On 16.08.21 г. 13:20, David Sterba wrote:

On Fri, Aug 13, 2021 at 10:33:05PM -0600, Chris Murphy wrote:

quoted

I get the following call trace about 0.6s after dnf completes an
update which I imagine deletes many files. I'll try to reproduce and
get /proc/lock_stat

[   95.674507] kernel: BUG: MAX_LOCKDEP_CHAIN_HLOCKS too low!

The message "BUG: MAX_LOCKDEP_CHAIN_HLOCKS too low" is related to
lockdep and not a btrfs problem, but it appears from time to time and as
Johannes said either increase the config variable, or ignore it.

But is not a bug if code triggers it? I.e I think it's a signal of
misbehaving code. CC'ed PeterZ who can hopefully shed some light on this.