You have been subscribed to a public bug:
from dmesg
1.313665] UBSAN: shift-out-of-bounds in
/build/linux-JjvoxS/linux-5.15.0/drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_device_queue_manager.c:997:32
[ 1.313668] shift exponent 64 is too large for 64-bit type 'long long
unsigned int'
there is also a crash but i don't know if it's related.
1.313671] Call Trace:
[ 1.313672] <TASK>
[ 1.313673] show_stack+0x52/0x5c
[ 1.313676] dump_stack_lvl+0x4a/0x63
[ 1.313678] dump_stack+0x10/0x16
[ 1.313679] ubsan_epilogue+0x9/0x49
[ 1.313680] __ubsan_handle_shift_out_of_bounds.cold+0x61/0xef
[ 1.313682] initialize_nocpsch.cold+0x15/0x59 [amdgpu]
[ 1.313841] device_queue_manager_init+0x208/0x3b0 [amdgpu]
[ 1.313946] kgd2kfd_device_init.cold+0x1af/0x483 [amdgpu]
[ 1.314079] amdgpu_amdkfd_device_init+0x135/0x170 [amdgpu]
[ 1.314178] amdgpu_device_ip_init+0x681/0x6a4 [amdgpu]
[ 1.314323] amdgpu_device_init.cold+0x25b/0x7db [amdgpu]
[ 1.314463] ? do_pci_enable_device+0xdb/0x110
[ 1.314466] amdgpu_driver_load_kms+0x1e/0x270 [amdgpu]
[ 1.314556] amdgpu_pci_probe+0x1ce/0x260 [amdgpu]
[ 1.314642] local_pci_probe+0x48/0x90
[ 1.314644] pci_device_probe+0x119/0x1f0
[ 1.314645] really_probe+0x21f/0x420
[ 1.314647] __driver_probe_device+0x119/0x190
[ 1.314648] driver_probe_device+0x23/0xc0
[ 1.314650] __driver_attach+0xbd/0x1f0
[ 1.314651] ? __device_attach_driver+0x120/0x120
[ 1.314652] bus_for_each_dev+0x7c/0xd0
[ 1.314654] driver_attach+0x1e/0x30
[ 1.314655] bus_add_driver+0x148/0x220
[ 1.314656] driver_register+0x95/0x100
[ 1.314657] __pci_register_driver+0x68/0x70
[ 1.314659] amdgpu_init+0x7c/0x1000 [amdgpu]
[ 1.314747] ? 0xffffffffc0fc0000
[ 1.314748] do_one_initcall+0x46/0x1e0
[ 1.314750] ? kmem_cache_alloc_trace+0x19e/0x2e0
[ 1.314752] do_init_module+0x52/0x260
[ 1.314753] load_module+0xb2b/0xbc0
[ 1.314754] __do_sys_finit_module+0xbf/0x120
[ 1.314756] __x64_sys_finit_module+0x18/0x20
[ 1.314757] do_syscall_64+0x59/0xc0
[ 1.314758] ? ksys_lseek+0x85/0xc0
[ 1.314759] ? exit_to_user_mode_prepare+0x37/0xb0
[ 1.314761] ? syscall_exit_to_user_mode+0x27/0x50
[ 1.314762] ? __x64_sys_lseek+0x18/0x20
[ 1.314763] ? do_syscall_64+0x69/0xc0
[ 1.314764] entry_SYSCALL_64_after_hwframe+0x61/0xcb
[ 1.314766] RIP: 0033:0x7fef87ab8a3d
[ 1.314767] Code: 5b 41 5c c3 66 0f 1f 84 00 00 00 00 00 f3 0f 1e fa 48 89
f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01
f0 ff ff 73 01 c3 48 8b 0d c3 a3 0f 00 f7 d8 64 89 01 48
[ 1.314768] RSP: 002b:00007fffe67aa7c8 EFLAGS: 00000246 ORIG_RAX:
0000000000000139
[ 1.314770] RAX: ffffffffffffffda RBX: 000055d1d00078a0 RCX: 00007fef87ab8a3d
[ 1.314771] RDX: 0000000000000000 RSI: 00007fef87c4f441 RDI: 0000000000000018
[ 1.314772] RBP: 0000000000020000 R08: 0000000000000000 R09: 0000000000000002
[ 1.314772] R10: 0000000000000018 R11: 0000000000000246 R12: 00007fef87c4f441
[ 1.314773] R13: 000055d1d002ea30 R14: 000055d1d0011600 R15: 000055d1d002eb10
[ 1.314774] </TASK>
[ 1.314778]
================================================================================
gpu is Amd r 380 tonga 4Gb.
I was adviced to try 6.0 series of kernel and I dont see the ubsan error there
but I would like to get the fixes backported 5.15 series.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
** Tags: bot-comment
--
5.15.0-53-UBSAN: shift-out-of-bounds in amdgpu
https://bugs.launchpad.net/bugs/1998285
You received this bug notification because you are a member of Kernel Packages,
which is subscribed to linux in Ubuntu.
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
