The Linux kernel's md/raid10 component contained a logic error where the 'still_degraded' flag was inverted after being converted from an integer to a boolean in the raid10_sync_request() function. This inversion caused the recovery process to incorrectly clear bitmap bits needed by a still-missing device. Consequently, when the missing device was re-added, its recovery skipped all regions written during the degraded state, leading to silent data corruption as the array was incorrectly marked as in-sync. The issue was identified by comparing it to the correct logic in the raid1 implementation and has been resolved by restoring the correct boolean assignment.

The vulnerability can cause silent data corruption in RAID10 arrays during device recovery. Specifically, data written while the array was degraded may be skipped during recovery, resulting in stale data being considered valid. This undermines data integrity and can lead to undetected data loss or corruption. There is no indication of active exploitation in the wild.

A fix has been implemented to correct the boolean assignment of the 'still_degraded' flag in the Linux kernel's md/raid10 subsystem. Users should apply the official kernel update that includes this fix to prevent silent data corruption during RAID10 recovery. Patch status is not explicitly stated in the input, so users should consult the vendor or kernel mailing lists for the specific fixed kernel versions and apply updates accordingly.