CVE-2026-99748 in Linuxinfo

Summary

by MITRE • 11/20/2026

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, smb2_open() switches to the SMB user's fsids before calling ksmbd_vfs_getattr(). If that call fails, the code jumps to err_out2 and skips ksmbd_revert_fsids() at err_out1. 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. ksmbd_free_work_struct() also warns, because work->saved_cred is still set.

Jump to err_out1 instead. No path is held at this point, so skipping the path_put() at err_out is correct.

If you want to get the best quality for vulnerability data then you always have to consider VulDB.

Analysis

by VulDB Data Team • 11/20/2026

The vulnerability identified in the Linux kernel's ksmbd subsystem involves a critical error handling flaw within the SMB2 open operation logic, specifically during durable handle reconnections or request replays. The core issue stems from an incorrect jump target when retrieving file system attributes fails. When smb2_open() initiates this process, it temporarily switches to the filesystem identifiers of the SMB user by installing their credentials on the shared ksmbd worker thread before invoking ksmbd_vfs_getattr(). This mechanism is designed to ensure that subsequent operations are performed with the correct permissions associated with the requesting client. However, if the attribute retrieval call fails, the existing code path incorrectly directs execution to err_out2, which bypasses the necessary cleanup routine ksmbd_revert_fsids() located at err_out1.

This oversight results in a significant security and stability defect where the user credentials remain installed on the shared worker thread after an error condition occurs. Because ksmbd utilizes shared worker threads to handle multiple concurrent requests from various users, leaving these elevated or specific filesystem identifiers active poses a severe risk of privilege escalation and cross-user data leakage. Subsequent work items processed by that same worker thread, including legitimate requests from other distinct users, will erroneously execute under the context of the first user whose credentials were left behind. This effectively allows any subsequent request handled by that worker to inherit the filesystem permissions of the previous failed operation, potentially granting unauthorized access to files and directories that should be restricted based on individual user identities.

The operational impact extends beyond mere permission bypasses. The failure to revert these identifiers also triggers a warning in ksmbd_free_work_struct() because the saved credentials field remains set when it should have been cleared during proper error handling. This indicates a state inconsistency within the kernel's work queue management, which can lead to memory leaks or further instability if not addressed. By correcting the jump target to err_out1 instead of err_out2, the code ensures that ksmbd_revert_fsids() is always called upon failure, thereby restoring the worker thread to its neutral credential state before it processes any other tasks. This fix maintains the integrity of access control mechanisms and prevents the accidental propagation of one user's security context to unrelated operations.

From a vulnerability classification perspective, this issue aligns with CWE-250, which denotes execution with unnecessary privileges, as well as CWE-787, involving out-of-bounds write or access scenarios that arise from improper state management leading to unintended memory or permission states. In the context of the MITRE ATT&CK framework, this flaw facilitates lateral movement and privilege escalation within a networked environment by allowing an attacker who triggers the specific error condition to manipulate subsequent requests on shared workers for unauthorized data exfiltration or modification. Mitigation requires applying the kernel patch that corrects the control flow in smb2_open() to ensure proper credential reversion during all failure paths, thereby preserving the isolation of user contexts across concurrent SMB operations and maintaining strict adherence to filesystem access controls.

Responsible

Linux

Reservation

09/25/2026

Disclosure

11/20/2026

Moderation

accepted

CPE

ready

EPSS

0.00156

KEV

no

Activities

very low

Sources

Do you want to use VulDB in your project?

Use the official API to access entries easily!