)]}' { "log": [ { "commit": "6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f", "tree": "c5ffb71a1a0bf7beb138c6bf3e9e2621313abe29", "parents": [ "fbf33803618ad4f531f78fe15cf328fe6c7f9978" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:18:59 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:18:59 2026 -0800" }, "message": "Linux 7.0-rc1\n" }, { "commit": "fbf33803618ad4f531f78fe15cf328fe6c7f9978", "tree": "321a5d344815c73a2471590e09e198e501fe499e", "parents": [ "75e1f66a9ed09f29c6883ea379c174e8cf31f7cc", "693680b9add63dbebb2505a553ff52f8c706c8c0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:12:04 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:12:04 2026 -0800" }, "message": "Merge tag \u0027fsverity-for-linus\u0027 of git://git.kernel.org/pub/scm/fs/fsverity/linux\n\nPull fsverity fixes from Eric Biggers:\n\n - Fix a build error on parisc\n\n - Remove the non-large-folio-aware function fsverity_verify_page()\n\n* tag \u0027fsverity-for-linus\u0027 of git://git.kernel.org/pub/scm/fs/fsverity/linux:\n fsverity: fix build error by adding fsverity_readahead() stub\n fsverity: remove fsverity_verify_page()\n f2fs: make f2fs_verify_cluster() partially large-folio-aware\n f2fs: remove unnecessary ClearPageUptodate in f2fs_verify_cluster()\n" }, { "commit": "75e1f66a9ed09f29c6883ea379c174e8cf31f7cc", "tree": "396c12dec90c6404bc7d93d1ea94e149dbd23309", "parents": [ "aaf96df9593bf4ab1b73c17891e4efe7570fdef3", "beeebffc807531f69445356180238500f56951cc" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:09:33 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 13:09:33 2026 -0800" }, "message": "Merge tag \u0027libcrypto-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux\n\nPull crypto library fix from Eric Biggers:\n \"Fix a big endian specific issue in the PPC64-optimized AES code\"\n\n* tag \u0027libcrypto-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux:\n lib/crypto: powerpc/aes: Fix rndkey_from_vsx() on big endian CPUs\n" }, { "commit": "aaf96df9593bf4ab1b73c17891e4efe7570fdef3", "tree": "3d76982f4ead106d876cf4eb91e3f4f7c3d8ff23", "parents": [ "746b9ef5d5ccbded13bdc1f9575fb587fe13794e" ], "author": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Tue Feb 17 13:10:46 2026 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 12:11:33 2026 -0800" }, "message": "CREDITS: Add -next to Stephen Rothwell\u0027s entry\n\nStephen retired and stepped back from -next maintainership, update his\nentry in CREDITS to recognise his 18 years of hard work making it what\nit is today and all the impact it\u0027s had on our development process.\n\nAlso update to his current GnuPG key while we\u0027re here.\n\nAcked-by: Stephen Rothwell \u003csfr@canb.auug.org.au\u003e\nAcked-by: SeongJae Park \u003csj@kernel.org\u003e\nReviewed-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\nAcked-by: Krzysztof Kozlowski \u003ckrzk@kernel.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "746b9ef5d5ccbded13bdc1f9575fb587fe13794e", "tree": "e093f15f374e4abd87c74047dc3eea029e47a762", "parents": [ "fd1d6b9d13f35dccbacbae25ed53593cd9086f84" ], "author": { "name": "Arnd Bergmann", "email": "arnd@kernel.org", "time": "Tue Feb 17 08:26:49 2026 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 12:09:23 2026 -0800" }, "message": "x509: select CONFIG_CRYPTO_LIB_SHA256\n\nThe x509 public key code gained a dependency on the sha256 hash\nimplementation, causing a rare link time failure in randconfig\nbuilds:\n\n arm-linux-gnueabi-ld: crypto/asymmetric_keys/x509_public_key.o: in function `x509_get_sig_params\u0027:\n x509_public_key.c:(.text.x509_get_sig_params+0x12): undefined reference to `sha256\u0027\n arm-linux-gnueabi-ld: (sha256): Unknown destination type (ARM/Thumb) in crypto/asymmetric_keys/x509_public_key.o\n x509_public_key.c:(.text.x509_get_sig_params+0x12): dangerous relocation: unsupported relocation\n\nSelect the necessary library code from Kconfig.\n\nFixes: 2c62068ac86b (\"x509: Separately calculate sha256 for blacklist\")\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: David Howells \u003cdhowells@redhat.com\u003e\nReviewed-by: Eric Biggers \u003cebiggers@kernel.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "fd1d6b9d13f35dccbacbae25ed53593cd9086f84", "tree": "78b00798727e519dc815f5ca7832634305470bf1", "parents": [ "5f2eac7767493bf23e6552db82ab25de0dccd54f" ], "author": { "name": "Haiyue Wang", "email": "haiyuewa@163.com", "time": "Sun Feb 22 20:11:00 2026 +0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 12:05:31 2026 -0800" }, "message": "xz: fix arm fdt compile error for kmalloc replacement\n\nAlign to the commit bf4afc53b77a (\"Convert \u0027alloc_obj\u0027 family to use the\nnew default GFP_KERNEL argument\") update the \u0027kmalloc_obj\u0027 declaration\nfor userspace to fix below compile error:\n\n In file included from arch/arm/boot/compressed/../../../../lib/decompress_unxz.c:241,\n from arch/arm/boot/compressed/decompress.c:56:\n arch/arm/boot/compressed/../../../../lib/xz/xz_dec_stream.c: In function \u0027xz_dec_init\u0027:\n arch/arm/boot/compressed/../../../../lib/xz/xz_dec_stream.c:787:28: error: implicit declaration of function \u0027kmalloc_obj\u0027; did you mean \u0027kmalloc\u0027? [-Wimplicit-function-declaration]\n 787 | struct xz_dec *s \u003d kmalloc_obj(*s);\n | ^~~~~~~~~~~\n | kmalloc\n\nSigned-off-by: Haiyue Wang \u003chaiyuewa@163.com\u003e\nFixes: 69050f8d6d07 (\"treewide: Replace kmalloc with kmalloc_obj for non-scalar types\")\nFixes: bf4afc53b77a (\"Convert \u0027alloc_obj\u0027 family to use the new default GFP_KERNEL argument\")\nReviewed-by: Kees Cook \u003ckees@kernel.org\u003e\nAcked-by: Lasse Collin \u003classe.collin@tukaani.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5f2eac7767493bf23e6552db82ab25de0dccd54f", "tree": "255c7ca204c0daf0597b9135012b14208d0d7922", "parents": [ "1dd419145d090f8fdf149cbb39dea6d968659dd2", "969c3cca0f3b88682cd833cee4cf01b0915629a3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 09:43:11 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 09:43:11 2026 -0800" }, "message": "Merge tag \u0027rtc-7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux\n\nPull RTC updates from Alexandre Belloni:\n\n - loongson: Loongson-2K0300 support\n\n - s35390a: nvmem support\n\n - zynqmp: rework calibration\n\n* tag \u0027rtc-7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:\n rtc: ds1390: fix number of bytes read from RTC\n rtc: class: Remove duplicate check for alarm\n rtc: optee: simplify OP-TEE context match\n rtc: interface: Alarm race handling should not discard preceding error\n rtc: s35390a: implement nvmem support\n rtc: loongson: Add Loongson-2K0300 support\n dt-bindings: rtc: loongson: Document Loongson-2K0300 compatible\n dt-bindings: rtc: loongson: Correct Loongson-1C interrupts property\n dt-bindings: rtc: renesas,rz-rtca3: Add RZ/V2N support\n dt-bindings: rtc: cpcap: convert to schema\n rtc: zynqmp: use dynamic max and min offset ranges\n rtc: zynqmp: rework set_offset\n rtc: zynqmp: rework read_offset\n rtc: zynqmp: check calibration max value\n rtc: zynqmp: correct frequency value\n rtc: amlogic-a4: Remove IRQF_ONESHOT\n rtc: pcf8563: use correct of_node for output clock\n rtc: max31335: use correct CONFIG symbol in IS_REACHABLE()\n rtc: nvvrs: Add ARCH_TEGRA to the NV VRS RTC driver\n" }, { "commit": "1dd419145d090f8fdf149cbb39dea6d968659dd2", "tree": "c5390477ec8356046ba72bb88bf053148a0b9a6d", "parents": [ "d2ba6e9c0ae54b3d0973e23d8806cd9a16b9faef", "97b281d7edb2ae662365be2809cd728470119720" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 08:43:31 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 08:43:31 2026 -0800" }, "message": "Merge tag \u0027rust-fixes-7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux\n\nPull rust fixes from Miguel Ojeda:\n \"Toolchain and infrastructure:\n\n - Pass \u0027-Zunstable-options\u0027 flag required by the future Rust 1.95.0\n\n - Fix \u0027objtool\u0027 warning for Rust 1.84.0\n\n \u0027kernel\u0027 crate:\n\n - \u0027irq\u0027 module: add missing bound detected by the future Rust 1.95.0\n\n - \u0027list\u0027 module: add missing \u0027unsafe\u0027 blocks and placeholder safety\n comments to macros (an issue for future callers within the crate)\n\n \u0027pin-init\u0027 crate:\n\n - Clean Clippy warning that changed behavior in the future Rust\n 1.95.0\"\n\n* tag \u0027rust-fixes-7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux:\n rust: list: Add unsafe blocks for container_of and safety comments\n rust: pin-init: replace clippy `expect` with `allow`\n rust: irq: add `\u0027static` bounds to irq callbacks\n objtool/rust: add one more `noreturn` Rust function\n rust: kbuild: pass `-Zunstable-options` for Rust 1.95.0\n" }, { "commit": "d2ba6e9c0ae54b3d0973e23d8806cd9a16b9faef", "tree": "91c968b293638433f4b7ac6066588c92e4e0cd4c", "parents": [ "189f164e573e18d9f8876dbd3ad8fcbe11f93037", "75f3cf0d1561b0a1cfa97ee35d78ca0118e584d5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 08:40:13 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 08:40:13 2026 -0800" }, "message": "Merge tag \u0027trace-rv-7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull runtime verifier fix from Steven Rostedt:\n\n - Fix multiple definition of __pcpu_unique_da_mon_this\n\n After refactoring monitors, we used static per-cpu variables with the\n same names across different per-cpu monitors. This is explicitly\n disallowed for modules on some architectures (alpha) or if\n CONFIG_DEBUG_FORCE_WEAK_PER_CPU is enabled (e.g. Fedora\u0027s debug\n kernel). Make sure all those variables have different names to avoid\n compilation issues.\n\n* tag \u0027trace-rv-7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n rv: Fix multiple definition of __pcpu_unique_da_mon_this\n" }, { "commit": "189f164e573e18d9f8876dbd3ad8fcbe11f93037", "tree": "733382723a15ea29b7c9f3dde3bcb281f0b3c287", "parents": [ "32a92f8c89326985e05dce8b22d3f0aa07a3e1bd" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Sat Feb 21 23:46:04 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Feb 22 08:26:33 2026 -0800" }, "message": "Convert remaining multi-line kmalloc_obj/flex GFP_KERNEL uses\n\nConversion performed via this Coccinelle script:\n\n // SPDX-License-Identifier: GPL-2.0-only\n // Options: --include-headers-for-types --all-includes --include-headers --keep-comments\n virtual patch\n\n @gfp depends on patch \u0026\u0026 !(file in \"tools\") \u0026\u0026 !(file in \"samples\")@\n identifier ALLOC \u003d {kmalloc_obj,kmalloc_objs,kmalloc_flex,\n \t\t kzalloc_obj,kzalloc_objs,kzalloc_flex,\n\t\t kvmalloc_obj,kvmalloc_objs,kvmalloc_flex,\n\t\t kvzalloc_obj,kvzalloc_objs,kvzalloc_flex};\n @@\n\n \tALLOC(...\n -\t\t, GFP_KERNEL\n \t)\n\n $ make coccicheck MODE\u003dpatch COCCI\u003dgfp.cocci\n\nBuild and boot tested x86_64 with Fedora 42\u0027s GCC and Clang:\n\nLinux version 6.19.0+ (user@host) (gcc (GCC) 15.2.1 20260123 (Red Hat 15.2.1-7), GNU ld version 2.44-12.fc42) #1 SMP PREEMPT_DYNAMIC 1970-01-01\nLinux version 6.19.0+ (user@host) (clang version 20.1.8 (Fedora 20.1.8-4.fc42), LLD 20.1.8) #1 SMP PREEMPT_DYNAMIC 1970-01-01\n\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "32a92f8c89326985e05dce8b22d3f0aa07a3e1bd", "tree": "65f84985b9ed2d5cf3c5243aca78d9428e25c312", "parents": [ "323bbfcf1ef8836d0d2ad9e2c1f1c684f0e3b5b3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 20:03:00 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 20:03:00 2026 -0800" }, "message": "Convert more \u0027alloc_obj\u0027 cases to default GFP_KERNEL arguments\n\nThis converts some of the visually simpler cases that have been split\nover multiple lines. I only did the ones that are easy to verify the\nresulting diff by having just that final GFP_KERNEL argument on the next\nline.\n\nSomebody should probably do a proper coccinelle script for this, but for\nme the trivial script actually resulted in an assertion failure in the\nmiddle of the script. I probably had made it a bit _too_ trivial.\n\nSo after fighting that far a while I decided to just do some of the\nsyntactically simpler cases with variations of the previous \u0027sed\u0027\nscripts.\n\nThe more syntactically complex multi-line cases would mostly really want\nwhitespace cleanup anyway.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "323bbfcf1ef8836d0d2ad9e2c1f1c684f0e3b5b3", "tree": "515dca34daa8d18ec26c2c3da3096c6c23955c95", "parents": [ "bf4afc53b77aeaa48b5409da5c8da6bb4eff7f43" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 17:06:51 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 17:09:51 2026 -0800" }, "message": "Convert \u0027alloc_flex\u0027 family to use the new default GFP_KERNEL argument\n\nThis is the exact same thing as the \u0027alloc_obj()\u0027 version, only much\nsmaller because there are a lot fewer users of the *alloc_flex()\ninterface.\n\nAs with alloc_obj() version, this was done entirely with mindless brute\nforce, using the same script, except using \u0027flex\u0027 in the pattern rather\nthan \u0027objs*\u0027.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "bf4afc53b77aeaa48b5409da5c8da6bb4eff7f43", "tree": "01fdd9d27f1b272bef0127966e08eac44d134d0a", "parents": [ "e19e1b480ac73c3e62ffebbca1174f0f511f43e7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 16:37:42 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 17:09:51 2026 -0800" }, "message": "Convert \u0027alloc_obj\u0027 family to use the new default GFP_KERNEL argument\n\nThis was done entirely with mindless brute force, using\n\n git grep -l \u0027\\\u003ck[vmz]*alloc_objs*(.*, GFP_KERNEL)\u0027 |\n xargs sed -i \u0027s/\\(alloc_objs*(.*\\), GFP_KERNEL)/\\1)/\u0027\n\nto convert the new alloc_obj() users that had a simple GFP_KERNEL\nargument to just drop that argument.\n\nNote that due to the extreme simplicity of the scripting, any slightly\nmore complex cases spread over multiple lines would not be triggered:\nthey definitely exist, but this covers the vast bulk of the cases, and\nthe resulting diff is also then easier to check automatically.\n\nFor the same reason the \u0027flex\u0027 versions will be done as a separate\nconversion.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "e19e1b480ac73c3e62ffebbca1174f0f511f43e7", "tree": "4f83448fd9923f10b4bc4caff48785df5b1c23b9", "parents": [ "fa5c82f4d2bbde10e9fd3a32aecacfe3813919ba" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 16:14:11 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 17:09:50 2026 -0800" }, "message": "add default_gfp() helper macro and use it in the new *alloc_obj() helpers\n\nMost simple allocations use GFP_KERNEL, and with the new allocation\nhelpers being introduced, let\u0027s just take advantage of that to simplify\nthat default case.\n\nIt\u0027s a numbers game:\n\n git grep \u0027alloc_obj(\u0027 |\n\tsed \u0027s/.*\\(GFP_[_A-Z]*\\).*/\\1/\u0027 |\n\tsort | uniq -c | sort -n | tail\n\nshows that about 90% of all those new allocator instances just use that\nstandard GFP_KERNEL.\n\nThose helpers are already macros, and we can easily just make it be the\ndefault case when the gfp argument is missing.\n\nAnd yes, we could do that for all the legacy interfaces too, but let\u0027s\nkeep it to just the new ones at least for now, since those all got\nconverted recently anyway, so this is not any \"extra\" noise outside of\nthat limited conversion.\n\nAnd, in fact, I want to do this before doing the -rc1 release, exactly\nso that we don\u0027t get extra merge conflicts.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "fa5c82f4d2bbde10e9fd3a32aecacfe3813919ba", "tree": "5b67eac20681c85579c4af0236c92bc32ddf31ad", "parents": [ "8934827db5403eae57d4537114a9ff88b0a8460f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 15:12:09 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 15:12:09 2026 -0800" }, "message": "slab.h: disable completely broken overflow handling in flex allocations\n\nCommit 69050f8d6d07 (\"treewide: Replace kmalloc with kmalloc_obj for\nnon-scalar types\") started using the new allocation helpers, and in the\nprocess showed that they were completely non-working.\n\nThe overflow logic in overflows_flex_counter_type() is completely the\nwrong way around, and that broke __alloc_flex() completely. By chance,\nthe resulting code was then such a mess that clang generated\nsufficiently garbage code that objtool warned about it all. Which made\nit somewhat quicker to narrow things down.\n\nWhile fixing overflows_flex_counter_type() would presumably fix this\nall, I\u0027m excising the whole broken overflow logic from __alloc_flex(),\nbecause we don\u0027t want that kind of code in basic allocation functions\nanyway.\n\nThat (no longer) broken overflows_flex_counter_type() thing needs to be\ninserted into the actual __set_flex_counter() logic in the unlikely case\nthat we ever want this at all. And made conditional.\n\nFixes: 81cee9166a90 (\"compiler_types: Introduce __flex_counter() and family\")\nFixes: 69050f8d6d07 (\"treewide: Replace kmalloc with kmalloc_obj for non-scalar types\")\nCc: Kees Cook \u003ckees@kernel.org\u003e\nLink: https://lore.kernel.org/all/CAHk-\u003dwhEd020BYzGTzYrENjD9Z5_82xx6h8HsQvH5xDSnv0\u003dHw@mail.gmail.com/\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "8934827db5403eae57d4537114a9ff88b0a8460f", "tree": "5167aa7e16b786b9135e19d508b234054fa6e8ce", "parents": [ "c7decec2f2d2ab0366567f9e30c0e1418cece43f", "7a70c15bd1449f1eb30991772edce37b41e496fb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 11:02:58 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 11:02:58 2026 -0800" }, "message": "Merge tag \u0027kmalloc_obj-treewide-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux\n\nPull kmalloc_obj conversion from Kees Cook:\n \"This does the tree-wide conversion to kmalloc_obj() and friends using\n coccinelle, with a subsequent small manual cleanup of whitespace\n alignment that coccinelle does not handle.\n\n This uncovered a clang bug in __builtin_counted_by_ref(), so the\n conversion is preceded by disabling that for current versions of\n clang. The imminent clang 22.1 release has the fix.\n\n I\u0027ve done allmodconfig build tests for x86_64, arm64, i386, and arm. I\n did defconfig builds for alpha, m68k, mips, parisc, powerpc, riscv,\n s390, sparc, sh, arc, csky, xtensa, hexagon, and openrisc\"\n\n* tag \u0027kmalloc_obj-treewide-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:\n kmalloc_obj: Clean up after treewide replacements\n treewide: Replace kmalloc with kmalloc_obj for non-scalar types\n compiler_types: Disable __builtin_counted_by_ref for Clang\n" }, { "commit": "c7decec2f2d2ab0366567f9e30c0e1418cece43f", "tree": "50312739ad43d0655ea71c942d848db2ff123e8e", "parents": [ "3544d5ce36f403db6e5c994f526101c870ffe9fe", "dbf0108347bdb5d4ccef8910555b16c1f1a505f8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:51:08 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:51:08 2026 -0800" }, "message": "Merge tag \u0027perf-tools-for-v7.0-1-2026-02-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools\n\nPull perf tools updates from Arnaldo Carvalho de Melo:\n\n - Introduce \u0027perf sched stats\u0027 tool with record/report/diff workflows\n using schedstat counters\n\n - Add a faster libdw based addr2line implementation and allow selecting\n it or its alternatives via \u0027perf config addr2line.style\u003d\u0027\n\n - Data-type profiling fixes and improvements including the ability to\n select fields using \u0027perf report\u0027\u0027s -F/-fields, e.g.:\n\n \u0027perf report --fields overhead,type\u0027\n\n - Add \u0027perf test\u0027 regression tests for Data-type profiling with C and\n Rust workloads\n\n - Fix srcline printing with inlines in callchains, make sure this has\n coverage in \u0027perf test\u0027\n\n - Fix printing of leaf IP in LBR callchains\n\n - Fix display of metrics without sufficient permission in \u0027perf stat\u0027\n\n - Print all machines in \u0027perf kvm report -vvv\u0027, not just the host\n\n - Switch from SHA-1 to BLAKE2s for build ID generation, remove SHA-1\n code\n\n - Fix \u0027perf report\u0027s histogram entry collapsing with \u0027-F\u0027 option\n\n - Use system\u0027s cacheline size instead of a hardcoded value in \u0027perf\n report\u0027\n\n - Allow filtering conversion by time range in \u0027perf data\u0027\n\n - Cover conversion to CTF using \u0027perf data\u0027 in \u0027perf test\u0027\n\n - Address newer glibc const-correctness (-Werror\u003ddiscarded-qualifiers)\n issues\n\n - Fixes and improvements for ARM\u0027s CoreSight support, simplify ARM SPE\n event config in \u0027perf mem\u0027, update docs for \u0027perf c2c\u0027 including the\n ARM events it can be used with\n\n - Build support for generating metrics from arch specific python\n script, add extra AMD, Intel, ARM64 metrics using it\n\n - Add AMD Zen 6 events and metrics\n\n - Add JSON file with OpenHW Risc-V CVA6 hardware counters\n\n - Add \u0027perf kvm\u0027 stats live testing\n\n - Add more \u0027perf stat\u0027 tests to \u0027perf test\u0027\n\n - Fix segfault in `perf lock contention -b/--use-bpf`\n\n - Fix various \u0027perf test\u0027 cases for s390\n\n - Build system cleanups, bump minimum shellcheck version to 0.7.2\n\n - Support building the capstone based annotation routines as a plugin\n\n - Allow passing extra Clang flags via EXTRA_BPF_FLAGS\n\n* tag \u0027perf-tools-for-v7.0-1-2026-02-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools: (255 commits)\n perf test script: Add python script testing support\n perf test script: Add perl script testing support\n perf script: Allow the generated script to be a path\n perf test: perf data --to-ctf testing\n perf test: Test pipe mode with data conversion --to-json\n perf json: Pipe mode --to-ctf support\n perf json: Pipe mode --to-json support\n perf check: Add libbabeltrace to the listed features\n perf build: Allow passing extra Clang flags via EXTRA_BPF_FLAGS\n perf test data_type_profiling.sh: Skip just the Rust tests if code_with_type workload is missing\n tools build: Fix feature test for rust compiler\n perf libunwind: Fix calls to thread__e_machine()\n perf stat: Add no-affinity flag\n perf evlist: Reduce affinity use and move into iterator, fix no affinity\n perf evlist: Missing TPEBS close in evlist__close()\n perf evlist: Special map propagation for tool events that read on 1 CPU\n perf stat-shadow: In prepare_metric fix guard on reading NULL perf_stat_evsel\n Revert \"perf tool_pmu: More accurately set the cpus for tool events\"\n tools build: Emit dependencies file for test-rust.bin\n tools build: Make test-rust.bin be removed by the \u0027clean\u0027 target\n ...\n" }, { "commit": "3544d5ce36f403db6e5c994f526101c870ffe9fe", "tree": "7259b303e612a67a0ce98108a713f49826723848", "parents": [ "981361604566a28517a518c317943d9b7c392217", "e3a22b532080bf32a4d4eb6172c8cb878123e13c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:25:42 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:25:42 2026 -0800" }, "message": "Merge tag \u0027cocci-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jlawall/linux\n\nPull coccinelle updates from Julia Lawall:\n \"This simplifies and clarifies the handling of output generated by\n Coccinelle that is sent to standard error.\n\n By default, this goes to /dev/null. Remind the user of that and\n encourage them to provide another file name (Benjamin Philip)\"\n\n* tag \u0027cocci-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jlawall/linux:\n Documentation: Coccinelle: document debug log handling\n scripts: coccicheck: warn on unset debug file\n scripts: coccicheck: simplify debug file handling\n" }, { "commit": "981361604566a28517a518c317943d9b7c392217", "tree": "7e3836f811aff8b3dba09914f2142f89412d628c", "parents": [ "f9d66e64a2bcb979d47eb7d67aa7e9b454fd5d15", "8c1f92ca8bca3ce2d2c085571af89503bc7bc7c4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:20:32 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:20:32 2026 -0800" }, "message": "Merge tag \u0027ntb-7.0\u0027 of https://github.com/jonmason/ntb\n\nPull NTB (PCIe non-transparent bridge) updates from Jon Mason:\n \"NTB updates include debugfs improvements, correctness fixes, cleanups,\n and new hardware support:\n\n ntb_transport QP stats are converted to seq_file, a tx_memcpy_offload\n module parameter is introduced with associated ordering fixes, and a\n debugfs queue name truncation bug is corrected.\n\n Additional fixes address format specifier mismatches in ntb_tool and\n boundary conditions in the Switchtec driver, while unused MSI helpers\n are removed and the codebase migrates to dma_map_phys().\n\n Intel Gen6 (Diamond Rapids) NTB support is also added\"\n\n* tag \u0027ntb-7.0\u0027 of https://github.com/jonmason/ntb:\n NTB: ntb_transport: Use seq_file for QP stats debugfs\n NTB: ntb_transport: Fix too small buffer for debugfs_name\n ntb/ntb_tool: correct sscanf format for u64 and size_t in tool_peer_mw_trans_write\n ntb: intel: Add Intel Gen6 NTB support for DiamondRapids\n NTB/msi: Remove unused functions\n ntb: ntb_hw_switchtec: Increase MAX_MWS limit to 256\n ntb: ntb_hw_switchtec: Fix array-index-out-of-bounds access\n ntb: ntb_hw_switchtec: Fix shift-out-of-bounds for 0 mw lut\n NTB: epf: allow built-in build\n ntb: migrate to dma_map_phys instead of map_page\n NTB: ntb_transport: Add \u0027tx_memcpy_offload\u0027 module option\n NTB: ntb_transport: Remove unused \u0027retries\u0027 field from ntb_queue_entry\n" }, { "commit": "f9d66e64a2bcb979d47eb7d67aa7e9b454fd5d15", "tree": "ed6bc2b311de8f14f8d07678f44e3d8133e21569", "parents": [ "817c16e5655958c34035a10c8c4b93e33e529624", "ea129e55c9e06a51a93c3f5ef3e32a6cfa3f8ec7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:05:49 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 10:05:49 2026 -0800" }, "message": "Merge tag \u0027io_uring-20260221\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux\n\nPull io_uring fixes from Jens Axboe:\n\n - A fix for a missing URING_CMD128 opcode check, fixing an issue with\n the SQE mixed mode support introduced in 6.19. Merged late due to\n having multiple dependencies\n\n - Add sqe-\u003ecmd size checking for big SQEs, similar to what we have for\n normal sized SQEs\n\n - Fix a race condition in zcrx, that leads to a double free\n\n* tag \u0027io_uring-20260221\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:\n io_uring: Add size check for sqe-\u003ecmd\n io_uring: add IORING_OP_URING_CMD128 to opcode checks\n io_uring/zcrx: fix user_ref race between scrub and refill paths\n" }, { "commit": "817c16e5655958c34035a10c8c4b93e33e529624", "tree": "eaa39143acd6347772a608c633b9807bcc33ac87", "parents": [ "4cf44657887b4c41374981d0afb2ca302b189e15", "f043a93fff9e3e3e648b6525483f59104b0819fa" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:58:22 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:58:22 2026 -0800" }, "message": "Merge tag \u0027fixes-2026-02-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock\n\nPull memblock fix from Mike Rapoport:\n \"Fix detection of NUMA node for CXL windows\n\n phys_to_target_node() may assign a CXL Fixed Memory Window to the\n wrong NUMA node when a CXL node resides in the gap of discontinuous\n System RAM node.\n\n Fix this by checking both numa_meminfo and numa_reserved_meminfo,\n preferring the reserved NID when the address appears in both\"\n\n* tag \u0027fixes-2026-02-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock:\n mm: numa_memblks: Identify the accurate NUMA ID of CFMW\n" }, { "commit": "4cf44657887b4c41374981d0afb2ca302b189e15", "tree": "f7bd3423c50ce9843335994b460969c32edac099", "parents": [ "8eb604d4ee8bf6183b00b8a96f0007b1be28ca9d", "640c9dc72f21f325700a4b0f839ad568ff21c697" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:38:59 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:38:59 2026 -0800" }, "message": "Merge tag \u0027sched_ext-for-7.0-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext\n\nPull sched_ext fixes from Tejun Heo:\n\n - Various bug fixes for the example schedulers and selftests\n\n* tag \u0027sched_ext-for-7.0-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:\n tools/sched_ext: fix getopt not re-parsed on restart\n tools/sched_ext: scx_userland: fix data races on shared counters\n tools/sched_ext: scx_pair: fix stride \u003d\u003d 0 crash on single-CPU systems\n tools/sched_ext: scx_central: fix CPU_SET and skeleton leak on early exit\n tools/sched_ext: scx_userland: fix stale data on restart\n tools/sched_ext: scx_flatcg: fix potential stack overflow from VLA in fcg_read_stats\n selftests/sched_ext: Fix rt_stall flaky failure\n tools/sched_ext: scx_userland: fix restart and stats thread lifecycle bugs\n tools/sched_ext: scx_central: fix sched_setaffinity() call with the set size\n tools/sched_ext: scx_flatcg: zero-initialize stats counter array\n" }, { "commit": "8eb604d4ee8bf6183b00b8a96f0007b1be28ca9d", "tree": "bf87c95b5cd230c442c62a9897f3075ffc2b8f5f", "parents": [ "d79526b89571ae447c1a5cfd3d627efa07098348", "a09dc10d1353f0e92c21eae2a79af1c2b1ddcde8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:11:32 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Feb 21 09:11:32 2026 -0800" }, "message": "Merge tag \u0027v7.0-rc-part2-ksmbd-server-fixes\u0027 of git://git.samba.org/ksmbd\n\nPull smb server fixes from Steve French:\n \"Two small fixes:\n\n - fix potential deadlock\n\n - minor cleanup\"\n\n* tag \u0027v7.0-rc-part2-ksmbd-server-fixes\u0027 of git://git.samba.org/ksmbd:\n ksmbd: call ksmbd_vfs_kern_path_end_removing() on some error paths\n smb: server: Remove duplicate include of misc.h\n" }, { "commit": "e3a22b532080bf32a4d4eb6172c8cb878123e13c", "tree": "4e2253aeb78dc46bc4adcf6b096c7c0b22fde57b", "parents": [ "bb1c9ccf740eae3f116a2bb5d26c2bfb80cc8c7e" ], "author": { "name": "Benjamin Philip", "email": "benjamin.philip495@gmail.com", "time": "Wed Jan 07 00:38:36 2026 +0530" }, "committer": { "name": "Julia Lawall", "email": "Julia.Lawall@inria.fr", "time": "Sat Feb 21 17:22:45 2026 +0100" }, "message": "Documentation: Coccinelle: document debug log handling\n\nThe current debug documentation does not mention that logs are printed\nto stdout unless DEBUG_FILE is set. It also doesn\u0027t mention that\nCoccinelle cannot overwrite debug files.\n\nDocument this behaviour in the examples and reference it in the\ndebugging section.\n\nSigned-off-by: Benjamin Philip \u003cbenjamin.philip495@gmail.com\u003e\nSigned-off-by: Julia Lawall \u003cjulia.lawall@inria.fr\u003e\n" }, { "commit": "bb1c9ccf740eae3f116a2bb5d26c2bfb80cc8c7e", "tree": "11fe19a116794621e68b937fa14cf0161c461dc2", "parents": [ "8952cfe431cf5b1d615f054f698cd74034bd1385" ], "author": { "name": "Benjamin Philip", "email": "benjamin.philip495@gmail.com", "time": "Wed Jan 07 00:38:35 2026 +0530" }, "committer": { "name": "Julia Lawall", "email": "Julia.Lawall@inria.fr", "time": "Sat Feb 21 17:22:30 2026 +0100" }, "message": "scripts: coccicheck: warn on unset debug file\n\ncoccicheck prints debug logs to stdout unless a debug file has been set.\nThis makes it hard to read coccinelle\u0027s suggested changes, especially\nfor someone new to coccicheck.\n\nFrom this commit, we warn about this behaviour from within the script on\nan unset debug file. Explicitly setting the debug file to /dev/null\nsuppresses the warning while keeping the default.\n\nSigned-off-by: Benjamin Philip \u003cbenjamin.philip495@gmail.com\u003e\nSigned-off-by: Julia Lawall \u003cjulia.lawall@inria.fr\u003e\n" }, { "commit": "8952cfe431cf5b1d615f054f698cd74034bd1385", "tree": "b9c2b884529cb62f07afbca7fa099bbaa2de2c50", "parents": [ "d79526b89571ae447c1a5cfd3d627efa07098348" ], "author": { "name": "Benjamin Philip", "email": "benjamin.philip495@gmail.com", "time": "Wed Jan 07 00:38:34 2026 +0530" }, "committer": { "name": "Julia Lawall", "email": "Julia.Lawall@inria.fr", "time": "Sat Feb 21 17:22:00 2026 +0100" }, "message": "scripts: coccicheck: simplify debug file handling\n\nThis commit separates handling unset files and pre-existing files. It\nalso eliminates a duplicated check for unset files in run_cmd_parmap().\n\nSigned-off-by: Benjamin Philip \u003cbenjamin.philip495@gmail.com\u003e\nSigned-off-by: Julia Lawall \u003cjulia.lawall@inria.fr\u003e\n" }, { "commit": "7a70c15bd1449f1eb30991772edce37b41e496fb", "tree": "24a343690f7afa3e084e26b2470a8382f743ee59", "parents": [ "69050f8d6d075dc01af7a5f2f550a8067510366f" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Sat Feb 21 00:12:19 2026 -0800" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Sat Feb 21 01:02:52 2026 -0800" }, "message": "kmalloc_obj: Clean up after treewide replacements\n\nCoccinelle doesn\u0027t handle re-indenting line escapes. Fix the 2 places\nwhere these got misaligned.\n\nRemove 2 now-redundant type casts, found with:\n$ git grep -P \u0027struct (\\S+).*\\)\\s*k\\S+alloc_(objs?|flex)\\(struct \\1\u0027\n\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "69050f8d6d075dc01af7a5f2f550a8067510366f", "tree": "bb265f94d9dfa7876c06a5d9f88673d496a15341", "parents": [ "d39a1d7486d98668dd34aaa6732aad7977c45f5a" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Feb 20 23:49:23 2026 -0800" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Sat Feb 21 01:02:28 2026 -0800" }, "message": "treewide: Replace kmalloc with kmalloc_obj for non-scalar types\n\nThis is the result of running the Coccinelle script from\nscripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to\navoid scalar types (which need careful case-by-case checking), and\ninstead replace kmalloc-family calls that allocate struct or union\nobject instances:\n\nSingle allocations:\tkmalloc(sizeof(TYPE), ...)\nare replaced with:\tkmalloc_obj(TYPE, ...)\n\nArray allocations:\tkmalloc_array(COUNT, sizeof(TYPE), ...)\nare replaced with:\tkmalloc_objs(TYPE, COUNT, ...)\n\nFlex array allocations:\tkmalloc(struct_size(PTR, FAM, COUNT), ...)\nare replaced with:\tkmalloc_flex(*PTR, FAM, COUNT, ...)\n\n(where TYPE may also be *VAR)\n\nThe resulting allocations no longer return \"void *\", instead returning\n\"TYPE *\".\n\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "d39a1d7486d98668dd34aaa6732aad7977c45f5a", "tree": "e3d0917dd485ceaca58f9c557550933a9acb2ab9", "parents": [ "d79526b89571ae447c1a5cfd3d627efa07098348" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Feb 20 13:15:58 2026 -0800" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Sat Feb 21 01:01:14 2026 -0800" }, "message": "compiler_types: Disable __builtin_counted_by_ref for Clang\n\nUnfortunately, there is a corner case of __builtin_counted_by_ref()\nusage that crashes[1] Clang since support was introduced in Clang 19.\nDisable it prior to Clang 22. Found while tested kmalloc_obj treewide\nrefactoring (via kmalloc_flex() usage).\n\nLink: https://github.com/llvm/llvm-project/issues/182575 [1]\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "640c9dc72f21f325700a4b0f839ad568ff21c697", "tree": "fff0d9135f60b2fc5cf70b8763002d01ba697d39", "parents": [ "f892f9f99464bead942a75d2b00dda6be07de97f" ], "author": { "name": "David Carlier", "email": "devnexen@gmail.com", "time": "Wed Feb 18 19:22:35 2026 +0000" }, "committer": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Fri Feb 20 17:17:38 2026 -1000" }, "message": "tools/sched_ext: fix getopt not re-parsed on restart\n\nAfter goto restart, optind retains its advanced position from the\nprevious getopt loop, causing getopt() to immediately return -1.\nThis silently drops all command-line options on the restarted skeleton.\n\nReset optind to 1 at the restart label so options are re-parsed.\n\nAffected schedulers: scx_simple, scx_central, scx_flatcg, scx_pair,\nscx_sdt, scx_cpu0.\n\nSigned-off-by: David Carlier \u003cdevnexen@gmail.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n" }, { "commit": "f892f9f99464bead942a75d2b00dda6be07de97f", "tree": "421ad80d41357a0fb00addd07ea1452cf43476cb", "parents": [ "625be3456b3ced6e2dca6166962c0cf6cc2e546d" ], "author": { "name": "David Carlier", "email": "devnexen@gmail.com", "time": "Wed Feb 18 19:22:23 2026 +0000" }, "committer": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Fri Feb 20 17:17:31 2026 -1000" }, "message": "tools/sched_ext: scx_userland: fix data races on shared counters\n\nThe stats thread reads nr_vruntime_enqueues, nr_vruntime_dispatches,\nnr_vruntime_failed, and nr_curr_enqueued concurrently with the main\nthread writing them, with no synchronization.\n\nUse __atomic builtins with relaxed ordering for all accesses to these\ncounters to eliminate the data races.\n\nOnly display accuracy is affected, not scheduling correctness.\n\nSigned-off-by: David Carlier \u003cdevnexen@gmail.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n" }, { "commit": "d79526b89571ae447c1a5cfd3d627efa07098348", "tree": "ccd22bc5f6cd63cf78289c72ffdd682319362b9f", "parents": [ "0de6219fd74440199fb0bfc6ce02bb8bdb8e9466", "40534d19ed2afb880ecf202dab26a8e7a5808d16" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:14:36 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:14:36 2026 -0800" }, "message": "Merge tag \u0027spi-fix-v7.0-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi\n\nPull spi fixes from Mark Brown:\n \"There\u0027s a relatively large but ultimately simple fix for spidev here\n which addresses some ABBA races by simplifying down to just using a\n single lock, it\u0027s not clear to me that there was ever any benefit in\n having the two separate locks in the first place.\n\n We also have simple missing error check fix in in the wpcm-fiu driver\"\n\n* tag \u0027spi-fix-v7.0-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:\n spi: spidev: fix lock inversion between spi_lock and buf_lock\n spi: wpcm-fiu: Fix potential NULL pointer dereference in wpcm_fiu_probe()\n" }, { "commit": "0de6219fd74440199fb0bfc6ce02bb8bdb8e9466", "tree": "fe7436bdd094024b8a21624fef5522e1f64e6064", "parents": [ "3f6eb5a6d27b56ea44c2de9f9028d803aeccbfe8", "09ad01a530bb6ad260bda4fa56bab84619d90968" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:11:55 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:11:55 2026 -0800" }, "message": "Merge tag \u0027regulator-fix-v7.0-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator\n\nPull regulator fixes from Mark Brown:\n \"A few driver specific fixes, plus a patch from Bjorn which removes a\n fixed limit on regulator names that was breaking some Qualcomm\n systems\"\n\n* tag \u0027regulator-fix-v7.0-merge-window\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:\n regulator: s2mps11: fix pctrlsel macro usage in s2mpg10_of_parse_cb()\n regulator: s2mps11: drop redundant sanity checks in s2mpg10_of_parse_cb()\n regulator: core: Remove regulator supply_name length limit\n regulator: mt6363: Fix interrmittent timeout\n" }, { "commit": "3f6eb5a6d27b56ea44c2de9f9028d803aeccbfe8", "tree": "f20895da656cafd5064eefd500484c2b3053e374", "parents": [ "433b23a3dad2b9423fc0d415c6d43add73ac7788", "8c746e22096579897d1f8f74dbb6b17a6862fb6d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:05:52 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 17:05:52 2026 -0800" }, "message": "Merge tag \u0027pci-v7.0-fixes-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci\n\nPull pci fixes from Bjorn Helgaas:\n\n - Fix bridge window selection bug that prevented resource assignment\n (Kai-Heng Feng)\n\n - Fix bridge window sizing, which failed to assign resources for\n windows containing only optional resources (ROMs, SR-IOV BARs, etc)\n (Ilpo Järvinen)\n\n - Select CONFIGFS_FS when PCI_EPF_TEST is enabled to avoid a link error\n (Arnd Bergmann)\n\n - Fix recently merged Endpoint inbound submapping feature (Koichiro\n Den)\n\n* tag \u0027pci-v7.0-fixes-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci:\n PCI: dwc: ep: Always clear IB maps on BAR update\n PCI: dwc: ep: Return after clearing BAR-match inbound mapping\n PCI: endpoint: pci-epf-test: Select configfs\n PCI: Account fully optional bridge windows correctly\n PCI: Validate window resource type in pbus_select_window_for_type()\n" }, { "commit": "433b23a3dad2b9423fc0d415c6d43add73ac7788", "tree": "8fd0054b77e06860feb934154d9bbe8440cbe103", "parents": [ "7e8d85235677b6571857c26854ad1d4edc64c50c", "1afafbaf749d8e8ec53f8e38efdc731131902b5b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 16:18:48 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 16:18:48 2026 -0800" }, "message": "Merge tag \u0027dmi-for-v7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging\n\nPull dmi update from Jean Delvare:\n\n - include product_family info in dmi-id modalias\n\n* tag \u0027dmi-for-v7.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:\n firmware/dmi: Include product_family info to modalias\n" }, { "commit": "7e8d85235677b6571857c26854ad1d4edc64c50c", "tree": "1b5467c61e72ef3f63430fe9533ff81d6545eeb8", "parents": [ "99e447220b938dfed6488db95a2930b57ea849ba", "fbd03587ba732c612b8a569d1cf5bed72bd3a27c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 16:10:54 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 16:10:54 2026 -0800" }, "message": "Merge tag \u0027gpio-fixes-for-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fixes from Bartosz Golaszewski:\n\n - add a missing IS_ERR() check in gpio-nomadik\n\n - fix a NULL-pointer dereference in GPIO character device code\n\n - restore label matching in swnode-lookup due to reported regressions\n in existing users (this will get removed again once we audit and\n update all drivers)\n\n - fix remove path in GPIO sysfs code\n\n - normalize the return value of gpio_chip::get() in gpio-amd-fch\n\n* tag \u0027gpio-fixes-for-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpio: amd-fch: ionly return allowed values from amd_fch_gpio_get()\n gpio: sysfs: fix chip removal with GPIOs exported over sysfs\n gpio: swnode: restore the swnode-name-against-chip-label matching\n gpio: cdev: Avoid NULL dereference in linehandle_create()\n gpio: nomadik: Add missing IS_ERR() check\n" }, { "commit": "99e447220b938dfed6488db95a2930b57ea849ba", "tree": "ea511783a2d5048525b88002f1fd4d29794e4740", "parents": [ "bc1d4e705f48f001f3a5480f04067c48bd00bcf0", "709cc48d3d01facaeb1eec3d93e1e1fb2fb21717" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:54:48 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:54:48 2026 -0800" }, "message": "Merge tag \u0027i2c-for-7.0-part2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull more i2c updates from Wolfram Sang:\n \"Designware:\n - refactor the transfer path to support I2C_M_STOP\n - handle pm runtime by using the active auto try macros\n - handle controllers lacking explicit START and STOP conditions\n - general cleanups\n\n Other i2c drivers:\n - qualcomm: add support for qcs8300-cci\n - amd8111: general cleanups\n - cp2112: add DT bindings\"\n\n* tag \u0027i2c-for-7.0-part2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n dt-bindings: i2c: Add CP2112 HID USB to SMBus Bridge\n i2c: amd8111: switch to devm_ functions\n i2c: amd8111: Remove spaces in MODULE_* macros\n i2c: designware-platdrv: fix cleanup on probe failure\n i2c: designware-platdrv: simplify reset control\n dt-bindings: i2c: qcom-cci: Document qcs8300 compatible\n i2c: designware: Remove dead code in AMD ISP case\n i2c: designware: Support of controller with IC_EMPTYFIFO_HOLD_MASTER disabled\n i2c: designware: Use runtime PM macro for auto-cleanup\n i2c: designware: Implement I2C_M_STOP support\n" }, { "commit": "bc1d4e705f48f001f3a5480f04067c48bd00bcf0", "tree": "d4da92d831ded487a1a71fce1fb9930f3f24e6a6", "parents": [ "d4a292c5f8e65d2784b703c67179f4f7d0c7846c", "89339d889d63a4f981f185d048796762e86aa6c0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:47:44 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:47:44 2026 -0800" }, "message": "Merge tag \u0027sound-fix-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"Here are a bunch of updates, but there should be no big surprises;\n mostly device-specific quirks and fix-ups or non-code changes:\n\n - Quirks for ASoC AMD, HD-audio and USB-audio\n\n - Fixes in ASoC fsl, rockchip, renesas, aw codecs\n\n - Fixes for USB-audio packet handling in the implicit feedback mode\n\n - Updates of SPDX license IDs in some files\"\n\n* tag \u0027sound-fix-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (28 commits)\n ASoC: rockchip: i2s-tdm: Use param rate if not provided by set_sysclk\n ALSA: hda/hdmi: Add quirk for TUXEDO IBS14G6\n ASoC: dt-bindings: asahi-kasei,ak5558: Fix the supply names\n ASoC: dt-bindings: asahi-kasei,ak4458: Fix the supply names\n ASoC: dt-bindings: asahi-kasei,ak4458: set unevaluatedProperties:false\n ASoC: amd: amd_sdw: add machine driver quirk for Lenovo models\n ASoC: amd: acp: Add ACP7.0 match entries for Realtek parts\n ALSA: echoaudio: Add SPDX ids to some files\n ALSA: isa: Add SPDX id lines to some files\n ALSA: core: Add SPDX license id to files\n ASoC: tas2783A: add explicit port prepare handling\n ASoC: renesas: rz-ssi: Fix playback and capture\n ALSA: hda/realtek: Fix headset mic on ASUS Zenbook 14 UX3405MA\n ALSA: hda/conexant: Fix headphone jack handling on Acer Swift SF314\n ASoC: qcom: sm8250: Add quinary MI2S support\n ASoC: amd: yc: Add DMI quirk for ASUS Vivobook Pro 15X M6501RR\n ALSA: usb-audio: Avoid potentially repeated XRUN error messages\n ALSA: usb-audio: Add sanity check for OOB writes at silencing\n ALSA: usb-audio: Optimize the copy of packet sizes for implicit fb handling\n ALSA: usb-audio: Update the number of packets properly at receiving\n ...\n" }, { "commit": "d4a292c5f8e65d2784b703c67179f4f7d0c7846c", "tree": "0e3da362e92fea4acb77ceb7af1abe5713d90eac", "parents": [ "eee3666c92406fd8e5e3084b0b3129528dfe9557", "ae9e8654579709c2f10b8c86a8467e1710d4599f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:36:38 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:36:38 2026 -0800" }, "message": "Merge tag \u0027drm-next-2026-02-21\u0027 of https://gitlab.freedesktop.org/drm/kernel\n\nPull drm fixes from Dave Airlie:\n \"This is the fixes and cleanups for the end of the merge window, it\u0027s\n nearly all amdgpu, with some amdkfd, then a pagemap core fix, i915/xe\n display fixes, and some xe driver fixes.\n\n Nothing seems out of the ordinary, except amdgpu is a little more\n volume than usual.\n\n pagemap:\n - drm/pagemap: pass pagemap_addr by reference\n\n amdgpu:\n - DML 2.1 fixes\n - Panel replay fixes\n - Display writeback fixes\n - MES 11 old firmware compat fix\n - DC CRC improvements\n - DPIA fixes\n - XGMI fixes\n - ASPM fix\n - SMU feature bit handling fixes\n - DC LUT fixes\n - RAS fixes\n - Misc memory leak in error path fixes\n - SDMA queue reset fixes\n - PG handling fixes\n - 5 level GPUVM page table fix\n - SR-IOV fix\n - Queue reset fix\n - SMU 13.x fixes\n - DC resume lag fix\n - MPO fixes\n - DCN 3.6 fix\n - VSDB fixes\n - HWSS clean up\n - Replay fixes\n - DCE cursor fixes\n - DCN 3.5 SR DDR5 latency fixes\n - HPD fixes\n - Error path unwind fixes\n - SMU13/14 mode1 reset fixes\n - PSP 15 updates\n - SMU 15 updates\n - Sync fix in amdgpu_dma_buf_move_notify()\n - HAINAN fix\n - PSP 13.x fix\n - GPUVM locking fix\n - Fixes for DC analog support\n - DC FAMS fixes\n - DML 2.1 fixes\n - eDP fixes\n - Misc DC fixes\n - Fastboot fix\n - 3DLUT fixes\n - GPUVM fixes\n - 64bpp format fix\n - Fix for MacBooks with switchable gfx\n\n amdkfd:\n - Fix possible double deletion of validate list\n - Event setup fix\n - Device disconnect regression fix\n - APU GTT as VRAM fix\n - Fix piority inversion with MQDs\n - NULL check fix\n\n radeon:\n - HAINAN fix\n\n i915/xe display:\n - Regresion fix for HDR 4k displays (#15503)\n - Fixup for Dell XPS 13 7390 eDP rate limit\n - Memory leak fix on ACPI _DSM handling\n - Add missing slice count check during DP mode validation\n\n xe:\n - drm/xe: Prevent VFs from exposing the CCS mode sysfs file\n - SRIOV related fixes\n - PAT cache fix\n - MMIO read fix\n - W/a fixes\n - Adjust type of xe_modparam.force_vram_bar_size\n - Wedge mode fix\n - HWMon fix\n\n* tag \u0027drm-next-2026-02-21\u0027 of https://gitlab.freedesktop.org/drm/kernel: (143 commits)\n drm/amd/display: Remove unneeded DAC link encoder register\n drm/amd/display: Enable DAC in DCE link encoder\n drm/amd/display: Set CRTC source for DAC using registers\n drm/amd/display: Initialize DAC in DCE link encoder using VBIOS\n drm/amd/display: Turn off DAC in DCE link encoder using VBIOS\n drm/amd/display: Don\u0027t call find_analog_engine() twice\n drm/amdgpu: fix 4-level paging if GMC supports 57-bit VA v2\n drm/amdgpu: keep vga memory on MacBooks with switchable graphics\n drm/amdgpu: Set atomics to true for xgmi\n drm/amdkfd: Check for NULL return values\n drm/amd/display: Use same max plane scaling limits for all 64 bpp formats\n drm/amdgpu: Set vmid0 PAGE_TABLE_DEPTH for GFX12.1\n drm/amdkfd: Disable MQD queue priority\n drm/amd/display: Remove conditional for shaper 3DLUT power-on\n drm/amd/display: Check return of shaper curve to HW format\n drm/amd/display: Correct logic check error for fastboot\n drm/amd/display: Skip eDP detection when no sink\n Revert \"drm/amd/display: Add Gfx Base Case For Linear Tiling Handling\"\n Revert \"drm/amd/display: Correct hubp GfxVersion verification\"\n Revert \"drm/amd/display: Add Handling for gfxversion DcGfxBase\"\n ...\n" }, { "commit": "eee3666c92406fd8e5e3084b0b3129528dfe9557", "tree": "fca3226a3b69cfbf2b99222528a2a06e7498436b", "parents": [ "68010e7b3daf0c2cf91eccb329703e82d1ef5aff", "0f967cb8ea04486953f85b249b42dda3de842053" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:24:21 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:24:21 2026 -0800" }, "message": "Merge tag \u0027fbdev-for-7.0-rc1-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/linux-fbdev\n\nPull more fbdev updates from Helge Deller:\n \"Code cleanups for the au1100fb fbdev driver (Uwe Kleine-König)\"\n\n* tag \u0027fbdev-for-7.0-rc1-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/linux-fbdev:\n fbdev: au1100fb: Replace license boilerplate by SPDX header\n fbdev: au1100fb: Fold au1100fb.h into its only user\n fbdev: au1100fb: Replace custom printk wrappers by pr_*\n fbdev: au1100fb: Make driver compilable on non-mips platforms\n fbdev: au1100fb: Use proper conversion specifiers in printk formats\n fbdev: au1100fb: Mark several local functions as static\n fbdev: au1100fb: Don\u0027t store device specific data in global variables\n" }, { "commit": "68010e7b3daf0c2cf91eccb329703e82d1ef5aff", "tree": "6720eb4715428aaaa44ad368da9ebee7dac4852d", "parents": [ "b3f1da2a4d851b8e1ccf932e52c6772fe2253a47", "9678e53179aa7e907360f5b5b275769008a69b80" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:05:26 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 15:05:26 2026 -0800" }, "message": "Merge tag \u0027trace-v7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull tracing fixes from Steven Rostedt:\n\n - Fix possible dereference of uninitialized pointer\n\n When validating the persistent ring buffer on boot up, if the first\n validation fails, a reference to \"head_page\" is performed in the\n error path, but it skips over the initialization of that variable.\n Move the initialization before the first validation check.\n\n - Fix use of event length in validation of persistent ring buffer\n\n On boot up, the persistent ring buffer is checked to see if it is\n valid by several methods. One being to walk all the events in the\n memory location to make sure they are all valid. The length of the\n event is used to move to the next event. This length is determined by\n the data in the buffer. If that length is corrupted, it could\n possibly make the next event to check located at a bad memory\n location.\n\n Validate the length field of the event when doing the event walk.\n\n - Fix function graph on archs that do not support use of ftrace_ops\n\n When an architecture defines HAVE_DYNAMIC_FTRACE_WITH_ARGS, it means\n that its function graph tracer uses the ftrace_ops of the function\n tracer to call its callbacks. This allows a single registered\n callback to be called directly instead of checking the callback\u0027s\n meta data\u0027s hash entries against the function being traced.\n\n For architectures that do not support this feature, it must always\n call the loop function that tests each registered callback (even if\n there\u0027s only one). The loop function tests each callback\u0027s meta data\n against its hash of functions and will call its callback if the\n function being traced is in its hash map.\n\n The issue was that there was no check against this and the direct\n function was being called even if the architecture didn\u0027t support it.\n This meant that if function tracing was enabled at the same time as a\n callback was registered with the function graph tracer, its callback\n would be called for every function that the function tracer also\n traced, even if the callback\u0027s meta data only wanted to be called\n back for a small subset of functions.\n\n Prevent the direct calling for those architectures that do not\n support it.\n\n - Fix references to trace_event_file for hist files\n\n The hist files used event_file_data() to get a reference to the\n associated trace_event_file the histogram was attached to. This would\n return a pointer even if the trace_event_file is about to be freed\n (via RCU). Instead it should use the event_file_file() helper that\n returns NULL if the trace_event_file is marked to be freed so that no\n new references are added to it.\n\n - Wake up hist poll readers when an event is being freed\n\n When polling on a hist file, the task is only awoken when a hist\n trigger is triggered. This means that if an event is being freed\n while there\u0027s a task waiting on its hist file, it will need to wait\n until the hist trigger occurs to wake it up and allow the freeing to\n happen. Note, the event will not be completely freed until all\n references are removed, and a hist poller keeps a reference. But it\n should still be woken when the event is being freed.\n\n* tag \u0027trace-v7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n tracing: Wake up poll waiters for hist files when removing an event\n tracing: Fix checking of freed trace_event_file for hist files\n fgraph: Do not call handlers direct when not using ftrace_ops\n tracing: ring-buffer: Fix to check event length before using\n ring-buffer: Fix possible dereference of uninitialized pointer\n" }, { "commit": "b3f1da2a4d851b8e1ccf932e52c6772fe2253a47", "tree": "6e307f51e8fc402284da126257489a5293ca2706", "parents": [ "233a0c0f4498309f98d98a4b6543fa05a12d9587", "ecb7c2484cfc83a93658907580035a8adf1e0a92" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:57:09 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:57:09 2026 -0800" }, "message": "Merge tag \u0027for-7.0-rc1-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux\n\nPull btrfs fixes from David Sterba:\n\n - multiple error handling fixes of unexpected conditions\n\n - reset block group size class once it becomes empty so that\n its class can be changed\n\n - error message level adjustments\n\n - fixes of returned error values\n\n - use correct block reserve for delayed refs\n\n* tag \u0027for-7.0-rc1-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:\n btrfs: fix invalid leaf access in btrfs_quota_enable() if ref key not found\n btrfs: fix lost error return in btrfs_find_orphan_roots()\n btrfs: fix lost return value on error in finish_verity()\n btrfs: change unaligned root messages to error level in btrfs_validate_super()\n btrfs: use the correct type to initialize block reserve for delayed refs\n btrfs: do not ASSERT() when the fs flips RO inside btrfs_repair_io_failure()\n btrfs: reset block group size class when it becomes empty\n btrfs: replace BUG() with error handling in __btrfs_balance()\n btrfs: handle unexpected exact match in btrfs_set_inode_index_count()\n" }, { "commit": "233a0c0f4498309f98d98a4b6543fa05a12d9587", "tree": "9dd60d8bc27f9e77d5718cd8f3432fcd54489d3a", "parents": [ "219d7660c57ec7457a6020f3d025cfe1139ffc64", "99853d9daef240a45e161d0e33487cda4810b999" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:46:31 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:46:31 2026 -0800" }, "message": "Merge tag \u0027ecryptfs-7.0-rc1-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs\n\nPull ecryptfs updates from Tyler Hicks:\n \"This consists of some really minor typo fixes that fell through the\n cracks and some more recent code cleanups:\n\n - Comment typo fixes\n\n - Removal of an unused function declaration\n\n - Use strscpy() instead of the deprecated strcpy()\n\n - Use string copying helpers instead of memcpy() and manually\n terminating strings\"\n\n* tag \u0027ecryptfs-7.0-rc1-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs:\n ecryptfs: Replace memcpy + NUL termination in ecryptfs_copy_filename\n ecryptfs: Drop redundant NUL terminations after calling ecryptfs_to_hex\n ecryptfs: Replace memcpy + NUL termination in ecryptfs_new_file_context\n ecryptfs: Replace strcpy with strscpy in ecryptfs_validate_options\n ecryptfs: Replace strcpy with strscpy in ecryptfs_cipher_code_to_string\n ecryptfs: Replace strcpy with strscpy in ecryptfs_set_default_crypt_stat_vals\n ecryptfs: simplify list initialization in ecryptfs_parse_packet_set()\n ecryptfs: Remove unused declartion ecryptfs_fill_zeros()\n ecryptfs: Fix packet format comment in parse_tag_67_packet()\n ecryptfs: comment typo fix\n ecryptfs: keystore: Fix typo \u0027the the\u0027 in comment\n" }, { "commit": "8c1f92ca8bca3ce2d2c085571af89503bc7bc7c4", "tree": "a1c884e6cee606b88d656dfd7c1ddc8cc9e6502d", "parents": [ "6a4b50585d74fe45d3ade1e3e86ba8aae79761a5" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Wed Jan 07 13:24:58 2026 +0900" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB: ntb_transport: Use seq_file for QP stats debugfs\n\nThe ./qp*/stats debugfs file for each NTB transport QP is currently\nimplemented with a hand-crafted kmalloc() buffer and a series of\nscnprintf() calls. This is a pre-seq_file style pattern and makes future\nextensions easy to truncate.\n\nConvert the stats file to use the seq_file helpers via\nDEFINE_SHOW_ATTRIBUTE(), which simplifies the code and lets the seq_file\ncore handle buffering and partial reads.\n\nReviewed-by: Frank Li \u003cFrank.Li@nxp.com\u003e\nReviewed-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "6a4b50585d74fe45d3ade1e3e86ba8aae79761a5", "tree": "f9f067eb4214554163d92230bebc465fedda5c9b", "parents": [ "21fbdc4d0b1e7f9cabacc3587d07c62e01c7b5e8" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Wed Jan 07 13:24:57 2026 +0900" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB: ntb_transport: Fix too small buffer for debugfs_name\n\nThe buffer used for \"qp%d\" was only 4 bytes, which truncates names like\n\"qp10\" to \"qp1\" and causes multiple queues to share the same directory.\n\nEnlarge the buffer and use sizeof() to avoid truncation.\n\nFixes: fce8a7bb5b4b (\"PCI-Express Non-Transparent Bridge Support\")\nCc: \u003cstable@vger.kernel.org\u003e # v3.9+\nReviewed-by: Frank Li \u003cFrank.Li@nxp.com\u003e\nReviewed-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "21fbdc4d0b1e7f9cabacc3587d07c62e01c7b5e8", "tree": "7a36b66242c91f820ce295f5f1d5b544a788e5b6", "parents": [ "4921811678e93a83cbfebc14814a165ae794bf1d" ], "author": { "name": "yangqixiao", "email": "yangqixiao@inspur.com", "time": "Tue Dec 30 20:46:56 2025 +0800" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb/ntb_tool: correct sscanf format for u64 and size_t in tool_peer_mw_trans_write\n\nThe sscanf() call in tool_peer_mw_trans_write() uses \"%lli:%zi\" to parse\nuser input into \u0027u64 addr\u0027 and \u0027size_t wsize\u0027. This is incorrect:\n\n - \"%lli\" expects a signed long long *, but \u0027addr\u0027 is u64 (unsigned).\n Input like \"0x8000000000000000\" is misinterpreted as negative,\n leading to corrupted address values.\n\n - \"%zi\" expects a signed ssize_t *, but \u0027wsize\u0027 is size_t (unsigned).\n Input of \"-1\" is successfully parsed and stored as SIZE_MAX\n (e.g., 0xFFFFFFFFFFFFFFFF), which may cause buffer overflows\n or infinite loops in subsequent memory operations.\n\nFix by using format specifiers that match the actual variable types:\n - \"%llu\" for u64 (supports hex/decimal, standard for kernel u64 parsing)\n - \"%zu\" for size_t (standard and safe; rejects negative input)\n\nSigned-off-by: yangqixiao \u003cyangqixiao@inspur.com\u003e\nReviewed-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "4921811678e93a83cbfebc14814a165ae794bf1d", "tree": "a62d99ba9d988d8d03d21ecc86d696a04006f3ca", "parents": [ "7bd27439a578bb7724a9f09f240337ab95d68d2b" ], "author": { "name": "Dave Jiang", "email": "dave.jiang@intel.com", "time": "Thu Jan 08 14:21:01 2026 -0700" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb: intel: Add Intel Gen6 NTB support for DiamondRapids\n\nAdd DiamondRapids NTB support by adding the DID and adjust the changed\nPPD0 offset.\n\nSigned-off-by: Dave Jiang \u003cdave.jiang@intel.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "7bd27439a578bb7724a9f09f240337ab95d68d2b", "tree": "fe3eb00e134df1faf386f4585f8cdd7693c5bd6e", "parents": [ "b1f4077465b2110d8a486cd6f1aed0c2569c339a" ], "author": { "name": "Dr. David Alan Gilbert", "email": "linux@treblig.org", "time": "Thu Feb 20 02:03:57 2025 +0000" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB/msi: Remove unused functions\n\nntbm_msi_free_irq() and ntb_msi_peer_addr() were both added in 2019\u0027s\ncommit 26b3a37b9284 (\"NTB: Introduce MSI library\")\nbut have remained unused.\n\nRemove them, and the ntbm_msi_callback_match() helper that\nwas used by ntbm_msi_free_irq().\n\nSigned-off-by: Dr. David Alan Gilbert \u003clinux@treblig.org\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "b1f4077465b2110d8a486cd6f1aed0c2569c339a", "tree": "df1e28c553934f6d83ef95e1c56de2a1aa6bdc8d", "parents": [ "c8ba7ad2cc1c7b90570aa347b8ebbe279f1eface" ], "author": { "name": "Maciej Grochowski", "email": "Maciej.Grochowski@sony.com", "time": "Thu Feb 13 14:53:19 2025 -0800" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb: ntb_hw_switchtec: Increase MAX_MWS limit to 256\n\nMicrochip NTB switchtec devices supports up to 512 LUT\u0027s across all\nNT partitions. This patch enable symmetric NTB configuration to utilize\nall 512 memory windows across 2 peers partitions.\n\nSigned-off-by: Maciej Grochowski \u003cMaciej.Grochowski@sony.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "c8ba7ad2cc1c7b90570aa347b8ebbe279f1eface", "tree": "2baaf5b41173a8eaf1f1db05e8bb7aaefbfc1d50", "parents": [ "186615f8855a0be4ee7d3fcd09a8ecc10e783b08" ], "author": { "name": "Maciej Grochowski", "email": "Maciej.Grochowski@sony.com", "time": "Thu Feb 13 14:53:18 2025 -0800" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb: ntb_hw_switchtec: Fix array-index-out-of-bounds access\n\nNumber of MW LUTs depends on NTB configuration and can be set to MAX_MWS,\nThis patch protects against invalid index out of bounds access to mw_sizes\nWhen invalid access print message to user that configuration is not valid.\n\nSigned-off-by: Maciej Grochowski \u003cMaciej.Grochowski@sony.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "186615f8855a0be4ee7d3fcd09a8ecc10e783b08", "tree": "5b582376ee2a39f3c7dbc5303435d4a0a0de8f25", "parents": [ "f39e7d0e7f0dc3f8886d957d268eb5c5b28ddf97" ], "author": { "name": "Maciej Grochowski", "email": "Maciej.Grochowski@sony.com", "time": "Thu Feb 13 14:53:17 2025 -0800" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb: ntb_hw_switchtec: Fix shift-out-of-bounds for 0 mw lut\n\nNumber of MW LUTs depends on NTB configuration and can be set to zero,\nin such scenario rounddown_pow_of_two will cause undefined behaviour and\nshould not be performed.\nThis patch ensures that rounddown_pow_of_two is called on valid value.\n\nSigned-off-by: Maciej Grochowski \u003cMaciej.Grochowski@sony.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "f39e7d0e7f0dc3f8886d957d268eb5c5b28ddf97", "tree": "ac9c178264d1042c5f4c8667023227808f860206", "parents": [ "c722e055d8630683177c46738f41fe5fbb0e5045" ], "author": { "name": "Baruch Siach", "email": "baruch@tkos.co.il", "time": "Mon Nov 03 09:44:55 2025 +0200" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB: epf: allow built-in build\n\nntb_hw_epf works just as well when built into the kernel image. Don\u0027t\nforce module build.\n\nSigned-off-by: Baruch Siach \u003cbaruch@tkos.co.il\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "c722e055d8630683177c46738f41fe5fbb0e5045", "tree": "84ebadcdf6adbec61d7f0a3b6639aa1b46f109a5", "parents": [ "322617a06c97153f7b0681ecaa55490abccff7fa" ], "author": { "name": "Chu Guangqing", "email": "chuguangqing@inspur.com", "time": "Fri Oct 24 14:20:42 2025 +0800" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "ntb: migrate to dma_map_phys instead of map_page\n\nAfter introduction of dma_map_phys(), there is no need to convert\nfrom physical address to struct page in order to map page. So let\u0027s\nuse it directly.\n\nSigned-off-by: Chu Guangqing \u003cchuguangqing@inspur.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "322617a06c97153f7b0681ecaa55490abccff7fa", "tree": "47e4a6d2907f0a721721815bc75320c5185a77b5", "parents": [ "b36490b5fb9866295cc13808b04a968b13acbab3" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Thu Oct 23 16:21:05 2025 +0900" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB: ntb_transport: Add \u0027tx_memcpy_offload\u0027 module option\n\nSome platforms (e.g. R-Car S4) do not gain from using a DMAC on TX path\nin ntb_transport and end up CPU-bound on memcpy_toio(). Add a module\nparameter \u0027tx_memcpy_offload\u0027 that moves the TX memcpy_toio() and\ndescriptor writes to a per-QP kernel thread. It is disabled by default.\n\nThis change also fixes a rare ordering hazard in ntb_tx_copy_callback(),\nthat was observed on R-Car S4 once throughput improved with the new\nmodule parameter: the DONE flag write to the peer MW, which is WC\nmapped, could be observed after the DB/MSI trigger. Both operations are\nposted PCIe MWr (often via different OB iATUs), so WC buffering and\nbridges may reorder visibility. Insert dma_mb() to enforce store-\u003eload\nordering and then read back hdr-\u003eflags to flush the posted write before\nringing the doorbell / issuing MSI.\n\nWhile at it, update tx_index with WRITE_ONCE() at the earlier possible\nlocation to make ntb_transport_tx_free_entry() robust.\n\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "b36490b5fb9866295cc13808b04a968b13acbab3", "tree": "776c2d1a3f7c5876c8176f7d59f7ff8c8abe8c42", "parents": [ "05f7e89ab9731565d8a62e3b5d1ec206485eeb0b" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Thu Oct 23 16:21:04 2025 +0900" }, "committer": { "name": "Jon Mason", "email": "jdmason@kudzu.us", "time": "Fri Feb 20 17:31:55 2026 -0500" }, "message": "NTB: ntb_transport: Remove unused \u0027retries\u0027 field from ntb_queue_entry\n\nDrop the unused field \u0027retries\u0027 from struct ntb_queue_entry for\nsimplicity\u0027s sake.\n\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\n" }, { "commit": "969c3cca0f3b88682cd833cee4cf01b0915629a3", "tree": "51060ba998dcd7534c21b4437ad3875dc9f60bb4", "parents": [ "76e8919aef2dca25a07a0d6480e486ebe5e9ea79" ], "author": { "name": "Andreas Gabriel-Platschek", "email": "andi.platschek@gmail.com", "time": "Mon Feb 09 06:34:39 2026 +0100" }, "committer": { "name": "Alexandre Belloni", "email": "alexandre.belloni@bootlin.com", "time": "Fri Feb 20 23:12:25 2026 +0100" }, "message": "rtc: ds1390: fix number of bytes read from RTC\n\nThe spi_write_then_read() reads 8 bytes starting from\nDS1390_REG_SECONDS (\u003d\u003d 0x01), so the last byte read would already\nbe part of the alarm (Tenths and Hundredths of Seconds) feature.\n\nHowever 7 bytes are engouh -- seconds (0x01), minutes (0x02), hours (0x03),\nday (0x04), date (0x05), month/century (0x06) and year (0x07).\n\nSigned-off-by: Andreas Gabriel-Platschek \u003candi.platschek@gmail.com\u003e\nLink: https://patch.msgid.link/20260209053439.313825-1-andi.platschek@gmail.com\nSigned-off-by: Alexandre Belloni \u003calexandre.belloni@bootlin.com\u003e\n" }, { "commit": "76e8919aef2dca25a07a0d6480e486ebe5e9ea79", "tree": "4b028a34732709b16897e3a5b668a1b896cec57b", "parents": [ "d1f64cdfafdbd3fe220a59b85f13e125f49ce73e" ], "author": { "name": "Jinjie Ruan", "email": "ruanjinjie@huawei.com", "time": "Thu Jan 22 17:00:31 2026 +0800" }, "committer": { "name": "Alexandre Belloni", "email": "alexandre.belloni@bootlin.com", "time": "Fri Feb 20 23:12:17 2026 +0100" }, "message": "rtc: class: Remove duplicate check for alarm\n\nIn __devm_rtc_register_device(), the callee rtc_initialize_alarm()\nwill check the alarm, there is no need to check in advance,\nso remove it.\n\nSigned-off-by: Jinjie Ruan \u003cruanjinjie@huawei.com\u003e\nLink: https://patch.msgid.link/20260122090031.3871746-1-ruanjinjie@huawei.com\nSigned-off-by: Alexandre Belloni \u003calexandre.belloni@bootlin.com\u003e\n" }, { "commit": "d1f64cdfafdbd3fe220a59b85f13e125f49ce73e", "tree": "579e7d1a0ae13626938d48d0dc314a05d2096394", "parents": [ "81be22cd4ace020045cc6d31255c6f7c071eb7c0" ], "author": { "name": "Rouven Czerwinski", "email": "rouven.czerwinski@linaro.org", "time": "Mon Jan 26 11:11:26 2026 +0100" }, "committer": { "name": "Alexandre Belloni", "email": "alexandre.belloni@bootlin.com", "time": "Fri Feb 20 23:12:10 2026 +0100" }, "message": "rtc: optee: simplify OP-TEE context match\n\nSimplify the TEE implementor ID match by returning the boolean\nexpression directly instead of going through an if/else.\n\nSigned-off-by: Rouven Czerwinski \u003crouven.czerwinski@linaro.org\u003e\nLink: https://patch.msgid.link/20260126-optee-simplify-context-match-v1-3-d4104e526cb6@linaro.org\nSigned-off-by: Alexandre Belloni \u003calexandre.belloni@bootlin.com\u003e\n" }, { "commit": "219d7660c57ec7457a6020f3d025cfe1139ffc64", "tree": "cadc37189c7e1b9d529a7b83805dabcc65ef1f07", "parents": [ "43257b2ebd1f08536e3a8c06e2946af902015162", "08020dbe3125e936429e7966bf072e08fa964f36" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:06:06 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 14:06:06 2026 -0800" }, "message": "Merge tag \u0027apparmor-pr-2026-02-18\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor\n\nPull AppArmor updates from John Johansen:\n \"Features:\n - add .kunitconfig\n - audit execpath in userns mediation\n - add support loading per permission tagging\n\n Cleanups:\n - remove unused percpu critical sections in buffer management\n - document the buffer hold, add an overflow guard\n - split xxx_in_ns into its two separate semantic use cases\n - remove apply_modes_to_perms from label_match\n - refactor/cleanup cred helper fns.\n - guard against free attachment/data routines being called with NULL\n - drop in_atomic flag in common_mmap, common_file_perm, and cleanup\n - make str table more generic and be able to have multiple entries\n - Replace deprecated strcpy with memcpy in gen_symlink_name\n - Replace deprecated strcpy in d_namespace_path\n - Replace sprintf/strcpy with scnprintf/strscpy in aa_policy_init\n - replace sprintf with snprintf in aa_new_learning_profile\n\n Bug Fixes:\n - fix cast in format string DEBUG statement\n - fix make aa_labelmatch return consistent\n - fix fmt string type error in process_strs_entry\n - fix kernel-doc comments for inview\n - fix invalid deref of rawdata when export_binary is unset\n - avoid per-cpu hold underflow in aa_get_buffer\n - fix fast path cache check for unix sockets\n - fix rlimit for posix cpu timers\n - fix label and profile debug macros\n - move check for aa_null file to cover all cases\n - return -ENOMEM in unpack_perms_table upon alloc failure\n - fix boolean argument in apparmor_mmap_file\n - Fix \u0026 Optimize table creation from possibly unaligned memory\n - Allow apparmor to handle unaligned dfa tables\n - fix NULL deref in aa_sock_file_perm\n - fix NULL pointer dereference in __unix_needs_revalidation\n - fix signedness bug in unpack_tags()\"\n\n* tag \u0027apparmor-pr-2026-02-18\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jj/linux-apparmor: (34 commits)\n apparmor: fix signedness bug in unpack_tags()\n apparmor: fix cast in format string DEBUG statement\n apparmor: fix aa_label to return state from compount and component match\n apparmor: fix fmt string type error in process_strs_entry\n apparmor: fix kernel-doc comments for inview\n apparmor: fix invalid deref of rawdata when export_binary is unset\n apparmor: add .kunitconfig\n apparmor: cleanup remove unused percpu critical sections in buffer management\n apparmor: document the buffer hold, add an overflow guard\n apparmor: avoid per-cpu hold underflow in aa_get_buffer\n apparmor: split xxx_in_ns into its two separate semantic use cases\n apparmor: make label_match return a consistent value\n apparmor: remove apply_modes_to_perms from label_match\n apparmor: fix fast path cache check for unix sockets\n apparmor: fix rlimit for posix cpu timers\n apparmor: refactor/cleanup cred helper fns.\n apparmor: fix label and profile debug macros\n apparmor: move check for aa_null file to cover all cases\n apparmor: guard against free routines being called with a NULL\n apparmor: return -ENOMEM in unpack_perms_table upon alloc failure\n ...\n" }, { "commit": "81be22cd4ace020045cc6d31255c6f7c071eb7c0", "tree": "1d08f8a43b094c5ede7d2faa6d27cd53eb740126", "parents": [ "8eeb611b3d16241559bbf81ae5d938368c87818c" ], "author": { "name": "Anthony Pighin (Nokia)", "email": "anthony.pighin@nokia.com", "time": "Tue Nov 25 17:35:19 2025 +0000" }, "committer": { "name": "Alexandre Belloni", "email": "alexandre.belloni@bootlin.com", "time": "Fri Feb 20 22:49:17 2026 +0100" }, "message": "rtc: interface: Alarm race handling should not discard preceding error\n\nCommit 795cda8338ea (\"rtc: interface: Fix long-standing race when setting\nalarm\") should not discard any errors from the preceding validations.\n\nPrior to that commit, if the alarm feature was disabled, or the\nset_alarm failed, a meaningful error code would be returned to the\ncaller for further action.\n\nAfter, more often than not, the __rtc_read_time will cause a success\nreturn code instead, misleading the caller.\n\nAn example of this is when timer_enqueue is called for a rtc-abx080x\ndevice. Since that driver does not clear the alarm feature bit, but\ninstead relies on the set_alarm operation to return invalid, the discard\nof the return code causes very different behaviour; i.e.\n hwclock: select() to /dev/rtc0 to wait for clock tick timed out\n\nFixes: 795cda8338ea (\"rtc: interface: Fix long-standing race when setting alarm\")\nSigned-off-by: Anthony Pighin (Nokia) \u003canthony.pighin@nokia.com\u003e\nReviewed-by: Esben Haabendal \u003cesben@geanix.com\u003e\nTested-by: Nick Bowler \u003cnbowler@draconx.ca\u003e\nLink: https://patch.msgid.link/BN0PR08MB6951415A751F236375A2945683D1A@BN0PR08MB6951.namprd08.prod.outlook.com\nSigned-off-by: Alexandre Belloni \u003calexandre.belloni@bootlin.com\u003e\n" }, { "commit": "43257b2ebd1f08536e3a8c06e2946af902015162", "tree": "40f2cb98fcad60cc3f1e8aa222b915be60d63aeb", "parents": [ "a95f71ad3e2e224277508e006580c333d0a5fe36", "c732084c891d89a1a0824126b3075999a981bc88" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:51:07 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:51:07 2026 -0800" }, "message": "Merge tag \u0027kmalloc_obj-prep-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux\n\nPull kmalloc_obj prep from Kees Cook:\n \"Fixes for return types to prepare for the kmalloc_obj treewide\n conversion, that haven\u0027t yet appeared during the merge window:\n dm-crypt, dm-zoned, drm/msm, and arm64 kvm\"\n\n* tag \u0027kmalloc_obj-prep-v7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:\n KVM: arm64: vgic: Handle const qualifier from gic_kvm_info allocation type\n drm/msm: Adjust msm_iommu_pagetable_prealloc_allocate() allocation type\n dm: dm-zoned: Adjust dmz_load_mapping() allocation type\n dm-crypt: Adjust crypt_alloc_tfms_aead() allocation type\n" }, { "commit": "ae9e8654579709c2f10b8c86a8467e1710d4599f", "tree": "ff6f132cedc127b03ee29a362c801af533946089", "parents": [ "7a4a583db62818a362c5721a7b4624248f967647", "2d01d88a532b8847757d3015c66beca315f496bf" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 06:30:06 2026 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 06:30:54 2026 +1000" }, "message": "Merge tag \u0027drm-xe-next-fixes-2026-02-19\u0027 of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next\n\nUAPI Changes:\n - drm/xe: Prevent VFs from exposing the CCS mode sysfs file (Nareshkumar)\n\nCross-subsystem Changes:\n - drm/pagemap: pass pagemap_addr by reference (Arnd)\n\nDriver Changes:\n - SRIOV related fixes (Michal, Piotr)\n - PAT cache fix (Jia)\n - MMIO read fix (Shuicheng)\n - W/a fixes (Roper)\n - Adjust type of xe_modparam.force_vram_bar_size (Shuicheng)\n - Wedge mode fix (Raag)\n - HWMon fix (Karthik)\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patch.msgid.link/aZeR6CXDRbeudIVR@intel.com\n" }, { "commit": "a95f71ad3e2e224277508e006580c333d0a5fe36", "tree": "7dabac863a327ff169b21e32d54d9eed324546a9", "parents": [ "f283371efd6a0ec1713349cd8c12fa5ffb19c60a", "561f60dfd1c10acaaf81bf523de2fe1ff39e2645" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:21:00 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:21:00 2026 -0800" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux\n\nPull ARM updates from Russell King:\n\n - avoid %pK for ARM MM prints\n\n - implement ARCH_HAS_CC_CAN_LINK to ensure runnable user progs\n\n - handle BE8 and BE32 for user progs\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux:\n ARM: 9470/1: Handle BE8 vs BE32 in ARCH_CC_CAN_LINK\n ARM: 9469/1: Implement ARCH_HAS_CC_CAN_LINK\n ARM: 9467/1: mm: Don\u0027t use %pK through printk\n" }, { "commit": "7a4a583db62818a362c5721a7b4624248f967647", "tree": "27c875ecc74e76ab343a14926f4576409b69a6f9", "parents": [ "7b5570df351317f9c74719e8a23fa3aec0739178", "777a02812f739af9e349cd0f695400a1d84053f4" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 06:13:33 2026 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 06:13:35 2026 +1000" }, "message": "Merge tag \u0027drm-intel-next-fixes-2026-02-20\u0027 of https://gitlab.freedesktop.org/drm/i915/kernel into drm-next\n\n- Add missing slice count check during DP mode validation\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Joonas Lahtinen \u003cjoonas.lahtinen@linux.intel.com\u003e\nLink: https://patch.msgid.link/aZgb43cDcTPBYssk@jlahtine-mobl\n" }, { "commit": "f283371efd6a0ec1713349cd8c12fa5ffb19c60a", "tree": "c530b22b9faf051019cd8025201bc16e5c3ace0a", "parents": [ "a27a5c0f08039475e21f3517e1bfe4a05bb82887", "948a013a54c47d5eba06e644b99d4927a8bc62f8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:04:40 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 12:04:40 2026 -0800" }, "message": "Merge tag \u0027efi-fixes-for-v7.0-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi\n\nPull EFI fixes from Ard Biesheuvel:\n \"Mixed bag of EFI tweaks and bug fixes:\n\n - Add a missing symbol export spotted by Arnd\u0027s randconfig testing\n\n - Fix kexec from a kernel booted with \u0027noefi\u0027\n\n - Fix memblock handling of the unaccepted memory table\n\n - Constify an occurrence of struct efivar_operations\n\n - Add Ilias as EFI reviewer\"\n\n* tag \u0027efi-fixes-for-v7.0-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi:\n efi: Align unaccepted memory range to page boundary\n efi: Fix reservation of unaccepted memory table\n MAINTAINERS: Add a reviewer entry for EFI\n efi: stmm: Constify struct efivar_operations\n x86/kexec: Copy ACPI root pointer address from config table\n efi: export sysfb_primary_display for EDID\n" }, { "commit": "7b5570df351317f9c74719e8a23fa3aec0739178", "tree": "769b9db8998ed965b83e6446461b42aa6ec97829", "parents": [ "3e2138510b2ad556a12a7b284a025a23992e4aa1", "3ce2ad267079fcdd2c087bbbfb4f91bec527a4d5" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 05:55:40 2026 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Feb 21 05:59:28 2026 +1000" }, "message": "Merge tag \u0027amd-drm-next-6.20-2026-02-19\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-next\n\namd-drm-next-6.20-2026-02-19:\n\namdgpu:\n- Fixes for DC analog support\n- DC FAMS fixes\n- DML 2.1 fixes\n- eDP fixes\n- Misc DC fixes\n- Fastboot fix\n- 3DLUT fixes\n- GPUVM fixes\n- 64bpp format fix\n- XGMI fix\n- Fix for MacBooks with switchable gfx\n\namdkfd:\n- Fix piority inversion with MQDs\n- NULL check fix\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patch.msgid.link/20260219172807.2451298-1-alexander.deucher@amd.com\n" }, { "commit": "a27a5c0f08039475e21f3517e1bfe4a05bb82887", "tree": "cb44e638433c2357108b743a8d9153a23d99aa44", "parents": [ "9806790115d1949ee0bd6ef8533293b5a293a1bb", "729a2e8e9ac47099a967567389cc9d73ef4194ca" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:44:39 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:44:39 2026 -0800" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fixes from Will Deacon:\n \"Two arm64 fixes: one fixes a warning that started showing up with\n gcc 16 and the other fixes a lockup in udelay() when running on a\n vCPU loaded on a CPU with the new-fangled WFIT instruction:\n\n - Fix compiler warning from huge_pte_clear() with GCC 16\n\n - Fix hang in udelay() on systems with WFIT by consistently using the\n virtual counter to calculate the delta\"\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: hugetlbpage: avoid unused-but-set-parameter warning (gcc-16)\n arm64: Force the use of CNTVCT_EL0 in __delay()\n" }, { "commit": "9806790115d1949ee0bd6ef8533293b5a293a1bb", "tree": "7fcdc6333a76ecfec0294264859ca757c9034e74", "parents": [ "9abf79529f4c4bc43464dcb82bfcc16297f90cc6", "123d2e75a14e8a21cecac3f3452221ba901f8111" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:24:45 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:24:45 2026 -0800" }, "message": "Merge tag \u0027s390-7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux\n\nPull s390 fixes from Heiko Carstens:\n\n - Make KEXEC_SIG available again for CONFIG_MODULES\u003dn\n\n - The s390 topology code used to call rebuild_sched_domains() before\n common code scheduling domains were setup. This was silently ignored\n by common code, but now results in a warning. Address by avoiding the\n early call\n\n - Convert debug area lock from spinlock to raw spinlock to address\n lockdep warnings\n\n - The recent 3490 tape device driver rework resulted in a different\n device driver name, which is visible via sysfs for user space. This\n breaks at least one user space application. Change the device driver\n name back to its old name to fix this\n\n* tag \u0027s390-7.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:\n s390/tape: Fix device driver name\n s390/debug: Convert debug area lock from a spinlock to a raw spinlock\n s390/smp: Avoid calling rebuild_sched_domains() early\n s390/kexec: Make KEXEC_SIG available when CONFIG_MODULES\u003dn\n" }, { "commit": "9abf79529f4c4bc43464dcb82bfcc16297f90cc6", "tree": "d4148aac086303bfebd786723e407716d5cb0f2e", "parents": [ "c8cb804a8add8d6f41b3805441b2ad16e30fbcc0", "0d4b3ca115b5436d7e68453527566db3e87ab7f0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:16:09 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 09:16:09 2026 -0800" }, "message": "Merge tag \u0027xtensa-20260219\u0027 of https://github.com/jcmvbkbc/linux-xtensa\n\nPull Xtensa update from Max Filippov:\n\n - fix unhandled case in the load/store fault handler\n in configurations with MMU\n\n* tag \u0027xtensa-20260219\u0027 of https://github.com/jcmvbkbc/linux-xtensa:\n xtensa: align: validate access in fast_load_store\n" }, { "commit": "c8cb804a8add8d6f41b3805441b2ad16e30fbcc0", "tree": "4e7d74208d7a2a623ce45f3f182297ccd046769d", "parents": [ "d31558c077d8be422b65e97974017c030b4bd91a", "5043d7ed39dfdce3107d6642aa31bb7736547104" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 08:57:35 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 08:57:35 2026 -0800" }, "message": "Merge tag \u0027for-linus-7.0-rc1a-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip\n\nPull xen fix from Juergen Gross:\n \"A single patch fixing a boot regression when running as a Xen PV\n guest. This issue was introduced in this merge window\"\n\n* tag \u0027for-linus-7.0-rc1a-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:\n x86/xen: Fix Xen PV guest boot\n" }, { "commit": "d31558c077d8be422b65e97974017c030b4bd91a", "tree": "6b894308c43fb455ce02e515dd6bd91d1412835d", "parents": [ "8bf22c33e7a172fbc72464f4cc484d23a6b412ba", "158ebb578cd5f7881fdc7c4ecebddcf9463f91fd" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 08:48:31 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 20 08:48:31 2026 -0800" }, "message": "Merge tag \u0027hyperv-next-signed-20260218\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux\n\nPull Hyper-V updates from Wei Liu:\n\n - Debugfs support for MSHV statistics (Nuno Das Neves)\n\n - Support for the integrated scheduler (Stanislav Kinsburskii)\n\n - Various fixes for MSHV memory management and hypervisor status\n handling (Stanislav Kinsburskii)\n\n - Expose more capabilities and flags for MSHV partition management\n (Anatol Belski, Muminul Islam, Magnus Kulke)\n\n - Miscellaneous fixes to improve code quality and stability (Carlos\n López, Ethan Nelson-Moore, Li RongQing, Michael Kelley, Mukesh\n Rathor, Purna Pavan Chandra Aekkaladevi, Stanislav Kinsburskii, Uros\n Bizjak)\n\n - PREEMPT_RT fixes for vmbus interrupts (Jan Kiszka)\n\n* tag \u0027hyperv-next-signed-20260218\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux: (34 commits)\n mshv: Handle insufficient root memory hypervisor statuses\n mshv: Handle insufficient contiguous memory hypervisor status\n mshv: Introduce hv_deposit_memory helper functions\n mshv: Introduce hv_result_needs_memory() helper function\n mshv: Add SMT_ENABLED_GUEST partition creation flag\n mshv: Add nested virtualization creation flag\n Drivers: hv: vmbus: Simplify allocation of vmbus_evt\n mshv: expose the scrub partition hypercall\n mshv: Add support for integrated scheduler\n mshv: Use try_cmpxchg() instead of cmpxchg()\n x86/hyperv: Fix error pointer dereference\n x86/hyperv: Reserve 3 interrupt vectors used exclusively by MSHV\n Drivers: hv: vmbus: Use kthread for vmbus interrupts on PREEMPT_RT\n x86/hyperv: Remove ASM_CALL_CONSTRAINT with VMMCALL insn\n x86/hyperv: Use savesegment() instead of inline asm() to save segment registers\n mshv: fix SRCU protection in irqfd resampler ack handler\n mshv: make field names descriptive in a header struct\n x86/hyperv: Update comment in hyperv_cleanup()\n mshv: clear eventfd counter on irqfd shutdown\n x86/hyperv: Use memremap()/memunmap() instead of ioremap_cache()/iounmap()\n ...\n" }, { "commit": "75f3cf0d1561b0a1cfa97ee35d78ca0118e584d5", "tree": "5ee5bba40a6d73b56b826df9841b61c65d2ae06d", "parents": [ "403faa575738a7f92267b2ca2ee56cd1b9373078" ], "author": { "name": "Mikhail Gavrilov", "email": "mikhail.v.gavrilov@gmail.com", "time": "Mon Feb 16 22:27:07 2026 +0500" }, "committer": { "name": "Gabriele Monaco", "email": "gmonaco@redhat.com", "time": "Fri Feb 20 13:12:00 2026 +0100" }, "message": "rv: Fix multiple definition of __pcpu_unique_da_mon_this\n\nThe refactoring in commit 30984ccf31b7 (\"rv: Refactor da_monitor to\nminimise macros\") replaced per-monitor unique variable names\n(da_mon_##name) with a fixed name (da_mon_this).\n\nWhile this works for \u0027static\u0027 variables (each translation unit gets its\nown copy), DEFINE_PER_CPU internally generates a non-static dummy\nvariable __pcpu_unique_\u003cn\u003e for each per-cpu definition. The requirement\nfor this variable to be unique although static exists for modules on\nspecific architectures (alpha) and if the kernel is built with\nCONFIG_DEBUG_FORCE_WEAK_PER_CPU (e.g. Fedora\u0027s debug kernel).\n\nWhen multiple per-cpu monitors (e.g. sco and sts) are built-in\nsimultaneously, they all produce the same __pcpu_unique_da_mon_this\nsymbol, causing a link error:\n\n ld: kernel/trace/rv/monitors/sts/sts.o: multiple definition of\n `__pcpu_unique_da_mon_this\u0027;\n kernel/trace/rv/monitors/sco/sco.o: first defined here\n\nFix this by introducing a DA_MON_NAME macro that expands to a\nper-monitor unique name (da_mon_\u003cMONITOR_NAME\u003e) via the existing\nCONCATENATE helper. This restores the uniqueness that was present\nbefore the refactoring.\n\nFixes: 30984ccf31b7 (\"rv: Refactor da_monitor to minimise macros\")\nReviewed-by: Gabriele Monaco \u003cgmonaco@redhat.com\u003e\nSigned-off-by: Mikhail Gavrilov \u003cmikhail.v.gavrilov@gmail.com\u003e\nLink: https://lore.kernel.org/r/20260216172707.1441516-1-mikhail.v.gavrilov@gmail.com\nSigned-off-by: Gabriele Monaco \u003cgmonaco@redhat.com\u003e\n" }, { "commit": "97b281d7edb2ae662365be2809cd728470119720", "tree": "26d882dc2e5ca6a765b68988aea0abfaa2319fcc", "parents": [ "a58b8764aed9648357b1c5b6368c9943ba33b7f9" ], "author": { "name": "Philipp Stanner", "email": "phasta@kernel.org", "time": "Mon Feb 16 14:16:15 2026 +0100" }, "committer": { "name": "Miguel Ojeda", "email": "ojeda@kernel.org", "time": "Fri Feb 20 01:59:51 2026 +0100" }, "message": "rust: list: Add unsafe blocks for container_of and safety comments\n\nimpl_list_item_mod.rs calls container_of! without unsafe blocks at a\ncouple of places. Since container_of! is unsafe, the blocks are strictly\nnecessary.\n\nThe problem was so far not visible because the \"unsafe-op-in-unsafe-fn\"\ncheck is a lint rather than a hard compiler error, and Rust suppresses\nlints triggered inside of a macro from another crate.\n\nThus, the error becomes only visible once someone from within the kernel\ncrate tries to use linked lists:\n\n error[E0133]: call to unsafe function `core::ptr::mut_ptr::\u003cimpl *mut T\u003e::byte_sub`\n is unsafe and requires unsafe block\n --\u003e rust/kernel/lib.rs:252:29\n |\n 252 | let container_ptr \u003d field_ptr.byte_sub(offset).cast::\u003c$Container\u003e();\n | ^^^^^^^^^^^^^^^^^^^^^^^^^^ call to unsafe function\n |\n ::: rust/kernel/drm/jq.rs:98:1\n |\n 98 | / impl_list_item! {\n 99 | | impl ListItem\u003c0\u003e for BasicItem { using ListLinks { self.links }; }\n 100 | | }\n | |_- in this macro invocation\n |\n note: an unsafe function restricts its caller, but its body is safe by default\n --\u003e rust/kernel/list/impl_list_item_mod.rs:216:13\n |\n 216 | unsafe fn view_value(me: *mut $crate::list::ListLinks\u003c$num\u003e) -\u003e *const Self {\n | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n |\n ::: rust/kernel/drm/jq.rs:98:1\n |\n 98 | / impl_list_item! {\n 99 | | impl ListItem\u003c0\u003e for BasicItem { using ListLinks { self.links }; }\n 100 | | }\n | |_- in this macro invocation\n \u003d note: requested on the command line with `-D unsafe-op-in-unsafe-fn`\n \u003d note: this error originates in the macro `$crate::container_of` which comes\n from the expansion of the macro `impl_list_item`\n\nTherefore, add unsafe blocks to container_of! calls to fix the issue.\n\n[ As discussed, let\u0027s fix the build for those that want to use the\n macro within the `kernel` crate now and we can discuss the proper\n safety comments afterwards. Thus I removed the ones from the patch.\n\n However, we cannot just avoid the comments with `CLIPPY\u003d1`, so I\n provided placeholders for now, like we did in the past. They were\n also needed for an `unsafe impl`.\n\n While I am not happy about it, it isn\u0027t worse than the current\n status (the comments were meant to be there), and at least this\n shows what is missing -- our pre-existing \"good first issue\" [1]\n may motivate new contributors to complete them properly.\n\n Finally, I moved one of the existing safety comments one line down\n so that Clippy could locate it.\n\n Link: https://github.com/Rust-for-Linux/linux/issues/351 [1]\n\n - Miguel ]\n\nCc: stable@vger.kernel.org\nFixes: c77f85b347dd (\"rust: list: remove OFFSET constants\")\nSuggested-by: Alice Ryhl \u003caliceryhl@google.com\u003e\nSigned-off-by: Philipp Stanner \u003cphasta@kernel.org\u003e\nReviewed-by: Gary Guo \u003cgary@garyguo.net\u003e\nReviewed-by: Alice Ryhl \u003caliceryhl@google.com\u003e\nLink: https://patch.msgid.link/20260216131613.45344-3-phasta@kernel.org\n[ Fixed formatting. Reworded to fix the lint suppression\n explanation. Indent build error. - Miguel ]\nSigned-off-by: Miguel Ojeda \u003cojeda@kernel.org\u003e\n" }, { "commit": "8c746e22096579897d1f8f74dbb6b17a6862fb6d", "tree": "3991f8f87c86e87d866944e1524f6e9d3f1a367a", "parents": [ "88a71941b90ddda44f4105e354e82a89b0389bc6" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Mon Feb 02 23:54:07 2026 +0900" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Thu Feb 19 15:34:19 2026 -0600" }, "message": "PCI: dwc: ep: Always clear IB maps on BAR update\n\ndw_pcie_ep_set_bar() currently tears down existing inbound mappings only\nwhen either the previous or the new struct pci_epf_bar uses submaps\n(num_submap !\u003d 0). If both the old and new mappings are BAR Match Mode,\nreprogramming the same ATU index is sufficient, so no explicit teardown\nwas needed.\n\nHowever, some callers may reuse the same struct pci_epf_bar instance and\nupdate it in place before calling set_bar() again. In that case\nep_func-\u003eepf_bar[bar] and the passed-in epf_bar can point to the same\nobject, so we cannot reliably distinguish BAR Match Mode -\u003e BAR Match Mode\nfrom Address Match Mode -\u003e BAR Match Mode. As a result, the conditional\nteardown based on num_submap becomes unreliable and existing inbound maps\nmay be left active.\n\nCall dw_pcie_ep_clear_ib_maps() unconditionally before reprogramming the\nBAR so that in-place updates are handled correctly.\n\nThis introduces a behavioral change in a corner case: if a BAR\nreprogramming attempt fails (especially for the long-standing BAR Match\nMode -\u003e BAR Match Mode update case), the previously programmed inbound\nmapping will already have been torn down. This should be acceptable, since\nthe caller observes the error and should not use the BAR for any real\ntransactions in that case.\n\nWhile at it, document that the existing update parameter check is\nbest-effort for in-place updates.\n\nFixes: cc839bef7727 (\"PCI: dwc: ep: Support BAR subrange inbound mapping via Address Match Mode iATU\")\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nReviewed-by: Niklas Cassel \u003ccassel@kernel.org\u003e\nLink: https://patch.msgid.link/20260202145407.503348-3-den@valinux.co.jp\n" }, { "commit": "88a71941b90ddda44f4105e354e82a89b0389bc6", "tree": "eaec49f70fa3e1900fdd427885747fbcd06e6d65", "parents": [ "5ddb66967924e38c680e6a304585c0f553681a3a" ], "author": { "name": "Koichiro Den", "email": "den@valinux.co.jp", "time": "Mon Feb 02 23:54:06 2026 +0900" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Thu Feb 19 15:34:19 2026 -0600" }, "message": "PCI: dwc: ep: Return after clearing BAR-match inbound mapping\n\ndw_pcie_ep_clear_ib_maps() first checks whether the inbound mapping for a\nBAR is in BAR Match Mode (tracked via ep_func-\u003ebar_to_atu[bar]). Once\nfound, the iATU region is disabled and the bookkeeping is cleared.\n\nBAR Match Mode and Address Match Mode mappings are mutually exclusive for a\ngiven BAR, so there is nothing left for the Address Match Mode teardown\npath to do after the BAR Match Mode mapping has been removed.\n\nReturn early after clearing the BAR Match Mode mapping to avoid running the\nAddress Match Mode teardown path. This makes the helper\u0027s intention\nexplicit and helps detect incorrect use of pci_epc_set_bar().\n\nSuggested-by: Niklas Cassel \u003ccassel@kernel.org\u003e\nSigned-off-by: Koichiro Den \u003cden@valinux.co.jp\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nReviewed-by: Niklas Cassel \u003ccassel@kernel.org\u003e\nLink: https://patch.msgid.link/20260202145407.503348-2-den@valinux.co.jp\n" }, { "commit": "5ddb66967924e38c680e6a304585c0f553681a3a", "tree": "84a0239299db88199cb0078c4e5d5d00554ec332", "parents": [ "a3b93b42238b3bb9eefc25729f567e66eea66a56" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Wed Feb 11 08:07:40 2026 +0100" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Thu Feb 19 15:34:19 2026 -0600" }, "message": "PCI: endpoint: pci-epf-test: Select configfs\n\nLike some of the other endpoint modules, pci-epf-test now also uses\nconfigfs, but is missing an indication in Kconfig:\n\n arm-linux-gnueabi-ld: drivers/pci/endpoint/functions/pci-epf-test.o: in function `pci_epf_test_add_cfs\u0027:\n pci-epf-test.c:(.text.pci_epf_test_add_cfs+0x2c): undefined reference to `config_group_init_type_name\u0027\n\nSelect the symbol as needed.\n\nFixes: ffcc4850a161 (\"PCI: endpoint: pci-epf-test: Allow overriding default BAR sizes\")\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nCloses: https://lore.kernel.org/oe-kbuild-all/202602180706.VtXkmtqL-lkp@intel.com/\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nReviewed-by: Niklas Cassel \u003ccassel@kernel.org\u003e\nReviewed-by: Damien Le Moal \u003cdlemoal@kernel.org\u003e\nLink: https://patch.msgid.link/20260211070812.4087119-1-arnd@kernel.org\n" }, { "commit": "a3b93b42238b3bb9eefc25729f567e66eea66a56", "tree": "f1d962ce7e94dfd21f08f6241a186b947c962f39", "parents": [ "e5f72cb9cea599dc9f5a9b80a33560a1d06f01cc" ], "author": { "name": "Ilpo Järvinen", "email": "ilpo.jarvinen@linux.intel.com", "time": "Thu Feb 19 00:34:18 2026 +0200" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Thu Feb 19 15:33:40 2026 -0600" }, "message": "PCI: Account fully optional bridge windows correctly\n\npbus_size_mem_optional() adds dev_res-\u003eadd_size of a bridge window into\nchildren_add_size when the window has a non-optional part. However, if the\nbridge window is fully optional, only r_size is added (which is zero for\nsuch a window).\n\nAlso, a second dev_res entry will be added by pci_dev_res_add_to_list()\ninto realloc_head for the bridge window (resulting in triggering the\nrealloc_head-must-be-fully-consumed sanity check after a single pass of the\nresource assignment algorithm):\n\n WARNING: drivers/pci/setup-bus.c:2153 at pci_assign_unassigned_root_bus_resources+0xa5/0x260\n\nCorrect these problems by always adding dev_res-\u003eadd_size for bridge\nwindows and not calling pci_dev_res_add_to_list() if the dev_res entry\nexists.\n\nFixes: 6a5e64c75e82 (\"PCI: Add pbus_mem_size_optional() to handle optional sizes\")\nReported-by: RavitejaX Veesam \u003cravitejax.veesam@intel.com\u003e\nSigned-off-by: Ilpo Järvinen \u003cilpo.jarvinen@linux.intel.com\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nTested-by: RavitejaX Veesam \u003cravitejax.veesam@intel.com\u003e\nLink: https://patch.msgid.link/20260218223419.22366-1-ilpo.jarvinen@linux.intel.com\n" }, { "commit": "9678e53179aa7e907360f5b5b275769008a69b80", "tree": "d508b7ecbb56ab93b4f2c93536832b44b2b0320b", "parents": [ "f0a0da1f907e8488826d91c465f7967a56a95aca" ], "author": { "name": "Petr Pavlu", "email": "petr.pavlu@suse.com", "time": "Thu Feb 19 17:27:02 2026 +0100" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Thu Feb 19 15:25:11 2026 -0500" }, "message": "tracing: Wake up poll waiters for hist files when removing an event\n\nThe event_hist_poll() function attempts to verify whether an event file is\nbeing removed, but this check may not occur or could be unnecessarily\ndelayed. This happens because hist_poll_wakeup() is currently invoked only\nfrom event_hist_trigger() when a hist command is triggered. If the event\nfile is being removed, no associated hist command will be triggered and a\nwaiter will be woken up only after an unrelated hist command is triggered.\n\nFix the issue by adding a call to hist_poll_wakeup() in\nremove_event_file_dir() after setting the EVENT_FILE_FL_FREED flag. This\nensures that a task polling on a hist file is woken up and receives\nEPOLLERR.\n\nCc: stable@vger.kernel.org\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nCc: Tom Zanussi \u003czanussi@kernel.org\u003e\nAcked-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nLink: https://patch.msgid.link/20260219162737.314231-3-petr.pavlu@suse.com\nFixes: 1bd13edbbed6 (\"tracing/hist: Add poll(POLLIN) support on hist file\")\nSigned-off-by: Petr Pavlu \u003cpetr.pavlu@suse.com\u003e\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "f0a0da1f907e8488826d91c465f7967a56a95aca", "tree": "45707f5d62201a87187a64add25b6b44b42da23f", "parents": [ "f4ff9f646a4d373f9e895c2f0073305da288bc0a" ], "author": { "name": "Petr Pavlu", "email": "petr.pavlu@suse.com", "time": "Thu Feb 19 17:27:01 2026 +0100" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Thu Feb 19 15:23:49 2026 -0500" }, "message": "tracing: Fix checking of freed trace_event_file for hist files\n\nThe event_hist_open() and event_hist_poll() functions currently retrieve\na trace_event_file pointer from a file struct by invoking\nevent_file_data(), which simply returns file-\u003ef_inode-\u003ei_private. The\nfunctions then check if the pointer is NULL to determine whether the event\nis still valid. This approach is flawed because i_private is assigned when\nan eventfs inode is allocated and remains set throughout its lifetime.\nInstead, the code should call event_file_file(), which checks for\nEVENT_FILE_FL_FREED. Using the incorrect access function may result in the\ncode potentially opening a hist file for an event that is being removed or\nbecoming stuck while polling on this file.\n\nCorrect the access method to event_file_file() in both functions.\n\nCc: stable@vger.kernel.org\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nCc: Tom Zanussi \u003czanussi@kernel.org\u003e\nLink: https://patch.msgid.link/20260219162737.314231-2-petr.pavlu@suse.com\nFixes: 1bd13edbbed6 (\"tracing/hist: Add poll(POLLIN) support on hist file\")\nSigned-off-by: Petr Pavlu \u003cpetr.pavlu@suse.com\u003e\nAcked-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "f4ff9f646a4d373f9e895c2f0073305da288bc0a", "tree": "1696644927339697198cb7e595d2628c98dff3cf", "parents": [ "912b0ee248c529a4f45d1e7f568dc1adddbf2a4a" ], "author": { "name": "Steven Rostedt", "email": "rostedt@goodmis.org", "time": "Wed Feb 18 10:42:44 2026 -0500" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Thu Feb 19 15:21:22 2026 -0500" }, "message": "fgraph: Do not call handlers direct when not using ftrace_ops\n\nThe function graph tracer was modified to us the ftrace_ops of the\nfunction tracer. This simplified the code as well as allowed more features\nof the function graph tracer.\n\nNot all architectures were converted over as it required the\nimplementation of HAVE_DYNAMIC_FTRACE_WITH_ARGS to implement. For those\narchitectures, it still did it the old way where the function graph tracer\nhandle was called by the function tracer trampoline. The handler then had\nto check the hash to see if the registered handlers wanted to be called by\nthat function or not.\n\nIn order to speed up the function graph tracer that used ftrace_ops, if\nonly one callback was registered with function graph, it would call its\nfunction directly via a static call.\n\nNow, if the architecture does not support the use of using ftrace_ops and\nstill has the ftrace function trampoline calling the function graph\nhandler, then by doing a direct call it removes the check against the\nhandler\u0027s hash (list of functions it wants callbacks to), and it may call\nthat handler for functions that the handler did not request calls for.\n\nOn 32bit x86, which does not support the ftrace_ops use with function\ngraph tracer, it shows the issue:\n\n ~# trace-cmd start -p function -l schedule\n ~# trace-cmd show\n # tracer: function_graph\n #\n # CPU DURATION FUNCTION CALLS\n # | | | | | | |\n 2) * 11898.94 us | schedule();\n 3) # 1783.041 us | schedule();\n 1) | schedule() {\n ------------------------------------------\n 1) bash-8369 \u003d\u003e kworker-7669\n ------------------------------------------\n 1) | schedule() {\n ------------------------------------------\n 1) kworker-7669 \u003d\u003e bash-8369\n ------------------------------------------\n 1) + 97.004 us | }\n 1) | schedule() {\n [..]\n\nNow by starting the function tracer is another instance:\n\n ~# trace-cmd start -B foo -p function\n\nThis causes the function graph tracer to trace all functions (because the\nfunction trace calls the function graph tracer for each on, and the\nfunction graph trace is doing a direct call):\n\n ~# trace-cmd show\n # tracer: function_graph\n #\n # CPU DURATION FUNCTION CALLS\n # | | | | | | |\n 1) 1.669 us | } /* preempt_count_sub */\n 1) + 10.443 us | } /* _raw_spin_unlock_irqrestore */\n 1) | tick_program_event() {\n 1) | clockevents_program_event() {\n 1) 1.044 us | ktime_get();\n 1) 6.481 us | lapic_next_event();\n 1) + 10.114 us | }\n 1) + 11.790 us | }\n 1) ! 181.223 us | } /* hrtimer_interrupt */\n 1) ! 184.624 us | } /* __sysvec_apic_timer_interrupt */\n 1) | irq_exit_rcu() {\n 1) 0.678 us | preempt_count_sub();\n\nWhen it should still only be tracing the schedule() function.\n\nTo fix this, add a macro FGRAPH_NO_DIRECT to be set to 0 when the\narchitecture does not support function graph use of ftrace_ops, and set to\n1 otherwise. Then use this macro to know to allow function graph tracer to\ncall the handlers directly or not.\n\nCc: stable@vger.kernel.org\nCc: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nCc: Mark Rutland \u003cmark.rutland@arm.com\u003e\nLink: https://patch.msgid.link/20260218104244.5f14dade@gandalf.local.home\nFixes: cc60ee813b503 (\"function_graph: Use static_call and branch to optimize entry function\")\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "912b0ee248c529a4f45d1e7f568dc1adddbf2a4a", "tree": "becc73482af4fecf9344d9dd3bc4d67db74dfae9", "parents": [ "f1547779402c4cd67755c33616b7203baa88420b" ], "author": { "name": "Masami Hiramatsu (Google)", "email": "mhiramat@kernel.org", "time": "Mon Feb 16 18:30:15 2026 +0900" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Thu Feb 19 15:21:12 2026 -0500" }, "message": "tracing: ring-buffer: Fix to check event length before using\n\nCheck the event length before adding it for accessing next index in\nrb_read_data_buffer(). Since this function is used for validating\npossibly broken ring buffers, the length of the event could be broken.\nIn that case, the new event (e + len) can point a wrong address.\nTo avoid invalid memory access at boot, check whether the length of\neach event is in the possible range before using it.\n\nCc: stable@vger.kernel.org\nCc: Mathieu Desnoyers \u003cmathieu.desnoyers@efficios.com\u003e\nFixes: 5f3b6e839f3c (\"ring-buffer: Validate boot range memory events\")\nLink: https://patch.msgid.link/177123421541.142205.9414352170164678966.stgit@devnote2\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "f1547779402c4cd67755c33616b7203baa88420b", "tree": "45761ce249770cee732fdbece161579089671973", "parents": [ "3c6e577d5ae705edebed9882ff474d7a48a47dd2" ], "author": { "name": "Daniil Dulov", "email": "d.dulov@aladdin.ru", "time": "Fri Feb 13 13:01:30 2026 +0300" }, "committer": { "name": "Steven Rostedt (Google)", "email": "rostedt@goodmis.org", "time": "Thu Feb 19 15:20:41 2026 -0500" }, "message": "ring-buffer: Fix possible dereference of uninitialized pointer\n\nThere is a pointer head_page in rb_meta_validate_events() which is not\ninitialized at the beginning of a function. This pointer can be dereferenced\nif there is a failure during reader page validation. In this case the control\nis passed to \"invalid\" label where the pointer is dereferenced in a loop.\n\nTo fix the issue initialize orig_head and head_page before calling\nrb_validate_buffer.\n\nFound by Linux Verification Center (linuxtesting.org) with SVACE.\n\nCc: stable@vger.kernel.org\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nReported-by: Dan Carpenter \u003cdan.carpenter@linaro.org\u003e\nAcked-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e\nLink: https://patch.msgid.link/20260213100130.2013839-1-d.dulov@aladdin.ru\nCloses: https://lore.kernel.org/r/202406130130.JtTGRf7W-lkp@intel.com/\nFixes: 5f3b6e839f3c (\"ring-buffer: Validate boot range memory events\")\nSigned-off-by: Daniil Dulov \u003cd.dulov@aladdin.ru\u003e\nSigned-off-by: Steven Rostedt (Google) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "8bf22c33e7a172fbc72464f4cc484d23a6b412ba", "tree": "7034d84f09ee8c239574adec764ddae7594775f0", "parents": [ "4f13d0dabc87fb585b96d90cc4b29f67a2995405", "571dcbeb8e635182bb825ae758399831805693c2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 19 10:39:08 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 19 10:39:08 2026 -0800" }, "message": "Merge tag \u0027net-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Jakub Kicinski:\n \"Including fixes from Netfilter.\n\n Current release - new code bugs:\n\n - net: fix backlog_unlock_irq_restore() vs CONFIG_PREEMPT_RT\n\n - eth: mlx5e: XSK, Fix unintended ICOSQ change\n\n - phy_port: correctly recompute the port\u0027s linkmodes\n\n - vsock: prevent child netns mode switch from local to global\n\n - couple of kconfig fixes for new symbols\n\n Previous releases - regressions:\n\n - nfc: nci: fix false-positive parameter validation for packet data\n\n - net: do not delay zero-copy skbs in skb_attempt_defer_free()\n\n Previous releases - always broken:\n\n - mctp: ensure our nlmsg responses to user space are zero-initialised\n\n - ipv6: ioam: fix heap buffer overflow in __ioam6_fill_trace_data()\n\n - fixes for ICMP rate limiting\n\n Misc:\n\n - intel: fix PCI device ID conflict between i40e and ipw2200\"\n\n* tag \u0027net-7.0-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (85 commits)\n net: nfc: nci: Fix parameter validation for packet data\n net/mlx5e: Use unsigned for mlx5e_get_max_num_channels\n net/mlx5e: Fix deadlocks between devlink and netdev instance locks\n net/mlx5e: MACsec, add ASO poll loop in macsec_aso_set_arm_event\n net/mlx5: Fix misidentification of write combining CQE during poll loop\n net/mlx5e: Fix misidentification of ASO CQE during poll loop\n net/mlx5: Fix multiport device check over light SFs\n bonding: alb: fix UAF in rlb_arp_recv during bond up/down\n bnge: fix reserving resources from FW\n eth: fbnic: Advertise supported XDP features.\n rds: tcp: fix uninit-value in __inet_bind\n net/rds: Fix NULL pointer dereference in rds_tcp_accept_one\n octeontx2-af: Fix default entries mcam entry action\n net/mlx5e: XSK, Fix unintended ICOSQ change\n ipv6: icmp: icmpv6_xrlim_allow() optimization if net.ipv6.icmp.ratelimit is zero\n ipv4: icmp: icmpv4_xrlim_allow() optimization if net.ipv4.icmp_ratelimit is zero\n ipv6: icmp: remove obsolete code in icmpv6_xrlim_allow()\n inet: move icmp_global_{credit,stamp} to a separate cache line\n icmp: prevent possible overflow in icmp_global_allow()\n selftests/net: packetdrill: add ipv4-mapped-ipv6 tests\n ...\n" }, { "commit": "4f13d0dabc87fb585b96d90cc4b29f67a2995405", "tree": "a84b90a0f0473bd81ce7af04582fbfc5342cc603", "parents": [ "2b7a25df823dc7d8f56f8ce7c2d2dac391cea9c2", "1e5c009126952f673ffa2427acbd69e57493f0d2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 19 10:36:54 2026 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Feb 19 10:36:54 2026 -0800" }, "message": "Merge tag \u0027bpf-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf\n\nPull bpf fixes from Alexei Starovoitov:\n\n - Fix invalid write loop logic in libbpf\u0027s bpf_linker__add_buf() (Amery\n Hung)\n\n - Fix a potential use-after-free of BTF object (Anton Protopopov)\n\n - Add feature detection to libbpf and avoid moving arena global\n variables on older kernels (Emil Tsalapatis)\n\n - Remove extern declaration of bpf_stream_vprintk() from libbpf headers\n (Ihor Solodrai)\n\n - Fix truncated netlink dumps in bpftool (Jakub Kicinski)\n\n - Fix map_kptr grace period wait in bpf selftests (Kumar Kartikeya\n Dwivedi)\n\n - Remove hexdump dependency while building bpf selftests (Matthieu\n Baerts)\n\n - Complete fsession support in BPF trampolines on riscv (Menglong Dong)\n\n* tag \u0027bpf-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:\n selftests/bpf: Remove hexdump dependency\n libbpf: Remove extern declaration of bpf_stream_vprintk()\n selftests/bpf: Use vmlinux.h in test_xdp_meta\n bpftool: Fix truncated netlink dumps\n libbpf: Delay feature gate check until object prepare time\n libbpf: Do not use PROG_TYPE_TRACEPOINT program for feature gating\n bpf: Add a map/btf from a fd array more consistently\n selftests/bpf: Fix map_kptr grace period wait\n selftests/bpf: enable fsession_test on riscv64\n selftests/bpf: Adjust selftest due to function rename\n bpf, riscv: add fsession support for trampolines\n bpf: Fix a potential use-after-free of BTF object\n bpf, riscv: introduce emit_store_stack_imm64() for trampoline\n libbpf: Fix invalid write loop logic in bpf_linker__add_buf()\n libbpf: Add gating for arena globals relocation feature\n" }, { "commit": "c732084c891d89a1a0824126b3075999a981bc88", "tree": "2ed25ef62e113baec24e3d459d294835886044ab", "parents": [ "48634a9ea06a453a3f0635a7063dcc63f6729436" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Feb 06 14:30:23 2026 -0800" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Thu Feb 19 10:14:02 2026 -0800" }, "message": "KVM: arm64: vgic: Handle const qualifier from gic_kvm_info allocation type\n\nIn preparation for making the kmalloc family of allocators type aware,\nwe need to make sure that the returned type from the allocation matches\nthe type of the variable being assigned. (Before, the allocator would\nalways return \"void *\", which can be implicitly cast to any pointer type.)\n\nThe assigned type is \"struct gic_kvm_info\", but the returned type,\nwhile matching, is const qualified. To get them exactly matching, just\nuse the dereferenced pointer for the sizeof().\n\nLink: https://patch.msgid.link/20260206223022.it.052-kees@kernel.org\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "48634a9ea06a453a3f0635a7063dcc63f6729436", "tree": "bd9972071f350ef44be136cfe6dee953bc0e68c4", "parents": [ "a3fab954405e61f593f57bdb3510e304c6a0aa80" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Feb 06 14:21:52 2026 -0800" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Thu Feb 19 10:07:01 2026 -0800" }, "message": "drm/msm: Adjust msm_iommu_pagetable_prealloc_allocate() allocation type\n\nIn preparation for making the kmalloc family of allocators type aware,\nwe need to make sure that the returned type from the allocation matches\nthe type of the variable being assigned. (Before, the allocator would\nalways return \"void *\", which can be implicitly cast to any pointer type.)\n\nThe assigned type is \"void **\" but the returned type will be \"void ***\".\nThese are the same allocation size (pointer size), but the types do not\nmatch. Adjust the allocation type to match the assignment.\n\nReviewed-by: Dmitry Baryshkov \u003cdmitry.baryshkov@oss.qualcomm.com\u003e\nLink: https://patch.msgid.link/20260206222151.work.016-kees@kernel.org\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "a3fab954405e61f593f57bdb3510e304c6a0aa80", "tree": "e4a127e898cdac9f85e3c18deca954eaec7b2842", "parents": [ "c6fcccd5398775624582777776113dd220db2bbd" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Apr 25 23:17:08 2025 -0700" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Thu Feb 19 10:05:20 2026 -0800" }, "message": "dm: dm-zoned: Adjust dmz_load_mapping() allocation type\n\nIn preparation for making the kmalloc family of allocators type aware,\nwe need to make sure that the returned type from the allocation matches\nthe type of the variable being assigned. (Before, the allocator would\nalways return \"void *\", which can be implicitly cast to any pointer type.)\n\nThe assigned type is \"struct dmz_mblock **\" but the returned type will\nbe \"struct dmz_mblk **\". These are the same allocation size (pointer\nsize), but the types do not match. Adjust the allocation type to match\nthe assignment.\n\nLink: https://patch.msgid.link/20250426061707.work.587-kees@kernel.org\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "c6fcccd5398775624582777776113dd220db2bbd", "tree": "083a0296ed54052c4851b6b6c7cfcb197df5884e", "parents": [ "2b7a25df823dc7d8f56f8ce7c2d2dac391cea9c2" ], "author": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Fri Apr 25 23:16:30 2025 -0700" }, "committer": { "name": "Kees Cook", "email": "kees@kernel.org", "time": "Thu Feb 19 10:05:20 2026 -0800" }, "message": "dm-crypt: Adjust crypt_alloc_tfms_aead() allocation type\n\nIn preparation for making the kmalloc family of allocators type aware,\nwe need to make sure that the returned type from the allocation matches\nthe type of the variable being assigned. (Before, the allocator would\nalways return \"void *\", which can be implicitly cast to any pointer type.)\n\nThe assigned type is \"struct crypto_skcipher **\" but the returned type\nwill be \"struct crypto_aead **\". These are the same allocation size\n(pointer size), but the types don\u0027t match. Adjust the allocation type\nto match the assignment.\n\nLink: https://patch.msgid.link/20250426061629.work.266-kees@kernel.org\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\n" }, { "commit": "571dcbeb8e635182bb825ae758399831805693c2", "tree": "36f9bbf30544f2d9af0b64498f41024687445ee2", "parents": [ "7997bca6c2b00310161fbcd628dc0161fd3b040c" ], "author": { "name": "Michael Thalmeier", "email": "michael.thalmeier@hale.at", "time": "Wed Feb 18 09:30:00 2026 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:32:51 2026 -0800" }, "message": "net: nfc: nci: Fix parameter validation for packet data\n\nSince commit 9c328f54741b (\"net: nfc: nci: Add parameter validation for\npacket data\") communication with nci nfc chips is not working any more.\n\nThe mentioned commit tries to fix access of uninitialized data, but\nfailed to understand that in some cases the data packet is of variable\nlength and can therefore not be compared to the maximum packet length\ngiven by the sizeof(struct).\n\nFixes: 9c328f54741b (\"net: nfc: nci: Add parameter validation for packet data\")\nCc: stable@vger.kernel.org\nSigned-off-by: Michael Thalmeier \u003cmichael.thalmeier@hale.at\u003e\nReported-by: syzbot+740e04c2a93467a0f8c8@syzkaller.appspotmail.com\nLink: https://patch.msgid.link/20260218083000.301354-1-michael.thalmeier@hale.at\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "7997bca6c2b00310161fbcd628dc0161fd3b040c", "tree": "a19375288eea1b81ada5d776053556cb443cac99", "parents": [ "e6834a4c474697df23ab9948fd3577b26bf48656", "57a94d4b22b0c6cc5d601e6b6238d78fb923d991" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:26:39 2026 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:23 2026 -0800" }, "message": "Merge branch \u0027mlx5-misc-fixes-2026-02-18\u0027\n\nTariq Toukan says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlx5 misc fixes 2026-02-18\n\nThis patchset provides misc bug fixes from the team to the mlx5\ncore and Eth drivers.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20260218072904.1764634-1-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "57a94d4b22b0c6cc5d601e6b6238d78fb923d991", "tree": "a19375288eea1b81ada5d776053556cb443cac99", "parents": [ "83ac0304a2d77519dae1e54c9713cbe1aedf19c9" ], "author": { "name": "Cosmin Ratiu", "email": "cratiu@nvidia.com", "time": "Wed Feb 18 09:29:04 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:19 2026 -0800" }, "message": "net/mlx5e: Use unsigned for mlx5e_get_max_num_channels\n\nThe max number of channels is always an unsigned int, use the correct\ntype to fix compilation errors done with strict type checking, e.g.:\n\nerror: call to ‘__compiletime_assert_1110’ declared with attribute\n error: min(mlx5e_get_devlink_param_num_doorbells(mdev),\n mlx5e_get_max_num_channels(mdev)) signedness error\n\nFixes: 74a8dadac17e (\"net/mlx5e: Preparations for supporting larger number of channels\")\nSigned-off-by: Cosmin Ratiu \u003ccratiu@nvidia.com\u003e\nReviewed-by: Dragos Tatulea \u003cdtatulea@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-7-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "83ac0304a2d77519dae1e54c9713cbe1aedf19c9", "tree": "c0eae92a12d861d7a8a479dde2a12149e4c93d28", "parents": [ "9854b243ce4225328d0b32fdc998d35b6952d3f7" ], "author": { "name": "Cosmin Ratiu", "email": "cratiu@nvidia.com", "time": "Wed Feb 18 09:29:03 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:16 2026 -0800" }, "message": "net/mlx5e: Fix deadlocks between devlink and netdev instance locks\n\nIn the mentioned \"Fixes\" commit, various work tasks triggering devlink\nhealth reporter recovery were switched to use netdev_trylock to protect\nagainst concurrent tear down of the channels being recovered. But this\nhad the side effect of introducing potential deadlocks because of\nincorrect lock ordering.\n\nThe correct lock order is described by the init flow:\nprobe_one -\u003e mlx5_init_one (acquires devlink lock)\n-\u003e mlx5_init_one_devl_locked -\u003e mlx5_register_device\n-\u003e mlx5_rescan_drivers_locked -...-\u003e mlx5e_probe -\u003e _mlx5e_probe\n-\u003e register_netdev (acquires rtnl lock)\n-\u003e register_netdevice (acquires netdev lock)\n\u003d\u003e devlink lock -\u003e rtnl lock -\u003e netdev lock.\n\nBut in the current recovery flow, the order is wrong:\nmlx5e_tx_err_cqe_work (acquires netdev lock)\n-\u003e mlx5e_reporter_tx_err_cqe -\u003e mlx5e_health_report\n-\u003e devlink_health_report (acquires devlink lock \u003d\u003e boom!)\n-\u003e devlink_health_reporter_recover\n-\u003e mlx5e_tx_reporter_recover -\u003e mlx5e_tx_reporter_recover_from_ctx\n-\u003e mlx5e_tx_reporter_err_cqe_recover\n\nThe same pattern exists in:\nmlx5e_reporter_rx_timeout\nmlx5e_reporter_tx_ptpsq_unhealthy\nmlx5e_reporter_tx_timeout\n\nFix these by moving the netdev_trylock calls from the work handlers\nlower in the call stack, in the respective recovery functions, where\nthey are actually necessary.\n\nFixes: 8f7b00307bf1 (\"net/mlx5e: Convert mlx5 netdevs to instance locking\")\nSigned-off-by: Cosmin Ratiu \u003ccratiu@nvidia.com\u003e\nReviewed-by: Dragos Tatulea \u003cdtatulea@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-6-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "9854b243ce4225328d0b32fdc998d35b6952d3f7", "tree": "a3c7f1cc038ee2a550b77cdefbd02deb3005d9b8", "parents": [ "d451994ebc7d4392610bd4b2ab339b255deb4143" ], "author": { "name": "Gal Pressman", "email": "gal@nvidia.com", "time": "Wed Feb 18 09:29:02 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:14 2026 -0800" }, "message": "net/mlx5e: MACsec, add ASO poll loop in macsec_aso_set_arm_event\n\nThe macsec_aso_set_arm_event function calls mlx5_aso_poll_cq once\nwithout a retry loop. If the CQE is not immediately available after\nposting the WQE, the function fails unnecessarily.\n\nUse read_poll_timeout() to poll 3-10 usecs for CQE, consistent with\nother ASO polling code paths in the driver.\n\nFixes: 739cfa34518e (\"net/mlx5: Make ASO poll CQ usable in atomic context\")\nSigned-off-by: Gal Pressman \u003cgal@nvidia.com\u003e\nReviewed-by: Jianbo Liu \u003cjianbol@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-5-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d451994ebc7d4392610bd4b2ab339b255deb4143", "tree": "b036ee67349d7ce0f48b5724f330a1d6b6f7013b", "parents": [ "ae3cb71e6c4dbda0c0b7c10475b744377813c7bd" ], "author": { "name": "Gal Pressman", "email": "gal@nvidia.com", "time": "Wed Feb 18 09:29:01 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:11 2026 -0800" }, "message": "net/mlx5: Fix misidentification of write combining CQE during poll loop\n\nThe write combining completion poll loop uses usleep_range() which can\nsleep much longer than requested due to scheduler latency. Under load,\nwe witnessed a 20ms+ delay until the process was rescheduled, causing\nthe jiffies based timeout to expire while the thread is sleeping.\n\nThe original do-while loop structure (poll, sleep, check timeout) would\nexit without a final poll when waking after timeout, missing a CQE that\narrived during sleep.\n\nInstead of the open-coded while loop, use the kernel\u0027s poll_timeout_us()\nwhich always performs an additional check after the sleep expiration,\nand is less error-prone.\n\nNote: poll_timeout_us() doesn\u0027t accept a sleep range, by passing 10\nsleep_us the sleep range effectively changes from 2-10 to 3-10 usecs.\n\nFixes: d98995b4bf98 (\"net/mlx5: Reimplement write combining test\")\nSigned-off-by: Gal Pressman \u003cgal@nvidia.com\u003e\nReviewed-by: Jianbo Liu \u003cjianbol@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-4-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "ae3cb71e6c4dbda0c0b7c10475b744377813c7bd", "tree": "37d27d739ed99ec931ab358e30e8ad4f44901fb3", "parents": [ "47bf2e813817159f4d195be83a9b5a640ee6baec" ], "author": { "name": "Gal Pressman", "email": "gal@nvidia.com", "time": "Wed Feb 18 09:29:00 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:08 2026 -0800" }, "message": "net/mlx5e: Fix misidentification of ASO CQE during poll loop\n\nThe ASO completion poll loop uses usleep_range() which can sleep much\nlonger than requested due to scheduler latency. Under load, we witnessed\na 20ms+ delay until the process was rescheduled, causing the jiffies\nbased timeout to expire while the thread is sleeping.\n\nThe original do-while loop structure (poll, sleep, check timeout) would\nexit without a final poll when waking after timeout, missing a CQE that\narrived during sleep.\n\nInstead of the open-coded while loop, use the kernel\u0027s\nread_poll_timeout() which always performs an additional check after the\nsleep expiration, and is less error-prone.\n\nNote: read_poll_timeout() doesn\u0027t accept a sleep range, by passing 10\nsleep_us the sleep range effectively changes from 2-10 to 3-10 usecs.\n\nFixes: 739cfa34518e (\"net/mlx5: Make ASO poll CQ usable in atomic context\")\nFixes: 7e3fce82d945 (\"net/mlx5e: Overcome slow response for first macsec ASO WQE\")\nSigned-off-by: Gal Pressman \u003cgal@nvidia.com\u003e\nReviewed-by: Jianbo Liu \u003cjianbol@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-3-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "47bf2e813817159f4d195be83a9b5a640ee6baec", "tree": "9466c839425e37b5c858964241944f68b50175a3", "parents": [ "e6834a4c474697df23ab9948fd3577b26bf48656" ], "author": { "name": "Shay Drory", "email": "shayd@nvidia.com", "time": "Wed Feb 18 09:28:59 2026 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:05 2026 -0800" }, "message": "net/mlx5: Fix multiport device check over light SFs\n\nDriver is using num_vhca_ports capability to distinguish between\nmultiport master device and multiport slave device. num_vhca_ports is a\ncapability the driver sets according to the MAX num_vhca_ports\ncapability reported by FW. On the other hand, light SFs doesn\u0027t set the\nabove capbility.\n\nThis leads to wrong results whenever light SFs is checking whether he is\na multiport master or slave.\n\nTherefore, use the MAX capability to distinguish between master and\nslave devices.\n\nFixes: e71383fb9cd1 (\"net/mlx5: Light probe local SFs\")\nSigned-off-by: Shay Drory \u003cshayd@nvidia.com\u003e\nReviewed-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Jacob Keller \u003cJacob.e.keller@intel.com\u003e\nLink: https://patch.msgid.link/20260218072904.1764634-2-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e6834a4c474697df23ab9948fd3577b26bf48656", "tree": "612dc7815629329a05d0a417ba2c7b437f4a32f5", "parents": [ "604530085b2ef484843c723a105b6fd3218b4710" ], "author": { "name": "Hangbin Liu", "email": "liuhangbin@gmail.com", "time": "Wed Feb 18 06:09:19 2026 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:30:01 2026 -0800" }, "message": "bonding: alb: fix UAF in rlb_arp_recv during bond up/down\n\nThe ALB RX path may access rx_hashtbl concurrently with bond\nteardown. During rapid bond up/down cycles, rlb_deinitialize()\nfrees rx_hashtbl while RX handlers are still running, leading\nto a null pointer dereference detected by KASAN.\n\nHowever, the root cause is that rlb_arp_recv() can still be accessed\nafter setting recv_probe to NULL, which is actually a use-after-free\n(UAF) issue. That is the reason for using the referenced commit in the\nFixes tag.\n\n[ 214.174138] Oops: general protection fault, probably for non-canonical address 0xdffffc000000001d: 0000 [#1] SMP KASAN PTI\n[ 214.186478] KASAN: null-ptr-deref in range [0x00000000000000e8-0x00000000000000ef]\n[ 214.194933] CPU: 30 UID: 0 PID: 2375 Comm: ping Kdump: loaded Not tainted 6.19.0-rc8+ #2 PREEMPT(voluntary)\n[ 214.205907] Hardware name: Dell Inc. PowerEdge R730/0WCJNT, BIOS 2.14.0 01/14/2022\n[ 214.214357] RIP: 0010:rlb_arp_recv+0x505/0xab0 [bonding]\n[ 214.220320] Code: 0f 85 2b 05 00 00 48 b8 00 00 00 00 00 fc ff df 40 0f b6 ed 48 c1 e5 06 49 03 ad 78 01 00 00 48 8d 7d 28 48 89 fa 48 c1 ea 03 \u003c0f\u003e b6\n 04 02 84 c0 74 06 0f 8e 12 05 00 00 80 7d 28 00 0f 84 8c 00\n[ 214.241280] RSP: 0018:ffffc900073d8870 EFLAGS: 00010206\n[ 214.247116] RAX: dffffc0000000000 RBX: ffff888168556822 RCX: ffff88816855681e\n[ 214.255082] RDX: 000000000000001d RSI: dffffc0000000000 RDI: 00000000000000e8\n[ 214.263048] RBP: 00000000000000c0 R08: 0000000000000002 R09: ffffed11192021c8\n[ 214.271013] R10: ffff8888c9010e43 R11: 0000000000000001 R12: 1ffff92000e7b119\n[ 214.278978] R13: ffff8888c9010e00 R14: ffff888168556822 R15: ffff888168556810\n[ 214.286943] FS: 00007f85d2d9cb80(0000) GS:ffff88886ccb3000(0000) knlGS:0000000000000000\n[ 214.295966] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[ 214.302380] CR2: 00007f0d047b5e34 CR3: 00000008a1c2e002 CR4: 00000000001726f0\n[ 214.310347] Call Trace:\n[ 214.313070] \u003cIRQ\u003e\n[ 214.315318] ? __pfx_rlb_arp_recv+0x10/0x10 [bonding]\n[ 214.320975] bond_handle_frame+0x166/0xb60 [bonding]\n[ 214.326537] ? __pfx_bond_handle_frame+0x10/0x10 [bonding]\n[ 214.332680] __netif_receive_skb_core.constprop.0+0x576/0x2710\n[ 214.339199] ? __pfx_arp_process+0x10/0x10\n[ 214.343775] ? sched_balance_find_src_group+0x98/0x630\n[ 214.349513] ? __pfx___netif_receive_skb_core.constprop.0+0x10/0x10\n[ 214.356513] ? arp_rcv+0x307/0x690\n[ 214.360311] ? __pfx_arp_rcv+0x10/0x10\n[ 214.364499] ? __lock_acquire+0x58c/0xbd0\n[ 214.368975] __netif_receive_skb_one_core+0xae/0x1b0\n[ 214.374518] ? __pfx___netif_receive_skb_one_core+0x10/0x10\n[ 214.380743] ? lock_acquire+0x10b/0x140\n[ 214.385026] process_backlog+0x3f1/0x13a0\n[ 214.389502] ? process_backlog+0x3aa/0x13a0\n[ 214.394174] __napi_poll.constprop.0+0x9f/0x370\n[ 214.399233] net_rx_action+0x8c1/0xe60\n[ 214.403423] ? __pfx_net_rx_action+0x10/0x10\n[ 214.408193] ? lock_acquire.part.0+0xbd/0x260\n[ 214.413058] ? sched_clock_cpu+0x6c/0x540\n[ 214.417540] ? mark_held_locks+0x40/0x70\n[ 214.421920] handle_softirqs+0x1fd/0x860\n[ 214.426302] ? __pfx_handle_softirqs+0x10/0x10\n[ 214.431264] ? __neigh_event_send+0x2d6/0xf50\n[ 214.436131] do_softirq+0xb1/0xf0\n[ 214.439830] \u003c/IRQ\u003e\n\nThe issue is reproducible by repeatedly running\nip link set bond0 up/down while receiving ARP messages, where\nrlb_arp_recv() can race with rlb_deinitialize() and dereference\na freed rx_hashtbl entry.\n\nFix this by setting recv_probe to NULL and then calling\nsynchronize_net() to wait for any concurrent RX processing to finish.\nThis ensures that no RX handler can access rx_hashtbl after it is freed\nin bond_alb_deinitialize().\n\nReported-by: Liang Li \u003cliali@redhat.com\u003e\nFixes: 3aba891dde38 (\"bonding: move processing of recv handlers into handle_frame()\")\nReviewed-by: Nikolay Aleksandrov \u003cnikolay@nvidia.com\u003e\nAcked-by: Jay Vosburgh \u003cjv@jvosburgh.net\u003e\nSigned-off-by: Hangbin Liu \u003cliuhangbin@gmail.com\u003e\nLink: https://patch.msgid.link/20260218060919.101574-1-liuhangbin@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "604530085b2ef484843c723a105b6fd3218b4710", "tree": "57d93e094890e50da739eeed5fd23a94f804a6c1", "parents": [ "e977fcb3a318b53b47f23b44ac237fceb1b731fe" ], "author": { "name": "Vikas Gupta", "email": "vikas.gupta@broadcom.com", "time": "Wed Feb 18 10:57:55 2026 +0530" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Feb 19 09:29:59 2026 -0800" }, "message": "bnge: fix reserving resources from FW\n\nHWRM_FUNC_CFG is used to reserve resources, whereas HWRM_FUNC_QCFG is\nintended for querying resource information from the firmware.\nSince __bnge_hwrm_reserve_pf_rings() reserves resources for a specific\nPF, the command type should be HWRM_FUNC_CFG.\n\nFixes: 627c67f038d2 (\"bng_en: Add resource management support\")\nSigned-off-by: Vikas Gupta \u003cvikas.gupta@broadcom.com\u003e\nReviewed-by: Bhargava Chenna Marreddy \u003cbhargava.marreddy@broadcom.com\u003e\nReviewed-by: Simon Horman \u003chorms@kernel.org\u003e\nLink: https://patch.msgid.link/20260218052755.4097468-1-vikas.gupta@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3ce2ad267079fcdd2c087bbbfb4f91bec527a4d5", "tree": "769b9db8998ed965b83e6446461b42aa6ec97829", "parents": [ "4bd8b5f8bcb57b430c35494d8a2471ce5fd7661d" ], "author": { "name": "Timur Kristóf", "email": "timur.kristof@gmail.com", "time": "Mon Feb 02 12:25:08 2026 +0100" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Thu Feb 19 12:16:12 2026 -0500" }, "message": "drm/amd/display: Remove unneeded DAC link encoder register\n\nNot needed anymore since we use the VBIOS function.\n\nSigned-off-by: Timur Kristóf \u003ctimur.kristof@gmail.com\u003e\nTested-by: Mauro Rossi \u003cissor.oruam@gmail.com\u003e\nReviewed-by: Alex Hung \u003calex.hung@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "4bd8b5f8bcb57b430c35494d8a2471ce5fd7661d", "tree": "68c12bb6b1a31896cdc91dff585ff9b25c0e1b40", "parents": [ "cbced93894d145239c83881d7fd953b7392c23a8" ], "author": { "name": "Timur Kristóf", "email": "timur.kristof@gmail.com", "time": "Mon Feb 02 12:25:07 2026 +0100" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Thu Feb 19 12:16:12 2026 -0500" }, "message": "drm/amd/display: Enable DAC in DCE link encoder\n\nEnsure that the DAC output is enabled at the correct time by\nmoving it to the DCE link encoder similarly to how digital\noutputs are enabled.\n\nThis also removes the call to DAC1EncoderControl from the DCE\nHWSS, which always felt like it was a hacky solution.\n\nFixes: 0fbe321a93ce (\"drm/amd/display: Implement DCE analog link encoders (v2)\")\nSigned-off-by: Timur Kristóf \u003ctimur.kristof@gmail.com\u003e\nTested-by: Mauro Rossi \u003cissor.oruam@gmail.com\u003e\nReviewed-by: Alex Hung \u003calex.hung@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" } ], "next": "cbced93894d145239c83881d7fd953b7392c23a8" }