{"id":"CVE-2026-80682","aliases":[],"url":"https://o3.security/vulnerability/CVE-2026-80682","summary":"riscv/mm: use physical alignment for vmemmap_start_pfn","details":"In the Linux kernel, the following vulnerability has been resolved:\n\nriscv/mm: use physical alignment for vmemmap_start_pfn\n\nRISC-V computes vmemmap_start_pfn by rounding phys_ram_base down to\nVMEMMAP_ADDR_ALIGN.  That alignment must therefore be expressed in the\nphysical-address domain.\n\nCommit 476849b0fba4 (\"riscv/mm: align vmemmap to maximal folio size\")\nattempted to account for the maximal folio alignment by feeding\nMAX_FOLIO_VMEMMAP_ALIGN directly into VMEMMAP_ADDR_ALIGN.  However,\nMAX_FOLIO_VMEMMAP_ALIGN is measured in bytes of struct page storage,\nwhereas VMEMMAP_ADDR_ALIGN is used to align a physical address.\n\nThe mask-based compound_info encoding requires pfn_to_page(0) to be\nnaturally aligned to MAX_FOLIO_VMEMMAP_ALIGN.  Commit 9f94db4c7eaa\n(\"mm/sparse: check memmap alignment for compound_info_has_mask()\") added a\ncheck for that requirement and exposed the unit mismatch on systems such\nas QEMU virt, where the DRAM base is not aligned to MAX_FOLIO_NR_PAGES *\nPAGE_SIZE.\n\nHere is the log:\n[    0.000000][    C0] ------------[ cut here ]------------\n[    0.000000][    C0] WARNING: mm/sparse.c:365 at sparse_init+0x58a/0x6fe, CPU#0: swapper/0\n[    0.000000][    C0] Modules linked in:\n[    0.000000][    C0] CPU: 0 UID: 0 PID: 0 Comm: swapper Not tainted 7.2.0-rc3-g1d8304bdd65f #2 PREEMPT \n[    0.000000][    C0] Hardware name: riscv-virtio,qemu (DT)\n[    0.000000][    C0] epc : sparse_init+0x58a/0x6fe\n[    0.000000][    C0]  ra : sparse_init+0x58a/0x6fe\n[    0.000000][    C0] epc : ffffffff86851c88 ra : ffffffff86851c88 sp : ffffffff88807a30\n[    0.000000][    C0]  gp : ffffffff8a3bf240 tp : ffffffff88842080 t0 : ff600000ffab6000\n[    0.000000][    C0]  t1 : 000000017fab6000 t2 : 65203a6573726363 s0 : ffffffff88807bc0\n[    0.000000][    C0]  s1 : 000000000e000000 a0 : 0000000000000007 a1 : 0000000000000000\n[    0.000000][    C0]  a2 : 0000000000000002 a3 : ffffffff86851c88 a4 : 0000000000000000\n[    0.000000][    C0]  a5 : ffffffff88843080 a6 : 0000000000000003 a7 : 0000000000000000\n[    0.000000][    C0]  s2 : ff60000000000000 s3 : 0040000000000000 s4 : 0004000000000000\n[    0.000000][    C0]  s5 : ffffffff8a4d92e0 s6 : ff600000ffab55e0 s7 : ffffffff88384d00\n[    0.000000][    C0]  s8 : 0000000000000003 s9 : ffffffff88384cc1 s10: ffffffff88384cc0\n[    0.000000][    C0]  s11: ffffffff8a4daae0 t3 : ffffffff915e8b20 t4 : ffffffff915e8b20\n[    0.000000][    C0]  t5 : ffffffff915e8b20 t6 : ffffffff915e8bc8 ssp : 0000000000000000\n[    0.000000][    C0] status: 0000000200000100 badaddr: ffffffff86851c88 cause: 0000000000000003\n[    0.000000][    C0] [<ffffffff86851c88>] sparse_init+0x58a/0x6fe\n[    0.000000][    C0] [<ffffffff8683d396>] mm_core_init_early+0x116/0x1e30\n[    0.000000][    C0] [<ffffffff86801edc>] start_kernel+0xd2/0x848\n\nConvert MAX_FOLIO_VMEMMAP_ALIGN to the equivalent physical alignment\nbefore using it in VMEMMAP_ADDR_ALIGN.  This keeps the existing\nround_down() logic while making the resulting vmemmap base satisfy the\nmask-alignment requirement.","published":"2026-08-28T06:52:49.498Z","modified":"2026-08-29T03:47:17.064247166Z","cvss":null,"epss":null,"cisaKev":null,"exploitsKnown":null,"affectedPackages":[{"ecosystem":"Linux","name":"Kernel","fixedVersion":"7.1.8"}],"fix":null,"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/11ad86830a78ea4e1942afc73c10e4ca8e69e03f"},{"type":"WEB","url":"https://git.kernel.org/stable/c/45ebe4540817cb540a59e4dc04014ac5dddc9990"},{"type":"ADVISORY","url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/80xxx/CVE-2026-80682.json"},{"type":"ADVISORY","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-80682"}],"provenance":{"sources":["OSV.dev","FIRST.org (EPSS)"],"lastVerified":"2026-08-29T03:47:17.064247166Z"}}