On Thu, Mar 10, 2022 at 03:21:15PM +1100, Ryan Tierney wrote: > Hi Linux kernel team, > > We have had multiple servers lock up over the course of a few months on > kernel version 5.10.92 > Looking to find out any information to help mitigate this. > > I've also found a related report here back in 2021: https://lore.kernel.org/all/757b684a-67b5-999b-7f2d-b55fb1c61fd8@xxxxxxxxxx/T/ > > Please see the stack trace obtained during this lockup > > [2136172.975892] ------------[ cut here ]------------ > [2136172.975896] kernel BUG at include\/linux\/swapops.h:204! > [2136172.981268] invalid opcode: 0000 [#1] SMP NOPTI > [2136172.985983] CPU: 49 PID: 1672949 Comm: apt-cache Not tainted 5.10.0-11-amd64 #1 Debian 5.10.92-1 > [2136172.994944] Hardware name: Supermicro AS -1114S-WN10RT\/H12SSW-NTR, BIOS 2.3 10\/28\/2021 > [2136173.003042] RIP: e030:__migration_entry_wait+0xf9\/0x100 > [2136173.008444] Code: 0f 45 c2 41 8b 40 34 85 c0 74 9f 8d 50 01 f0 41 0f b1 50 34 75 f1 48 89 ef e8 03 e0 e4 ff 66 90 5b 4c 89 c7 5d e9 27 48 f7 ff <0f> 0b 0f 1f 44 00 00 0f 1f 44 00 00 49 89 f9 48 8b 3e e8 08 39 d8 > [2136173.027394] RSP: e02b:ffffc90075a4fdb8 EFLAGS: 00010246 > [2136173.032795] RAX: 000fffffc0000000 RBX: ffff88807b0e6968 RCX: ffffea00006a6f87 > [2136173.040114] RDX: 0000000000000000 RSI: ffff88807b0e6968 RDI: fffffffffcac8400 > [2136173.047429] RBP: ffffea0001ec39a8 R08: ffffea00006a6f40 R09: ffff88804679f2c0 > [2136173.054745] R10: 000ffffffffff000 R11: 0000000000000000 R12: ffffc90075a4fe40 > [2136173.062060] R13: 7c0000000001a9bd R14: fff0000000000fff R15: 0000000000000000 > [2136173.069378] FS: 00007f2cc7eee800(0000) GS:ffff888117440000(0000) knlGS:0000000000000000 > [2136173.077651] CS: e030 DS: 0000 ES: 0000 CR0: 0000000080050033 > [2136173.083575] CR2: 00007f2cc4b2df8c CR3: 000000008d396000 CR4: 0000000000050660 > [2136173.090893] Call Trace: > [2136173.093527] do_swap_page+0x66f\/0x900 > [2136173.097370] handle_mm_fault+0xd7d\/0x1bf0 > [2136173.101584] ? xfs_file_read_iter+0x6e\/0xd0 [xfs] > [2136173.106468] do_user_addr_fault+0x1b8\/0x3f0 > [2136173.110835] exc_page_fault+0x78\/0x160 > [2136173.114766] ? asm_exc_page_fault+0x8\/0x30 > [2136173.119038] asm_exc_page_fault+0x1e\/0x30 > [2136173.123224] RIP: e033:0x7f2cc8930fcb > [2136173.126977] Code: 8b 46 08 48 8d 14 80 48 8d 04 50 48 8d 44 85 00 48 39 c3 74 5a 8b 43 10 48 8d 14 80 48 8d 04 50 48 8d 5c 85 00 48 39 eb 74 45 <8b> 33 4c 01 c6 0f b7 46 fe 49 39 c4 75 c7 4c 89 e2 4c 89 44 24 08 > [2136173.145919] RSP: e02b:00007ffd2a3f3e80 EFLAGS: 00010206 > [2136173.151323] RAX: 0000000000000005 RBX: 00007f2cc4b2df8c RCX: 0000000000000006 > [2136173.158632] RDX: 00007f2cc351b2e6 RSI: 00007ffd2a3f3fd0 RDI: 00007f2cc351b2e6 > [2136173.165948] RBP: 00007f2cc339b000 R08: 00007f2cc339b000 R09: 0000000000000de0 > [2136173.173262] R10: 00007f2cc4c8bc7c R11: 00000000206e6f68 R12: 0000000000000005 > [2136173.180580] R13: 00007ffd2a3f3f20 R14: 00007f2cc4b0d410 R15: 000055ab3740af08 > [2136173.187892] Modules linked in: xen_acpi_processor xen_gntdev xen_evtchn binfmt_misc xenfs xen_privcmd nls_ascii nls_cp437 vfat fat ghash_clmulni_intel aesni_intel libaes crypto_simd cryptd glue_helper wmi_bmof efi_pstore pcspkr ast drm_vram_helper drm_ttm_helper ttm ccp drm_kms_helper rng_core rndis_host cdc_ether cec usbnet i2c_algo_bit joydev mii k10temp sp5100_tco watchdog ipmi_ssif evdev bridge acpi_ipmi ipmi_si 8021q ipmi_devintf garp ipmi_msghandler stp mrp llc button bonding loop psmouse uhci_hcd ohci_hcd ehci_hcd drbd lru_cache drm fuse configfs efivarfs ip_tables x_tables autofs4 xfs raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor hid_generic usbhid hid raid6_pq libcrc32c crc32c_generic raid0 multipath linear raid1 raid10 md_mod crc32_pclmul crc32c_intel ahci libahci xhci_pci nvme xhci_hcd libata nvme_core i40e t10_pi usbcore bnxt_en scsi_mod crc_t10dif crct10dif_generic crct10dif_pclmul crct10dif_common ptp usb_common i2c_piix4 pps_core wmi > [2136173.274260] ---[ end trace 37e8c1af7f6e782a ]--- > This looks like a filesystem issue, have you tried asking on the linux-fsdevel mailing list? And does it also happen on 5.16? thanks, greg k-h
