CVE-2026-93074 (GCVE-0-2026-93074)
Vulnerability from cvelistv5
Published
2026-09-17 16:10
Modified
2026-10-19 09:05
Severity ?
VLAI Severity ?
EPSS score ?
Summary
In the Linux kernel, the following vulnerability has been resolved:
dax/fsdev: use __va(phys) for kaddr in direct_access
Use __va(phys) instead of virt_addr + linear_offset for the kaddr
return in __fsdev_dax_direct_access(). The previous code added a
device-linear byte offset to virt_addr (which is __va of ranges[0]),
but for multi-range devices with physical gaps between ranges, this
linear arithmetic crosses the gap and produces a wrong kernel virtual
address. Using __va(phys) where phys comes from dax_pgoff_to_phys()
is correct for any range layout because the direct map translates
each physical address independently.
This leaves dev_dax->virt_addr write-only, so remove the field
(suggested by Dave Jiang).
References
Impacted products
{
"containers": {
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"product": "Linux",
"programFiles": [
"drivers/dax/dax-private.h",
"drivers/dax/fsdev.c"
],
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"vendor": "Linux",
"versions": [
{
"lessThan": "7b642bd3d39105eef4d5908970726c5fda291b53",
"status": "affected",
"version": "759455848df0b9ac3acabdbedcdc4a55af67935f",
"versionType": "git"
},
{
"lessThan": "ff7c73fca793bd5c29a15ba735b0886f62f3a840",
"status": "affected",
"version": "759455848df0b9ac3acabdbedcdc4a55af67935f",
"versionType": "git"
}
]
},
{
"defaultStatus": "affected",
"product": "Linux",
"programFiles": [
"drivers/dax/dax-private.h",
"drivers/dax/fsdev.c"
],
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"vendor": "Linux",
"versions": [
{
"status": "affected",
"version": "7.1"
},
{
"lessThan": "7.1",
"status": "unaffected",
"version": "0",
"versionType": "semver"
},
{
"lessThanOrEqual": "7.2.*",
"status": "unaffected",
"version": "7.2.6",
"versionType": "semver"
},
{
"lessThanOrEqual": "*",
"status": "unaffected",
"version": "7.3",
"versionType": "original_commit_for_fix"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*",
"versionEndExcluding": "7.2.6",
"versionStartIncluding": "7.1",
"vulnerable": true
},
{
"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*",
"versionEndExcluding": "7.3",
"versionStartIncluding": "7.1",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "In the Linux kernel, the following vulnerability has been resolved:\n\ndax/fsdev: use __va(phys) for kaddr in direct_access\n\nUse __va(phys) instead of virt_addr + linear_offset for the kaddr\nreturn in __fsdev_dax_direct_access(). The previous code added a\ndevice-linear byte offset to virt_addr (which is __va of ranges[0]),\nbut for multi-range devices with physical gaps between ranges, this\nlinear arithmetic crosses the gap and produces a wrong kernel virtual\naddress. Using __va(phys) where phys comes from dax_pgoff_to_phys()\nis correct for any range layout because the direct map translates\neach physical address independently.\n\nThis leaves dev_dax-\u003evirt_addr write-only, so remove the field\n(suggested by Dave Jiang)."
}
],
"metrics": [
{
"cvssV3_1": {
"baseScore": 7.8,
"baseSeverity": "HIGH",
"vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"version": "3.1"
},
"scenarios": [
{
"lang": "en",
"value": "AV:L - The wrong kaddr is computed in __fsdev_dax_direct_access() from a device pgoff supplied by dax_iomap_pgoff() during local VFS DAX I/O (dax_iomap_rw, dax_zero_iter, dax_fault_iter). No network protocol carries that pgoff.\nAC:L - Once a gapped multi-range fsdev_dax device is bound, any I/O whose pgoff dax_pgoff_to_phys() places in a later range makes virt_addr+offset a deterministic wrong linear-map address; the attacker chooses the file offset that yields that pgoff, with no race.\nPR:L - dax_match_type() never auto-binds DAXDRV_FSDEV_TYPE (admin new_id) and fs_dax_get() is the mount-time holder, but dax_direct_access() then has no capability check, so an unprivileged user with file access on the mounted FS-DAX tree reaches the bad kaddr via read/write/fallocate.\nUI:N - The attacker triggers the bad kaddr solely by their own read, write, fallocate, or page-fault on files they can already open; this is not attacker-supplied media that a victim must mount.\nS:U - dax_copy_from_iter(), dax_copy_to_iter(), and fsdev_write_dax() use the wrong host-kernel __va() pointer; impact stays in the host kernel and does not cross a KVM, IOMMU, or other VM-escape boundary.\nC:H - dax_iomap_iter() calls dax_copy_to_iter() from virt_addr+offset, which for a later range is __va of the physical gap (another dax mapping in the same region, or System RAM), so a file read discloses that memory.\nI:H - dax_copy_from_iter() and fsdev_dax_zero_page_range()\u0027s fsdev_write_dax() write attacker bytes or PAGE_SIZE zeros through the same wrong kaddr into the gap, giving a kernel write into adjacent DAX or RAM.\nA:H - memcpy_flushcache() or copy through a kaddr that lands in an unmapped gap oopses the kernel; even a successful wrong-address write can panic by corrupting adjacent structures."
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-10-19T09:05:10.055Z",
"orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"shortName": "Linux"
},
"references": [
{
"url": "https://git.kernel.org/stable/c/7b642bd3d39105eef4d5908970726c5fda291b53"
},
{
"url": "https://git.kernel.org/stable/c/ff7c73fca793bd5c29a15ba735b0886f62f3a840"
}
],
"title": "dax/fsdev: use __va(phys) for kaddr in direct_access",
"x_generator": {
"engine": "bippy-1.2.0"
}
}
},
"cveMetadata": {
"assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"assignerShortName": "Linux",
"cveId": "CVE-2026-93074",
"datePublished": "2026-09-17T16:10:57.051Z",
"dateReserved": "2026-09-17T15:57:05.661Z",
"dateUpdated": "2026-10-19T09:05:10.055Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
Loading…
Loading…
Sightings
| Author | Source | Type | Date |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or seen somewhere by the user.
- Confirmed: The vulnerability is confirmed from an analyst perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: This vulnerability was exploited and seen by the user reporting the sighting.
- Patched: This vulnerability was successfully patched by the user reporting the sighting.
- Not exploited: This vulnerability was not exploited or seen by the user reporting the sighting.
- Not confirmed: The user expresses doubt about the veracity of the vulnerability.
- Not patched: This vulnerability was not successfully patched by the user reporting the sighting.
Loading…
Loading…