$linuxjunkies
>

soft lockup

also: softlockup

A kernel warning that occurs when a CPU is stuck running kernel code for too long without yielding, often indicating a performance problem or infinite loop rather than a hardware failure.

A soft lockup is a kernel detection mechanism that triggers when a CPU spends excessive time executing kernel code without rescheduling tasks or allowing other processes to run. The kernel uses a watchdog timer to detect this condition and prints a warning message, but does not force a reboot (unlike a hard lockup).

When a soft lockup is detected, you'll typically see kernel log messages like BUG: soft lockup - CPU#0 stuck for 22s!. This usually indicates a performance bottleneck, resource contention, or potentially buggy kernel code that's consuming CPU in an inefficient way.

Common causes include: high interrupt load, excessive system calls, or poorly written device drivers. Unlike a hard lockup (complete system freeze), a soft lockup warning allows the system to continue running and be investigated. You can check kernel logs with dmesg or journalctl to diagnose the problematic code.

Related terms