Description of problem: Jun 22 15:02:07 fangorn kernel: BUG: soft lockup - CPU#1 stuck for 61s! [mcelog:5421] Jun 22 15:02:07 fangorn kernel: Modules linked in: ipv6 cpufreq_ondemand powernow_k8 freq_table dm_multipath uinput cfi_cmdset_0002 cfi_util jedec_probe cfi_probe gen_probe ck804xrom mtd forcedeth chipreg i2c_nforce2 map_funcs pata_amd pcspkr i2c_core ata_generic pata_acpi sata_nv raid456 async_xor async_memcpy async_tx xor raid1 [last unloaded: scsi_wait_scan] Jun 22 15:02:07 fangorn kernel: CPU 1: Jun 22 15:02:07 fangorn kernel: Modules linked in: ipv6 cpufreq_ondemand powernow_k8 freq_table dm_multipath uinput cfi_cmdset_0002 cfi_util jedec_probe cfi_probe gen_probe ck804xrom mtd forcedeth chipreg i2c_nforce2 map_funcs pata_amd pcspkr i2c_core ata_generic pata_acpi sata_nv raid456 async_xor async_memcpy async_tx xor raid1 [last unloaded: scsi_wait_scan] Jun 22 15:02:07 fangorn kernel: Pid: 5421, comm: mcelog Not tainted 2.6.27.24-170.2.68.fc10.x86_64 #1 empty Jun 22 15:02:07 fangorn kernel: RIP: 0010:[<ffffffff810625cc>] [<ffffffff810625cc>] smp_call_function_mask+0x170/0x1dd Jun 22 15:02:07 fangorn kernel: RSP: 0018:ffff880117e51d88 EFLAGS: 00000202 Jun 22 15:02:07 fangorn kernel: RAX: ffff880117e51e48 RBX: ffff880117e51e68 RCX: 00000000000000fc Jun 22 15:02:07 fangorn kernel: RDX: ffffffffffffffff RSI: 00000000000008fc RDI: 0000000000000292 Jun 22 15:02:07 fangorn kernel: RBP: 0000000000000001 R08: ffff880117e50000 R09: ffff88003a34b9c0 Jun 22 15:02:07 fangorn kernel: R10: 0000000000000000 R11: 0000006017e51e80 R12: ffff88002803b5a0 Jun 22 15:02:07 fangorn kernel: R13: ffff8800a695d000 R14: ffff880117e50000 R15: ffffffff816df990 Jun 22 15:02:07 fangorn kernel: FS: 00007f8d89f526f0(0000) GS:ffff880127c04980(0000) knlGS:0000000000000000 Jun 22 15:02:07 fangorn kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b Jun 22 15:02:07 fangorn kernel: CR2: 00000031eb283070 CR3: 000000010d0e1000 CR4: 00000000000006e0 Jun 22 15:02:07 fangorn kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 Jun 22 15:02:07 fangorn kernel: DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Jun 22 15:02:07 fangorn kernel: Jun 22 15:02:07 fangorn kernel: Call Trace: Jun 22 15:02:07 fangorn kernel: [<ffffffff810625c4>] ? smp_call_function_mask+0x168/0x1dd Jun 22 15:02:07 fangorn kernel: [<ffffffff8101a876>] ? collect_tscs+0x0/0x33 Jun 22 15:02:07 fangorn kernel: [<ffffffff810c2e6b>] ? chrdev_open+0x0/0x169 Jun 22 15:02:07 fangorn kernel: [<ffffffff81333bf2>] ? _spin_lock_irq+0x21/0x26 Jun 22 15:02:07 fangorn kernel: [<ffffffff813322b3>] ? wait_for_common+0x15a/0x174 Jun 22 15:02:07 fangorn kernel: [<ffffffff8103a69b>] ? default_wake_function+0x0/0xf Jun 22 15:02:07 fangorn kernel: [<ffffffff810ca2ef>] ? do_filp_open+0x3e4/0x7db Jun 22 15:02:07 fangorn kernel: [<ffffffff8101a876>] ? collect_tscs+0x0/0x33 Jun 22 15:02:07 fangorn kernel: [<ffffffff81062654>] ? smp_call_function+0x1b/0x1d Jun 22 15:02:07 fangorn kernel: [<ffffffff81046663>] ? on_each_cpu+0x18/0x46 Jun 22 15:02:07 fangorn kernel: [<ffffffff8101ab3b>] ? mce_read+0x127/0x1b9 Jun 22 15:02:07 fangorn kernel: [<ffffffff810c0f04>] ? vfs_read+0xa8/0x102 Jun 22 15:02:07 fangorn kernel: [<ffffffff810c1022>] ? sys_read+0x47/0x6e Jun 22 15:02:07 fangorn kernel: [<ffffffff8101027a>] ? system_call_fastpath+0x16/0x1b Jun 22 15:02:07 fangorn kernel: Version-Release number of selected component (if applicable): Linux fangorn.howardsilvan.com 2.6.27.24-170.2.68.fc10.x86_64 #1 SMP Wed May 20 22:47:23 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux How reproducible: It operates as a heavy-use mailserver (postfix, dovecot, ASSP)
After a few days, maybe a week, other lockups/crashes would yet occur with the "noapic" setting. So apparently "noapic" only made the problem occur less-frequently, but it was not a true workaround. The fix was to review all CMOS/BIOS settings, especially those under ACPI. I enabled "ACPI 3.0" instead of "ACPI 2.0". I disabled NMI. There were a few other changes that I can't recall from memory. After making these CMOS/BIOS setting changes then I haven't had another problem.