)]}' { "log": [ { "commit": "677dd2f8bcf765f6b6fdc5b07ec4df706b670f85", "tree": "b9d6056af1e6097d403875020848f10fa79e8728", "parents": [ "600c2f9512b5c8b44be541d4aa8edade696591df" ], "author": { "name": "Ming Wang", "email": "wangming01@loongson.cn", "time": "Wed Mar 04 09:19:14 2026 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Apr 27 16:58:36 2026 +0100" }, "message": "LoongArch: Increase MAX_MEMORY_RANGES to 1024\n\nOn some LoongArch systems, especially multi-socket machines like 3D6000,\nthe current limit of MAX_MEMORY_RANGES (64) is too small to accommodate\nall memory range entries in /proc/iomem.\n\nWhen the limit is reached, get_memory_ranges() stops parsing, which causes\nkexec to fail with \"locate_hole failed\" because it cannot correctly identify\nthe crash kernel reserved memory or necessary RAM segments.\n\nIncrease MAX_MEMORY_RANGES to 1024 to support systems with complex\nmemory layouts.\n\nSigned-off-by: Ming Wang \u003cwangming01@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "600c2f9512b5c8b44be541d4aa8edade696591df", "tree": "13a78e254014c947cdeabc83cb3f1e20a0458e06", "parents": [ "e18a71faae081d075c8d3ec559aa68a59295c739" ], "author": { "name": "Sourabh Jain", "email": "sourabhjain@linux.ibm.com", "time": "Tue Feb 10 10:35:36 2026 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Feb 17 16:07:39 2026 +0000" }, "message": "ppc64: Reserve FDT memory for full elfcorehdr memory size\n\nWith --hotplug support, extra memory is added to elfcorehdr kexec\nsegment to accommodate additional resources during memory hotplug\nevents. However, the FDT is not updated with the same size, which\ncan lead to elfcorehdr corruption in the kdump kernel.\n\nFix this by updating the elfcorehdr size in the FDT memory\nreservation entry to use its actual memsz instead of the current\nbuffer size.\n\nFixes: b8408c95e7c0 (\"powerpc/kexec_load: add hotplug support\")\nCc: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nCc: Simon Horman \u003chorms@kernel.org\u003e\nSigned-off-by: Sourabh Jain \u003csourabhjain@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "e18a71faae081d075c8d3ec559aa68a59295c739", "tree": "cf886d7a798c5194dbfb68d0162e76f3abc9181b", "parents": [ "1633ed5ee80ef9641e5544d7ae69be7459ff0465" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Fri Dec 05 10:05:36 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 12 10:02:53 2025 +0000" }, "message": "UKI: Fix the size of real payload\n\nAccording to the PE file specification, each section\u0027s SizeOfRawData\nmust be a multiple of FileAlignment (usually 512). So when ukify builds\na UKI image, it pads the kernel image, initrd, etc. with zeros aligned\nto 512 bytes. The actual payload size is recorded in VirtualSize.\n\nSince the checksum includes the trailing zeros, this causes issues when\nloading a signed x86 bzImage, which is stored in the UKI\u0027s .linux\nsection.\n\nCredit goes to Philipp, who analysed and pointed out this issue to me.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nCc: Philipp Rudo \u003cprudo@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "1633ed5ee80ef9641e5544d7ae69be7459ff0465", "tree": "d3d57d66abcc8c93c10134357f138441409a9155", "parents": [ "393c449aec3d0af62e813e9f80c4a6f58a521378" ], "author": { "name": "Shivang Upadhyay", "email": "shivangu@linux.ibm.com", "time": "Tue Dec 02 15:52:36 2025 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Dec 03 17:25:18 2025 +0000" }, "message": "ppc64: Fix minor build warnings\n\nppc64 target compiles warning free now.\n\nCc: Sourabh Jain \u003csourabhjain@linux.ibm.com\u003e\nSigned-off-by: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "393c449aec3d0af62e813e9f80c4a6f58a521378", "tree": "f89fd0a8f37fa475a174ac1ef34b511d94948b36", "parents": [ "71d6fd99af7e0dfaf13c67d96d31c0fcb0fa44a2" ], "author": { "name": "Pnina Feder", "email": "pnina.feder@mobileye.com", "time": "Thu Nov 06 14:03:44 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:49:59 2025 +0000" }, "message": "util_lib: Add direct map fallback in vaddr_to_offset()\n\nThe vmcore-dmesg tool could fail with the message:\n\"No program header covering vaddr 0x%llx found kexec bug?\"\n\nThis occurred when a virtual address belonged to the kernel’s direct\nmapping region, which may not be covered by any PT_LOAD segment in\nthe vmcore ELF headers.\n\nAdd a direct-map fallback in vaddr_to_offset() that converts such\nvirtual addresses using the known page and physical offsets. This\nallows resolving these addresses correctly.\n\nTested on Linux 6.16 (RISC-V)\n\nSigned-off-by: Pnina Feder \u003cpnina.feder@mobileye.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "71d6fd99af7e0dfaf13c67d96d31c0fcb0fa44a2", "tree": "157081c64578dd412d3d45482bdcd4968e7673ba", "parents": [ "4dc0397796757745934b0715c0c33260053ed738" ], "author": { "name": "Qiang Ma", "email": "maqianga@uniontech.com", "time": "Tue Nov 04 10:59:59 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:45:39 2025 +0000" }, "message": "kexec: add kexec flag to support debug printing\n\nThis add KEXEC_DEBUG to kexec_flags so that it can be passed\nto kernel when \u0027-d\u0027 is added with kexec_load interface. With that\nflag enabled, kernel can enable the debugging message printing.\n\nThis patch requires support from the kexec_load debugging message\nof the Linux kernel[1].\n\n[1]: https://lore.kernel.org/kexec/20251103063440.1681657-1-maqianga@uniontech.com/\n\nSigned-off-by: Qiang Ma \u003cmaqianga@uniontech.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "4dc0397796757745934b0715c0c33260053ed738", "tree": "ae4c0e673f004d6ef18d530b1ca9b72567489bf6", "parents": [ "bf4aa2a1f365d79181392c2ed4d2e8de21b0c399" ], "author": { "name": "Shivang Upadhyay", "email": "shivangu@linux.ibm.com", "time": "Wed Oct 22 19:16:09 2025 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:42:19 2025 +0000" }, "message": "ppc64: life the dtb and initrd restriction\n\nThe dtb and initrd restriction is now applied only when performing a\nkdump.\n\nSigned-off-by: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "bf4aa2a1f365d79181392c2ed4d2e8de21b0c399", "tree": "e9694fcecfefa5241d4a9b58b2cecd0c8394ae9d", "parents": [ "39631c8fd64f34db6437013597b8f3256dd90d3b" ], "author": { "name": "Shivang Upadhyay", "email": "shivangu@linux.ibm.com", "time": "Wed Oct 22 19:16:08 2025 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:41:26 2025 +0000" }, "message": "Add arch_do_unload hook for arch-specific cleanup\n\nProvide a new `arch_do_unload` hook that allows architectures to perform\ncleanup actions during kexec unload. On ppc64, this hook resets the boot\nCPU.\n\nSigned-off-by: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "39631c8fd64f34db6437013597b8f3256dd90d3b", "tree": "45e87aaa81ebb96f526f01049073dd5c6d28796c", "parents": [ "32f664bfa479f4a0ce5b00331a8d04ad069d2492" ], "author": { "name": "Shivang Upadhyay", "email": "shivangu@linux.ibm.com", "time": "Wed Oct 22 19:16:07 2025 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:40:52 2025 +0000" }, "message": "ppc64: handle reboot CPU in case of user provided DTB\n\nWhen restricting available CPUs for the next kexec\u0027ed kernel, ensure the\nreboot CPU is part of that available CPU\u0027s in new DTB. ppc64 requires\nthe booting CPU to be present in the new kernel’s DTB; otherwise, the\nsystem will panic during bring-up.\n\nExample Scenario\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nCurrent Kernel\n--------------\nPhysical CPUs: [1] [2] [3] [4] [5] [6] [7] [8]\nAvailable set: ^ ^ ^ ^ ^ ^ ^ ^\n | | | | | | | |\n +---+---+---+---+---+---+---+\n All CPUs available\n\nNext Kernel (limited CPU set)\n-----------------------------\nPhysical CPUs: [4] [5] [6] [7]\nLogical IDs: [0] [1] [2] [3]\n\nMapping:\n Logical 0 → Physical 4\n Logical 1 → Physical 5\n Logical 2 → Physical 6\n Logical 3 → Physical 7\n\nIf next kernel should only boot on CPUs 6 and 7\n------------------------------------------------\nSelected CPUs:\n Physical CPUs: [6] [7]\n Logical IDs: [2] [3]\n\nRequired reboot CPU \u003d Logical ID [3]\n → Physical CPU [7]\n\nSigned-off-by: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "32f664bfa479f4a0ce5b00331a8d04ad069d2492", "tree": "e4efeb6a6f810c4618f12820c0ff8ff7cb356ab5", "parents": [ "86c3d1f7b646832e962af371b23af18a4f3c2105" ], "author": { "name": "Shivang Upadhyay", "email": "shivangu@linux.ibm.com", "time": "Wed Oct 22 19:16:06 2025 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:40:00 2025 +0000" }, "message": "ppc64: ensure /memreserve/ sections exist in user-provided FDT\n\nUser-provided FDTs may lack `/memreserve/` sections, leading to hangs\nduring kexec. To fix this, allocate a new FDT buffer and use libfdt to\nadd reserve entries for the initrd, RTAS/OPAL regions (via\n/proc/device-tree), and the FDT itself.\n\nSigned-off-by: Shivang Upadhyay \u003cshivangu@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "86c3d1f7b646832e962af371b23af18a4f3c2105", "tree": "e4fb802f6d168da7c56187534c4856888c1afacf", "parents": [ "eb8609a293631e9f62269a4f1d84f47fef446f07" ], "author": { "name": "Mason Rocha", "email": "mrocha@turretllc.us", "time": "Tue Oct 21 21:07:04 2025 -0500" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:37:40 2025 +0000" }, "message": "kexec/ifdown.c: Hide error if sockets are disabled\n\nPrevents the message \"Function not implemented\" from being logged when\na system with networking support disabled, as there couldn\u0027t possibly be\nany interfaces to bring down to the point where we need to make sure the\nuser knows that the interfaces were not brought down.\n\nSigned-off-by: Mason Rocha \u003cmrocha@turretllc.us\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "eb8609a293631e9f62269a4f1d84f47fef446f07", "tree": "9748e02b43718440a9e39c49d17642925c5bff56", "parents": [ "8ac3a7e08957fcb3fa59c9fc4e9c283b9a072ce6" ], "author": { "name": "Mason Rocha", "email": "mrocha@turretllc.us", "time": "Tue Oct 21 21:07:03 2025 -0500" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:37:40 2025 +0000" }, "message": "kexec/ifdown.c: Use AF_NETLINK instead of AF_INET\n\nOn embedded systems, there is no guarantee that INET support has been\nbuilt in. However, as long as network support is built into the kernel,\nwe can use the netlink protocol to create a socket to still then use for\nthe ioctl calls since they don\u0027t require any certain family. This will\nhelp prevent EAFNOSUPPORT from being returned and printed to the console\non these systems.\n\nSigned-off-by: Mason Rocha \u003cmrocha@turretllc.us\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "8ac3a7e08957fcb3fa59c9fc4e9c283b9a072ce6", "tree": "aa45e5a58f117421129dea1c2c4029d5c321bad0", "parents": [ "b89799f7f20d4b2476e99e95e6795450f2f1c111" ], "author": { "name": "John Paul Adrian Glaubitz", "email": "glaubitz@physik.fu-berlin.de", "time": "Wed Oct 22 13:44:13 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:27:49 2025 +0000" }, "message": "kexec-tools: powerpc: Fix pointer declarations in read_memory_region_limits()\n\nFixes the following two build errors on 32-bit PowerPC:\n\nkexec/arch/ppc/kexec-ppc.c: In function \u0027read_memory_region_limits\u0027:\nkexec/arch/ppc/kexec-ppc.c:106:11: error: assignment to \u0027long long unsigned int *\u0027 from incompatible pointer type \u0027long unsigned int *\u0027 [-Wincompatible-pointer-types]\n 106 | p \u003d (unsigned long*)buf;\n | ^\n\nkexec/arch/ppc/kexec-ppc.c: In function \u0027read_memory_region_limits\u0027:\nkexec/arch/ppc/kexec-ppc.c:112:19: error: assignment to \u0027long unsigned int *\u0027 from incompatible pointer type \u0027long long unsigned int *\u0027 [-Wincompatible-pointer-types]\n 112 | p \u003d (unsigned long long *)p + 1;\n | ^\n\nSigned-off-by: John Paul Adrian Glaubitz \u003cglaubitz@physik.fu-berlin.de\u003e\nReviewed-by: Khalid Aziz \u003ckhalid@gonehiking.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b89799f7f20d4b2476e99e95e6795450f2f1c111", "tree": "d6c78ed1e6d4b20e5a7a403a787a123a18d60592", "parents": [ "36294ec99439a9f5fcefdffcb78d669c6a4005b7" ], "author": { "name": "John Paul Adrian Glaubitz", "email": "glaubitz@physik.fu-berlin.de", "time": "Wed Oct 22 13:44:12 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:27:44 2025 +0000" }, "message": "kexec-tools: powerpc: Fix function signature of comparefunc()\n\nFixes the following build error on 32-bit PowerPC:\n\nkexec/arch/ppc/fs2dt.c: In function \u0027putnode\u0027:\nkexec/arch/ppc/fs2dt.c:338:51: error: passing argument 4 of \u0027scandir\u0027 from incompatible pointer type [-Wincompatible-pointer-types]\n 338 | numlist \u003d scandir(pathname, \u0026namelist, 0, comparefunc);\n | ^~~~~~~~~~~\n | |\n | int (*)(const void *, const void *)\n\nSigned-off-by: John Paul Adrian Glaubitz \u003cglaubitz@physik.fu-berlin.de\u003e\nReviewed-by: Khalid Aziz \u003ckhalid@gonehiking.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "36294ec99439a9f5fcefdffcb78d669c6a4005b7", "tree": "e99ca1f59dbc0aa5be68e1279d48dfb2ef9a26c4", "parents": [ "15d78e5799eea7ec5ea9c5897ae95aaa0ce8970c" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:22:37 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Nov 11 13:23:12 2025 +0000" }, "message": "kexec-tools 2.0.32.git\n\nAdd .git to version so it doesn\u0027t look like a release.\nThis is just so when people build code from git it can\nbe identified as such from the version string.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "15d78e5799eea7ec5ea9c5897ae95aaa0ce8970c", "tree": "efb4027c5cff03365661e465d427e97a58c0962f", "parents": [ "63a6986d7aaedebd972044e9e880ba9ea5ee7e50" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Oct 16 09:17:02 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Oct 16 09:17:02 2025 +0100" }, "message": "kexec-tools 2.0.32\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "63a6986d7aaedebd972044e9e880ba9ea5ee7e50", "tree": "d9e272e60fa7e4c34bd1036b8671075303bdf59c", "parents": [ "591ebf60649de53dd7a9c916ef4e831626ab1810" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Oct 09 10:53:09 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Oct 09 10:53:09 2025 +0100" }, "message": "kexec-tools 2.0.32-rc1\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "591ebf60649de53dd7a9c916ef4e831626ab1810", "tree": "2b8c2d2d89ebf8f9f677b847cfb823f1b1a83e08", "parents": [ "2842172e14056ae1e054a3541bff153b6af94f34" ], "author": { "name": "Khalid Aziz", "email": "khalid@gonehiking.org", "time": "Mon Sep 29 13:11:31 2025 -0600" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 30 09:29:54 2025 +0100" }, "message": "COPYING: Update with current text for GPL v2\n\nText for GPL v2 has been updated to remove old mailing address. This\nupdates the license text in COPYING file with the current license text\nfrom https://www.gnu.org/licenses/old-licenses/gpl-2.0.txt\n\nSigned-off-by: Khalid Aziz \u003ckhalid@gonehiking.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "2842172e14056ae1e054a3541bff153b6af94f34", "tree": "e6b149ec7131230cf9051d097f62c9defcd07ad7", "parents": [ "c343815e7e0ae37f28a1010d625e7f5acf424e8e" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:14 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:46:08 2025 +0100" }, "message": "kexec/arch/arm64/crashdump-arm64.c: remove extra whitespace\n\nRemove extra whitespace.\n\nFixes: aecc554e7ba8 (\"Correct string specifiers\")\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "c343815e7e0ae37f28a1010d625e7f5acf424e8e", "tree": "3ee5bf4c31dcf999c8db794732b07a9322baa7b3", "parents": [ "b7b2a13547a0e06c2de01822ed5655f6e8a551ee" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:13 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:46:05 2025 +0100" }, "message": "kexec/arch/i386/kexec-x86-common.c: remove duplicate \u003cstdio.h\u003e\n\nRemove duplicate \u003cstdio.h\u003e.\n\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b7b2a13547a0e06c2de01822ed5655f6e8a551ee", "tree": "f0a2e0463778a1bd7c156156bf3b7c4e65d9a022", "parents": [ "9c12e71d712a346b52db17ddc16e81a8629d2f79" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:12 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:46:01 2025 +0100" }, "message": "util_lib/elf_info.c: fix typo: prink -\u003e printk\n\nFix typo: prink -\u003e printk.\n\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "9c12e71d712a346b52db17ddc16e81a8629d2f79", "tree": "aca15be79e40535b69f58f699210fed6a79581a7", "parents": [ "c837f1971fab244e30165f4469a2dd75c0393bb7" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:11 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:45:57 2025 +0100" }, "message": "kexec/arch/i386/x86-linux-setup.c: replace %d with %u\n\n...because e820[i].type is unsigned.\n\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "c837f1971fab244e30165f4469a2dd75c0393bb7", "tree": "5722224f3c6994b6253265d1f5a158e017e2af5f", "parents": [ "288e352ed7c8db5caf8750c6d2ca3ba01b011568" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:10 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:45:53 2025 +0100" }, "message": "kexec/kexec-elf-exec.c: Replace %lux with %lx\n\nReplace %lux, which prints a decimal number followed by literal \u0027x\u0027,\nwith %lx, which prints the hexadecimal one. The latter is more useful\nfor the (address) ranges.\n\nFixes: d6bc88c0696 (\"elf: Support ELF loading with relocation\")\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "288e352ed7c8db5caf8750c6d2ca3ba01b011568", "tree": "03dd13267f7fefd60e268cdd7efc32baafcb4aa0", "parents": [ "0b5aecb85a21148b9bf79d8708b9c48809160c7f" ], "author": { "name": "Askar Safin", "email": "safinaskar@zohomail.com", "time": "Tue Aug 19 21:30:09 2025 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Sep 29 10:45:33 2025 +0100" }, "message": "Fix printf string specifiers, otherwise kexec doesn\u0027t work on my laptop\n\nTL;DR: this patch fixes regression.\nThis patch should be backported to all distributions, which packaged v2.0.31, otherwise\nkexec doesn\u0027t work at all at my laptop with pretty common setup with v2.0.31.\nv2.0.31 is broken without this patch.\n\nRecently I found a bug: kexec doesn\u0027t work at my laptop.\nHere is my setup:\n* Laptop Dell Precision 7780\n* CPU Intel Raptor Lake-S\n* i915 GPU\n* x86_64\n* UEFI\n* Linux v6.16-rc5\n* linux-firmware 20250708\n* kexec-tools v2.0.31\n* I do kexec from this kernel to the same kernel\n* Kernel is UEFI PE image\n* Secure boot is disabled\n* Kernel config is minimized Debian config\n* I use command \u0027kexec --debug --kexec-syscall -l /disk/vmlinuz --initrd\u003d/disk/initramfs.cpio.gz --append\u003d\"...\"\u0027, then I execute \u0027kexec --debug -e\u0027\n\nWhen I execute \"kexec --debug -e\", I see black screen instead of proper boot of kexec\u0027d kernel.\n\nOkay, why the bug happens? I suspect this is because of \"%lux\"\nstring specifiers, which are totally wrong. The author meant \"print (and scan) in hexademical\"\nhere, but this specifier prints (and scans) number in decimal, followed by literal \"x\". Oops.\nAnd this seems to break kexec.\n\nChanging this to \"%lx\" or \"%llx\" will not work, because uint64_t is\ndefined differently on different architectures. So I use PRIx64 to fix this.\n\nI tested on my laptop that this patch actually fixes the bug.\n\nFixes: aecc554e7b (\"Correct string specifiers\")\nReviewed-by: Andy Shevchenko \u003candy@kernel.org\u003e\nSigned-off-by: Askar Safin \u003csafinaskar@zohomail.com\u003e\nTested-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "0b5aecb85a21148b9bf79d8708b9c48809160c7f", "tree": "792f800e56481be4dc6dc218042862a96f521a27", "parents": [ "efb97b509b1a43fce8fe150135cf8167aa0db440" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 15:20:16 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "workflow: Add loongarch64 build\n\nAdd loongarch64 to build matrix.\n\nA gcc-14 package is installed as there is no generic gcc package\nfor loongarch in Ubuntu 24.04. Hopefully this will be resolved in\nUbuntu 26.04.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "efb97b509b1a43fce8fe150135cf8167aa0db440", "tree": "99de5d558557731305b7b37b5e3966b2bb795c81", "parents": [ "19d1b275c64987f39c5090a03626014876846049" ], "author": { "name": "Khem Raj", "email": "raj.khem@gmail.com", "time": "Thu Sep 04 20:22:52 2025 -0700" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "Make the segment base match pinned section address\n\nET_EXEC uses image base of 0x400000, but the build forces\nsection VMAs like .text \u003d 0x10000. LLD now errors when any\nsection address is below the image base unless you explicitly\nset the base. (Older LLD tolerated it.)\n\nTo fix this\n\nuse a tiny linker script to keep .text at 0x10000 and avoid .note overlap\n\nWith hardcoded -Ttext,0x10000 and newer toolchains emitting\n.note.gnu.property before .text, the link can fail with an overlap.\nProvide a minimal linker script that:\n - sets the image base to 0x10000,\n - places .text first at exactly 0x10000,\n - moves .note.gnu.property after .text,\n - optionally fixes .bss at 0x12000.\n\nWorks with both ld.bfd and LLD 21+.\n\nFixes\n| x86_64-yoe-linux-ld.lld: error: section \u0027.text\u0027 address (0x10000) is smaller than image base (0x400000); specify --image-base\n| x86_64-yoe-linux-ld.lld: error: section \u0027.bss\u0027 address (0x12000) is smaller than image base (0x400000); specify --image-base\n\nSigned-off-by: Khem Raj \u003craj.khem@gmail.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "19d1b275c64987f39c5090a03626014876846049", "tree": "ea28c90f9753faecea6d2d1ac1fa32b77502fe70", "parents": [ "f1d66280a48b7575d9e22538f4587237e14cf432" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:55 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Remove \u0027kexec_file\u0027 cmdline parameters when using --reuse-cmdline option\n\nOn the LoongArch architecture, when loading the kernel via kexec_file,\nthe \u0027kexec_file\u0027prefix is automatically appended to the command line.\nConsequently, if the \u0027--reuse-cmdline\u0027 option is used to inherit the\ncurrent command line, this prefix must be detected and removed.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "f1d66280a48b7575d9e22538f4587237e14cf432", "tree": "9e70c85a25b825226f37f54bfab00a002034bf78", "parents": [ "b7c35c42c3db7e628378540890ed0b975629b0e1" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:54 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Add kexec_file_load syscall\n\nCreate prepare_kexec_file_options() function to prepare the options\nto kexec_file_load syscall, and it would be used in pei_loongarch_load().\n\nCurrently, pez(vmlinuz.efi), pei(vmlinux.efi) and elf(vmlinux) format\nimages are supported.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b7c35c42c3db7e628378540890ed0b975629b0e1", "tree": "493e1f076a997cc492f58ae0aeae086c7ccd5bd3", "parents": [ "13e701e72ecf7bc7679808cb01d280559d1d2e86" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:53 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch/pez: Fix kernel_fd handling when kexec_file is supported\n\nAfter the current pez kernel image is decompressed and loaded, this kernel_fd\nwill be released. When kexec_file is added subsequently, the kernel detects\nthat the content in kernel_fd is empty, causing the kernel loading to fail.\nTherefore, kernel_fd is rewritten for processing.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "13e701e72ecf7bc7679808cb01d280559d1d2e86", "tree": "effa7d66eb0f0572d5d26e74f008738ef9fd6b63", "parents": [ "919005b26c8f75e052c0a8a87f20ccb9935ff131" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:52 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Add pe_hdr-\u003emachine check for pei format images\n\nAdd the pe_hdr-\u003emachine check to ensure that the kernel image is of the 64-bit\nLoongArch architecture.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "919005b26c8f75e052c0a8a87f20ccb9935ff131", "tree": "c8135c33dd30a2364f8e5558fb379abbbed612b3", "parents": [ "2d3f8185f5d7353fe8b7fb988c5b2b2447b6722c" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:51 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Fix the use of loongarch_image_header in ELF format\n\nloongarch_image_header should not be used in the ELF format. When the CONFIG_EFI_STUB\nconfiguration is turned off, the ELF kernel image will not add loongarch_image_heade\nprocessing, as can be seen from the kernel code arch/loongarch/kernel/head.S.\n\nThus, use the elf header content to fix this issue.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "2d3f8185f5d7353fe8b7fb988c5b2b2447b6722c", "tree": "d7a3487cc28a1449bb6eed651822b7a2f47920f5", "parents": [ "024073e74dc72a1c80c444b4ee079e087bef9a92" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:50 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Change initrd allocation to top-down\n\nWhen loading a secondary kernel with KASLR enabled, the randomized relocation\nof the kernel image may overlap the initrd memory region, causing data corruption\nand boot failure. This occurs because KASLR dynamically shifts the kernel\u0027s load\naddress, which could accidentally target the initrd\u0027s pre-allocated space.\n\nModify the initrd allocation strategy from bottom-up to top-down within the\nreserved memory region. By placing the initrd at higher addresses first, we\nminimize the risk of the relocated kernel overwriting it.\n\nReported-by: Chenghao Duan \u003cduanchenghao@kylinos.cn\u003e\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "024073e74dc72a1c80c444b4ee079e087bef9a92", "tree": "7aff7ede3353f80052a801191d15ccdf4dc60cd9", "parents": [ "d9c82c188646d04bec29608a0db3cfc974aa48a7" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:49 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Enforce relocatable kernel check for crash dump\n\nWithout enabling the RELOCATABLE configuration, LoongArch is a non-PIE kernel\nand cannot be loaded to run at any appropriate address. So the CRASH_DUMP\nfeature depends on RELOCATABLE.\n\n$ cat arch/loongarch/Kconfig\nconfig ARCH_SELECTS_CRASH_DUMP\n def_bool y\n depends on CRASH_DUMP\n select RELOCATABLE\n\nThe relocatable kernel is determined by checking if there is a la_abs section.\nCurrently, only the elf format has been checked, while pei/pez is in the FIXME\nstate.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "d9c82c188646d04bec29608a0db3cfc974aa48a7", "tree": "05791cf8a1eeaf8319e29ca306527f1ce0d7cd94", "parents": [ "e2bfd6df147c87e04cfec378ae20f243cb09456e" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:48 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Terminate the cmdline string using \u0027\\0\u0027\n\nTerminate the cmdline string using \u0027\\0\u0027 instead of 0, improve code readability.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "e2bfd6df147c87e04cfec378ae20f243cb09456e", "tree": "b85aba588fc71dec29bedda683af2ec2b2ca928d", "parents": [ "39ba2bd288af1100b837d299174b71dc9f7201fc" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:47 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "LoongArch: Fix comments\n\nloongarch_load_other_segments() mainly loads initrd and cmdline. Fix this\nannotation.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "39ba2bd288af1100b837d299174b71dc9f7201fc", "tree": "b5ea96cdad09b81bf5589e640e5109a3c5ba1187", "parents": [ "b8f6c2a98d2c3cd85bf8b4a530b9b29f40ca0a9a" ], "author": { "name": "Youling Tang", "email": "tangyouling@kylinos.cn", "time": "Tue Sep 16 09:46:46 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 17:46:00 2025 +0100" }, "message": "kexec: Handle removal of multiple \u0027crashkernel\u0027 parameters\n\nWhen the kernel command line contains multiple \u0027crashkernel\u0027 parameters\n(e.g., `crashkernel\u003d1G,high crashkernel\u003d256M,low`), the original\n`remove_parameter()` function only removed the first instance. This\nleft residual parameters that caused conflicts during kexec operations.\n\nSigned-off-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b8f6c2a98d2c3cd85bf8b4a530b9b29f40ca0a9a", "tree": "853f0f27ebf4788c6a218d83195f5b34f58e5cad", "parents": [ "daa29443819d3045338792b5ba950ed90e79d7a5" ], "author": { "name": "Chenghao Duan", "email": "duanchenghao@kylinos.cn", "time": "Mon Sep 22 10:10:57 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Sep 23 15:08:15 2025 +0100" }, "message": "LoongArch: Change COMMAND_LINE_SIZE to 4096\n\nThe default COMMAND_LINE_SIZE inherited from asm-generic is 512, which\nis too small for modern use cases. For example, when the parameter path\nin the cmdline of the kdump first kernel is too long, the limit is\neasily exceeded when loading the second kernel.\n\nIncrease the command line size to 4096 bytes, In LoongArch, this is\nconsistent with the upstream configuration of the Linux kernel as of\ncommit f7794a4d92ad (\"LoongArch: Increase COMMAND_LINE_SIZE up to 4096\").\n\nThis patch is necessary because the Linux kernel has been changed to\n4096. When the kernel cmdline size exceeds 512, the --reuse-cmdline\noption of the kexec command will inevitably result in an error.\n\nThe COMMAND_LINE_SIZE in kexec/arch/loongarch/crashdump-loongarch.h\nis unused, so it is removed here.\n\nSuggested-by: Youling Tang \u003ctangyouling@kylinos.cn\u003e\nSigned-off-by: Chenghao Duan \u003cduanchenghao@kylinos.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "daa29443819d3045338792b5ba950ed90e79d7a5", "tree": "da74d67eaa05cc9f41add58ba8c563ad4299c61e", "parents": [ "8322826fa7b04a5c0f023eda78d69dd1413a1412" ], "author": { "name": "Nicholas Sielicki", "email": "opensource@nslick.com", "time": "Sun Jun 22 06:12:53 2025 -0700" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Jun 23 17:09:07 2025 +0100" }, "message": "kexec/zstd: fix static zstd build\n\n60ce9bdf introduced a local reimplementation of ZSTD_isFrame to this\ntree, as the definition in libzstd is only available for static linking.\nWhen a static build of libzstd is in-fact available to kexec-tools, this\nwas previously causing multiple-definition errors at link time.\n\nprefer the ZSTD_isFrame from libzstd, if it\u0027s available, by adding\n__attribute__((weak)) on the local definition.\n\nSigned-off-by: Nicholas Sielicki \u003copensource@nslick.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "8322826fa7b04a5c0f023eda78d69dd1413a1412", "tree": "9346503f77dbd17963771c81489adb7a41c4073f", "parents": [ "b257f311f5f0d32969f58098b03458b5bd728419" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Tue May 13 18:32:40 2025 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed May 14 11:40:56 2025 +0100" }, "message": "kexec/zboot: Add boundary check on PE header offset\n\nIf a buildroot enviorment does not provide libzstd and kexec_file_load a\nzstd compressed kernel, it will suffer from coredump.\n\nThe following is observed from the coredump\n\nCore was generated by `/sbin/kexec -s -p --command-line\u003dBOOT_IMAGE\u003d\\(hd0,gpt2\\)/vmlinuz-6.15.0-0.rc5.250509g9c69f8884904.47.fc43.aarch64\\ ro\\ rootflags\u003dsubvol\u003droot\\ irqpoll\\ nr_cpus\u003d1\\ reset_devices\\ cgroup_disable\u003dmemory\\ udev.children-max\u003d2\\ panic\u003d10\\ swiotlb\u003dnoforce\\ novmcoredd\\ cma\u003d0\\ hugetlb_cma\u003d0\\ sbsa_gwdt.pretimeout\u003d0 --initrd\u003d/boot/initramfs-6.15.0-0.rc5.250509g9c69f8884904.47.fc43.aarch64kdump.img /boot/vmlinuz-6.15.0-0.rc5.250509g9c69f8884904.47.fc43.aarch64\u0027.\nProgram terminated with signal SIGSEGV, Segmentation fault.\n143\t\tif (!!memcmp(buf, \"PE\\0\\0\", 4)) {\n(gdb)\n\nInstead of coredump, it is better to exit elegantly by adding checkout\non the PE header offset.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b257f311f5f0d32969f58098b03458b5bd728419", "tree": "cc53f259078c9ef67140be4148af51cec7061c26", "parents": [ "363087cd156f870dfd07fef04077696daa77acaa" ], "author": { "name": "Song Shuai", "email": "songshuaishuai@tinylab.org", "time": "Tue Apr 22 18:23:00 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Apr 23 16:30:59 2025 +0100" }, "message": "RISC-V: Support loading Image binary file\n\nAdd image-riscv file_type to probe/load Image file type,\n\nAs for kexec_load, find the pbase aligned text_offset from image header\nand prepare segments for this syscall.\n\nfor kexec_file_load, setup the related options and let kernel part to\ndeal with the Image.\n\nSigned-off-by: Song Shuai \u003csongshuaishuai@tinylab.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "363087cd156f870dfd07fef04077696daa77acaa", "tree": "60a8e0f16a117cf55e00b9c1a91f687aeae896cf", "parents": [ "027e6e237427b00b986d9c51fe65944b8fdcc85b" ], "author": { "name": "Song Shuai", "email": "songshuaishuai@tinylab.org", "time": "Tue Apr 22 18:22:59 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Apr 23 16:30:24 2025 +0100" }, "message": "RISC-V: Separate elf_riscv_find_pbase out\n\nThe is the preparative patch for RISC-V kexec Image file support.\n\nSeparate the elf_riscv_find_pbase() function out to\nallow kernel_load syscall load Image binary file.\n\nSigned-off-by: Song Shuai \u003csongshuaishuai@tinylab.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "027e6e237427b00b986d9c51fe65944b8fdcc85b", "tree": "e0bb31ce28fc653276ab0444dd5cc988cf5cfe97", "parents": [ "b3fd54b022fe5ba09ebff184d4d22466672946cf" ], "author": { "name": "Li Zhengyu", "email": "lizhengyu3@huawei.com", "time": "Tue Apr 22 18:22:58 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Apr 23 16:30:12 2025 +0100" }, "message": "RISC-V: Enable kexec_file_load syscall\n\nCreate prepare_kexec_file_options() function to prepare the options\nto kexec_file_load syscall, and it would be used in elf_riscv_load()\nor the future image_riscv_load().\n\nThe patch comes from the RISC-V Linux kernel_file_load support[1],\nSo its author should be Li Zhengyu.\n\n[1]: https://lore.kernel.org/all/20220408100914.150110-1-lizhengyu3@huawei.com/\n\nSigned-off-by: Song Shuai \u003csongshuaishuai@tinylab.org\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b3fd54b022fe5ba09ebff184d4d22466672946cf", "tree": "d288ec8a7eee56ce68b9c8a20ee3a7c2292179cf", "parents": [ "1bda78d605044b060404ad54904e5a4f97a87bdb" ], "author": { "name": "Nick Kossifidis", "email": "mick@ics.forth.gr", "time": "Tue Apr 22 18:22:57 2025 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Apr 23 16:29:49 2025 +0100" }, "message": "RISC-V: Add support for riscv kexec/kdump on kexec-tools\n\nThis patch adds support for loading the ELF kernel image. It parses\nthe current/provided device tree to determine the system\u0027s memory\nlayout, and /proc/iomem for the various kernel segments.\n\nTested on Qemu\u0027s rv64 virt machine and SoC of T-Head RISC-V Xuantie\n910 CPU.\n\nNow, some history: The first stab at supporting kexec-tools on RISC-V\nwas done by Nick Kossifidis. The initial patch has since then had a\nnumber of improvements/fixes by other authors. Given, this is the\nfirst commit for RISC-V, carrying the fixes/changes commits in the\nupstream tree does not really add anything (bisectability).\n\nInstead all the fixes that were applied to Nick\u0027s first commit is\noutlined below:\n\nYixun Lan, and Xianting Tian:\n * Fixed a failure to fail to find free memory area for dtb load when\n using initrd image [1].\n * Fixed memory range size calculation in\n kexec/arch/riscv/crashdump-riscv.c:85\n\nSimon Horman:\n * workflow: Add riscv64\n\n Add riscv64 to matrix of build architectures.\n * RISC-V: distribute purgatory/riscv/Makefile\n\n Include purgatory/riscv/Makefile in distribution tarball.\n\n Local patch as it is planned to suggest this as a fix for the\n patch that introduced this problem. [2]\n\nSong Shuai:\n * RISC-V: Fix the undeclared ‘EM_RISCV’ build failure\n\n Use local `elf.h` instead of `linux/elf.h` to fix this build\n error:\n\n ```\n kexec/arch/riscv/crashdump-riscv.c:17:13: error: ‘EM_RISCV’ undeclared here (not in a function); did you mean ‘EM_CRIS’?\n .machine \u003d EM_RISCV,\n ^~~~~~~~\n EM_CRIS\n ```\n\n * RISC-V: Correct the usage of command line option\n\n RISC-V process OPT_CMDLINE with the \"command-line\" partten, but\n the riscv_opts_usage shows the \"cmdline\" option. So correct the\n usage\u0027s output.\n\n * RISC-V: Use linux,usable-memory-range for crash kernel\n\n Now we use \"memeory::linux,usable-memory\" to indicate the\n available memory for the crash kernel.\n\n While booting with UEFI, the crash kernel would use efi.memmap to\n re-populate memblock and then first kernel\u0027s memory would be\n corrputed. Consequently, the /proc/vmcore file failed to create in\n my local test.\n\n And according to \"chosen\" dtschema [3], the available memory for\n the crash kernel should be held via\n \"chosen::linux,usable-memory-range\" property which will re-cap\n memblock even after UEFI\u0027s re-population.\n\n * RISC-V: Get memory ranges from iomem\n\n When booting with UEFI, Linux marks the Runtime Code/Data memory\n as no-map and then exports it to \"Reserved\" iomem_resource.\n\n Kexc-tools uses dtb_get_memory_ranges() function to get memory\n ranges via parsing dtb, but it can\u0027t see the Reserved EFI Runtime\n memory. That would corrupt EFI Runtime memory and fail the kexeced\n kernel to deal EFI stuff.\n\n In my test, the kexeced kernel warned \"efi: System table signature\n incorrect!\" and then paniced at efi_call_rts() due to the null\n efi.runtime.\n\n So we should use /proc/iomem to get memory ranges.\n\nBjörn Töpel:\n * Massaged this commit message!\n * Fixed up the build, by adding missing RV stub.\n * RISC-V: Only cap the upper/end usable memory window\n\n When loading the initrd in the kexec_load flow, memory for the\n segments are searched from end to start. Only the max_usable\n should be capped, after a successful initrd addtion.\n\n Currently min/max usable is set to the same value, making it\n impossible from subsequent segment allocations to success.\n\n * RISC-V: Make get_memory_ranges() properly exclude \"Reserved\"\n regions\n\n The get_memory_ranges() did not exclude \"Reserved\" regions from\n \"System RAM\" regions. It simply added \"Reserved\" as IOMEM, and\n IOMEM is not considered when looking for holes to place kexec\n segments.\n\n Instead, do a two pass of the /proc/iomem file. First pass, adds\n all the \"System RAM\" memory, and the second pass removes all\n intersecting \"Reserved\" regions.\n\n[1] https://lore.kernel.org/linux-riscv/CALecT5gQWn0PRO4Q24b6qkrfVE5OxsCp65TuhWTb30ceK_OJ0A@mail.gmail.com/\n[2] https://lore.kernel.org/kexec/20221020031548.47587-1-xianting.tian@linux.alibaba.com/\n[3] https://github.com/devicetree-org/dt-schema/blob/main/dtschema/schemas/chosen.yaml\n\nTested-by: Yixun Lan \u003cyixun.lan@gmail.com\u003e\nCo-developed-by: Xianting Tian \u003cxianting.tian@linux.alibaba.com\u003e\nCo-developed-by: Yixun Lan \u003cyixun.lan@gmail.com\u003e\nSigned-off-by: Nick Kossifidis \u003cmick@ics.forth.gr\u003e\nSigned-off-by: Song Shuai \u003csongshuaishuai@tinylab.org\u003e\nSigned-off-by: Björn Töpel \u003cbjorn@rivosinc.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "1bda78d605044b060404ad54904e5a4f97a87bdb", "tree": "5d80c9951126a5c193661587b127d072cc527566", "parents": [ "14f48ea80e510e5af79ca6c46fa28e99a3695d1c" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 22 11:14:13 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Apr 23 16:28:26 2025 +0100" }, "message": "workflow: update to Ubuntu 24.04\n\nBump workflow to use latest LTS release.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "14f48ea80e510e5af79ca6c46fa28e99a3695d1c", "tree": "de18a5551ae4fefd499799b73b686c708097f095", "parents": [ "3baa1baebf917392849e7176888ca48d8730d2dc" ], "author": { "name": "Ross Lagerwall", "email": "ross.lagerwall@citrix.com", "time": "Thu Apr 17 09:31:40 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 22 10:24:04 2025 +0100" }, "message": "x86_64: Use the x86-64 level for purgatory\n\nThe compiler may be configured by default to use a newer\nmicroarchitecture level such as x86-64-v3. Set purgatory to compile\nusing the baseline x86-64 level since the environment isn\u0027t set up\ncorrectly to use SSE and AVX instructions. At the same time, be a bit\nparanoid and add some additional flags to ensure that the compiler\ndoesn\u0027t use any such instructions.\n\nSigned-off-by: Ross Lagerwall \u003cross.lagerwall@citrix.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "3baa1baebf917392849e7176888ca48d8730d2dc", "tree": "50acb165c81cd0fe305e463de836dba9a7a34f4f", "parents": [ "d0d0a5991526233484b5d3c3efd04d81c6a0fc73" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:25:05 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 22 10:23:38 2025 +0100" }, "message": "kexec-tools 2.0.31.git\n\nAdd .git to version so it doesn\u0027t look like a release.\nThis is just so when people build code from git it can\nbe identified as such from the version string.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "d0d0a5991526233484b5d3c3efd04d81c6a0fc73", "tree": "b32f50ce680efffd9297cc5fead4c8feb2da0414", "parents": [ "f80553597da191a8ccabe3f6762611c6d60f0f95" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 22 10:09:15 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 22 10:09:15 2025 +0100" }, "message": "kexec-tools 2.0.31\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "f80553597da191a8ccabe3f6762611c6d60f0f95", "tree": "52944b36a80fea168dd09ea9100530a58de52904", "parents": [ "a7fcd424c4c80dea5a2fd5ffa274ffeb8129c790" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Apr 14 11:35:26 2025 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Apr 14 11:35:26 2025 +0100" }, "message": "kexec-tools 2.0.31-rc1\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "a7fcd424c4c80dea5a2fd5ffa274ffeb8129c790", "tree": "7545ae7a9037f0085c345552894b1f1a5bd36927", "parents": [ "60ce9bdf0e5cf37ee3db3586ffc8b17f8cbb5aa3" ], "author": { "name": "Jeremy Linton", "email": "jeremy.linton@arm.com", "time": "Thu Dec 12 15:13:25 2024 -0600" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 13 12:37:48 2024 +0000" }, "message": "kexec: Enable zstd in kexec decompression paths\n\nThe kexec decompression routines support detecting zlib or lzma\ncompressed images. Lets add the ability to automatically decompress\nzstd ones as well. The zstd attempt is made after gzip because its\nfairly inexpensive relative lzma, and gzip remains a popular option.\n\nSigned-off-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nAcked-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "60ce9bdf0e5cf37ee3db3586ffc8b17f8cbb5aa3", "tree": "a4ac1bf8057174ab993ddf52e02a9cabecc2eee5", "parents": [ "aecc554e7ba8bd449dceaf3eeecacc09b9b77fc4" ], "author": { "name": "Jeremy Linton", "email": "jeremy.linton@arm.com", "time": "Thu Dec 12 15:13:24 2024 -0600" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 13 12:34:16 2024 +0000" }, "message": "zstd: Add zstd decompression logic\n\nzstd is becoming a popular zlib replacement because it both tends to\nget a better compression ratio as well performs considerably\nbetter. As such, zstd is now one of the options that can be used to\ncompress the Linux kernel.\n\nLets enable it by using a system provided shared library, creating the\nrequired boilerplate to match the existing zlib/lzma function\nsignatures, and creating build options to enable/disable it.\n\nSigned-off-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nAcked-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "aecc554e7ba8bd449dceaf3eeecacc09b9b77fc4", "tree": "f6634bb3ec0bcf23aa92c9f9c1dc73bc2b8ff52d", "parents": [ "6aecc32c6db895b1c0b1f522d82c8810ece65542" ], "author": { "name": "Jeremy Linton", "email": "jeremy.linton@arm.com", "time": "Tue Dec 10 20:58:46 2024 -0600" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 13 12:33:14 2024 +0000" }, "message": "Correct string specifiers\n\nRecent GCC\u0027s on fedora complain about the use of\n%lld for uint64_t because its unsigned. Convert a number\nof these and similar calls.\n\nSigned-off-by: Jeremy Linton \u003cjeremy.linton@arm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "6aecc32c6db895b1c0b1f522d82c8810ece65542", "tree": "0cac7323d86b0af4319be2866392a86aa9d1e0cd", "parents": [ "455f55c7c02931b22c818572ed28378c02591397" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Fri Dec 06 10:44:43 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 13 12:33:13 2024 +0000" }, "message": "pe-zboot: Truncate the trailing zero if Image is signed\n\n*** Issue ***\nIn the linux kernel drivers/firmware/efi/libstub/Makefile.zboot, the\noriginal Image is padded with zero, using the following instruction:\n\ttruncate -s $$(hexdump -s16 -n4 -e \u0027\"%u\"\u0027 $\u003c) $@\n\nHence pe-zboot.c decompress and gets Image plus trailing zeroes.\n\nThese trailing zeroes don\u0027t affect loading the original PE file. But\nthey do raise an issue during the signature verification. The root cause is\nthat the kernel function:\n\tstatic int pefile_digest_pe_contents(const void *pebuf, unsigned int pelen,\n\t\t\t\t\t struct pefile_context *ctx,\n\t\t\t\t\t struct shash_desc *desc)\ntreats [pebuf, pebuf+pelen] as valid payload, which includes the\ntrailing zeroes. But that is not the truth.\n\n*** Solution ***\nIn practice, the table of attribute certificates come at the end of a\nPE file. This patch utilizes that fact and truncates at the boundary of the\ncertificate table to get the original Image.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nTo: kexec@lists.infradead.org\n" }, { "commit": "455f55c7c02931b22c818572ed28378c02591397", "tree": "0ac1469f5088d3e9f202c5dd6e6f48ca552ab212", "parents": [ "0475ec5d32469a486713f9ed7023645bbbbb8769" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Fri Dec 06 10:44:42 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Dec 13 12:33:12 2024 +0000" }, "message": "PE: Extract get_pehdr_offset() for reuse\n\nThe get_pehdr_offset() function helps skip the DOS stub in a PE file to\nlocate the start of the PE header. Since it can be used in multiple C\nfiles, it is being moved to a header file.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "0475ec5d32469a486713f9ed7023645bbbbb8769", "tree": "fbf27cb5247cf6f054c627e7075fb0e0546ddc61", "parents": [ "c3a991a169e7f0338f627f1f461998f279d17759" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Tue Nov 05 11:57:07 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:26:52 2024 +0000" }, "message": "x86_64: Support UKI image format\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nAcked-by: Dave Young \u003cdyoung@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "c3a991a169e7f0338f627f1f461998f279d17759", "tree": "c8e529a38c4d70fbf8bf49fc6de4bb901a241e2d", "parents": [ "99d32ffdf1efb5e4d0b2c6f84f5186f5cf36b0bf" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Tue Nov 05 11:57:06 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:26:41 2024 +0000" }, "message": "kexec: Create a temporary file to hold .linux section in uki_probe()\n\n**Problem**\nIn the arm64 case, a temporary file for kernel is created in\npez_arm64_probe() and passed down to info.kernel_fd later.\n\nHowever in the x86_64 case, nobody passes a temporary fd to info.kernel_fd,\nso finally the fd of the UKI image is passed to syscall. That is not the\nexpected behavior.\n\n**Situation**\nThe current framework of kexec-tools passes kernel fd to syscall.\nMeanwhile info.kernel_fd presents each probe with an opportunity to\noverwhelm the existing kernel fd.\n\n**Solution**\nIn the UKI probe, a temporary file is created to hold the .linux\nsection. Later, if no other probe methods set info.kernel_fd, it will be\nset in the UKI load method. This way, the correct fd can be passed to\nthe syscall.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nAcked-by: Dave Young \u003cdyoung@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "99d32ffdf1efb5e4d0b2c6f84f5186f5cf36b0bf", "tree": "b10628439d0321b669a17b5470988a7795645c17", "parents": [ "21361a91dedd802dcb198ef4d96c325dc6aaed02" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Tue Nov 05 11:57:05 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:26:33 2024 +0000" }, "message": "UKI: Split out the routine to create temporary fd\n\nThe process of creating temporary fd for initrd can be reused. Hence\nsplitting it out for later use.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nAcked-by: Dave Young \u003cdyoung@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "21361a91dedd802dcb198ef4d96c325dc6aaed02", "tree": "a2a5c37bf191c30f76c5685611c14a97bcfd370a", "parents": [ "2c8ca05212ff9557db0301a600587cd6d661fe31" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:25:05 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:25:05 2024 +0000" }, "message": "kexec-tools 2.0.30.git\n\nAdd .git to version so it doesn\u0027t look like a release.\nThis is just so when people build code from git it can\nbe identified as such from the version string.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "2c8ca05212ff9557db0301a600587cd6d661fe31", "tree": "81fb030b6e72ed665956ca496351c82f74d01164", "parents": [ "a7fcf2b90916d8732d52144884171b71749d27a5" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:09:51 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Dec 02 13:09:51 2024 +0000" }, "message": "kexec-tools 2.0.30\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "a7fcf2b90916d8732d52144884171b71749d27a5", "tree": "66d39da2536a621104706d57cad7513371646e3c", "parents": [ "07a6217af3cfb7b7952be22a58d4a3a365e422bb" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Nov 15 09:46:53 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Nov 15 09:46:53 2024 +0000" }, "message": "kexec-tools 2.0.30-rc1\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "07a6217af3cfb7b7952be22a58d4a3a365e422bb", "tree": "fd1dfd4338d7c7d5bbdee555598a4c8c95ac0fd1", "parents": [ "ec3e4dfeec22a2c78d6c7fbccf006bf4a7ba635d" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 13 17:44:37 2024 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Nov 15 09:41:59 2024 +0000" }, "message": "kexec: x86: Include pref_address when looking for a memory hole\n\nWhen the kernel is built relocatable, the starting address can be\nanything above 1Mb on x86. However, the startup_*() entries consider\nthat the minumum address for the kernel for the decompression has to\nbe at least LOAD_PHYSICAL_ADDR which is turn the value that is provided\nas pref_address in boot protocol. The boot protocol itself says:\n\n This field, if nonzero, represents a preferred load address for the\n kernel. A relocating bootloader should attempt to load at this\n address if possible.\n\nBesides that the code in the kernel (arch/x86/kernel/kexec-bzimage64.c)\nhas these lines (in bzImage64_load() function):\n\n if (header-\u003epref_address \u003c MIN_KERNEL_LOAD_ADDR)\n kbuf.buf_min \u003d MIN_KERNEL_LOAD_ADDR;\n else\n kbuf.buf_min \u003d header-\u003epref_address;\n\nAll that said, do the same in kexec tools. Without this patch\nthe relocatable kernel may end up in the memory hole that is\nnot enough for in-place decompression and Bad Things will happen\nas it\u0027s proven on Intel Merrifield, that has a reserved memory\nblock starting at 64Mb.\n\nNote, it doesn\u0027t mean that kernel has no issues itself in this particular\nstage, but at least we may work around some corner cases for kexec.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "ec3e4dfeec22a2c78d6c7fbccf006bf4a7ba635d", "tree": "e06922da39762bec09a3e04c1b8c307dbb8cfe63", "parents": [ "fb9302eb9bc80f4d88db69e78235407f644b46c5" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 06 19:58:07 2024 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Nov 07 13:49:03 2024 +0000" }, "message": "kexec: x86: Use init_size in 32-bit case as well\n\nIf the region of the memory is not enough for the decompressed\nkernel, e.g., it follows by reserved memory, Bad Things may happen\nas it\u0027s proven to be in practice on Intel Merrifield platform.\n\nInclude decompressed kernel size into the calculations for finding\nthe appropriate hole in memory to load kernel to.\n\nNote, this has been already done for 64-bit case, this just\nsynchronises the approach for 32-bit case.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "fb9302eb9bc80f4d88db69e78235407f644b46c5", "tree": "c372ee928968936a03796cf7e08e45483db8c6c8", "parents": [ "13e0a9327a42d78b0c0d0f672e02d8013187251a" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Mon Nov 04 21:07:44 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Nov 06 09:23:02 2024 +0000" }, "message": "Fix fd leakage in uki_image_probe()\n\nThere is a fd leakage in the fail path when writing to initrd fd in\nuki_image_probe(). Fix it by this patch.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "13e0a9327a42d78b0c0d0f672e02d8013187251a", "tree": "241b4b508a0bb80d08596c69dfabf00608613d35", "parents": [ "cd360f9f1aacd11a111e5da84708fc3181212181" ], "author": { "name": "Chen Haixiang", "email": "chenhaixiang3@huawei.com", "time": "Mon Oct 21 20:00:25 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Nov 04 09:18:01 2024 +0000" }, "message": "Before adding to usablemem_rgns, check if the memory range is already included.\n\nWhen kexec_iomem_for_each_line() reads from /proc/iomem,\nconcurrent modifications to /proc/iomem may lead to\nusablemem_rgns recording duplicate Crash kernel segments.\nThis can result in the number of retrieved Crash kernel\nsegments exceeding CRASH_MAX_RESERVED_RANGES, triggering\na realloc of the crash_reserved_mem in usablemem_rgns,\nwhich could crash the process. We should ensure that each\nrange added to usablemem_rgns is unique to prevent these issues.\n\nReviewed-by: Louhongxiang \u003clouhongxiang@huawei.com\u003e\nReviewed-by: wangbin \u003cwangbin224@huawei.com\u003e\nReviewed-by: yangyanchao \u003cyangyanchao6@huawei.com\u003e\nSigned-off-by: chenhaixiang \u003cchenhaixiang3@huawei.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "cd360f9f1aacd11a111e5da84708fc3181212181", "tree": "c20088b24320f8e94ce055d6c6f817f9903c7228", "parents": [ "0c72c4426ff542229e8b3c2a75600e7353376730" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Wed Oct 16 19:34:15 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 21 11:06:16 2024 +0100" }, "message": "arm64: Support UKI image format\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "0c72c4426ff542229e8b3c2a75600e7353376730", "tree": "d291c30724009becdaaaf9307a9d3ea2368402e5", "parents": [ "14fde160b906a2f0e5f6c55a4200efb7bcccc228" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Wed Oct 16 19:34:14 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 21 11:06:00 2024 +0100" }, "message": "kexec: Introduce UKI image parser\n\nA UKI image is a PE file that consists of several sections, typically\nincluding: .text, .data, .linux, .initrd, .cmdline, and others.\n\nThe kernel image is stored in the .linux section, which is one of the\nformats currently recognized by kexec-tools. Therefore, the UKI parser\ncan be used to strip away the UKI layer, allowing the other parser to\ncontinue the procession of the kernel image.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "14fde160b906a2f0e5f6c55a4200efb7bcccc228", "tree": "93078f4f53105eb5804efeddc8f0713bf27c9796", "parents": [ "ba1ad3017946aa49c5296cf9ae0b8b426d1f8d07" ], "author": { "name": "Pingfan Liu", "email": "piliu@redhat.com", "time": "Wed Oct 16 19:34:13 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 21 11:05:31 2024 +0100" }, "message": "kexec: Introduce implicit_initrd_fd to pass internal initrd information\n\nAt present, initrd information is passed through --initrd args, and\nhandled until image load. But for the UKI case, the initrd may be stored\nimplicitly in the .initrd section.\n\nIt means that initrd is perceived at the probe stage. And there should\nbe a way to carry this information forward to the load stage. In oder to\nimplement that, introducing a global implicit_initrd_fd.\n\nSigned-off-by: Pingfan Liu \u003cpiliu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "ba1ad3017946aa49c5296cf9ae0b8b426d1f8d07", "tree": "72c285e519653a99a9551abd35215bb8b912a8e3", "parents": [ "d2f42971661b61a1baad853cdfea54e785ba59ca" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Mon Oct 14 18:31:29 2024 +0300" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 14 17:21:09 2024 +0100" }, "message": "build: Update configure.ac to follow the changes\n\nRunning bootstrap on Debian spills a few problems in configure.ac.\nUpdate it accordingly.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "d2f42971661b61a1baad853cdfea54e785ba59ca", "tree": "f99ea9bf6dd839162739975a9b4649177d117926", "parents": [ "07821da7cf96928c3b3800992e5eaca9b7aa7ec9" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Mon Oct 14 18:31:28 2024 +0300" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 14 17:20:57 2024 +0100" }, "message": "util_lib/elf_info: Fix format specifiers for fprintf()\n\nCompiler is not happy when wrong specifier is being used.\nFix them all.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "07821da7cf96928c3b3800992e5eaca9b7aa7ec9", "tree": "9c133680e78c983efb2615f000e5b6ac19dec8b4", "parents": [ "8767f1e1b25e13ff8743a755499e91400fcbefd0" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Mon Oct 14 18:31:27 2024 +0300" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 14 17:20:38 2024 +0100" }, "message": "kexec: i386: Fix format specifiers for various printf() and scanf()\n\nCompiler is not happy when wrong specifier is being used.\nFix them all.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "8767f1e1b25e13ff8743a755499e91400fcbefd0", "tree": "708c3c13e42b9edfa8277f505c8925ebe7376afc", "parents": [ "5b76649e0c2367c98d3c87feaadff693e4bfacd7" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Mon Oct 14 18:31:26 2024 +0300" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Oct 14 17:19:59 2024 +0100" }, "message": "kexec: i386: Fix 32-bit right shifts on 32-bit architectures\n\n32-bit compilation makes compiler unhappy about too big right shifts.\nUse a similar trick to Linux kernel project by replacing foo \u003e\u003e 32 by\n(foo \u003e\u003e 16) \u003e\u003e 16.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "5b76649e0c2367c98d3c87feaadff693e4bfacd7", "tree": "59370e5995f59dd76efaa16aa6981b48e66f2b05", "parents": [ "8dd984687b7961e6a810334424627c82d2a11684" ], "author": { "name": "Julien Olivain", "email": "ju.o@free.fr", "time": "Fri Jul 19 19:47:44 2024 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 23 10:56:05 2024 +0100" }, "message": "s390: add missing limits.h header inclusion\n\nCommit b84fcdd48c [1] introduced a usage of PATH_MAX in file\n\"kexec-s390.c\". This constant is defined in the \u003climits.h\u003e header.\nSee [2].\n\nWith some libc, such as musl, kexec-tools fails to build with output:\n\n kexec/arch/s390/kexec-s390.c: In function \u0027remove_offline_memory\u0027:\n kexec/arch/s390/kexec-s390.c:120:19: error: \u0027PATH_MAX\u0027 undeclared (first use in this function)\n 120 | char path[PATH_MAX];\n | ^~~~~~~~\n\nThe issue can be reproduced, for example, with Buildroot Linux, using\na s390 gcc toolchain with the musl libc and the following commands:\n\n git clone -b 2024.05.1 --depth 1 https://gitlab.com/buildroot.org/buildroot.git\n cd buildroot\n cat \u003e .config \u003c\u003cEOF\n BR2_s390x\u003dy\n BR2_TOOLCHAIN_BUILDROOT_MUSL\u003dy\n BR2_PACKAGE_KEXEC\u003dy\n EOF\n make olddefconfig\n make\n\nNote: in this test, kexec-tools v2.0.28 is used.\n\nThis commit fixes the issue by adding the missing header inclusion.\n\n[1] https://git.kernel.org/pub/scm/utils/kernel/kexec/kexec-tools.git/commit/?id\u003db84fcdd48cbe70e54e66f2cf1b251e0db58f04f4\n[2] https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/limits.h.html\n\nSigned-off-by: Julien Olivain \u003cju.o@free.fr\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "8dd984687b7961e6a810334424627c82d2a11684", "tree": "cf1b774d6dcafb45e3e6d1835af7382706a89093", "parents": [ "992c419f6cbf2a477a02d6e8e6d447194ffb80dc" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 23 10:53:46 2024 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 23 10:53:46 2024 +0100" }, "message": "kexec-tools 2.0.29.git\n\nAdd .git to version so it doesn\u0027t look like a release.\nThis is just so when people build code from git it can\nbe identified as such from the version string.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "992c419f6cbf2a477a02d6e8e6d447194ffb80dc", "tree": "1a64eabc545f81d5631053020e06e6b32ba20a30", "parents": [ "920ae8851837a6945dcb3736b045a132d7e5bfa8" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jul 18 10:21:19 2024 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jul 18 10:21:19 2024 +0100" }, "message": "kexec-tools 2.0.29\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "920ae8851837a6945dcb3736b045a132d7e5bfa8", "tree": "5789eb5fb1982b95b00149ea9b2dfd206e561c9a", "parents": [ "0be146b657715164c1154f9ca7f886171d3a79f2" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Jul 12 10:19:55 2024 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Jul 12 10:19:55 2024 +0100" }, "message": "kexec-tools 2.0.29-rc1\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "0be146b657715164c1154f9ca7f886171d3a79f2", "tree": "ad6f9ccdb48d1e78877724eed0f02b8c21db50de", "parents": [ "b8408c95e7c03ea0a0398c9caa872b0b004a373c" ], "author": { "name": "Sourabh Jain", "email": "sourabhjain@linux.ibm.com", "time": "Sun Jul 07 20:54:56 2024 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 09 08:26:32 2024 +0100" }, "message": "doc/hotplug: update man and --help\n\nUpdate the man page and --help option to make the description of the\n--hotplug option easier to understand.\n\nAcked-by: Baoquan He \u003cbhe@redhat.com\u003e\nAcked-by: Hari Bathini \u003chbathini@linux.ibm.com\u003e\nSigned-off-by: Sourabh Jain \u003csourabhjain@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "b8408c95e7c03ea0a0398c9caa872b0b004a373c", "tree": "c26ef0dd2d96121ff694583e1d1d0aa62ac05478", "parents": [ "3c47f384f1c402722bd6662c060c0af688478095" ], "author": { "name": "Sourabh Jain", "email": "sourabhjain@linux.ibm.com", "time": "Sun Jul 07 20:54:55 2024 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 09 08:26:07 2024 +0100" }, "message": "powerpc/kexec_load: add hotplug support\n\nKernel commits b741092d5976 (\"powerpc/crash: add crash CPU hotplug\nsupport\") and 849599b702ef (\"powerpc/crash: add crash memory hotplug\nsupport\") added crash CPU/Memory hotplug support on PowerPC. This patch\nextends that support for the kexec_load syscall.\n\nDuring CPU/Memory hotplug events on PowerPC, two kexec segments,\nelfcorehdr, and FDT, get updated by the kernel. To ensure the kernel\ncan safely update these two kexec segments for the kdump image loaded\nusing the kexec_load system call, the following changes are made:\n\n1. Extra size is allocated for both elfcorehdr and FDT to accommodate\n additional resources in the future. For the elfcorehdr, the size hint\n is taken from /sys/kernel/crash_elfcorehdr_size sysfs, while for FDT,\n extra size is allocated to hold possible CPU nodes.\n\n2. Both elfcorehdr and FDT are skipped from SHA calculation.\n\nAcked-by: Hari Bathini \u003chbathini@linux.ibm.com\u003e\nSigned-off-by: Sourabh Jain \u003csourabhjain@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "3c47f384f1c402722bd6662c060c0af688478095", "tree": "dc6b54106e7ddab1ffd937d8ad92a0154c5414c6", "parents": [ "4fd0553cdfe397920d418d8b90ac52f8f1303a2d" ], "author": { "name": "Sourabh Jain", "email": "sourabhjain@linux.ibm.com", "time": "Sun Jul 07 20:54:54 2024 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jul 09 08:25:19 2024 +0100" }, "message": "kexec_load: Use new kexec flag for hotplug support\n\nKernel commit 79365026f869 (crash: add a new kexec flag for hotplug\nsupport) has introduced a new kexec flag to generalize hotplug support.\nThe newly introduced kexec flags for hotplug allow architectures to\nexclude all the required kexec segments from SHA calculation so that\nthe kernel can update them on hotplug events. This was not possible\nearlier with the KEXEC_UPDATE_ELFCOREHDR kexec flags since it was added\nonly for the elfcorehdr segment.\n\nTo enable architectures to control the list of kexec segments to exclude\nwhen hotplug support is enabled, add a new architecture-specific\nfunction named arch_do_exclude_segment. During the SHA calculation, this\nfunction gets called to let the architecture decide whether a specific\nkexec segment should be considered for SHA calculation or not.\n\nNote: To avoid breaking backward compatibility, the new kexec flag\nKEXEC_CRASH_HOTPLUG_SUPPORT is not used for x86 for now.\n\nSigned-off-by: Sourabh Jain \u003csourabhjain@linux.ibm.com\u003e\nAcked-by: Baoquan He \u003cbhe@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "4fd0553cdfe397920d418d8b90ac52f8f1303a2d", "tree": "002ee6304185dd1799f087b7c15e82b60133de95", "parents": [ "2e8a93af1e852d7b197b8561e433058c668fcc9c" ], "author": { "name": "Khem Raj", "email": "raj.khem@gmail.com", "time": "Wed May 15 21:35:05 2024 -0700" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Jun 18 17:29:20 2024 +0100" }, "message": "x86-linux-setup.c: Use POSIX basename API\n\nMusl C library only supports POSIX basename function. while glibc has\nboth GNU extention as well as POSIX basename implemented. Switch to\nusing posix version, so it can work across musl and glibc\n\nbasename prototype has been removed from string.h from latest musl [1]\ncompilers e.g. clang-18/GCC-14 flags the absense of prototype as error.\ntherefore include libgen.h for providing it.\n\n[1] https://git.musl-libc.org/cgit/musl/commit/?id\u003d725e17ed6dff4d0cd22487bb64470881e86a92e7\n\nSigned-off-by: Khem Raj \u003craj.khem@gmail.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "2e8a93af1e852d7b197b8561e433058c668fcc9c", "tree": "12ac8c333c27c081afda962ec1de9bdd1709116d", "parents": [ "69ccfa12d5e3851e4538711db1f696a4f2180558" ], "author": { "name": "Ming Wang", "email": "wangming01@loongson.cn", "time": "Mon May 20 09:45:53 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Jun 03 18:55:58 2024 +0100" }, "message": "kexec: loongarch: fix load command line segment error\n\nWhen kernel cmdline reserved multiple crash kernel segments, like:\n\"crashkernel\u003d4G,high crashkernel\u003d256M,low\".\nCmdline segment may be loaded in locations larger than 4G of memory.\nSo the parameter \u0027buf_max\u0027 of add_buffer function should be\n\u0027hole_max\u0027.\n\nSigned-off-by: Hongchen Zhang \u003czhanghongchen@loongson.cn\u003e\nSigned-off-by: Ming Wang \u003cwangming01@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "69ccfa12d5e3851e4538711db1f696a4f2180558", "tree": "521492ee3723a685b674d28a9af62779ec7db813", "parents": [ "3194d1eba9c182602eabc107c874045f414f6341" ], "author": { "name": "Ming Wang", "email": "wangming01@loongson.cn", "time": "Mon May 20 09:32:41 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Jun 03 18:54:31 2024 +0100" }, "message": "kexec: loongarch: add multi crash kernel segment support\n\nProduction kernel may reserved multiple crash kernel segments, like:\n\u0027crashkernel\u003d4G,high crashkernel\u003d256M,low\u0027.\nSo we should pass multiple mem\u003d\u003csize\u003e@\u003caddress\u003e parameters in capture kernel.\n\nSigned-off-by: Hongchen Zhang \u003czhanghongchen@loongson.cn\u003e\nSigned-off-by: Ming Wang \u003cwangming01@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "3194d1eba9c182602eabc107c874045f414f6341", "tree": "c64de2b35474fbc3833aec8e7931f17b8edc7286", "parents": [ "2a3e54e9f734ce08dee20b4ff1bdd5037f1b0193" ], "author": { "name": "Hongchen Zhang", "email": "zhanghongchen@loongson.cn", "time": "Mon May 20 09:20:16 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Mon Jun 03 18:52:38 2024 +0100" }, "message": "LoongArch: fix kernel image size error\n\nUse kernel_size instead of kernel image size is wrong, we should get\nthe kernel image size from loongarch image header.\n\nSigned-off-by: Ming Wang \u003cwangming01@loongson.cn\u003e\nSigned-off-by: Hongchen Zhang \u003czhanghongchen@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "2a3e54e9f734ce08dee20b4ff1bdd5037f1b0193", "tree": "5a0dac0554f31435c5567d78037afe7706f8d8fb", "parents": [ "f53bcef5063c70960b2b8eb8990722d5c9d1b99e" ], "author": { "name": "Haiqing Bai", "email": "Haiqing.Bai@windriver.com", "time": "Wed Apr 24 14:17:27 2024 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Apr 26 18:38:36 2024 +0100" }, "message": "ARM: Fix add_buffer_phys_virt() align issue\n\nWhen \"CONFIG_ARM_LPAE\" is enabled,3 level page table\nis used by MMU, the \"SECTION_SIZE\" is defined with\n(1 \u003c\u003c 21), but \u0027add_buffer_phys_virt()\u0027 hardcode this\nto (1 \u003c\u003c 20).\n\nSuggested-By: fredrik.markstrom@gmail.com\nSigned-off-by: Haiqing Bai \u003cHaiqing.Bai@windriver.com\u003e\nSigned-off-by: Alexander Kanavin \u003calex@linutronix.de\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "f53bcef5063c70960b2b8eb8990722d5c9d1b99e", "tree": "b58c7aad9d92c119c20b6c5cfe9c88269ad88ca3", "parents": [ "623ceb333c18df6f75d46940d9e1934004bc8ae0" ], "author": { "name": "Coiby Xu", "email": "coxu@redhat.com", "time": "Mon Apr 08 14:37:55 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Tue Apr 23 11:13:57 2024 +0100" }, "message": "Fix incorrect Free Software Foundation address in the license\n\nAs detected by rpmlint,\n kexec-tools.x86_64: E: incorrect-fsf-address /usr/share/licenses/kexec-tools/COPYING\n The Free Software Foundation address in this file seems to be outdated or\n misspelled. Ask upstream to update the address, or if this is a license file,\n possibly the entire file with a new copy available from the FSF.\n\nReplace current one with\nhttps://www.gnu.org/licenses/old-licenses/gpl-2.0.txt\nto fix this issue.\n\nSigned-off-by: Coiby Xu \u003ccoxu@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "623ceb333c18df6f75d46940d9e1934004bc8ae0", "tree": "196b3d63bf4df212415ed4618658cdb019bf5681", "parents": [ "9d9cf8de8b2ad8273861a30476a46f34cd34871a" ], "author": { "name": "Baoquan He", "email": "bhe@redhat.com", "time": "Thu Mar 21 15:30:37 2024 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Mar 22 20:45:54 2024 +0000" }, "message": "util_lib/elf_info.c: fix a warning\n\nThere\u0027s a incorrect array operation in function scan_vmcoreinfo(), it\nwill cause below warning message.\n\n-----------------------\nutil_lib/elf_info.c: In function ‘scan_vmcoreinfo’:\nutil_lib/elf_info.c:360:43: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow\u003d]\n 360 | temp_buf[len + 1] \u003d \u0027\\0\u0027;\n | ~~~~~~~~~~~~~~~~~~^~~~~~\nutil_lib/elf_info.c:319:14: note: at offset 1024 into destination object ‘temp_buf’ of size 1024\n 319 | char temp_buf[1024];\n | ^~~~~~~~\n---------------------\n\nFix it to avoid oob access of array.\n\nSigned-off-by: Baoquan He \u003cbhe@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "9d9cf8de8b2ad8273861a30476a46f34cd34871a", "tree": "89546eae8bd562ec813724f9df6e0c5685ff0ed9", "parents": [ "6ee2ac1bf7395a66c50365a329ae0a1519094c92" ], "author": { "name": "Baoquan He", "email": "bhe@redhat.com", "time": "Tue Nov 14 23:20:30 2023 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Mar 15 09:32:30 2024 +0100" }, "message": "kexec_file: add kexec_file flag to support debug printing\n\nThis add KEXEC_FILE_DEBUG to kexec_file_flags so that it can be passed\nto kernel when \u0027-d\u0027 is added with kexec_file_load interface. With that\nflag enabled, kernel can enable the debugging message printing.\n\nSigned-off-by: Baoquan He \u003cbhe@redhat.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "6ee2ac1bf7395a66c50365a329ae0a1519094c92", "tree": "7a775e8fe80b5ba0345019a02f9741e468a9122a", "parents": [ "328de8e00e298f00d7ba6b25dc3950147e9642e6" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Feb 02 09:44:09 2024 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Feb 07 09:50:55 2024 +0000" }, "message": "workflow: update to use checkout@v4\n\nUpdate to use checkout@v4.\n\nThis addresses the following warning that appears in GitHub runs:\n\n\"Node.js 16 actions are deprecated. Please update the following actions\n to use Node.js 20: actions/checkout@v3. For more information\n see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.\n\nLink: https://github.com/horms/kexec-tools/actions/runs/7753454923\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "328de8e00e298f00d7ba6b25dc3950147e9642e6", "tree": "a7f6e49daee75a853a6f196f9225b51196c0eabb", "parents": [ "94fbe64fb22d61726ca0c0996987574b6c783c19" ], "author": { "name": "Michel Lind", "email": "salimma@fedoraproject.org", "time": "Tue Jan 30 04:14:31 2024 -0600" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Fri Feb 02 09:39:06 2024 +0100" }, "message": "Fix building on x86_64 with binutils 2.41\n\nNewer versions of the GNU assembler (observed with binutils 2.41) will\ncomplain about the \".arch i386\" in files assembled with \"as --64\",\nwith the message \"Error: 64bit mode not supported on \u0027i386\u0027\".\n\nFix by moving \".arch i386\" below the relevant \".code32\" directive, so\nthat the assembler is no longer expecting 64-bit instructions to be used\nby the time that the \".arch i386\" directive is encountered.\n\nBased on similar iPXE fix:\nhttps://github.com/ipxe/ipxe/commit/6ca597eee\n\nSigned-off-by: Michel Lind \u003cmichel@michel-slm.name\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "94fbe64fb22d61726ca0c0996987574b6c783c19", "tree": "98e8b3ccac73caa72a4735292bb31a7ca09d38c8", "parents": [ "21ea4569856aecf7467b6a2035a10ca89e8f7789" ], "author": { "name": "Jiri Bohac", "email": "jbohac@suse.cz", "time": "Tue Jan 16 18:14:31 2024 +0100" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jan 25 16:54:50 2024 +0000" }, "message": "kexec: don\u0027t use kexec_file_load on XEN\n\nSince commit 29fe5067ed07 (\"kexec: make -a the default\")\nkexec tries the kexec_file_load syscall first and only falls back to kexec_load on\nselected error codes.\n\nThis effectively breaks kexec on XEN, unless -c is pecified to force the kexec_load\nsyscall.\n\nThe XEN-specific functions (xen_kexec_load / xen_kexec_unload) are only called\nfrom my_load / k_unload, i.e. the kexec_load code path.\n\nWith -p (panic kernel) kexec_file_load on XEN fails with -EADDRNOTAVAIL (crash\nkernel reservation is ignored by the kernel on XEN), which is not in the list\nof return codes that cause the fallback to kexec_file.\n\nWithout -p kexec_file_load actualy leads to a kernel oops on v6.4.0\n(needs to be dubugged separately).\n\nSigned-off-by: Jiri Bohac \u003cjbohac@suse.cz\u003e\nFixes: 29fe5067ed07 (\"kexec: make -a the default\")\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "21ea4569856aecf7467b6a2035a10ca89e8f7789", "tree": "bd078cbcf210f0b067e451e18f7211c4ce606fe5", "parents": [ "adef8a8e4bfda3cf43e8ab87c3b210cdf723ae33" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jan 25 16:54:40 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jan 25 16:54:40 2024 +0000" }, "message": "kexec-tools 2.0.28.git\n\nAdd .git to version so it doesn\u0027t look like a release.\nThis is just so when people build code from git it can\nbe identified as such from the version string.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "adef8a8e4bfda3cf43e8ab87c3b210cdf723ae33", "tree": "4c2d490d69c6276273fe57e584d2b0281064bbd4", "parents": [ "5d7bc25cf15b872d92366e44fdc7da6a44f8b12b" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jan 11 11:28:33 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Jan 11 11:28:33 2024 +0000" }, "message": "kexec-tools 2.0.28\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "5d7bc25cf15b872d92366e44fdc7da6a44f8b12b", "tree": "29eb7787584b51b573b4ab5ab3b8f342741f661d", "parents": [ "549466430ae68f7df6cfa65088fe55d4ef926d8e" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Jan 03 10:08:39 2024 +0000" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Jan 03 10:08:39 2024 +0000" }, "message": "kexec-tools 2.0.28-rc1\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "549466430ae68f7df6cfa65088fe55d4ef926d8e", "tree": "b8555a80c644fc6bd01564ddcde3ccbc4fca3318", "parents": [ "ba0ac0efe29927d135d073176bba559a25733b43" ], "author": { "name": "WANG Rui", "email": "wangrui@loongson.cn", "time": "Sat Nov 25 14:52:32 2023 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Sat Dec 02 12:42:17 2023 +0000" }, "message": "LoongArch: Load vmlinux.efi to the link address\n\nCurrently, kexec loads vmlinux.efi to address 0 instead of the link\naddress. This causes kexec to fail to boot the new vmlinux.efi on qemu.\n\n pei_loongarch_load: kernel_segment: 0000000000000000\n pei_loongarch_load: kernel_entry: 00000000013f1000\n pei_loongarch_load: image_size: 0000000001ca0000\n pei_loongarch_load: text_offset: 0000000000200000\n pei_loongarch_load: phys_offset: 0000000000000000\n pei_loongarch_load: PE format: yes\n loongarch_load_other_segments:333: command_line: kexec console\u003dttyS0,115200\n kexec_load: entry \u003d 0x13f1000 flags \u003d 0x1020000\n nr_segments \u003d 2\n segment[0].buf \u003d 0x7fffeea38010\n segment[0].bufsz \u003d 0x1b55200\n segment[0].mem \u003d (nil)\n segment[0].memsz \u003d 0x1ca0000\n segment[1].buf \u003d 0x5555570940b0\n segment[1].bufsz \u003d 0x200\n segment[1].mem \u003d 0x1ca0000\n segment[1].memsz \u003d 0x4000\n\nThis patch constrains the range of the kernel segment by `hole_min`\nand `hole_max` to place vmlinux.efi exactly at the link address.\n\n pei_loongarch_load: kernel_segment: 0000000000200000\n pei_loongarch_load: kernel_entry: 00000000013f1000\n pei_loongarch_load: image_size: 0000000001ca0000\n pei_loongarch_load: text_offset: 0000000000200000\n pei_loongarch_load: phys_offset: 0000000000000000\n pei_loongarch_load: PE format: yes\n loongarch_load_other_segments:339: command_line: kexec console\u003dttyS0,115200\n kexec_load: entry \u003d 0x13f1000 flags \u003d 0x1020000\n nr_segments \u003d 2\n segment[0].buf \u003d 0x7ffff2028010\n segment[0].bufsz \u003d 0x1b55200\n segment[0].mem \u003d 0x200000\n segment[0].memsz \u003d 0x1ca0000\n segment[1].buf \u003d 0x555557498098\n segment[1].bufsz \u003d 0x200\n segment[1].mem \u003d 0x1ea0000\n segment[1].memsz \u003d 0x4000\n\nSigned-off-by: WANG Rui \u003cwangrui@loongson.cn\u003e\nReviewed-by: Huacai Chen \u003cchenhuacai@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "ba0ac0efe29927d135d073176bba559a25733b43", "tree": "a4a6f67fb39aa91708b6552fb7cb9b2cd20e5138", "parents": [ "74dfaefd631694d771a38ede308da2083eae3a3b" ], "author": { "name": "WANG Rui", "email": "wangrui@loongson.cn", "time": "Fri Nov 24 16:54:10 2023 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Sat Dec 02 12:42:17 2023 +0000" }, "message": "LoongArch: Fix an issue with relocatable vmlinux\n\nNormally vmlinux for LoongArch is of ET_EXEC type, while if built with\nCONFIG_RELOCATABLE (this is PIE) and Clang, it will be of ET_DYN type.\nMeanwhile, physical address field of segments in vmlinux has actually\nthe same value as virtual address field.\n\nSimilar to arm64, this patch allows to unconditionally skip the check\non LoongArch.\n\nLink: https://github.com/ClangBuiltLinux/linux/issues/1963\nSigned-off-by: WANG Rui \u003cwangrui@loongson.cn\u003e\nReviewed-by: Huacai Chen \u003cchenhuacai@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "74dfaefd631694d771a38ede308da2083eae3a3b", "tree": "75bfff2725956be8b60a9b323b7c69c781d237f3", "parents": [ "22dcf5cb940a1f1174d601ad7dacf16b91ca6b44" ], "author": { "name": "Laurent Vivier", "email": "laurent@vivier.eu", "time": "Sat Apr 22 11:59:04 2023 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Sat Dec 02 12:42:17 2023 +0000" }, "message": "m68k: fix getrandom() use with uclibc\n\nWith uclibc, getrandom() is only defined with _GNU_SOURCE, fix that:\n\nkexec/arch/m68k/bootinfo.c: In function \u0027bootinfo_add_rng_seed\u0027:\nkexec/arch/m68k/bootinfo.c:231:13: warning: implicit declaration of function \u0027getrandom\u0027; did you mean \u0027srandom\u0027? [-Wimplicit-function-declaration]\n 231 | if (getrandom(bi-\u003erng_seed.data, RNG_SEED_LEN, GRND_NONBLOCK) !\u003d RNG_SEED_LEN) {\n | ^~~~~~~~~\n | srandom\nkexec/arch/m68k/bootinfo.c:231:56: error: \u0027GRND_NONBLOCK\u0027 undeclared (first use in this function)\n 231 | if (getrandom(bi-\u003erng_seed.data, RNG_SEED_LEN, GRND_NONBLOCK) !\u003d RNG_SEED_LEN) {\n | ^~~~~~~~~~~~~\n\nFixes: b9de05184816 (\"m68k: pass rng seed via BI_RNG_SEED\")\nCc: Jason@zx2c4.com\nSigned-off-by: Laurent Vivier \u003claurent@vivier.eu\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "22dcf5cb940a1f1174d601ad7dacf16b91ca6b44", "tree": "d8d479af005e6d0ede131f7a677749432fa41045", "parents": [ "44e7b73c331f557fb1e6cddada11205a7a96de46" ], "author": { "name": "WANG Rui", "email": "wangrui@loongson.cn", "time": "Sat Nov 25 15:26:43 2023 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Sat Dec 02 12:42:17 2023 +0000" }, "message": "lzma: Relax memory limit for lzma decompressor\n\nThe kexec cannot load LZMA compressed vmlinuz.efi on LoongArch.\n\n Try LZMA decompression.\n lzma_decompress_file: read on /tmp/Image4yyfhM of 65536 bytes failed\n pez_prepare: decompressed size 8563960\n pez_prepare: done\n Cannot load vmlinuz.efi\n\nThe root cause is that lzma decompressor requires more memory usage,\nwhich exceeds the current 64M limit.\n\nReported-by: Huacai Chen \u003cchenhuacai@kernel.org\u003e\nSigned-off-by: WANG Rui \u003cwangrui@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "44e7b73c331f557fb1e6cddada11205a7a96de46", "tree": "f666a13e96511ae6a17588c24071490133c9d43d", "parents": [ "74d66d405f30fb247da9684f6e29bbef8a4f56cf" ], "author": { "name": "Aditya Gupta", "email": "adityag@linux.ibm.com", "time": "Thu Nov 16 14:11:37 2023 +0530" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Thu Nov 16 11:32:00 2023 +0000" }, "message": "kexec: ppc64: print help to stdout instead of stderr\n\nCurrently \u0027kexec --help\u0027 on powerpc64 prints the generic help/usage to\nstdout, and the powerpc64 specific options to stderr\n\nThat is, if the stdout of \u0027kexec --help\u0027 is redirected to some file,\nsome of the help options will not be redirected, and instead printed on\nthe terminal/stderr:\n\n [root@machine kexec-tools]# kexec --help \u003e /tmp/out\n --command-line\u003d\u003cCommand line\u003e command line to append.\n --append\u003d\u003cCommand line\u003e same as --command-line.\n --ramdisk\u003d\u003cfilename\u003e Initial RAM disk.\n --initrd\u003d\u003cfilename\u003e same as --ramdisk.\n --devicetreeblob\u003d\u003cfilename\u003e Specify device tree blob file.\n Not applicable while using\n --kexec-file-syscall.\n --dtb\u003d\u003cfilename\u003e same as --devicetreeblob.\n elf support is still broken\n --elf64-core-headers Prepare core headers in ELF64 format\n --dt-no-old-root Do not reuse old kernel root\u003d param.\n while creating flatten device tree.\n\nFix this inconsistency by writing powerpc64 specific options to stdout,\nsimilar to the generic \u0027kexec --help\u0027\n\nWith the proposed changes, it is like this (nothing printed to stderr):\n\n [root@machine kexec-tools]# ./build/sbin/kexec --help \u003e /tmp/out\n\nReported-by: Srikar Dronamraju \u003csrikar@linux.vnet.ibm.com\u003e\nSigned-off-by: Aditya Gupta \u003cadityag@linux.ibm.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "74d66d405f30fb247da9684f6e29bbef8a4f56cf", "tree": "d21674d48ff3d3401649118949c5c78392d6d435", "parents": [ "ab3a70af85679bbc5efe63057c7f65365ed6e748" ], "author": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Oct 11 10:54:27 2023 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Oct 11 10:54:27 2023 +0200" }, "message": "workflow: update to Ubuntu 22.04\n\nBump workflow to use latest LTS release.\n\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "ab3a70af85679bbc5efe63057c7f65365ed6e748", "tree": "ce06a19ee252e3927cca7ae21865d41e28cc7081", "parents": [ "6419b008fde783fd0cc2cc266bd1c9cf35e99a0e" ], "author": { "name": "Ming Wang", "email": "wangming01@loongson.cn", "time": "Wed Oct 11 11:56:21 2023 +0800" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Oct 11 10:51:59 2023 +0200" }, "message": "kexec/loongarch64: fix \u0027make dist\u0027 file loss issue\n\nThe Makefile omits the iomem.h file, causing the archive file\ngenerated by \u0027make dist\u0027 to lose iomem.h. This patch is used to\nfix this problem.\n\nSigned-off-by: Ming Wang \u003cwangming01@loongson.cn\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "6419b008fde783fd0cc2cc266bd1c9cf35e99a0e", "tree": "83fde1863fd2dce031c5337e42f2c50dfd9daaaf", "parents": [ "118b567ce74af750535c5b77ab8c97b0e1a6b931" ], "author": { "name": "Julien Olivain", "email": "ju.o@free.fr", "time": "Sat Sep 23 18:46:06 2023 +0200" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Oct 04 14:06:49 2023 +0200" }, "message": "kexec: provide a memfd_create() wrapper if not present in libc\n\nCommit 714fa115 \"kexec/arm64: Simplify the code for zImage\" introduced\na use of the memfd_create() system call, included in version\nkexec-tools v2.0.27.\n\nThis system call was introduced in kernel commit [1], first included\nin kernel v3.17 (released on 2014-10-05).\n\nThe memfd_create() glibc wrapper function was added much later in\ncommit [2], first included in glibc version 2.27 (released on\n2018-02-01).\n\nThis direct use memfd_create() introduced a requirement on\nKernel \u003e\u003d 3.17 and glibc \u003e\u003d 2.27.\n\nThere is old toolchains like [3] for example (which ships gcc 7.3.1,\nglibc 2.25 and includes kernel v4.10 headers), that can still be used\nto build newer kernels. Even if such toolchains can be seen as\noutdated, they are is still claimed as supported by recent kernel.\nFor example, Kernel v6.5.5 has a requirement on gcc version 5.1 and\ngreater. See [4].\n\nMoreover, kexec-tools \u003c\u003d 2.0.26 could be compiled using recent\ntoolchains with alternative libc (e.g. uclibc-ng, musl) which are not\nproviding the memfd_create() wrapper.\n\nWhen compiling kexec-tools v2.0.27 with a toolchain not providing the\nmemfd_create() syscall wrapper, the compilation fail with message:\n\n kexec/kexec.c: In function \u0027copybuf_memfd\u0027:\n kexec/kexec.c:645:7: warning: implicit declaration of function \u0027memfd_create\u0027; did you mean \u0027SYS_memfd_create\u0027? [-Wimplicit-function-declaration]\n fd \u003d memfd_create(\"kernel\", MFD_ALLOW_SEALING);\n ^~~~~~~~~~~~\n SYS_memfd_create\n kexec/kexec.c:645:30: error: \u0027MFD_ALLOW_SEALING\u0027 undeclared (first use in this function); did you mean \u0027_PC_ALLOC_SIZE_MIN\u0027?\n fd \u003d memfd_create(\"kernel\", MFD_ALLOW_SEALING);\n ^~~~~~~~~~~~~~~~~\n _PC_ALLOC_SIZE_MIN\n\nIn order to let kexec-tools compile in a wider range of configurations,\nthis commit adds a memfd_create() function check in autoconf configure\nscript, and adds a system call wrapper which will be used if the\nfunction is not available. With this commit, the environment\nrequirement is relaxed to only kernel \u003e\u003d v3.17.\n\nNote: this issue was found in kexec-tools integration in Buildroot [5]\nusing the command \"utils/test-pkg -a -p kexec\", which tests many\ntoolchain/arch combinations.\n\n[1] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id\u003d9183df25fe7b194563db3fec6dc3202a5855839c\n[2] https://sourceware.org/git/?p\u003dglibc.git;a\u003dcommitdiff;h\u003d59d2cbb1fe4b8601d5cbd359c3806973eab6c62d\n[3] https://releases.linaro.org/components/toolchain/binaries/7.3-2018.05/aarch64-linux-gnu/gcc-linaro-7.3.1-2018.05-x86_64_aarch64-linux-gnu.tar.xz\n[4] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/Documentation/process/changes.rst?h\u003dv6.5.5#n32\n[5] https://buildroot.org/\n\nSigned-off-by: Julien Olivain \u003cju.o@free.fr\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" }, { "commit": "118b567ce74af750535c5b77ab8c97b0e1a6b931", "tree": "80ff172b812ec329cb47f258b62c55ede8cac99b", "parents": [ "d59d17f372398d270f9d0f1aaf37f5ad1c4a601d" ], "author": { "name": "Eric DeVolder", "email": "eric.devolder@oracle.com", "time": "Wed Sep 27 14:11:36 2023 -0400" }, "committer": { "name": "Simon Horman", "email": "horms@kernel.org", "time": "Wed Oct 04 14:04:48 2023 +0200" }, "message": "crashdump/x86: set the elfcorehdr segment size for hotplug\n\nFor hotplug, the elfcorehdr segment must be sized appropriately\nto allow a growing number of CPUs or memory regions. Use the size\nreported by the kernel via /sys/kernel/crash_elfcorehdr_sz.\n\nSigned-off-by: Eric DeVolder \u003ceric.devolder@oracle.com\u003e\nSigned-off-by: Simon Horman \u003chorms@kernel.org\u003e\n" } ], "next": "d59d17f372398d270f9d0f1aaf37f5ad1c4a601d" }