Linux Kernel Deadlock Vulnerability in PCI Error Recovery on s390

A deadlock vulnerability has been identified in the Linux kernel's handling of PCI error recovery on s390 systems. This issue arises because the PCI error recovery process blocks configuration accesses, leading to a hang in tasks that manage health recovery for certain devices. The deadlock can be reproduced by simultaneously executing PCI error recovery and requesting a configuration dump for the same device, which the current error recovery process does not handle properly.

Impact

The vulnerability can cause tasks to be blocked for an extended period, leading to a hung state in the system's task management. This was observed with tasks being blocked for over 122 seconds.

Reproduction

The vulnerability can be reproduced by injecting a PCI error recovery event for a specific device while simultaneously requesting a configuration dump for that same device. This can be done using the 'zpcictl --reset-fw <BDF>' command to trigger the error recovery, followed by a 'devlink health dump show pci/<BDF> reporter fw_fatal' command to request the configuration dump. The resulting deadlock can be observed in the system's task management logs.

Remediation

Users can apply the latest patches available in the Linux kernel stable tree to address this vulnerability.