RT kernel: kernel warningWARNING: CPU: 17 PID: 118273 at kernel/sched/deadline.c:794 setup_new_dl_entity+0x12c/0x1e8

[Colin Ian King]

ARM64, Ubuntu Noble, RT kernel 6.8.1.1002-realtime in VM (using virt-manager), hosted on a 24 proc ARM server with 24 processor ARM64 client server installation.

using stress-ng V0.18.00 (see https://github.com/ColinIanKing or https://launchpad.net/~colin-king/+archive/ubuntu/stress-ng for latest debs) on ARM64 RT kernel I'm seeing a lot of kernel warning messages when running:

sudo stress-ng --class scheduler --perm 2 -t 1 --progress

[ 180.532908] ------------[ cut here ]------------
[ 180.532926] WARNING: CPU: 23 PID: 5830 at kernel/sched/deadline.c:794 setup_new_dl_entity+0x12c/0x1e8
[ 180.532972] Modules linked in: qrtr cfg80211 binfmt_misc nls_iso8859_1 dm_multipath efi_pstore nfnetlink dmi_sysfs qemu_fw_cfg ip_tables x_tables autofs4 btrfs blake2b_generic raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor xor_neon raid6_pq libcrc32c raid1 raid0 crct10dif_ce polyval_ce polyval_generic ghash_ce sm4 sha2_ce arm_smccc_trng sha256_arm64 sha1_ce virtio_rng xhci_pci xhci_pci_renesas aes_neon_bs aes_neon_blk aes_ce_blk aes_ce_cipher
[ 180.533137] CPU: 23 PID: 5830 Comm: stress-ng-clone Tainted: G W 6.8.1-1002-realtime #2-Ubuntu
[ 180.533149] Hardware name: QEMU KVM Virtual Machine, BIOS 2023.05-2ubuntu0.1 02/12/2024
[ 180.533156] pstate: 804000c5 (Nzcv daIF +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 180.533166] pc : setup_new_dl_entity+0x12c/0x1e8
[ 180.533179] lr : enqueue_dl_entity+0x610/0x928
[ 180.533192] sp : ffff800085e43620
[ 180.533196] x29: ffff800085e43620 x28: ffff5d40cf51c340 x27: ffff5d40cf51c340
[ 180.533212] x26: ffff5d40c6b93700 x25: ffffc42e213b7000 x24: 0000000000000017
[ 180.533228] x23: 0000000000000000 x22: 0000000000000017 x21: ffff5d40ff3e4080
[ 180.533242] x20: ffff5d40ff3e4080 x19: ffff5d40cd3401f0 x18: ffff800085da50b0
[ 180.533258] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000
[ 180.533273] x14: 0000000000000000 x13: 0000000000000000 x12: 0000000000000000
[ 180.533287] x11: 0000000000000000 x10: 0000000000000000 x9 : ffffc42e1e3595d4
[ 180.533302] x8 : 0000000000000000 x7 : 0000000000000000 x6 : 0000000000000000
[ 180.533315] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffffc42e1e359518
[ 180.533329] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff5d40cf6581f0
[ 180.533345] Call trace:
[ 180.533349] setup_new_dl_entity+0x12c/0x1e8
[ 180.533362] enqueue_dl_entity+0x610/0x928
[ 180.533375] enqueue_task_dl+0xbc/0x250
[ 180.533388] __do_set_cpus_allowed+0x104/0x2a0
[ 180.533400] __set_cpus_allowed_ptr_locked+0x180/0x1f0
[ 180.533413] __set_cpus_allowed_ptr+0x74/0xf0
[ 180.533425] migrate_enable+0x74/0x148
[ 180.533436] obj_cgroup_charge+0x190/0x2b8
[ 180.533447] kmem_cache_alloc+0x2dc/0x330
[ 180.533459] anon_vma_clone+0x64/0x1d8
[ 180.533469] anon_vma_fork+0x34/0x1a0
[ 180.533477] dup_mmap+0x5c0/0x738
[ 180.533489] dup_mm.constprop.0+0x6c/0x138
[ 180.533499] copy_process+0xd7c/0x1020
[ 180.533510] kernel_clone+0xa4/0x4a0
[ 180.533521] __do_sys_clone+0x78/0xe0
[ 180.533532] __arm64_sys_clone+0x2c/0x58
[ 180.533543] invoke_syscall+0x7c/0x128
[ 180.533553] el0_svc_common.constprop.0+0x4c/0x140
[ 180.533563] do_el0_svc+0x28/0x58
[ 180.533571] el0_svc+0x44/0x1c0
[ 180.533584] el0t_64_sync_handler+0x148/0x158
[ 180.533596] el0t_64_sync+0x1b0/0x1b8
[ 180.533606] ---[ end trace 0000000000000000 ]---

[Kevin Becker (kevinbecker)]

I have not been able to reproduce this one yet, neither in bare metal, LXD VMs (multipass) nor QEMU VMs on arm64, both linux-realtime and linux (generic). I'll continue working on this issue to try to reproduce it.

[Colin Ian King (colin-king)]

I'll see if I can find any other ways to repro this.