Hi, I am getting stuff freezing randomly since 6.14-rc. I do not have a clear way to reproduce this, but it tends to happen when I delete stuff on /tmp (tmpfs) using KDE's dolphin. I usually loose the ability to switch vtys or launch new terminals, so by chance I had dmesg running and got the call trace below when things started to go south. At this point dolphin was defunct state, and was consuming 100% CPU. 'perf top' was showing xas_find, __lock_acquire, lock_release and sched_clock_noinstr with highest percentage. Since I also have current bcachefs merged in my kernel I already checked with Kent, but nothing seems to point to bcachefs. Also my other (non-desktop) systems running the same kernel do not show this behavior. Hopefully someone in here can make sense of this ... [19136.543931] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks: [19136.543938] rcu: Tasks blocked on level-1 rcu_node (CPUs 16-31): P314703 [19136.543943] rcu: (detected by 29, t=21002 jiffies, g=8366533, q=26504 ncpus=32) [19136.543946] task:KIO::WorkerThre state:R running task stack:0 pid:314703 tgid:314656 ppid:3984 task_flags:0x400040 flags:0x00004006 [19136.543951] Call Trace: [19136.543953] <TASK> [19136.543958] __schedule+0x784/0x1520 [19136.543963] ? __schedule+0x784/0x1520 [19136.543966] ? __schedule+0x784/0x1520 [19136.543969] preempt_schedule_irq+0x52/0x90 [19136.543972] raw_irqentry_exit_cond_resched+0x2f/0x40 [19136.543975] irqentry_exit+0x3e/0x50 [19136.543977] irqentry_exit+0x3e/0x50 [19136.543979] ? sysvec_apic_timer_interrupt+0x48/0x90 [19136.543982] ? asm_sysvec_apic_timer_interrupt+0x1f/0x30 [19136.543985] ? local_clock_noinstr+0x10/0xc0 [19136.543987] ? local_clock+0x14/0x30 [19136.543990] ? __lock_acquire+0x1fd/0x6c0 [19136.543995] ? local_clock+0x14/0x30 [19136.543997] ? lock_release+0x120/0x470 [19136.544000] ? find_get_entries+0x76/0x2e0 [19136.544004] ? find_get_entries+0xfb/0x2e0 [19136.544006] ? find_get_entries+0x76/0x2e0 [19136.544011] ? shmem_undo_range+0x35f/0x520 [19136.544027] ? shmem_evict_inode+0x135/0x290 [19136.544029] ? lock_release+0x120/0x470 [19136.544031] ? evict+0x1a8/0x340 [19136.544036] ? evict+0x1c5/0x340 [19136.544038] ? lock_release+0x120/0x470 [19136.544040] ? iput+0x1f2/0x290 [19136.544044] ? iput+0x1fa/0x290 [19136.544047] ? do_unlinkat+0x1e7/0x2c0 [19136.544051] ? __x64_sys_unlink+0x20/0x30 [19136.544053] ? x64_sys_call+0xa9a/0x20a0 [19136.544055] ? do_syscall_64+0x58/0xf0 [19136.544057] ? entry_SYSCALL_64_after_hwframe+0x50/0x58 [19136.544063] </TASK> [19157.614008] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks: [19157.614014] rcu: Tasks blocked on level-1 rcu_node (CPUs 16-31): P314703 [19157.614019] rcu: (detected by 8, t=21002 jiffies, g=8366553, q=40992 ncpus=32) [19157.614022] task:KIO::WorkerThre state:R running task stack:0 pid:314703 tgid:314656 ppid:3984 task_flags:0x400040 flags:0x00004006 [19157.614027] Call Trace: [19157.614029] <TASK> [19157.614033] __schedule+0x784/0x1520 [19157.614041] ? irqentry_exit+0x3e/0x50 [19157.614043] ? sysvec_apic_timer_interrupt+0x48/0x90 [19157.614046] ? asm_sysvec_apic_timer_interrupt+0x1f/0x30 [19157.614048] ? find_get_entries+0x76/0x2e0 [19157.614052] ? local_clock_noinstr+0x10/0xc0 [19157.614055] ? local_clock+0x14/0x30 [19157.614057] ? __lock_acquire+0x1fd/0x6c0 [19157.614061] ? local_clock_noinstr+0x10/0xc0 [19157.614063] ? local_clock+0x14/0x30 [19157.614065] ? lock_release+0x120/0x470 [19157.614068] ? find_get_entries+0x76/0x2e0 [19157.614071] ? find_get_entries+0x76/0x2e0 [19157.614073] ? find_get_entries+0xfb/0x2e0 [19157.614075] ? find_get_entries+0x76/0x2e0 [19157.614080] ? shmem_undo_range+0x35f/0x520 [19157.614096] ? shmem_evict_inode+0x135/0x290 [19157.614098] ? lock_release+0x120/0x470 [19157.614100] ? evict+0x1a8/0x340 [19157.614105] ? evict+0x1c5/0x340 [19157.614107] ? lock_release+0x120/0x470 [19157.614109] ? iput+0x1f2/0x290 [19157.614113] ? iput+0x1fa/0x290 [19157.614115] ? do_unlinkat+0x1e7/0x2c0 [19157.614120] ? __x64_sys_unlink+0x20/0x30 [19157.614122] ? x64_sys_call+0xa9a/0x20a0 [19157.614124] ? do_syscall_64+0x58/0xf0 [19157.614127] ? entry_SYSCALL_64_after_hwframe+0x50/0x58 [19157.614132] </TASK> Kind regards Malte
