgi?id\u003d91481\n\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "df4839fdc9b3c922586b945f062f38cbbda022bb", "tree": "8ff48911fd93882d61dbaee4ee086e9300717972", "parents": [ "1c960876be7cffd2798a9e2be090e0a5afaee895" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 16 10:37:28 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:40 2015 +1000" }, "message": "drm/dp/mst: fixup handling hotplug on port removal.\n\noutput ports should always have a connector, unless\nin the rare case connector allocation fails in the\ndriver.\n\nIn this case we only need to teardown the pdt,\nand free the struct, and there is no need to\nsend a hotplug msg.\n\nIn the case were we add the port to the destroy\nlist we need to send a hotplug if we destroy\nany connectors, so userspace knows to reprobe\nstuff.\n\nthis patch also handles port-\u003econnector allocation\nfailing which should be a rare event, but makes\nthe code consistent.\n\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "1c960876be7cffd2798a9e2be090e0a5afaee895", "tree": "6ad6d29068955e5471e5fcec6bbf10701532cf87", "parents": [ "0c6dadbe79c54818ed4c268297b663f0c4d1ef98" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Wed Sep 16 11:04:49 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:34:40 2015 +1000" }, "message": "drm/dp/mst: don\u0027t pass port into the path builder function\n\nThis is unnecessary and it makes it easier to see what is needed\nfrom port.\n\nalso add blank line to make things nicer.\n\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "0c6dadbe79c54818ed4c268297b663f0c4d1ef98", "tree": "09a4aefc9cf9155c2429c3490fa1f8782acee47b", "parents": [ "03f9abb28c3e10a1d0e08df8bca69e0606b23ea6" ], "author": { "name": "Alex Deucher", "email": "alexdeucher@gmail.com", "time": "Wed Sep 30 14:47:38 2015 -0400" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:31:00 2015 +1000" }, "message": "drm/radeon: drop radeon_fb_helper_set_par\n\nIt was just a wrapper around drm_fb_helper_set_par that\ncalled cursor_set2 in addition. Now that the core handles\nthis, drop this radeon specific version.\n\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Michel Dänzer \u003cmichel.daenzer@amd.com\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "03f9abb28c3e10a1d0e08df8bca69e0606b23ea6", "tree": "9a4c60a30b1d58e55c74e53fad29dd01decaaddd", "parents": [ "ec403b89e3152f93199b7b1813148bcdf6829311" ], "author": { "name": "Alex Deucher", "email": "alexdeucher@gmail.com", "time": "Wed Sep 30 14:47:37 2015 -0400" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 15:30:41 2015 +1000" }, "message": "drm: handle cursor_set2 in restore_fbdev_mode\n\nIf a driver uses the cursor_set2 crtc callback rather than\ncursor_set, use that. This fixes the fbdev helper for drivers\nthat use cursor_set2.\n\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Michel Dänzer \u003cmichel.daenzer@amd.com\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "bde17b90dd9712cb61a7ab0c1ccd0f7f6aa57957", "tree": "d4597bfde4f5d4a3a5729e5534c44993c6216352", "parents": [ "1bca1000fa71a1092947b4a51928abe80a3316d2", "676bd99178cd962ed24ffdad222b7069d330a969" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 22:20:11 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 22:20:11 2015 -0400" }, "message": "Merge branch \u0027akpm\u0027 (patches from Andrew)\n\nMerge misc fixes from Andrew Morton:\n \"12 fixes\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e:\n dmapool: fix overflow condition in pool_find_page()\n thermal: avoid division by zero in power allocator\n memcg: remove pcp_counter_lock\n kprobes: use _do_fork() in samples to make them work again\n drivers/input/joystick/Kconfig: zhenhua.c needs BITREVERSE\n memcg: make mem_cgroup_read_stat() unsigned\n memcg: fix dirty page migration\n dax: fix NULL pointer in __dax_pmd_fault()\n mm: hugetlbfs: skip shared VMAs when unmapping private pages to satisfy a fault\n mm/slab: fix unexpected index mapping result of kmalloc_size(INDEX_NODE+1)\n userfaultfd: remove kernel header include from uapi header\n arch/x86/include/asm/efi.h: fix build failure\n" }, { "commit": "1bca1000fa71a1092947b4a51928abe80a3316d2", "tree": "d229750e2baeeba923722697fd0c40d4288442fc", "parents": [ "3deaa4f531506a12ac4860ccd83cb6cbcb15a7eb", "eb6d1c287ae1f7221248d5be26a5b1560073c09e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 22:06:40 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 22:06:40 2015 -0400" }, "message": "Merge tag \u0027pm+acpi-4.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management and ACPI fixes from Rafael Wysocki:\n \"These are fixes mostly, for a few changes made in this cycle (the\n intel_idle driver, the OPP library, the ACPI EC driver, turbostat) and\n for some issues that have just been discovered (ACPI PCI IRQ\n management, PCI power management documentation, turbostat), with a\n couple of cleanups on top of them.\n\n Specifics:\n\n - intel_idle driver fixup for the recently added Skylake chips\n support (Len Brown).\n\n - Operating Performance Points (OPP) library fix related to the\n recently added support for new DT bindings and a fix for a typo in\n a comment (Viresh Kumar, Stephen Boyd).\n\n - ACPI EC driver fix for a recently introduced memory leak in an\n error code path (Lv Zheng).\n\n - ACPI PCI IRQ management fix for the issue where an ISA IRQ is\n shared with a PCI device which requires it to be configured in a\n different way and may cause an interrupt storm to happen as a\n result with an extra ACPI SCI IRQ handling simplification on top of\n it (Jiang Liu).\n\n - Update of the PCI power management documentation that became\n outdated and started to actively confuse the readers to make it\n actually reflect the code (Rafael J Wysocki).\n\n - turbostat fixes including an IVB Xeon regression fix (related to\n the --debug command line option), Skylake adjustment for the TSC\n running at a frequency that doesn\u0027t match the base one exactly, and\n a Knights Landing quirk to account for the fact that it only\n updates APERF and MPERF every 1024 clock cycles plus bumping up the\n turbostat version number (Len Brown, Hubert Chrzaniuk)\"\n\n* tag \u0027pm+acpi-4.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n tools/power turbosat: update version number\n tools/power turbostat: SKL: Adjust for TSC difference from base frequency\n tools/power turbostat: KNL workaround for %Busy and Avg_MHz\n tools/power turbostat: IVB Xeon: fix --debug regression\n ACPI / PCI: Remove duplicated penalty on SCI IRQ\n ACPI, PCI, irq: Do not share PCI IRQ with ISA IRQ\n ACPI / EC: Fix a memory leak issue in acpi_ec_query()\n PM / OPP: Fix typo modifcation -\u003e modification\n PCI / PM: Update runtime PM documentation for PCI devices\n PM / OPP: of_property_count_u32_elems() can return errors\n intel_idle: Skylake Client Support - updated\n" }, { "commit": "3deaa4f531506a12ac4860ccd83cb6cbcb15a7eb", "tree": "9bad90fe23c061fab0ab1df71621875c866db6cf", "parents": [ "ccf70ddcbe9984cee406be2bacfedd5e4776919d", "b84f78782052ee4516903e5d0566a5eee365b771" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:55:35 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:55:35 2015 -0400" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nPull networking fixes from David Miller:\n\n1) Fix regression in SKB partial checksum handling, from Pravin B\n Shalar.\n\n2) Fix VLAN inside of VXLAN handling in i40e driver, from Jesse\n Brandeburg.\n\n3) Cure softlockups during accept() in SCTP, from Karl Heiss.\n\n4) MSG_PEEK should return multiple SKBs worth of data in AF_UNIX, from\n Aaron Conole.\n\n5) IPV6 erroneously ignores output interface specifier in lookup key for\n route lookups, fix from David Ahern.\n\n6) In Marvell DSA driver, forward unknown frames to CPU port, from\n Andrew Lunn.\n\n7) Mission flow flag initializations in some code paths, from David\n Ahern.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:\n net: Initialize flow flags in input path\n net: dsa: fix preparation of a port STP update\n testptp: Silence compiler warnings on ppc64\n net/mlx4: Handle return codes in mlx4_qp_attach_common\n dsa: mv88e6xxx: Enable forwarding for unknown to the CPU port\n skbuff: Fix skb checksum partial check.\n net: ipv6: Add RT6_LOOKUP_F_IFACE flag if oif is set\n net sysfs: Print link speed as signed integer\n bna: fix error handling\n af_unix: return data from multiple SKBs on recv() with MSG_PEEK flag\n af_unix: Convert the unix_sk macro to an inline function for type safety\n net: sctp: Don\u0027t use 64 kilobyte lookup table for four elements\n l2tp: protect tunnel-\u003edel_work by ref_count\n net/ibm/emac: bump version numbers for correct work with ethtool\n sctp: Prevent soft lockup when sctp_accept() is called during a timeout event\n sctp: Whitespace fix\n i40e/i40evf: check for stopped admin queue\n i40e: fix VLAN inside VXLAN\n r8169: fix handling rtl_readphy result\n net: hisilicon: fix handling platform_get_irq result\n" }, { "commit": "676bd99178cd962ed24ffdad222b7069d330a969", "tree": "ca76fb3382ada58d059dae22fb60f18b7437c8d4", "parents": [ "44241628bb207ec211bebd156aaf69470d90c209" ], "author": { "name": "Robin Murphy", "email": "robin.murphy@arm.com", "time": "Thu Oct 01 15:37:19 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "dmapool: fix overflow condition in pool_find_page()\n\nIf a DMA pool lies at the very top of the dma_addr_t range (as may\nhappen with an IOMMU involved), the calculated end address of the pool\nwraps around to zero, and page lookup always fails.\n\nTweak the relevant calculation to be overflow-proof.\n\nSigned-off-by: Robin Murphy \u003crobin.murphy@arm.com\u003e\nCc: Arnd Bergmann \u003carnd@arndb.de\u003e\nCc: Marek Szyprowski \u003cm.szyprowski@samsung.com\u003e\nCc: Sumit Semwal \u003csumit.semwal@linaro.org\u003e\nCc: Sakari Ailus \u003csakari.ailus@iki.fi\u003e\nCc: Russell King \u003crmk+kernel@arm.linux.org.uk\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "44241628bb207ec211bebd156aaf69470d90c209", "tree": "2b0c143661cf3d9fbc66f5241854df85f0164a78", "parents": [ "ef510194cefe0cd369ef73419cd65b0a5bb4fb5b" ], "author": { "name": "Andrea Arcangeli", "email": "aarcange@redhat.com", "time": "Thu Oct 01 15:37:16 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "thermal: avoid division by zero in power allocator\n\nDuring boot I get a div by zero Oops regression starting in v4.3-rc3.\n\nSigned-off-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nReviewed-by: Javi Merino \u003cjavi.merino@arm.com\u003e\nCc: Zhang Rui \u003crui.zhang@intel.com\u003e\nCc: Eduardo Valentin \u003cedubezval@gmail.com\u003e\nCc: Daniel Kurtz \u003cdjkurtz@chromium.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ef510194cefe0cd369ef73419cd65b0a5bb4fb5b", "tree": "0763e369e1af60e90021edb2852da05d821f1442", "parents": [ "54aea4542980a3ed580426a81c5af799df4d610d" ], "author": { "name": "Greg Thelen", "email": "gthelen@google.com", "time": "Thu Oct 01 15:37:13 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "memcg: remove pcp_counter_lock\n\nCommit 733a572e66d2 (\"memcg: make mem_cgroup_read_{stat|event}() iterate\npossible cpus instead of online\") removed the last use of the per memcg\npcp_counter_lock but forgot to remove the variable.\n\nKill the vestigial variable.\n\nSigned-off-by: Greg Thelen \u003cgthelen@google.com\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.com\u003e\nAcked-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "54aea4542980a3ed580426a81c5af799df4d610d", "tree": "6f9ca54438333b0ee829ec57f56712359447d33b", "parents": [ "09a59a9d57a9d6f49510c93304d6e105deb83b93" ], "author": { "name": "Petr Mladek", "email": "pmladek@suse.com", "time": "Thu Oct 01 15:37:11 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "kprobes: use _do_fork() in samples to make them work again\n\nCommit 3033f14ab78c (\"clone: support passing tls argument via C rather\nthan pt_regs magic\") introduced _do_fork() that allowed to pass @tls\nparameter.\n\nThe old do_fork() is defined only for architectures that are not ready\nto use this way and do not define HAVE_COPY_THREAD_TLS.\n\nLet\u0027s use _do_fork() in the kprobe examples to make them work again on\nall architectures.\n\nSigned-off-by: Petr Mladek \u003cpmladek@suse.com\u003e\nCc: Ingo Molnar \u003cmingo@redhat.com\u003e\nCc: Masami Hiramatsu \u003cmasami.hiramatsu.pt@hitachi.com\u003e\nCc: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: \"H. Peter Anvin\" \u003chpa@zytor.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Thiago Macieira \u003cthiago.macieira@intel.com\u003e\nCc: Jiri Kosina \u003cjkosina@suse.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "09a59a9d57a9d6f49510c93304d6e105deb83b93", "tree": "adf118e46559e078361486776f707b46e53bbd50", "parents": [ "484ebb3b8c8b27dd2171696462a3116edb9ff801" ], "author": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Thu Oct 01 15:37:08 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "drivers/input/joystick/Kconfig: zhenhua.c needs BITREVERSE\n\nIt uses bitrev8(), so it must ensure that lib/bitrev.o gets included in\nvmlinux.\n\nCc: Fengguang Wu \u003cfengguang.wu@gmail.com\u003e\nCc: yalin wang \u003cyalin.wang2010@gmail.com\u003e\nCc: Dmitry Torokhov \u003cdmitry.torokhov@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "484ebb3b8c8b27dd2171696462a3116edb9ff801", "tree": "1056eab0960e2ec868c6a5cfd3076d05ee5887a3", "parents": [ "0610c25daa3e76e38ad5a8fae683a89ff9f71798" ], "author": { "name": "Greg Thelen", "email": "gthelen@google.com", "time": "Thu Oct 01 15:37:05 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "memcg: make mem_cgroup_read_stat() unsigned\n\nmem_cgroup_read_stat() returns a page count by summing per cpu page\ncounters. The summing is racy wrt. updates, so a transient negative\nsum is possible. Callers don\u0027t want negative values:\n\n - mem_cgroup_wb_stats() doesn\u0027t want negative nr_dirty or nr_writeback.\n This could confuse dirty throttling.\n\n - oom reports and memory.stat shouldn\u0027t show confusing negative usage.\n\n - tree_usage() already avoids negatives.\n\nAvoid returning negative page counts from mem_cgroup_read_stat() and\nconvert it to unsigned.\n\n[akpm@linux-foundation.org: fix old typo while we\u0027re in there]\nSigned-off-by: Greg Thelen \u003cgthelen@google.com\u003e\nCc: Johannes Weiner \u003channes@cmpxchg.org\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\t[4.2+]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "0610c25daa3e76e38ad5a8fae683a89ff9f71798", "tree": "4e37a050e59f20773ec59af2191b9bc6b9da66fe", "parents": [ "8346c416d17bf5b4ea1508662959bb62e73fd6a5" ], "author": { "name": "Greg Thelen", "email": "gthelen@google.com", "time": "Thu Oct 01 15:37:02 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "memcg: fix dirty page migration\n\nThe problem starts with a file backed dirty page which is charged to a\nmemcg. Then page migration is used to move oldpage to newpage.\n\nMigration:\n - copies the oldpage\u0027s data to newpage\n - clears oldpage.PG_dirty\n - sets newpage.PG_dirty\n - uncharges oldpage from memcg\n - charges newpage to memcg\n\nClearing oldpage.PG_dirty decrements the charged memcg\u0027s dirty page\ncount.\n\nHowever, because newpage is not yet charged, setting newpage.PG_dirty\ndoes not increment the memcg\u0027s dirty page count. After migration\ncompletes newpage.PG_dirty is eventually cleared, often in\naccount_page_cleaned(). At this time newpage is charged to a memcg so\nthe memcg\u0027s dirty page count is decremented which causes underflow\nbecause the count was not previously incremented by migration. This\nunderflow causes balance_dirty_pages() to see a very large unsigned\nnumber of dirty memcg pages which leads to aggressive throttling of\nbuffered writes by processes in non root memcg.\n\nThis issue:\n - can harm performance of non root memcg buffered writes.\n - can report too small (even negative) values in\n memory.stat[(total_)dirty] counters of all memcg, including the root.\n\nTo avoid polluting migrate.c with #ifdef CONFIG_MEMCG checks, introduce\npage_memcg() and set_page_memcg() helpers.\n\nTest:\n 0) setup and enter limited memcg\n mkdir /sys/fs/cgroup/test\n echo 1G \u003e /sys/fs/cgroup/test/memory.limit_in_bytes\n echo $$ \u003e /sys/fs/cgroup/test/cgroup.procs\n\n 1) buffered writes baseline\n dd if\u003d/dev/zero of\u003d/data/tmp/foo bs\u003d1M count\u003d1k\n sync\n grep ^dirty /sys/fs/cgroup/test/memory.stat\n\n 2) buffered writes with compaction antagonist to induce migration\n yes 1 \u003e /proc/sys/vm/compact_memory \u0026\n rm -rf /data/tmp/foo\n dd if\u003d/dev/zero of\u003d/data/tmp/foo bs\u003d1M count\u003d1k\n kill %\n sync\n grep ^dirty /sys/fs/cgroup/test/memory.stat\n\n 3) buffered writes without antagonist, should match baseline\n rm -rf /data/tmp/foo\n dd if\u003d/dev/zero of\u003d/data/tmp/foo bs\u003d1M count\u003d1k\n sync\n grep ^dirty /sys/fs/cgroup/test/memory.stat\n\n (speed, dirty residue)\n unpatched patched\n 1) 841 MB/s 0 dirty pages 886 MB/s 0 dirty pages\n 2) 611 MB/s -33427456 dirty pages 793 MB/s 0 dirty pages\n 3) 114 MB/s -33427456 dirty pages 891 MB/s 0 dirty pages\n\n Notice that unpatched baseline performance (1) fell after\n migration (3): 841 -\u003e 114 MB/s. In the patched kernel, post\n migration performance matches baseline.\n\nFixes: c4843a7593a9 (\"memcg: add per cgroup dirty page accounting\")\nSigned-off-by: Greg Thelen \u003cgthelen@google.com\u003e\nReported-by: Dave Hansen \u003cdave.hansen@intel.com\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.com\u003e\nAcked-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\t[4.2+]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "8346c416d17bf5b4ea1508662959bb62e73fd6a5", "tree": "059bcaf04ab8cce9fdc5b3d6698044da7ade9de8", "parents": [ "2f84a8990ebbe235c59716896e017c6b2ca1200f" ], "author": { "name": "Ross Zwisler", "email": "ross.zwisler@linux.intel.com", "time": "Thu Oct 01 15:36:59 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "dax: fix NULL pointer in __dax_pmd_fault()\n\nCommit 46c043ede471 (\"mm: take i_mmap_lock in unmap_mapping_range() for\nDAX\") moved some code in __dax_pmd_fault() that was responsible for\nzeroing newly allocated PMD pages. The new location didn\u0027t properly set\nup \u0027kaddr\u0027, so when run this code resulted in a NULL pointer BUG.\n\nFix this by getting the correct \u0027kaddr\u0027 via bdev_direct_access().\n\nSigned-off-by: Ross Zwisler \u003cross.zwisler@linux.intel.com\u003e\nReported-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\nReviewed-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\nCc: Alexander Viro \u003cviro@zeniv.linux.org.uk\u003e\nCc: Matthew Wilcox \u003cwilly@linux.intel.com\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Dave Chinner \u003cdavid@fromorbit.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "2f84a8990ebbe235c59716896e017c6b2ca1200f", "tree": "556282bc5eab5c75394236f796d5120d48718f99", "parents": [ "03a2d2a3eafe4015412cf4e9675ca0e2d9204074" ], "author": { "name": "Mel Gorman", "email": "mgorman@techsingularity.net", "time": "Thu Oct 01 15:36:57 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "mm: hugetlbfs: skip shared VMAs when unmapping private pages to satisfy a fault\n\nSunDong reported the following on\n\n https://bugzilla.kernel.org/show_bug.cgi?id\u003d103841\n\n\tI think I find a linux bug, I have the test cases is constructed. I\n\tcan stable recurring problems in fedora22(4.0.4) kernel version,\n\tarch for x86_64. I construct transparent huge page, when the parent\n\tand child process with MAP_SHARE, MAP_PRIVATE way to access the same\n\thuge page area, it has the opportunity to lead to huge page copy on\n\twrite failure, and then it will munmap the child corresponding mmap\n\tarea, but then the child mmap area with VM_MAYSHARE attributes, child\n\tprocess munmap this area can trigger VM_BUG_ON in set_vma_resv_flags\n\tfunctions (vma - \u003e vm_flags \u0026 VM_MAYSHARE).\n\nThere were a number of problems with the report (e.g. it\u0027s hugetlbfs that\ntriggers this, not transparent huge pages) but it was fundamentally\ncorrect in that a VM_BUG_ON in set_vma_resv_flags() can be triggered that\nlooks like this\n\n\t vma ffff8804651fd0d0 start 00007fc474e00000 end 00007fc475e00000\n\t next ffff8804651fd018 prev ffff8804651fd188 mm ffff88046b1b1800\n\t prot 8000000000000027 anon_vma (null) vm_ops ffffffff8182a7a0\n\t pgoff 0 file ffff88106bdb9800 private_data (null)\n\t flags: 0x84400fb(read|write|shared|mayread|maywrite|mayexec|mayshare|dontexpand|hugetlb)\n\t ------------\n\t kernel BUG at mm/hugetlb.c:462!\n\t SMP\n\t Modules linked in: xt_pkttype xt_LOG xt_limit [..]\n\t CPU: 38 PID: 26839 Comm: map Not tainted 4.0.4-default #1\n\t Hardware name: Dell Inc. PowerEdge R810/0TT6JF, BIOS 2.7.4 04/26/2012\n\t set_vma_resv_flags+0x2d/0x30\n\nThe VM_BUG_ON is correct because private and shared mappings have\ndifferent reservation accounting but the warning clearly shows that the\nVMA is shared.\n\nWhen a private COW fails to allocate a new page then only the process\nthat created the VMA gets the page -- all the children unmap the page.\nIf the children access that data in the future then they get killed.\n\nThe problem is that the same file is mapped shared and private. During\nthe COW, the allocation fails, the VMAs are traversed to unmap the other\nprivate pages but a shared VMA is found and the bug is triggered. This\npatch identifies such VMAs and skips them.\n\nSigned-off-by: Mel Gorman \u003cmgorman@techsingularity.net\u003e\nReported-by: SunDong \u003csund_sky@126.com\u003e\nReviewed-by: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Hugh Dickins \u003chughd@google.com\u003e\nCc: Naoya Horiguchi \u003cn-horiguchi@ah.jp.nec.com\u003e\nCc: David Rientjes \u003crientjes@google.com\u003e\nReviewed-by: Naoya Horiguchi \u003cn-horiguchi@ah.jp.nec.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": "03a2d2a3eafe4015412cf4e9675ca0e2d9204074", "tree": "b21da286900ea88617157887f564403a5957b901", "parents": [ "9ff42d10c3b3e26d9555878f31b9a2e5c24efa57" ], "author": { "name": "Joonsoo Kim", "email": "js1304@gmail.com", "time": "Thu Oct 01 15:36:54 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "mm/slab: fix unexpected index mapping result of kmalloc_size(INDEX_NODE+1)\n\nCommit description is copied from the original post of this bug:\n\n http://comments.gmane.org/gmane.linux.kernel.mm/135349\n\nKernels after v3.9 use kmalloc_size(INDEX_NODE + 1) to get the next\nlarger cache size than the size index INDEX_NODE mapping. In kernels\n3.9 and earlier we used malloc_sizes[INDEX_L3 + 1].cs_size.\n\nHowever, sometimes we can\u0027t get the right output we expected via\nkmalloc_size(INDEX_NODE + 1), causing a BUG().\n\nThe mapping table in the latest kernel is like:\n index \u003d {0, 1, 2 , 3, 4, 5, 6, n}\n size \u003d {0, 96, 192, 8, 16, 32, 64, 2^n}\nThe mapping table before 3.10 is like this:\n index \u003d {0 , 1 , 2, 3, 4 , 5 , 6, n}\n size \u003d {32, 64, 96, 128, 192, 256, 512, 2^(n+3)}\n\nThe problem on my mips64 machine is as follows:\n\n(1) When configured DEBUG_SLAB \u0026\u0026 DEBUG_PAGEALLOC \u0026\u0026 DEBUG_LOCK_ALLOC\n \u0026\u0026 DEBUG_SPINLOCK, the sizeof(struct kmem_cache_node) will be \"150\",\n and the macro INDEX_NODE turns out to be \"2\": #define INDEX_NODE\n kmalloc_index(sizeof(struct kmem_cache_node))\n\n(2) Then the result of kmalloc_size(INDEX_NODE + 1) is 8.\n\n(3) Then \"if(size \u003e\u003d kmalloc_size(INDEX_NODE + 1)\" will lead to \"size\n \u003d PAGE_SIZE\".\n\n(4) Then \"if ((size \u003e\u003d (PAGE_SIZE \u003e\u003e 3))\" test will be satisfied and\n \"flags |\u003d CFLGS_OFF_SLAB\" will be covered.\n\n(5) if (flags \u0026 CFLGS_OFF_SLAB)\" test will be satisfied and will go to\n \"cachep-\u003eslabp_cache \u003d kmalloc_slab(slab_size, 0u)\", and the result\n here may be NULL while kernel bootup.\n\n(6) Finally,\"BUG_ON(ZERO_OR_NULL_PTR(cachep-\u003eslabp_cache));\" causes the\n BUG info as the following shows (may be only mips64 has this problem):\n\nThis patch fixes the problem of kmalloc_size(INDEX_NODE + 1) and removes\nthe BUG by adding \u0027size \u003e\u003d 256\u0027 check to guarantee that all necessary\nsmall sized slabs are initialized regardless sequence of slab size in\nmapping table.\n\nFixes: e33660165c90 (\"slab: Use common kmalloc_index/kmalloc_size...\")\nSigned-off-by: Joonsoo Kim \u003ciamjoonsoo.kim@lge.com\u003e\nReported-by: Liuhailong \u003cliu.hailong6@zte.com.cn\u003e\nAcked-by: Christoph Lameter \u003ccl@linux.com\u003e\nCc: Pekka Enberg \u003cpenberg@kernel.org\u003e\nCc: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "9ff42d10c3b3e26d9555878f31b9a2e5c24efa57", "tree": "83c7c568df204d7a2c07396b1edd975df3b052ef", "parents": [ "a523841ee4e506fa1f05ff3a85b1e6d8176a3d4d" ], "author": { "name": "Andre Przywara", "email": "andre.przywara@arm.com", "time": "Thu Oct 01 15:36:51 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "userfaultfd: remove kernel header include from uapi header\n\nAs include/uapi/linux/userfaultfd.h is a user visible header file, it\nshould not include kernel-exclusive header files.\n\nSo trying to build the userfaultfd test program from the selftests\ndirectory fails, since it contains a reference to linux/compiler.h. As\nit turns out, that header is not really needed there, so we can simply\nremove it to fix that issue.\n\nSigned-off-by: Andre Przywara \u003candre.przywara@arm.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Shuah Khan \u003cshuahkh@osg.samsung.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "a523841ee4e506fa1f05ff3a85b1e6d8176a3d4d", "tree": "4eec3889a7283cb9549e4a123ca444d8b75e15c1", "parents": [ "f97b870eced0ec562f953d32eda03906c7dacad6" ], "author": { "name": "Andrey Ryabinin", "email": "ryabinin.a.a@gmail.com", "time": "Thu Oct 01 15:36:48 2015 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 21:42:35 2015 -0400" }, "message": "arch/x86/include/asm/efi.h: fix build failure\n\nWith KMEMCHECK\u003dy, KASAN\u003dn:\n\n arch/x86/platform/efi/efi.c:673:3: error: implicit declaration of function `memcpy\u0027 [-Werror\u003dimplicit-function-declaration]\n arch/x86/platform/efi/efi_64.c:139:2: error: implicit declaration of function `memcpy\u0027 [-Werror\u003dimplicit-function-declaration]\n arch/x86/include/asm/desc.h:121:2: error: implicit declaration of function `memcpy\u0027 [-Werror\u003dimplicit-function-declaration]\n\nDon\u0027t #undef memcpy if KASAN\u003dn.\n\nFixes: 769a8089c1fd (\"x86, efi, kasan: #undef memset/memcpy/memmove per arch\")\nSigned-off-by: Andrey Ryabinin \u003cryabinin.a.a@gmail.com\u003e\nReported-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nReported-by: Sedat Dilek \u003csedat.dilek@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ec403b89e3152f93199b7b1813148bcdf6829311", "tree": "65c9140f32b844f9dd4f13aa62920c7b8f70103d", "parents": [ "62886a367b59d7740f9db89fd418ab3e450ab7c7", "b94be972538f4523f09243af7c726158f0bfae33" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:47:29 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:47:29 2015 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2015-10-01\u0027 of git://anongit.freedesktop.org/drm-intel into drm-fixes\n\na few i915 fixes for v4.3.\n\n* tag \u0027drm-intel-fixes-2015-10-01\u0027 of git://anongit.freedesktop.org/drm-intel:\n drm/i915: Call non-locking version of drm_kms_helper_poll_enable(), v2\n drm: Add a non-locking version of drm_kms_helper_poll_enable(), v2\n drm/i915: Consider HW CSB write pointer before resetting the sw read pointer\n drm/i915/skl: Don\u0027t call intel_prepare_ddi when encoder list isn\u0027t yet initialized.\n" }, { "commit": "62886a367b59d7740f9db89fd418ab3e450ab7c7", "tree": "ecb688ef9a5943715cb7c55373aeebc4256a9ac1", "parents": [ "8e592eab0401bd70a8ba8534b903145855a7b703", "575f9c8604e0b4c7b36fb41fc5fd280a3c336906" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:46:51 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:46:51 2015 +1000" }, "message": "Merge tag \u0027vmwgfx-fixes-4.3-151001\u0027 of git://people.freedesktop.org/~thomash/linux into drm-fixes\n\nA single commit to fix a command submission hang regression.\n\nPull request of 2015-10-01\n\n* tag \u0027vmwgfx-fixes-4.3-151001\u0027 of git://people.freedesktop.org/~thomash/linux:\n drm/vmwgfx: Fix a command submission hang regression\n" }, { "commit": "8e592eab0401bd70a8ba8534b903145855a7b703", "tree": "dbd1fccb7031c0547472aa2e4adce7d74a290862", "parents": [ "5b6386419c4f78526afe1615668b6444d7b66fd7", "cda374253f862bd0f43edda6935a48294fe8ed3e" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:46:18 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:46:18 2015 +1000" }, "message": "Merge branch \u0027exynos-drm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-fixes\n\n This pull request includes regression fixups, build warnings, and\n trivial cleanups which mostly remove some codes not used anymore.\n\n* \u0027exynos-drm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos:\n drm/exynos: Staticize local function in exynos_drm_gem.c\n drm/exynos: fimd: actually disable dp clock\n drm/exynos: dp: remove suspend/resume functions\n drm/exynos: remove unused mode_fixup() code\n drm/exynos: remove decon_mode_fixup()\n drm/exynos: remove fimd_mode_fixup()\n drm/exynos: rotator: Clock control is unused if !PM\n drm/exynos: fimc: Clock control is unused if !PM\n drm/exynos: Suspend/resume is unused if !PM\n drm/exynos: create a fake mmap offset with gem creation\n drm/exynos: remove call to drm_gem_free_mmap_offset()\n drm/exynos: Remove useless EXPORT_SYMBOL_GPLs\n drm/exynos: cleanup line feed in exynos_drm_gem_get_ioctl\n drm/exynos: cleanup function calling written twice\n drm/exynos: staticize exynos_drm_gem_init()\n drm/exynos: remove unnecessary NULL assignment\n drm/exynos: fix missed calling of drm_prime_gem_destroy()\n drm/exynos: fix layering violation of address\n" }, { "commit": "5b6386419c4f78526afe1615668b6444d7b66fd7", "tree": "74caa76b54cf773e66f76acad1959fcbbbc2c603", "parents": [ "8d0d94015e96b8853c4f7f06eac3f269e1b3d866", "3e4e3805643445b71b7ee1b84892e43d004e24e2" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:35:22 2015 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Oct 02 10:35:22 2015 +1000" }, "message": "Merge branch \u0027drm-fixes-4.3\u0027 of git://people.freedesktop.org/~agd5f/linux into drm-fixes\n\nradeon and amdgpu fixes for 4.3.\n- backlight s/r fixes\n- typo fix from Dan\n- vm debugging fix\n- remove import_gpu_mem after discussion with Daniel\n\n* \u0027drm-fixes-4.3\u0027 of git://people.freedesktop.org/~agd5f/linux:\n drm/amdgpu: only print meaningful VM faults\n drm/amdgpu/cgs: remove import_gpu_mem\n drm/amdgpu: Restore LCD backlight level on resume\n drm/radeon: Restore LCD backlight level on resume (\u003e\u003d R5xx)\n drm/amdgpu: signedness bug in amdgpu_cs_parser_init()\n" }, { "commit": "ccf70ddcbe9984cee406be2bacfedd5e4776919d", "tree": "f9aaf50eab789d1cd8ce79a2e9e5804d86decfbc", "parents": [ "46c8217c4a54c17dd4c000ad804fa1e223a10578", "d2922422c48df93f3edff7d872ee4f3191fefb08" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 16:43:25 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 16:43:25 2015 -0400" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull KVM fixes from Paolo Bonzini:\n \"(Relatively) a lot of reverts, mostly.\n\n Bugs have trickled in for a new feature in 4.2 (MTRR support in\n guests) so I\u0027m reverting it all; let\u0027s not make this -rc period busier\n for KVM than it\u0027s been so far. This covers the four reverts from me.\n\n The fifth patch is being reverted because Radim found a bug in the\n implementation of stable scheduler clock, *but* also managed to\n implement the feature entirely without hypervisor support. So instead\n of fixing the hypervisor side we can remove it completely; 4.4 will\n get the new implementation\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm:\n Use WARN_ON_ONCE for missing X86_FEATURE_NRIPS\n Update KVM homepage Url\n Revert \"KVM: SVM: use NPT page attributes\"\n Revert \"KVM: svm: handle KVM_X86_QUIRK_CD_NW_CLEARED in svm_get_mt_mask\"\n Revert \"KVM: SVM: Sync g_pat with guest-written PAT value\"\n Revert \"KVM: x86: apply guest MTRR virtualization on host reserved pages\"\n Revert \"KVM: x86: zero kvmclock_offset when vcpu0 initializes kvmclock system MSR\"\n" }, { "commit": "46c8217c4a54c17dd4c000ad804fa1e223a10578", "tree": "c188a31698428a416ab39382cb49d421bc666715", "parents": [ "f97b870eced0ec562f953d32eda03906c7dacad6", "2866196f294954ce9fa226825c8c1eaa64c7da8a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 16:38:52 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 16:38:52 2015 -0400" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma\n\nPull rdma fixes from Doug Ledford:\n - Fixes for mlx5 related issues\n - Fixes for ipoib multicast handling\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma:\n IB/ipoib: increase the max mcast backlog queue\n IB/ipoib: Make sendonly multicast joins create the mcast group\n IB/ipoib: Expire sendonly multicast joins\n IB/mlx5: Remove pa_lkey usages\n IB/mlx5: Remove support for IB_DEVICE_LOCAL_DMA_LKEY\n IB/iser: Add module parameter for always register memory\n xprtrdma: Replace global lkey with lkey local to PD\n" }, { "commit": "eb6d1c287ae1f7221248d5be26a5b1560073c09e", "tree": "5cf86434abe6802f9749c1cd22c95ab5c592fea4", "parents": [ "01351eb23c5047895c14c91a898546c289a3c35c", "e9a7bed40cbc09eee4b7a15509726a44565805be", "e1a2d49cd5ef551c51be95cc037033e9e582b0cd", "a9062dcec3697db5bdf064311ad097c469fd87ec" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:47 2015 +0200" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:47 2015 +0200" }, "message": "Merge branches \u0027pm-cpuidle\u0027, \u0027pm-opp\u0027 and \u0027pm-tools\u0027\n\n* pm-cpuidle:\n intel_idle: Skylake Client Support - updated\n\n* pm-opp:\n PM / OPP: Fix typo modifcation -\u003e modification\n PM / OPP: of_property_count_u32_elems() can return errors\n\n* pm-tools:\n tools/power turbosat: update version number\n tools/power turbostat: SKL: Adjust for TSC difference from base frequency\n tools/power turbostat: KNL workaround for %Busy and Avg_MHz\n tools/power turbostat: IVB Xeon: fix --debug regression\n" }, { "commit": "01351eb23c5047895c14c91a898546c289a3c35c", "tree": "9da0b8500659bcc304d762fccbd414884f864a29", "parents": [ "dd953d318daad9a3c1f9a6bf31430bf40163051e", "15b94fa32a422f4b97dc34e4b7060ec83d10bee5" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:35 2015 +0200" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:35 2015 +0200" }, "message": "Merge branch \u0027acpi-ec\u0027\n\n* acpi-ec:\n ACPI / EC: Fix a memory leak issue in acpi_ec_query()\n" }, { "commit": "dd953d318daad9a3c1f9a6bf31430bf40163051e", "tree": "1ee2cc222183032028d9ec08e2c29964938dc3f5", "parents": [ "9ffecb10283508260936b96022d4ee43a7798b4c", "a8360062ccfb4b891d3013d0e55826c8bcb02bfb", "d323efc786910bcc0c8f8b9f97780c70544ac4df" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:12 2015 +0200" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Oct 01 22:30:12 2015 +0200" }, "message": "Merge branches \u0027pm-pci\u0027 and \u0027acpi-pci\u0027\n\n* pm-pci:\n PCI / PM: Update runtime PM documentation for PCI devices\n\n* acpi-pci:\n ACPI / PCI: Remove duplicated penalty on SCI IRQ\n ACPI, PCI, irq: Do not share PCI IRQ with ISA IRQ\n" }, { "commit": "1a541b4e3cd6f5795022514114854b3e1345f24e", "tree": "89342e38161dc6e6d2b6c13df33c559e3e437cb6", "parents": [ "9ffecb10283508260936b96022d4ee43a7798b4c" ], "author": { "name": "Steve Capper", "email": "steve.capper@linaro.org", "time": "Thu Oct 01 13:06:07 2015 +0100" }, "committer": { "name": "Catalin Marinas", "email": "catalin.marinas@arm.com", "time": "Thu Oct 01 18:02:21 2015 +0100" }, "message": "arm64: Fix THP protection change logic\n\n6910fa1 (\"arm64: enable PTE type bit in the mask for pte_modify\") fixes\na problem whereby a large block of PROT_NONE mapped memory is\nincorrectly mapped as block descriptors when mprotect is called.\n\nUnfortunately, a subtle bug was introduced by this fix to the THP logic.\n\nIf one mmaps a large block of memory, then faults it such that it is\ncollapsed into THPs; resulting calls to mprotect on this area of memory\nwill lead to incorrect table descriptors being written instead of block\ndescriptors. This is because pmd_modify calls pte_modify which is now\nallowed to modify the type of the page table entry.\n\nThis patch reverts commit 6910fa16dbe142f6a0fd0fd7c249f9883ff7fc8a, and\nfixes the problem it was trying to address by adjusting PAGE_NONE to\nrepresent a table entry. Thus no change in pte type is required when\nmoving from PROT_NONE to a different protection.\n\nFixes: 6910fa16dbe1 (\"arm64: enable PTE type bit in the mask for pte_modify\")\nCc: \u003cstable@vger.kernel.org\u003e # 4.0+\nCc: Feng Kan \u003cfkan@apm.com\u003e\nReported-by: Ganapatrao Kulkarni \u003cGanapatrao.Kulkarni@caviumnetworks.com\u003e\nTested-by: Ganapatrao Kulkarni \u003cgkulkarni@caviumnetworks.com\u003e\nReviewed-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\nSigned-off-by: Steve Capper \u003csteve.capper@linaro.org\u003e\nSigned-off-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\n" }, { "commit": "1e16a8f11669c98a0adf5fb5f8522aebc1f69f71", "tree": "7938293c85575ee215228cb058c5cc53253704f8", "parents": [ "5b235dc2647e4977b17b5c41d959d0f455831c3f" ], "author": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Oct 01 15:45:44 2015 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Oct 01 15:45:44 2015 +0200" }, "message": "MIPS: BPF: Do all exports of symbols with FEXPORT().\n\nFEXPORT also marks the symbol as code using .type symbol, @function.\nWithout objdump -d will output only a hexdump for code following the\naffected symbols.\n\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "d2922422c48df93f3edff7d872ee4f3191fefb08", "tree": "c326791ccf4378c1793c187fbda326855aeec31c", "parents": [ "038161dea1eaeee80341134e2675b24656a71b59" ], "author": { "name": "Dirk Müller", "email": "dmueller@suse.com", "time": "Thu Oct 01 13:43:42 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 14:59:37 2015 +0200" }, "message": "Use WARN_ON_ONCE for missing X86_FEATURE_NRIPS\n\nThe cpu feature flags are not ever going to change, so warning\neverytime can cause a lot of kernel log spam\n(in our case more than 10GB/hour).\n\nThe warning seems to only occur when nested virtualization is\nenabled, so it\u0027s probably triggered by a KVM bug. This is a\nsensible and safe change anyway, and the KVM bug fix might not\nbe suitable for stable releases anyway.\n\nCc: stable@vger.kernel.org\nSigned-off-by: Dirk Mueller \u003cdmueller@suse.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "038161dea1eaeee80341134e2675b24656a71b59", "tree": "c6d74e7020e23dcc6d75172a6c3e1066b0575d32", "parents": [ "fc07e76ac7ffa3afd621a1c3858a503386a14281" ], "author": { "name": "Dirk Müller", "email": "dmueller@suse.com", "time": "Thu Oct 01 13:46:01 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 14:31:26 2015 +0200" }, "message": "Update KVM homepage Url\n\nThe old one appears to be a generic catch all page, which\nis unhelpful.\n\nSigned-off-by: Dirk Mueller \u003cdmueller@suse.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "f97b870eced0ec562f953d32eda03906c7dacad6", "tree": "232ce6b9dc47eecd3f69cf64b0018431397009cc", "parents": [ "9522f476d932eb23ebf8dbbc6b4e2f966f3fe911", "7c7feb2ebfc9c0552c51f0c050db1d1a004faac5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 07:57:27 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 07:57:27 2015 -0400" }, "message": "Merge tag \u0027upstream-4.3-rc4\u0027 of git://git.infradead.org/linux-ubifs\n\nPull UBI/UBIFS fixes from Richard Weinberger:\n \"This contains three bug fixes for both UBI and UBIFS\"\n\n* tag \u0027upstream-4.3-rc4\u0027 of git://git.infradead.org/linux-ubifs:\n UBI: return ENOSPC if no enough space available\n UBI: Validate data_size\n UBIFS: Kill unneeded locking in ubifs_init_security\n" }, { "commit": "9522f476d932eb23ebf8dbbc6b4e2f966f3fe911", "tree": "89a2e788915e614f29bd95ab9f57169412d72859", "parents": [ "dd36d7393d6310b0c1adefb22fba79c3cf8a577c", "0266715155379820c5c626e2cd4b8c91751036bc" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 07:50:08 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Oct 01 07:50:08 2015 -0400" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security\n\nPull key signing fixes from James Morris:\n \"Keyrings and modsign fixes from David Howells\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:\n MODSIGN: Change from CMS to PKCS#7 signing if the openssl is too old\n X.509: Don\u0027t strip leading 00\u0027s from key ID when constructing key description\n KEYS: Remove unnecessary header #inclusions from extract-cert.c\n KEYS: Fix race between key destruction and finding a keyring by name\n" }, { "commit": "fc07e76ac7ffa3afd621a1c3858a503386a14281", "tree": "666420f710cccb20cb17cded72f3595fb1aa49e1", "parents": [ "bcf166a9942c3aabd8b752a7c38a49f57c54cfb8" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:20:22 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:30:44 2015 +0200" }, "message": "Revert \"KVM: SVM: use NPT page attributes\"\n\nThis reverts commit 3c2e7f7de3240216042b61073803b61b9b3cfb22.\nInitializing the mapping from MTRR to PAT values was reported to\nfail nondeterministically, and it also caused extremely slow boot\n(due to caching getting disabled---bug 103321) with assigned devices.\n\nReported-by: Markus Trippelsdorf \u003cmarkus@trippelsdorf.de\u003e\nReported-by: Sebastian Schuette \u003cdracon@ewetel.net\u003e\nCc: stable@vger.kernel.org # 4.2+\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "bcf166a9942c3aabd8b752a7c38a49f57c54cfb8", "tree": "dba718263b0c1e8d61a4f721ea9af7823ba95772", "parents": [ "625422f60c55bbc368b8568ff925770b36bfc189" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:19:55 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:30:43 2015 +0200" }, "message": "Revert \"KVM: svm: handle KVM_X86_QUIRK_CD_NW_CLEARED in svm_get_mt_mask\"\n\nThis reverts commit 5492830370171b6a4ede8a3bfba687a8d0f25fa5.\nIt builds on the commit that is being reverted next.\n\nCc: stable@vger.kernel.org # 4.2+\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "625422f60c55bbc368b8568ff925770b36bfc189", "tree": "6f7a6f37adc224aa5723a4b4afcd85ae2178af72", "parents": [ "606decd67049217684e3cb5a54104d51ddd4ef35" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:28:15 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:30:43 2015 +0200" }, "message": "Revert \"KVM: SVM: Sync g_pat with guest-written PAT value\"\n\nThis reverts commit e098223b789b4a618dacd79e5e0dad4a9d5018d1,\nwhich has a dependency on other commits being reverted.\n\nCc: stable@vger.kernel.org # 4.2+\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "606decd67049217684e3cb5a54104d51ddd4ef35", "tree": "cac2e61715bd29bdb93ab713b2764329d0d9b94a", "parents": [ "9bac175d8ed0b1dd3d3611c0713666b724eeace3" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:12:47 2015 +0200" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Oct 01 13:30:42 2015 +0200" }, "message": "Revert \"KVM: x86: apply guest MTRR virtualization on host reserved pages\"\n\nThis reverts commit fd717f11015f673487ffc826e59b2bad69d20fe5.\nIt was reported to cause Machine Check Exceptions (bug 104091).\n\nReported-by: harn-solo@gmx.de\nCc: stable@vger.kernel.org # 4.2+\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "0ce3cc008ec04258b6a6314b09f1a6012810881a", "tree": "cd94bbe92edb33a27a2cc0028d0abe87b2a886e2", "parents": [ "a5caa209ba9c29c6421292e7879d2387a2ef39c9" ], "author": { "name": "Ard Biesheuvel", "email": "ard.biesheuvel@linaro.org", "time": "Fri Sep 25 23:02:19 2015 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Oct 01 12:51:28 2015 +0200" }, "message": "arm64/efi: Fix boot crash by not padding between EFI_MEMORY_RUNTIME regions\n\nThe new Properties Table feature introduced in UEFIv2.5 may\nsplit memory regions that cover PE/COFF memory images into\nseparate code and data regions. Since these regions only differ\nin the type (runtime code vs runtime data) and the permission\nbits, but not in the memory type attributes (UC/WC/WT/WB), the\nspec does not require them to be aligned to 64 KB.\n\nSince the relative offset of PE/COFF .text and .data segments\ncannot be changed on the fly, this means that we can no longer\npad out those regions to be mappable using 64 KB pages.\nUnfortunately, there is no annotation in the UEFI memory map\nthat identifies data regions that were split off from a code\nregion, so we must apply this logic to all adjacent runtime\nregions whose attributes only differ in the permission bits.\n\nSo instead of rounding each memory region to 64 KB alignment at\nboth ends, only round down regions that are not directly\npreceded by another runtime region with the same type\nattributes. Since the UEFI spec does not mandate that the memory\nmap be sorted, this means we also need to sort it first.\n\nNote that this change will result in all EFI_MEMORY_RUNTIME\nregions whose start addresses are not aligned to the OS page\nsize to be mapped with executable permissions (i.e., on kernels\ncompiled with 64 KB pages). However, since these mappings are\nonly active during the time that UEFI Runtime Services are being\ninvoked, the window for abuse is rather small.\n\nTested-by: Mark Salter \u003cmsalter@redhat.com\u003e\nTested-by: Mark Rutland \u003cmark.rutland@arm.com\u003e [UEFI 2.4 only]\nSigned-off-by: Ard Biesheuvel \u003card.biesheuvel@linaro.org\u003e\nSigned-off-by: Matt Fleming \u003cmatt.fleming@intel.com\u003e\nReviewed-by: Mark Salter \u003cmsalter@redhat.com\u003e\nReviewed-by: Mark Rutland \u003cmark.rutland@arm.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e # v4.0+\nCc: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\nCc: Leif Lindholm \u003cleif.lindholm@linaro.org\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Will Deacon \u003cwill.deacon@arm.com\u003e\nCc: linux-kernel@vger.kernel.org\nLink: http://lkml.kernel.org/r/1443218539-7610-3-git-send-email-matt@codeblueprint.co.uk\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "a5caa209ba9c29c6421292e7879d2387a2ef39c9", "tree": "26ce6df7867c604e2d30b2f6d28867133c2edb49", "parents": [ "dd36d7393d6310b0c1adefb22fba79c3cf8a577c" ], "author": { "name": "Matt Fleming", "email": "matt.fleming@intel.com", "time": "Fri Sep 25 23:02:18 2015 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Oct 01 12:51:28 2015 +0200" }, "message": "x86/efi: Fix boot crash by mapping EFI memmap entries bottom-up at runtime, instead of top-down\n\nBeginning with UEFI v2.5 EFI_PROPERTIES_TABLE was introduced\nthat signals that the firmware PE/COFF loader supports splitting\ncode and data sections of PE/COFF images into separate EFI\nmemory map entries. This allows the kernel to map those regions\nwith strict memory protections, e.g. EFI_MEMORY_RO for code,\nEFI_MEMORY_XP for data, etc.\n\nUnfortunately, an unwritten requirement of this new feature is\nthat the regions need to be mapped with the same offsets\nrelative to each other as observed in the EFI memory map. If\nthis is not done crashes like this may occur,\n\n BUG: unable to handle kernel paging request at fffffffefe6086dd\n IP: [\u003cfffffffefe6086dd\u003e] 0xfffffffefe6086dd\n Call Trace:\n [\u003cffffffff8104c90e\u003e] efi_call+0x7e/0x100\n [\u003cffffffff81602091\u003e] ? virt_efi_set_variable+0x61/0x90\n [\u003cffffffff8104c583\u003e] efi_delete_dummy_variable+0x63/0x70\n [\u003cffffffff81f4e4aa\u003e] efi_enter_virtual_mode+0x383/0x392\n [\u003cffffffff81f37e1b\u003e] start_kernel+0x38a/0x417\n [\u003cffffffff81f37495\u003e] x86_64_start_reservations+0x2a/0x2c\n [\u003cffffffff81f37582\u003e] x86_64_start_kernel+0xeb/0xef\n\nHere 0xfffffffefe6086dd refers to an address the firmware\nexpects to be mapped but which the OS never claimed was mapped.\nThe issue is that included in these regions are relative\naddresses to other regions which were emitted by the firmware\ntoolchain before the \"splitting\" of sections occurred at\nruntime.\n\nNeedless to say, we don\u0027t satisfy this unwritten requirement on\nx86_64 and instead map the EFI memory map entries in reverse\norder. The above crash is almost certainly triggerable with any\nkernel newer than v3.13 because that\u0027s when we rewrote the EFI\nruntime region mapping code, in commit d2f7cbe7b26a (\"x86/efi:\nRuntime services virtual mapping\"). For kernel versions before\nv3.13 things may work by pure luck depending on the\nfragmentation of the kernel virtual address space at the time we\nmap the EFI regions.\n\nInstead of mapping the EFI memory map entries in reverse order,\nwhere entry N has a higher virtual address than entry N+1, map\nthem in the same order as they appear in the EFI memory map to\npreserve this relative offset between regions.\n\nThis patch has been kept as small as possible with the intention\nthat it should be applied aggressively to stable and\ndistribution kernels. It is very much a bugfix rather than\nsupport for a new feature, since when EFI_PROPERTIES_TABLE is\nenabled we must map things as outlined above to even boot - we\nhave no way of asking the firmware not to split the code/data\nregions.\n\nIn fact, this patch doesn\u0027t even make use of the more strict\nmemory protections available in UEFI v2.5. That will come later.\n\nSuggested-by: Ard Biesheuvel \u003card.biesheuvel@linaro.org\u003e\nReported-by: Ard Biesheuvel \u003card.biesheuvel@linaro.org\u003e\nSigned-off-by: Matt Fleming \u003cmatt.fleming@intel.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nCc: Borislav Petkov \u003cbp@suse.de\u003e\nCc: Chun-Yi \u003cjlee@suse.com\u003e\nCc: Dave Young \u003cdyoung@redhat.com\u003e\nCc: H. Peter Anvin \u003chpa@zytor.com\u003e\nCc: James Bottomley \u003cJBottomley@Odin.com\u003e\nCc: Lee, Chun-Yi \u003cjlee@suse.com\u003e\nCc: Leif Lindholm \u003cleif.lindholm@linaro.org\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Matthew Garrett \u003cmjg59@srcf.ucam.org\u003e\nCc: Mike Galbraith \u003cefault@gmx.de\u003e\nCc: Peter Jones \u003cpjones@redhat.com\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: linux-kernel@vger.kernel.org\nLink: http://lkml.kernel.org/r/1443218539-7610-2-git-send-email-matt@codeblueprint.co.uk\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "95c2b17534654829db428f11bcf4297c059a2a7e", "tree": "aea31e60108daabb0efa0342ba302109d22ca70e", "parents": [ "d32dc9aa10c739363c775baf4499416b2e0dc11f" ], "author": { "name": "Ben Hutchings", "email": "ben@decadent.org.uk", "time": "Sat Sep 26 12:23:56 2015 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Thu Oct 01 12:18:13 2015 +0200" }, "message": "genirq: Fix race in register_irq_proc()\n\nPer-IRQ directories in procfs are created only when a handler is first\nadded to the irqdesc, not when the irqdesc is created. In the case of\na shared IRQ, multiple tasks can race to create a directory. This\nrace condition seems to have been present forever, but is easier to\nhit with async probing.\n\nSigned-off-by: Ben Hutchings \u003cben@decadent.org.uk\u003e\nLink: http://lkml.kernel.org/r/1443266636.2004.2.camel@decadent.org.uk\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "0bf6cd5b9531bcc29c0a5e504b6ce2984c6fd8d8", "tree": "bf6d257b11c79cda7f8f44e3fd8d9c535143fdfd", "parents": [ "f4829a9b7a61e159367350008a608b062c4f6840" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Sun Sep 27 21:01:51 2015 +0200" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Oct 01 10:10:57 2015 +0200" }, "message": "blk-mq: factor out a helper to iterate all tags for a request_queue\n\nAnd replace the blk_mq_tag_busy_iter with it - the driver use has been\nreplaced with a new helper a while ago, and internal to the block we\nonly need the new version.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "f4829a9b7a61e159367350008a608b062c4f6840", "tree": "a2f488a95324e0c0b558273f48b18413ab2fa221", "parents": [ "60de074ba1e8f327db19bc33d8530131ac01695d" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Sun Sep 27 21:01:50 2015 +0200" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Oct 01 10:10:55 2015 +0200" }, "message": "blk-mq: fix racy updates of rq-\u003eerrors\n\nblk_mq_complete_request may be a no-op if the request has already\nbeen completed by others means (e.g. a timeout or cancellation), but\ncurrently drivers have to set rq-\u003eerrors before calling\nblk_mq_complete_request, which might leave us with the wrong error value.\n\nAdd an error parameter to blk_mq_complete_request so that we can\ndefer setting rq-\u003eerrors until we known we won the race to complete the\nrequest.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "5b235dc2647e4977b17b5c41d959d0f455831c3f", "tree": "7f74a5a10f4c23aba51097e3b67f2b9c3a940640", "parents": [ "7a63076d9a31a6c2073da45021eeb4f89d2a8b56" ], "author": { "name": "Alban Bedel", "email": "albeu@free.fr", "time": "Fri Sep 04 14:29:16 2015 +0200" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Thu Oct 01 09:06:26 2015 +0200" }, "message": "MIPS: Fix the build on jz4740 after removing the custom gpio.h\n\nSomehow the wrong version of the patch to remove the use of custom\ngpio.h on mips has been merged. This patch add the missing fixes for a\nbuild error on jz4740 because linux/gpio.h doesn\u0027t provide any machine\nspecfics definitions anymore.\n\nSigned-off-by: Alban Bedel \u003calbeu@free.fr\u003e\nCc: Paul Burton \u003cpaul.burton@imgtec.com\u003e\nCc: Lars-Peter Clausen \u003clars@metafoo.de\u003e\nCc: Brian Norris \u003ccomputersforpeace@gmail.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Linus Walleij \u003clinus.walleij@linaro.org\u003e\nCc: linux-mips@linux-mips.org\nCc: linux-kernel@vger.kernel.org\nPatchwork: https://patchwork.linux-mips.org/patch/11089/\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "95c632f4e4e6365a20397f4fd04dcf27aab02958", "tree": "aacb57bca96a6cc58e8a194a95a72ce13cc2d833", "parents": [ "4ac86a6dcec1c3878de9747bf5a2aa4455be69e3", "7ba78053aacb89998a052843e3c56983c31d57f0" ], "author": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Oct 01 09:02:11 2015 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Oct 01 09:02:11 2015 +0200" }, "message": "Merge remote-tracking branch \u0027tglx/x86/urgent\u0027 into x86/urgent\n\nPick up the WCHAN fixes from Thomas.\n\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "cda374253f862bd0f43edda6935a48294fe8ed3e", "tree": "f32d65b2db097ac243e5b030c1072c75d42adf41", "parents": [ "3c79fb8c9424a24bf812b9a8cb4f19b781052b0b" ], "author": { "name": "Krzysztof Kozlowski", "email": "k.kozlowski@samsung.com", "time": "Thu Oct 01 14:21:40 2015 +0900" }, "committer": { "name": "Inki Dae", "email": "inki.dae@samsung.com", "time": "Thu Oct 01 14:29:51 2015 +0900" }, "message": "drm/exynos: Staticize local function in exynos_drm_gem.c\n\nThe exynos_drm_gem_mmap_buffer() is not used outside so make it static.\n\nSigned-off-by: Krzysztof Kozlowski \u003ck.kozlowski@samsung.com\u003e\nSigned-off-by: Inki Dae \u003cinki.dae@samsung.com\u003e\n" }, { "commit": "3c79fb8c9424a24bf812b9a8cb4f19b781052b0b", "tree": "227cd6128ae5b02905dc4aef924287a7ecc97dae", "parents": [ "148ba09c465cc54d8e68f041bf9a30332b315c39" ], "author": { "name": "Gustavo Padovan", "email": "gustavo.padovan@collabora.co.uk", "time": "Wed Sep 30 18:40:54 2015 -0300" }, "committer": { "name": "Inki Dae", "email": "inki.dae@samsung.com", "time": "Thu Oct 01 13:25:11 2015 +0900" }, "message": "drm/exynos: fimd: actually disable dp clock\n\nfimd_dp_clock_enable() was setting the always to enabled,\nthis patch fix this to actually use the value that is set to \u0027val\u0027.\n\nReported-by: Emilio López \u003cemilio.lopez@collabora.co.uk\u003e\nSigned-off-by: Gustavo Padovan \u003cgustavo.padovan@collabora.co.uk\u003e\nSigned-off-by: Inki Dae \u003cinki.dae@samsung.com\u003e\n" }, { "commit": "148ba09c465cc54d8e68f041bf9a30332b315c39", "tree": "78e37b518dd12b560dda4ea25e9f078806104bd9", "parents": [ "10d97d5869f9a50eb3142c7ee562ecc5b5b33f5b" ], "author": { "name": "Inki Dae", "email": "inki.dae@samsung.com", "time": "Wed Sep 30 20:12:29 2015 +0900" }, "committer": { "name": "Inki Dae", "email": "inki.dae@samsung.com", "time": "Thu Oct 01 13:23:13 2015 +0900" }, "message": "drm/exynos: dp: remove suspend/resume functions\n\nThis patch removes unnecessary pm suspend/resume functions.\n\nAll kms sub drivers will be controlled by top of Exynos drm driver\nand connector dpms so these sub drivers shouldn\u0027t have their own\npm interfaces.\n\nSigned-off-by: Inki Dae \u003cinki.dae@samsung.com\u003e\nReviewed-by: Gustavo Padovan \u003cgustavo.padovan@collabora.co.uk\u003e\n" }, { "commit": "8d0d94015e96b8853c4f7f06eac3f269e1b3d866", "tree": "90ef9121a318ccbe5cad390d665d2ad114dccbe2", "parents": [ "9ffecb10283508260936b96022d4ee43a7798b4c" ], "author": { "name": "Fabiano Fidêncio", "email": "fidencio@redhat.com", "time": "Thu Sep 24 15:18:34 2015 +0200" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Oct 01 13:57:21 2015 +1000" }, "message": "drm/qxl: recreate the primary surface when the bo is not primary\n\nWhen disabling/enabling a crtc the primary area must be updated\nindependently of which crtc has been disabled/enabled.\n\nResolves: https://bugzilla.redhat.com/show_bug.cgi?id\u003d1264735\n\nSigned-off-by: Fabiano Fidêncio \u003cfidencio@redhat.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "7b09a1bba4091a9d208481d7831682a1f3061ab9", "tree": "f9681f78e175c16b76ffcab80daac533aa6b5be6", "parents": [ "e87ffbdf06971a80ad2a11217200bdd936195af1" ], "author": { "name": "Robert Jarzmik", "email": "robert.jarzmik@free.fr", "time": "Wed Sep 30 19:42:15 2015 +0200" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Oct 01 07:44:31 2015 +0530" }, "message": "dmaengine: pxa_dma: fix residue corner case\n\nA very tiny temporal window exists in the residue calculation where :\n - upon entering residue calculation, the transfer is ongoing\n - when reading the current transfer pointer, it just changed to\n the \"finisher/linker\" descriptor\n\nIn this case, the residue returned is the whole transfer length instead\nof 0. Fix it.\n\nThis appears almost in one extreme case, where the driver is used\nby older clients which inquire for residue in interrupt context, such\nas the smsc91x ethernet driver, in a tight loop :\n interrupt_handler()\n dmaengine_submit()\n do {\n dmaengine_tx_status()\n } while (residue \u003e 0 || status !\u003d DMA_ERROR)\n\nSigned-off-by: Robert Jarzmik \u003crobert.jarzmik@free.fr\u003e\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "e87ffbdf06971a80ad2a11217200bdd936195af1", "tree": "68607591da1fef0f854285ac9e0e50bb90b5509f", "parents": [ "aa3ee5f569fda51e54c224c0df60e187e9c5e582" ], "author": { "name": "Robert Jarzmik", "email": "robert.jarzmik@free.fr", "time": "Wed Sep 30 19:42:14 2015 +0200" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Oct 01 07:44:25 2015 +0530" }, "message": "dmaengine: pxa_dma: fix the no-requestor case\n\nA very small number of devices don\u0027t use the flow control offered by\nrequestor lines. In these specific cases, the pxa dma driver should be\naware of that and not try to use a requestor line.\n\nSigned-off-by: Robert Jarzmik \u003crobert.jarzmik@free.fr\u003e\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "aa3ee5f569fda51e54c224c0df60e187e9c5e582", "tree": "df659cc3fe3093b686ec5fd3675358aedcad562b", "parents": [ "0be2136b67067617b36c70e525d7534108361e36" ], "author": { "name": "Axel Lin", "email": "axel.lin@ingics.com", "time": "Sat Sep 26 17:15:47 2015 +0800" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Oct 01 07:42:27 2015 +0530" }, "message": "dmaengine: zxdma: Fix off-by-one for testing valid pchan request\n\nThe valid pchan range is 0 ~ d-\u003edma_requests - 1.\n\nSigned-off-by: Axel Lin \u003caxel.lin@ingics.com\u003e\nReviewed-by: Jun Nie \u003cjun.nie@linaro.org\u003e\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "0be2136b67067617b36c70e525d7534108361e36", "tree": "f6d2c9fd2dcbf65a2a53aaac33d7949602ccb777", "parents": [ "a1cf09031e641d3cceaca4a4dd20ef6a785bc9b3" ], "author": { "name": "Ludovic Desroches", "email": "ludovic.desroches@atmel.com", "time": "Tue Sep 15 15:39:11 2015 +0200" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Oct 01 07:29:49 2015 +0530" }, "message": "dmaengine: at_xdmac: clean used descriptor\n\nWhen putting back a descriptor to the free descs list, some fields are\nnot set to 0, it can cause bugs if someone uses it without having this\nin mind.\nDescriptor are not put back one by one so it is easier to clean\ndescriptors when we request them.\n\nSigned-off-by: Ludovic Desroches \u003cludovic.desroches@atmel.com\u003e\nCc: stable@vger.kernel.org #4.2\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "a1cf09031e641d3cceaca4a4dd20ef6a785bc9b3", "tree": "e62a9e42a2db4de85ce78ea624cbaeb6f9e3d3a3", "parents": [ "6bea0f6d1c47b07be88dfd93f013ae05fcb3d8bf" ], "author": { "name": "Maxime Ripard", "email": "maxime.ripard@free-electrons.com", "time": "Tue Sep 15 15:36:00 2015 +0200" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Oct 01 07:29:40 2015 +0530" }, "message": "dmaengine: at_xdmac: change block increment addressing mode\n\nThe addressing mode we were using was not only incrementing the address at\neach microblock, but also at each data boundary, which was severely slowing\nthe transfer, without any benefit since we were not using the data stride.\n\nSwitch to the micro block increment only in order to get back to an\nacceptable performance level.\n\nSigned-off-by: Maxime Ripard \u003cmaxime.ripard@free-electrons.com\u003e\nSigned-off-by: Ludovic Desroches \u003cludovic.desroches@atmel.com\u003e\nFixes: 6007ccb57744 (\"dmaengine: xdmac: Add interleaved transfer support\")\nCc: stable@vger.kernel.org #4.2\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "7ba78053aacb89998a052843e3c56983c31d57f0", "tree": "de45a4545a91a08613ed1bc0fb4659241fba34d0", "parents": [ "eddd3826a1a0190e5235703d1e666affa4d13b96" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Sep 30 08:38:23 2015 +0000" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Sep 30 21:51:34 2015 +0200" }, "message": "x86/process: Unify 32bit and 64bit implementations of get_wchan()\n\nThe stack layout and the functionality is identical. Use the 64bit\nversion for all of x86.\n\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReviewed-by: Borislav Petkov \u003cbp@alien8.de\u003e\nReviewed-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nCc: Andrey Ryabinin \u003cryabinin.a.a@gmail.com\u003e\nCc: Andy Lutomirski \u003cluto@amacapital.net\u003e\nCc: Andrey Konovalov \u003candreyknvl@google.com\u003e\nCc: Kostya Serebryany \u003ckcc@google.com\u003e\nCc: Alexander Potapenko \u003cglider@google.com\u003e\nCc: kasan-dev \u003ckasan-dev@googlegroups.com\u003e\nCc: Denys Vlasenko \u003cdvlasenk@redhat.com\u003e\nCc: Andi Kleen \u003cak@linux.intel.com\u003e\nCc: Sasha Levin \u003csasha.levin@oracle.com\u003e\nCc: Wolfram Gloger \u003cwmglo@dent.med.uni-muenchen.de\u003e\nLink: http://lkml.kernel.org/r/20150930083302.779694618@linutronix.de\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "eddd3826a1a0190e5235703d1e666affa4d13b96", "tree": "65a236d7bfe928f827a49319f4be27f9d6cd3f21", "parents": [ "1e034743e918d195d339af340ae933727c072bce" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Sep 30 08:38:22 2015 +0000" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Sep 30 21:51:34 2015 +0200" }, "message": "x86/process: Add proper bound checks in 64bit get_wchan()\n\nDmitry Vyukov reported the following using trinity and the memory\nerror detector AddressSanitizer\n(https://code.google.com/p/address-sanitizer/wiki/AddressSanitizerForKernel).\n\n[ 124.575597] ERROR: AddressSanitizer: heap-buffer-overflow on\naddress ffff88002e280000\n[ 124.576801] ffff88002e280000 is located 131938492886538 bytes to\nthe left of 28857600-byte region [ffffffff81282e0a, ffffffff82e0830a)\n[ 124.578633] Accessed by thread T10915:\n[ 124.579295] inlined in describe_heap_address\n./arch/x86/mm/asan/report.c:164\n[ 124.579295] #0 ffffffff810dd277 in asan_report_error\n./arch/x86/mm/asan/report.c:278\n[ 124.580137] #1 ffffffff810dc6a0 in asan_check_region\n./arch/x86/mm/asan/asan.c:37\n[ 124.581050] #2 ffffffff810dd423 in __tsan_read8 ??:0\n[ 124.581893] #3 ffffffff8107c093 in get_wchan\n./arch/x86/kernel/process_64.c:444\n\nThe address checks in the 64bit implementation of get_wchan() are\nwrong in several ways:\n\n - The lower bound of the stack is not the start of the stack\n page. It\u0027s the start of the stack page plus sizeof (struct\n thread_info)\n\n - The upper bound must be:\n\n top_of_stack - TOP_OF_KERNEL_STACK_PADDING - 2 * sizeof(unsigned long).\n\n The 2 * sizeof(unsigned long) is required because the stack pointer\n points at the frame pointer. The layout on the stack is: ... IP FP\n ... IP FP. So we need to make sure that both IP and FP are in the\n bounds.\n\nFix the bound checks and get rid of the mix of numeric constants, u64\nand unsigned long. Making all unsigned long allows us to use the same\nfunction for 32bit as well.\n\nUse READ_ONCE() when accessing the stack. This does not prevent a\nconcurrent wakeup of the task and the stack changing, but at least it\navoids TOCTOU.\n\nAlso check task state at the end of the loop. Again that does not\nprevent concurrent changes, but it avoids walking for nothing.\n\nAdd proper comments while at it.\n\nReported-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nReported-by: Sasha Levin \u003csasha.levin@oracle.com\u003e\nBased-on-patch-from: Wolfram Gloger \u003cwmglo@dent.med.uni-muenchen.de\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReviewed-by: Borislav Petkov \u003cbp@alien8.de\u003e\nReviewed-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nCc: Andrey Ryabinin \u003cryabinin.a.a@gmail.com\u003e\nCc: Andy Lutomirski \u003cluto@amacapital.net\u003e\nCc: Andrey Konovalov \u003candreyknvl@google.com\u003e\nCc: Kostya Serebryany \u003ckcc@google.com\u003e\nCc: Alexander Potapenko \u003cglider@google.com\u003e\nCc: kasan-dev \u003ckasan-dev@googlegroups.com\u003e\nCc: Denys Vlasenko \u003cdvlasenk@redhat.com\u003e\nCc: Andi Kleen \u003cak@linux.intel.com\u003e\nCc: Wolfram Gloger \u003cwmglo@dent.med.uni-muenchen.de\u003e\nCc: stable@vger.kernel.org\nLink: http://lkml.kernel.org/r/20150930083302.694788319@linutronix.de\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\n" }, { "commit": "dd36d7393d6310b0c1adefb22fba79c3cf8a577c", "tree": "aedf3c14c57f3663dbe476155d24a0471f9b5f0f", "parents": [ "d4e842be2b697eb61c8863fe4b080bf0688198a2", "3cef072d3bbd4344823545e50d8cb240a6d4635d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 30 13:13:55 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 30 13:13:55 2015 -0400" }, "message": "Merge git://www.linux-watchdog.org/linux-watchdog\n\nPull watchdog fixes from Wim Van Sebroeck:\n \"This fixes:\n\n - module autoload for 3 OF platform drivers\n - poweroff behaviour on bcm2835 watchdog device\n - I2C dependencies for iTCO_wdt.c\"\n\n* git://www.linux-watchdog.org/linux-watchdog:\n watchdog: iTCO: Fix dependencies on I2C\n watchdog: bcm2835: Fix poweroff behaviour\n watchdog: Fix module autoload for OF platform driver\n" }, { "commit": "d4e842be2b697eb61c8863fe4b080bf0688198a2", "tree": "143f5c1f3bffc61fa59b6723c2c6bc7823093a5e", "parents": [ "70c8a00a09076ca2683ef24356fad2e55f93a427", "f491e70ccffa5d19aa51c958909320fa1f3905ed" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 30 13:11:42 2015 -0400" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 30 13:11:42 2015 -0400" }, "message": "Merge tag \u0027hwmon-for-linus-v4.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull hwmin fixes from Guenter Roeck:\n \"Fix module autoload for various drivers\"\n\n* tag \u0027hwmon-for-linus-v4.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n hwmon: (pwm-fan) Fix module autoload for OF platform driver\n hwmon: (gpio-fan) Fix module autoload for OF platform driver\n hwmon: (abx500) Fix module autoload for OF platform driver\n" } ], "next": "70c8a00a09076ca2683ef24356fad2e55f93a427" }