{"id":"CVE-2026-97942","summary":"x86/alternatives: Exclude text poking against change_page_attr()","details":"In the Linux kernel, the following vulnerability has been resolved:\n\nx86/alternatives: Exclude text poking against change_page_attr()\n\nFrom time to time, the following BUG can be observed\nin the x86 alternatives patching code [0]:\n\n  \u003e kernel BUG at arch/x86/kernel/alternative.c:2576!\n  \u003e Oops: invalid opcode: 0000 [#1] SMP NOPTI\n  \u003e CPU: 0 UID: 0 PID: 355 Comm: (udev-worker) Not tainted 7.1.3-1-default #1 PREEMPT(full) openSUSE Tumbleweed  8c1795b03ec64f997e57a8ad38b1161e3b98da64\n  \u003e Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS unknown 02/02/2022\n  \u003e RIP: 0010:__text_poke+0x2aa/0x450\n  \u003e Call Trace:\n  \u003e  \u003cTASK\u003e\n  \u003e  smp_text_poke_batch_finish+0x2a7/0x320\n  \u003e  __static_call_transform+0xb7/0x220\n  \u003e  arch_static_call_transform+0x5b/0xb0\n  \u003e  __static_call_init+0xe9/0x270\n  \u003e  static_call_module_notify+0x11f/0x150\n  \u003e  notifier_call_chain+0x61/0xe0\n  \u003e  blocking_notifier_call_chain_robust+0x63/0xc0\n  \u003e  load_module+0x1c92/0x20c0\n  \u003e  init_module_from_file+0xd8/0x140\n  \u003e  idempotent_init_module+0x100/0x2f0\n  \u003e  __x64_sys_finit_module+0x71/0xe0\n  \u003e  do_syscall_64+0xe1/0x610\n  \u003e  entry_SYSCALL_64_after_hwframe+0x76/0x7e\n\nwhich matches the following BUG_ON() in alternative.c:\n\n\t/*\n\t * If something went wrong, crash and burn since recovery paths are not\n\t * implemented.\n\t */\n\tBUG_ON(!pages[0] || (cross_page_boundary && !pages[1]));\n\nThis can happen if vmalloc_to_page() fails, for any reason. Such can happen\nif text poking races with CPA, which can possibly result in the collapsing\nof page tables (or breaking of PMD hugepages). It is not a problem for most\nusers of vmalloc_to_page() (they solely own the vmalloc'd range) but, when\nCONFIG_ARCH_HAS_EXECMEM_ROX=y, various modules own a single execmem vmalloc\nrange, and can call set_memory_*() in parallel on it. This can happen to\nrace against __text_poke and cause havoc in vmalloc_to_page().\n\nFix it by excluding against CPA using the init_mm mmap read lock.\n\n[ dhansen: Fix up SoB ordering. The actual code flow here was:\n\t   Pedro=\u003eLorenzo=\u003eMike=\u003eMe which is reflected in the SoB chain\n\t   now. I *believe* Mike simply picked up Lorenzo's update to\n\t   Pedro's post from the Link ]","modified":"2026-09-26T03:48:41.499075089Z","published":"2026-09-25T10:22:54.188Z","database_specific":{"osv_generated_from":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/97xxx/CVE-2026-97942.json","cna_assigner":"Linux"},"references":[{"type":"PACKAGE","url":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git"},{"type":"WEB","url":"https://git.kernel.org/stable/c/1587d3394e254639cc36516256031334095e6ef3"},{"type":"WEB","url":"https://git.kernel.org/stable/c/281e6f536f2f3f95d91938c5bd7ba9bcb4c1049d"},{"type":"WEB","url":"https://git.kernel.org/stable/c/89c60435b90d32ab5e3a39da3ae73d463d07debe"},{"type":"ADVISORY","url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/97xxx/CVE-2026-97942.json"},{"type":"ADVISORY","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-97942"}],"affected":[{"ranges":[{"type":"GIT","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","events":[{"introduced":"64f6a4e10c05ed527f0f24b7954964255e0d3535"},{"fixed":"281e6f536f2f3f95d91938c5bd7ba9bcb4c1049d"},{"fixed":"89c60435b90d32ab5e3a39da3ae73d463d07debe"},{"fixed":"1587d3394e254639cc36516256031334095e6ef3"}]}],"database_specific":{"source":"https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-97942.json"}},{"package":{"name":"Kernel","ecosystem":"Linux"},"ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"6.15.0"},{"fixed":"6.18.53"}]},{"type":"ECOSYSTEM","events":[{"introduced":"6.19.0"},{"fixed":"7.2.7"}]}],"database_specific":{"source":"https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-97942.json"}}],"schema_version":"1.9.0"}