{"dataType":"CVE_RECORD","dataVersion":"5.2","cveMetadata":{"cveId":"CVE-2026-80656","assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","state":"PUBLISHED","assignerShortName":"Linux","dateReserved":"2026-08-26T14:34:25.780Z","datePublished":"2026-08-28T06:49:02.596Z","dateUpdated":"2026-08-29T06:21:45.266Z"},"containers":{"cna":{"providerMetadata":{"orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux","dateUpdated":"2026-08-29T06:21:45.266Z"},"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nhfsplus: Add a sanity check for btree node size\n\nSyzbot reported an uninit-value bug in [1] with a corrupted HFS+ image,\nduring the file system mounting process, specifically while loading the\ncatalog, a corrupted node_size value of 1 caused the rec_off argument\npassed to hfs_bnode_read_u16() (within hfs_bnode_find()) to be excessively\nlarge. Consequently, the function failed to return a valid value to\ninitialize the off variable, triggering the bug [1].\n\nEvery node starts from BTree node descriptor: struct hfs_bnode_desc.\nSo, the size of node cannot be lesser than that. However, technical\nspecification declares that: \"The node size (which is expressed in bytes)\nmust be power of two, from 512 through 32,768, inclusive.\" Add a check\nfor btree node size base on technical specification.\n\n[1]\nBUG: KMSAN: uninit-value in hfsplus_bnode_find+0x141c/0x1600 fs/hfsplus/bnode.c:584\n hfsplus_bnode_find+0x141c/0x1600 fs/hfsplus/bnode.c:584\n hfsplus_btree_open+0x169a/0x1e40 fs/hfsplus/btree.c:382\n hfsplus_fill_super+0x111f/0x2770 fs/hfsplus/super.c:553\n get_tree_bdev_flags+0x6e6/0x920 fs/super.c:1694\n get_tree_bdev+0x38/0x50 fs/super.c:1717\n hfsplus_get_tree+0x35/0x40 fs/hfsplus/super.c:709\n vfs_get_tree+0xb3/0x5d0 fs/super.c:1754\n fc_mount fs/namespace.c:1193 [inline]"}],"metrics":[{"cvssV3_1":{"version":"3.1","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","baseScore":7.8,"baseSeverity":"HIGH"},"scenarios":[{"lang":"en","value":"AV:L - The vulnerable hfs_btree_open→hfs_bnode_find path is reached only via local mount(2)/vfs_get_tree on a block/loop/USB device (hfsplus_fill_super); there is no network-facing consumer of hfsplus btree parsing.\nAC:L - A crafted on-disk btree header node_size of 1 is a valid power-of-two and deterministically makes rec_off negative (wrapped to a huge u32), so hfs_bnode_read_u16 skips filling its stack buffer and KMSAN reports uninit-value at mount without races or external preconditions.\nPR:L - hfsplus is FS_REQUIRES_DEV without FS_USERNS_MOUNT; the highest-impact scenario is an unprivileged local user supplying an HFS+ USB/SD image that udisks2 or a desktop automounter mounts, or a loop image attached through session polkit helpers—no real root in init_ns is needed by the attacker.\nUI:N - The attacker inserts removable media or supplies the image that gets mounted; mount-time catalog/extent btree loading fires automatically during fill_super with no separate victim action beyond routine automount behavior.\nS:U - Uninitialized-value consumption and any mount-time parser fault stay within the host kernel parsing the attacker's filesystem; no VM escape, IOMMU bypass, or sandbox boundary is crossed.\nC:H - Before the fix, invalid node_size lets hfs_bnode_read return without initializing the hfs_bnode_read_u16 buffer; the resulting garbage off value drives subsequent btree record-offset parsing and bounded node reads that consume uninitialized stack bytes—a memory-safety defect class treated as arbitrary read potential per kernel CVSS guidance.\nI:H - Attacker-controlled sub-512 power-of-two node_size corrupts btree rec_off arithmetic; uninitialized or crafted record offsets can pass hfs_bnode_find checks and seed inconsistent in-memory btree nodes whose corrupted offsets propagate into further mount-time bmap and page-cache accesses.\nA:H - Syzbot/KMSAN builds panic on the uninit-value at hfsplus_bnode_find; even on production kernels a crafted image causes mount failure after kernel error logging, and repeated automount of malicious removable media can deny access to the volume and waste mount resources."}]}],"affected":[{"product":"Linux","vendor":"Linux","defaultStatus":"unaffected","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["fs/hfsplus/btree.c","include/linux/hfs_common.h"],"versions":[{"version":"8ad2c6a36ac4328072377906a47ea0bff11e4032","lessThan":"306265eb9384d2c224f1b14d83a29323f0eab6d1","status":"affected","versionType":"git"},{"version":"8ad2c6a36ac4328072377906a47ea0bff11e4032","lessThan":"3f95e2661574ff13f099dd13456751933c280628","status":"affected","versionType":"git"}]},{"product":"Linux","vendor":"Linux","defaultStatus":"affected","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","programFiles":["fs/hfsplus/btree.c","include/linux/hfs_common.h"],"versions":[{"version":"7.1","status":"affected"},{"version":"0","lessThan":"7.1","status":"unaffected","versionType":"semver"},{"version":"7.1.5","lessThanOrEqual":"7.1.*","status":"unaffected","versionType":"semver"},{"version":"7.2","lessThanOrEqual":"*","status":"unaffected","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"operator":"OR","negate":false,"cpeMatch":[{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"7.1","versionEndExcluding":"7.1.5"},{"vulnerable":true,"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionStartIncluding":"7.1","versionEndExcluding":"7.2"}]}]}],"references":[{"url":"https://git.kernel.org/stable/c/306265eb9384d2c224f1b14d83a29323f0eab6d1"},{"url":"https://git.kernel.org/stable/c/3f95e2661574ff13f099dd13456751933c280628"}],"title":"hfsplus: Add a sanity check for btree node size","x_generator":{"engine":"bippy-1.2.0"}}}}