{ "dataType": "CVE_RECORD", "dataVersion": "5.2", "cveMetadata": { "cveId": "CVE-2025-39941", "assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "state": "PUBLISHED", "assignerShortName": "Linux", "dateReserved": "2025-04-16T07:20:57.148Z", "datePublished": "2025-10-04T07:31:04.080Z", "dateUpdated": "2026-08-05T12:06:15.944Z" }, "containers": { "cna": { "providerMetadata": { "orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "shortName": "Linux", "dateUpdated": "2026-08-05T12:06:15.944Z" }, "descriptions": [ { "lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nzram: fix slot write race condition\n\nParallel concurrent writes to the same zram index result in leaked\nzsmalloc handles. Schematically we can have something like this:\n\nCPU0 CPU1\nzram_slot_lock()\nzs_free(handle)\nzram_slot_lock()\n\t\t\t\tzram_slot_lock()\n\t\t\t\tzs_free(handle)\n\t\t\t\tzram_slot_lock()\n\ncompress\t\t\tcompress\nhandle = zs_malloc()\t\thandle = zs_malloc()\nzram_slot_lock\nzram_set_handle(handle)\nzram_slot_lock\n\t\t\t\tzram_slot_lock\n\t\t\t\tzram_set_handle(handle)\n\t\t\t\tzram_slot_lock\n\nEither CPU0 or CPU1 zsmalloc handle will leak because zs_free() is done\ntoo early. In fact, we need to reset zram entry right before we set its\nnew handle, all under the same slot lock scope." } ], "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 code is the zram block-device write path (`zram_submit_bio` → `zram_write_page`), reached only through local block I/O — writes to /dev/zramN, a zram-backed filesystem, or zram swap. There is no network-facing consumer of this path.\nAC:L - The attacker controls both sides of the race: two threads issuing overlapping O_DIRECT or sub-page writes to the same 4 KiB index in a loop, with the window widened by the compress/zs_malloc stage and further by `zram_bvec_write_partial()`'s read-modify-write. The reporter reproduced it with an ordinary fio/blktests workload, and no attacker-uncontrollable condition is involved.\nPR:L - No capability check exists on the write path; the attacker only needs write access to a zram-backed device or filesystem, which unprivileged users routinely have on the dominant zram deployments (Android/ChromeOS swap, systemd zram-generator, embedded/automotive zram-backed /tmp, /var, or overlay upper dirs), or via membership in the `disk` group. Real root in the init namespace is not required.\nUI:N - The attacker's own two threads create and win the race entirely on their own. No victim action, mount, or file-open by another user is needed.\nS:U - The corruption is confined to the zram driver's slot table and zsmalloc pool within the kernel's own security authority. No VM, IOMMU, or sandbox boundary is crossed.\nC:H - The flag/handle desync yields two distinct kernel-memory disclosures into userspace-readable pages: a sticky ZRAM_SAME slot holding a real handle fills a whole 4 KiB page with a raw kernel heap pointer, and a sticky ZRAM_HUGE slot holding a small compressed object makes `read_incompressible_page()` `copy_page()` up to ~4 KiB of adjacent kernel memory past the end of the zspage.\nI:H - The race corrupts kernel slot metadata (handle, ZRAM_SAME/ZRAM_HUGE flags, obj_size) and the pool accounting counters, so stored page contents are silently replaced by kernel pointers or out-of-bounds heap data — corrupting file blocks or swapped-out anonymous pages belonging to other, potentially privileged, processes.\nA:H - `zs_obj_read_begin(pool, handle, NULL)` on a desynced ZRAM_HUGE slot whose object spans two zpdescs memcpys into a NULL `local_copy`, producing a kernel oops, alongside `WARN_ON` splats on decompression failure. Additionally, every won race permanently orphans a zsmalloc handle, an unbounded and repeatable leak that exhausts memory on the very device backing system swap." } ] } ], "affected": [ { "product": "Linux", "vendor": "Linux", "defaultStatus": "unaffected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": [ "drivers/block/zram/zram_drv.c" ], "versions": [ { "version": "71268035f5d734ad6373d953298bd5779985497a", "lessThan": "ff750e9f2c4d63854c33967d1646b5e89a9a19a2", "status": "affected", "versionType": "git" }, { "version": "71268035f5d734ad6373d953298bd5779985497a", "lessThan": "ce4be9e4307c5a60701ff6e0cafa74caffdc54ce", "status": "affected", "versionType": "git" } ] }, { "product": "Linux", "vendor": "Linux", "defaultStatus": "affected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": [ "drivers/block/zram/zram_drv.c" ], "versions": [ { "version": "6.14", "status": "affected" }, { "version": "0", "lessThan": "6.14", "status": "unaffected", "versionType": "semver" }, { "version": "6.16.9", "lessThanOrEqual": "6.16.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.17", "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": "6.14", "versionEndExcluding": "6.16.9" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.14", "versionEndExcluding": "6.17" } ] } ] } ], "references": [ { "url": "https://git.kernel.org/stable/c/ff750e9f2c4d63854c33967d1646b5e89a9a19a2" }, { "url": "https://git.kernel.org/stable/c/ce4be9e4307c5a60701ff6e0cafa74caffdc54ce" } ], "title": "zram: fix slot write race condition", "x_generator": { "engine": "bippy-1.2.0" } } } }