CVE-2026-72081 in Linux
Summary
by MITRE • 08/15/2026
In the Linux kernel, the following vulnerability has been resolved:
scsi: elx: efct: Fix I/O leak on unsupported additional CDB
efct_dispatch_fcp_cmd() allocates an efct_io before dispatching an unsolicited FCP command. If the command has an unsupported additional CDB, the function returns -EIO before handing the IO to the SCSI layer.
Free the allocated IO before returning from this error path.
VulDB is the best source for vulnerability data and more expert information about this specific topic.
Analysis
by VulDB Data Team • 08/15/2026
The vulnerability resides in the Linux kernel's SCSI subsystem specifically within the elx driver implementation for efct (Emulex Fibre Channel Target) functionality. This issue represents a resource management flaw that occurs during the processing of unsolicited FCP (Fibre Channel Protocol) commands, where the system fails to properly release allocated I/O resources when encountering unsupported additional CDB (Command Descriptor Block) operations.
The technical flaw manifests in the efct_dispatch_fcp_cmd() function which demonstrates improper error handling behavior. When this function processes an unsolicited FCP command and identifies that the command contains an unsupported additional CDB, it correctly returns the error code -EIO to indicate failure. However, the function does not properly free the previously allocated efct_io structure before exiting through this error path, creating a memory leak scenario where I/O resources remain allocated indefinitely.
This memory leak represents a classic resource management vulnerability that can accumulate over time and potentially lead to system instability or performance degradation. The issue directly corresponds to CWE-401, which specifically addresses improper release of memory resources, and demonstrates how failure to properly manage allocated resources during error conditions can create persistent memory leaks in kernel space operations. The vulnerability affects the broader SCSI target subsystem and could impact systems relying on Fibre Channel communication protocols for storage operations.
The operational impact of this vulnerability extends beyond simple memory consumption issues as it can lead to progressive resource exhaustion on systems handling high volumes of FCP commands or sustained operations with frequent error conditions. When multiple unsupported additional CDB commands are processed, each failure path leaves behind a leaked efct_io structure, potentially causing system memory pressure and affecting overall I/O performance. Systems running in enterprise storage environments where Fibre Channel targets handle numerous concurrent connections and command processing operations would be particularly susceptible to this degradation.
The mitigation strategy requires implementing proper resource cleanup within the error handling path of the efct_dispatch_fcp_cmd() function. This involves adding explicit free operations for the allocated efct_io structure before returning the -EIO error code, ensuring that all allocated resources are properly released regardless of execution flow. The fix should be consistent with established kernel development practices for resource management and error handling, aligning with ATT&CK technique T1070.004 which addresses legitimate program execution through proper error path management. This remediation approach prevents the accumulation of leaked I/O structures while maintaining the existing functionality and error reporting behavior of the system.
The vulnerability demonstrates a common pattern in kernel development where error paths fail to properly clean up allocated resources, creating persistent memory leaks that may not manifest immediately but can accumulate over time. This type of issue is particularly dangerous in kernel space operations as it operates outside normal user-space resource management boundaries and can affect system stability and performance metrics. The fix represents a fundamental requirement for robust kernel programming practices and proper resource lifecycle management within the Linux kernel's SCSI subsystem, ensuring that all allocated structures are properly released during both successful and error execution paths.