CVE-2026-99746 in Linux
Summary
by MITRE • 11/20/2026
In the Linux kernel, the following vulnerability has been resolved:
ASoC: SOF: Intel: hda-sdw-bpt: Stop active DMA before closing BPT streams
When sdw_enable_stream() fails in intel_ace2x_bpt_send_async() after hda_sdw_bpt_send_async() has already started the TX and RX DMA streams, intel_ace2x_bpt_close_stream() invokes hda_sdw_bpt_close() directly without calling hda_sdw_bpt_wait(). As a result, hda_sdw_bpt_dma_deprepare() tears down the DMA buffers and BDL while the DMA hardware is still running.
Disable any running TX or RX BPT DMA stream at the start of hda_sdw_bpt_close() before unmapping channels and depreparing DMA resources.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 11/20/2026
The identified vulnerability resides within the Linux kernel's Advanced System on Chip (ASoC) Sound Open Firmware subsystem, specifically in the Intel HDA-SoundWire bridge driver code responsible for managing bidirectional port streams. This issue manifests as a race condition and improper resource cleanup sequence during stream closure operations. The core technical flaw occurs when the function sdw_enable_stream fails to activate a SoundWire stream after hda_sdw_bpt_send_async has already initiated Transmission and Reception Direct Memory Access DMA streams. In this specific error path, the subsequent call to intel_ace2x_bpt_close_stream incorrectly invokes hda_sdw_bpt_close directly without first executing hda_sdw_bpt_wait. This omission means that the system proceeds to tear down DMA buffers and Buffer Descriptor Lists while the hardware DMA engines are still actively processing data or holding locks on memory regions, creating a critical synchronization failure in the driver's lifecycle management.
The operational impact of this vulnerability is significant for system stability and security integrity. By allowing DMA teardown operations to proceed while the hardware remains active, the kernel risks accessing freed or invalid memory buffers. This can lead to use-after-free conditions where the CPU attempts to read from or write to memory locations that are no longer mapped or valid due to concurrent DMA activity. Such scenarios frequently result in system crashes, kernel panics, or data corruption within audio processing pipelines. Furthermore, if an attacker can trigger this specific error path through crafted input or timing manipulation, they may exploit the resulting undefined behavior to achieve arbitrary code execution with kernel privileges, effectively bypassing standard memory safety protections inherent in modern operating systems.
From a vulnerability classification perspective, this issue aligns closely with CWE-362, which describes concurrent access using shared resources without adequate locking, and CWE-416, referring to use after free scenarios arising from improper resource management. In the context of the MITRE ATT&CK framework for enterprise security, this flaw relates to techniques involving exploitation of software vulnerabilities during system initialization or shutdown phases, potentially allowing an adversary to escalate privileges by disrupting normal driver operations. The lack of proper synchronization between DMA hardware state and software buffer management represents a classic example of race condition vulnerability in low-level device drivers where timing dependencies are not strictly enforced.
To mitigate this risk, the recommended remediation involves modifying the hda_sdw_bpt_close function to explicitly disable any running Transmission or Reception Bidirectional Port DMA streams at the very beginning of its execution flow. This ensures that hardware activity is halted before the driver proceeds to unmap channels and deprepare DMA resources. By enforcing a strict order of operations where active DMA engines are stopped prior to buffer teardown, the system prevents access to invalid memory regions and eliminates the race condition window. System administrators should apply kernel updates containing this patch immediately to restore proper resource lifecycle management within the ASoC SOF Intel driver stack, thereby maintaining both system stability and security posture against potential exploitation of these synchronization flaws.