)]}' { "log": [ { "commit": "54ecb8f7028c5eb3d740bb82b0f1d90f2df63c5c", "tree": "1cc9e9e69fa376b595d7d4ab1ca0fc5e66bd4633", "parents": [ "bb48a59135926ece9b1361e8b96b33fc658830bc" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:35:40 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:35:40 2019 -0700" }, "message": "Linux 5.4-rc1\n" }, { "commit": "bb48a59135926ece9b1361e8b96b33fc658830bc", "tree": "c19fe42b0c1c923b088a9250f7fbcdab9e8f266d", "parents": [ "80b29b6b8cd7479a67f5e338195dbc121b30c879", "d4e204948fe3e0dc8e1fbf3f8f3290c9c2823be3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:25:24 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:25:24 2019 -0700" }, "message": "Merge tag \u0027for-5.4-rc1-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux\n\nPull btrfs fixes from David Sterba:\n \"A bunch of fixes that accumulated in recent weeks, mostly material for\n stable.\n\n Summary:\n\n - fix for regression from 5.3 that prevents to use balance convert\n with single profile\n\n - qgroup fixes: rescan race, accounting leak with multiple writers,\n potential leak after io failure recovery\n\n - fix for use after free in relocation (reported by KASAN)\n\n - other error handling fixups\"\n\n* tag \u0027for-5.4-rc1-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:\n btrfs: qgroup: Fix reserved data space leak if we have multiple reserve calls\n btrfs: qgroup: Fix the wrong target io_tree when freeing reserved data space\n btrfs: Fix a regression which we can\u0027t convert to SINGLE profile\n btrfs: relocation: fix use-after-free on dead relocation roots\n Btrfs: fix race setting up and completing qgroup rescan workers\n Btrfs: fix missing error return if writeback for extent buffer never started\n btrfs: adjust dirty_metadata_bytes after writeback failure of extent buffer\n Btrfs: fix selftests failure due to uninitialized i_mode in test inodes\n" }, { "commit": "80b29b6b8cd7479a67f5e338195dbc121b30c879", "tree": "6a48612f4b33233b7fd295ab095fcf9e281ea7d5", "parents": [ "cef0aa0ce8592f68fb093b2be0d341a568ff9890", "9af032a30172e119a5935f802b066631f8ded2d6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:16:17 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:16:17 2019 -0700" }, "message": "Merge tag \u0027csky-for-linus-5.4-rc1\u0027 of git://github.com/c-sky/csky-linux\n\nPull csky updates from Guo Ren:\n \"This round of csky subsystem just some fixups:\n\n - Fix mb() synchronization problem\n\n - Fix dma_alloc_coherent with PAGE_SO attribute\n\n - Fix cache_op failed when cross memory ZONEs\n\n - Optimize arch_sync_dma_for_cpu/device with dma_inv_range\n\n - Fix ioremap function losing\n\n - Fix arch_get_unmapped_area() implementation\n\n - Fix defer cache flush for 610\n\n - Support kernel non-aligned access\n\n - Fix 610 vipt cache flush mechanism\n\n - Fix add zero_fp fixup perf backtrace panic\n\n - Move static keyword to the front of declaration\n\n - Fix csky_pmu.max_period assignment\n\n - Use generic free_initrd_mem()\n\n - entry: Remove unneeded need_resched() loop\"\n\n* tag \u0027csky-for-linus-5.4-rc1\u0027 of git://github.com/c-sky/csky-linux:\n csky: Move static keyword to the front of declaration\n csky: entry: Remove unneeded need_resched() loop\n csky: Fixup csky_pmu.max_period assignment\n csky: Fixup add zero_fp fixup perf backtrace panic\n csky: Use generic free_initrd_mem()\n csky: Fixup 610 vipt cache flush mechanism\n csky: Support kernel non-aligned access\n csky: Fixup defer cache flush for 610\n csky: Fixup arch_get_unmapped_area() implementation\n csky: Fixup ioremap function losing\n csky: Optimize arch_sync_dma_for_cpu/device with dma_inv_range\n csky/dma: Fixup cache_op failed when cross memory ZONEs\n csky: Fixup dma_alloc_coherent with PAGE_SO attribute\n csky: Fixup mb() synchronization problem\n" }, { "commit": "cef0aa0ce8592f68fb093b2be0d341a568ff9890", "tree": "9bd2d26d5f12c81f470c5f61d6fd0345a98362c6", "parents": [ "cf4f493b102399adea3cf65cdde7161c17fb605c", "9bfd7319e8d353b8b81c4cfd4d7eced71adbfbb5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:04:28 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 10:04:28 2019 -0700" }, "message": "Merge tag \u0027armsoc-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc\n\nPull ARM SoC fixes from Olof Johansson:\n \"A few fixes that have trickled in through the merge window:\n\n - Video fixes for OMAP due to panel-dpi driver removal\n\n - Clock fixes for OMAP that broke no-idle quirks + nfsroot on DRA7\n\n - Fixing arch version on ASpeed ast2500\n\n - Two fixes for reset handling on ARM SCMI\"\n\n* tag \u0027armsoc-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:\n ARM: aspeed: ast2500 is ARMv6K\n reset: reset-scmi: add missing handle initialisation\n firmware: arm_scmi: reset: fix reset_state assignment in scmi_domain_reset\n bus: ti-sysc: Remove unpaired sysc_clkdm_deny_idle()\n ARM: dts: logicpd-som-lv: Fix i2c2 and i2c3 Pin mux\n ARM: dts: am3517-evm: Fix missing video\n ARM: dts: logicpd-torpedo-baseboard: Fix missing video\n ARM: omap2plus_defconfig: Fix missing video\n bus: ti-sysc: Fix handling of invalid clocks\n bus: ti-sysc: Fix clock handling for no-idle quirks\n" }, { "commit": "cf4f493b102399adea3cf65cdde7161c17fb605c", "tree": "b3ae5844c4bc69e1428a9ed4db7c01ad786bce43", "parents": [ "c710364f78afdef8c2ed07556d0743c5a30ed429", "8ed4889eb83179dbc9a105cfed65cc42ecb61097" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 09:29:53 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 09:29:53 2019 -0700" }, "message": "Merge tag \u0027trace-v5.4-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull tracing fixes from Steven Rostedt:\n \"A few more tracing fixes:\n\n - Fix a buffer overflow by checking nr_args correctly in probes\n\n - Fix a warning that is reported by clang\n\n - Fix a possible memory leak in error path of filter processing\n\n - Fix the selftest that checks for failures, but wasn\u0027t failing\n\n - Minor clean up on call site output of a memory trace event\"\n\n* tag \u0027trace-v5.4-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n selftests/ftrace: Fix same probe error test\n mm, tracing: Print symbol name for call_site in trace events\n tracing: Have error path in predicate_parse() free its allocated memory\n tracing: Fix clang -Wint-in-bool-context warnings in IF_ASSIGN macro\n tracing/probe: Fix to check the difference of nr_args before adding probe\n" }, { "commit": "c710364f78afdef8c2ed07556d0743c5a30ed429", "tree": "38173e6818c3c2e4414a767dd81e60d1d9ff4309", "parents": [ "97f9a3c4eee55b0178b518ae7114a6a53372913d", "e51df6ce668a8f75ce27f83ce0f60103c568c375" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 09:21:53 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 30 09:21:53 2019 -0700" }, "message": "Merge tag \u0027mmc-v5.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull more MMC updates from Ulf Hansson:\n \"A couple more updates/fixes for MMC:\n\n - sdhci-pci: Add Genesys Logic GL975x support\n\n - sdhci-tegra: Recover loss in throughput for DMA\n\n - sdhci-of-esdhc: Fix DMA bug\"\n\n* tag \u0027mmc-v5.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: host: sdhci-pci: Add Genesys Logic GL975x support\n mmc: tegra: Implement -\u003eset_dma_mask()\n mmc: sdhci: Let drivers define their DMA mask\n mmc: sdhci-of-esdhc: set DMA snooping based on DMA coherence\n mmc: sdhci: improve ADMA error reporting\n" }, { "commit": "9af032a30172e119a5935f802b066631f8ded2d6", "tree": "d6a5d4c7cf5d297fd7963c64521002e9b404f0b0", "parents": [ "a2139d3b4fd7ef26a363a1b1eb6cd55be2c1bcd1" ], "author": { "name": "Krzysztof Wilczynski", "email": "kw@linux.com", "time": "Tue Sep 03 13:36:51 2019 +0200" }, "committer": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Mon Sep 30 11:50:49 2019 +0800" }, "message": "csky: Move static keyword to the front of declaration\n\nMove the static keyword to the front of declaration of\ncsky_pmu_of_device_ids, and resolve the following compiler\nwarning that can be seen when building with warnings\nenabled (W\u003d1):\n\narch/csky/kernel/perf_event.c:1340:1: warning:\n ‘static’ is not at beginning of declaration [-Wold-style-declaration]\n\nSigned-off-by: Krzysztof Wilczynski \u003ckw@linux.com\u003e\nSigned-off-by: Guo Ren \u003cguoren@kernel.org\u003e\n" }, { "commit": "a2139d3b4fd7ef26a363a1b1eb6cd55be2c1bcd1", "tree": "5fda9e52e3217670fa16135ee2d3da44c6f7d89c", "parents": [ "3a09d8e2893b2403a043890e5832966e8640feaf" ], "author": { "name": "Valentin Schneider", "email": "valentin.schneider@arm.com", "time": "Mon Sep 23 15:36:14 2019 +0100" }, "committer": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Mon Sep 30 11:49:47 2019 +0800" }, "message": "csky: entry: Remove unneeded need_resched() loop\n\nSince the enabling and disabling of IRQs within preempt_schedule_irq()\nis contained in a need_resched() loop, we don\u0027t need the outer arch\ncode loop.\n\nSigned-off-by: Valentin Schneider \u003cvalentin.schneider@arm.com\u003e\nSigned-off-by: Guo Ren \u003cguoren@kernel.org\u003e\n" }, { "commit": "97f9a3c4eee55b0178b518ae7114a6a53372913d", "tree": "03b87ee366ed32ac87b3e0919c7919c5223fda09", "parents": [ "1eb80d6ffb1759968374606c1e36ea88e043e66d", "dc925a36060e8cef050a9d05c64dae1c30dc5027" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:52:52 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:52:52 2019 -0700" }, "message": "Merge tag \u0027char-misc-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc\n\nPull Documentation/process update from Greg KH:\n \"Here are two small Documentation/process/embargoed-hardware-issues.rst\n file updates that missed my previous char/misc pull request.\n\n The first one adds an Intel representative for the process, and the\n second one cleans up the text a bit more when it comes to how the\n disclosure rules work, as it was a bit confusing to some companies\"\n\n* tag \u0027char-misc-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:\n Documentation/process: Clarify disclosure rules\n Documentation/process: Volunteer as the ambassador for Intel\n" }, { "commit": "1eb80d6ffb1759968374606c1e36ea88e043e66d", "tree": "94e24dd276ac0016e7934dabb8763624f55a8c5a", "parents": [ "7edee5229c8f4e075fe71274620bb11ead885c9b", "473ef57ad8edc25efd083a583a5f6604b47d3822" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:42:07 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:42:07 2019 -0700" }, "message": "Merge branch \u0027work.misc\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs\n\nPull more vfs updates from Al Viro:\n \"A couple of misc patches\"\n\n* \u0027work.misc\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:\n afs dynroot: switch to simple_dir_operations\n fs/handle.c - fix up kerneldoc\n" }, { "commit": "7edee5229c8f4e075fe71274620bb11ead885c9b", "tree": "c72fabd92b7bb9074f20fee69c9e2a3f5acbbdc9", "parents": [ "3f2dc2798b81531fd93a3b9b7c39da47ec689e55", "a016e2794fc3a245a91946038dd8f34d65e53cc3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:37:32 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:37:32 2019 -0700" }, "message": "Merge tag \u00275.4-rc-smb3-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6\n\nPull more cifs updates from Steve French:\n \"Fixes from the recent SMB3 Test events and Storage Developer\n Conference (held the last two weeks).\n\n Here are nine smb3 patches including an important patch for debugging\n traces with wireshark, with three patches marked for stable.\n\n Additional fixes from last week to better handle some newly discovered\n reparse points, and a fix the create/mkdir path for setting the mode\n more atomically (in SMB3 Create security descriptor context), and one\n for path name processing are still being tested so are not included\n here\"\n\n* tag \u00275.4-rc-smb3-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6:\n CIFS: Fix oplock handling for SMB 2.1+ protocols\n smb3: missing ACL related flags\n smb3: pass mode bits into create calls\n smb3: Add missing reparse tags\n CIFS: fix max ea value size\n fs/cifs/sess.c: Remove set but not used variable \u0027capabilities\u0027\n fs/cifs/smb2pdu.c: Make SMB2_notify_init static\n smb3: fix leak in \"open on server\" perf counter\n smb3: allow decryption keys to be dumped by admin for debugging\n" }, { "commit": "3a09d8e2893b2403a043890e5832966e8640feaf", "tree": "223e7acd6759948b7561d310b447ae766dd795a6", "parents": [ "48ede51fd94fe9251058fc85626b2aeb5cbb5884" ], "author": { "name": "Mao Han", "email": "han_mao@c-sky.com", "time": "Wed Sep 25 17:23:02 2019 +0800" }, "committer": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Mon Sep 30 10:26:33 2019 +0800" }, "message": "csky: Fixup csky_pmu.max_period assignment\n\nThe csky_pmu.max_period has type u64, and BIT() can only return\n32 bits unsigned long on C-SKY. The initialization for max_period\nwill be incorrect when count_width is bigger than 32.\n\nUse BIT_ULL()\n\nSigned-off-by: Mao Han \u003chan_mao@c-sky.com\u003e\nSigned-off-by: Guo Ren \u003cren_guo@c-sky.com\u003e\n" }, { "commit": "48ede51fd94fe9251058fc85626b2aeb5cbb5884", "tree": "0937cf6de34f7cb04f87c1501a8d7af8cd1e27f8", "parents": [ "fdbdcddc2c93096e9b956de930d2d710a1342502" ], "author": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Wed Sep 25 19:56:16 2019 +0800" }, "committer": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Mon Sep 30 10:26:32 2019 +0800" }, "message": "csky: Fixup add zero_fp fixup perf backtrace panic\n\nWe need set fp zero to let backtrace know the end. The patch fixup perf\ncallchain panic problem, because backtrace didn\u0027t know what is the end\nof fp.\n\nSigned-off-by: Guo Ren \u003cren_guo@c-sky.com\u003e\nReported-by: Mao Han \u003chan_mao@c-sky.com\u003e\n" }, { "commit": "fdbdcddc2c93096e9b956de930d2d710a1342502", "tree": "949586e648753b082c0c5159de63bf0cccca6956", "parents": [ "4ad35c1f56386c8e7019c921bba1af109fde9693" ], "author": { "name": "Mike Rapoport", "email": "rppt@linux.ibm.com", "time": "Wed Aug 28 16:35:19 2019 +0300" }, "committer": { "name": "Guo Ren", "email": "ren_guo@c-sky.com", "time": "Mon Sep 30 10:26:24 2019 +0800" }, "message": "csky: Use generic free_initrd_mem()\n\nThe csky implementation of free_initrd_mem() is an open-coded version of\nfree_reserved_area() without poisoning.\n\nRemove it and make csky use the generic version of free_initrd_mem().\n\nSigned-off-by: Mike Rapoport \u003crppt@linux.ibm.com\u003e\nSigned-off-by: Guo Ren \u003cguoren@kernel.org\u003e\n" }, { "commit": "3f2dc2798b81531fd93a3b9b7c39da47ec689e55", "tree": "075660db24621f4be8e24882bcaa88e15bc797f4", "parents": [ "a3c0e7b1fe1fc62bba5f591c4bc404eea96823b8", "02f03c4206c1b2a7451d3b3546f86c9c783eac13" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:25:39 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 19:25:39 2019 -0700" }, "message": "Merge branch \u0027entropy\u0027\n\nMerge active entropy generation updates.\n\nThis is admittedly partly \"for discussion\". We need to have a way\nforward for the boot time deadlocks where user space ends up waiting for\nmore entropy, but no entropy is forthcoming because the system is\nentirely idle just waiting for something to happen.\n\nWhile this was triggered by what is arguably a user space bug with\nGDM/gnome-session asking for secure randomness during early boot, when\nthey didn\u0027t even need any such truly secure thing, the issue ends up\nbeing that our \"getrandom()\" interface is prone to that kind of\nconfusion, because people don\u0027t think very hard about whether they want\nto block for sufficient amounts of entropy.\n\nThe approach here-in is to decide to not just passively wait for entropy\nto happen, but to start actively collecting it if it is missing. This\nis not necessarily always possible, but if the architecture has a CPU\ncycle counter, there is a fair amount of noise in the exact timings of\nreasonably complex loads.\n\nWe may end up tweaking the load and the entropy estimates, but this\nshould be at least a reasonable starting point.\n\nAs part of this, we also revert the revert of the ext4 IO pattern\nimprovement that ended up triggering the reported lack of external\nentropy.\n\n* getrandom() active entropy waiting:\n Revert \"Revert \"ext4: make __ext4_get_inode_loc plug\"\"\n random: try to actively add entropy rather than passively wait for it\n" }, { "commit": "02f03c4206c1b2a7451d3b3546f86c9c783eac13", "tree": "5b158f08f9372030255b520f77f36570b1d69d04", "parents": [ "50ee7529ec4500c88f8664560770a7a1b65db72b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 17:59:23 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 17:59:23 2019 -0700" }, "message": "Revert \"Revert \"ext4: make __ext4_get_inode_loc plug\"\"\n\nThis reverts commit 72dbcf72156641fde4d8ea401e977341bfd35a05.\n\nInstead of waiting forever for entropy that may just not happen, we now\ntry to actively generate entropy when required, and are thus hopefully\navoiding the problem that caused the nice ext4 IO pattern fix to be\nreverted.\n\nSo revert the revert.\n\nCc: Ahmed S. Darwish \u003cdarwish.07@gmail.com\u003e\nCc: Ted Ts\u0027o \u003ctytso@mit.edu\u003e\nCc: Willy Tarreau \u003cw@1wt.eu\u003e\nCc: Alexander E. Patrakov \u003cpatrakov@gmail.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "50ee7529ec4500c88f8664560770a7a1b65db72b", "tree": "122866c9cd0b4e2425309dc4696c728682434ee1", "parents": [ "4d856f72c10ecb060868ed10ff1b1453943fc6c8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 16:53:52 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 17:38:52 2019 -0700" }, "message": "random: try to actively add entropy rather than passively wait for it\n\nFor 5.3 we had to revert a nice ext4 IO pattern improvement, because it\ncaused a bootup regression due to lack of entropy at bootup together\nwith arguably broken user space that was asking for secure random\nnumbers when it really didn\u0027t need to.\n\nSee commit 72dbcf721566 (Revert \"ext4: make __ext4_get_inode_loc plug\").\n\nThis aims to solve the issue by actively generating entropy noise using\nthe CPU cycle counter when waiting for the random number generator to\ninitialize. This only works when you have a high-frequency time stamp\ncounter available, but that\u0027s the case on all modern x86 CPU\u0027s, and on\nmost other modern CPU\u0027s too.\n\nWhat we do is to generate jitter entropy from the CPU cycle counter\nunder a somewhat complex load: calling the scheduler while also\nguaranteeing a certain amount of timing noise by also triggering a\ntimer.\n\nI\u0027m sure we can tweak this, and that people will want to look at other\nalternatives, but there\u0027s been a number of papers written on jitter\nentropy, and this should really be fairly conservative by crediting one\nbit of entropy for every timer-induced jump in the cycle counter. Not\nbecause the timer itself would be all that unpredictable, but because\nthe interaction between the timer and the loop is going to be.\n\nEven if (and perhaps particularly if) the timer actually happens on\nanother CPU, the cacheline interaction between the loop that reads the\ncycle counter and the timer itself firing is going to add perturbations\nto the cycle counter values that get mixed into the entropy pool.\n\nAs Thomas pointed out, with a modern out-of-order CPU, even quite simple\nloops show a fair amount of hard-to-predict timing variability even in\nthe absense of external interrupts. But this tries to take that further\nby actually having a fairly complex interaction.\n\nThis is not going to solve the entropy issue for architectures that have\nno CPU cycle counter, but it\u0027s not clear how (and if) that is solvable,\nand the hardware in question is largely starting to be irrelevant. And\nby doing this we can at least avoid some of the even more contentious\napproaches (like making the entropy waiting time out in order to avoid\nthe possibly unbounded waiting).\n\nCc: Ahmed Darwish \u003cdarwish.07@gmail.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\nCc: Nicholas Mc Guire \u003chofrat@opentech.at\u003e\nCc: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Kees Cook \u003ckeescook@chromium.org\u003e\nCc: Willy Tarreau \u003cw@1wt.eu\u003e\nCc: Alexander E. Patrakov \u003cpatrakov@gmail.com\u003e\nCc: Lennart Poettering \u003cmzxreary@0pointer.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "9bfd7319e8d353b8b81c4cfd4d7eced71adbfbb5", "tree": "862e64535dab42603f79d3d7969ceeb74c8e6269", "parents": [ "a4207a1c5edc3112b3538769024c7b2ad42b9fe9", "a4c8723a162e6244fb01944fbf446750575dba59" ], "author": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Sun Sep 29 11:19:25 2019 -0700" }, "committer": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Sun Sep 29 11:20:48 2019 -0700" }, "message": "Merge tag \u0027fixes-5.4-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap into arm/fixes\n\nFixes for omap variants\n\nFew fixes for ti-sysc interconnect target module driver for no-idle\nquirks that caused nfsroot to fail on some dra7 boards.\n\nAnd let\u0027s fixes to get LCD working again for logicpd board that got\nbroken a while back with removal of panel-dpi driver. We need to now\nuse generic CONFIG_DRM_PANEL_SIMPLE instead.\n\n* tag \u0027fixes-5.4-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:\n bus: ti-sysc: Remove unpaired sysc_clkdm_deny_idle()\n ARM: dts: logicpd-som-lv: Fix i2c2 and i2c3 Pin mux\n ARM: dts: am3517-evm: Fix missing video\n ARM: dts: logicpd-torpedo-baseboard: Fix missing video\n ARM: omap2plus_defconfig: Fix missing video\n bus: ti-sysc: Fix handling of invalid clocks\n bus: ti-sysc: Fix clock handling for no-idle quirks\n\nLink: https://lore.kernel.org/r/pull-1568819401-72461@atomide.com\nSigned-off-by: Olof Johansson \u003colof@lixom.net\u003e\n" }, { "commit": "a4207a1c5edc3112b3538769024c7b2ad42b9fe9", "tree": "739b56fe8aba751719cb7c19f528ce2bd0093f24", "parents": [ "b74d957f6317e60924d0d3c0c01750814c24611b", "61423712dbb86e02af4aa5de65b9041493c92cac" ], "author": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Sun Sep 29 11:19:18 2019 -0700" }, "committer": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Sun Sep 29 11:20:41 2019 -0700" }, "message": "Merge tag \u0027scmi-fixes-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux into arm/fixes\n\nARM SCMI fixes for v5.4\n\nCouple of fixes: one in scmi reset driver initialising missed scmi handle\nand an other in scmi reset API implementation fixing the assignment of\nreset state\n\n* tag \u0027scmi-fixes-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux:\n reset: reset-scmi: add missing handle initialisation\n firmware: arm_scmi: reset: fix reset_state assignment in scmi_domain_reset\n\nLink: https://lore.kernel.org/r/20190918142139.GA4370@bogus\nSigned-off-by: Olof Johansson \u003colof@lixom.net\u003e\n" }, { "commit": "a3c0e7b1fe1fc62bba5f591c4bc404eea96823b8", "tree": "494409b373e8d43e72c88d228e77f3d3aaf46a91", "parents": [ "939ca9f1751d1d65424f80b9284b6c18e78c7f4e", "4c806b897d6075bfa5067e524fb058c57ab64e7b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:33:41 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:33:41 2019 -0700" }, "message": "Merge tag \u0027libnvdimm-fixes-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm\n\nMore libnvdimm updates from Dan Williams:\n\n - Complete the reworks to interoperate with powerpc dynamic huge page\n sizes\n\n - Fix a crash due to missed accounting for the powerpc \u0027struct\n page\u0027-memmap mapping granularity\n\n - Fix badblock initialization for volatile (DRAM emulated) pmem ranges\n\n - Stop triggering request_key() notifications to userspace when\n NVDIMM-security is disabled / not present\n\n - Miscellaneous small fixups\n\n* tag \u0027libnvdimm-fixes-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:\n libnvdimm/region: Enable MAP_SYNC for volatile regions\n libnvdimm: prevent nvdimm from requesting key when security is disabled\n libnvdimm/region: Initialize bad block for volatile namespaces\n libnvdimm/nfit_test: Fix acpi_handle redefinition\n libnvdimm/altmap: Track namespace boundaries in altmap\n libnvdimm: Fix endian conversion issues \n libnvdimm/dax: Pick the right alignment default when creating dax devices\n powerpc/book3s64: Export has_transparent_hugepage() related functions.\n" }, { "commit": "939ca9f1751d1d65424f80b9284b6c18e78c7f4e", "tree": "fc09b1cee556f8652eb86e18a0190337c1d2d1b7", "parents": [ "9ecb3e10a9f2d3443c3bcd0dba42e7e3e2e67618", "6c375eccded41df8033ed55a1b785531b304fc67" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:24:23 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:24:23 2019 -0700" }, "message": "Merge branch \u0027linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal\n\nPull thermal SoC updates from Eduardo Valentin:\n \"This is a really small pull in the midst of a lot of pending patches.\n\n We are in the middle of restructuring how we are maintaining the\n thermal subsystem, as per discussion in our last LPC. For now, I am\n sending just some changes that were pending in my tree. Looking\n forward to get a more streamlined process in the next merge window\"\n\n* \u0027linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal:\n thermal: db8500: Rewrite to be a pure OF sensor\n thermal: db8500: Use dev helper variable\n thermal: db8500: Finalize device tree conversion\n thermal: thermal_mmio: remove some dead code\n" }, { "commit": "9ecb3e10a9f2d3443c3bcd0dba42e7e3e2e67618", "tree": "5fc1d25681485da17739fa2287bd8998791d8a79", "parents": [ "4d2af08ed08ce87c4fd2379fa857153631ee8537", "11af27f494086188620e7768e421894af93c126a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:20:16 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:20:16 2019 -0700" }, "message": "Merge branch \u0027i2c/for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull more i2c updates from Wolfram Sang:\n\n - make Lenovo Yoga C630 boot now that the dependencies are merged\n\n - restore BlockProcessCall for i801, accidently removed in this merge\n window\n\n - a bugfix for the riic driver\n\n - an improvement to the slave-eeprom driver which should have been in\n the first pull request but sadly got lost in the process\n\n* \u0027i2c/for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n i2c: slave-eeprom: Add read only mode\n i2c: i801: Bring back Block Process Call support for certain platforms\n i2c: riic: Clear NACK in tend isr\n i2c: qcom-geni: Disable DMA processing on the Lenovo Yoga C630\n" }, { "commit": "4d2af08ed08ce87c4fd2379fa857153631ee8537", "tree": "6e2c3891290440fb902145fa7a11837f7d17f694", "parents": [ "02dc96ef6c25f990452c114c59d75c368a1f4c8f", "2a78f9962565e53b78363eaf516eb052009e8020" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:00:14 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Sep 29 10:00:14 2019 -0700" }, "message": "Merge tag \u0027iommu-fixes-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu\n\nPull iommu fixes from Joerg Roedel:\n \"A couple of fixes for the AMD IOMMU driver have piled up:\n\n - Some fixes for the reworked IO page-table which caused memory leaks\n or did not allow to downgrade mappings under some conditions.\n\n - Locking fixes to fix a couple of possible races around accessing\n \u0027struct protection_domain\u0027. The races got introduced when the\n dma-ops path became lock-less in the fast-path\"\n\n* tag \u0027iommu-fixes-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:\n iommu/amd: Lock code paths traversing protection_domain-\u003edev_list\n iommu/amd: Lock dev_data in attach/detach code paths\n iommu/amd: Check for busy devices earlier in attach_device()\n iommu/amd: Take domain-\u003elock for complete attach/detach path\n iommu/amd: Remove amd_iommu_devtable_lock\n iommu/amd: Remove domain-\u003eupdated\n iommu/amd: Wait for completion of IOTLB flush in attach_device\n iommu/amd: Unmap all L7 PTEs when downgrading page-sizes\n iommu/amd: Introduce first_pte_l7() helper\n iommu/amd: Fix downgrading default page-sizes in alloc_pte()\n iommu/amd: Fix pages leak in free_pagetable()\n" }, { "commit": "dc925a36060e8cef050a9d05c64dae1c30dc5027", "tree": "55f97dbc8e3f3b808c2f07ca0ee3f8b8aeb1a07f", "parents": [ "38c7a30a9d5ff8dd0c7a9ab1441aa5a96cf60afa" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Sep 25 10:29:49 2019 +0200" }, "committer": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Sun Sep 29 12:43:18 2019 +0200" }, "message": "Documentation/process: Clarify disclosure rules\n\nThe role of the contact list provided by the disclosing party and how it\naffects the disclosure process and the ability to include experts into\nthe development process is not really well explained.\n\nNeither is it entirely clear when the disclosing party will be informed\nabout the fact that a developer who is not covered by an employer NDA needs\nto be brought in and disclosed.\n\nExplain the role of the contact list and the information policy along with\nan eventual conflict resolution better.\n\nReported-by: Dave Hansen \u003cdave.hansen@linux.intel.com\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nAcked-by: Dave Hansen \u003cdave.hansen@linux.intel.com\u003e\nLink: https://lore.kernel.org/r/alpine.DEB.2.21.1909251028390.10825@nanos.tec.linutronix.de\nSigned-off-by: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\n" }, { "commit": "02dc96ef6c25f990452c114c59d75c368a1f4c8f", "tree": "bddda0591191f65931195d3171743507d6cae7d6", "parents": [ "edf445ad7c8d58c2784a5b733790e80999093d8f", "faeacb6ddb13b7a020b50b9246fe098653cfbd6e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 17:47:33 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 17:47:33 2019 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from David Miller:\n\n 1) Sanity check URB networking device parameters to avoid divide by\n zero, from Oliver Neukum.\n\n 2) Disable global multicast filter in NCSI, otherwise LLDP and IPV6\n don\u0027t work properly. Longer term this needs a better fix tho. From\n Vijay Khemka.\n\n 3) Small fixes to selftests (use ping when ping6 is not present, etc.)\n from David Ahern.\n\n 4) Bring back rt_uses_gateway member of struct rtable, it\u0027s semantics\n were not well understood and trying to remove it broke things. From\n David Ahern.\n\n 5) Move usbnet snaity checking, ignore endpoints with invalid\n wMaxPacketSize. From Bjørn Mork.\n\n 6) Missing Kconfig deps for sja1105 driver, from Mao Wenan.\n\n 7) Various small fixes to the mlx5 DR steering code, from Alaa Hleihel,\n Alex Vesker, and Yevgeny Kliteynik\n\n 8) Missing CAP_NET_RAW checks in various places, from Ori Nimron.\n\n 9) Fix crash when removing sch_cbs entry while offloading is enabled,\n from Vinicius Costa Gomes.\n\n10) Signedness bug fixes, generally in looking at the result given by\n of_get_phy_mode() and friends. From Dan Crapenter.\n\n11) Disable preemption around BPF_PROG_RUN() calls, from Eric Dumazet.\n\n12) Don\u0027t create VRF ipv6 rules if ipv6 is disabled, from David Ahern.\n\n13) Fix quantization code in tcp_bbr, from Kevin Yang.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (127 commits)\n net: tap: clean up an indentation issue\n nfp: abm: fix memory leak in nfp_abm_u32_knode_replace\n tcp: better handle TCP_USER_TIMEOUT in SYN_SENT state\n sk_buff: drop all skb extensions on free and skb scrubbing\n tcp_bbr: fix quantization code to not raise cwnd if not probing bandwidth\n mlxsw: spectrum_flower: Fail in case user specifies multiple mirror actions\n Documentation: Clarify trap\u0027s description\n mlxsw: spectrum: Clear VLAN filters during port initialization\n net: ena: clean up indentation issue\n NFC: st95hf: clean up indentation issue\n net: phy: micrel: add Asym Pause workaround for KSZ9021\n net: socionext: ave: Avoid using netdev_err() before calling register_netdev()\n ptp: correctly disable flags on old ioctls\n lib: dimlib: fix help text typos\n net: dsa: microchip: Always set regmap stride to 1\n nfp: flower: fix memory leak in nfp_flower_spawn_vnic_reprs\n nfp: flower: prevent memory leak in nfp_flower_spawn_phy_reprs\n net/sched: Set default of CONFIG_NET_TC_SKB_EXT to N\n vrf: Do not attempt to create IPv6 mcast rule if IPv6 is disabled\n net: sched: sch_sfb: don\u0027t call qdisc_put() while holding tree lock\n ...\n" }, { "commit": "edf445ad7c8d58c2784a5b733790e80999093d8f", "tree": "2ffeaee2454bf0b03530c22ac23b4eb5edb4d89d", "parents": [ "a2953204b576ea3ba4afd07b917811d50fc49778", "76e654cc91bbe627aa6067916f02a4d3ac041620" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:26:47 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:26:47 2019 -0700" }, "message": "Merge branch \u0027hugepage-fallbacks\u0027 (hugepatch patches from David Rientjes)\n\nMerge hugepage allocation updates from David Rientjes:\n \"We (mostly Linus, Andrea, and myself) have been discussing offlist how\n to implement a sane default allocation strategy for hugepages on NUMA\n platforms.\n\n With these reverts in place, the page allocator will happily allocate\n a remote hugepage immediately rather than try to make a local hugepage\n available. This incurs a substantial performance degradation when\n memory compaction would have otherwise made a local hugepage\n available.\n\n This series reverts those reverts and attempts to propose a more sane\n default allocation strategy specifically for hugepages. Andrea\n acknowledges this is likely to fix the swap storms that he originally\n reported that resulted in the patches that removed __GFP_THISNODE from\n hugepage allocations.\n\n The immediate goal is to return 5.3 to the behavior the kernel has\n implemented over the past several years so that remote hugepages are\n not immediately allocated when local hugepages could have been made\n available because the increased access latency is untenable.\n\n The next goal is to introduce a sane default allocation strategy for\n hugepages allocations in general regardless of the configuration of\n the system so that we prevent thrashing of local memory when\n compaction is unlikely to succeed and can prefer remote hugepages over\n remote native pages when the local node is low on memory.\"\n\nNote on timing: this reverts the hugepage VM behavior changes that got\nintroduced fairly late in the 5.3 cycle, and that fixed a huge\nperformance regression for certain loads that had been around since\n4.18.\n\nAndrea had this note:\n\n \"The regression of 4.18 was that it was taking hours to start a VM\n where 3.10 was only taking a few seconds, I reported all the details\n on lkml when it was finally tracked down in August 2018.\n\n https://lore.kernel.org/linux-mm/20180820032640.9896-2-aarcange@redhat.com/\n\n __GFP_THISNODE in MADV_HUGEPAGE made the above enterprise vfio\n workload degrade like in the \"current upstream\" above. And it still\n would have been that bad as above until 5.3-rc5\"\n\nwhere the bad behavior ends up happening as you fill up a local node,\nand without that change, you\u0027d get into the nasty swap storm behavior\ndue to compaction working overtime to make room for more memory on the\nnodes.\n\nAs a result 5.3 got the two performance fix reverts in rc5.\n\nHowever, David Rientjes then noted that those performance fixes in turn\nregressed performance for other loads - although not quite to the same\ndegree. He suggested reverting the reverts and instead replacing them\nwith two small changes to how hugepage allocations are done (patch\ndescriptions rephrased by me):\n\n - \"avoid expensive reclaim when compaction may not succeed\": just admit\n that the allocation failed when you\u0027re trying to allocate a huge-page\n and compaction wasn\u0027t successful.\n\n - \"allow hugepage fallback to remote nodes when madvised\": when that\n node-local huge-page allocation failed, retry without forcing the\n local node.\n\nbut by then I judged it too late to replace the fixes for a 5.3 release.\nSo 5.3 was released with behavior that harked back to the pre-4.18 logic.\n\nBut now we\u0027re in the merge window for 5.4, and we can see if this\nalternate model fixes not just the horrendous swap storm behavior, but\nalso restores the performance regression that the late reverts caused.\n\nFingers crossed.\n\n* emailed patches from David Rientjes \u003crientjes@google.com\u003e:\n mm, page_alloc: allow hugepage fallback to remote nodes when madvised\n mm, page_alloc: avoid expensive reclaim when compaction may not succeed\n Revert \"Revert \"Revert \"mm, thp: consolidate THP gfp handling into alloc_hugepage_direct_gfpmask\"\"\n Revert \"Revert \"mm, thp: restore node-local hugepage allocations\"\"\n" }, { "commit": "8ed4889eb83179dbc9a105cfed65cc42ecb61097", "tree": "5a75f2d6833fe3c07277bfc243c519b56ae3aadd", "parents": [ "f7d6316fb43735ae8af969c2e582fbee85709483" ], "author": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Fri Sep 27 11:10:22 2019 -0400" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Sat Sep 28 17:13:40 2019 -0400" }, "message": "selftests/ftrace: Fix same probe error test\n\nThe \"same probe\" selftest that tests that adding the same probe fails\ndoesn\u0027t add the same probe and passes, which fails the test.\n\nFixes: b78b94b82122 (\"selftests/ftrace: Update kprobe event error testcase\")\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "f7d6316fb43735ae8af969c2e582fbee85709483", "tree": "dee179100411da66f4f41bc622d7ea4c83c65f7d", "parents": [ "96c5c6e6a5b6db592acae039fed54b5c8844cd35" ], "author": { "name": "Changbin Du", "email": "changbin.du@gmail.com", "time": "Sat Sep 14 18:32:15 2019 +0800" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Sat Sep 28 17:13:39 2019 -0400" }, "message": "mm, tracing: Print symbol name for call_site in trace events\n\nTo improve the readability of raw slab trace points, print the call_site ip\nusing \u0027%pS\u0027. Then we can grep events with function names.\n\n[002] .... 808.188897: kmem_cache_free: call_site\u003dputname+0x47/0x50 ptr\u003d00000000cef40c80\n[002] .... 808.188898: kfree: call_site\u003dsecurity_cred_free+0x42/0x50 ptr\u003d0000000062400820\n[002] .... 808.188904: kmem_cache_free: call_site\u003dput_cred_rcu+0x88/0xa0 ptr\u003d0000000058d74ef8\n[002] .... 808.188913: kmem_cache_alloc: call_site\u003dprepare_creds+0x26/0x100 ptr\u003d0000000058d74ef8 bytes_req\u003d168 bytes_alloc\u003d576 gfp_flags\u003dGFP_KERNEL\n[002] .... 808.188917: kmalloc: call_site\u003dsecurity_prepare_creds+0x77/0xa0 ptr\u003d0000000062400820 bytes_req\u003d8 bytes_alloc\u003d336 gfp_flags\u003dGFP_KERNEL|__GFP_ZERO\n[002] .... 808.188920: kmem_cache_alloc: call_site\u003dgetname_flags+0x4f/0x1e0 ptr\u003d00000000cef40c80 bytes_req\u003d4096 bytes_alloc\u003d4480 gfp_flags\u003dGFP_KERNEL\n[002] .... 808.188925: kmem_cache_free: call_site\u003dputname+0x47/0x50 ptr\u003d00000000cef40c80\n[002] .... 808.188926: kfree: call_site\u003dsecurity_cred_free+0x42/0x50 ptr\u003d0000000062400820\n[002] .... 808.188931: kmem_cache_free: call_site\u003dput_cred_rcu+0x88/0xa0 ptr\u003d0000000058d74ef8\n\nLink: http://lkml.kernel.org/r/20190914103215.23301-1-changbin.du@gmail.com\n\nSigned-off-by: Changbin Du \u003cchangbin.du@gmail.com\u003e\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "96c5c6e6a5b6db592acae039fed54b5c8844cd35", "tree": "70a42fe11bc971e3048624439a8d7c7ef11a9004", "parents": [ "968e5170939662341242812b9c82ef51cf140a33" ], "author": { "name": "Navid Emamdoost", "email": "navid.emamdoost@gmail.com", "time": "Fri Sep 20 17:57:59 2019 -0500" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Sat Sep 28 17:13:39 2019 -0400" }, "message": "tracing: Have error path in predicate_parse() free its allocated memory\n\nIn predicate_parse, there is an error path that is not going to\nout_free instead it returns directly which leads to a memory leak.\n\nLink: http://lkml.kernel.org/r/20190920225800.3870-1-navid.emamdoost@gmail.com\n\nSigned-off-by: Navid Emamdoost \u003cnavid.emamdoost@gmail.com\u003e\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "968e5170939662341242812b9c82ef51cf140a33", "tree": "55b5f965485179f55046d23c82426240baf62f4d", "parents": [ "d2aea95a1a4d195d939d16303700921be318a2b9" ], "author": { "name": "Nathan Chancellor", "email": "natechancellor@gmail.com", "time": "Thu Sep 26 09:22:59 2019 -0700" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Sat Sep 28 17:13:39 2019 -0400" }, "message": "tracing: Fix clang -Wint-in-bool-context warnings in IF_ASSIGN macro\n\nAfter r372664 in clang, the IF_ASSIGN macro causes a couple hundred\nwarnings along the lines of:\n\nkernel/trace/trace_output.c:1331:2: warning: converting the enum\nconstant to a boolean [-Wint-in-bool-context]\nkernel/trace/trace.h:409:3: note: expanded from macro\n\u0027trace_assign_type\u0027\n IF_ASSIGN(var, ent, struct ftrace_graph_ret_entry,\n ^\nkernel/trace/trace.h:371:14: note: expanded from macro \u0027IF_ASSIGN\u0027\n WARN_ON(id \u0026\u0026 (entry)-\u003etype !\u003d id); \\\n ^\n264 warnings generated.\n\nThis warning can catch issues with constructs like:\n\n if (state \u003d\u003d A || B)\n\nwhere the developer really meant:\n\n if (state \u003d\u003d A || state \u003d\u003d B)\n\nThis is currently the only occurrence of the warning in the kernel\ntree across defconfig, allyesconfig, allmodconfig for arm32, arm64,\nand x86_64. Add the implicit \u0027!\u003d 0\u0027 to the WARN_ON statement to fix\nthe warnings and find potential issues in the future.\n\nLink: https://github.com/llvm/llvm-project/commit/28b38c277a2941e9e891b2db30652cfd962f070b\nLink: https://github.com/ClangBuiltLinux/linux/issues/686\nLink: http://lkml.kernel.org/r/20190926162258.466321-1-natechancellor@gmail.com\n\nReviewed-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nSigned-off-by: Nathan Chancellor \u003cnatechancellor@gmail.com\u003e\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "d2aea95a1a4d195d939d16303700921be318a2b9", "tree": "adff5b3ea6cd7e23ff2bffa66af453fefd264b6a", "parents": [ "f8d7ab2bded897607bff6324d5c6ea6b4aecca0c" ], "author": { "name": "Masami Hiramatsu", "email": "mhiramat@kernel.org", "time": "Sat Sep 28 05:53:29 2019 -0400" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Sat Sep 28 17:07:53 2019 -0400" }, "message": "tracing/probe: Fix to check the difference of nr_args before adding probe\n\nSteven reported that a test triggered:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n BUG: KASAN: slab-out-of-bounds in trace_kprobe_create+0xa9e/0xe40\n Read of size 8 at addr ffff8880c4f25a48 by task ftracetest/4798\n\n CPU: 2 PID: 4798 Comm: ftracetest Not tainted 5.3.0-rc6-test+ #30\n Hardware name: Hewlett-Packard HP Compaq Pro 6300 SFF/339A, BIOS K01 v03.03 07/14/2016\n Call Trace:\n dump_stack+0x7c/0xc0\n ? trace_kprobe_create+0xa9e/0xe40\n print_address_description+0x6c/0x332\n ? trace_kprobe_create+0xa9e/0xe40\n ? trace_kprobe_create+0xa9e/0xe40\n __kasan_report.cold.6+0x1a/0x3b\n ? trace_kprobe_create+0xa9e/0xe40\n kasan_report+0xe/0x12\n trace_kprobe_create+0xa9e/0xe40\n ? print_kprobe_event+0x280/0x280\n ? match_held_lock+0x1b/0x240\n ? find_held_lock+0xac/0xd0\n ? fs_reclaim_release.part.112+0x5/0x20\n ? lock_downgrade+0x350/0x350\n ? kasan_unpoison_shadow+0x30/0x40\n ? __kasan_kmalloc.constprop.6+0xc1/0xd0\n ? trace_kprobe_create+0xe40/0xe40\n ? trace_kprobe_create+0xe40/0xe40\n create_or_delete_trace_kprobe+0x2e/0x60\n trace_run_command+0xc3/0xe0\n ? trace_panic_handler+0x20/0x20\n ? kasan_unpoison_shadow+0x30/0x40\n trace_parse_run_command+0xdc/0x163\n vfs_write+0xe1/0x240\n ksys_write+0xba/0x150\n ? __ia32_sys_read+0x50/0x50\n ? tracer_hardirqs_on+0x61/0x180\n ? trace_hardirqs_off_caller+0x43/0x110\n ? mark_held_locks+0x29/0xa0\n ? do_syscall_64+0x14/0x260\n do_syscall_64+0x68/0x260\n\nFix to check the difference of nr_args before adding probe\non existing probes. This also may set the error log index\nbigger than the number of command parameters. In that case\nit sets the error position is next to the last parameter.\n\nLink: http://lkml.kernel.org/r/156966474783.3478.13217501608215769150.stgit@devnote2\n\nFixes: ca89bc071d5e (\"tracing/kprobe: Add multi-probe per event support\")\nReported-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\nSigned-off-by: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "76e654cc91bbe627aa6067916f02a4d3ac041620", "tree": "338187c038bfaf4d2780ffce7e5ad6359dc97806", "parents": [ "b39d0ee2632d2f4fb180e8e4eba33736283f23de" ], "author": { "name": "David Rientjes", "email": "rientjes@google.com", "time": "Wed Sep 04 12:54:25 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:05:38 2019 -0700" }, "message": "mm, page_alloc: allow hugepage fallback to remote nodes when madvised\n\nFor systems configured to always try hard to allocate transparent\nhugepages (thp defrag setting of \"always\") or for memory that has been\nexplicitly madvised to MADV_HUGEPAGE, it is often better to fallback to\nremote memory to allocate the hugepage if the local allocation fails\nfirst.\n\nThe point is to allow the initial call to __alloc_pages_node() to attempt\nto defragment local memory to make a hugepage available, if possible,\nrather than immediately fallback to remote memory. Local hugepages will\nalways have a better access latency than remote (huge)pages, so an attempt\nto make a hugepage available locally is always preferred.\n\nIf memory compaction cannot be successful locally, however, it is likely\nbetter to fallback to remote memory. This could take on two forms: either\nallow immediate fallback to remote memory or do per-zone watermark checks.\nIt would be possible to fallback only when per-zone watermarks fail for\norder-0 memory, since that would require local reclaim for all subsequent\nfaults so remote huge allocation is likely better than thrashing the local\nzone for large workloads.\n\nIn this case, it is assumed that because the system is configured to try\nhard to allocate hugepages or the vma is advised to explicitly want to try\nhard for hugepages that remote allocation is better when local allocation\nand memory compaction have both failed.\n\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nCc: Stefan Priebe - Profihost AG \u003cs.priebe@profihost.ag\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill@shutemov.name\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b39d0ee2632d2f4fb180e8e4eba33736283f23de", "tree": "624651ef41b223a7846d46b68c67b0cfec20d306", "parents": [ "19deb7695e072deaff025e03de40c61b525bd57e" ], "author": { "name": "David Rientjes", "email": "rientjes@google.com", "time": "Wed Sep 04 12:54:22 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:05:38 2019 -0700" }, "message": "mm, page_alloc: avoid expensive reclaim when compaction may not succeed\n\nMemory compaction has a couple significant drawbacks as the allocation\norder increases, specifically:\n\n - isolate_freepages() is responsible for finding free pages to use as\n migration targets and is implemented as a linear scan of memory\n starting at the end of a zone,\n\n - failing order-0 watermark checks in memory compaction does not account\n for how far below the watermarks the zone actually is: to enable\n migration, there must be *some* free memory available. Per the above,\n watermarks are not always suffficient if isolate_freepages() cannot\n find the free memory but it could require hundreds of MBs of reclaim to\n even reach this threshold (read: potentially very expensive reclaim with\n no indication compaction can be successful), and\n\n - if compaction at this order has failed recently so that it does not even\n run as a result of deferred compaction, looping through reclaim can often\n be pointless.\n\nFor hugepage allocations, these are quite substantial drawbacks because\nthese are very high order allocations (order-9 on x86) and falling back to\ndoing reclaim can potentially be *very* expensive without any indication\nthat compaction would even be successful.\n\nReclaim itself is unlikely to free entire pageblocks and certainly no\nreliance should be put on it to do so in isolation (recall lumpy reclaim).\nThis means we should avoid reclaim and simply fail hugepage allocation if\ncompaction is deferred.\n\nIt is also not helpful to thrash a zone by doing excessive reclaim if\ncompaction may not be able to access that memory. If order-0 watermarks\nfail and the allocation order is sufficiently large, it is likely better\nto fail the allocation rather than thrashing the zone.\n\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nCc: Stefan Priebe - Profihost AG \u003cs.priebe@profihost.ag\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill@shutemov.name\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "19deb7695e072deaff025e03de40c61b525bd57e", "tree": "cdbb63365d5f464d4e2660ba467ec062bd5f450b", "parents": [ "ac79f78dab892fcdc11fda8af5cc5e80d09dca8a" ], "author": { "name": "David Rientjes", "email": "rientjes@google.com", "time": "Wed Sep 04 12:54:20 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:05:38 2019 -0700" }, "message": "Revert \"Revert \"Revert \"mm, thp: consolidate THP gfp handling into alloc_hugepage_direct_gfpmask\"\"\n\nThis reverts commit 92717d429b38e4f9f934eed7e605cc42858f1839.\n\nSince commit a8282608c88e (\"Revert \"mm, thp: restore node-local hugepage\nallocations\"\") is reverted in this series, it is better to restore the\nprevious 5.2 behavior between the thp allocation and the page allocator\nrather than to attempt any consolidation or cleanup for a policy that is\nnow reverted. It\u0027s less risky during an rc cycle and subsequent patches\nin this series further modify the same policy that the pre-5.3 behavior\nimplements.\n\nConsolidation and cleanup can be done subsequent to a sane default page\nallocation strategy, so this patch reverts a cleanup done on a strategy\nthat is now reverted and thus is the least risky option.\n\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nCc: Stefan Priebe - Profihost AG \u003cs.priebe@profihost.ag\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill@shutemov.name\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ac79f78dab892fcdc11fda8af5cc5e80d09dca8a", "tree": "e774a6c484bb2272dbbcb3d7c9253dd92fac5d39", "parents": [ "4d856f72c10ecb060868ed10ff1b1453943fc6c8" ], "author": { "name": "David Rientjes", "email": "rientjes@google.com", "time": "Wed Sep 04 12:54:18 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 14:05:38 2019 -0700" }, "message": "Revert \"Revert \"mm, thp: restore node-local hugepage allocations\"\"\n\nThis reverts commit a8282608c88e08b1782141026eab61204c1e533f.\n\nThe commit references the original intended semantic for MADV_HUGEPAGE\nwhich has subsequently taken on three unique purposes:\n\n - enables or disables thp for a range of memory depending on the system\u0027s\n config (is thp \"enabled\" set to \"always\" or \"madvise\"),\n\n - determines the synchronous compaction behavior for thp allocations at\n fault (is thp \"defrag\" set to \"always\", \"defer+madvise\", or \"madvise\"),\n and\n\n - reverts a previous MADV_NOHUGEPAGE (there is no madvise mode to only\n clear previous hugepage advice).\n\nThese are the three purposes that currently exist in 5.2 and over the\npast several years that userspace has been written around. Adding a\nNUMA locality preference adds a fourth dimension to an already conflated\nadvice mode.\n\nBased on the semantic that MADV_HUGEPAGE has provided over the past\nseveral years, there exist workloads that use the tunable based on these\nprinciples: specifically that the allocation should attempt to\ndefragment a local node before falling back. It is agreed that remote\nhugepages typically (but not always) have a better access latency than\nremote native pages, although on Naples this is at parity for\nintersocket.\n\nThe revert commit that this patch reverts allows hugepage allocation to\nimmediately allocate remotely when local memory is fragmented. This is\ncontrary to the semantic of MADV_HUGEPAGE over the past several years:\nthat is, memory compaction should be attempted locally before falling\nback.\n\nThe performance degradation of remote hugepages over local hugepages on\nRome, for example, is 53.5% increased access latency. For this reason,\nthe goal is to revert back to the 5.2 and previous behavior that would\nattempt local defragmentation before falling back. With the patch that\nis reverted by this patch, we see performance degradations at the tail\nbecause the allocator happily allocates the remote hugepage rather than\neven attempting to make a local hugepage available.\n\nzone_reclaim_mode is not a solution to this problem since it does not\nonly impact hugepage allocations but rather changes the memory\nallocation strategy for *all* page allocations.\n\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nCc: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nCc: Stefan Priebe - Profihost AG \u003cs.priebe@profihost.ag\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill@shutemov.name\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "a2953204b576ea3ba4afd07b917811d50fc49778", "tree": "04eee2a327a8871ee991ece819cb8945728cc9ec", "parents": [ "f19e00ee84be14e840386cb4f3c0bda5b9cfb5ab", "253c892193ab58da6b1d94371285971b22c63260" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 13:43:00 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 13:43:00 2019 -0700" }, "message": "Merge tag \u0027powerpc-5.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Michael Ellerman:\n \"An assortment of fixes that were either missed by me, or didn\u0027t arrive\n quite in time for the first v5.4 pull.\n\n - Most notable is a fix for an issue with tlbie (broadcast TLB\n invalidation) on Power9, when using the Radix MMU. The tlbie can\n race with an mtpid (move to PID register, essentially MMU context\n switch) on another thread of the core, which can cause stores to\n continue to go to a page after it\u0027s unmapped.\n\n - A fix in our KVM code to add a missing barrier, the lack of which\n has been observed to cause missed IPIs and subsequently stuck CPUs\n in the host.\n\n - A change to the way we initialise PCR (Processor Compatibility\n Register) to make it forward compatible with future CPUs.\n\n - On some older PowerVM systems our H_BLOCK_REMOVE support could\n oops, fix it to detect such systems and fallback to the old\n invalidation method.\n\n - A fix for an oops seen on some machines when using KASAN on 32-bit.\n\n - A handful of other minor fixes, and two new selftests.\n\n Thanks to: Alistair Popple, Aneesh Kumar K.V, Christophe Leroy,\n Gustavo Romero, Joel Stanley, Jordan Niethe, Laurent Dufour, Michael\n Roth, Oliver O\u0027Halloran\"\n\n* tag \u0027powerpc-5.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/eeh: Fix eeh eeh_debugfs_break_device() with SRIOV devices\n powerpc/nvdimm: use H_SCM_QUERY hcall on H_OVERLAP error\n powerpc/nvdimm: Use HCALL error as the return value\n selftests/powerpc: Add test case for tlbie vs mtpidr ordering issue\n powerpc/mm: Fixup tlbie vs mtpidr/mtlpidr ordering issue on POWER9\n powerpc/book3s64/radix: Rename CPU_FTR_P9_TLBIE_BUG feature flag\n powerpc/book3s64/mm: Don\u0027t do tlbie fixup for some hardware revisions\n powerpc/pseries: Call H_BLOCK_REMOVE when supported\n powerpc/pseries: Read TLB Block Invalidate Characteristics\n KVM: PPC: Book3S HV: use smp_mb() when setting/clearing host_ipi flag\n powerpc/mm: Fix an Oops in kasan_mmu_init()\n powerpc/mm: Add a helper to select PAGE_KERNEL_RO or PAGE_READONLY\n powerpc/64s: Set reserved PCR bits\n powerpc: Fix definition of PCR bits to work with old binutils\n powerpc/book3s64/radix: Remove WARN_ON in destroy_context()\n powerpc/tm: Add tm-poison test\n" }, { "commit": "f19e00ee84be14e840386cb4f3c0bda5b9cfb5ab", "tree": "6881f4545023996284eb18004929a4ec147ba674", "parents": [ "9c5efe9ae7df78600c0ee7bcce27516eb687fa6e", "ca14c996afe7228ff9b480cf225211cc17212688" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 13:37:41 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 13:37:41 2019 -0700" }, "message": "Merge branch \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fix from Ingo Molnar:\n \"A kexec fix for the case when GCC_PLUGIN_STACKLEAK\u003dy is enabled\"\n\n* \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/purgatory: Disable the stackleak GCC plugin for the purgatory\n" }, { "commit": "9c5efe9ae7df78600c0ee7bcce27516eb687fa6e", "tree": "158cfb9720d876e68a14a4cccaffeb58fb7baac5", "parents": [ "aefcf2f4b58155d27340ba5f9ddbe9513da8286d", "4892f51ad54ddff2883a60b6ad4323c1f632a9d6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 12:39:07 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 12:39:07 2019 -0700" }, "message": "Merge branch \u0027sched-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull scheduler fixes from Ingo Molnar:\n\n - Apply a number of membarrier related fixes and cleanups, which fixes\n a use-after-free race in the membarrier code\n\n - Introduce proper RCU protection for tasks on the runqueue - to get\n rid of the subtle task_rcu_dereference() interface that was easy to\n get wrong\n\n - Misc fixes, but also an EAS speedup\n\n* \u0027sched-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n sched/fair: Avoid redundant EAS calculation\n sched/core: Remove double update_max_interval() call on CPU startup\n sched/core: Fix preempt_schedule() interrupt return comment\n sched/fair: Fix -Wunused-but-set-variable warnings\n sched/core: Fix migration to invalid CPU in __set_cpus_allowed_ptr()\n sched/membarrier: Return -ENOMEM to userspace on memory allocation failure\n sched/membarrier: Skip IPIs when mm-\u003emm_users \u003d\u003d 1\n selftests, sched/membarrier: Add multi-threaded test\n sched/membarrier: Fix p-\u003emm-\u003emembarrier_state racy load\n sched/membarrier: Call sync_core only before usermode for same mm\n sched/membarrier: Remove redundant check\n sched/membarrier: Fix private expedited registration check\n tasks, sched/core: RCUify the assignment of rq-\u003ecurr\n tasks, sched/core: With a grace period after finish_task_switch(), remove unnecessary code\n tasks, sched/core: Ensure tasks are available for a grace period after leaving the runqueue\n tasks: Add a count of task RCU users\n sched/core: Convert vcpu_is_preempted() from macro to an inline function\n sched/fair: Remove unused cfs_rq_clock_task() function\n" }, { "commit": "11af27f494086188620e7768e421894af93c126a", "tree": "4678ba6525f599857ac3303766d5999164c834f2", "parents": [ "fd4b204a0971854c35795ab60b3673a5b57dfebc" ], "author": { "name": "Björn Ardö", "email": "bjorn.ardo@axis.com", "time": "Fri Sep 06 16:06:09 2019 +0200" }, "committer": { "name": "Wolfram Sang", "email": "wsa@the-dreams.de", "time": "Sat Sep 28 20:44:12 2019 +0200" }, "message": "i2c: slave-eeprom: Add read only mode\n\nAdd read-only versions of all EEPROMs. These versions are read-only\non the i2c side, but can be written from the sysfs side.\n\nSigned-off-by: Björn Ardö \u003cbjorn.ardo@axis.com\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@the-dreams.de\u003e\n" }, { "commit": "fd4b204a0971854c35795ab60b3673a5b57dfebc", "tree": "7d8130fc60e9005c0eb3ddf4ebc89b514b2177a9", "parents": [ "a71e2ac1f32097fbb2beab098687a7a95c84543e" ], "author": { "name": "Jarkko Nikula", "email": "jarkko.nikula@linux.intel.com", "time": "Fri Sep 27 14:09:11 2019 +0300" }, "committer": { "name": "Wolfram Sang", "email": "wsa@the-dreams.de", "time": "Sat Sep 28 20:44:12 2019 +0200" }, "message": "i2c: i801: Bring back Block Process Call support for certain platforms\n\nCommit b84398d6d7f9 (\"i2c: i801: Use iTCO version 6 in Cannon Lake PCH\nand beyond\") looks like to drop by accident Block Write-Block Read Process\nCall support for Intel Sunrisepoint, Lewisburg, Denverton and Kaby Lake.\n\nThat support was added for above and newer platforms by the commit\n315cd67c9453 (\"i2c: i801: Add Block Write-Block Read Process Call\nsupport\") so bring it back for above platforms.\n\nFixes: b84398d6d7f9 (\"i2c: i801: Use iTCO version 6 in Cannon Lake PCH and beyond\")\nSigned-off-by: Jarkko Nikula \u003cjarkko.nikula@linux.intel.com\u003e\nReviewed-by: Alexander Sverdlin \u003calexander.sverdlin@nokia.com\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@the-dreams.de\u003e\n" }, { "commit": "a71e2ac1f32097fbb2beab098687a7a95c84543e", "tree": "0abdf47803a81d7e2bae28f50351d7ad65bd1c76", "parents": [ "127068abe85bf3dee50df51cb039a5a987a4a666" ], "author": { "name": "Chris Brandt", "email": "chris.brandt@renesas.com", "time": "Thu Sep 26 07:19:09 2019 -0500" }, "committer": { "name": "Wolfram Sang", "email": "wsa@the-dreams.de", "time": "Sat Sep 28 20:44:12 2019 +0200" }, "message": "i2c: riic: Clear NACK in tend isr\n\nThe NACKF flag should be cleared in INTRIICNAKI interrupt processing as\ndescription in HW manual.\n\nThis issue shows up quickly when PREEMPT_RT is applied and a device is\nprobed that is not plugged in (like a touchscreen controller). The result\nis endless interrupts that halt system boot.\n\nFixes: 310c18a41450 (\"i2c: riic: add driver\")\nCc: stable@vger.kernel.org\nReported-by: Chien Nguyen \u003cchien.nguyen.eb@rvc.renesas.com\u003e\nSigned-off-by: Chris Brandt \u003cchris.brandt@renesas.com\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@the-dreams.de\u003e\n" }, { "commit": "127068abe85bf3dee50df51cb039a5a987a4a666", "tree": "65c5b3adc3cdd995e5a2e7d02f1aba6c5eda7bba", "parents": [ "aefcf2f4b58155d27340ba5f9ddbe9513da8286d" ], "author": { "name": "Lee Jones", "email": "lee.jones@linaro.org", "time": "Thu Sep 05 20:24:12 2019 +0100" }, "committer": { "name": "Wolfram Sang", "email": "wsa@the-dreams.de", "time": "Sat Sep 28 19:47:04 2019 +0200" }, "message": "i2c: qcom-geni: Disable DMA processing on the Lenovo Yoga C630\n\nWe have a production-level laptop (Lenovo Yoga C630) which is exhibiting\na rather horrific bug. When I2C HID devices are being scanned for at\nboot-time the QCom Geni based I2C (Serial Engine) attempts to use DMA.\nWhen it does, the laptop reboots and the user never sees the OS.\n\nAttempts are being made to debug the reason for the spontaneous reboot.\nNo luck so far, hence the requirement for this hot-fix. This workaround\nwill be removed once we have a viable fix.\n\nSigned-off-by: Lee Jones \u003clee.jones@linaro.org\u003e\nTested-by: Bjorn Andersson \u003cbjorn.andersson@linaro.org\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@the-dreams.de\u003e\n" }, { "commit": "aefcf2f4b58155d27340ba5f9ddbe9513da8286d", "tree": "a8ae994ca71b2e43e3075c796165e5ed62e28e42", "parents": [ "f1f2f614d535564992f32e720739cb53cf03489f", "45893a0abee6b5fd52994a3a1095735aeaec472b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 08:14:15 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Sep 28 08:14:15 2019 -0700" }, "message": "Merge branch \u0027next-lockdown\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security\n\nPull kernel lockdown mode from James Morris:\n \"This is the latest iteration of the kernel lockdown patchset, from\n Matthew Garrett, David Howells and others.\n\n From the original description:\n\n This patchset introduces an optional kernel lockdown feature,\n intended to strengthen the boundary between UID 0 and the kernel.\n When enabled, various pieces of kernel functionality are restricted.\n Applications that rely on low-level access to either hardware or the\n kernel may cease working as a result - therefore this should not be\n enabled without appropriate evaluation beforehand.\n\n The majority of mainstream distributions have been carrying variants\n of this patchset for many years now, so there\u0027s value in providing a\n doesn\u0027t meet every distribution requirement, but gets us much closer\n to not requiring external patches.\n\n There are two major changes since this was last proposed for mainline:\n\n - Separating lockdown from EFI secure boot. Background discussion is\n covered here: https://lwn.net/Articles/751061/\n\n - Implementation as an LSM, with a default stackable lockdown LSM\n module. This allows the lockdown feature to be policy-driven,\n rather than encoding an implicit policy within the mechanism.\n\n The new locked_down LSM hook is provided to allow LSMs to make a\n policy decision around whether kernel functionality that would allow\n tampering with or examining the runtime state of the kernel should be\n permitted.\n\n The included lockdown LSM provides an implementation with a simple\n policy intended for general purpose use. This policy provides a coarse\n level of granularity, controllable via the kernel command line:\n\n lockdown\u003d{integrity|confidentiality}\n\n Enable the kernel lockdown feature. If set to integrity, kernel features\n that allow userland to modify the running kernel are disabled. If set to\n confidentiality, kernel features that allow userland to extract\n confidential information from the kernel are also disabled.\n\n This may also be controlled via /sys/kernel/security/lockdown and\n overriden by kernel configuration.\n\n New or existing LSMs may implement finer-grained controls of the\n lockdown features. Refer to the lockdown_reason documentation in\n include/linux/security.h for details.\n\n The lockdown feature has had signficant design feedback and review\n across many subsystems. This code has been in linux-next for some\n weeks, with a few fixes applied along the way.\n\n Stephen Rothwell noted that commit 9d1f8be5cf42 (\"bpf: Restrict bpf\n when kernel lockdown is in confidentiality mode\") is missing a\n Signed-off-by from its author. Matthew responded that he is providing\n this under category (c) of the DCO\"\n\n* \u0027next-lockdown\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security: (31 commits)\n kexec: Fix file verification on S390\n security: constify some arrays in lockdown LSM\n lockdown: Print current-\u003ecomm in restriction messages\n efi: Restrict efivar_ssdt_load when the kernel is locked down\n tracefs: Restrict tracefs when the kernel is locked down\n debugfs: Restrict debugfs when the kernel is locked down\n kexec: Allow kexec_file() with appropriate IMA policy when locked down\n lockdown: Lock down perf when in confidentiality mode\n bpf: Restrict bpf when kernel lockdown is in confidentiality mode\n lockdown: Lock down tracing and perf kprobes when in confidentiality mode\n lockdown: Lock down /proc/kcore\n x86/mmiotrace: Lock down the testmmiotrace module\n lockdown: Lock down module params that specify hardware parameters (eg. ioport)\n lockdown: Lock down TIOCSSERIAL\n lockdown: Prohibit PCMCIA CIS storage when the kernel is locked down\n acpi: Disable ACPI table override if the kernel is locked down\n acpi: Ignore acpi_rsdp kernel param when the kernel has been locked down\n ACPI: Limit access to custom_method when the kernel is locked down\n x86/msr: Restrict MSR access when the kernel is locked down\n x86: Lock down IO port access when the kernel is locked down\n ...\n" }, { "commit": "2a78f9962565e53b78363eaf516eb052009e8020", "tree": "7d33ccd62917126b896ba4e54b25c697ecc7babd", "parents": [ "ab7b2577f0d119052b98b8d913bad369ac2760eb" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:23:00 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:44:13 2019 +0200" }, "message": "iommu/amd: Lock code paths traversing protection_domain-\u003edev_list\n\nThe traversing of this list requires protection_domain-\u003elock to be taken\nto avoid nasty races with attach/detach code. Make sure the lock is held\non all code-paths traversing this list.\n\nReported-by: Filippo Sironi \u003csironi@amazon.de\u003e\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "ab7b2577f0d119052b98b8d913bad369ac2760eb", "tree": "fb3c380fff0299866ebdc89fd52010d1bc4782f8", "parents": [ "45e528d9c479aeef2d3d1db1e619b243f91e324f" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:22:59 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:44:04 2019 +0200" }, "message": "iommu/amd: Lock dev_data in attach/detach code paths\n\nMake sure that attaching a detaching a device can\u0027t race against each\nother and protect the iommu_dev_data with a spin_lock in these code\npaths.\n\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "45e528d9c479aeef2d3d1db1e619b243f91e324f", "tree": "4e25b8515f6b8319e6951ea5c4ea87afc050f22b", "parents": [ "f6c0bfce271b2dd613e8b8e009eefe89c1f788e8" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:22:58 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:43:58 2019 +0200" }, "message": "iommu/amd: Check for busy devices earlier in attach_device()\n\nCheck early in attach_device whether the device is already attached to a\ndomain. This also simplifies the code path so that __attach_device() can\nbe removed.\n\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "f6c0bfce271b2dd613e8b8e009eefe89c1f788e8", "tree": "c01bbfce2692ac4d1f7f620a26ca98c52adea28c", "parents": [ "3a11905b69eb026402448c750f97a0eadfa76b08" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:22:57 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:43:52 2019 +0200" }, "message": "iommu/amd: Take domain-\u003elock for complete attach/detach path\n\nThe code-paths before __attach_device() and __detach_device() are called\nalso access and modify domain state, so take the domain lock there too.\nThis allows to get rid of the __detach_device() function.\n\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "3a11905b69eb026402448c750f97a0eadfa76b08", "tree": "64d30ce5111b1080f063adeef861bad127eb6ad3", "parents": [ "f15d9a992f901d4f22db868adf800844d1cac9f2" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:22:56 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:43:46 2019 +0200" }, "message": "iommu/amd: Remove amd_iommu_devtable_lock\n\nThe lock is not necessary because the device table does not\ncontain shared state that needs protection. Locking is only\nneeded on an individual entry basis, and that needs to\nhappen on the iommu_dev_data level.\n\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "f15d9a992f901d4f22db868adf800844d1cac9f2", "tree": "2c3bb88ca50ec64df785f1343708c1c2b243a19e", "parents": [ "0b15e02f0cc4fb34a9160de7ba6db3a4013dc1b7" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Sep 25 15:22:55 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Sat Sep 28 14:43:36 2019 +0200" }, "message": "iommu/amd: Remove domain-\u003eupdated\n\nThis struct member was used to track whether a domain\nchange requires updates to the device-table and IOMMU cache\nflushes. The problem is, that access to this field is racy\nsince locking in the common mapping code-paths has been\neliminated.\n\nMove the updated field to the stack to get rid of all\npotential races and remove the field from the struct.\n\nFixes: 92d420ec028d (\"iommu/amd: Relax locking in dma_ops path\")\nReviewed-by: Filippo Sironi \u003csironi@amazon.de\u003e\nReviewed-by: Jerry Snitselaar \u003cjsnitsel@redhat.com\u003e\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "f1f2f614d535564992f32e720739cb53cf03489f", "tree": "7c363b95df8d63954e822c98462fd3a9a98794ae", "parents": [ "298fb76a5583900a155d387efaf37a8b39e5dea2", "2a7f0e53daf29ca6dc9fbe2a27158f13474ec1b5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 19:37:27 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 19:37:27 2019 -0700" }, "message": "Merge branch \u0027next-integrity\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity\n\nPull integrity updates from Mimi Zohar:\n \"The major feature in this time is IMA support for measuring and\n appraising appended file signatures. In addition are a couple of bug\n fixes and code cleanup to use struct_size().\n\n In addition to the PE/COFF and IMA xattr signatures, the kexec kernel\n image may be signed with an appended signature, using the same\n scripts/sign-file tool that is used to sign kernel modules.\n\n Similarly, the initramfs may contain an appended signature.\n\n This contained a lot of refactoring of the existing appended signature\n verification code, so that IMA could retain the existing framework of\n calculating the file hash once, storing it in the IMA measurement list\n and extending the TPM, verifying the file\u0027s integrity based on a file\n hash or signature (eg. xattrs), and adding an audit record containing\n the file hash, all based on policy. (The IMA support for appended\n signatures patch set was posted and reviewed 11 times.)\n\n The support for appended signature paves the way for adding other\n signature verification methods, such as fs-verity, based on a single\n system-wide policy. The file hash used for verifying the signature and\n the signature, itself, can be included in the IMA measurement list\"\n\n* \u0027next-integrity\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity:\n ima: ima_api: Use struct_size() in kzalloc()\n ima: use struct_size() in kzalloc()\n sefltest/ima: support appended signatures (modsig)\n ima: Fix use after free in ima_read_modsig()\n MODSIGN: make new include file self contained\n ima: fix freeing ongoing ahash_request\n ima: always return negative code for error\n ima: Store the measurement again when appraising a modsig\n ima: Define ima-modsig template\n ima: Collect modsig\n ima: Implement support for module-style appended signatures\n ima: Factor xattr_verify() out of ima_appraise_measurement()\n ima: Add modsig appraise_type option for module-style appended signatures\n integrity: Select CONFIG_KEYS instead of depending on it\n PKCS#7: Introduce pkcs7_get_digest()\n PKCS#7: Refactor verify_pkcs7_signature()\n MODSIGN: Export module signature definitions\n ima: initialize the \"template\" field with the default template\n" }, { "commit": "298fb76a5583900a155d387efaf37a8b39e5dea2", "tree": "55b903ec587e8ec470c13084938303f542139557", "parents": [ "8f744bdee4fefb17fac052c7418b830de2b59ac8", "e41f9efb85d38d95744b9f35b9903109032b93d4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 17:00:27 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 17:00:27 2019 -0700" }, "message": "Merge tag \u0027nfsd-5.4\u0027 of git://linux-nfs.org/~bfields/linux\n\nPull nfsd updates from Bruce Fields:\n \"Highlights:\n\n - Add a new knfsd file cache, so that we don\u0027t have to open and close\n on each (NFSv2/v3) READ or WRITE. This can speed up read and write\n in some cases. It also replaces our readahead cache.\n\n - Prevent silent data loss on write errors, by treating write errors\n like server reboots for the purposes of write caching, thus forcing\n clients to resend their writes.\n\n - Tweak the code that allocates sessions to be more forgiving, so\n that NFSv4.1 mounts are less likely to hang when a server already\n has a lot of clients.\n\n - Eliminate an arbitrary limit on NFSv4 ACL sizes; they should now be\n limited only by the backend filesystem and the maximum RPC size.\n\n - Allow the server to enforce use of the correct kerberos credentials\n when a client reclaims state after a reboot.\n\n And some miscellaneous smaller bugfixes and cleanup\"\n\n* tag \u0027nfsd-5.4\u0027 of git://linux-nfs.org/~bfields/linux: (34 commits)\n sunrpc: clean up indentation issue\n nfsd: fix nfs read eof detection\n nfsd: Make nfsd_reset_boot_verifier_locked static\n nfsd: degraded slot-count more gracefully as allocation nears exhaustion.\n nfsd: handle drc over-allocation gracefully.\n nfsd: add support for upcall version 2\n nfsd: add a \"GetVersion\" upcall for nfsdcld\n nfsd: Reset the boot verifier on all write I/O errors\n nfsd: Don\u0027t garbage collect files that might contain write errors\n nfsd: Support the server resetting the boot verifier\n nfsd: nfsd_file cache entries should be per net namespace\n nfsd: eliminate an unnecessary acl size limit\n Deprecate nfsd fault injection\n nfsd: remove duplicated include from filecache.c\n nfsd: Fix the documentation for svcxdr_tmpalloc()\n nfsd: Fix up some unused variable warnings\n nfsd: close cached files prior to a REMOVE or RENAME that would replace target\n nfsd: rip out the raparms cache\n nfsd: have nfsd_test_lock use the nfsd_file cache\n nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache\n ...\n" }, { "commit": "8f744bdee4fefb17fac052c7418b830de2b59ac8", "tree": "aab0dfba240bf48ed67c32b60a9edb372c7afee9", "parents": [ "9977b1a71488742606376c09e19e0074e4403cdf", "a62a8ef9d97da23762a588592c8b8eb50a8deb6a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 15:54:24 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 15:54:24 2019 -0700" }, "message": "Merge tag \u0027virtio-fs-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse\n\nPull fuse virtio-fs support from Miklos Szeredi:\n \"Virtio-fs allows exporting directory trees on the host and mounting\n them in guest(s).\n\n This isn\u0027t actually a new filesystem, but a glue layer between the\n fuse filesystem and a virtio based back-end.\n\n It\u0027s similar in functionality to the existing virtio-9p solution, but\n significantly faster in benchmarks and has better POSIX compliance.\n Further permformance improvements can be achieved by sharing the page\n cache between host and guest, allowing for faster I/O and reduced\n memory use.\n\n Kata Containers have been including the out-of-tree virtio-fs (with\n the shared page cache patches as well) since version 1.7 as an\n experimental feature. They have been active in development and plan to\n switch from virtio-9p to virtio-fs as their default solution. There\n has been interest from other sources as well.\n\n The userspace infrastructure is slated to be merged into qemu once the\n kernel part hits mainline.\n\n This was developed by Vivek Goyal, Dave Gilbert and Stefan Hajnoczi\"\n\n* tag \u0027virtio-fs-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:\n virtio-fs: add virtiofs filesystem\n virtio-fs: add Documentation/filesystems/virtiofs.rst\n fuse: reserve values for mapping protocol\n" }, { "commit": "9977b1a71488742606376c09e19e0074e4403cdf", "tree": "08d35c71199ea2cd7adef98dd52ef917014d08e7", "parents": [ "568d850e3c6015acec8f854f5be97766497a676b", "aafee43b72863f1f70aeaf1332d049916e8df239" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 15:10:34 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 15:10:34 2019 -0700" }, "message": "Merge tag \u00279p-for-5.4\u0027 of git://github.com/martinetd/linux\n\nPull 9p updates from Dominique Martinet:\n \"Some of the usual small fixes and cleanup.\n\n Small fixes all around:\n - avoid overlayfs copy-up for PRIVATE mmaps\n - KUMSAN uninitialized warning for transport error\n - one syzbot memory leak fix in 9p cache\n - internal API cleanup for v9fs_fill_super\"\n\n* tag \u00279p-for-5.4\u0027 of git://github.com/martinetd/linux:\n 9p/vfs_super.c: Remove unused parameter data in v9fs_fill_super\n 9p/cache.c: Fix memory leak in v9fs_cache_session_get_cookie\n 9p: Transport error uninitialized\n 9p: avoid attaching writeback_fid on mmap with type PRIVATE\n" }, { "commit": "568d850e3c6015acec8f854f5be97766497a676b", "tree": "8749c8e8950c90f3da7535bf13e9cb000ebf4fa9", "parents": [ "70570a6418be2bd28be30fda700e57a81df7282b", "c82dd6d078a2bb29d41eda032bb96d05699a524d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 13:08:36 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 13:08:36 2019 -0700" }, "message": "Merge tag \u0027riscv/for-v5.4-rc1-b\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux\n\nPull more RISC-V updates from Paul Walmsley:\n \"Some additional RISC-V updates.\n\n This includes one significant fix:\n\n - Prevent interrupts from being unconditionally re-enabled during\n exception handling if they were disabled in the context in which\n the exception occurred\n\n Also a few other fixes:\n\n - Fix a build error when sparse memory support is manually enabled\n\n - Prevent CPUs beyond CONFIG_NR_CPUS from being enabled in early boot\n\n And a few minor improvements:\n\n - DT improvements: in the FU540 SoC DT files, improve U-Boot\n compatibility by adding an \"ethernet0\" alias, drop an unnecessary\n property from the DT files, and add support for the PWM device\n\n - KVM preparation: add a KVM-related macro for future RISC-V KVM\n support, and export some symbols required to build KVM support as\n modules\n\n - defconfig additions: build more drivers by default for QEMU\n configurations\"\n\n* tag \u0027riscv/for-v5.4-rc1-b\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:\n riscv: Avoid interrupts being erroneously enabled in handle_exception()\n riscv: dts: sifive: Drop \"clock-frequency\" property of cpu nodes\n riscv: dts: sifive: Add ethernet0 to the aliases node\n RISC-V: Export kernel symbols for kvm\n KVM: RISC-V: Add KVM_REG_RISCV for ONE_REG interface\n arch/riscv: disable excess harts before picking main boot hart\n RISC-V: Enable VIRTIO drivers in RV64 and RV32 defconfig\n RISC-V: Fix building error when CONFIG_SPARSEMEM_MANUAL\u003dy\n riscv: dts: Add DT support for SiFive FU540 PWM driver\n" }, { "commit": "70570a6418be2bd28be30fda700e57a81df7282b", "tree": "09b817a3c93a86aa8024e90d714010b4f6233ad5", "parents": [ "8bbe0dec38e147a50e9dd5f585295f7e68e0f2d0", "91d99a724e9c60e14332c26ab2284bf696b94c8e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 13:02:19 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 13:02:19 2019 -0700" }, "message": "Merge tag \u0027nios2-v5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/lftan/nios2\n\nPull nios2 fix from Ley Foon Tan:\n \"Make sure the command line buffer is NUL-terminated\"\n\n* tag \u0027nios2-v5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/lftan/nios2:\n nios2: force the string buffer NULL-terminated\n" }, { "commit": "8bbe0dec38e147a50e9dd5f585295f7e68e0f2d0", "tree": "0781fd7dd090bbccd88f1f58caf6387a585b9ebe", "parents": [ "e37e3bc7e265d05d00f14079767537699cf6bd46", "fd3edd4a9066f28de99a16685a586d68a9f551f8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:44:26 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:44:26 2019 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull more KVM updates from Paolo Bonzini:\n \"x86 KVM changes:\n\n - The usual accuracy improvements for nested virtualization\n\n - The usual round of code cleanups from Sean\n\n - Added back optimizations that were prematurely removed in 5.2 (the\n bare minimum needed to fix the regression was in 5.3-rc8, here\n comes the rest)\n\n - Support for UMWAIT/UMONITOR/TPAUSE\n\n - Direct L2-\u003eL0 TLB flushing when L0 is Hyper-V and L1 is KVM\n\n - Tell Windows guests if SMT is disabled on the host\n\n - More accurate detection of vmexit cost\n\n - Revert a pvqspinlock pessimization\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm: (56 commits)\n KVM: nVMX: cleanup and fix host 64-bit mode checks\n KVM: vmx: fix build warnings in hv_enable_direct_tlbflush() on i386\n KVM: x86: Don\u0027t check kvm_rebooting in __kvm_handle_fault_on_reboot()\n KVM: x86: Drop ____kvm_handle_fault_on_reboot()\n KVM: VMX: Add error handling to VMREAD helper\n KVM: VMX: Optimize VMX instruction error and fault handling\n KVM: x86: Check kvm_rebooting in kvm_spurious_fault()\n KVM: selftests: fix ucall on x86\n Revert \"locking/pvqspinlock: Don\u0027t wait if vCPU is preempted\"\n kvm: nvmx: limit atomic switch MSRs\n kvm: svm: Intercept RDPRU\n kvm: x86: Add \"significant index\" flag to a few CPUID leaves\n KVM: x86/mmu: Skip invalid pages during zapping iff root_count is zero\n KVM: x86/mmu: Explicitly track only a single invalid mmu generation\n KVM: x86/mmu: Revert \"KVM: x86/mmu: Remove is_obsolete() call\"\n KVM: x86/mmu: Revert \"Revert \"KVM: MMU: reclaim the zapped-obsolete page first\"\"\n KVM: x86/mmu: Revert \"Revert \"KVM: MMU: collapse TLB flushes when zap all pages\"\"\n KVM: x86/mmu: Revert \"Revert \"KVM: MMU: zap pages in batch\"\"\n KVM: x86/mmu: Revert \"Revert \"KVM: MMU: add tracepoint for kvm_mmu_invalidate_all_pages\"\"\n KVM: x86/mmu: Revert \"Revert \"KVM: MMU: show mmu_valid_gen in shadow page related tracepoints\"\"\n ...\n" }, { "commit": "e37e3bc7e265d05d00f14079767537699cf6bd46", "tree": "86387fbfaa379e067ef4d2fa53a284c30dfdf699", "parents": [ "738f531d877ac2b228b25354dfa4da6e79a2c369", "da635e7abe3f4ec9a8270ca4f5ba946d1a43f678" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:19:47 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:19:47 2019 -0700" }, "message": "Merge tag \u0027pwm/for-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm\n\nPull pwm updates from Thierry Reding:\n \"Besides one new driver being added for the PWM controller found in\n various Spreadtrum SoCs, this series of changes brings a slew of,\n mostly minor, fixes and cleanups for existing drivers, as well as some\n enhancements to the core code.\n\n Lastly, Uwe is added to the PWM subsystem entry of the MAINTAINERS\n file, making official his role as a reviewer\"\n\n* tag \u0027pwm/for-5.4-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm: (34 commits)\n MAINTAINERS: Add myself as reviewer for the PWM subsystem\n MAINTAINERS: Add patchwork link for PWM entry\n MAINTAINERS: Add a selection of PWM related keywords to the PWM entry\n pwm: mediatek: Add MT7629 compatible string\n dt-bindings: pwm: Update bindings for MT7629 SoC\n pwm: mediatek: Update license and switch to SPDX tag\n pwm: mediatek: Use pwm_mediatek as common prefix\n pwm: mediatek: Allocate the clks array dynamically\n pwm: mediatek: Remove the has_clks field\n pwm: mediatek: Drop the check for of_device_get_match_data()\n pwm: atmel: Consolidate driver data initialization\n pwm: atmel: Remove unneeded check for match data\n pwm: atmel: Remove platform_device_id and use only dt bindings\n pwm: stm32-lp: Add check in case requested period cannot be achieved\n pwm: Ensure pwm_apply_state() doesn\u0027t modify the state argument\n pwm: fsl-ftm: Don\u0027t update the state for the caller of pwm_apply_state()\n pwm: sun4i: Don\u0027t update the state for the caller of pwm_apply_state()\n pwm: rockchip: Don\u0027t update the state for the caller of pwm_apply_state()\n pwm: Let pwm_get_state() return the last implemented state\n pwm: Introduce local struct pwm_chip in pwm_apply_state()\n ...\n" }, { "commit": "738f531d877ac2b228b25354dfa4da6e79a2c369", "tree": "cea724f0bee623d6694a6cf434992727a8475864", "parents": [ "47db9b9a6eba4c5b0872220c8c8ff787a4b06ab0", "bda521624e75c665c407b3d9cece6e7a28178cd8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:08:24 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 12:08:24 2019 -0700" }, "message": "Merge tag \u0027for-5.4/io_uring-2019-09-27\u0027 of git://git.kernel.dk/linux-block\n\nPull more io_uring updates from Jens Axboe:\n \"Just two things in here:\n\n - Improvement to the io_uring CQ ring wakeup for batched IO (me)\n\n - Fix wrong comparison in poll handling (yangerkun)\n\n I realize the first one is a little late in the game, but it felt\n pointless to hold it off until the next release. Went through various\n testing and reviews with Pavel and peterz\"\n\n* tag \u0027for-5.4/io_uring-2019-09-27\u0027 of git://git.kernel.dk/linux-block:\n io_uring: make CQ ring wakeups be more efficient\n io_uring: compare cached_cq_tail with cq.head in_io_uring_poll\n" }, { "commit": "faeacb6ddb13b7a020b50b9246fe098653cfbd6e", "tree": "b931fd338ed0aa11db4ad875f355f88459928955", "parents": [ "78beef629fd95be4ed853b2d37b832f766bd96ca" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Fri Sep 27 10:40:39 2019 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:58:35 2019 +0200" }, "message": "net: tap: clean up an indentation issue\n\nThere is a statement that is indented too deeply, remove\nthe extraneous tab.\n\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "47db9b9a6eba4c5b0872220c8c8ff787a4b06ab0", "tree": "10f87c30b63fd9d2ce9d7ccc51e7667be8be05e7", "parents": [ "d0e00bc5ada53bda296ce8bfffc2f2be9eb22632", "8d6996630c03d7ceeabe2611378fea5ca1c3f1b3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:58:03 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:58:03 2019 -0700" }, "message": "Merge tag \u0027for-linus-2019-09-27\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"A few fixes/changes to round off this merge window. This contains:\n\n - Small series making some functional tweaks to blk-iocost (Tejun)\n\n - Elevator switch locking fix (Ming)\n\n - Kill redundant call in blk-wbt (Yufen)\n\n - Fix flush timeout handling (Yufen)\"\n\n* tag \u0027for-linus-2019-09-27\u0027 of git://git.kernel.dk/linux-block:\n block: fix null pointer dereference in blk_mq_rq_timed_out()\n rq-qos: get rid of redundant wbt_update_limits()\n iocost: bump up default latency targets for hard disks\n iocost: improve nr_lagging handling\n iocost: better trace vrate changes\n block: don\u0027t release queue\u0027s sysfs lock during switching elevator\n blk-mq: move lockdep_assert_held() into elevator_exit\n" }, { "commit": "78beef629fd95be4ed853b2d37b832f766bd96ca", "tree": "4f720286e599426e5e946e6203e8126d756fb547", "parents": [ "a41e8a88b06ee39fad4cb4a8ccf822563560a89c" ], "author": { "name": "Navid Emamdoost", "email": "navid.emamdoost@gmail.com", "time": "Thu Sep 26 20:51:46 2019 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:55:51 2019 +0200" }, "message": "nfp: abm: fix memory leak in nfp_abm_u32_knode_replace\n\nIn nfp_abm_u32_knode_replace if the allocation for match fails it should\ngo to the error handling instead of returning. Updated other gotos to\nhave correct errno returned, too.\n\nSigned-off-by: Navid Emamdoost \u003cnavid.emamdoost@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e51df6ce668a8f75ce27f83ce0f60103c568c375", "tree": "3e42543d66433f9383c50d32c3b7647a6c55838a", "parents": [ "b960bc448a252428bacca271f3416a8bda3b599b" ], "author": { "name": "Ben Chuang", "email": "ben.chuang@genesyslogic.com.tw", "time": "Wed Sep 11 15:23:44 2019 +0800" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Fri Sep 27 20:48:20 2019 +0200" }, "message": "mmc: host: sdhci-pci: Add Genesys Logic GL975x support\n\nAdd support for the GL9750 and GL9755 chipsets.\n\nEnable v4 mode and wait 5ms after set 1.8V signal enable for GL9750/\nGL9755. Fix the value of SDHCI_MAX_CURRENT register and use the vendor\ntuning flow for GL9750.\n\nCo-developed-by: Michael K Johnson \u003cjohnsonm@danlj.org\u003e\nSigned-off-by: Michael K Johnson \u003cjohnsonm@danlj.org\u003e\nSigned-off-by: Ben Chuang \u003cben.chuang@genesyslogic.com.tw\u003e\nAcked-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "a41e8a88b06ee39fad4cb4a8ccf822563560a89c", "tree": "4cb7c743f2425c421a5604ae6232843a4055c949", "parents": [ "174e23810cd3183dc2ca3f5166ef965a55eaaf54" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Thu Sep 26 15:42:51 2019 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:42:24 2019 +0200" }, "message": "tcp: better handle TCP_USER_TIMEOUT in SYN_SENT state\n\nYuchung Cheng and Marek Majkowski independently reported a weird\nbehavior of TCP_USER_TIMEOUT option when used at connect() time.\n\nWhen the TCP_USER_TIMEOUT is reached, tcp_write_timeout()\nbelieves the flow should live, and the following condition\nin tcp_clamp_rto_to_user_timeout() programs one jiffie timers :\n\n remaining \u003d icsk-\u003eicsk_user_timeout - elapsed;\n if (remaining \u003c\u003d 0)\n return 1; /* user timeout has passed; fire ASAP */\n\nThis silly situation ends when the max syn rtx count is reached.\n\nThis patch makes sure we honor both TCP_SYNCNT and TCP_USER_TIMEOUT,\navoiding these spurious SYN packets.\n\nFixes: b701a99e431d (\"tcp: Add tcp_clamp_rto_to_user_timeout() helper to improve accuracy\")\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nReported-by: Yuchung Cheng \u003cycheng@google.com\u003e\nReported-by: Marek Majkowski \u003cmarek@cloudflare.com\u003e\nCc: Jon Maxwell \u003cjmaxwell37@gmail.com\u003e\nLink: https://marc.info/?l\u003dlinux-netdev\u0026m\u003d156940118307949\u0026w\u003d2\nAcked-by: Jon Maxwell \u003cjmaxwell37@gmail.com\u003e\nTested-by: Marek Majkowski \u003cmarek@cloudflare.com\u003e\nSigned-off-by: Marek Majkowski \u003cmarek@cloudflare.com\u003e\nAcked-by: Yuchung Cheng \u003cycheng@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "174e23810cd3183dc2ca3f5166ef965a55eaaf54", "tree": "a9ffbf41f4c75ccef3332330253b37e21a84104f", "parents": [ "6b3656a60f2067738d1a423328199720806f0c44" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Thu Sep 26 20:37:05 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:40:19 2019 +0200" }, "message": "sk_buff: drop all skb extensions on free and skb scrubbing\n\nNow that we have a 3rd extension, add a new helper that drops the\nextension space and use it when we need to scrub an sk_buff.\n\nAt this time, scrubbing clears secpath and bridge netfilter data, but\nretains the tc skb extension, after this patch all three get cleared.\n\nNAPI reuse/free assumes we can only have a secpath attached to skb, but\nit seems better to clear all extensions there as well.\n\nv2: add unlikely hint (Eric Dumazet)\n\nFixes: 95a7233c452a (\"net: openvswitch: Set OvS recirc_id from tc chain index\")\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6b3656a60f2067738d1a423328199720806f0c44", "tree": "c4f1c8c356afb67f007c8fccf97e4d159d804bd2", "parents": [ "94e7e5da388de3b25d6066748323438bc9b37625" ], "author": { "name": "Kevin(Yudong) Yang", "email": "yyd@google.com", "time": "Thu Sep 26 10:30:05 2019 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:37:50 2019 +0200" }, "message": "tcp_bbr: fix quantization code to not raise cwnd if not probing bandwidth\n\nThere was a bug in the previous logic that attempted to ensure gain cycling\ngets inflight above BDP even for small BDPs. This code correctly raised and\nlowered target inflight values during the gain cycle. And this code\ncorrectly ensured that cwnd was raised when probing bandwidth. However, it\ndid not correspondingly ensure that cwnd was *not* raised in this way when\n*not* probing for bandwidth. The result was that small-BDP flows that were\nalways cwnd-bound could go for many cycles with a fixed cwnd, and not probe\nor yield bandwidth at all. This meant that multiple small-BDP flows could\nfail to converge in their bandwidth allocations.\n\nFixes: 3c346b233c68 (\"tcp_bbr: fix bw probing to raise in-flight data for very small BDPs\")\nSigned-off-by: Kevin(Yudong) Yang \u003cyyd@google.com\u003e\nAcked-by: Neal Cardwell \u003cncardwell@google.com\u003e\nAcked-by: Yuchung Cheng \u003cycheng@google.com\u003e\nAcked-by: Soheil Hassas Yeganeh \u003csoheil@google.com\u003e\nAcked-by: Priyaranjan Jha \u003cpriyarjha@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d0e00bc5ada53bda296ce8bfffc2f2be9eb22632", "tree": "7b00f11f0406f21ae1ebf5f376a963f439fddeef", "parents": [ "7bccb9f10c8f36ee791769b531ed4d28f6379aae", "0f84d1d18c46d0f995962c876c8b2900fd183fd7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:35:13 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:35:13 2019 -0700" }, "message": "Merge branch \u0027for-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux\n\nPull thermal management updates from Zhang Rui:\n\n - Add Amit Kucheria as thermal subsystem Reviewer (Amit Kucheria)\n\n - Fix a use after free bug when unregistering thermal zone devices (Ido\n Schimmel)\n\n - Fix thermal core framework to use put_device() when device_register()\n fails (Yue Hu)\n\n - Enable intel_pch_thermal and MMIO RAPL support for Intel Icelake\n platform (Srinivas Pandruvada)\n\n - Add clock operations in qorip thermal driver, for some platforms with\n clock control like i.MX8MQ (Anson Huang)\n\n - A couple of trivial fixes and cleanups for thermal core and different\n soc thermal drivers (Amit Kucheria, Christophe JAILLET, Chuhong Yuan,\n Fuqian Huang, Kelsey Skunberg, Nathan Huckleberry, Rishi Gupta,\n Srinivas Kandagatla)\n\n* \u0027for-5.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux:\n MAINTAINERS: Add Amit Kucheria as reviewer for thermal\n thermal: Add some error messages\n thermal: Fix use-after-free when unregistering thermal zone device\n thermal/drivers/core: Use put_device() if device_register() fails\n thermal_hwmon: Sanitize thermal_zone type\n thermal: intel: Use dev_get_drvdata\n thermal: intel: int3403: replace printk(KERN_WARN...) with pr_warn(...)\n thermal: intel: int340x_thermal: Remove unnecessary acpi_has_method() uses\n thermal: int340x: processor_thermal: Add Ice Lake support\n drivers: thermal: qcom: tsens: Fix memory leak from qfprom read\n thermal: tegra: Fix a typo\n thermal: rcar_gen3_thermal: Replace devm_add_action() followed by failure action with devm_add_action_or_reset()\n thermal: armada: Fix -Wshift-negative-value\n dt-bindings: thermal: qoriq: Add optional clocks property\n thermal: qoriq: Use __maybe_unused instead of #if CONFIG_PM_SLEEP\n thermal: qoriq: Use devm_platform_ioremap_resource() instead of of_iomap()\n thermal: qoriq: Fix error path of calling qoriq_tmu_register_tmu_zone fail\n thermal: qoriq: Add clock operations\n drivers: thermal: processor_thermal_device: Export sysfs interface for TCC offset\n" }, { "commit": "94e7e5da388de3b25d6066748323438bc9b37625", "tree": "659187395bef3c1b7b049e533e426d0a34ee6597", "parents": [ "4208966f65f520d7f392dbaa62e39a8fa88ffb95", "52feb8b588f6d23673dd7cc2b44b203493b627f6" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:33:19 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:33:19 2019 +0200" }, "message": "Merge branch \u0027mlxsw-Various-fixes\u0027\n\nIdo Schimmel says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlxsw: Various fixes\n\nThis patchset includes two small fixes for the mlxsw driver and one\npatch which clarifies recently introduced devlink-trap documentation.\n\nPatch #1 clears the port\u0027s VLAN filters during port initialization. This\nensures that the drop reason reported to the user is consistent. The\nproblem is explained in detail in the commit message.\n\nPatch #2 clarifies the description of one of the traps exposed via\ndevlink-trap.\n\nPatch #3 from Danielle forbids the installation of a tc filter with\nmultiple mirror actions since this is not supported by the device. The\nfailure is communicated to the user via extack.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "52feb8b588f6d23673dd7cc2b44b203493b627f6", "tree": "659187395bef3c1b7b049e533e426d0a34ee6597", "parents": [ "44bde514eb13ac32b20442880e8175584af7592c" ], "author": { "name": "Danielle Ratson", "email": "danieller@mellanox.com", "time": "Thu Sep 26 14:43:40 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:33:19 2019 +0200" }, "message": "mlxsw: spectrum_flower: Fail in case user specifies multiple mirror actions\n\nThe ASIC can only mirror a packet to one port, but when user is trying\nto set more than one mirror action, it doesn\u0027t fail.\n\nAdd a check if more than one mirror action was specified per rule and if so,\nfail for not being supported.\n\nFixes: d0d13c1858a11 (\"mlxsw: spectrum_acl: Add support for mirror action\")\nSigned-off-by: Danielle Ratson \u003cdanieller@mellanox.com\u003e\nAcked-by: Jiri Pirko \u003cjiri@mellanox.com\u003e\nSigned-off-by: Ido Schimmel \u003cidosch@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "44bde514eb13ac32b20442880e8175584af7592c", "tree": "26369a06ceb5c5073992196deca11e399c746e0d", "parents": [ "979b9b251ae06e3408153bd7b9342a290d65e826" ], "author": { "name": "Ido Schimmel", "email": "idosch@mellanox.com", "time": "Thu Sep 26 14:43:39 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:33:19 2019 +0200" }, "message": "Documentation: Clarify trap\u0027s description\n\nAlex noted that the below description might not be obvious to all users.\nClarify it by adding an example.\n\nFixes: f3047ca01f12 (\"Documentation: Add devlink-trap documentation\")\nReported-by: Alex Kushnarov \u003calexanderk@mellanox.com\u003e\nReviewed-by: Alex Kushnarov \u003calexanderk@mellanox.com\u003e\nAcked-by: Jiri Pirko \u003cjiri@mellanox.com\u003e\nSigned-off-by: Ido Schimmel \u003cidosch@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "979b9b251ae06e3408153bd7b9342a290d65e826", "tree": "fbc6206678f3a9c8a8316ff093fd1715201e452d", "parents": [ "4208966f65f520d7f392dbaa62e39a8fa88ffb95" ], "author": { "name": "Ido Schimmel", "email": "idosch@mellanox.com", "time": "Thu Sep 26 14:43:38 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:33:19 2019 +0200" }, "message": "mlxsw: spectrum: Clear VLAN filters during port initialization\n\nWhen a port is created, its VLAN filters are not cleared by the\nfirmware. This causes tagged packets to be later dropped by the ingress\nSTP filters, which default to DISCARD state.\n\nThe above did not matter much until commit b5ce611fd96e (\"mlxsw:\nspectrum: Add devlink-trap support\") where we exposed the drop reason to\nusers.\n\nWithout this patch, the drop reason users will see is not consistent. If\na port is enslaved to a VLAN-aware bridge and a packet with an invalid\nVLAN tries to ingress the bridge, it will be dropped due to ingress STP\nfilter. If the VLAN is later enabled and then disabled, the packet will\nbe dropped by the ingress VLAN filter despite the above being a\nseemingly NOP operation.\n\nFix this by clearing all the VLAN filters during port initialization.\nAdjust the test accordingly.\n\nFixes: b5ce611fd96e (\"mlxsw: spectrum: Add devlink-trap support\")\nReported-by: Alex Kushnarov \u003calexanderk@mellanox.com\u003e\nTested-by: Alex Kushnarov \u003calexanderk@mellanox.com\u003e\nAcked-by: Jiri Pirko \u003cjiri@mellanox.com\u003e\nSigned-off-by: Ido Schimmel \u003cidosch@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4208966f65f520d7f392dbaa62e39a8fa88ffb95", "tree": "d87a098a02501ef6bea7251af3abd1186d88bf60", "parents": [ "6ba5bbba95f789d76ce3bf440ee02fdaf52ec486" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Thu Sep 26 12:22:52 2019 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:32:02 2019 +0200" }, "message": "net: ena: clean up indentation issue\n\nThere memset is indented incorrectly, remove the extraneous tabs.\n\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6ba5bbba95f789d76ce3bf440ee02fdaf52ec486", "tree": "87a59013517e4f3094bc2ad72a549b34b88cd29b", "parents": [ "407d8098cb1ab338199f4753162799a488d87d23" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Thu Sep 26 12:13:06 2019 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:31:18 2019 +0200" }, "message": "NFC: st95hf: clean up indentation issue\n\nThe return statement is indented incorrectly, add in a missing\ntab and remove an extraneous space after the return\n\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b960bc448a252428bacca271f3416a8bda3b599b", "tree": "a0fa007a4abf5d671dbfb2a2473ac04bf7504c8a", "parents": [ "4ee7dde4c777f14cb0f98dd201491bf6cc15899b" ], "author": { "name": "Nicolin Chen", "email": "nicoleotsuka@gmail.com", "time": "Mon Sep 23 12:08:10 2019 +0200" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Fri Sep 27 20:30:13 2019 +0200" }, "message": "mmc: tegra: Implement -\u003eset_dma_mask()\n\nThe SDHCI controller on Tegra186 supports 40-bit addressing, which is\nusually enough to address all of system memory. However, if the SDHCI\ncontroller is behind an IOMMU, the address space can go beyond. This\nhappens on Tegra186 and later where the ARM SMMU has an input address\nspace of 48 bits. If the DMA API is backed by this ARM SMMU, the top-\ndown IOVA allocator will cause IOV addresses to be returned that the\nSDHCI controller cannot access.\n\nUnfortunately, prior to the introduction of the -\u003eset_dma_mask() host\noperation, the SDHCI core would set either a 64-bit DMA mask if the\ncontroller claimed to support 64-bit addressing, or a 32-bit DMA mask\notherwise.\n\nSince the full 64 bits cannot be addressed on Tegra, this had to be\nworked around in commit 68481a7e1c84 (\"mmc: tegra: Mark 64 bit dma\nbroken on Tegra186\") by setting the SDHCI_QUIRK2_BROKEN_64_BIT_DMA\nquirk, which effectively restricts the DMA mask to 32 bits.\n\nOne disadvantage of this is that dma_map_*() APIs will now try to use\nthe swiotlb to bounce DMA to addresses beyond of the controller\u0027s DMA\nmask. This in turn caused degraded performance and can lead to\nsituations where the swiotlb buffer is exhausted, which in turn leads\nto DMA transfers to fail.\n\nWith the recent introduction of the -\u003eset_dma_mask() host operation,\nthis can now be properly fixed. For each generation of Tegra, the exact\nsupported DMA mask can be configured. This kills two birds with one\nstone: it avoids the use of bounce buffers because system memory never\nexceeds the addressable memory range of the SDHCI controllers on these\ndevices, and at the same time when an IOMMU is involved, it prevents\nIOV addresses from being allocated beyond the addressible range of the\ncontrollers.\n\nSince the DMA mask is now properly handled, the 64-bit DMA quirk can be\nremoved.\n\nSigned-off-by: Nicolin Chen \u003cnicoleotsuka@gmail.com\u003e\n[treding@nvidia.com: provide more background in commit message]\nTested-by: Nicolin Chen \u003cnicoleotsuka@gmail.com\u003e\nAcked-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\nCc: stable@vger.kernel.org # v4.15 +\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "4ee7dde4c777f14cb0f98dd201491bf6cc15899b", "tree": "8c06f6f8d918b5c58362470db9164f1250cbef6c", "parents": [ "121bd08b029e03404c451bb237729cdff76eafed" ], "author": { "name": "Adrian Hunter", "email": "adrian.hunter@intel.com", "time": "Mon Sep 23 12:08:09 2019 +0200" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Fri Sep 27 20:30:13 2019 +0200" }, "message": "mmc: sdhci: Let drivers define their DMA mask\n\nAdd host operation -\u003eset_dma_mask() so that drivers can define their own\nDMA masks.\n\nSigned-off-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nTested-by: Nicolin Chen \u003cnicoleotsuka@gmail.com\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\nCc: stable@vger.kernel.org # v4.15 +\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "121bd08b029e03404c451bb237729cdff76eafed", "tree": "5e09c72d876a495cd91d40d75cc6d630217c84c5", "parents": [ "d1c536e3177390da43d99f20143b810c35433d1f" ], "author": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Sun Sep 22 11:26:58 2019 +0100" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Fri Sep 27 20:30:13 2019 +0200" }, "message": "mmc: sdhci-of-esdhc: set DMA snooping based on DMA coherence\n\nWe must not unconditionally set the DMA snoop bit; if the DMA API is\nassuming that the device is not DMA coherent, and the device snoops the\nCPU caches, the device can see stale cache lines brought in by\nspeculative prefetch.\n\nThis leads to the device seeing stale data, potentially resulting in\ncorrupted data transfers. Commonly, this results in a descriptor fetch\nerror such as:\n\nmmc0: ADMA error\nmmc0: sdhci: \u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d SDHCI REGISTER DUMP \u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmmc0: sdhci: Sys addr: 0x00000000 | Version: 0x00002202\nmmc0: sdhci: Blk size: 0x00000008 | Blk cnt: 0x00000001\nmmc0: sdhci: Argument: 0x00000000 | Trn mode: 0x00000013\nmmc0: sdhci: Present: 0x01f50008 | Host ctl: 0x00000038\nmmc0: sdhci: Power: 0x00000003 | Blk gap: 0x00000000\nmmc0: sdhci: Wake-up: 0x00000000 | Clock: 0x000040d8\nmmc0: sdhci: Timeout: 0x00000003 | Int stat: 0x00000001\nmmc0: sdhci: Int enab: 0x037f108f | Sig enab: 0x037f108b\nmmc0: sdhci: ACmd stat: 0x00000000 | Slot int: 0x00002202\nmmc0: sdhci: Caps: 0x35fa0000 | Caps_1: 0x0000af00\nmmc0: sdhci: Cmd: 0x0000333a | Max curr: 0x00000000\nmmc0: sdhci: Resp[0]: 0x00000920 | Resp[1]: 0x001d8a33\nmmc0: sdhci: Resp[2]: 0x325b5900 | Resp[3]: 0x3f400e00\nmmc0: sdhci: Host ctl2: 0x00000000\nmmc0: sdhci: ADMA Err: 0x00000009 | ADMA Ptr: 0x000000236d43820c\nmmc0: sdhci: \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\nmmc0: error -5 whilst initialising SD card\n\nbut can lead to other errors, and potentially direct the SDHCI\ncontroller to read/write data to other memory locations (e.g. if a valid\ndescriptor is visible to the device in a stale cache line.)\n\nFix this by ensuring that the DMA snoop bit corresponds with the\nbehaviour of the DMA API. Since the driver currently only supports DT,\nuse of_dma_is_coherent(). Note that device_get_dma_attr() can not be\nused as that risks re-introducing this bug if/when the driver is\nconverted to ACPI.\n\nSigned-off-by: Russell King \u003crmk+kernel@armlinux.org.uk\u003e\nAcked-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "d1c536e3177390da43d99f20143b810c35433d1f", "tree": "bb9f00e274d2a1c74cb86badd55935f1947d9dd4", "parents": [ "3c6a6910a81eae3566bb5fef6ea0f624382595e6" ], "author": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Sun Sep 22 11:26:53 2019 +0100" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Fri Sep 27 20:30:13 2019 +0200" }, "message": "mmc: sdhci: improve ADMA error reporting\n\nADMA errors are potentially data corrupting events; although we print\nthe register state, we do not usefully print the ADMA descriptors.\nWorse than that, we print them by referencing their virtual address\nwhich is meaningless when the register state gives us the DMA address\nof the failing descriptor.\n\nPrint the ADMA descriptors giving their DMA addresses rather than their\nvirtual addresses, and print them using SDHCI_DUMP() rather than DBG().\n\nWe also do not show the correct value of the interrupt status register;\nthe register dump shows the current value, after we have cleared the\npending interrupts we are going to service. What is more useful is to\nprint the interrupts that _were_ pending at the time the ADMA error was\nencountered. Fix that too.\n\nSigned-off-by: Russell King \u003crmk+kernel@armlinux.org.uk\u003e\nAcked-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "407d8098cb1ab338199f4753162799a488d87d23", "tree": "fdae4189240b8df288ec16d52c21d7dcc36d35c2", "parents": [ "fd4a8093ec0bd37d450587d50f3e10f0af57cc47" ], "author": { "name": "Hans Andersson", "email": "hans.andersson@cellavision.se", "time": "Thu Sep 26 09:54:37 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:27:26 2019 +0200" }, "message": "net: phy: micrel: add Asym Pause workaround for KSZ9021\n\nThe Micrel KSZ9031 PHY may fail to establish a link when the Asymmetric\nPause capability is set. This issue is described in a Silicon Errata\n(DS80000691D or DS80000692D), which advises to always disable the\ncapability.\n\nMicrel KSZ9021 has no errata, but has the same issue with Asymmetric Pause.\nThis patch apply the same workaround as the one for KSZ9031.\n\nFixes: 3aed3e2a143c (\"net: phy: micrel: add Asym Pause workaround\")\nSigned-off-by: Hans Andersson \u003chans.andersson@cellavision.se\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "fd4a8093ec0bd37d450587d50f3e10f0af57cc47", "tree": "fa16e3eaffc4e470ce696eda6225e0e274bb64ce", "parents": [ "2df4de1681767df900e15e34195bbf7dc1b23e06" ], "author": { "name": "Kunihiko Hayashi", "email": "hayashi.kunihiko@socionext.com", "time": "Thu Sep 26 15:35:10 2019 +0900" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:26:28 2019 +0200" }, "message": "net: socionext: ave: Avoid using netdev_err() before calling register_netdev()\n\nUntil calling register_netdev(), ndev-\u003edev_name isn\u0027t specified, and\nnetdev_err() displays \"(unnamed net_device)\".\n\n ave 65000000.ethernet (unnamed net_device) (uninitialized): invalid phy-mode setting\n ave: probe of 65000000.ethernet failed with error -22\n\nThis replaces netdev_err() with dev_err() before calling register_netdev().\n\nSigned-off-by: Kunihiko Hayashi \u003chayashi.kunihiko@socionext.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2df4de1681767df900e15e34195bbf7dc1b23e06", "tree": "1ded2106e3e0ff1cf96b4ec30b43b17a8516ea9f", "parents": [ "991ad2b24da2fa7b4ba9775ca1fed2d660d54ab0" ], "author": { "name": "Jacob Keller", "email": "jacob.e.keller@intel.com", "time": "Wed Sep 25 19:28:19 2019 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:25:32 2019 +0200" }, "message": "ptp: correctly disable flags on old ioctls\n\nCommit 415606588c61 (\"PTP: introduce new versions of IOCTLs\",\n2019-09-13) introduced new versions of the PTP ioctls which actually\nvalidate that the flags are acceptable values.\n\nAs part of this, it cleared the flags value using a bitwise\nand+negation, in an attempt to prevent the old ioctl from accidentally\nenabling new features.\n\nThis is incorrect for a couple of reasons. First, it results in\naccidentally preventing previously working flags on the request ioctl.\nBy clearing the \"valid\" flags, we now no longer allow setting the\nenable, rising edge, or falling edge flags.\n\nSecond, if we add new additional flags in the future, they must not be\nset by the old ioctl. (Since the flag wasn\u0027t checked before, we could\npotentially break userspace programs which sent garbage flag data.\n\nThe correct way to resolve this is to check for and clear all but the\noriginally valid flags.\n\nCreate defines indicating which flags are correctly checked and\ninterpreted by the original ioctls. Use these to clear any bits which\nwill not be correctly interpreted by the original ioctls.\n\nIn the future, new flags must be added to the VALID_FLAGS macros, but\n*not* to the V1_VALID_FLAGS macros. In this way, new features may be\nexposed over the v2 ioctls, but without breaking previous userspace\nwhich happened to not clear the flags value properly. The old ioctl will\ncontinue to behave the same way, while the new ioctl gains the benefit\nof using the flags fields.\n\nCc: Richard Cochran \u003crichardcochran@gmail.com\u003e\nCc: Felipe Balbi \u003cfelipe.balbi@linux.intel.com\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nCc: Christopher Hall \u003cchristopher.s.hall@intel.com\u003e\nSigned-off-by: Jacob Keller \u003cjacob.e.keller@intel.com\u003e\nAcked-by: Richard Cochran \u003crichardcochran@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "991ad2b24da2fa7b4ba9775ca1fed2d660d54ab0", "tree": "e0d7a1ab5c1e0ecc8d51517774d0846274b122cc", "parents": [ "a3aa6e65beebf3780026753ebf39db19f4c92990" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Wed Sep 25 17:20:42 2019 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:23:37 2019 +0200" }, "message": "lib: dimlib: fix help text typos\n\nFix help text typos for DIMLIB.\n\nFixes: 4f75da3666c0 (\"linux/dim: Move implementation to .c files\")\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nCc: Uwe Kleine-König \u003cuwe@kleine-koenig.org\u003e\nCc: Tal Gilboa \u003ctalgi@mellanox.com\u003e\nCc: Saeed Mahameed \u003csaeedm@mellanox.com\u003e\nAcked-by: Uwe Kleine-König \u003cuwe@kleine-koenig.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a3aa6e65beebf3780026753ebf39db19f4c92990", "tree": "d919147ab33e385154001904fc0dfe4d156c65a3", "parents": [ "c5f095baa880036702595a09de38e7a0ca420c0b" ], "author": { "name": "Marek Vasut", "email": "marex@denx.de", "time": "Thu Sep 26 00:08:42 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:21:07 2019 +0200" }, "message": "net: dsa: microchip: Always set regmap stride to 1\n\nThe regmap stride is set to 1 for regmap describing 8bit registers already.\nHowever, for 16/32/64bit registers, the stride is 2/4/8 respectively. This\nis not correct, as the switch protocol supports unaligned register reads\nand writes and the KSZ87xx even uses such unaligned register accesses to\nread e.g. MIB counter.\n\nThis patch fixes MIB counter access on KSZ87xx.\n\nSigned-off-by: Marek Vasut \u003cmarex@denx.de\u003e\nCc: Andrew Lunn \u003candrew@lunn.ch\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nCc: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nCc: George McCollister \u003cgeorge.mccollister@gmail.com\u003e\nCc: Tristram Ha \u003cTristram.Ha@microchip.com\u003e\nCc: Vivien Didelot \u003cvivien.didelot@savoirfairelinux.com\u003e\nCc: Woojung Huh \u003cwoojung.huh@microchip.com\u003e\nFixes: 46558d601cb6 (\"net: dsa: microchip: Initial SPI regmap support\")\nFixes: 255b59ad0db2 (\"net: dsa: microchip: Factor out regmap config generation into common header\")\nReviewed-by: George McCollister \u003cgeorge.mccollister@gmail.com\u003e\nTested-by: George McCollister \u003cgeorge.mccollister@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7bccb9f10c8f36ee791769b531ed4d28f6379aae", "tree": "71984804c7c76fcb12c5209009facb27d5e9d431", "parents": [ "289991ce1cac18e7cd489902986ef986baa49568", "ca2fc5efffde5a3827adfb0ab6a51b6f1c64d5ff" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:17:38 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:17:38 2019 -0700" }, "message": "Merge tag \u0027linux-watchdog-5.4-rc1\u0027 of git://www.linux-watchdog.org/linux-watchdog\n\nPull watchdog updates from Wim Van Sebroeck:\n\n - addition of AST2600, i.MX7ULP and F81803 watchdog support\n\n - removal of the w90x900 and ks8695 drivers\n\n - ziirave_wdt improvements\n\n - small fixes and improvements\n\n* tag \u0027linux-watchdog-5.4-rc1\u0027 of git://www.linux-watchdog.org/linux-watchdog: (51 commits)\n watchdog: f71808e_wdt: Add F81803 support\n watchdog: qcom: remove unnecessary variable from private storage\n watchdog: qcom: support pre-timeout when the bark irq is available\n watchdog: imx_sc: this patch just fixes whitespaces\n watchdog: apseed: Add access_cs0 option for alt-boot\n watchdog: aspeed: add support for dual boot\n watchdog: orion_wdt: use timer1 as a pretimeout\n watchdog: Add i.MX7ULP watchdog support\n dt-bindings: watchdog: Add i.MX7ULP bindings\n dt-bindings: watchdog: sun4i: Add the watchdog clock\n dt-bindings: watchdog: sun4i: Add the watchdog interrupts\n dt-bindings: watchdog: Convert Allwinner watchdog to a schema\n dt-bindings: watchdog: Add YAML schemas for the generic watchdog bindings\n watchdog: aspeed: Add support for AST2600\n dt-bindings: watchdog: Add ast2600 compatible\n watchdog: ziirave_wdt: Update checked I2C functionality mask\n watchdog: ziirave_wdt: Drop ziirave_firm_write_block_data()\n watchdog: ziirave_wdt: Fix DOWNLOAD_START payload\n watchdog: ziirave_wdt: Drop status polling code\n watchdog: ziirave_wdt: Fix RESET_PROCESSOR payload\n ...\n" }, { "commit": "c5f095baa880036702595a09de38e7a0ca420c0b", "tree": "01d704e7f7f6972db774075ee808f7f5a8d03a0d", "parents": [ "8ce39eb5a67aee25d9f05b40b673c95b23502e3e", "9b05b6e11d5e93a3a517cadc12b9836e0470c255" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:15:00 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:15:00 2019 +0200" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf\n\nPablo Neira Ayuso says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nNetfilter fixes for net\n\nThe following patchset contains Netfilter fixes for net:\n\n1) Add NFT_CHAIN_POLICY_UNSET to replace hardcoded -1 to\n specify that the chain policy is unset. The chain policy\n field is actually defined as an 8-bit unsigned integer.\n\n2) Remove always true condition reported by smatch in\n chain policy check.\n\n3) Fix element lookup on dynamic sets, from Florian Westphal.\n\n4) Use __u8 in ebtables uapi header, from Masahiro Yamada.\n\n5) Bogus EBUSY when removing flowtable after chain flush,\n from Laura Garcia Liebana.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "289991ce1cac18e7cd489902986ef986baa49568", "tree": "b04c83b7d931c91fd71d68d9e5c6d572b324a320", "parents": [ "0cd81d77d0569f1dc1e39abeea93c6184e9b5b54", "3e2cb6d89325dc36a03937a2b82fc7eb902c96b0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:13:35 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:13:35 2019 -0700" }, "message": "Merge tag \u0027drm-next-2019-09-27\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"Fixes built up over the past 1.5 weeks or so, it\u0027s two weeks of\n amdgpu, some core cleanups and some panfrost fixes. I also finally\n figured out why my desktop was slow to do a bunch of stuff (someone\n gave it an IPv6 address which can\u0027t reach anything!).\n\n core:\n - Some cleanups and fixes in the self-refresh helpers\n - Some cleanups and fixes in the atomic helpers\n\n amdgpu:\n - Fix a 64 bit divide\n - Prevent a memory leak in a failure case in dc\n - Load proper gfx firmware on navi14 variants\n - Add more navi12 and navi14 PCI ids\n - Misc fixes for renoir\n - Fix bandwidth issues with multiple displays on vega20\n - Support for Dali\n - Fix a possible oops with KFD on hawaii\n - Fix for backlight level after resume on some APUs\n - Other misc fixes\n\n panfrost:\n - Multiple panfrost fixes for regulator support and page fault\n handling\"\n\n* tag \u0027drm-next-2019-09-27\u0027 of git://anongit.freedesktop.org/drm/drm: (34 commits)\n drm/amd/display: prevent memory leak\n drm/amdgpu/gfx10: add support for wks firmware loading\n drm/amdgpu/display: include slab.h in dcn21_resource.c\n drm/amdgpu/display: fix 64 bit divide\n drm/panfrost: Prevent race when handling page fault\n drm/panfrost: Remove NULL checks for regulator\n drm/panfrost: Fix regulator_get_optional() misuse\n drm: Measure Self Refresh Entry/Exit times to avoid thrashing\n drm: Fix kerneldoc and remove unused struct member in self_refresh helper\n drm/atomic: Rename crtc_state-\u003epageflip_flags to async_flip\n drm/atomic: Reject FLIP_ASYNC unconditionally\n drm/atomic: Take the atomic toys away from X\n drm/amdgpu: flag navi12 and 14 as experimental for 5.4\n drm/kms: Duct-tape for mode object lifetime checks\n drm/amdgpu: add navi12 pci id\n drm/amdgpu: add navi14 PCI ID for work station SKU\n drm/amdkfd: Swap trap temporary registers in gfx10 trap handler\n drm/amd/powerplay: implement sysfs for getting dpm clock\n drm/amd/display: Restore backlight brightness after system resume\n drm/amd/display: Implement voltage limitation for dali\n ...\n" }, { "commit": "8ce39eb5a67aee25d9f05b40b673c95b23502e3e", "tree": "f5ac35b09a619450d16779f24e66bbe8f3cfb427", "parents": [ "8572cea1461a006bce1d06c0c4b0575869125fa4" ], "author": { "name": "Navid Emamdoost", "email": "navid.emamdoost@gmail.com", "time": "Wed Sep 25 14:05:09 2019 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:12:03 2019 +0200" }, "message": "nfp: flower: fix memory leak in nfp_flower_spawn_vnic_reprs\n\nIn nfp_flower_spawn_vnic_reprs in the loop if initialization or the\nallocations fail memory is leaked. Appropriate releases are added.\n\nFixes: b94524529741 (\"nfp: flower: add per repr private data for LAG offload\")\nSigned-off-by: Navid Emamdoost \u003cnavid.emamdoost@gmail.com\u003e\nAcked-by: Jakub Kicinski \u003cjakub.kicinski@netronome.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8572cea1461a006bce1d06c0c4b0575869125fa4", "tree": "2474340d64175ca072193f1d2c98e87fb8f9ff3e", "parents": [ "dfe5999dc03e321d08190772c98843284d5cf419" ], "author": { "name": "Navid Emamdoost", "email": "navid.emamdoost@gmail.com", "time": "Wed Sep 25 13:24:02 2019 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:10:52 2019 +0200" }, "message": "nfp: flower: prevent memory leak in nfp_flower_spawn_phy_reprs\n\nIn nfp_flower_spawn_phy_reprs, in the for loop over eth_tbl if any of\nintermediate allocations or initializations fail memory is leaked.\nrequiered releases are added.\n\nFixes: b94524529741 (\"nfp: flower: add per repr private data for LAG offload\")\nSigned-off-by: Navid Emamdoost \u003cnavid.emamdoost@gmail.com\u003e\nAcked-by: Jakub Kicinski \u003cjakub.kicinski@netronome.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "dfe5999dc03e321d08190772c98843284d5cf419", "tree": "ac8bf4a5f9709c41c13d3a9371e53480835c734b", "parents": [ "dac91170f8e9c73784af5fad6225e954b795601c" ], "author": { "name": "Paul Blakey", "email": "paulb@mellanox.com", "time": "Wed Sep 25 18:02:35 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:08:28 2019 +0200" }, "message": "net/sched: Set default of CONFIG_NET_TC_SKB_EXT to N\n\nThis a new feature, it is preferred that it defaults to N.\nWe will probe the feature support from userspace before actually using it.\n\nFixes: 95a7233c452a (\u0027net: openvswitch: Set OvS recirc_id from tc chain index\u0027)\nSigned-off-by: Paul Blakey \u003cpaulb@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "dac91170f8e9c73784af5fad6225e954b795601c", "tree": "f153f5702dc01ec960ffd052cfd5b7b7acdf5e3a", "parents": [ "3c30819dc68ab9498216421b6846123900c0a6d3" ], "author": { "name": "David Ahern", "email": "dsahern@gmail.com", "time": "Wed Sep 25 07:53:19 2019 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 20:07:05 2019 +0200" }, "message": "vrf: Do not attempt to create IPv6 mcast rule if IPv6 is disabled\n\nA user reported that vrf create fails when IPv6 is disabled at boot using\n\u0027ipv6.disable\u003d1\u0027:\n https://bugzilla.kernel.org/show_bug.cgi?id\u003d204903\n\nThe failure is adding fib rules at create time. Add RTNL_FAMILY_IP6MR to\nthe check in vrf_fib_rule if ipv6_mod_enabled is disabled.\n\nFixes: e4a38c0c4b27 (\"ipv6: add vrf table handling code for ipv6 mcast\")\nSigned-off-by: David Ahern \u003cdsahern@gmail.com\u003e\nCc: Patrick Ruddy \u003cpruddy@vyatta.att-mail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0cd81d77d0569f1dc1e39abeea93c6184e9b5b54", "tree": "4de744e901974a07202b6849877688ba1cb86581", "parents": [ "ea1e2bbec6edb34fd13032bdfc80276526e87967", "4720101fab62d0453babb0287b58a9c5bf78fb80" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:05:49 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 11:05:49 2019 -0700" }, "message": "Merge tag \u0027ntb-5.4\u0027 of git://github.com/jonmason/ntb\n\nPull NTB updates from Jon Mason:\n \"A few bugfixes and support for new AMD NTB hardware\"\n\n* tag \u0027ntb-5.4\u0027 of git://github.com/jonmason/ntb:\n NTB: fix IDT Kconfig typos/spellos\n ntb_hw_amd: Add memory window support for new AMD hardware\n ntb_hw_amd: Add a new NTB PCI device ID\n NTB: ntb_transport: remove redundant assignment to rc\n ntb_hw_switchtec: make ntb_mw_set_trans() work when addr \u003d\u003d 0\n ntb: point to right memory window index\n" }, { "commit": "ea1e2bbec6edb34fd13032bdfc80276526e87967", "tree": "ad0f1dd8918c98378bd2a9c20e1949bc20546204", "parents": [ "da05b5ea12c1e50b2988a63470d6b69434796f8b" ], "author": { "name": "Jarkko Sakkinen", "email": "jarkko.sakkinen@linux.intel.com", "time": "Fri Sep 27 17:18:05 2019 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 27 10:17:53 2019 -0700" }, "message": "keys: Add Jarkko Sakkinen as co-maintainer\n\nTo address a major procedural concern on Linus\u0027s part the keyrings needs\na co-maintainer.\n\nSuggested-by: David Howells \u003cdhowells@redhat.com\u003e\nSigned-off-by: Jarkko Sakkinen \u003cjarkko.sakkinen@linux.intel.com\u003e\nSigned-off-by: David Howells \u003cdhowells@redhat.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3c30819dc68ab9498216421b6846123900c0a6d3", "tree": "2da4d8143b0a9f171cb38ba70ca9c1f339b24cef", "parents": [ "5c7ff18149fe9f2302f6910cacc80b6183fb0729", "768fb61fcc13b2acaca758275d54c09a65e2968b" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 16:23:32 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 16:23:32 2019 +0200" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf\n\nDaniel Borkmann says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\npull-request: bpf 2019-09-27\n\nThe following pull-request contains BPF updates for your *net* tree.\n\nThe main changes are:\n\n1) Fix libbpf\u0027s BTF dumper to not skip anonymous enum definitions, from Andrii.\n\n2) Fix BTF verifier issues when handling the BTF of vmlinux, from Alexei.\n\n3) Fix nested calls into bpf_event_output() from TCP sockops BPF\n programs, from Allan.\n\n4) Fix NULL pointer dereference in AF_XDP\u0027s xsk map creation when\n allocation fails, from Jonathan.\n\n5) Remove unneeded 64 byte alignment requirement of the AF_XDP UMEM\n headroom, from Bjorn.\n\n6) Remove unused XDP_OPTIONS getsockopt() call which results in an error\n on older kernels, from Toke.\n\n7) Fix a client/server race in tcp_rtt BPF kselftest case, from Stanislav.\n\n8) Fix indentation issue in BTF\u0027s btf_enum_check_kflag_member(), from Colin.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d4e204948fe3e0dc8e1fbf3f8f3290c9c2823be3", "tree": "06439f29ac19eb4436afbeb2c4b3e0a4503dbb38", "parents": [ "bab32fc069ce8829c416e8737c119f62a57970f9" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Mon Sep 16 20:02:39 2019 +0800" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Sep 27 15:24:34 2019 +0200" }, "message": "btrfs: qgroup: Fix reserved data space leak if we have multiple reserve calls\n\n[BUG]\nThe following script can cause btrfs qgroup data space leak:\n\n mkfs.btrfs -f $dev\n mount $dev -o nospace_cache $mnt\n\n btrfs subv create $mnt/subv\n btrfs quota en $mnt\n btrfs quota rescan -w $mnt\n btrfs qgroup limit 128m $mnt/subv\n\n for (( i \u003d 0; i \u003c 3; i++)); do\n # Create 3 64M holes for latter fallocate to fail\n truncate -s 192m $mnt/subv/file\n xfs_io -c \"pwrite 64m 4k\" $mnt/subv/file \u003e /dev/null\n xfs_io -c \"pwrite 128m 4k\" $mnt/subv/file \u003e /dev/null\n sync\n\n # it\u0027s supposed to fail, and each failure will leak at least 64M\n # data space\n xfs_io -f -c \"falloc 0 192m\" $mnt/subv/file \u0026\u003e /dev/null\n rm $mnt/subv/file\n sync\n done\n\n # Shouldn\u0027t fail after we removed the file\n xfs_io -f -c \"falloc 0 64m\" $mnt/subv/file\n\n[CAUSE]\nBtrfs qgroup data reserve code allow multiple reservations to happen on\na single extent_changeset:\nE.g:\n\tbtrfs_qgroup_reserve_data(inode, \u0026data_reserved, 0, SZ_1M);\n\tbtrfs_qgroup_reserve_data(inode, \u0026data_reserved, SZ_1M, SZ_2M);\n\tbtrfs_qgroup_reserve_data(inode, \u0026data_reserved, 0, SZ_4M);\n\nBtrfs qgroup code has its internal tracking to make sure we don\u0027t\ndouble-reserve in above example.\n\nThe only pattern utilizing this feature is in the main while loop of\nbtrfs_fallocate() function.\n\nHowever btrfs_qgroup_reserve_data()\u0027s error handling has a bug in that\non error it clears all ranges in the io_tree with EXTENT_QGROUP_RESERVED\nflag but doesn\u0027t free previously reserved bytes.\n\nThis bug has a two fold effect:\n- Clearing EXTENT_QGROUP_RESERVED ranges\n This is the correct behavior, but it prevents\n btrfs_qgroup_check_reserved_leak() to catch the leakage as the\n detector is purely EXTENT_QGROUP_RESERVED flag based.\n\n- Leak the previously reserved data bytes.\n\nThe bug manifests when N calls to btrfs_qgroup_reserve_data are made and\nthe last one fails, leaking space reserved in the previous ones.\n\n[FIX]\nAlso free previously reserved data bytes when btrfs_qgroup_reserve_data\nfails.\n\nFixes: 524725537023 (\"btrfs: qgroup: Introduce btrfs_qgroup_reserve_data function\")\nCC: stable@vger.kernel.org # 4.4+\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "bab32fc069ce8829c416e8737c119f62a57970f9", "tree": "af6b046ff374a309ad974bf68ab4ea586d391856", "parents": [ "fab273595507a9ec7035df6d5512a955d80a80ba" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Mon Sep 16 20:02:38 2019 +0800" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Sep 27 15:24:28 2019 +0200" }, "message": "btrfs: qgroup: Fix the wrong target io_tree when freeing reserved data space\n\n[BUG]\nUnder the following case with qgroup enabled, if some error happened\nafter we have reserved delalloc space, then in error handling path, we\ncould cause qgroup data space leakage:\n\nFrom btrfs_truncate_block() in inode.c:\n\n\tret \u003d btrfs_delalloc_reserve_space(inode, \u0026data_reserved,\n\t\t\t\t\t block_start, blocksize);\n\tif (ret)\n\t\tgoto out;\n\n again:\n\tpage \u003d find_or_create_page(mapping, index, mask);\n\tif (!page) {\n\t\tbtrfs_delalloc_release_space(inode, data_reserved,\n\t\t\t\t\t block_start, blocksize, true);\n\t\tbtrfs_delalloc_release_extents(BTRFS_I(inode), blocksize, true);\n\t\tret \u003d -ENOMEM;\n\t\tgoto out;\n\t}\n\n[CAUSE]\nIn the above case, btrfs_delalloc_reserve_space() will call\nbtrfs_qgroup_reserve_data() and mark the io_tree range with\nEXTENT_QGROUP_RESERVED flag.\n\nIn the error handling path, we have the following call stack:\nbtrfs_delalloc_release_space()\n|- btrfs_free_reserved_data_space()\n |- btrsf_qgroup_free_data()\n |- __btrfs_qgroup_release_data(reserved\u003d@reserved, free\u003d1)\n |- qgroup_free_reserved_data(reserved\u003d@reserved)\n |- clear_record_extent_bits();\n |- freed +\u003d changeset.bytes_changed;\n\nHowever due to a completion bug, qgroup_free_reserved_data() will clear\nEXTENT_QGROUP_RESERVED flag in BTRFS_I(inode)-\u003eio_failure_tree, other\nthan the correct BTRFS_I(inode)-\u003eio_tree.\nSince io_failure_tree is never marked with that flag,\nbtrfs_qgroup_free_data() will not free any data reserved space at all,\ncausing a leakage.\n\nThis type of error handling can only be triggered by errors outside of\nqgroup code. So EDQUOT error from qgroup can\u0027t trigger it.\n\n[FIX]\nFix the wrong target io_tree.\n\nReported-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nFixes: bc42bda22345 (\"btrfs: qgroup: Fix qgroup reserved space underflow by only freeing reserved ranges\")\nCC: stable@vger.kernel.org # 4.14+\nReviewed-by: Nikolay Borisov \u003cnborisov@suse.com\u003e\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "8d6996630c03d7ceeabe2611378fea5ca1c3f1b3", "tree": "2663815e5f56768c84d80685d1c4dc44156a9840", "parents": [ "2af2783f2ea4f273598557b247e320509247df80" ], "author": { "name": "Yufen Yu", "email": "yuyufen@huawei.com", "time": "Fri Sep 27 16:19:55 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Fri Sep 27 07:01:25 2019 -0600" }, "message": "block: fix null pointer dereference in blk_mq_rq_timed_out()\n\nWe got a null pointer deference BUG_ON in blk_mq_rq_timed_out()\nas following:\n\n[ 108.825472] BUG: kernel NULL pointer dereference, address: 0000000000000040\n[ 108.827059] PGD 0 P4D 0\n[ 108.827313] Oops: 0000 [#1] SMP PTI\n[ 108.827657] CPU: 6 PID: 198 Comm: kworker/6:1H Not tainted 5.3.0-rc8+ #431\n[ 108.829503] Workqueue: kblockd blk_mq_timeout_work\n[ 108.829913] RIP: 0010:blk_mq_check_expired+0x258/0x330\n[ 108.838191] Call Trace:\n[ 108.838406] bt_iter+0x74/0x80\n[ 108.838665] blk_mq_queue_tag_busy_iter+0x204/0x450\n[ 108.839074] ? __switch_to_asm+0x34/0x70\n[ 108.839405] ? blk_mq_stop_hw_queue+0x40/0x40\n[ 108.839823] ? blk_mq_stop_hw_queue+0x40/0x40\n[ 108.840273] ? syscall_return_via_sysret+0xf/0x7f\n[ 108.840732] blk_mq_timeout_work+0x74/0x200\n[ 108.841151] process_one_work+0x297/0x680\n[ 108.841550] worker_thread+0x29c/0x6f0\n[ 108.841926] ? rescuer_thread+0x580/0x580\n[ 108.842344] kthread+0x16a/0x1a0\n[ 108.842666] ? kthread_flush_work+0x170/0x170\n[ 108.843100] ret_from_fork+0x35/0x40\n\nThe bug is caused by the race between timeout handle and completion for\nflush request.\n\nWhen timeout handle function blk_mq_rq_timed_out() try to read\n\u0027req-\u003eq-\u003emq_ops\u0027, the \u0027req\u0027 have completed and reinitiated by next\nflush request, which would call blk_rq_init() to clear \u0027req\u0027 as 0.\n\nAfter commit 12f5b93145 (\"blk-mq: Remove generation seqeunce\"),\nnormal requests lifetime are protected by refcount. Until \u0027rq-\u003eref\u0027\ndrop to zero, the request can really be free. Thus, these requests\ncannot been reused before timeout handle finish.\n\nHowever, flush request has defined .end_io and rq-\u003eend_io() is still\ncalled even if \u0027rq-\u003eref\u0027 doesn\u0027t drop to zero. After that, the \u0027flush_rq\u0027\ncan be reused by the next flush request handle, resulting in null\npointer deference BUG ON.\n\nWe fix this problem by covering flush request with \u0027rq-\u003eref\u0027.\nIf the refcount is not zero, flush_end_io() return and wait the\nlast holder recall it. To record the request status, we add a new\nentry \u0027rq_status\u0027, which will be used in flush_end_io().\n\nCc: Christoph Hellwig \u003chch@infradead.org\u003e\nCc: Keith Busch \u003ckeith.busch@intel.com\u003e\nCc: Bart Van Assche \u003cbvanassche@acm.org\u003e\nCc: stable@vger.kernel.org # v4.18+\nReviewed-by: Ming Lei \u003cming.lei@redhat.com\u003e\nReviewed-by: Bob Liu \u003cbob.liu@oracle.com\u003e\nSigned-off-by: Yufen Yu \u003cyuyufen@huawei.com\u003e\n\n-------\nv2:\n - move rq_status from struct request to struct blk_flush_queue\nv3:\n - remove unnecessary \u0027{}\u0027 pair.\nv4:\n - let spinlock to protect \u0027fq-\u003erq_status\u0027\nv5:\n - move rq_status after flush_running_idx member of struct blk_flush_queue\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "5c7ff18149fe9f2302f6910cacc80b6183fb0729", "tree": "737cf41f5cb7364febd3fa4e43bc6194de6c6ab2", "parents": [ "05733434ee9ae6548723a808647248583e347cca", "e3ae1f96accd21405715fe9c56b4d83bc7d96d44" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:13:55 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:15:16 2019 +0200" }, "message": "Merge branch \u0027qdisc-destroy\u0027\n\nVlad Buslov says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nFix Qdisc destroy issues caused by adding fine-grained locking to filter API\n\nTC filter API unlocking introduced several new fine-grained locks. The\nchange caused sleeping-while-atomic BUGs in several Qdiscs that call cls\nAPIs which need to obtain new mutex while holding sch tree spinlock. This\nseries fixes affected Qdiscs by ensuring that cls API that became sleeping\nis only called outside of sch tree lock critical section.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nAcked-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e3ae1f96accd21405715fe9c56b4d83bc7d96d44", "tree": "737cf41f5cb7364febd3fa4e43bc6194de6c6ab2", "parents": [ "c2999f7fb05b87da4060e38150c70fa46794d82b" ], "author": { "name": "Vlad Buslov", "email": "vladbu@mellanox.com", "time": "Tue Sep 24 18:51:18 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:13:55 2019 +0200" }, "message": "net: sched: sch_sfb: don\u0027t call qdisc_put() while holding tree lock\n\nRecent changes that removed rtnl dependency from rules update path of tc\nalso made tcf_block_put() function sleeping. This function is called from\nops-\u003edestroy() of several Qdisc implementations, which in turn is called by\nqdisc_put(). Some Qdiscs call qdisc_put() while holding sch tree spinlock,\nwhich results sleeping-while-atomic BUG.\n\nSteps to reproduce for sfb:\n\ntc qdisc add dev ens1f0 handle 1: root sfb\ntc qdisc add dev ens1f0 parent 1:10 handle 50: sfq perturb 10\ntc qdisc change dev ens1f0 root handle 1: sfb\n\nResulting dmesg:\n\n[ 7265.938717] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:909\n[ 7265.940152] in_atomic(): 1, irqs_disabled(): 0, pid: 28579, name: tc\n[ 7265.941455] INFO: lockdep is turned off.\n[ 7265.942744] CPU: 11 PID: 28579 Comm: tc Tainted: G W 5.3.0-rc8+ #721\n[ 7265.944065] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017\n[ 7265.945396] Call Trace:\n[ 7265.946709] dump_stack+0x85/0xc0\n[ 7265.947994] ___might_sleep.cold+0xac/0xbc\n[ 7265.949282] __mutex_lock+0x5b/0x960\n[ 7265.950543] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 7265.951803] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 7265.953022] tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 7265.954248] tcf_block_put_ext.part.0+0x21/0x50\n[ 7265.955478] tcf_block_put+0x50/0x70\n[ 7265.956694] sfq_destroy+0x15/0x50 [sch_sfq]\n[ 7265.957898] qdisc_destroy+0x5f/0x160\n[ 7265.959099] sfb_change+0x175/0x330 [sch_sfb]\n[ 7265.960304] tc_modify_qdisc+0x324/0x840\n[ 7265.961503] rtnetlink_rcv_msg+0x170/0x4b0\n[ 7265.962692] ? netlink_deliver_tap+0x95/0x400\n[ 7265.963876] ? rtnl_dellink+0x2d0/0x2d0\n[ 7265.965064] netlink_rcv_skb+0x49/0x110\n[ 7265.966251] netlink_unicast+0x171/0x200\n[ 7265.967427] netlink_sendmsg+0x224/0x3f0\n[ 7265.968595] sock_sendmsg+0x5e/0x60\n[ 7265.969753] ___sys_sendmsg+0x2ae/0x330\n[ 7265.970916] ? ___sys_recvmsg+0x159/0x1f0\n[ 7265.972074] ? do_wp_page+0x9c/0x790\n[ 7265.973233] ? __handle_mm_fault+0xcd3/0x19e0\n[ 7265.974407] __sys_sendmsg+0x59/0xa0\n[ 7265.975591] do_syscall_64+0x5c/0xb0\n[ 7265.976753] entry_SYSCALL_64_after_hwframe+0x49/0xbe\n[ 7265.977938] RIP: 0033:0x7f229069f7b8\n[ 7265.979117] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 \u003c48\u003e 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83 ec 28 89 5\n4\n[ 7265.981681] RSP: 002b:00007ffd7ed2d158 EFLAGS: 00000246 ORIG_RAX: 000000000000002e\n[ 7265.983001] RAX: ffffffffffffffda RBX: 000000005d813ca1 RCX: 00007f229069f7b8\n[ 7265.984336] RDX: 0000000000000000 RSI: 00007ffd7ed2d1c0 RDI: 0000000000000003\n[ 7265.985682] RBP: 0000000000000000 R08: 0000000000000001 R09: 000000000165c9a0\n[ 7265.987021] R10: 0000000000404eda R11: 0000000000000246 R12: 0000000000000001\n[ 7265.988309] R13: 000000000047f640 R14: 0000000000000000 R15: 0000000000000000\n\nIn sfb_change() function use qdisc_purge_queue() instead of\nqdisc_tree_flush_backlog() to properly reset old child Qdisc and save\npointer to it into local temporary variable. Put reference to Qdisc after\nsch tree lock is released in order not to call potentially sleeping cls API\nin atomic section. This is safe to do because Qdisc has already been reset\nby qdisc_purge_queue() inside sch tree lock critical section.\n\nReported-by: syzbot+ac54455281db908c581e@syzkaller.appspotmail.com\nFixes: c266f64dbfa2 (\"net: sched: protect block state with mutex\")\nSuggested-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: Vlad Buslov \u003cvladbu@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c2999f7fb05b87da4060e38150c70fa46794d82b", "tree": "6a60ecd89a1e9f4d1094b6d5c9072017f7e8f4dd", "parents": [ "4ce70b4aed5752332b268909336b351721965dc4" ], "author": { "name": "Vlad Buslov", "email": "vladbu@mellanox.com", "time": "Tue Sep 24 18:51:17 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:13:55 2019 +0200" }, "message": "net: sched: multiq: don\u0027t call qdisc_put() while holding tree lock\n\nRecent changes that removed rtnl dependency from rules update path of tc\nalso made tcf_block_put() function sleeping. This function is called from\nops-\u003edestroy() of several Qdisc implementations, which in turn is called by\nqdisc_put(). Some Qdiscs call qdisc_put() while holding sch tree spinlock,\nwhich results sleeping-while-atomic BUG.\n\nSteps to reproduce for multiq:\n\ntc qdisc add dev ens1f0 root handle 1: multiq\ntc qdisc add dev ens1f0 parent 1:10 handle 50: sfq perturb 10\nethtool -L ens1f0 combined 2\ntc qdisc change dev ens1f0 root handle 1: multiq\n\nResulting dmesg:\n\n[ 5539.419344] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:909\n[ 5539.420945] in_atomic(): 1, irqs_disabled(): 0, pid: 27658, name: tc\n[ 5539.422435] INFO: lockdep is turned off.\n[ 5539.423904] CPU: 21 PID: 27658 Comm: tc Tainted: G W 5.3.0-rc8+ #721\n[ 5539.425400] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017\n[ 5539.426911] Call Trace:\n[ 5539.428380] dump_stack+0x85/0xc0\n[ 5539.429823] ___might_sleep.cold+0xac/0xbc\n[ 5539.431262] __mutex_lock+0x5b/0x960\n[ 5539.432682] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 5539.434103] ? __nla_validate_parse+0x51/0x840\n[ 5539.435493] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 5539.436903] tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 5539.438327] tcf_block_put_ext.part.0+0x21/0x50\n[ 5539.439752] tcf_block_put+0x50/0x70\n[ 5539.441165] sfq_destroy+0x15/0x50 [sch_sfq]\n[ 5539.442570] qdisc_destroy+0x5f/0x160\n[ 5539.444000] multiq_tune+0x14a/0x420 [sch_multiq]\n[ 5539.445421] tc_modify_qdisc+0x324/0x840\n[ 5539.446841] rtnetlink_rcv_msg+0x170/0x4b0\n[ 5539.448269] ? netlink_deliver_tap+0x95/0x400\n[ 5539.449691] ? rtnl_dellink+0x2d0/0x2d0\n[ 5539.451116] netlink_rcv_skb+0x49/0x110\n[ 5539.452522] netlink_unicast+0x171/0x200\n[ 5539.453914] netlink_sendmsg+0x224/0x3f0\n[ 5539.455304] sock_sendmsg+0x5e/0x60\n[ 5539.456686] ___sys_sendmsg+0x2ae/0x330\n[ 5539.458071] ? ___sys_recvmsg+0x159/0x1f0\n[ 5539.459461] ? do_wp_page+0x9c/0x790\n[ 5539.460846] ? __handle_mm_fault+0xcd3/0x19e0\n[ 5539.462263] __sys_sendmsg+0x59/0xa0\n[ 5539.463661] do_syscall_64+0x5c/0xb0\n[ 5539.465044] entry_SYSCALL_64_after_hwframe+0x49/0xbe\n[ 5539.466454] RIP: 0033:0x7f1fe08177b8\n[ 5539.467863] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 \u003c48\u003e 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83 ec 28 89 5\n4\n[ 5539.470906] RSP: 002b:00007ffe812de5d8 EFLAGS: 00000246 ORIG_RAX: 000000000000002e\n[ 5539.472483] RAX: ffffffffffffffda RBX: 000000005d8135e3 RCX: 00007f1fe08177b8\n[ 5539.474069] RDX: 0000000000000000 RSI: 00007ffe812de640 RDI: 0000000000000003\n[ 5539.475655] RBP: 0000000000000000 R08: 0000000000000001 R09: 000000000182e9b0\n[ 5539.477203] R10: 0000000000404eda R11: 0000000000000246 R12: 0000000000000001\n[ 5539.478699] R13: 000000000047f640 R14: 0000000000000000 R15: 0000000000000000\n\nRearrange locking in multiq_tune() in following ways:\n\n- In loop that removes Qdiscs from disabled queues, call\n qdisc_purge_queue() instead of qdisc_tree_flush_backlog() on Qdisc that\n is being destroyed. Save the Qdisc in temporary allocated array and call\n qdisc_put() on each element of the array after sch tree lock is released.\n This is safe to do because Qdiscs have already been reset by\n qdisc_purge_queue() inside sch tree lock critical section.\n\n- Do the same change for second loop that initializes Qdiscs for newly\n enabled queues in multiq_tune() function. Since sch tree lock is obtained\n and released on each iteration of this loop, just call qdisc_put()\n directly outside of critical section. Don\u0027t verify that old Qdisc is not\n noop_qdisc before releasing reference to it because such check is already\n performed by qdisc_put*() functions.\n\nFixes: c266f64dbfa2 (\"net: sched: protect block state with mutex\")\nSigned-off-by: Vlad Buslov \u003cvladbu@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4ce70b4aed5752332b268909336b351721965dc4", "tree": "7788cc6485d64e5801da621e4ca43709fec67860", "parents": [ "05733434ee9ae6548723a808647248583e347cca" ], "author": { "name": "Vlad Buslov", "email": "vladbu@mellanox.com", "time": "Tue Sep 24 18:51:16 2019 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:13:55 2019 +0200" }, "message": "net: sched: sch_htb: don\u0027t call qdisc_put() while holding tree lock\n\nRecent changes that removed rtnl dependency from rules update path of tc\nalso made tcf_block_put() function sleeping. This function is called from\nops-\u003edestroy() of several Qdisc implementations, which in turn is called by\nqdisc_put(). Some Qdiscs call qdisc_put() while holding sch tree spinlock,\nwhich results sleeping-while-atomic BUG.\n\nSteps to reproduce for htb:\n\ntc qdisc add dev ens1f0 root handle 1: htb default 12\ntc class add dev ens1f0 parent 1: classid 1:1 htb rate 100kbps ceil 100kbps\ntc qdisc add dev ens1f0 parent 1:1 handle 40: sfq perturb 10\ntc class add dev ens1f0 parent 1:1 classid 1:2 htb rate 100kbps ceil 100kbps\n\nResulting dmesg:\n\n[ 4791.148551] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:909\n[ 4791.151354] in_atomic(): 1, irqs_disabled(): 0, pid: 27273, name: tc\n[ 4791.152805] INFO: lockdep is turned off.\n[ 4791.153605] CPU: 19 PID: 27273 Comm: tc Tainted: G W 5.3.0-rc8+ #721\n[ 4791.154336] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017\n[ 4791.155075] Call Trace:\n[ 4791.155803] dump_stack+0x85/0xc0\n[ 4791.156529] ___might_sleep.cold+0xac/0xbc\n[ 4791.157251] __mutex_lock+0x5b/0x960\n[ 4791.157966] ? console_unlock+0x363/0x5d0\n[ 4791.158676] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 4791.159395] ? tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 4791.160103] tcf_chain0_head_change_cb_del.isra.0+0x1b/0xf0\n[ 4791.160815] tcf_block_put_ext.part.0+0x21/0x50\n[ 4791.161530] tcf_block_put+0x50/0x70\n[ 4791.162233] sfq_destroy+0x15/0x50 [sch_sfq]\n[ 4791.162936] qdisc_destroy+0x5f/0x160\n[ 4791.163642] htb_change_class.cold+0x5df/0x69d [sch_htb]\n[ 4791.164505] tc_ctl_tclass+0x19d/0x480\n[ 4791.165360] rtnetlink_rcv_msg+0x170/0x4b0\n[ 4791.166191] ? netlink_deliver_tap+0x95/0x400\n[ 4791.166907] ? rtnl_dellink+0x2d0/0x2d0\n[ 4791.167625] netlink_rcv_skb+0x49/0x110\n[ 4791.168345] netlink_unicast+0x171/0x200\n[ 4791.169058] netlink_sendmsg+0x224/0x3f0\n[ 4791.169771] sock_sendmsg+0x5e/0x60\n[ 4791.170475] ___sys_sendmsg+0x2ae/0x330\n[ 4791.171183] ? ___sys_recvmsg+0x159/0x1f0\n[ 4791.171894] ? do_wp_page+0x9c/0x790\n[ 4791.172595] ? __handle_mm_fault+0xcd3/0x19e0\n[ 4791.173309] __sys_sendmsg+0x59/0xa0\n[ 4791.174024] do_syscall_64+0x5c/0xb0\n[ 4791.174725] entry_SYSCALL_64_after_hwframe+0x49/0xbe\n[ 4791.175435] RIP: 0033:0x7f0aa41497b8\n[ 4791.176129] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 \u003c48\u003e 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83 ec 28 89 5\n4\n[ 4791.177532] RSP: 002b:00007fff4e37d588 EFLAGS: 00000246 ORIG_RAX: 000000000000002e\n[ 4791.178243] RAX: ffffffffffffffda RBX: 000000005d8132f7 RCX: 00007f0aa41497b8\n[ 4791.178947] RDX: 0000000000000000 RSI: 00007fff4e37d5f0 RDI: 0000000000000003\n[ 4791.179662] RBP: 0000000000000000 R08: 0000000000000001 R09: 00000000020149a0\n[ 4791.180382] R10: 0000000000404eda R11: 0000000000000246 R12: 0000000000000001\n[ 4791.181100] R13: 000000000047f640 R14: 0000000000000000 R15: 0000000000000000\n\nIn htb_change_class() function save parent-\u003eleaf.q to local temporary\nvariable and put reference to it after sch tree lock is released in order\nnot to call potentially sleeping cls API in atomic section. This is safe to\ndo because Qdisc has already been reset by qdisc_purge_queue() inside sch\ntree lock critical section.\n\nFixes: c266f64dbfa2 (\"net: sched: protect block state with mutex\")\nSigned-off-by: Vlad Buslov \u003cvladbu@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "05733434ee9ae6548723a808647248583e347cca", "tree": "7a9fe3ea13f277233a5871b7ac7008bdda4c51f3", "parents": [ "4e1e83be1cd4379eee1571ced88e0b0c1069b0af" ], "author": { "name": "Ka-Cheong Poon", "email": "ka-cheong.poon@oracle.com", "time": "Tue Sep 24 08:51:16 2019 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:10:55 2019 +0200" }, "message": "net/rds: Check laddr_check before calling it\n\nIn rds_bind(), laddr_check is called without checking if it is NULL or\nnot. And rs_transport should be reset if rds_add_bound() fails.\n\nFixes: c5c1a030a7db (\"net/rds: An rds_sock is added too early to the hash table\")\nReported-by: syzbot+fae39afd2101a17ec624@syzkaller.appspotmail.com\nSigned-off-by: Ka-Cheong Poon \u003cka-cheong.poon@oracle.com\u003e\nAcked-by: Santosh Shilimkar \u003csantosh.shilimkar@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4e1e83be1cd4379eee1571ced88e0b0c1069b0af", "tree": "a3c9c3a820ff3b8cc28671d498fad022142f74ff", "parents": [ "2b6fd3ea438c742d162a40a124b0181922633163", "f6c0f5d209fa80eb808e08aa4206f6e264041ef6" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:05:02 2019 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 27 12:05:02 2019 +0200" }, "message": "Merge branch \u0027SO_PRIORITY\u0027\n\nEric Dumazet says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\ntcp: provide correct skb-\u003epriority\n\nSO_PRIORITY socket option requests TCP egress packets\nto contain a user provided value.\n\nTCP manages to send most packets with the requested values,\nnotably for TCP_ESTABLISHED state, but fails to do so for\nfew packets.\n\nThese packets are control packets sent on behalf\nof SYN_RECV or TIME_WAIT states.\n\nNote that to test this with packetdrill, it is a bit\nof a hassle, since packetdrill can not verify priority\nof egress packets, other than indirect observations,\nusing for example sch_prio on its tunnel device.\n\nThe bad skb priorities cause problems for GCP,\nas this field is one of the keys used in routing.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" } ], "next": "f6c0f5d209fa80eb808e08aa4206f6e264041ef6" }