Control: tags -1 + moreinfo Hi,
> Mar 09 21:25:15 sonic kernel: <TASK> > Mar 09 21:25:15 sonic kernel: ? __die_body.cold+0x1a/0x1f > Mar 09 21:25:15 sonic kernel: ? page_fault_oops+0xd2/0x2b0 > Mar 09 21:25:15 sonic kernel: ? srso_alias_return_thunk+0x5/0x7f > Mar 09 21:25:15 sonic kernel: ? search_bpf_extables+0x5b/0x80 > Mar 09 21:25:15 sonic kernel: ? exc_page_fault+0xca/0x170 > Mar 09 21:25:15 sonic kernel: ? asm_exc_page_fault+0x22/0x30 > Mar 09 21:25:15 sonic kernel: ? __vunmap_range_noflush+0x211/0x470 > Mar 09 21:25:15 sonic kernel: ? psi_group_change+0x152/0x370 > Mar 09 21:25:15 sonic kernel: remove_vm_area+0x91/0xb0 > Mar 09 21:25:15 sonic kernel: __vunmap+0xb8/0x340 > Mar 09 21:25:15 sonic kernel: free_work+0x21/0x30 > Mar 09 21:25:15 sonic kernel: process_one_work+0x1c4/0x380 > Mar 09 21:25:15 sonic kernel: worker_thread+0x4d/0x380 > Mar 09 21:25:15 sonic kernel: ? rescuer_thread+0x3a0/0x3a0 > Mar 09 21:25:15 sonic kernel: kthread+0xd7/0x100 > Mar 09 21:25:15 sonic kernel: ? kthread_complete_and_exit+0x20/0x20 > Mar 09 21:25:15 sonic kernel: ret_from_fork+0x1f/0x30 > Mar 09 21:25:15 sonic kernel: </TASK> This stack does not seem to be directly related to zfs, but a normal page free routine. Maybe you can do some comparison (e.g. with or w/o zfs module installed) and see if this could be consistently reproduced. Thanks, Shengqi Chen
