I have done the following
Steps to reproduce
No exact steps, occurs under moderate to heavy workload after several hours (this case was a gitlab runner node).
We've seen this signature in several issues that we closed with the 0.8.0.0 stability fixes. Those fixes apparently did not address this selinux_inode_free_security failure mode, so let track this problem down and eliminate it.
Current behavior
Container runs fine for several hours, then hangs.
Expected behavior
Container doesn't panic.
Environment
- OS: 26
- Xcode: N/A
- Container: 0.8.0
Relevant log output
[124676.642940] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
[124676.642989] Mem abort info:
[124676.643006] ESR = 0x0000000096000046
[124676.643025] EC = 0x25: DABT (current EL), IL = 32 bits
[124676.643047] SET = 0, FnV = 0
[124676.643064] EA = 0, S1PTW = 0
[124676.643076] FSC = 0x06: level 2 translation fault
[124676.643094] Data abort info:
[124676.643110] ISV = 0, ISS = 0x00000046, ISS2 = 0x00000000
[124676.643126] CM = 0, WnR = 1, TnD = 0, TagAccess = 0
[124676.643146] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[124676.643164] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000109cf9000
[124676.643188] [0000000000000008] pgd=08000001058df003, p4d=08000001058df003, pud=0800000109d71003, pmd=0000000000000000
[124676.643227] Internal error: Oops: 0000000096000046 [#1] SMP
[124676.643255] CPU: 4 UID: 0 PID: 207 Comm: xxx Not tainted 6.12.28 #1
[124676.643282] pstate: a1400005 (NzCv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--)
[124676.643310] pc : selinux_inode_free_security+0x78/0xbc
[124676.644336] lr : security_inode_free+0x58/0x74
[124676.644356] sp : ffff800080c33a80
[124676.644371] x29: ffff800080c33a80 x28: ffff008081c98000 x27: 0000000000000000
[124676.644406] x26: 0000000000000000 x25: 0000000000000000 x24: ffffdce02a75f0f0
[124676.644437] x23: 0000000000000001 x22: ffff800080c33cb8 x21: ffff0080afd22130
[124676.644470] x20: ffff0080a5f1f1e0 x19: ffff0080afd22128 x18: 0000000000000000
[124676.644495] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000
[124676.644527] x14: 000000000000002f x13: ffffdce02abc05a0 x12: 0000000000000000
[124676.644559] x11: ffff0080804023d8 x10: ffff008080402238 x9 : 0000000000000000
[124676.644591] x8 : ffff0081731c5308 x7 : 0000000000000003 x6 : ffff800080c339e0
[124676.644624] x5 : 0000000000000000 x4 : 0000000000000000 x3 : 0000000000000001
[124676.644656] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff0080b4e98980
[124676.644688] Call trace:
[124676.644700] selinux_inode_free_security+0x78/0xbc
[124676.644721] security_inode_free+0x58/0x74
[124676.644737] __destroy_inode+0x74/0x1ec
[124676.645110] destroy_inode+0x2c/0x70
[124676.645128] evict+0x180/0x2d4
[124676.645147] iput+0xd4/0x228
[124676.645163] dentry_unlink_inode+0xe0/0x144
[124676.645180] __dentry_kill+0x78/0x1fc
[124676.645192] shrink_dentry_list+0x8c/0xf0
[124676.645207] shrink_dcache_parent+0xe8/0x158
[124676.645226] shrink_dcache_for_umount+0x38/0x188
[124676.645245] generic_shutdown_super+0x20/0x15c
[124676.645271] kill_anon_super+0x1c/0xd0
[124676.645288] deactivate_locked_super+0x50/0x160
[124676.645307] deactivate_super+0x68/0x7c
[124676.645322] cleanup_mnt+0x94/0x11c
[124676.645342] __cleanup_mnt+0x10/0x18
[124676.645359] task_work_run+0xb0/0x118
[124676.645420] do_notify_resume+0x138/0x13c
[124676.645436] el0_svc+0x100/0x1b8
[124676.645458] el0t_64_sync_handler+0x134/0x150
[124676.645481] el0t_64_sync+0x160/0x164
[124676.645498] Code: 88047c03 35ffff84 35000221 a9408261 (f9000420)
[124676.645518] ---[ end trace 0000000000000000 ]---
[124676.645537] Kernel panic - not syncing: Oops: Fatal exception
[124676.645566] SMP: stopping secondary CPUs
[124676.645696] Kernel Offset: 0x5cdfa9e00000 from 0xffff800080000000
[124676.645830] PHYS_OFFSET: 0xffffff8080000000
[124676.645858] CPU features: 0x00,00000013,00222908,4203f403
[124676.645893] Memory Limit: none
[124676.645913] ---[ end Kernel panic - not syncing: Oops: Fatal exception ]---
Code of Conduct
I have done the following
Steps to reproduce
No exact steps, occurs under moderate to heavy workload after several hours (this case was a gitlab runner node).
We've seen this signature in several issues that we closed with the 0.8.0.0 stability fixes. Those fixes apparently did not address this
selinux_inode_free_securityfailure mode, so let track this problem down and eliminate it.Current behavior
Container runs fine for several hours, then hangs.
Expected behavior
Container doesn't panic.
Environment
Relevant log output
Code of Conduct