CVE-2026-99748: ksmbd: revert fsids on durable reconnect and replay failure
In the Linux kernel, the following vulnerability has been resolved:
ksmbd: revert fsids on durable reconnect and replay failure
On a durable handle reconnect or a request replay, smb2open() switches to the SMB user's fsids before calling ksmbdvfsgetattr(). If that call fails, the code jumps to errout2 and skips ksmbdrevertfsids() at errout1. The user's credentials then stay installed on the shared ksmbd worker. Later work on that worker, including other users' requests, runs with the first user's fsuid, fsgid and supplementary groups. ksmbdfreeworkstruct() also warns, because work->savedcred is still set.
Jump to errout1 instead. No path is held at this point, so skipping the pathput() at errout is correct.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.7 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5 - Upgrade
Upgrade
debian/linuxto a version that resolves this vulnerability.Fixed in 6.1.176-1Fixed in 6.1.192-2Fixed in 7.3.5-1
Event History
Frequently Asked Questions
Who is exposed to this issue?
Systems using the Linux kernel's ksmbd SMB server are exposed on the affected durable-handle reconnect and request-replay error paths. The risk is greatest where a shared ksmbd worker subsequently processes requests for different SMB users.
What must occur for the credential leak to be triggered?
An SMB durable handle reconnect or request replay must cause smb2_open() to switch to an SMB user's fsids, followed by a failure from ksmbd_vfs_getattr(). The failure path then leaves that user's fsuid, fsgid, and supplementary groups installed on the shared worker.
What is the security impact after the failure?
Later work handled by the same ksmbd worker, including requests from other users, can run under the first user's credentials. The affected path also leaves work->saved_cred set, which causes a warning when ksmbd_free_work_struct() runs.
How can administrators tell whether this condition has occurred?
A warning from ksmbd_free_work_struct() indicating that work->saved_cred remains set is an indicator of the affected failure path. The issue requires a ksmbd_vfs_getattr() failure during durable reconnect or replay processing.