Libreoffice causes my System to freeze?!

General system Applications
1

I am baffled by a very peculiar problem, every time I open any .odt, .docx, .pdf with either Writer or Draw they show the little floating box with the loading bar. But then at the half way point my system completely freezes. Then killing it with a long press of the on/off button is necessary to be able to restart/do anything.
I have tried different files, opened them with Okular, Onlyoffice etc., its only Libreoffice that completely freaks out. I only have one extension installed, that enables better image export. (ancient I know but it still works and the in built functionality is too bad for my needs: https://extensions.libreoffice.org/en/extensions/show/export-as-images).
Trying to open Writre/Draw without any document shows nothing, Writer shortly displays the blue icon in the taskbar but that soon vanishes. The process soffice.bin then appears as a zombie process in Ksysguard.
I have uninstalled fresh and moved to still, then back again, the error persists nonetheless.

So as this is pretty much the first time I want/have to diagnose whats even happening, my primary question is, where do I find or how to I trigger the creation of logs that can be helpful in this case?

Im on the latest updates available (for EOS KDE) and my specs should be quite enough for it. [Ryzen 5900X - 64GB RAM - RX 6800XT - 200GB free space on the home disk]

Thanks, any help is greatly appreciated.

2

No expert, but have you tried starting Libreoffice in safe mode?

https://help.libreoffice.org/latest/en-US/text/shared/01/profile_safe_mode.html

3

Starting it in safemode kinda works, I can start the Programms, Writer etc. but when I try to apply a full factory reset, it just freezes again. Tried twice, it clears the contents of that window but the borders remain. Then it either freezes that Window or my System.

Shutting my System down while it wasnt frozen but LibreOffice was unresponsive, takes a long time.
It gets stuck at the step of
“A stop job is running for User Manager for UID 1000 (x/x mins)” for five to ten minutes.
Also
“systemd-shutdown[1]: waiting for process 69694 (soffice.bin)” takes a while to finish.

4

I’m no LO expert and can’t think of a next step.
but you might consider removing the add on that you are using, just to see if that changes te behaviour. Or to get in touch with the LO community, to see if they know how to analyse this problem.

5

Im gonna try to get help from their forum.
But just as a note, the extension worked fine, used it multiple times without issue.

6

Do other users on your system have the same problem?

Have you tried to remove the libreoffice config?

mv $HOME/.config/libreoffice/ $HOME/.config/libreoffice.bak

7

Can I just check, do you have hardware acceleration enabled? If yes, can you try disabling it?

Tools - Options - View - [Graphics Output → Use Hardware Acceleration]

8

Changing users doesnt really help, the loading banner reaches 100% and it doesnt freeze anything, but the Program doesnt actually open.
Ive removed the config, but as I tried the reset again, the libreoffice-startmenu is unresposive and cant be killed even with shift+delete in Ksysguard, and i dont know how else to kill it. So as I cant open another instance, ill restart my pc, which might take a while as it gets stuck. (as mentioned above)

9

Starting it in safe-mode without acceleration works, but it also works with acceleration. Also deleting the extension doesnt help.
I however cant use it normally at all.
Either I open a document with LO and it freezes/crashes my System, or I open LO by itself and it just dies in the background/doesnt start and freezes the process to complete unresponsiveness.

10

My only other thought, and while you’ve confirmed that switching users doesn’t help, - delete the entirety of the LibreOffice folder in ~/.config. Journalctl is a great tool that may help in pinpointing if there’s something else that’s happening in your system that precipitates the crash.

11

Just a guess to point into a direction:

whenever hard crashes happen I think of sth very basic, like drivers.

I can see RX 6800XT in your hardware descripiton. Is this a nVidia graphic? In the past buggy nVidia drivers were the culprit for crashes, freezes etc.

Have you checked for newer drivers (or an older version if already running the current one?)

12

Its a last gen mid range AMD card and Im running it with the normal AMDGPU drivers. I do have ROCm stuff from the AUR but I dont think that has anything to do with the problem cause disabling acceleration doesnt help.

13

Deleting the LO folder in .config didnt do anything.

So, I checked journalctl and this is the last bit before my System dies:

Nov 21 18:06:42 UserName kernel: amdgpu 0000:2f:00.0: amdgpu: bo 0000000063ca2c31 va 0x0800000000-0x0800000001 conflict with 0x0800000000-0x0800000200
Nov 21 18:06:42 UserName kernel: amdgpu: Failed to map VA 0x800000000000 in vm. ret -22
Nov 21 18:06:42 UserName kernel: amdgpu: Failed to map bo to gpuvm
Nov 21 18:06:42 UserName kernel: ------------[ cut here ]------------
Nov 21 18:06:42 UserName kernel: refcount_t: addition on 0; use-after-free.
Nov 21 18:06:42 UserName kernel: WARNING: CPU: 9 PID: 4096 at lib/refcount.c:25 refcount_warn_saturate+0xe5/0x110
Nov 21 18:06:42 UserName kernel: Modules linked in: snd_seq_dummy snd_hrtimer snd_seq snd_seq_device xt_CHECKSUM xt_MASQUERADE xt_conntrack ipt_REJECT xt_tcpudp nft_compat nf_nat_tftp nf_conntrack_tftp bridge stp llc wireguard curve25519_x86_64 libchacha20poly13>
Nov 21 18:06:42 UserName kernel:  joydev mac_hid loop fuse ip_tables x_tables dm_crypt cbc encrypted_keys trusted asn1_encoder tee uas usb_storage usbhid crct10dif_pclmul crc32_pclmul polyval_clmulni polyval_generic gf128mul ghash_clmulni_intel sha512_ssse3 aesn>
Nov 21 18:06:42 UserName kernel: CPU: 9 PID: 4096 Comm: soffice.bin Not tainted 6.6.2-arch1-1 #1 11215f9ba7ddfb51644674a5b2ced71612c62fe9
Nov 21 18:06:42 UserName kernel: Hardware name: Micro-Star International Co., Ltd. MS-7C37/MPG X570 GAMING PLUS (MS-7C37), BIOS A.L2 06/12/2023
Nov 21 18:06:42 UserName kernel: RIP: 0010:refcount_warn_saturate+0xe5/0x110
Nov 21 18:06:42 UserName kernel: Code: 82 a7 ff 0f 0b e9 cb 43 77 00 80 3d bf be 98 01 00 0f 85 5e ff ff ff 48 c7 c7 f8 28 69 af c6 05 ab be 98 01 01 e8 3b 82 a7 ff <0f> 0b e9 a4 43 77 00 48 c7 c7 50 29 69 af c6 05 8f be 98 01 01 e8
Nov 21 18:06:42 UserName kernel: RSP: 0018:ffffc90007eafc60 EFLAGS: 00010282
Nov 21 18:06:42 UserName kernel: RAX: 0000000000000000 RBX: ffff88816e198158 RCX: 0000000000000027
Nov 21 18:06:42 UserName kernel: RDX: ffff888fdec61708 RSI: 0000000000000001 RDI: ffff888fdec61700
Nov 21 18:06:42 UserName kernel: RBP: ffff888136f15000 R08: 0000000000000000 R09: ffffc90007eafae8
Nov 21 18:06:42 UserName kernel: R10: 0000000000000003 R11: ffff88901f2a36a8 R12: ffff8881dbd09b38
Nov 21 18:06:42 UserName kernel: R13: ffff8881dbd09b40 R14: ffff8881db52e640 R15: ffff88816e198000
Nov 21 18:06:42 UserName kernel: FS:  00007f79aead5300(0000) GS:ffff888fdec40000(0000) knlGS:0000000000000000
Nov 21 18:06:42 UserName kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Nov 21 18:06:42 UserName kernel: CR2: 0000562ef664f000 CR3: 00000001289c2000 CR4: 0000000000f50ee0
Nov 21 18:06:42 UserName kernel: PKRU: 55555554
Nov 21 18:06:42 UserName kernel: Call Trace:
Nov 21 18:06:42 UserName kernel:  <TASK>
Nov 21 18:06:42 UserName kernel:  ? refcount_warn_saturate+0xe5/0x110
Nov 21 18:06:42 UserName kernel:  ? __warn+0x81/0x130
Nov 21 18:06:42 UserName kernel:  ? refcount_warn_saturate+0xe5/0x110
Nov 21 18:06:42 UserName kernel:  ? report_bug+0x171/0x1a0
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? console_unlock+0x78/0x120
Nov 21 18:06:42 UserName kernel:  ? handle_bug+0x3c/0x80
Nov 21 18:06:42 UserName kernel:  ? exc_invalid_op+0x17/0x70
Nov 21 18:06:42 UserName kernel:  ? asm_exc_invalid_op+0x1a/0x20
Nov 21 18:06:42 UserName kernel:  ? refcount_warn_saturate+0xe5/0x110
Nov 21 18:06:42 UserName kernel:  dma_resv_add_fence+0x18f/0x1e0
Nov 21 18:06:42 UserName kernel:  amdgpu_amdkfd_gpuvm_acquire_process_vm+0x212/0x530 [amdgpu 0401721894ca8f32d5d0b424349ce03960632e80]
Nov 21 18:06:42 UserName kernel:  kfd_process_device_init_vm+0xb0/0x320 [amdgpu 0401721894ca8f32d5d0b424349ce03960632e80]
Nov 21 18:06:42 UserName kernel:  kfd_ioctl_acquire_vm+0x89/0xc0 [amdgpu 0401721894ca8f32d5d0b424349ce03960632e80]
Nov 21 18:06:42 UserName kernel:  kfd_ioctl+0x3cc/0x4e0 [amdgpu 0401721894ca8f32d5d0b424349ce03960632e80]
Nov 21 18:06:42 UserName kernel:  ? __pfx_kfd_ioctl_acquire_vm+0x10/0x10 [amdgpu 0401721894ca8f32d5d0b424349ce03960632e80]
Nov 21 18:06:42 UserName kernel:  __x64_sys_ioctl+0x97/0xd0
Nov 21 18:06:42 UserName kernel:  do_syscall_64+0x60/0x90
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? mntput_no_expire+0x4a/0x250
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? __call_rcu_common+0xf4/0x740
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? syscall_exit_to_user_mode+0x2b/0x40
Nov 21 18:06:42 UserName kernel:  ? srso_alias_return_thunk+0x5/0x7f
Nov 21 18:06:42 UserName kernel:  ? do_syscall_64+0x6c/0x90
Nov 21 18:06:42 UserName kernel:  ? do_syscall_64+0x6c/0x90
Nov 21 18:06:42 UserName kernel:  entry_SYSCALL_64_after_hwframe+0x6e/0xd8
Nov 21 18:06:42 UserName kernel: RIP: 0033:0x7f79b89ef3af
Nov 21 18:06:42 UserName kernel: Code: 00 48 89 44 24 18 31 c0 48 8d 44 24 60 c7 04 24 10 00 00 00 48 89 44 24 08 48 8d 44 24 20 48 89 44 24 10 b8 10 00 00 00 0f 05 <89> c2 3d 00 f0 ff ff 77 18 48 8b 44 24 18 64 48 2b 04 25 28 00 00
Nov 21 18:06:42 UserName kernel: RSP: 002b:00007fffe71fd340 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
Nov 21 18:06:42 UserName kernel: RAX: ffffffffffffffda RBX: 00007fffe71fd430 RCX: 00007f79b89ef3af
Nov 21 18:06:42 UserName kernel: RDX: 00007fffe71fd4b0 RSI: 0000000040084b15 RDI: 000000000000001d
Nov 21 18:06:42 UserName kernel: RBP: 00007fffe71fd4b0 R08: 000000000000000d R09: 0000000000000001
Nov 21 18:06:42 UserName kernel: R10: 0000000000000001 R11: 0000000000000246 R12: 0000000040084b15
Nov 21 18:06:42 UserName kernel: R13: 000000000000001d R14: 0000562ef65f7620 R15: 00007f796b2be180
Nov 21 18:06:42 UserName kernel:  </TASK>
Nov 21 18:06:42 UserName kernel: ---[ end trace 0000000000000000 ]---
Nov 21 18:06:42 UserName kernel: BUG: unable to handle page fault for address: 000000000090c4e0
Nov 21 18:06:42 UserName kernel: #PF: supervisor write access in kernel mode
Nov 21 18:06:42 UserName kernel: #PF: error_code(0x0002) - not-present page

I have no idea how to read what this says, so any help interpreting would be greatly appreciated.
Though GPT4 mentions one if the cause for this problem could be my gpus memory. Is there a way to test gpu-memory like it is for normal ram? But then again, I havent experienced any issues with any other program, I would think a hardware problem with the vram would likely manifest elsewhere as well.

14

Deleting the .config/ folder libreoffice does actually do something, it makes LO “only” freeze into complete unresponsiveness instead of crashing my system. After rebooting because I couldnt find any other way to stop/kill/terminate the running zombie process, a second start with the now recreated config causes a complete crash.
Also the process being in the background dead and frozen, still has an impact on my graphics card. For some reasin it prevents stable diffusion (ComfyUI) or llama.cpp from working, they get stuck without throwing an error. They dont load anything into gpu memory either. Really bizarrely weird behaviour.

The freezing/crashing also happens when I use the AppImage instead of the repo package.

15

No solution but perhaps helpful in case you need a reboot: