!oNSIfazDqEcwhcOjSL:matrix.org

disko

387 Members
disko - declarative disk partitioning - https://github.com/nix-community/disko101 Servers

Load older messages

SenderMessageTime
14 Mar 2026
@matthewcroughan:defenestrate.itmatthewcroughanMaybe it's only on ZFS because of ARC 18:42:47
@lassulus:lassul.uslassulushmm, maybe?18:43:05
@lassulus:lassul.uslassulusbut we could disable that for the image builder maybe?18:43:19
@matthewcroughan:defenestrate.itmatthewcroughanpossibly as a kernel param.. yeah18:43:28
@lassulus:lassul.uslassulusbut not sure it's ARC, why would it be? you can copy more files than you have memory on ZFS usually?18:44:36
@matthewcroughan:defenestrate.itmatthewcroughanVMs18:44:48
@matthewcroughan:defenestrate.itmatthewcroughanweird stuff18:44:52
@matthewcroughan:defenestrate.itmatthewcroughanbut I can't reproduce it with ext4 atm18:45:00
@matthewcroughan:defenestrate.itmatthewcroughan Hah, okay so I just changed everything to zfs in the make-disk-image test 18:48:12
@matthewcroughan:defenestrate.itmatthewcroughanimmediate kernel panics18:48:15
@amadaluzia:unredacted.orgamadaluzia -> 4d2.org changed their display name from amadaluzia to amadaluzia[uorg].18:50:34
@matthewcroughan:defenestrate.itmatthewcroughanI think it's probably ARC..18:53:30
@matthewcroughan:defenestrate.itmatthewcroughanI have a toy big closure, and it didn't happen with only 1024 of memory 18:53:51
@matthewcroughan:defenestrate.itmatthewcroughanbut I bet if I get the size closer to 10G it will happen18:54:03
@matthewcroughan:defenestrate.itmatthewcroughanI can imagine what's happening tbh18:56:02
@matthewcroughan:defenestrate.itmatthewcroughan xargs -P 8 -I '{}' cp --recursive '{}' /mnt/nix/store 18:56:15
@matthewcroughan:defenestrate.itmatthewcroughan maybe if you're lucky, your {} gets filled with store paths that are small enough to succee 18:56:28
@matthewcroughan:defenestrate.itmatthewcroughan * maybe if you're lucky, your {} gets filled with store paths that are small enough to succeed 18:56:30
@matthewcroughan:defenestrate.itmatthewcroughan maybe if you're unlucky, you get 8 paths in {} that are too big to copy in one go for your current invocation 18:56:51
@matthewcroughan:defenestrate.itmatthewcroughanso I can't reproduce it because my toplevel is very different to what I can create in the test18:57:12
@matthewcroughan:defenestrate.itmatthewcroughanhehe, I know19:00:50
@matthewcroughan:defenestrate.itmatthewcroughandiffoscope19:00:51
@matthewcroughan:defenestrate.itmatthewcroughanah fuck it all19:02:56
@matthewcroughan:defenestrate.itmatthewcroughanit's only happening in kernel 6_19, 6_1819:03:08
@matthewcroughan:defenestrate.itmatthewcroughanCaught in the act19:04:48
@matthewcroughan:defenestrate.itmatthewcroughan 
nixos-disko-images> +++ xargs -P 8 -I '{}' cp --recursive '{}' /mnt/nix/store
nixos-disko-images> [   45.634884] Kernel panic - not syncing: Out of memory: compulsory panic_on_oom is enabled
nixos-disko-images> [   45.637327] CPU: 15 UID: 0 PID: 1928 Comm: cp Tainted: P           O        6.18.16 #1-NixOS PREEMPT(voluntary) 
nixos-disko-images> [   45.640188] Tainted: [P]=PROPRIETARY_MODULE, [O]=OOT_MODULE
nixos-disko-images> [   45.641794] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS unknown 02/02/2022
nixos-disko-images> [   45.644168] Call Trace:
nixos-disko-images> [   45.644972]  <TASK>
nixos-disko-images> [   45.645686]  dump_stack_lvl+0x5d/0x80
nixos-disko-images> [   45.646829]  vpanic+0xdb/0x2d0
nixos-disko-images> [   45.647822]  panic+0x6b/0x6b
nixos-disko-images> [   45.648745]  out_of_memory.cold+0x58/0x78
nixos-disko-images> [   45.649981]  __alloc_frozen_pages_noprof+0xfa1/0x1120
nixos-disko-images> [   45.651495]  alloc_pages_mpol+0x86/0x170
nixos-disko-images> [   45.652706]  folio_alloc_noprof+0x5b/0xb0
nixos-disko-images> [   45.653898]  __filemap_get_folio+0x1db/0x600
nixos-disko-images> [   45.655197]  ? page_cache_ra_unbounded+0x19f/0x260
nixos-disko-images> [   45.656666]  filemap_fault+0x112/0xf50
nixos-disko-images> [   45.657746]  ? filemap_map_pages+0x4ee/0x770
nixos-disko-images> [   45.659020]  __do_fault+0x34/0x1d0
nixos-disko-images> [   45.660068]  do_fault+0x350/0x5c0
nixos-disko-images> [   45.661093]  __handle_mm_fault+0x8d8/0xf40
nixos-disko-images> [   45.662365]  ? pick_next_task_fair+0x146/0x260
nixos-disko-images> [   45.663656]  handle_mm_fault+0xe7/0x2e0
nixos-disko-images> [   45.664751]  do_user_addr_fault+0x21a/0x690
nixos-disko-images> [   45.665975]  exc_page_fault+0x6a/0x150
nixos-disko-images> [   45.667109]  asm_exc_page_fault+0x26/0x30
nixos-disko-images> [   45.668339] RIP: 0033:0x5630228ed340
nixos-disko-images> [   45.669415] Code: Unable to access opcode bytes at 0x5630228ed316.
nixos-disko-images> [   45.671204] RSP: 002b:00007ffd040fe0f8 EFLAGS: 00010246
nixos-disko-images> [   45.672812] RAX: 0000000000011901 RBX: 0000000000011976 RCX: 0000000000000000
nixos-disko-images> [   45.674865] RDX: 0000000000011976 RSI: 00007f83903bf000 RDI: 0000000000000005
nixos-disko-images> [   45.676785] RBP: 00007ffd040fe120 R08: 0000000000000000 R09: 0000000000011976
nixos-disko-images> [   45.678651] R10: 0000000000000000 R11: 0000000000000000 R12: 00007f83903bf000
nixos-disko-images> [   45.680269] R13: 0000000000000000 R14: 0000000000000005 R15: 00007f83903bf000
nixos-disko-images> [   45.681855]  </TASK>
nixos-disko-images> [   45.682705] Kernel Offset: 0x19a00000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
nixos-disko-images> [   46.397862] Rebooting in 1 seconds..
19:04:50
@matthewcroughan:defenestrate.itmatthewcroughansomething something page cache19:05:12
@lassulus:lassul.uslassuluscan you check if https://github.com/nix-community/disko/pull/1234 makes it better? or worse? :D19:15:18
@matthewcroughan:defenestrate.itmatthewcroughanhttps://github.com/nixos/nixpkgs/issues/49519819:15:27
@matthewcroughan:defenestrate.itmatthewcroughanThe issue doesn't happen in 6.12, and does in anything higher19:15:54

Show newer messages

Back to Room ListRoom Version: 10