)]}' { "log": [ { "commit": "8d3dea210042f54b952b481838c1e7dfc4ec751d", "tree": "027a4c249cab159e5e287f9c49289db0a662aa11", "parents": [ "6004b044f77121d09cacf47073c19dc106da0e9d", "8fa5070833886268e4fb646daaca99f725b378e9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 16:06:31 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 16:06:31 2024 -0800" }, "message": "Merge tag \u0027mips-fixes_6.8_2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux\n\nPull MIPS fixes from Thomas Bogendoerfer:\n\n - Fix for broken ipv6 checksums\n\n - Fix handling of exceptions in delay slots\n\n* tag \u0027mips-fixes_6.8_2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:\n mm/memory: Use exception ip to search exception tables\n MIPS: Clear Cause.BD in instruction_pointer_set\n ptrace: Introduce exception_ip arch hook\n MIPS: Add \u0027memory\u0027 clobber to csum_ipv6_magic() inline assembler\n" }, { "commit": "6004b044f77121d09cacf47073c19dc106da0e9d", "tree": "eb257598a56b23f0fe7af675fe423ce93df66b80", "parents": [ "1f3a3e2aaeb4e6ba9b6df6f2e720131765b23b82", "bb6f4dbe2639d5b8a9fde4bfb6fefecfd3f18df3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 16:02:36 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 16:02:36 2024 -0800" }, "message": "Merge tag \u0027landlock-6.8-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mic/linux\n\nPull landlock test fixes from Mickaël Salaün:\n \"Fix build issues for tests, and improve test compatibility\"\n\n* tag \u0027landlock-6.8-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mic/linux:\n selftests/landlock: Fix capability for net_test\n selftests/landlock: Fix fs_test build with old libc\n selftests/landlock: Fix net_test build with old libc\n" }, { "commit": "1f3a3e2aaeb4e6ba9b6df6f2e720131765b23b82", "tree": "8f84ebe0aecfbd6c12364ce6fb9f82eeb683e336", "parents": [ "91f842ffe6ca1e97a3966e9e499c3ac6fbcc4bc4", "2f6397e448e689adf57e6788c90f913abd7e1af8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 15:47:02 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 15:47:02 2024 -0800" }, "message": "Merge tag \u0027for-6.8-rc4-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux\n\nPull btrfs fixes from David Sterba:\n \"A few regular fixes and one fix for space reservation regression since\n 6.7 that users have been reporting:\n\n - fix over-reservation of metadata chunks due to not keeping proper\n balance between global block reserve and delayed refs reserve; in\n practice this leaves behind empty metadata block groups, the\n workaround is to reclaim them by using the \u0027-musage\u003d1\u0027 balance\n filter\n\n - other space reservation fixes:\n - do not delete unused block group if it may be used soon\n - do not reserve space for checksums for NOCOW files\n\n - fix extent map assertion failure when writing out free space inode\n\n - reject encoded write if inode has nodatasum flag set\n\n - fix chunk map leak when loading block group zone info\"\n\n* tag \u0027for-6.8-rc4-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:\n btrfs: don\u0027t refill whole delayed refs block reserve when starting transaction\n btrfs: zoned: fix chunk map leak when loading block group zone info\n btrfs: reject encoded write if inode has nodatasum flag set\n btrfs: don\u0027t reserve space for checksums when writing to nocow files\n btrfs: add new unused block groups to the list of unused block groups\n btrfs: do not delete unused block group if it may be used soon\n btrfs: add and use helper to check if block group is used\n btrfs: don\u0027t drop extent_map for free space inode on write error\n" }, { "commit": "91f842ffe6ca1e97a3966e9e499c3ac6fbcc4bc4", "tree": "c6bfd68bd577b7621ed815a04386c1cba871379f", "parents": [ "7e90b5c295ec1e47c8ad865429f046970c549a66", "829388b725f8d266ccec32a2f446717d8693eaba" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 15:34:03 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Feb 14 15:34:03 2024 -0800" }, "message": "Merge tag \u0027linux_kselftest-kunit-fixes-6.8-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest\n\nPull KUnit fix from Shuah Khan:\n \"One important fix to unregister kunit_bus when KUnit module is\n unloaded.\n\n Not doing so causes an error when KUnit module tries to re-register\n the bus when it gets reloaded\"\n\n* tag \u0027linux_kselftest-kunit-fixes-6.8-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:\n kunit: device: Unregister the kunit_bus on shutdown\n" }, { "commit": "2f6397e448e689adf57e6788c90f913abd7e1af8", "tree": "d00f9a55d59b1dded35f94f46fa4b65f76b3e452", "parents": [ "88e81a67773017a2b93f6d5fe7c78bb0c5a6e4dd" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Fri Feb 02 14:32:17 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Tue Feb 13 18:39:09 2024 +0100" }, "message": "btrfs: don\u0027t refill whole delayed refs block reserve when starting transaction\n\nSince commit 28270e25c69a (\"btrfs: always reserve space for delayed refs\nwhen starting transaction\") we started not only to reserve metadata space\nfor the delayed refs a caller of btrfs_start_transaction() might generate\nbut also to try to fully refill the delayed refs block reserve, because\nthere are several case where we generate delayed refs and haven\u0027t reserved\nspace for them, relying on the global block reserve. Relying too much on\nthe global block reserve is not always safe, and can result in hitting\n-ENOSPC during transaction commits or worst, in rare cases, being unable\nto mount a filesystem that needs to do orphan cleanup or anything that\nrequires modifying the filesystem during mount, and has no more\nunallocated space and the metadata space is nearly full. This was\nexplained in detail in that commit\u0027s change log.\n\nHowever the gap between the reserved amount and the size of the delayed\nrefs block reserve can be huge, so attempting to reserve space for such\na gap can result in allocating many metadata block groups that end up\nnot being used. After a recent patch, with the subject:\n\n \"btrfs: add new unused block groups to the list of unused block groups\"\n\nWe started to add new block groups that are unused to the list of unused\nblock groups, to avoid having them around for a very long time in case\nthey are never used, because a block group is only added to the list of\nunused block groups when we deallocate the last extent or when mounting\nthe filesystem and the block group has 0 bytes used. This is not a problem\nintroduced by the commit mentioned earlier, it always existed as our\nmetadata space reservations are, most of the time, pessimistic and end up\nnot using all the space they reserved, so we can occasionally end up with\none or two unused metadata block groups for a long period. However after\nthat commit mentioned earlier, we are just more pessimistic in the\nmetadata space reservations when starting a transaction and therefore the\nissue is more likely to happen.\n\nThis however is not always enough because we might create unused metadata\nblock groups when reserving metadata space at a high rate if there\u0027s\nalways a gap in the delayed refs block reserve and the cleaner kthread\nisn\u0027t triggered often enough or is busy with other work (running delayed\niputs, cleaning deleted roots, etc), not to mention the block group\u0027s\nallocated space is only usable for a new block group after the transaction\nused to remove it is committed.\n\nA user reported that he\u0027s getting a lot of allocated metadata block groups\nbut the usage percentage of metadata space was very low compared to the\ntotal allocated space, specially after running a series of block group\nrelocations.\n\nSo for now stop trying to refill the gap in the delayed refs block reserve\nand reserve space only for the delayed refs we are expected to generate\nwhen starting a transaction.\n\nCC: stable@vger.kernel.org # 6.7+\nReported-by: Ivan Shapovalov \u003cintelfx@intelfx.name\u003e\nLink: https://lore.kernel.org/linux-btrfs/9cdbf0ca9cdda1b4c84e15e548af7d7f9f926382.camel@intelfx.name/\nLink: https://lore.kernel.org/linux-btrfs/CAL3q7H6802ayLHUJFztzZAVzBLJAGdFx\u003d6FHNNy87+obZXXZpQ@mail.gmail.com/\nTested-by: Ivan Shapovalov \u003cintelfx@intelfx.name\u003e\nReported-by: Heddxh \u003cg311571057@gmail.com\u003e\nLink: https://lore.kernel.org/linux-btrfs/CAE93xANEby6RezOD\u003dzcofENYZOT-wpYygJyauyUAZkLv6XVFOA@mail.gmail.com/\nReviewed-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "88e81a67773017a2b93f6d5fe7c78bb0c5a6e4dd", "tree": "dfc87ddb3c57ac49f85bc68d237b84fdc17361f1", "parents": [ "1bd96c92c6a0a4d43815eb685c15aa4b78879dc9" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Mon Feb 12 21:50:53 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Tue Feb 13 18:38:19 2024 +0100" }, "message": "btrfs: zoned: fix chunk map leak when loading block group zone info\n\nAt btrfs_load_block_group_zone_info() we never drop a reference on the\nchunk map we have looked up, therefore leaking a reference on it. So\nadd the missing btrfs_free_chunk_map() at the end of the function.\n\nFixes: 7dc66abb5a47 (\"btrfs: use a dedicated data structure for chunk maps\")\nReported-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nTested-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Anand Jain \u003canand.jain@oracle.com\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: David Sterba \u003cdsterba@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "1bd96c92c6a0a4d43815eb685c15aa4b78879dc9", "tree": "2e76ef3f5b5ba62834ce44eaac8a7a4b6b6ef492", "parents": [ "feefe1f49d26bad9d8997096e3a200280fa7b1c5" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Fri Feb 02 12:09:22 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Tue Feb 13 18:38:05 2024 +0100" }, "message": "btrfs: reject encoded write if inode has nodatasum flag set\n\nCurrently we allow an encoded write against inodes that have the NODATASUM\nflag set, either because they are NOCOW files or they were created while\nthe filesystem was mounted with \"-o nodatasum\". This results in having\ncompressed extents without corresponding checksums, which is a filesystem\ninconsistency reported by \u0027btrfs check\u0027.\n\nFor example, running btrfs/281 with MOUNT_OPTIONS\u003d\"-o nodatacow\" triggers\nthis and \u0027btrfs check\u0027 errors out with:\n\n [1/7] checking root items\n [2/7] checking extents\n [3/7] checking free space tree\n [4/7] checking fs roots\n root 256 inode 257 errors 1040, bad file extent, some csum missing\n root 256 inode 258 errors 1040, bad file extent, some csum missing\n ERROR: errors found in fs roots\n (...)\n\nSo reject encoded writes if the target inode has NODATASUM set.\n\nCC: stable@vger.kernel.org # 6.1+\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: David Sterba \u003cdsterba@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "feefe1f49d26bad9d8997096e3a200280fa7b1c5", "tree": "beebaee9d360b2b4f399dae22c2f887c071616f0", "parents": [ "12c5128f101bfa47a08e4c0e1a75cfa2d0872bcd" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Wed Jan 31 17:18:04 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Tue Feb 13 18:36:35 2024 +0100" }, "message": "btrfs: don\u0027t reserve space for checksums when writing to nocow files\n\nCurrently when doing a write to a file we always reserve metadata space\nfor inserting data checksums. However we don\u0027t need to do it if we have\na nodatacow file (-o nodatacow mount option or chattr +C) or if checksums\nare disabled (-o nodatasum mount option), as in that case we are only\nadding unnecessary pressure to metadata reservations.\n\nFor example on x86_64, with the default node size of 16K, a 4K buffered\nwrite into a nodatacow file is reserving 655360 bytes of metadata space,\nas it\u0027s accounting for checksums. After this change, which stops reserving\nspace for checksums if we have a nodatacow file or checksums are disabled,\nwe only need to reserve 393216 bytes of metadata.\n\nCC: stable@vger.kernel.org # 6.1+\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: David Sterba \u003cdsterba@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "7e90b5c295ec1e47c8ad865429f046970c549a66", "tree": "7b095a8dd3035b175f9d6509b5940fa2c5c0b784", "parents": [ "c664e16bb1ba1c8cf1d7ecf3df5fd83bbb8ac15a", "b5f319360371087d52070d8f3fc7789e80ce69a6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Feb 13 08:38:57 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Feb 13 08:38:57 2024 -0800" }, "message": "Merge tag \u0027trace-tools-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull tracing tooling fixes from Steven Rostedt:\n \"RTLA:\n\n - rtla tools are exiting with a positive value when usage() is\n called. Make them return 0 if the usage was called via -h/--help\n\n - the -P priority sets the sched priority for rtla workload. When the\n SCHED_OTHER scheduler is selected, it sets the rt_priority instead\n of the nice parameter. Setting the nice value is the correct thing,\n so fix it\n\n - rtla is failing to compile with clang due to unsupported options\n from gcc. Adjusting the compiler/linker options makes clang work\n properly\n\n - Remove the sched_getattr() unused function on utils.c\n\n - Fixes for variable initialization and size, reported by clang\n\n Verification:\n\n - rv is failing to compile with clang due to unsupported options from\n gcc. Adjusting the compiler/linker options makes clang work\n properly\n\n - Fix an uninitialized variable on in_kernel.c reported by clang\"\n\n* tag \u0027trace-tools-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n tools/rtla: Exit with EXIT_SUCCESS when help is invoked\n tools/rtla: Replace setting prio with nice for SCHED_OTHER\n tools/rv: Fix curr_reactor uninitialized variable\n tools/rv: Fix Makefile compiler options for clang\n tools/rtla: Remove unused sched_getattr() function\n tools/rtla: Fix clang warning about mount_point var size\n tools/rtla: Fix uninitialized bucket/data-\u003ebucket_size warning\n tools/rtla: Fix Makefile compiler options for clang\n" }, { "commit": "c664e16bb1ba1c8cf1d7ecf3df5fd83bbb8ac15a", "tree": "d3711a8d2219030949d90d4eb7738c51a3c140cb", "parents": [ "716f4aaa7b48a55c73d632d0657b35342b1fefd7", "c23de7ceae59e4ca5894c3ecf4f785c50c0fa428" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Feb 12 14:11:30 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Feb 12 14:11:30 2024 -0800" }, "message": "Merge tag \u0027docs-6.8-fixes2\u0027 of git://git.lwn.net/linux\n\nPull documentation fix from Jonathan Corbet:\n \"A single fix to the kernel_feat extension for a bug that will crash\n the docs build in some situations\"\n\n* tag \u0027docs-6.8-fixes2\u0027 of git://git.lwn.net/linux:\n docs: kernel_feat.py: fix build error for missing files\n" }, { "commit": "8fa5070833886268e4fb646daaca99f725b378e9", "tree": "5d391de742ad041dbf2f46b87762dd15566d7349", "parents": [ "9d6e21ddf20293b3880ae55b9d14de91c5891c59" ], "author": { "name": "Jiaxun Yang", "email": "jiaxun.yang@flygoat.com", "time": "Fri Feb 02 12:30:28 2024 +0000" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Mon Feb 12 23:04:42 2024 +0100" }, "message": "mm/memory: Use exception ip to search exception tables\n\nOn architectures with delay slot, instruction_pointer() may differ\nfrom where exception was triggered.\n\nUse exception_ip we just introduced to search exception tables to\nget rid of the problem.\n\nFixes: 4bce37a68ff8 (\"mips/mm: Convert to using lock_mm_and_find_vma()\")\nReported-by: Xi Ruoyao \u003cxry111@xry111.site\u003e\nLink: https://lore.kernel.org/r/75e9fd7b08562ad9b456a5bdaacb7cc220311cc9.camel@xry111.site/\nSuggested-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nSigned-off-by: Jiaxun Yang \u003cjiaxun.yang@flygoat.com\u003e\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\n" }, { "commit": "9d6e21ddf20293b3880ae55b9d14de91c5891c59", "tree": "9635e60f51ed938cfc63626b3a875ee3409fc991", "parents": [ "11ba1728be3edb6928791f4c622f154ebe228ae6" ], "author": { "name": "Jiaxun Yang", "email": "jiaxun.yang@flygoat.com", "time": "Fri Feb 02 12:30:27 2024 +0000" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Mon Feb 12 23:04:40 2024 +0100" }, "message": "MIPS: Clear Cause.BD in instruction_pointer_set\n\nClear Cause.BD after we use instruction_pointer_set to override\nEPC.\n\nThis can prevent exception_epc check against instruction code at\nnew return address.\nIt won\u0027t be considered as \"in delay slot\" after epc being overridden\nanyway.\n\nSigned-off-by: Jiaxun Yang \u003cjiaxun.yang@flygoat.com\u003e\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\n" }, { "commit": "11ba1728be3edb6928791f4c622f154ebe228ae6", "tree": "f60c7001335271870aaa3e6f3bef8996b22a1e9c", "parents": [ "d55347bfe4e66dce2e1e7501e5492f4af3e315f8" ], "author": { "name": "Jiaxun Yang", "email": "jiaxun.yang@flygoat.com", "time": "Fri Feb 02 12:30:26 2024 +0000" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Mon Feb 12 23:04:34 2024 +0100" }, "message": "ptrace: Introduce exception_ip arch hook\n\nOn architectures with delay slot, architecture level instruction\npointer (or program counter) in pt_regs may differ from where\nexception was triggered.\n\nIntroduce exception_ip hook to invoke architecture code and determine\nactual instruction pointer to the exception.\n\nLink: https://lore.kernel.org/lkml/00d1b813-c55f-4365-8d81-d70258e10b16@app.fastmail.com/\nSigned-off-by: Jiaxun Yang \u003cjiaxun.yang@flygoat.com\u003e\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\n" }, { "commit": "d55347bfe4e66dce2e1e7501e5492f4af3e315f8", "tree": "2847f4768a0d61f1c728f91ff30b6cb2ddbac6bb", "parents": [ "841c35169323cd833294798e58b9bf63fa4fa1de" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Sun Feb 11 08:08:37 2024 -0800" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Mon Feb 12 23:03:19 2024 +0100" }, "message": "MIPS: Add \u0027memory\u0027 clobber to csum_ipv6_magic() inline assembler\n\nAfter \u0027lib: checksum: Use aligned accesses for ip_fast_csum and\ncsum_ipv6_magic tests\u0027 was applied, the test_csum_ipv6_magic unit test\nstarted failing for all mips platforms, both little and bit endian.\nOddly enough, adding debug code into test_csum_ipv6_magic() made the\nproblem disappear.\n\nThe gcc manual says:\n\n\"The \"memory\" clobber tells the compiler that the assembly code performs\n memory reads or writes to items other than those listed in the input\n and output operands (for example, accessing the memory pointed to by one\n of the input parameters)\n\"\n\nThis is definitely the case for csum_ipv6_magic(). Indeed, adding the\n\u0027memory\u0027 clobber fixes the problem.\n\nCc: Charlie Jenkins \u003ccharlie@rivosinc.com\u003e\nCc: Palmer Dabbelt \u003cpalmer@rivosinc.com\u003e\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nReviewed-by: Charlie Jenkins \u003ccharlie@rivosinc.com\u003e\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\n" }, { "commit": "716f4aaa7b48a55c73d632d0657b35342b1fefd7", "tree": "575f57874f065abc388dd26eec652e0caabd648f", "parents": [ "841c35169323cd833294798e58b9bf63fa4fa1de", "46f5ab762d048dad224436978315cbc2fa79c630" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Feb 12 07:15:45 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Feb 12 07:15:45 2024 -0800" }, "message": "Merge tag \u0027vfs-6.8-rc5.fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs\n\nPull vfs fixes from Christian Brauner:\n\n - Fix performance regression introduced by moving the security\n permission hook out of do_clone_file_range() and into its caller\n vfs_clone_file_range().\n\n This causes the security hook to be called in situation were it\n wasn\u0027t called before as the fast permission checks were left in\n do_clone_file_range().\n\n Fix this by merging the two implementations back together and\n restoring the old ordering: fast permission checks first, expensive\n ones later.\n\n - Tweak mount_setattr() permission checking so that mount properties on\n the real rootfs can be changed.\n\n When we added mount_setattr() we added additional checks compared to\n legacy mount(2). If the mount had a parent then verify that the\n caller and the mount namespace the mount is attached to match and if\n not make sure that it\u0027s an anonymous mount.\n\n But the real rootfs falls into neither category. It is neither an\n anoymous mount because it is obviously attached to the initial mount\n namespace but it also obviously doesn\u0027t have a parent mount. So that\n means legacy mount(2) allows changing mount properties on the real\n rootfs but mount_setattr(2) blocks this. This causes regressions (See\n the commit for details).\n\n Fix this by relaxing the check. If the mount has a parent or if it\n isn\u0027t a detached mount, verify that the mount namespaces of the\n caller and the mount are the same. Technically, we could probably\n write this even simpler and check that the mount namespaces match if\n it isn\u0027t a detached mount. But the slightly longer check makes it\n clearer what conditions one needs to think about.\n\n* tag \u0027vfs-6.8-rc5.fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:\n fs: relax mount_setattr() permission checks\n remap_range: merge do_clone_file_range() into vfs_clone_file_range()\n" }, { "commit": "b5f319360371087d52070d8f3fc7789e80ce69a6", "tree": "8b690c0cef849975dc2807a5aab472fd33b14c07", "parents": [ "14f08c976ffe0d2117c6199c32663df1cbc45c65" ], "author": { "name": "John Kacur", "email": "jkacur@redhat.com", "time": "Fri Feb 02 19:16:07 2024 -0500" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 10:59:09 2024 +0100" }, "message": "tools/rtla: Exit with EXIT_SUCCESS when help is invoked\n\nFix rtla so that the following commands exit with 0 when help is invoked\n\nrtla osnoise top -h\nrtla osnoise hist -h\nrtla timerlat top -h\nrtla timerlat hist -h\n\nLink: https://lore.kernel.org/linux-trace-devel/20240203001607.69703-1-jkacur@redhat.com\n\nCc: stable@vger.kernel.org\nFixes: 1eeb6328e8b3 (\"rtla/timerlat: Add timerlat hist mode\")\nSigned-off-by: John Kacur \u003cjkacur@redhat.com\u003e\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "14f08c976ffe0d2117c6199c32663df1cbc45c65", "tree": "fa6ad468768f163cf19bf1201a5e0c3a09de679c", "parents": [ "61ec586bc0815959d3314cf7ce242529c977b357" ], "author": { "name": "limingming3", "email": "limingming890315@gmail.com", "time": "Wed Feb 07 14:51:42 2024 +0800" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 10:56:36 2024 +0100" }, "message": "tools/rtla: Replace setting prio with nice for SCHED_OTHER\n\nSince the sched_priority for SCHED_OTHER is always 0, it makes no\nsence to set it.\nSetting nice for SCHED_OTHER seems more meaningful.\n\nLink: https://lkml.kernel.org/r/20240207065142.1753909-1-limingming3@lixiang.com\n\nCc: stable@vger.kernel.org\nFixes: b1696371d865 (\"rtla: Helper functions for rtla\")\nSigned-off-by: limingming3 \u003climingming3@lixiang.com\u003e\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "61ec586bc0815959d3314cf7ce242529c977b357", "tree": "afaa5693eb726a5736329343e642aa45113b342e", "parents": [ "f9b2c87105c989a7b259c6da87673ada96dce2f8" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:34 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:58:36 2024 +0100" }, "message": "tools/rv: Fix curr_reactor uninitialized variable\n\nclang is reporting:\n\n$ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n\nclang -O -g -DVERSION\u003d\\\"6.8.0-rc3\\\" -flto\u003dauto -fexceptions\n\t-fstack-protector-strong -fasynchronous-unwind-tables\n\t-fstack-clash-protection -Wall -Werror\u003dformat-security\n\t-Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS\n\t$(pkg-config --cflags libtracefs) -I include\n\t-c -o src/in_kernel.o src/in_kernel.c\n[...]\n\nsrc/in_kernel.c:227:6: warning: variable \u0027curr_reactor\u0027 is used uninitialized whenever \u0027if\u0027 condition is true [-Wsometimes-uninitialized]\n 227 | if (!end)\n | ^~~~\nsrc/in_kernel.c:242:9: note: uninitialized use occurs here\n 242 | return curr_reactor;\n | ^~~~~~~~~~~~\nsrc/in_kernel.c:227:2: note: remove the \u0027if\u0027 if its condition is always false\n 227 | if (!end)\n | ^~~~~~~~~\n 228 | goto out_free;\n | ~~~~~~~~~~~~~\nsrc/in_kernel.c:221:6: warning: variable \u0027curr_reactor\u0027 is used uninitialized whenever \u0027if\u0027 condition is true [-Wsometimes-uninitialized]\n 221 | if (!start)\n | ^~~~~~\nsrc/in_kernel.c:242:9: note: uninitialized use occurs here\n 242 | return curr_reactor;\n | ^~~~~~~~~~~~\nsrc/in_kernel.c:221:2: note: remove the \u0027if\u0027 if its condition is always false\n 221 | if (!start)\n | ^~~~~~~~~~~\n 222 | goto out_free;\n | ~~~~~~~~~~~~~\nsrc/in_kernel.c:215:20: note: initialize the variable \u0027curr_reactor\u0027 to silence this warning\n 215 | char *curr_reactor;\n | ^\n | \u003d NULL\n2 warnings generated.\n\nWhich is correct. Setting curr_reactor to NULL avoids the problem.\n\nLink: https://lkml.kernel.org/r/3a35551149e5ee0cb0950035afcb8082c3b5d05b.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nCc: Donald Zickus \u003cdzickus@redhat.com\u003e\nFixes: 6d60f89691fc (\"tools/rv: Add in-kernel monitor interface\")\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "f9b2c87105c989a7b259c6da87673ada96dce2f8", "tree": "78c01cff8b4a22581b18cf900736d104b0363a54", "parents": [ "084ce16df0f060efd371092a09a7ae74a536dc11" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:33 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:58:08 2024 +0100" }, "message": "tools/rv: Fix Makefile compiler options for clang\n\nThe following errors are showing up when compiling rv with clang:\n\n $ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n [...]\n clang -O -g -DVERSION\u003d\\\"6.8.0-rc1\\\" -flto\u003dauto -ffat-lto-objects\n -fexceptions -fstack-protector-strong -fasynchronous-unwind-tables\n -fstack-clash-protection -Wall -Werror\u003dformat-security\n -Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS\n -Wno-maybe-uninitialized $(pkg-config --cflags libtracefs)\n -I include -c -o src/utils.o src/utils.c\n clang: warning: optimization flag \u0027-ffat-lto-objects\u0027 is not supported [-Wignored-optimization-argument]\n warning: unknown warning option \u0027-Wno-maybe-uninitialized\u0027; did you mean \u0027-Wno-uninitialized\u0027? [-Wunknown-warning-option]\n 1 warning generated.\n\n clang -o rv -ggdb src/in_kernel.o src/rv.o src/trace.o src/utils.o $(pkg-config --libs libtracefs)\n src/in_kernel.o: file not recognized: file format not recognized\n clang: error: linker command failed with exit code 1 (use -v to see invocation)\n make: *** [Makefile:110: rv] Error 1\n\nSolve these issues by:\n - removing -ffat-lto-objects and -Wno-maybe-uninitialized if using clang\n - informing the linker about -flto\u003dauto\n\nLink: https://lkml.kernel.org/r/ed94a8ddc2ca8c8ef663cfb7ae9dd196c4a66b33.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nFixes: 4bc4b131d44c (\"rv: Add rv tool\")\nSuggested-by: Donald Zickus \u003cdzickus@redhat.com\u003e\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "084ce16df0f060efd371092a09a7ae74a536dc11", "tree": "d97dd432faefcce3f9f6c643db2f4baedacbd2fe", "parents": [ "30369084ac6e27479a347899e74f523e6ca29b89" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:32 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:57:37 2024 +0100" }, "message": "tools/rtla: Remove unused sched_getattr() function\n\nClang is reporting:\n\n$ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n[...]\nclang -O -g -DVERSION\u003d\\\"6.8.0-rc3\\\" -flto\u003dauto -fexceptions -fstack-protector-strong -fasynchronous-unwind-tables -fstack-clash-protection -Wall -Werror\u003dformat-security -Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS $(pkg-config --cflags libtracefs) -c -o src/utils.o src/utils.c\nsrc/utils.c:241:19: warning: unused function \u0027sched_getattr\u0027 [-Wunused-function]\n 241 | static inline int sched_getattr(pid_t pid, struct sched_attr *attr,\n | ^~~~~~~~~~~~~\n1 warning generated.\n\nWhich is correct, so remove the unused function.\n\nLink: https://lkml.kernel.org/r/eaed7ba122c4ae88ce71277c824ef41cbf789385.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nCc: Donald Zickus \u003cdzickus@redhat.com\u003e\nFixes: b1696371d865 (\"rtla: Helper functions for rtla\")\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "30369084ac6e27479a347899e74f523e6ca29b89", "tree": "29e3d4ead88920a1087590830a51a8cc5e780277", "parents": [ "64dc40f7523369912d7adb22c8cb655f71610505" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:31 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:57:14 2024 +0100" }, "message": "tools/rtla: Fix clang warning about mount_point var size\n\nclang is reporting this warning:\n\n$ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n[...]\nclang -O -g -DVERSION\u003d\\\"6.8.0-rc3\\\" -flto\u003dauto -fexceptions\n\t-fstack-protector-strong -fasynchronous-unwind-tables\n\t-fstack-clash-protection -Wall -Werror\u003dformat-security\n\t-Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS\n\t$(pkg-config --cflags libtracefs) -c -o src/utils.o src/utils.c\n\nsrc/utils.c:548:66: warning: \u0027fscanf\u0027 may overflow; destination buffer in argument 3 has size 1024, but the corresponding specifier may require size 1025 [-Wfortify-source]\n 548 | while (fscanf(fp, \"%*s %\" STR(MAX_PATH) \"s %99s %*s %*d %*d\\n\", mount_point, type) \u003d\u003d 2) {\n | ^\n\nIncrease mount_point variable size to MAX_PATH+1 to avoid the overflow.\n\nLink: https://lkml.kernel.org/r/1b46712e93a2f4153909514a36016959dcc4021c.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nCc: Donald Zickus \u003cdzickus@redhat.com\u003e\nFixes: a957cbc02531 (\"rtla: Add -C cgroup support\")\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "64dc40f7523369912d7adb22c8cb655f71610505", "tree": "cc0acdff7bf64685fab4cd67a87901c758299684", "parents": [ "bc4cbc9d260ba8358ca63662919f4bb223cb603b" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:30 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:56:22 2024 +0100" }, "message": "tools/rtla: Fix uninitialized bucket/data-\u003ebucket_size warning\n\nWhen compiling rtla with clang, I am getting the following warnings:\n\n$ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n\n[..]\nclang -O -g -DVERSION\u003d\\\"6.8.0-rc3\\\" -flto\u003dauto -fexceptions\n\t-fstack-protector-strong -fasynchronous-unwind-tables\n\t-fstack-clash-protection -Wall -Werror\u003dformat-security\n\t-Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS\n\t$(pkg-config --cflags libtracefs)\n\t-c -o src/osnoise_hist.o src/osnoise_hist.c\nsrc/osnoise_hist.c:138:6: warning: variable \u0027bucket\u0027 is used uninitialized whenever \u0027if\u0027 condition is false [-Wsometimes-uninitialized]\n 138 | if (data-\u003ebucket_size)\n | ^~~~~~~~~~~~~~~~~\nsrc/osnoise_hist.c:149:6: note: uninitialized use occurs here\n 149 | if (bucket \u003c entries)\n | ^~~~~~\nsrc/osnoise_hist.c:138:2: note: remove the \u0027if\u0027 if its condition is always true\n 138 | if (data-\u003ebucket_size)\n | ^~~~~~~~~~~~~~~~~~~~~~\n 139 | bucket \u003d duration / data-\u003ebucket_size;\nsrc/osnoise_hist.c:132:12: note: initialize the variable \u0027bucket\u0027 to silence this warning\n 132 | int bucket;\n | ^\n | \u003d 0\n1 warning generated.\n\n[...]\n\nclang -O -g -DVERSION\u003d\\\"6.8.0-rc3\\\" -flto\u003dauto -fexceptions\n\t-fstack-protector-strong -fasynchronous-unwind-tables\n\t-fstack-clash-protection -Wall -Werror\u003dformat-security\n\t-Wp,-D_FORTIFY_SOURCE\u003d2 -Wp,-D_GLIBCXX_ASSERTIONS\n\t$(pkg-config --cflags libtracefs)\n\t-c -o src/timerlat_hist.o src/timerlat_hist.c\nsrc/timerlat_hist.c:181:6: warning: variable \u0027bucket\u0027 is used uninitialized whenever \u0027if\u0027 condition is false [-Wsometimes-uninitialized]\n 181 | if (data-\u003ebucket_size)\n | ^~~~~~~~~~~~~~~~~\nsrc/timerlat_hist.c:204:6: note: uninitialized use occurs here\n 204 | if (bucket \u003c entries)\n | ^~~~~~\nsrc/timerlat_hist.c:181:2: note: remove the \u0027if\u0027 if its condition is always true\n 181 | if (data-\u003ebucket_size)\n | ^~~~~~~~~~~~~~~~~~~~~~\n 182 | bucket \u003d latency / data-\u003ebucket_size;\nsrc/timerlat_hist.c:175:12: note: initialize the variable \u0027bucket\u0027 to silence this warning\n 175 | int bucket;\n | ^\n | \u003d 0\n1 warning generated.\n\nThis is a legit warning, but data-\u003ebucket_size is always \u003e 0 (see\ntimerlat_hist_parse_args()), so the if is not necessary.\n\nRemove the unneeded if (data-\u003ebucket_size) to avoid the warning.\n\nLink: https://lkml.kernel.org/r/6e1b1665cd99042ae705b3e0fc410858c4c42346.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nCc: Donald Zickus \u003cdzickus@redhat.com\u003e\nFixes: 1eeb6328e8b3 (\"rtla/timerlat: Add timerlat hist mode\")\nFixes: 829a6c0b5698 (\"rtla/osnoise: Add the hist mode\")\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "bc4cbc9d260ba8358ca63662919f4bb223cb603b", "tree": "4592d12bcbfbcea910d90368cee9bd8f9ccf8ee5", "parents": [ "841c35169323cd833294798e58b9bf63fa4fa1de" ], "author": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Tue Feb 06 12:05:29 2024 +0100" }, "committer": { "name": "Daniel Bristot de Oliveira", "email": "bristot@kernel.org", "time": "Mon Feb 12 09:49:13 2024 +0100" }, "message": "tools/rtla: Fix Makefile compiler options for clang\n\nThe following errors are showing up when compiling rtla with clang:\n\n $ make HOSTCC\u003dclang CC\u003dclang LLVM_IAS\u003d1\n [...]\n\n clang -O -g -DVERSION\u003d\\\"6.8.0-rc1\\\" -flto\u003dauto -ffat-lto-objects\n\t-fexceptions -fstack-protector-strong\n\t-fasynchronous-unwind-tables -fstack-clash-protection -Wall\n\t-Werror\u003dformat-security -Wp,-D_FORTIFY_SOURCE\u003d2\n\t-Wp,-D_GLIBCXX_ASSERTIONS -Wno-maybe-uninitialized\n\t$(pkg-config --cflags libtracefs) -c -o src/utils.o src/utils.c\n\n clang: warning: optimization flag \u0027-ffat-lto-objects\u0027 is not supported [-Wignored-optimization-argument]\n warning: unknown warning option \u0027-Wno-maybe-uninitialized\u0027; did you mean \u0027-Wno-uninitialized\u0027? [-Wunknown-warning-option]\n 1 warning generated.\n\n clang -o rtla -ggdb src/osnoise.o src/osnoise_hist.o src/osnoise_top.o\n src/rtla.o src/timerlat_aa.o src/timerlat.o src/timerlat_hist.o\n src/timerlat_top.o src/timerlat_u.o src/trace.o src/utils.o $(pkg-config --libs libtracefs)\n\n src/osnoise.o: file not recognized: file format not recognized\n clang: error: linker command failed with exit code 1 (use -v to see invocation)\n make: *** [Makefile:110: rtla] Error 1\n\nSolve these issues by:\n - removing -ffat-lto-objects and -Wno-maybe-uninitialized if using clang\n - informing the linker about -flto\u003dauto\n\nLink: https://lore.kernel.org/linux-trace-kernel/567ac1b94effc228ce9a0225b9df7232a9b35b55.1707217097.git.bristot@kernel.org\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Bill Wendling \u003cmorbo@google.com\u003e\nCc: Justin Stitt \u003cjustinstitt@google.com\u003e\nFixes: 1a7b22ab15eb (\"tools/rtla: Build with EXTRA_{C,LD}FLAGS\")\nSuggested-by: Donald Zickus \u003cdzickus@redhat.com\u003e\nSigned-off-by: Daniel Bristot de Oliveira \u003cbristot@kernel.org\u003e\n" }, { "commit": "841c35169323cd833294798e58b9bf63fa4fa1de", "tree": "75d93259d492e4776a766abd5c2264deeca12815", "parents": [ "2766f59ca44e517a1d6226979c784b026f0e89c2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 12:18:13 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 12:18:13 2024 -0800" }, "message": "Linux 6.8-rc4\n" }, { "commit": "2766f59ca44e517a1d6226979c784b026f0e89c2", "tree": "56860cd5dceff62ed6c625c49707b16704fb5af1", "parents": [ "c021e191cf3da2fea766d784546f4c5acbfcd137", "dad6a09f3148257ac1773cd90934d721d68ab595" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 11:44:14 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 11:44:14 2024 -0800" }, "message": "Merge tag \u0027timers_urgent_for_v6.8_rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull timer fix from Borislav Petkov:\n\n - Make sure a warning is issued when a hrtimer gets queued after the\n timers have been migrated on the CPU down path and thus said timer\n will get ignored\n\n* tag \u0027timers_urgent_for_v6.8_rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n hrtimer: Report offline hrtimer enqueue\n" }, { "commit": "c021e191cf3da2fea766d784546f4c5acbfcd137", "tree": "e6dcf7a55d8fa0fc9010ba8f0e3c6fce24015c71", "parents": [ "7521f258ea303c827434c101884b62a2b137a942", "f6a1892585cd19e63c4ef2334e26cd536d5b678d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 11:41:51 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 11 11:41:51 2024 -0800" }, "message": "Merge tag \u0027x86_urgent_for_v6.8_rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Borislav Petkov:\n\n - Correct the minimum CPU family for Transmeta Crusoe in Kconfig so\n that such hw can boot again\n\n - Do not take into accout XSTATE buffer size info supplied by userspace\n when constructing a sigreturn frame\n\n - Switch get_/put_user* to EX_TYPE_UACCESS exception handling when an\n MCE is encountered so that it can be properly recovered from instead\n of simply panicking\n\n* tag \u0027x86_urgent_for_v6.8_rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/Kconfig: Transmeta Crusoe is CPU family 5, not 6\n x86/fpu: Stop relying on userspace for info to fault in xsave buffer\n x86/lib: Revert to _ASM_EXTABLE_UA() for {get,put}_user() fixups\n" }, { "commit": "7521f258ea303c827434c101884b62a2b137a942", "tree": "e4c054ffaa11dce0476352e75273e5e1db9127b0", "parents": [ "a5b6244cf87c50358f5562b8f07f7ac35fc7f6b0", "5bc09b397cbf1221f8a8aacb1152650c9195b02b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 15:28:07 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 15:28:07 2024 -0800" }, "message": "Merge tag \u0027mm-hotfixes-stable-2024-02-10-11-16\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm\n\nPull misc fixes from Andrew Morton:\n \"21 hotfixes. 12 are cc:stable and the remainder pertain to post-6.7\n issues or aren\u0027t considered to be needed in earlier kernel versions\"\n\n* tag \u0027mm-hotfixes-stable-2024-02-10-11-16\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)\n nilfs2: fix potential bug in end_buffer_async_write\n mm/damon/sysfs-schemes: fix wrong DAMOS tried regions update timeout setup\n nilfs2: fix hang in nilfs_lookup_dirty_data_buffers()\n MAINTAINERS: Leo Yan has moved\n mm/zswap: don\u0027t return LRU_SKIP if we have dropped lru lock\n fs,hugetlb: fix NULL pointer dereference in hugetlbs_fill_super\n mailmap: switch email address for John Moon\n mm: zswap: fix objcg use-after-free in entry destruction\n mm/madvise: don\u0027t forget to leave lazy MMU mode in madvise_cold_or_pageout_pte_range()\n arch/arm/mm: fix major fault accounting when retrying under per-VMA lock\n selftests: core: include linux/close_range.h for CLOSE_RANGE_* macros\n mm/memory-failure: fix crash in split_huge_page_to_list from soft_offline_page\n mm: memcg: optimize parent iteration in memcg_rstat_updated()\n nilfs2: fix data corruption in dsync block recovery for small block sizes\n mm/userfaultfd: UFFDIO_MOVE implementation should use ptep_get()\n exit: wait_task_zombie: kill the no longer necessary spin_lock_irq(siglock)\n fs/proc: do_task_stat: use sig-\u003estats_lock to gather the threads/children stats\n fs/proc: do_task_stat: move thread_group_cputime_adjusted() outside of lock_task_sighand()\n getrusage: use sig-\u003estats_lock rather than lock_task_sighand()\n getrusage: move thread_group_cputime_adjusted() outside of lock_task_sighand()\n ...\n" }, { "commit": "a5b6244cf87c50358f5562b8f07f7ac35fc7f6b0", "tree": "57496f15a09a3aacf2d4e99822e7e73ed64a0927", "parents": [ "a38ff5bbf9799dd944c762412aaebc54818407f5", "5f63a493b99c848ad5200402bebe26211e00025a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 08:02:48 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 08:02:48 2024 -0800" }, "message": "Merge tag \u0027block-6.8-2024-02-10\u0027 of git://git.kernel.dk/linux\n\nPull block fixes from Jens Axboe:\n\n - NVMe pull request via Keith:\n - Update a potentially stale firmware attribute (Maurizio)\n - Fixes for the recent verbose error logging (Keith, Chaitanya)\n - Protection information payload size fix for passthrough (Francis)\n\n - Fix for a queue freezing issue in virtblk (Yi)\n\n - blk-iocost underflow fix (Tejun)\n\n - blk-wbt task detection fix (Jan)\n\n* tag \u0027block-6.8-2024-02-10\u0027 of git://git.kernel.dk/linux:\n virtio-blk: Ensure no requests in virtqueues before deleting vqs.\n blk-iocost: Fix an UBSAN shift-out-of-bounds warning\n nvme: use ns-\u003ehead-\u003epi_size instead of t10_pi_tuple structure size\n nvme-core: fix comment to reflect right functions\n nvme: move passthrough logging attribute to head\n blk-wbt: Fix detection of dirty-throttled tasks\n nvme-host: fix the updating of the firmware version\n" }, { "commit": "a38ff5bbf9799dd944c762412aaebc54818407f5", "tree": "4a89dae3d71435e593fac78ea3e2f80900c7ac08", "parents": [ "5a7ec87063c0fbb7706fdccb0cc890757da6f4a1", "7ed4380009e96d9e9c605e12822e987b35b05648" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 07:56:39 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 07:56:39 2024 -0800" }, "message": "Merge tag \u0027firewire-fixes-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394\n\nPull firewire fix from Takashi Sakamoto:\n \"A change to accelerate the device detection step in some cases.\n\n In the self-identification step after bus-reset, all nodes in the same\n bus broadcast selfID packet including the value of gap count. The\n value is related to the cable hops between nodes, and used to\n calculate the subaction gap and the arbitration reset gap.\n\n When each node has the different value of the gap count, the\n asynchronous communication between them is unreliable, since an\n asynchronous transaction could be interrupted by another asynchronous\n transaction before completion. The gap count inconsistency can be\n resolved by several ways; e.g. the transfer of PHY configuration\n packet and generation of bus-reset.\n\n The current implementation of firewire stack can correctly detect the\n gap count inconsistency, however the recovery action from the\n inconsistency tends to be delayed after reading configuration ROM of\n root node. This results in the long time to probe devices in some\n combinations of hardware.\n\n Here the stack is changed to schedule the action as soon as possible\"\n\n* tag \u0027firewire-fixes-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394:\n firewire: core: send bus reset promptly on gap count error\n" }, { "commit": "5a7ec87063c0fbb7706fdccb0cc890757da6f4a1", "tree": "9518c968509f2a630941f0d9ecb2000ab476c866", "parents": [ "4a7bbe7519b6a5d189feeba1d417c8ce9f7e852b", "108a020c64434fed4b69762879d78cd24088b4c7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 07:53:41 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 10 07:53:41 2024 -0800" }, "message": "Merge tag \u00276.8-rc3-ksmbd-server-fixes\u0027 of git://git.samba.org/ksmbd\n\nPull smb server fixes from Steve French:\n \"Two ksmbd server fixes:\n\n - memory leak fix\n\n - a minor kernel-doc fix\"\n\n* tag \u00276.8-rc3-ksmbd-server-fixes\u0027 of git://git.samba.org/ksmbd:\n ksmbd: free aux buffer if ksmbd_iov_pin_rsp_read fails\n ksmbd: Add kernel-doc for ksmbd_extract_sharename() function\n" }, { "commit": "4a7bbe7519b6a5d189feeba1d417c8ce9f7e852b", "tree": "8d9d23f24baef8b3cf6c114feaea89592eae441a", "parents": [ "ca00c700c521916bacf7d490a2a5fd908059a927", "17e94b2585417e04dabc2f13bc03b4665ae687f3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:15:26 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:15:26 2024 -0800" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fixes from James Bottomley:\n \"Three small driver fixes and one core fix.\n\n The core fix being a fixup to the one in the last pull request which\n didn\u0027t entirely move checking of scsi_host_busy() out from under the\n host lock\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: ufs: core: Remove the ufshcd_release() in ufshcd_err_handling_prepare()\n scsi: ufs: core: Fix shift issue in ufshcd_clear_cmd()\n scsi: lpfc: Use unsigned type for num_sge\n scsi: core: Move scsi_host_busy() out of host lock if it is for per-command\n" }, { "commit": "ca00c700c521916bacf7d490a2a5fd908059a927", "tree": "e7225332ce3a2c38469a3ec4836fcad090335356", "parents": [ "e1e3f530a17f808c70f506f11ba7dabbfd8eb14a", "a5cc98eba2592d6e3c5a4351319595ddde2a5901" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:09:30 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:09:30 2024 -0800" }, "message": "Merge tag \u00276.8-rc3-smb3-client-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6\n\nPull smb client fixes from Steve French:\n\n - reconnect fix\n\n - multichannel channel selection fix\n\n - minor mount warning fix\n\n - reparse point fix\n\n - null pointer check improvement\n\n* tag \u00276.8-rc3-smb3-client-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6:\n smb3: clarify mount warning\n cifs: handle cases where multiple sessions share connection\n cifs: change tcon status when need_reconnect is set on it\n smb: client: set correct d_type for reparse points under DFS mounts\n smb3: add missing null server pointer check\n" }, { "commit": "e1e3f530a17f808c70f506f11ba7dabbfd8eb14a", "tree": "8212b45eafd77aaa48824523b8a0586aa150991c", "parents": [ "a2343df3fb1e62ef7c5c0a65c06621980f257d5b", "07045648c07c5632e0dfd5ce084d3cd0cec0258a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:05:02 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 17:05:02 2024 -0800" }, "message": "Merge tag \u0027ceph-for-6.8-rc4\u0027 of https://github.com/ceph/ceph-client\n\nPull ceph fixes from Ilya Dryomov:\n \"Some fscrypt-related fixups (sparse reads are used only for encrypted\n files) and two cap handling fixes from Xiubo and Rishabh\"\n\n* tag \u0027ceph-for-6.8-rc4\u0027 of https://github.com/ceph/ceph-client:\n ceph: always check dir caps asynchronously\n ceph: prevent use-after-free in encode_cap_msg()\n ceph: always set initial i_blkbits to CEPH_FSCRYPT_BLOCK_SHIFT\n libceph: just wait for more data to be available on the socket\n libceph: rename read_sparse_msg_*() to read_partial_sparse_msg_*()\n libceph: fail sparse-read if the data length doesn\u0027t match\n" }, { "commit": "a2343df3fb1e62ef7c5c0a65c06621980f257d5b", "tree": "936fc0aba7daecc771cbbcd494a043678c97b610", "parents": [ "4356e9f841f7fbb945521cef3577ba394c65f3fc", "622cd3daa8eae37359a6fd3c07c36d19f66606b5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 16:59:49 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 16:59:49 2024 -0800" }, "message": "Merge tag \u0027ntfs3_for_6.8\u0027 of https://github.com/Paragon-Software-Group/linux-ntfs3\n\nPull ntfs3 fixes from Konstantin Komarov:\n \"Fixed:\n - size update for compressed file\n - some logic errors, overflows\n - memory leak\n - some code was refactored\n\n Added:\n - implement super_operations::shutdown\n\n Improved:\n - alternative boot processing\n - reduced stack usage\"\n\n* tag \u0027ntfs3_for_6.8\u0027 of https://github.com/Paragon-Software-Group/linux-ntfs3: (28 commits)\n fs/ntfs3: Slightly simplify ntfs_inode_printk()\n fs/ntfs3: Add ioctl operation for directories (FITRIM)\n fs/ntfs3: Fix oob in ntfs_listxattr\n fs/ntfs3: Fix an NULL dereference bug\n fs/ntfs3: Update inode-\u003ei_size after success write into compressed file\n fs/ntfs3: Fixed overflow check in mi_enum_attr()\n fs/ntfs3: Correct function is_rst_area_valid\n fs/ntfs3: Use i_size_read and i_size_write\n fs/ntfs3: Prevent generic message \"attempt to access beyond end of device\"\n fs/ntfs3: use non-movable memory for ntfs3 MFT buffer cache\n fs/ntfs3: Use kvfree to free memory allocated by kvmalloc\n fs/ntfs3: Disable ATTR_LIST_ENTRY size check\n fs/ntfs3: Fix c/mtime typo\n fs/ntfs3: Add NULL ptr dereference checking at the end of attr_allocate_frame()\n fs/ntfs3: Add and fix comments\n fs/ntfs3: ntfs3_forced_shutdown use int instead of bool\n fs/ntfs3: Implement super_operations::shutdown\n fs/ntfs3: Drop suid and sgid bits as a part of fpunch\n fs/ntfs3: Add file_modified\n fs/ntfs3: Correct use bh_read\n ...\n" }, { "commit": "4356e9f841f7fbb945521cef3577ba394c65f3fc", "tree": "87bb7c1de6e813f9d7e9e57bb54c9116d3f83137", "parents": [ "9ed18b0b7765ec3aa5f10843d1c3daaa1a225337" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 12:39:31 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 15:57:48 2024 -0800" }, "message": "work around gcc bugs with \u0027asm goto\u0027 with outputs\n\nWe\u0027ve had issues with gcc and \u0027asm goto\u0027 before, and we created a\n\u0027asm_volatile_goto()\u0027 macro for that in the past: see commits\n3f0116c3238a (\"compiler/gcc4: Add quirk for \u0027asm goto\u0027 miscompilation\nbug\") and a9f180345f53 (\"compiler/gcc4: Make quirk for\nasm_volatile_goto() unconditional\").\n\nThen, much later, we ended up removing the workaround in commit\n43c249ea0b1e (\"compiler-gcc.h: remove ancient workaround for gcc PR\n58670\") because we no longer supported building the kernel with the\naffected gcc versions, but we left the macro uses around.\n\nNow, Sean Christopherson reports a new version of a very similar\nproblem, which is fixed by re-applying that ancient workaround. But the\nproblem in question is limited to only the \u0027asm goto with outputs\u0027\ncases, so instead of re-introducing the old workaround as-is, let\u0027s\nrename and limit the workaround to just that much less common case.\n\nIt looks like there are at least two separate issues that all hit in\nthis area:\n\n (a) some versions of gcc don\u0027t mark the asm goto as \u0027volatile\u0027 when it\n has outputs:\n\n https://gcc.gnu.org/bugzilla/show_bug.cgi?id\u003d98619\n https://gcc.gnu.org/bugzilla/show_bug.cgi?id\u003d110420\n\n which is easy to work around by just adding the \u0027volatile\u0027 by hand.\n\n (b) Internal compiler errors:\n\n https://gcc.gnu.org/bugzilla/show_bug.cgi?id\u003d110422\n\n which are worked around by adding the extra empty \u0027asm\u0027 as a\n barrier, as in the original workaround.\n\nbut the problem Sean sees may be a third thing since it involves bad\ncode generation (not an ICE) even with the manually added \u0027volatile\u0027.\n\nbut the same old workaround works for this case, even if this feels a\nbit like voodoo programming and may only be hiding the issue.\n\nReported-and-tested-by: Sean Christopherson \u003cseanjc@google.com\u003e\nLink: https://lore.kernel.org/all/20240208220604.140859-1-seanjc@google.com/\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Uros Bizjak \u003cubizjak@gmail.com\u003e\nCc: Jakub Jelinek \u003cjakub@redhat.com\u003e\nCc: Andrew Pinski \u003cquic_apinski@quicinc.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "a5cc98eba2592d6e3c5a4351319595ddde2a5901", "tree": "77a022edba3c72cd2b9f06532ccb3685080608a6", "parents": [ "a39c757bf0596b17482a507f31c3ef0af0d1d2b4" ], "author": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Tue Feb 06 23:57:18 2024 -0600" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Feb 09 14:43:27 2024 -0600" }, "message": "smb3: clarify mount warning\n\nWhen a user tries to use the \"sec\u003dkrb5p\" mount parameter to encrypt\ndata on connection to a server (when authenticating with Kerberos), we\nindicate that it is not supported, but do not note the equivalent\nrecommended mount parameter (\"sec\u003dkrb5,seal\") which turns on encryption\nfor that mount (and uses Kerberos for auth). Update the warning message.\n\nReviewed-by: Shyam Prasad N \u003csprasad@microsoft.com\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "a39c757bf0596b17482a507f31c3ef0af0d1d2b4", "tree": "01e463edb4a322bb3cc05be5a30c95a48fa379ea", "parents": [ "c6e02eefd6ace3da3369c764f15429f5647056af" ], "author": { "name": "Shyam Prasad N", "email": "sprasad@microsoft.com", "time": "Tue Feb 06 15:00:47 2024 +0000" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Feb 09 14:43:25 2024 -0600" }, "message": "cifs: handle cases where multiple sessions share connection\n\nBased on our implementation of multichannel, it is entirely\npossible that a server struct may not be found in any channel\nof an SMB session.\n\nIn such cases, we should be prepared to move on and search for\nthe server struct in the next session.\n\nSigned-off-by: Shyam Prasad N \u003csprasad@microsoft.com\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "c6e02eefd6ace3da3369c764f15429f5647056af", "tree": "acbd8039c464288f46b498353727f0900435fbed", "parents": [ "55c7788c37242702868bfac7861cdf0c358d6c3d" ], "author": { "name": "Shyam Prasad N", "email": "sprasad@microsoft.com", "time": "Tue Feb 06 15:00:46 2024 +0000" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Feb 09 14:43:23 2024 -0600" }, "message": "cifs: change tcon status when need_reconnect is set on it\n\nWhen a tcon is marked for need_reconnect, the intention\nis to have it reconnected.\n\nThis change adjusts tcon-\u003estatus in cifs_tree_connect\nwhen need_reconnect is set. Also, this change has a minor\ncorrection in resetting need_reconnect on success. It makes\nsure that it is done with tc_lock held.\n\nSigned-off-by: Shyam Prasad N \u003csprasad@microsoft.com\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "12c5128f101bfa47a08e4c0e1a75cfa2d0872bcd", "tree": "730afd9b65e859d2d765b58ab7474d062f5f4e63", "parents": [ "f4a9f219411f318ae60d6ff7f129082a75686c6c" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu Jan 25 09:53:19 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Feb 09 20:29:22 2024 +0100" }, "message": "btrfs: add new unused block groups to the list of unused block groups\n\nSpace reservations for metadata are, most of the time, pessimistic as we\nreserve space for worst possible cases - where tree heights are at the\nmaximum possible height (8), we need to COW every extent buffer in a tree\npath, need to split extent buffers, etc.\n\nFor data, we generally reserve the exact amount of space we are going to\nallocate. The exception here is when using compression, in which case we\nreserve space matching the uncompressed size, as the compression only\nhappens at writeback time and in the worst possible case we need that\namount of space in case the data is not compressible.\n\nThis means that when there\u0027s not available space in the corresponding\nspace_info object, we may need to allocate a new block group, and then\nthat block group might not be used after all. In this case the block\ngroup is never added to the list of unused block groups and ends up\nnever being deleted - except if we unmount and mount again the fs, as\nwhen reading block groups from disk we add unused ones to the list of\nunused block groups (fs_info-\u003eunused_bgs). Otherwise a block group is\nonly added to the list of unused block groups when we deallocate the\nlast extent from it, so if no extent is ever allocated, the block group\nis kept around forever.\n\nThis also means that if we have a bunch of tasks reserving space in\nparallel we can end up allocating many block groups that end up never\nbeing used or kept around for too long without being used, which has\nthe potential to result in ENOSPC failures in case for example we over\nallocate too many metadata block groups and then end up in a state\nwithout enough unallocated space to allocate a new data block group.\n\nThis is more likely to happen with metadata reservations as of kernel\n6.7, namely since commit 28270e25c69a (\"btrfs: always reserve space for\ndelayed refs when starting transaction\"), because we started to always\nreserve space for delayed references when starting a transaction handle\nfor a non-zero number of items, and also to try to reserve space to fill\nthe gap between the delayed block reserve\u0027s reserved space and its size.\n\nSo to avoid this, when finishing the creation a new block group, add the\nblock group to the list of unused block groups if it\u0027s still unused at\nthat time. This way the next time the cleaner kthread runs, it will delete\nthe block group if it\u0027s still unused and not needed to satisfy existing\nspace reservations.\n\nReported-by: Ivan Shapovalov \u003cintelfx@intelfx.name\u003e\nLink: https://lore.kernel.org/linux-btrfs/9cdbf0ca9cdda1b4c84e15e548af7d7f9f926382.camel@intelfx.name/\nCC: stable@vger.kernel.org # 6.7+\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nReviewed-by: Boris Burkov \u003cboris@bur.io\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "f4a9f219411f318ae60d6ff7f129082a75686c6c", "tree": "f2724a935e2b78b834d043d89b2cf26fbc129416", "parents": [ "1693d5442c458ae8d5b0d58463b873cd879569ed" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu Jan 25 09:53:14 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Feb 09 20:29:16 2024 +0100" }, "message": "btrfs: do not delete unused block group if it may be used soon\n\nBefore deleting a block group that is in the list of unused block groups\n(fs_info-\u003eunused_bgs), we check if the block group became used before\ndeleting it, as extents from it may have been allocated after it was added\nto the list.\n\nHowever even if the block group was not yet used, there may be tasks that\nhave only reserved space and have not yet allocated extents, and they\nmight be relying on the availability of the unused block group in order\nto allocate extents. The reservation works first by increasing the\n\"bytes_may_use\" field of the corresponding space_info object (which may\nfirst require flushing delayed items, allocating a new block group, etc),\nand only later a task does the actual allocation of extents.\n\nFor metadata we usually don\u0027t end up using all reserved space, as we are\npessimistic and typically account for the worst cases (need to COW every\nsingle node in a path of a tree at maximum possible height, etc). For\ndata we usually reserve the exact amount of space we\u0027re going to allocate\nlater, except when using compression where we always reserve space based\non the uncompressed size, as compression is only triggered when writeback\nstarts so we don\u0027t know in advance how much space we\u0027ll actually need, or\nif the data is compressible.\n\nSo don\u0027t delete an unused block group if the total size of its space_info\nobject minus the block group\u0027s size is less then the sum of used space and\nspace that may be used (space_info-\u003ebytes_may_use), as that means we have\ntasks that reserved space and may need to allocate extents from the block\ngroup. In this case, besides skipping the deletion, re-add the block group\nto the list of unused block groups so that it may be reconsidered later,\nin case the tasks that reserved space end up not needing to allocate\nextents from it.\n\nAllowing the deletion of the block group while we have reserved space, can\nresult in tasks failing to allocate metadata extents (-ENOSPC) while under\na transaction handle, resulting in a transaction abort, or failure during\nwriteback for the case of data extents.\n\nCC: stable@vger.kernel.org # 6.0+\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nReviewed-by: Boris Burkov \u003cboris@bur.io\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: David Sterba \u003cdsterba@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "1693d5442c458ae8d5b0d58463b873cd879569ed", "tree": "a1730d989c2cb5c86f0ae9e144ac1dc5b5dfe8fe", "parents": [ "5571e41ec6e56e35f34ae9f5b3a335ef510e0ade" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu Jan 25 09:53:06 2024 +0000" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Feb 09 20:29:14 2024 +0100" }, "message": "btrfs: add and use helper to check if block group is used\n\nAdd a helper function to determine if a block group is being used and make\nuse of it at btrfs_delete_unused_bgs(). This helper will also be used in\nfuture code changes.\n\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nReviewed-by: Boris Burkov \u003cboris@bur.io\u003e\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: David Sterba \u003cdsterba@suse.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "5571e41ec6e56e35f34ae9f5b3a335ef510e0ade", "tree": "ceaf74cb3273d670fef2c114a0905453799b1d0f", "parents": [ "e03ee2fe873eb68c1f9ba5112fee70303ebf9dfb" ], "author": { "name": "Josef Bacik", "email": "josef@toxicpanda.com", "time": "Wed Jan 31 14:27:25 2024 -0500" }, "committer": { "name": "David Sterba", "email": "dsterba@suse.com", "time": "Fri Feb 09 20:28:28 2024 +0100" }, "message": "btrfs: don\u0027t drop extent_map for free space inode on write error\n\nWhile running the CI for an unrelated change I hit the following panic\nwith generic/648 on btrfs_holes_spacecache.\n\nassertion failed: block_start !\u003d EXTENT_MAP_HOLE, in fs/btrfs/extent_io.c:1385\n------------[ cut here ]------------\nkernel BUG at fs/btrfs/extent_io.c:1385!\ninvalid opcode: 0000 [#1] PREEMPT SMP NOPTI\nCPU: 1 PID: 2695096 Comm: fsstress Kdump: loaded Tainted: G W 6.8.0-rc2+ #1\nRIP: 0010:__extent_writepage_io.constprop.0+0x4c1/0x5c0\nCall Trace:\n \u003cTASK\u003e\n extent_write_cache_pages+0x2ac/0x8f0\n extent_writepages+0x87/0x110\n do_writepages+0xd5/0x1f0\n filemap_fdatawrite_wbc+0x63/0x90\n __filemap_fdatawrite_range+0x5c/0x80\n btrfs_fdatawrite_range+0x1f/0x50\n btrfs_write_out_cache+0x507/0x560\n btrfs_write_dirty_block_groups+0x32a/0x420\n commit_cowonly_roots+0x21b/0x290\n btrfs_commit_transaction+0x813/0x1360\n btrfs_sync_file+0x51a/0x640\n __x64_sys_fdatasync+0x52/0x90\n do_syscall_64+0x9c/0x190\n entry_SYSCALL_64_after_hwframe+0x6e/0x76\n\nThis happens because we fail to write out the free space cache in one\ninstance, come back around and attempt to write it again. However on\nthe second pass through we go to call btrfs_get_extent() on the inode to\nget the extent mapping. Because this is a new block group, and with the\nfree space inode we always search the commit root to avoid deadlocking\nwith the tree, we find nothing and return a EXTENT_MAP_HOLE for the\nrequested range.\n\nThis happens because the first time we try to write the space cache out\nwe hit an error, and on an error we drop the extent mapping. This is\nnormal for normal files, but the free space cache inode is special. We\nalways expect the extent map to be correct. Thus the second time\nthrough we end up with a bogus extent map.\n\nSince we\u0027re deprecating this feature, the most straightforward way to\nfix this is to simply skip dropping the extent map range for this failed\nrange.\n\nI shortened the test by using error injection to stress the area to make\nit easier to reproduce. With this patch in place we no longer panic\nwith my error injection test.\n\nCC: stable@vger.kernel.org # 4.14+\nReviewed-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nSigned-off-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nSigned-off-by: David Sterba \u003cdsterba@suse.com\u003e\n" }, { "commit": "9ed18b0b7765ec3aa5f10843d1c3daaa1a225337", "tree": "bcd64ea4f057baae900fea6b41373242479ba52c", "parents": [ "ca8a66738a444e05fb256ca0081224d066318401", "3951f6add519a8e954bf78691a412f65b24f4715" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:19:36 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:19:36 2024 -0800" }, "message": "Merge tag \u0027riscv-for-linus-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux\n\nPull RISC-V fixes from Palmer Dabbelt:\n\n - fix missing TLB flush during early boot on SPARSEMEM_VMEMMAP\n configurations\n\n - fixes to correctly implement the break-before-make behavior requried\n by the ISA for NAPOT mappings\n\n - fix a missing TLB flush on intermediate mapping changes\n\n - fix build warning about a missing declaration of overflow_stack\n\n - fix performace regression related to incorrect tracking of completed\n batch TLB flushes\n\n* tag \u0027riscv-for-linus-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:\n riscv: Fix arch_tlbbatch_flush() by clearing the batch cpumask\n riscv: declare overflow_stack as exported from traps.c\n riscv: Fix arch_hugetlb_migration_supported() for NAPOT\n riscv: Flush the tlb when a page directory is freed\n riscv: Fix hugetlb_mask_last_page() when NAPOT is enabled\n riscv: Fix set_huge_pte_at() for NAPOT mapping\n riscv: mm: execute local TLB flush after populating vmemmap\n" }, { "commit": "ca8a66738a444e05fb256ca0081224d066318401", "tree": "9c5239bdede78afa276637b9b43a6dfcd69494a8", "parents": [ "6dc512a0a22e817100a0522d8f0985c8ffc86456", "44dc5c41b5b1267d4dd037d26afc0c4d3a568acb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:13:19 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:13:19 2024 -0800" }, "message": "Merge tag \u0027trace-v6.8-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull tracing fixes from Steven Rostedt:\n\n - Fix broken direct trampolines being called when another callback is\n attached the same function.\n\n ARM 64 does not support FTRACE_WITH_REGS, and when it added direct\n trampoline calls from ftrace, it removed the \"WITH_REGS\" flag from\n the ftrace_ops for direct trampolines. This broke x86 as x86 requires\n direct trampolines to have WITH_REGS.\n\n This wasn\u0027t noticed because direct trampolines work as long as the\n function it is attached to is not shared with other callbacks (like\n the function tracer). When there are other callbacks, a helper\n trampoline is called, to call all the non direct callbacks and when\n it returns, the direct trampoline is called.\n\n For x86, the direct trampoline sets a flag in the regs field to tell\n the x86 specific code to call the direct trampoline. But this only\n works if the ftrace_ops had WITH_REGS set. ARM does things\n differently that does not require this. For now, set WITH_REGS if the\n arch supports WITH_REGS (which ARM does not), and this makes it work\n for both ARM64 and x86.\n\n - Fix wasted memory in the saved_cmdlines logic.\n\n The saved_cmdlines is a cache that maps PIDs to COMMs that tracing\n can use. Most trace events only save the PID in the event. The\n saved_cmdlines file lists PIDs to COMMs so that the tracing tools can\n show an actual name and not just a PID for each event. There\u0027s an\n array of PIDs that map to a small set of saved COMM strings. The\n array is set to PID_MAX_DEFAULT which is usually set to 32768. When a\n PID comes in, it will add itself to this array along with the index\n into the COMM array (note if the system allows more than\n PID_MAX_DEFAULT, this cache is similar to cache lines as an update of\n a PID that has the same PID_MAX_DEFAULT bits set will flush out\n another task with the same matching bits set).\n\n A while ago, the size of this cache was changed to be dynamic and the\n array was moved into a structure and created with kmalloc(). But this\n new structure had the size of 131104 bytes, or 0x20020 in hex. As\n kmalloc allocates in powers of two, it was actually allocating\n 0x40000 bytes (262144) leaving 131040 bytes of wasted memory. The\n last element of this structure was a pointer to the COMM string array\n which defaulted to just saving 128 COMMs.\n\n By changing the last field of this structure to a variable length\n string, and just having it round up to fill the allocated memory, the\n default size of the saved COMM cache is now 8190. This not only uses\n the wasted space, but actually saves space by removing the extra\n allocation for the COMM names.\n\n* tag \u0027trace-v6.8-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n tracing: Fix wasted memory in saved_cmdlines logic\n ftrace: Fix DIRECT_CALLS to use SAVE_REGS by default\n" }, { "commit": "6dc512a0a22e817100a0522d8f0985c8ffc86456", "tree": "d45b24ec2f25f5e74d10e8c3df40b22cfbd598e6", "parents": [ "e6f39a90de9213693db19aeb2ddea54163f104d7", "9efd24ec5599ed485b7c4d9aeb731141f6285167" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:04:26 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 11:04:26 2024 -0800" }, "message": "Merge tag \u0027probes-fixes-v6.8-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull probes fixes from Masami Hiramatsu:\n\n - remove unnecessary initial values of kprobes local variables\n\n - probe-events parser bug fixes:\n\n - calculate the argument size and format string after setting type\n information from BTF, because BTF can change the size and format\n string.\n\n - show $comm parse error correctly instead of failing silently.\n\n* tag \u0027probes-fixes-v6.8-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n kprobes: Remove unnecessary initial values of variables\n tracing/probes: Fix to set arg size and fmt after setting type from BTF\n tracing/probes: Fix to show a parse error for bad type for $comm\n" }, { "commit": "e6f39a90de9213693db19aeb2ddea54163f104d7", "tree": "ef9694fe767eee0ae0f03d79f683cfc8c3de8019", "parents": [ "5ddfc2460611567cea5ed097f9534e7e8e84744b", "1ad55cecf22f05f1c884adf63cc09d3c3e609ebf" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:40:50 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:40:50 2024 -0800" }, "message": "Merge tag \u0027efi-fixes-for-v6.8-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi\n\nPull EFI fixes from Ard Biesheuvel:\n \"The only notable change here is the patch that changes the way we deal\n with spurious errors from the EFI memory attribute protocol. This will\n be backported to v6.6, and is intended to ensure that we will not\n paint ourselves into a corner when we tighten this further in order to\n comply with MS requirements on signed EFI code.\n\n Note that this protocol does not currently exist in x86 production\n systems in the field, only in Microsoft\u0027s fork of OVMF, but it will be\n mandatory for Windows logo certification for x86 PCs in the future.\n\n - Tighten ELF relocation checks on the RISC-V EFI stub\n\n - Give up if the new EFI memory attributes protocol fails spuriously\n on x86\n\n - Take care not to place the kernel in the lowest 16 MB of DRAM on\n x86\n\n - Omit special purpose EFI memory from memblock\n\n - Some fixes for the CXL CPER reporting code\n\n - Make the PE/COFF layout of mixed-mode capable images comply with a\n strict interpretation of the spec\"\n\n* tag \u0027efi-fixes-for-v6.8-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi:\n x86/efistub: Use 1:1 file:memory mapping for PE/COFF .compat section\n cxl/trace: Remove unnecessary memcpy\u0027s\n cxl/cper: Fix errant CPER prints for CXL events\n efi: Don\u0027t add memblocks for soft-reserved memory\n efi: runtime: Fix potential overflow of soft-reserved region size\n efi/libstub: Add one kernel-doc comment\n x86/efistub: Avoid placing the kernel below LOAD_PHYSICAL_ADDR\n x86/efistub: Give up if memory attribute protocol returns an error\n riscv/efistub: Tighten ELF relocation check\n riscv/efistub: Ensure GP-relative addressing is not used\n" }, { "commit": "5ddfc2460611567cea5ed097f9534e7e8e84744b", "tree": "6e19be8e6a3976d4a1bbdbc65cb26b368c3be669", "parents": [ "5ca243c23ee122ed983a2af3d454a4b58677d2f5", "67057f48df79a3d73683385f521215146861684b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:37:59 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:37:59 2024 -0800" }, "message": "Merge tag \u0027pci-v6.8-fixes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci\n\nPull pci fixes from Bjorn Helgaas:\n\n - Fix an unintentional truncation of DWC MSI-X address to 32 bits and\n update similar MSI code to match (Dan Carpenter)\n\n* tag \u0027pci-v6.8-fixes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci:\n PCI: dwc: Clean up dw_pcie_ep_raise_msi_irq() alignment\n PCI: dwc: Fix a 64bit bug in dw_pcie_ep_raise_msix_irq()\n" }, { "commit": "5ca243c23ee122ed983a2af3d454a4b58677d2f5", "tree": "176e31e1f0bc1eb6a60a34deb997aae13eb636a9", "parents": [ "eb747bcc3611c7b3df6c8e5826f5003daeea6e9f", "34cf8c657cf0365791cdc658ddbca9cc907726ce" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:35:39 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:35:39 2024 -0800" }, "message": "Merge tag \u0027hwmon-for-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull hwmon fixes from Guenter Roeck:\n\n - coretemp: Various fixes, and increase number of supported CPU cores\n\n - aspeed-pwm-tacho: Add missing mutex protection\n\n* tag \u0027hwmon-for-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n hwmon: (coretemp) Enlarge per package core count limit\n hwmon: (coretemp) Fix bogus core_id to attr name mapping\n hwmon: (coretemp) Fix out-of-bounds memory access\n hwmon: (aspeed-pwm-tacho) mutex for tach reading\n" }, { "commit": "eb747bcc3611c7b3df6c8e5826f5003daeea6e9f", "tree": "c91ff3b4c2e4b7b82118bbc29f24166148affd72", "parents": [ "3760081ff414c4fec37396499f5b298464583b03", "cc9432c4fb159a3913e0ce3173b8218cd5bad2e0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:33:54 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:33:54 2024 -0800" }, "message": "Merge tag \u0027mmc-v6.8-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull MMC fixes from Ulf Hansson:\n \"MMC core:\n - Allow non-sleeping read-only slot-gpio\n\n MMC host:\n - sdhci-pci-o2micro: Fix a warm reboot BIOS issue\"\n\n* tag \u0027mmc-v6.8-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: slot-gpio: Allow non-sleeping GPIO ro\n mmc: sdhci-pci-o2micro: Fix a warm reboot issue that disk can\u0027t be detected by BIOS\n" }, { "commit": "3760081ff414c4fec37396499f5b298464583b03", "tree": "90cf95ddc4e567833c167c756ba79f0047a6db0e", "parents": [ "4a8e4b3c2741a46c7c03c1726a124a750ba06905", "c41336f4d69057cbf88fed47951379b384540df5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:29:50 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:29:50 2024 -0800" }, "message": "Merge tag \u0027pmdomain-v6.8-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/linux-pm\n\nPull pmdomain fixes from Ulf Hansson:\n \"Core:\n - Move the unused cleanup to a _sync initcall\n\n Providers:\n - mediatek: Fix race conditions at probe/remove with genpd\n - renesas: r8a77980-sysc: CR7 must be always on\"\n\n* tag \u0027pmdomain-v6.8-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/linux-pm:\n pmdomain: mediatek: fix race conditions with genpd\n pmdomain: renesas: r8a77980-sysc: CR7 must be always on\n pmdomain: core: Move the unused cleanup to a _sync initcall\n" }, { "commit": "4a8e4b3c2741a46c7c03c1726a124a750ba06905", "tree": "054f664b147b5781838af1d76525edef35f814ce", "parents": [ "c76b766ec50d3d43e2dacea53a733b285f4b730d", "2526dffc6d65cffa32b88556bd68e4e72e889a55" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:27:56 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 10:27:56 2024 -0800" }, "message": "Merge tag \u0027gpio-fixes-for-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fix from Bartosz Golaszewski:\n\n - remove the new GPIO device from the global list unconditionally in\n error path in core GPIOLIB\n\n* tag \u0027gpio-fixes-for-v6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpio: remove GPIO device from the list unconditionally in error path\n" }, { "commit": "c76b766ec50d3d43e2dacea53a733b285f4b730d", "tree": "0bb4e915151d414858f79b725aaf3a9554e23253", "parents": [ "1f719a2f3fa67665578c759ac34fd3d3690c1a20", "311520887d7cad2d5494933bd19bf85eb5398ffc" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 09:57:12 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 09 09:57:12 2024 -0800" }, "message": "Merge tag \u0027drm-fixes-2024-02-09\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"Regular weekly fixes, xe, amdgpu and msm are most of them, with some\n misc in i915, ivpu and nouveau, scattered but nothing too intense at\n this point.\n\n i915:\n - gvt: docs fix, uninit var, MAINTAINERS\n\n ivpu:\n - add aborted job status\n - disable d3 hot delay\n - mmu fixes\n\n nouveau:\n - fix gsp rpc size request\n - fix dma buffer leaks\n - use common code for gsp mem ctor\n\n xe:\n - Fix a loop in an error path\n - Fix a missing dma-fence reference\n - Fix a retry path on userptr REMAP\n - Workaround for a false gcc warning\n - Fix missing map of the usm batch buffer in the migrate vm.\n - Fix a memory leak.\n - Fix a bad assumption of used page size\n - Fix hitting a BUG() due to zero pages to map.\n - Remove some leftover async bind queue relics\n\n amdgpu:\n - Misc NULL/bounds check fixes\n - ODM pipe policy fix\n - Aborted suspend fixes\n - JPEG 4.0.5 fix\n - DCN 3.5 fixes\n - PSP fix\n - DP MST fix\n - Phantom pipe fix\n - VRAM vendor fix\n - Clang fix\n - SR-IOV fix\n\n msm:\n - DPU:\n - fix for kernel doc warnings and smatch warnings in dpu_encoder\n - fix for smatch warning in dpu_encoder\n - fix the bus bandwidth value for SDM670\n - DP:\n - fixes to handle unknown bpc case correctly for DP\n - fix for MISC0 programming\n - GPU:\n - dmabuf vmap fix\n - a610 UBWC corruption fix (incorrect hbb)\n - revert a commit that was making GPU recovery unreliable\"\n\n* tag \u0027drm-fixes-2024-02-09\u0027 of git://anongit.freedesktop.org/drm/drm: (43 commits)\n drm/xe: Remove TEST_VM_ASYNC_OPS_ERROR\n drm/xe/vm: don\u0027t ignore error when in_kthread\n drm/xe: Assume large page size if VMA not yet bound\n drm/xe/display: Fix memleak in display initialization\n drm/xe: Map both mem.kernel_bb_pool and usm.bb_pool\n drm/xe: circumvent bogus stringop-overflow warning\n drm/xe: Pick correct userptr VMA to repin on REMAP op failure\n drm/xe: Take a reference in xe_exec_queue_last_fence_get()\n drm/xe: Fix loop in vm_bind_ioctl_ops_unwind\n drm/amdgpu: Fix HDP flush for VFs on nbio v7.9\n drm/amd/display: Implement bounds check for stream encoder creation in DCN301\n drm/amd/display: Increase frame-larger-than for all display_mode_vba files\n drm/amd/display: Clear phantom stream count and plane count\n drm/amdgpu: Avoid fetching VRAM vendor info\n drm/amd/display: Disable ODM by default for DCN35\n drm/amd/display: Update phantom pipe enable / disable sequence\n drm/amd/display: Fix MST Null Ptr for RV\n drm/amdgpu: Fix shared buff copy to user\n drm/amd/display: Increase eval/entry delay for DCN35\n drm/amdgpu: remove asymmetrical irq disabling in jpeg 4.0.5 suspend\n ...\n" }, { "commit": "f6a1892585cd19e63c4ef2334e26cd536d5b678d", "tree": "0dad38381c9999a03cc65d7db9f44556123bdaed", "parents": [ "d877550eaf2dc9090d782864c96939397a3c6835" ], "author": { "name": "Aleksander Mazur", "email": "deweloper@wp.pl", "time": "Tue Jan 23 14:43:00 2024 +0100" }, "committer": { "name": "Borislav Petkov (AMD)", "email": "bp@alien8.de", "time": "Fri Feb 09 16:28:19 2024 +0100" }, "message": "x86/Kconfig: Transmeta Crusoe is CPU family 5, not 6\n\nThe kernel built with MCRUSOE is unbootable on Transmeta Crusoe. It shows\nthe following error message:\n\n This kernel requires an i686 CPU, but only detected an i586 CPU.\n Unable to boot - please use a kernel appropriate for your CPU.\n\nRemove MCRUSOE from the condition introduced in commit in Fixes, effectively\nchanging X86_MINIMUM_CPU_FAMILY back to 5 on that machine, which matches the\nCPU family given by CPUID.\n\n [ bp: Massage commit message. ]\n\nFixes: 25d76ac88821 (\"x86/Kconfig: Explicitly enumerate i686-class CPUs in Kconfig\")\nSigned-off-by: Aleksander Mazur \u003cdeweloper@wp.pl\u003e\nSigned-off-by: Borislav Petkov (AMD) \u003cbp@alien8.de\u003e\nAcked-by: H. Peter Anvin \u003chpa@zytor.com\u003e\nCc: \u003cstable@kernel.org\u003e\nLink: https://lore.kernel.org/r/20240123134309.1117782-1-deweloper@wp.pl\n" }, { "commit": "44dc5c41b5b1267d4dd037d26afc0c4d3a568acb", "tree": "d289ab40e32930d440d831ce892156c07bcc8e9d", "parents": [ "a8b9cf62ade1bf17261a979fc97e40c2d7842353" ], "author": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Fri Feb 09 06:36:22 2024 -0500" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Fri Feb 09 06:43:21 2024 -0500" }, "message": "tracing: Fix wasted memory in saved_cmdlines logic\n\nWhile looking at improving the saved_cmdlines cache I found a huge amount\nof wasted memory that should be used for the cmdlines.\n\nThe tracing data saves pids during the trace. At sched switch, if a trace\noccurred, it will save the comm of the task that did the trace. This is\nsaved in a \"cache\" that maps pids to comms and exposed to user space via\nthe /sys/kernel/tracing/saved_cmdlines file. Currently it only caches by\ndefault 128 comms.\n\nThe structure that uses this creates an array to store the pids using\nPID_MAX_DEFAULT (which is usually set to 32768). This causes the structure\nto be of the size of 131104 bytes on 64 bit machines.\n\nIn hex: 131104 \u003d 0x20020, and since the kernel allocates generic memory in\npowers of two, the kernel would allocate 0x40000 or 262144 bytes to store\nthis structure. That leaves 131040 bytes of wasted space.\n\nWorse, the structure points to an allocated array to store the comm names,\nwhich is 16 bytes times the amount of names to save (currently 128), which\nis 2048 bytes. Instead of allocating a separate array, make the structure\nend with a variable length string and use the extra space for that.\n\nThis is similar to a recommendation that Linus had made about eventfs_inode names:\n\n https://lore.kernel.org/all/20240130190355.11486-5-torvalds@linux-foundation.org/\n\nInstead of allocating a separate string array to hold the saved comms,\nhave the structure end with: char saved_cmdlines[]; and round up to the\nnext power of two over sizeof(struct saved_cmdline_buffers) + num_cmdlines * TASK_COMM_LEN\nIt will use this extra space for the saved_cmdline portion.\n\nNow, instead of saving only 128 comms by default, by using this wasted\nspace at the end of the structure it can save over 8000 comms and even\nsaves space by removing the need for allocating the other array.\n\nLink: https://lore.kernel.org/linux-trace-kernel/20240209063622.1f7b6d5f@rorschach.local.home\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nCc: Vincent Donnefort \u003cvdonnefort@google.com\u003e\nCc: Sven Schnelle \u003csvens@linux.ibm.com\u003e\nCc: Mete Durlu \u003cmeted@linux.ibm.com\u003e\nFixes: 939c7a4f04fcd (\"tracing: Introduce saved_cmdlines_size file\")\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "a8b9cf62ade1bf17261a979fc97e40c2d7842353", "tree": "816deca0cf2f58ba88a7e7cc8b97ac90b9d766db", "parents": [ "ca185770db914869ff9fe773bac5e0e5e4165b83" ], "author": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Wed Jan 10 09:13:06 2024 +0900" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Fri Feb 09 04:58:22 2024 -0500" }, "message": "ftrace: Fix DIRECT_CALLS to use SAVE_REGS by default\n\nThe commit 60c8971899f3 (\"ftrace: Make DIRECT_CALLS work WITH_ARGS\nand !WITH_REGS\") changed DIRECT_CALLS to use SAVE_ARGS when there\nare multiple ftrace_ops at the same function, but since the x86 only\nsupport to jump to direct_call from ftrace_regs_caller, when we set\nthe function tracer on the same target function on x86, ftrace-direct\ndoes not work as below (this actually works on arm64.)\n\nAt first, insmod ftrace-direct.ko to put a direct_call on\n\u0027wake_up_process()\u0027.\n\n # insmod kernel/samples/ftrace/ftrace-direct.ko\n # less trace\n...\n \u003cidle\u003e-0 [006] ..s1. 564.686958: my_direct_func: waking up rcu_preempt-17\n \u003cidle\u003e-0 [007] ..s1. 564.687836: my_direct_func: waking up kcompactd0-63\n \u003cidle\u003e-0 [006] ..s1. 564.690926: my_direct_func: waking up rcu_preempt-17\n \u003cidle\u003e-0 [006] ..s1. 564.696872: my_direct_func: waking up rcu_preempt-17\n \u003cidle\u003e-0 [007] ..s1. 565.191982: my_direct_func: waking up kcompactd0-63\n\nSetup a function filter to the \u0027wake_up_process\u0027 too, and enable it.\n\n # cd /sys/kernel/tracing/\n # echo wake_up_process \u003e set_ftrace_filter\n # echo function \u003e current_tracer\n # less trace\n...\n \u003cidle\u003e-0 [006] ..s3. 686.180972: wake_up_process \u003c-call_timer_fn\n \u003cidle\u003e-0 [006] ..s3. 686.186919: wake_up_process \u003c-call_timer_fn\n \u003cidle\u003e-0 [002] ..s3. 686.264049: wake_up_process \u003c-call_timer_fn\n \u003cidle\u003e-0 [002] d.h6. 686.515216: wake_up_process \u003c-kick_pool\n \u003cidle\u003e-0 [002] d.h6. 686.691386: wake_up_process \u003c-kick_pool\n\nThen, only function tracer is shown on x86.\nBut if you enable \u0027kprobe on ftrace\u0027 event (which uses SAVE_REGS flag)\non the same function, it is shown again.\n\n # echo \u0027p wake_up_process\u0027 \u003e\u003e dynamic_events\n # echo 1 \u003e events/kprobes/p_wake_up_process_0/enable\n # echo \u003e trace\n # less trace\n...\n \u003cidle\u003e-0 [006] ..s2. 2710.345919: p_wake_up_process_0: (wake_up_process+0x4/0x20)\n \u003cidle\u003e-0 [006] ..s3. 2710.345923: wake_up_process \u003c-call_timer_fn\n \u003cidle\u003e-0 [006] ..s1. 2710.345928: my_direct_func: waking up rcu_preempt-17\n \u003cidle\u003e-0 [006] ..s2. 2710.349931: p_wake_up_process_0: (wake_up_process+0x4/0x20)\n \u003cidle\u003e-0 [006] ..s3. 2710.349934: wake_up_process \u003c-call_timer_fn\n \u003cidle\u003e-0 [006] ..s1. 2710.349937: my_direct_func: waking up rcu_preempt-17\n\nTo fix this issue, use SAVE_REGS flag for multiple ftrace_ops flag of\ndirect_call by default.\n\nLink: https://lore.kernel.org/linux-trace-kernel/170484558617.178953.1590516949390270842.stgit@devnote2\n\nFixes: 60c8971899f3 (\"ftrace: Make DIRECT_CALLS work WITH_ARGS and !WITH_REGS\")\nCc: stable@vger.kernel.org\nCc: Florent Revest \u003crevest@chromium.org\u003e\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nReviewed-by: Mark Rutland \u003cmark.rutland@arm.com\u003e\nTested-by: Mark Rutland \u003cmark.rutland@arm.com\u003e [arm64]\nAcked-by: Jiri Olsa \u003cjolsa@kernel.org\u003e\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "311520887d7cad2d5494933bd19bf85eb5398ffc", "tree": "5246277f04cef32c7f105431bea220b8d53f4806", "parents": [ "b30bed9d0012f295843f57058b8927e80eac5c54", "8d35217149daa33358c284aca6a56d5ab92cfc6c" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:32:38 2024 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:32:38 2024 +1000" }, "message": "Merge tag \u0027drm-msm-fixes-2024-02-07\u0027 of https://gitlab.freedesktop.org/drm/msm into drm-fixes\n\nFixes for v6.8-rc4\n\nDPU:\n- fix for kernel doc warnings and smatch warnings in dpu_encoder\n- fix for smatch warning in dpu_encoder\n- fix the bus bandwidth value for SDM670\n\nDP:\n- fixes to handle unknown bpc case correctly for DP. The current code was\n spilling over into other bits of DP configuration register, had to be\n fixed to avoid the extra shifts which were causing the spill over\n- fix for MISC0 programming in DP driver to program the correct\n colorimetry value\n\nGPU:\n- dmabuf vmap fix\n- a610 UBWC corruption fix (incorrect hbb)\n- revert a commit that was making GPU recovery unreliable\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Rob Clark \u003crobdclark@gmail.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGv+tb1+_cp7ftxcMZbbxE9810rvxeaC50eL\u003dmsQ+zkm0g@mail.gmail.com\n" }, { "commit": "b30bed9d0012f295843f57058b8927e80eac5c54", "tree": "5bf786b23303f2ab3b6446204083e4fa2271860b", "parents": [ "9da93fe430aac36fb7342a61434f305c4d791a43", "534c8a5b9d5d41d30cdcac93cfa1bca5e17be009" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:21:16 2024 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:21:23 2024 +1000" }, "message": "Merge tag \u0027amd-drm-fixes-6.8-2024-02-08\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes\n\namd-drm-fixes-6.8-2024-02-08:\n\namdgpu:\n- Misc NULL/bounds check fixes\n- ODM pipe policy fix\n- Aborted suspend fixes\n- JPEG 4.0.5 fix\n- DCN 3.5 fixes\n- PSP fix\n- DP MST fix\n- Phantom pipe fix\n- VRAM vendor fix\n- Clang fix\n- SR-IOV fix\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240208165500.4887-1-alexander.deucher@amd.com\n" }, { "commit": "9da93fe430aac36fb7342a61434f305c4d791a43", "tree": "2303b4eec64ee23c8d632b889c62715d00ad3933", "parents": [ "60c16201b680598951b920ae9b6a6eba9164216f", "a99682e839af7be11a606bf802cba5b2bf93b8e9" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:17:57 2024 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:17:57 2024 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2024-02-08\u0027 of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes\n\n- Just includes gvt-fixes-2024-02-05\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Joonas Lahtinen \u003cjoonas.lahtinen@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/ZcTETgXsejwVwat6@jlahtine-mobl.ger.corp.intel.com\n" }, { "commit": "60c16201b680598951b920ae9b6a6eba9164216f", "tree": "bc277219fe76f885794445ab54944a1c20e72b4f", "parents": [ "6c2bf9ca24a4168558420fd9e95d375e66bd1d78", "bf4c27b8267d7848bb81fd41e6aa07aa662f07fb" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:12:01 2024 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:12:09 2024 +1000" }, "message": "Merge tag \u0027drm-xe-fixes-2024-02-08\u0027 of https://gitlab.freedesktop.org/drm/xe/kernel into drm-fixes\n\nDriver Changes:\n- Fix a loop in an error path\n- Fix a missing dma-fence reference\n- Fix a retry path on userptr REMAP\n- Workaround for a false gcc warning\n- Fix missing map of the usm batch buffer\n in the migrate vm.\n- Fix a memory leak.\n- Fix a bad assumption of used page size\n- Fix hitting a BUG() due to zero pages to map.\n- Remove some leftover async bind queue relics\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Thomas Hellstrom \u003cthomas.hellstrom@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/ZcS2LllawGifubsk@fedora\n" }, { "commit": "6c2bf9ca24a4168558420fd9e95d375e66bd1d78", "tree": "823bc85569ac11663fe4f26e5550e04fe81987fe", "parents": [ "d0399da9fb5f8e3d897b9776bffee2d3bfe20210", "5f8408aca66772d3aa9b4831577b2ac5ec41bcd9" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 10:52:43 2024 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Feb 09 11:11:21 2024 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2024-02-08\u0027 of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes\n\nA null pointer dereference fix for v3d, a TTM pool initialization fix,\nseveral fixes for nouveau around register size, DMA buffer leaks and API\nconsistency, a multiple fixes for ivpu around MMU setup, initialization\nand firmware interactions.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Maxime Ripard \u003cmripard@redhat.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/4wsi2i6kgkqdu7nzp4g7hxasbswnrmc5cakgf5zzvnix53u7lr@4rmp7hwblow3\n" }, { "commit": "1f719a2f3fa67665578c759ac34fd3d3690c1a20", "tree": "4821a92547378496b0dcf22986b8218993d1fd78", "parents": [ "b0d5d0f73761124f0ece33f5fec421e76a22a9fd", "63e4b9d693e0f8c28359c7ea81e1ee510864c37b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 15:09:29 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 15:09:29 2024 -0800" }, "message": "Merge tag \u0027net-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Paolo Abeni:\n \"Including fixes from WiFi and netfilter.\n\n Current release - regressions:\n\n - nic: intel: fix old compiler regressions\n\n - netfilter: ipset: missing gc cancellations fixed\n\n Current release - new code bugs:\n\n - netfilter: ctnetlink: fix filtering for zone 0\n\n Previous releases - regressions:\n\n - core: fix from address in memcpy_to_iter_csum()\n\n - netfilter: nfnetlink_queue: un-break NF_REPEAT\n\n - af_unix: fix memory leak for dead unix_(sk)-\u003eoob_skb in GC.\n\n - devlink: avoid potential loop in devlink_rel_nested_in_notify_work()\n\n - iwlwifi:\n - mvm: fix a battery life regression\n - fix double-free bug\n\n - mac80211: fix waiting for beacons logic\n\n - nic: nfp: flower: prevent re-adding mac index for bonded port\n\n Previous releases - always broken:\n\n - rxrpc: fix generation of serial numbers to skip zero\n\n - tipc: check the bearer type before calling tipc_udp_nl_bearer_add()\n\n - tunnels: fix out of bounds access when building IPv6 PMTU error\n\n - nic: hv_netvsc: register VF in netvsc_probe if NET_DEVICE_REGISTER\n missed\n\n - nic: atlantic: fix DMA mapping for PTP hwts ring\n\n Misc:\n\n - selftests: more fixes to deal with very slow hosts\"\n\n* tag \u0027net-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (80 commits)\n netfilter: nft_set_pipapo: remove scratch_aligned pointer\n netfilter: nft_set_pipapo: add helper to release pcpu scratch area\n netfilter: nft_set_pipapo: store index in scratch maps\n netfilter: nft_set_rbtree: skip end interval element from gc\n netfilter: nfnetlink_queue: un-break NF_REPEAT\n netfilter: nf_tables: use timestamp to check for set element timeout\n netfilter: nft_ct: reject direction for ct id\n netfilter: ctnetlink: fix filtering for zone 0\n s390/qeth: Fix potential loss of L3-IP@ in case of network issues\n netfilter: ipset: Missing gc cancellations fixed\n octeontx2-af: Initialize maps.\n net: ethernet: ti: cpsw: enable mac_managed_pm to fix mdio\n net: ethernet: ti: cpsw_new: enable mac_managed_pm to fix mdio\n netfilter: nft_set_pipapo: remove static in nft_pipapo_get()\n netfilter: nft_compat: restrict match/target protocol to u16\n netfilter: nft_compat: reject unused compat flag\n netfilter: nft_compat: narrow down revision to unsigned 8-bits\n net: intel: fix old compiler regressions\n MAINTAINERS: Maintainer change for rds\n selftests: cmsg_ipv6: repeat the exact packet\n ...\n" }, { "commit": "b0d5d0f73761124f0ece33f5fec421e76a22a9fd", "tree": "d1c1bdd7f8e3001b8ad450d4d7a25247c045d47e", "parents": [ "047371968ffc470769f541d6933e262dc7085456", "4451e8e8415e4ef48cdc763d66855f8c25fda94c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 15:07:06 2024 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 15:07:06 2024 -0800" }, "message": "Merge tag \u0027pinctrl-v6.8-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl\n\nPull pinctrl fix from Linus Walleij:\n \"A single fix for the AMD driver which affects developer laptops, the\n pinctrl/GPIO driver won\u0027t probe on some systems\"\n\n* tag \u0027pinctrl-v6.8-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:\n pinctrl: amd: Add IRQF_ONESHOT to the interrupt request\n" }, { "commit": "5f63a493b99c848ad5200402bebe26211e00025a", "tree": "147b2232b7ec28763c51cd777e14d6a7e471fdd7", "parents": [ "4ce6e2db00de8103a0687fb0f65fd17124a51aaa", "4054705215ad7e6dd8e4e6ff27c39abd7667f700" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Feb 08 15:05:18 2024 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Feb 08 15:05:18 2024 -0700" }, "message": "Merge tag \u0027nvme-6.8-2023-02-08\u0027 of git://git.infradead.org/nvme into block-6.8\n\nPull NVMe fixes from Keith:\n\n\"nvme fixes for Linux 6.8\n\n - Update a potentially stale firmware attribute (Maurizio)\n - Fixes for the recent verbose error logging (Keith, Chaitanya)\n - Protection information payload size fix for passthrough (Francis)\"\n\n* tag \u0027nvme-6.8-2023-02-08\u0027 of git://git.infradead.org/nvme:\n nvme: use ns-\u003ehead-\u003epi_size instead of t10_pi_tuple structure size\n nvme-core: fix comment to reflect right functions\n nvme: move passthrough logging attribute to head\n nvme-host: fix the updating of the firmware version\n" }, { "commit": "4ce6e2db00de8103a0687fb0f65fd17124a51aaa", "tree": "f1d2b668e079df800225d164e718e1d87906703e", "parents": [ "2a427b49d02995ea4a6ff93a1432c40fa4d36821" ], "author": { "name": "Yi Sun", "email": "yi.sun@unisoc.com", "time": "Mon Jan 29 16:52:50 2024 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Feb 08 11:30:26 2024 -0700" }, "message": "virtio-blk: Ensure no requests in virtqueues before deleting vqs.\n\nEnsure no remaining requests in virtqueues before resetting vdev and\ndeleting virtqueues. Otherwise these requests will never be completed.\nIt may cause the system to become unresponsive.\n\nFunction blk_mq_quiesce_queue() can ensure that requests have become\nin_flight status, but it cannot guarantee that requests have been\nprocessed by the device. Virtqueues should never be deleted before\nall requests become complete status.\n\nFunction blk_mq_freeze_queue() ensure that all requests in virtqueues\nbecome complete status. And no requests can enter in virtqueues.\n\nSigned-off-by: Yi Sun \u003cyi.sun@unisoc.com\u003e\nReviewed-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nLink: https://lore.kernel.org/r/20240129085250.1550594-1-yi.sun@unisoc.com\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "c23de7ceae59e4ca5894c3ecf4f785c50c0fa428", "tree": "b1704e727a3ac9864fddc94d1685aeaf3880043c", "parents": [ "d546978e0c07b6333fdbcbd81b2f2e058d4560b5" ], "author": { "name": "Vegard Nossum", "email": "vegard.nossum@oracle.com", "time": "Mon Feb 05 18:51:26 2024 +0100" }, "committer": { "name": "Jonathan Corbet", "email": "corbet@lwn.net", "time": "Thu Feb 08 11:05:35 2024 -0700" }, "message": "docs: kernel_feat.py: fix build error for missing files\n\nIf the directory passed to the \u0027.. kernel-feat::\u0027 directive does not\nexist or the get_feat.pl script does not find any files to extract\nfeatures from, Sphinx will report the following error:\n\n Sphinx parallel build error:\n UnboundLocalError: local variable \u0027fname\u0027 referenced before assignment\n make[2]: *** [Documentation/Makefile:102: htmldocs] Error 2\n\nThis is due to how I changed the script in c48a7c44a1d0 (\"docs:\nkernel_feat.py: fix potential command injection\"). Before that, the\nfilename passed along to self.nestedParse() in this case was weirdly\njust the whole get_feat.pl invocation.\n\nWe can fix it by doing what kernel_abi.py does -- just pass\nself.arguments[0] as \u0027fname\u0027.\n\nFixes: c48a7c44a1d0 (\"docs: kernel_feat.py: fix potential command injection\")\nCc: Justin Forbes \u003cjforbes@fedoraproject.org\u003e\nCc: Salvatore Bonaccorso \u003ccarnil@debian.org\u003e\nCc: Jani Nikula \u003cjani.nikula@intel.com\u003e\nCc: Mauro Carvalho Chehab \u003cmchehab@kernel.org\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Vegard Nossum \u003cvegard.nossum@oracle.com\u003e\nLink: https://lore.kernel.org/r/20240205175133.774271-2-vegard.nossum@oracle.com\nSigned-off-by: Jonathan Corbet \u003ccorbet@lwn.net\u003e\n" }, { "commit": "2a427b49d02995ea4a6ff93a1432c40fa4d36821", "tree": "2f585cabca99a8bc6bc13a133308dd6690a3b523", "parents": [ "f814bdda774c183b0cc15ec8f3b6e7c6f4527ba5" ], "author": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Mon Nov 20 12:25:56 2023 -1000" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Feb 08 10:11:39 2024 -0700" }, "message": "blk-iocost: Fix an UBSAN shift-out-of-bounds warning\n\nWhen iocg_kick_delay() is called from a CPU different than the one which set\nthe delay, @now may be in the past of @iocg-\u003edelay_at leading to the\nfollowing warning:\n\n UBSAN: shift-out-of-bounds in block/blk-iocost.c:1359:23\n shift exponent 18446744073709 is too large for 64-bit type \u0027u64\u0027 (aka \u0027unsigned long long\u0027)\n ...\n Call Trace:\n \u003cTASK\u003e\n dump_stack_lvl+0x79/0xc0\n __ubsan_handle_shift_out_of_bounds+0x2ab/0x300\n iocg_kick_delay+0x222/0x230\n ioc_rqos_merge+0x1d7/0x2c0\n __rq_qos_merge+0x2c/0x80\n bio_attempt_back_merge+0x83/0x190\n blk_attempt_plug_merge+0x101/0x150\n blk_mq_submit_bio+0x2b1/0x720\n submit_bio_noacct_nocheck+0x320/0x3e0\n __swap_writepage+0x2ab/0x9d0\n\nThe underflow itself doesn\u0027t really affect the behavior in any meaningful\nway; however, the past timestamp may exaggerate the delay amount calculated\nlater in the code, which shouldn\u0027t be a material problem given the nature of\nthe delay mechanism.\n\nIf @now is in the past, this CPU is racing another CPU which recently set up\nthe delay and there\u0027s nothing this CPU can contribute w.r.t. the delay.\nLet\u0027s bail early from iocg_kick_delay() in such cases.\n\nReported-by: Breno Leitão \u003cleitao@debian.org\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\nFixes: 5160a5a53c0c (\"blk-iocost: implement delay adjustment hysteresis\")\nLink: https://lore.kernel.org/r/ZVvc9L_CYk5LO1fT@slm.duckdns.org\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "55c7788c37242702868bfac7861cdf0c358d6c3d", "tree": "6967c092dd90ce6ccbea210625910968daec4c3b", "parents": [ "45be0882c5f91e1b92e645001dd1a53b3bd58c97" ], "author": { "name": "Paulo Alcantara", "email": "pc@manguebit.com", "time": "Fri Feb 02 12:38:24 2024 -0300" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Thu Feb 08 10:50:40 2024 -0600" }, "message": "smb: client: set correct d_type for reparse points under DFS mounts\n\nSend query dir requests with an info level of\nSMB_FIND_FILE_FULL_DIRECTORY_INFO rather than\nSMB_FIND_FILE_DIRECTORY_INFO when the client is generating its own\ninode numbers (e.g. noserverino) so that reparse tags still\ncan be parsed directly from the responses, but server won\u0027t\nsend UniqueId (server inode number)\n\nSigned-off-by: Paulo Alcantara \u003cpc@manguebit.com\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "45be0882c5f91e1b92e645001dd1a53b3bd58c97", "tree": "033a9272806d208479c9f55ace4c173da7fcfe60", "parents": [ "54be6c6c5ae8e0d93a6c4641cb7528eb0b6ba478" ], "author": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Mon Feb 05 14:43:17 2024 -0600" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Thu Feb 08 10:50:40 2024 -0600" }, "message": "smb3: add missing null server pointer check\n\nAddress static checker warning in cifs_ses_get_chan_index():\n warn: variable dereferenced before check \u0027server\u0027\nTo be consistent, and reduce risk, we should add another check\nfor null server pointer.\n\nFixes: 88675b22d34e (\"cifs: do not search for channel if server is terminating\")\nReported-by: Dan Carpenter \u003cdan.carpenter@linaro.org\u003e\nReviewed-by: Shyam Prasad N \u003csprasad@microsoft.com\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "9efd24ec5599ed485b7c4d9aeb731141f6285167", "tree": "6a31cd4e78f791970d3d35a3ca037973dad1f98c", "parents": [ "9a571c1e275cedacd48c66a6bddd0c23f1dffdbf" ], "author": { "name": "Li zeming", "email": "zeming@nfschina.com", "time": "Tue Sep 19 09:28:23 2023 +0800" }, "committer": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Thu Feb 08 23:29:29 2024 +0900" }, "message": "kprobes: Remove unnecessary initial values of variables\n\nri and sym is assigned first, so it does not need to initialize the\nassignment.\n\nLink: https://lore.kernel.org/all/20230919012823.7815-1-zeming@nfschina.com/\n\nSigned-off-by: Li zeming \u003czeming@nfschina.com\u003e\nAcked-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\n" }, { "commit": "9a571c1e275cedacd48c66a6bddd0c23f1dffdbf", "tree": "35ad2a45ca09aac4f328586cc24d4921fe6472fa", "parents": [ "8c427cc2fa73684ea140999e121b7b6c1c717632" ], "author": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Wed Jan 24 00:03:02 2024 +0900" }, "committer": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Thu Feb 08 23:26:25 2024 +0900" }, "message": "tracing/probes: Fix to set arg size and fmt after setting type from BTF\n\nSince the BTF type setting updates probe_arg::type, the type size\ncalculation and setting print-fmt should be done after that.\nWithout this fix, the argument size and print-fmt can be wrong.\n\nLink: https://lore.kernel.org/all/170602218196.215583.6417859469540955777.stgit@devnote2/\n\nFixes: b576e09701c7 (\"tracing/probes: Support function parameters if BTF is available\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e" }, { "commit": "8c427cc2fa73684ea140999e121b7b6c1c717632", "tree": "bac02f6a863b1ff0cde4370edb9c32e39850ecb9", "parents": [ "54be6c6c5ae8e0d93a6c4641cb7528eb0b6ba478" ], "author": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Wed Jan 24 00:02:34 2024 +0900" }, "committer": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Thu Feb 08 23:26:13 2024 +0900" }, "message": "tracing/probes: Fix to show a parse error for bad type for $comm\n\nFix to show a parse error for bad type (non-string) for $comm/$COMM and\nimmediate-string. With this fix, error_log file shows appropriate error\nmessage as below.\n\n /sys/kernel/tracing # echo \u0027p vfs_read $comm:u32\u0027 \u003e\u003e kprobe_events\nsh: write error: Invalid argument\n /sys/kernel/tracing # echo \u0027p vfs_read \\\"hoge\":u32\u0027 \u003e\u003e kprobe_events\nsh: write error: Invalid argument\n /sys/kernel/tracing # cat error_log\n\n[ 30.144183] trace_kprobe: error: $comm and immediate-string only accepts string type\n Command: p vfs_read $comm:u32\n ^\n[ 62.618500] trace_kprobe: error: $comm and immediate-string only accepts string type\n Command: p vfs_read \\\"hoge\":u32\n ^\nLink: https://lore.kernel.org/all/170602215411.215583.2238016352271091852.stgit@devnote2/\n\nFixes: 3dd1f7f24f8c (\"tracing: probeevent: Fix to make the type of $comm string\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e" }, { "commit": "63e4b9d693e0f8c28359c7ea81e1ee510864c37b", "tree": "b3ef30d3118b002cde6b9346b27327db85642b28", "parents": [ "2fe8a236436fe40d8d26a1af8d150fc80f04ee1a", "5a8cdf6fd860ac5e6d08d72edbcecee049a7fec4" ], "author": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 12:56:39 2024 +0100" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 12:56:40 2024 +0100" }, "message": "Merge tag \u0027nf-24-02-08\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/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) Narrow down target/match revision to u8 in nft_compat.\n\n2) Bail out with unused flags in nft_compat.\n\n3) Restrict layer 4 protocol to u16 in nft_compat.\n\n4) Remove static in pipapo get command that slipped through when\n reducing set memory footprint.\n\n5) Follow up incremental fix for the ipset performance regression,\n this includes the missing gc cancellation, from Jozsef Kadlecsik.\n\n6) Allow to filter by zone 0 in ctnetlink, do not interpret zone 0\n as no filtering, from Felix Huettner.\n\n7) Reject direction for NFT_CT_ID.\n\n8) Use timestamp to check for set element expiration while transaction\n is handled to prevent garbage collection from removing set elements\n that were just added by this transaction. Packet path and netlink\n dump/get path still use current time to check for expiration.\n\n9) Restore NF_REPEAT in nfnetlink_queue, from Florian Westphal.\n\n10) map_index needs to be percpu and per-set, not just percpu.\n At this time its possible for a pipapo set to fill the all-zero part\n with ones and take the \u0027might have bits set\u0027 as \u0027start-from-zero\u0027 area.\n From Florian Westphal. This includes three patches:\n\n - Change scratchpad area to a structure that provides space for a\n per-set-and-cpu toggle and uses it of the percpu one.\n\n - Add a new free helper to prepare for the next patch.\n\n - Remove the scratch_aligned pointer and makes AVX2 implementation\n use the exact same memory addresses for read/store of the matching\n state.\n\nnetfilter pull request 24-02-08\n\n* tag \u0027nf-24-02-08\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:\n netfilter: nft_set_pipapo: remove scratch_aligned pointer\n netfilter: nft_set_pipapo: add helper to release pcpu scratch area\n netfilter: nft_set_pipapo: store index in scratch maps\n netfilter: nft_set_rbtree: skip end interval element from gc\n netfilter: nfnetlink_queue: un-break NF_REPEAT\n netfilter: nf_tables: use timestamp to check for set element timeout\n netfilter: nft_ct: reject direction for ct id\n netfilter: ctnetlink: fix filtering for zone 0\n netfilter: ipset: Missing gc cancellations fixed\n netfilter: nft_set_pipapo: remove static in nft_pipapo_get()\n netfilter: nft_compat: restrict match/target protocol to u16\n netfilter: nft_compat: reject unused compat flag\n netfilter: nft_compat: narrow down revision to unsigned 8-bits\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://lore.kernel.org/r/20240208112834.1433-1-pablo@netfilter.org\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "5a8cdf6fd860ac5e6d08d72edbcecee049a7fec4", "tree": "f6300ea55eb89b9ae57553e9c2e976b75e275eb4", "parents": [ "47b1c03c3c1a119435480a1e73f27197dc59131d" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Thu Feb 08 10:31:29 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:24:02 2024 +0100" }, "message": "netfilter: nft_set_pipapo: remove scratch_aligned pointer\n\nuse -\u003escratch for both avx2 and the generic implementation.\n\nAfter previous change the scratch-\u003emap member is always aligned properly\nfor AVX2, so we can just use scratch-\u003emap in AVX2 too.\n\nThe alignoff delta is stored in the scratchpad so we can reconstruct\nthe correct address to free the area again.\n\nFixes: 7400b063969b (\"nft_set_pipapo: Introduce AVX2-based lookup implementation\")\nReviewed-by: Stefano Brivio \u003csbrivio@redhat.com\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "47b1c03c3c1a119435480a1e73f27197dc59131d", "tree": "9a282b9922e412e7fca78d783eb51b307acfe87f", "parents": [ "76313d1a4aa9e30d5b43dee5efd8bcd4d8250006" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Wed Feb 07 21:52:47 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nft_set_pipapo: add helper to release pcpu scratch area\n\nAfter next patch simple kfree() is not enough anymore, so add\na helper for it.\n\nReviewed-by: Stefano Brivio \u003csbrivio@redhat.com\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "76313d1a4aa9e30d5b43dee5efd8bcd4d8250006", "tree": "3cc09921d2355f2669dfe5e163351bfa90a1f55f", "parents": [ "60c0c230c6f046da536d3df8b39a20b9a9fd6af0" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Wed Feb 07 21:52:46 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nft_set_pipapo: store index in scratch maps\n\nPipapo needs a scratchpad area to keep state during matching.\nThis state can be large and thus cannot reside on stack.\n\nEach set preallocates percpu areas for this.\n\nOn each match stage, one scratchpad half starts with all-zero and the other\nis inited to all-ones.\n\nAt the end of each stage, the half that starts with all-ones is\nalways zero. Before next field is tested, pointers to the two halves\nare swapped, i.e. resmap pointer turns into fill pointer and vice versa.\n\nAfter the last field has been processed, pipapo stashes the\nindex toggle in a percpu variable, with assumption that next packet\nwill start with the all-zero half and sets all bits in the other to 1.\n\nThis isn\u0027t reliable.\n\nThere can be multiple sets and we can\u0027t be sure that the upper\nand lower half of all set scratch map is always in sync (lookups\ncan be conditional), so one set might have swapped, but other might\nnot have been queried.\n\nThus we need to keep the index per-set-and-cpu, just like the\nscratchpad.\n\nNote that this bug fix is incomplete, there is a related issue.\n\navx2 and normal implementation might use slightly different areas of the\nmap array space due to the avx2 alignment requirements, so\nm-\u003escratch (generic/fallback implementation) and -\u003escratch_aligned\n(avx) may partially overlap. scratch and scratch_aligned are not distinct\nobjects, the latter is just the aligned address of the former.\n\nAfter this change, write to scratch_align-\u003emap_index may write to\nscratch-\u003emap, so this issue becomes more prominent, we can set to 1\na bit in the supposedly-all-zero area of scratch-\u003emap[].\n\nA followup patch will remove the scratch_aligned and makes generic and\navx code use the same (aligned) area.\n\nIts done in a separate change to ease review.\n\nFixes: 3c4287f62044 (\"nf_tables: Add set type for arbitrary concatenation of ranges\")\nReviewed-by: Stefano Brivio \u003csbrivio@redhat.com\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "60c0c230c6f046da536d3df8b39a20b9a9fd6af0", "tree": "a2b4665771dff942fc3e783cab32ec3213d0e859", "parents": [ "f82777e8ce6c039cdcacbcf1eb8619b99a20c06d" ], "author": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Wed Feb 07 18:49:51 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nft_set_rbtree: skip end interval element from gc\n\nrbtree lazy gc on insert might collect an end interval element that has\nbeen just added in this transactions, skip end interval elements that\nare not yet active.\n\nFixes: f718863aca46 (\"netfilter: nft_set_rbtree: fix overlap expiration walk\")\nCc: stable@vger.kernel.org\nReported-by: lonial con \u003ckongln9170@gmail.com\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "f82777e8ce6c039cdcacbcf1eb8619b99a20c06d", "tree": "1c1ad39ff43d5aed2bb56bb3182e049173aaad84", "parents": [ "7395dfacfff65e9938ac0889dafa1ab01e987d15" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Tue Feb 06 17:54:18 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nfnetlink_queue: un-break NF_REPEAT\n\nOnly override userspace verdict if the ct hook returns something\nother than ACCEPT.\n\nElse, this replaces NF_REPEAT (run all hooks again) with NF_ACCEPT\n(move to next hook).\n\nFixes: 6291b3a67ad5 (\"netfilter: conntrack: convert nf_conntrack_update to netfilter verdicts\")\nReported-by: l.6diay@passmail.com\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "7395dfacfff65e9938ac0889dafa1ab01e987d15", "tree": "7fc7c3c79a50488229817f9555365262da45cb3b", "parents": [ "38ed1c7062ada30d7c11e7a7acc749bf27aa14aa" ], "author": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Tue Feb 06 00:11:40 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nf_tables: use timestamp to check for set element timeout\n\nAdd a timestamp field at the beginning of the transaction, store it\nin the nftables per-netns area.\n\nUpdate set backend .insert, .deactivate and sync gc path to use the\ntimestamp, this avoids that an element expires while control plane\ntransaction is still unfinished.\n\n.lookup and .update, which are used from packet path, still use the\ncurrent time to check if the element has expired. And .get path and dump\nalso since this runs lockless under rcu read size lock. Then, there is\nasync gc which also needs to check the current time since it runs\nasynchronously from a workqueue.\n\nFixes: c3e1b005ed1c (\"netfilter: nf_tables: add set element timeout support\")\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "38ed1c7062ada30d7c11e7a7acc749bf27aa14aa", "tree": "5ec0a4ce8640a267b1807e6cf157cfbd48ea7acb", "parents": [ "fa173a1b4e3fd1ab5451cbc57de6fc624c824b0a" ], "author": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Mon Feb 05 14:59:24 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:19 2024 +0100" }, "message": "netfilter: nft_ct: reject direction for ct id\n\nDirection attribute is ignored, reject it in case this ever needs to be\nsupported\n\nFixes: 3087c3f7c23b (\"netfilter: nft_ct: Add ct id support\")\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "fa173a1b4e3fd1ab5451cbc57de6fc624c824b0a", "tree": "f149bfb89ce89c32194f5178c0385c897a8c57d6", "parents": [ "27c5a095e2518975e20a10102908ae8231699879" ], "author": { "name": "Felix Huettner", "email": "felix.huettner@mail.schwarz", "time": "Mon Feb 05 09:59:59 2024 +0000" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:10:18 2024 +0100" }, "message": "netfilter: ctnetlink: fix filtering for zone 0\n\npreviously filtering for the default zone would actually skip the zone\nfilter and flush all zones.\n\nFixes: eff3c558bb7e (\"netfilter: ctnetlink: support filtering by zone\")\nReported-by: Ilya Maximets \u003ci.maximets@ovn.org\u003e\nCloses: https://lore.kernel.org/netdev/2032238f-31ac-4106-8f22-522e76df5a12@ovn.org/\nSigned-off-by: Felix Huettner \u003cfelix.huettner@mail.schwarz\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "2fe8a236436fe40d8d26a1af8d150fc80f04ee1a", "tree": "cbf394a87cb0b5553c29ba4278532adc53e4d985", "parents": [ "db010ff6700f24f96e91ed56ca29cb69335008ef" ], "author": { "name": "Alexandra Winter", "email": "wintera@linux.ibm.com", "time": "Tue Feb 06 09:58:49 2024 +0100" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 12:10:09 2024 +0100" }, "message": "s390/qeth: Fix potential loss of L3-IP@ in case of network issues\n\nSymptom:\nIn case of a bad cable connection (e.g. dirty optics) a fast sequence of\nnetwork DOWN-UP-DOWN-UP could happen. UP triggers recovery of the qeth\ninterface. In case of a second DOWN while recovery is still ongoing, it\ncan happen that the IP@ of a Layer3 qeth interface is lost and will not\nbe recovered by the second UP.\n\nProblem:\nWhen registration of IP addresses with Layer 3 qeth devices fails, (e.g.\nbecause of bad address format) the respective IP address is deleted from\nits hash-table in the driver. If registration fails because of a ENETDOWN\ncondition, the address should stay in the hashtable, so a subsequent\nrecovery can restore it.\n\n3caa4af834df (\"qeth: keep ip-address after LAN_OFFLINE failure\")\nfixes this for registration failures during normal operation, but not\nduring recovery.\n\nSolution:\nKeep L3-IP address in case of ENETDOWN in qeth_l3_recover_ip(). For\nconsistency with qeth_l3_add_ip() we also keep it in case of EADDRINUSE,\ni.e. for some reason the card already/still has this address registered.\n\nFixes: 4a71df50047f (\"qeth: new qeth device driver\")\nCc: stable@vger.kernel.org\nSigned-off-by: Alexandra Winter \u003cwintera@linux.ibm.com\u003e\nLink: https://lore.kernel.org/r/20240206085849.2902775-1-wintera@linux.ibm.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "27c5a095e2518975e20a10102908ae8231699879", "tree": "a5e70afb6b13bc244f8d2a00716293fefb004927", "parents": [ "ab0beafd52b98dfb8b8244b2c6794efbc87478db" ], "author": { "name": "Jozsef Kadlecsik", "email": "kadlec@netfilter.org", "time": "Sun Feb 04 16:26:42 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 12:09:23 2024 +0100" }, "message": "netfilter: ipset: Missing gc cancellations fixed\n\nThe patch fdb8e12cc2cc (\"netfilter: ipset: fix performance regression\nin swap operation\") missed to add the calls to gc cancellations\nat the error path of create operations and at module unload. Also,\nbecause the half of the destroy operations now executed by a\nfunction registered by call_rcu(), neither NFNL_SUBSYS_IPSET mutex\nor rcu read lock is held and therefore the checking of them results\nfalse warnings.\n\nFixes: 97f7cf1cd80e (\"netfilter: ipset: fix performance regression in swap operation\")\nReported-by: syzbot+52bbc0ad036f6f0d4a25@syzkaller.appspotmail.com\nReported-by: Brad Spengler \u003cspender@grsecurity.net\u003e\nReported-by: Стас Ничипорович \u003cstasn77@gmail.com\u003e\nTested-by: Brad Spengler \u003cspender@grsecurity.net\u003e\nTested-by: Стас Ничипорович \u003cstasn77@gmail.com\u003e\nSigned-off-by: Jozsef Kadlecsik \u003ckadlec@netfilter.org\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "db010ff6700f24f96e91ed56ca29cb69335008ef", "tree": "8b7da4e69a3ef697813e4d5987dae293f341347b", "parents": [ "03fa49a386b298d357b90c9c5599f8d00dfc425c" ], "author": { "name": "Ratheesh Kannoth", "email": "rkannoth@marvell.com", "time": "Tue Feb 06 08:10:00 2024 +0530" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 12:03:02 2024 +0100" }, "message": "octeontx2-af: Initialize maps.\n\nkmalloc_array() without __GFP_ZERO flag does not initialize\nmemory to zero. This causes issues. Use kcalloc() for maps and\nbitmap_zalloc() for bitmaps.\n\nFixes: dd7842878633 (\"octeontx2-af: Add new devlink param to configure maximum usable NIX block LFs\")\nSigned-off-by: Ratheesh Kannoth \u003crkannoth@marvell.com\u003e\nReviewed-by: Brett Creeley \u003cbcreeley@amd.com\u003e\nReviewed-by: Simon Horman \u003chorms@kernel.org\u003e\nLink: https://lore.kernel.org/r/20240206024000.1070260-1-rkannoth@marvell.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "03fa49a386b298d357b90c9c5599f8d00dfc425c", "tree": "71db0a1c6f0d74e21e6ccb5b7a35a000e2f7ea5e", "parents": [ "335bac1daae3fd9070d0f9f34d7d7ba708729256", "bc4ce46b1e3d1da4309405cd4afc7c0fcddd0b90" ], "author": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 11:33:22 2024 +0100" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 11:33:22 2024 +0100" }, "message": "Merge branch \u0027cpsw-enable-mac_managed_pm-to-fix-mdio\u0027\n\nSinthu Raja says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nCPSW: enable mac_managed_pm to fix mdio\n\nThis patch fix the resume/suspend issue on CPSW interface.\n\nReference from the foloowing patchwork:\nhttps://lore.kernel.org/netdev/20221014144729.1159257-2-shenwei.wang@nxp.com/T/\n\nV1: https://patchwork.kernel.org/project/netdevbpf/patch/20240122083414.6246-1-sinthu.raja@ti.com/\nV2: https://patchwork.kernel.org/project/netdevbpf/patch/20240122093326.7618-1-sinthu.raja@ti.com/\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://lore.kernel.org/r/20240206005928.15703-1-sinthu.raja@ti.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "bc4ce46b1e3d1da4309405cd4afc7c0fcddd0b90", "tree": "71db0a1c6f0d74e21e6ccb5b7a35a000e2f7ea5e", "parents": [ "9def04e759caa5a3d741891037ae99f81e2fff01" ], "author": { "name": "Sinthu Raja", "email": "sinthu.raja@ti.com", "time": "Tue Feb 06 06:29:28 2024 +0530" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 11:33:14 2024 +0100" }, "message": "net: ethernet: ti: cpsw: enable mac_managed_pm to fix mdio\n\nThe below commit introduced a WARN when phy state is not in the states:\nPHY_HALTED, PHY_READY and PHY_UP.\ncommit 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume() state\")\n\nWhen cpsw resumes, there have port in PHY_NOLINK state, so the below\nwarning comes out. Set mac_managed_pm be true to tell mdio that the phy\nresume/suspend is managed by the mac, to fix the following warning:\n\nWARNING: CPU: 0 PID: 965 at drivers/net/phy/phy_device.c:326 mdio_bus_phy_resume+0x140/0x144\nCPU: 0 PID: 965 Comm: sh Tainted: G O 6.1.46-g247b2535b2 #1\nHardware name: Generic AM33XX (Flattened Device Tree)\n unwind_backtrace from show_stack+0x18/0x1c\n show_stack from dump_stack_lvl+0x24/0x2c\n dump_stack_lvl from __warn+0x84/0x15c\n __warn from warn_slowpath_fmt+0x1a8/0x1c8\n warn_slowpath_fmt from mdio_bus_phy_resume+0x140/0x144\n mdio_bus_phy_resume from dpm_run_callback+0x3c/0x140\n dpm_run_callback from device_resume+0xb8/0x2b8\n device_resume from dpm_resume+0x144/0x314\n dpm_resume from dpm_resume_end+0x14/0x20\n dpm_resume_end from suspend_devices_and_enter+0xd0/0x924\n suspend_devices_and_enter from pm_suspend+0x2e0/0x33c\n pm_suspend from state_store+0x74/0xd0\n state_store from kernfs_fop_write_iter+0x104/0x1ec\n kernfs_fop_write_iter from vfs_write+0x1b8/0x358\n vfs_write from ksys_write+0x78/0xf8\n ksys_write from ret_fast_syscall+0x0/0x54\nException stack(0xe094dfa8 to 0xe094dff0)\ndfa0: 00000004 005c3fb8 00000001 005c3fb8 00000004 00000001\ndfc0: 00000004 005c3fb8 b6f6bba0 00000004 00000004 0059edb8 00000000 00000000\ndfe0: 00000004 bed918f0 b6f09bd3 b6e89a66\n\nCc: \u003cstable@vger.kernel.org\u003e # v6.0+\nFixes: 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume() state\")\nFixes: fba863b81604 (\"net: phy: make PHY PM ops a no-op if MAC driver manages PHY PM\")\nSigned-off-by: Sinthu Raja \u003csinthu.raja@ti.com\u003e\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "9def04e759caa5a3d741891037ae99f81e2fff01", "tree": "7f9ea48385a5c24b37ac10cd1f8e098c7db19677", "parents": [ "335bac1daae3fd9070d0f9f34d7d7ba708729256" ], "author": { "name": "Sinthu Raja", "email": "sinthu.raja@ti.com", "time": "Tue Feb 06 06:29:27 2024 +0530" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Feb 08 11:32:53 2024 +0100" }, "message": "net: ethernet: ti: cpsw_new: enable mac_managed_pm to fix mdio\n\nThe below commit introduced a WARN when phy state is not in the states:\nPHY_HALTED, PHY_READY and PHY_UP.\ncommit 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume() state\")\n\nWhen cpsw_new resumes, there have port in PHY_NOLINK state, so the below\nwarning comes out. Set mac_managed_pm be true to tell mdio that the phy\nresume/suspend is managed by the mac, to fix the following warning:\n\nWARNING: CPU: 0 PID: 965 at drivers/net/phy/phy_device.c:326 mdio_bus_phy_resume+0x140/0x144\nCPU: 0 PID: 965 Comm: sh Tainted: G O 6.1.46-g247b2535b2 #1\nHardware name: Generic AM33XX (Flattened Device Tree)\n unwind_backtrace from show_stack+0x18/0x1c\n show_stack from dump_stack_lvl+0x24/0x2c\n dump_stack_lvl from __warn+0x84/0x15c\n __warn from warn_slowpath_fmt+0x1a8/0x1c8\n warn_slowpath_fmt from mdio_bus_phy_resume+0x140/0x144\n mdio_bus_phy_resume from dpm_run_callback+0x3c/0x140\n dpm_run_callback from device_resume+0xb8/0x2b8\n device_resume from dpm_resume+0x144/0x314\n dpm_resume from dpm_resume_end+0x14/0x20\n dpm_resume_end from suspend_devices_and_enter+0xd0/0x924\n suspend_devices_and_enter from pm_suspend+0x2e0/0x33c\n pm_suspend from state_store+0x74/0xd0\n state_store from kernfs_fop_write_iter+0x104/0x1ec\n kernfs_fop_write_iter from vfs_write+0x1b8/0x358\n vfs_write from ksys_write+0x78/0xf8\n ksys_write from ret_fast_syscall+0x0/0x54\nException stack(0xe094dfa8 to 0xe094dff0)\ndfa0: 00000004 005c3fb8 00000001 005c3fb8 00000004 00000001\ndfc0: 00000004 005c3fb8 b6f6bba0 00000004 00000004 0059edb8 00000000 00000000\ndfe0: 00000004 bed918f0 b6f09bd3 b6e89a66\n\nCc: \u003cstable@vger.kernel.org\u003e # v6.0+\nFixes: 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume() state\")\nFixes: fba863b81604 (\"net: phy: make PHY PM ops a no-op if MAC driver manages PHY PM\")\nSigned-off-by: Sinthu Raja \u003csinthu.raja@ti.com\u003e\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "2526dffc6d65cffa32b88556bd68e4e72e889a55", "tree": "e0c859f3b7afe2fbc2eba5c9bc5529c69435d9dc", "parents": [ "54be6c6c5ae8e0d93a6c4641cb7528eb0b6ba478" ], "author": { "name": "Bartosz Golaszewski", "email": "bartosz.golaszewski@linaro.org", "time": "Mon Feb 05 11:22:29 2024 +0100" }, "committer": { "name": "Bartosz Golaszewski", "email": "bartosz.golaszewski@linaro.org", "time": "Thu Feb 08 10:33:03 2024 +0100" }, "message": "gpio: remove GPIO device from the list unconditionally in error path\n\nSince commit 48e1b4d369cf (\"gpiolib: remove the GPIO device from the list\nwhen it\u0027s unregistered\") we remove the GPIO device entry from the global\nlist (used to order devices by their GPIO ranges) when unregistering the\nchip, not when releasing the device. It will not happen when the last\nreference is put anymore. This means, we need to remove it in error path\nin gpiochip_add_data_with_key() unconditionally, without checking if the\ndevice\u0027s .release() callback is set.\n\nFixes: 48e1b4d369cf (\"gpiolib: remove the GPIO device from the list when it\u0027s unregistered\")\nSigned-off-by: Bartosz Golaszewski \u003cbartosz.golaszewski@linaro.org\u003e\n" }, { "commit": "ab0beafd52b98dfb8b8244b2c6794efbc87478db", "tree": "7257cbf2b7f87d321794b4ae7de3dcc37b9c0f79", "parents": [ "d694b754894c93fb4d71a7f3699439dec111decc" ], "author": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Fri Feb 02 10:09:34 2024 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Thu Feb 08 10:27:41 2024 +0100" }, "message": "netfilter: nft_set_pipapo: remove static in nft_pipapo_get()\n\nThis has slipped through when reducing memory footprint for set\nelements, remove it.\n\nFixes: 9dad402b89e8 (\"netfilter: nf_tables: expose opaque set element as struct nft_elem_priv\")\nReported-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "bf4c27b8267d7848bb81fd41e6aa07aa662f07fb", "tree": "8d9e5056dc5cd30edfa601db84c2358bc6e591b5", "parents": [ "9e3fc1d65d4e8cf302e289847ab165ad9358fdb2" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Mon Feb 05 20:50:10 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:51:19 2024 +0100" }, "message": "drm/xe: Remove TEST_VM_ASYNC_OPS_ERROR\n\nTEST_VM_ASYNC_OPS_ERROR is broken and unused. Remove for now and will\npull back in a later time when it is used, fixed, and properly hidden\nbehind a Kconfig option. Also fixup the supported flags value.\n\nFixes: dd08ebf6c352 (\"drm/xe: Introduce a new DRM driver for Intel GPUs\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240206045010.2981051-1-matthew.brost@intel.com\n(cherry picked from commit d9890c028d66a9e1ee3cccaa081ab5aedcbfe431)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "9e3fc1d65d4e8cf302e289847ab165ad9358fdb2", "tree": "cfd774f26d8911d6480a4f7a079db36341ab6cb7", "parents": [ "95c058c8ef1d5d9e39ab2039a5eea4d5b93f4117" ], "author": { "name": "Matthew Auld", "email": "matthew.auld@intel.com", "time": "Fri Feb 02 17:14:36 2024 +0000" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:51:04 2024 +0100" }, "message": "drm/xe/vm: don\u0027t ignore error when in_kthread\n\nIf GUP fails and we are in_kthread, we can have pinned \u003d 0 and ret \u003d 0.\nIf that happens we call sg_alloc_append_table_from_pages() with n_pages\n\u003d 0, which is not well behaved and can trigger:\n\nkernel BUG at include/linux/scatterlist.h:115!\n\ndepending on if the pages array happens to be zeroed or not. Even if we\ndon\u0027t hit that it crashes later when trying to dma_map the returned\ntable.\n\nFixes: dd08ebf6c352 (\"drm/xe: Introduce a new DRM driver for Intel GPUs\")\nSigned-off-by: Matthew Auld \u003cmatthew.auld@intel.com\u003e\nCc: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\nCc: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240202171435.427630-2-matthew.auld@intel.com\n(cherry picked from commit 8087199cd5951c1eba26003b3e4296dbb2110adf)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "95c058c8ef1d5d9e39ab2039a5eea4d5b93f4117", "tree": "558313e250f0074288783b7784bb9c2f08bcc7fe", "parents": [ "11572b3f68d9933fef5c1afef4c20041701d8025" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Mon Feb 05 15:17:14 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:50:50 2024 +0100" }, "message": "drm/xe: Assume large page size if VMA not yet bound\n\nThe calculation to determine max page size of a VMA during a REMAP\noperations assumes the VMA has been bound. This assumption is not true\nif the VMA is from an eariler operation in an array of binds. If a VMA\nhas not been bound use the maximum page size which will ensure the\nprevious / next REMAP operations are not incorrectly skipped.\n\nFixes: 8f33b4f054fc (\"drm/xe: Avoid doing rebinds\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240205231714.2956225-1-matthew.brost@intel.com\n(cherry picked from commit 5ad6af5c91e9b942c44b657122270d935db3a813)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "11572b3f68d9933fef5c1afef4c20041701d8025", "tree": "db49ef66a601868a3d186ed1aa743b36de29794b", "parents": [ "3aa3c5c249086ffc920e8f6d6a15bdd441153d45" ], "author": { "name": "Xiaoming Wang", "email": "xiaoming.wang@intel.com", "time": "Fri Feb 02 13:56:58 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:50:20 2024 +0100" }, "message": "drm/xe/display: Fix memleak in display initialization\n\nintel_power_domains_init is called twice in xe_device_probe:\n\n1) intel_power_domains_init()\n xe_display_init_nommio()\n xe_device_probe()\n\n2) intel_power_domains_init()\n intel_display_driver_probe_noirq()\n xe_display_init_noirq()\n xe_device_probe()\n\nIt needs remove one to avoid power_domains-\u003epower_wells double malloc.\n\nunreferenced object 0xffff88811150ee00 (size 512):\n comm \"systemd-udevd\", pid 506, jiffies 4294674198 (age 3605.560s)\n hex dump (first 32 bytes):\n 10 b4 9d a0 ff ff ff ff ff ff ff ff ff ff ff ff ................\n ff ff ff ff ff ff ff ff 00 00 00 00 00 00 00 00 ................\n backtrace:\n [\u003cffffffff8134b901\u003e] __kmem_cache_alloc_node+0x1c1/0x2b0\n [\u003cffffffff812c98b2\u003e] __kmalloc+0x52/0x150\n [\u003cffffffffa08b0033\u003e] __set_power_wells+0xc3/0x360 [xe]\n [\u003cffffffffa08562fc\u003e] xe_display_init_nommio+0x4c/0x70 [xe]\n [\u003cffffffffa07f0d1c\u003e] xe_device_probe+0x3c/0x5a0 [xe]\n [\u003cffffffffa082e48f\u003e] xe_pci_probe+0x33f/0x5a0 [xe]\n [\u003cffffffff817f2187\u003e] local_pci_probe+0x47/0xa0\n [\u003cffffffff817f3db3\u003e] pci_device_probe+0xc3/0x1f0\n [\u003cffffffff8192f2a2\u003e] really_probe+0x1a2/0x410\n [\u003cffffffff8192f598\u003e] __driver_probe_device+0x78/0x160\n [\u003cffffffff8192f6ae\u003e] driver_probe_device+0x1e/0x90\n [\u003cffffffff8192f92a\u003e] __driver_attach+0xda/0x1d0\n [\u003cffffffff8192c95c\u003e] bus_for_each_dev+0x7c/0xd0\n [\u003cffffffff8192e159\u003e] bus_add_driver+0x119/0x220\n [\u003cffffffff81930d00\u003e] driver_register+0x60/0x120\n [\u003cffffffffa05e50a0\u003e] 0xffffffffa05e50a0\n\nThe call to intel_power_domains_cleanup() needs to stay where it is for\nnow. The main issue is that while the init is called by the display\nside, shared by i915 and xe, the cleanup is called by a non-shared code\npath. Fixing that will be done as a separate commit.\n\nFixes: 44e694958b95 (\"drm/xe/display: Implement display support\")\nSigned-off-by: Xiaoming Wang \u003cxiaoming.wang@intel.com\u003e\n[ reword commit message and explain why the fini needs to stay\n where it is ]\nReviewed-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nSigned-off-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240202215658.561298-1-lucas.demarchi@intel.com\n(cherry picked from commit 86c99abb5f1b6fcd69fb268eeb2e34cb7c4f355c)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "3aa3c5c249086ffc920e8f6d6a15bdd441153d45", "tree": "2c68c9aa2314c6c4a576851aa9254c7bbeb96837", "parents": [ "90773aaf9129ea6f47915bd3c47da261abe6a447" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Thu Feb 01 19:34:40 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:49:59 2024 +0100" }, "message": "drm/xe: Map both mem.kernel_bb_pool and usm.bb_pool\n\nFor integrated devices we need to map both mem.kernel_bb_pool and\nusm.bb_pool to be able to run batches from both pools.\n\nFixes: a682b6a42d4d (\"drm/xe: Support device page faults on integrated platforms\")\nTested-by: Brian Welty \u003cbrian.welty@intel.com\u003e\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Brian Welty \u003cbrian.welty@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240202033440.2351862-1-matthew.brost@intel.com\n(cherry picked from commit 72f86ed3c88933d6fa09b036de93621ea71097a7)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "90773aaf9129ea6f47915bd3c47da261abe6a447", "tree": "40b23de68da004d0c705e34f1ccd71b28917b333", "parents": [ "21abf108a062fa0323077b5ba3d26e2c0bba9232" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Wed Jan 03 12:48:02 2024 +0100" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:49:41 2024 +0100" }, "message": "drm/xe: circumvent bogus stringop-overflow warning\n\ngcc-13 warns about an array overflow that it sees but that is\nprevented by the \"asid % NUM_PF_QUEUE\" calculation:\n\ndrivers/gpu/drm/xe/xe_gt_pagefault.c: In function \u0027xe_guc_pagefault_handler\u0027:\ninclude/linux/fortify-string.h:57:33: error: writing 16 bytes into a region of size 0 [-Werror\u003dstringop-overflow\u003d]\ninclude/linux/fortify-string.h:689:26: note: in expansion of macro \u0027__fortify_memcpy_chk\u0027\n 689 | #define memcpy(p, q, s) __fortify_memcpy_chk(p, q, s, \\\n | ^~~~~~~~~~~~~~~~~~~~\ndrivers/gpu/drm/xe/xe_gt_pagefault.c:341:17: note: in expansion of macro \u0027memcpy\u0027\n 341 | memcpy(pf_queue-\u003edata + pf_queue-\u003etail, msg, len * sizeof(u32));\n | ^~~~~~\ndrivers/gpu/drm/xe/xe_gt_types.h:102:25: note: at offset [1144, 265324] into destination object \u0027tile\u0027 of size 8\n\nI found that rewriting the assignment using pointer addition rather than the\nequivalent array index calculation prevents the warning, so use that instead.\n\nI sent a bug report against gcc for the false positive warning.\n\nFixes: dd08ebf6c352 (\"drm/xe: Introduce a new DRM driver for Intel GPUs\")\nLink: https://gcc.gnu.org/bugzilla/show_bug.cgi?id\u003d113214\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240103114819.2913937-1-arnd@kernel.org\n(cherry picked from commit 774ef5dfc95578a9079426d5106076dcd59c4dfa)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "21abf108a062fa0323077b5ba3d26e2c0bba9232", "tree": "bead2f4e4420b41b19ddcbea5bc68bb692cbe792", "parents": [ "fc29b6d5ab5395dcb9f35de71e0347f3a6bca542" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Wed Jan 31 16:48:49 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:49:29 2024 +0100" }, "message": "drm/xe: Pick correct userptr VMA to repin on REMAP op failure\n\nA REMAP op is composed of 3 VMA\u0027s - unmap, prev map, and next map. When\nop_execute fails with -EAGAIN we need to update the local VMA pointer to\nthe current op state and then repin the VMA if it is a userptr.\n\nFixes a failure seen in xe_vm.munmap-style-unbind-userptr-one-partial.\n\nFixes: b06d47be7c83 (\"drm/xe: Port Xe to GPUVA\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240201004849.2219558-3-matthew.brost@intel.com\n(cherry picked from commit 447f74d223b4f6cbab74963bf1099050c15374ce)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "fc29b6d5ab5395dcb9f35de71e0347f3a6bca542", "tree": "272df24e98fe1b3c712d9684ffe24f9875a3766d", "parents": [ "ddc7d4c584704666fe7088bbd9ec2d72d0f63e65" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Wed Jan 31 16:48:48 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:49:22 2024 +0100" }, "message": "drm/xe: Take a reference in xe_exec_queue_last_fence_get()\n\nTake a reference in xe_exec_queue_last_fence_get(). Also fix a reference\ncounting underflow bug VM bind and unbind.\n\nFixes: dd08ebf6c352 (\"drm/xe: Introduce a new DRM driver for Intel GPUs\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240201004849.2219558-2-matthew.brost@intel.com\n(cherry picked from commit a856b67a84169e065ebbeee50258936b1eacc9eb)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "ddc7d4c584704666fe7088bbd9ec2d72d0f63e65", "tree": "a9c3964669d6ccf9fbed24e2dd32fac08386a51c", "parents": [ "54be6c6c5ae8e0d93a6c4641cb7528eb0b6ba478" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Thu Feb 01 09:55:32 2024 -0800" }, "committer": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Feb 08 09:49:10 2024 +0100" }, "message": "drm/xe: Fix loop in vm_bind_ioctl_ops_unwind\n\nThe logic for the unwind loop is incorrect resulting in an infinite\nloop. Fix to unwind to go from the last operations list to he first.\n\nFixes: 617eebb9c480 (\"drm/xe: Fix array of binds\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20240201175532.2303168-1-matthew.brost@intel.com\n(cherry picked from commit 3acc1ff1a72fce00cdbd3ef1c27108a967fd5616)\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\n" }, { "commit": "047371968ffc470769f541d6933e262dc7085456", "tree": "b1ce900e42494b864dc721932ff8239adf2d5232", "parents": [ "860d7dcb20105af1fc7228a162886c421296cd86", "24c890dd712f6345e382256cae8c97abb0406b70" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 06:12:14 2024 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 06:12:14 2024 +0000" }, "message": "Merge tag \u0027v6.8-p3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6\n\nPull crypto fixes from Herbert Xu:\n \"Fix regressions in cbc and algif_hash, as well as an older\n NULL-pointer dereference in ccp\"\n\n* tag \u0027v6.8-p3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:\n crypto: algif_hash - Remove bogus SGL free on zero-length error path\n crypto: cbc - Ensure statesize is zero\n crypto: ccp - Fix null pointer dereference in __sev_platform_shutdown_locked\n" }, { "commit": "860d7dcb20105af1fc7228a162886c421296cd86", "tree": "b819855c890c4abdac8d9ec43204e4d2ac61458c", "parents": [ "547ab8fc4cb04a1a6b34377dd8fad34cd2c8a8e3", "ebd4acc0cbeae9efea15993b11b05bd32942f3f0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 06:08:37 2024 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 08 06:08:37 2024 +0000" }, "message": "Merge tag \u0027percpu-for-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dennis/percpu\n\nPull percpu fix from Dennis Zhou:\n\n - fix riscv wrong size passed to local_flush_tlb_range_asid()\n\n* tag \u0027percpu-for-6.8-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dennis/percpu:\n riscv: Fix wrong size passed to local_flush_tlb_range_asid()\n" }, { "commit": "5bc09b397cbf1221f8a8aacb1152650c9195b02b", "tree": "fdf36bc0a847f5c100fd5456ac310d863dbda22b", "parents": [ "b9e4bc1046d20e0623a80660ef8627448056f817" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@gmail.com", "time": "Sun Feb 04 01:16:45 2024 +0900" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Wed Feb 07 21:20:37 2024 -0800" }, "message": "nilfs2: fix potential bug in end_buffer_async_write\n\nAccording to a syzbot report, end_buffer_async_write(), which handles the\ncompletion of block device writes, may detect abnormal condition of the\nbuffer async_write flag and cause a BUG_ON failure when using nilfs2.\n\nNilfs2 itself does not use end_buffer_async_write(). But, the async_write\nflag is now used as a marker by commit 7f42ec394156 (\"nilfs2: fix issue\nwith race condition of competition between segments for dirty blocks\") as\na means of resolving double list insertion of dirty blocks in\nnilfs_lookup_dirty_data_buffers() and nilfs_lookup_node_buffers() and the\nresulting crash.\n\nThis modification is safe as long as it is used for file data and b-tree\nnode blocks where the page caches are independent. However, it was\nirrelevant and redundant to also introduce async_write for segment summary\nand super root blocks that share buffers with the backing device. This\nled to the possibility that the BUG_ON check in end_buffer_async_write\nwould fail as described above, if independent writebacks of the backing\ndevice occurred in parallel.\n\nThe use of async_write for segment summary buffers has already been\nremoved in a previous change.\n\nFix this issue by removing the manipulation of the async_write flag for\nthe remaining super root block buffer.\n\nLink: https://lkml.kernel.org/r/20240203161645.4992-1-konishi.ryusuke@gmail.com\nFixes: 7f42ec394156 (\"nilfs2: fix issue with race condition of competition between segments for dirty blocks\")\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@gmail.com\u003e\nReported-by: syzbot+5c04210f7c7f897c1e7f@syzkaller.appspotmail.com\nCloses: https://lkml.kernel.org/r/00000000000019a97c05fd42f8c8@google.com\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" } ], "next": "b9e4bc1046d20e0623a80660ef8627448056f817" }