Running TOS7 beta on two F4-SSD units, both same firmware, and found a few things that reproduce consistently on both. Posting here so TerraMaster support/devs can pick it up. Happy to provide more info if needed.
---
Bug 1 - iniparse crashes on every boot (both units)
Every boot, a bunch of iniparse processes segfault during early userspace init. Same crash on both units, different hardware, so it's not specific to one config. The kernel has CET/IBT enabled and this binary is jumping into non-executable memory.
Code: Select all
[ 11.925605] iniparse[581]: segfault at 7f69c1d32413 ip 00007f69c1d32413 sp 00007fffdb5ae840 error 15 in iniparse[2bd413,7f69c1a75000+2be000] likely on CPU 2 (core 2, socket 0)
[ 11.925624] Code: ff d5 59 5e 5f 5d 6a 05 5a 6a 0a 58 0f 05 41 ff e5 5d e8 3c ff ff ff 2f 70 72 6f 63 2f 73 65 6c 66 2f 65 78 65 00 00 01 00 00 <e8> 4a 00 00 00 83 f9 49 75 44 53 57 48 8d 4c 37 fd 5e 56 5b eb 2f
[ 11.925912] iniparse[582]: segfault at 7f92cf7d9413 ip 00007f92cf7d9413 sp 00007ffedac41e20 error 15 in iniparse[2bd413,7f92cf51c000+2be000] likely on CPU 3 (core 3, socket 0)
[ 11.927788] iniparse[583]: segfault at 7fb2827ae413 ip 00007fb2827ae413 sp 00007fffe8cc3f10 error 15 in iniparse[2bd413,7fb2824f1000+2be000] likely on CPU 2 (core 2, socket 0)
[ 11.928518] iniparse[586]: segfault at 7fadc650f413 ip 00007fadc650f413 sp 00007ffea2b771d0 error 15 in iniparse[2bd413,7fadc6252000+2be000] likely on CPU 3 (core 3, socket 0)
[ 11.928847] iniparse[584]: segfault at 7f7952e37413 ip 00007f7952e37413 sp 00007ffd3d900b40 error 15 in iniparse[2bd413,7f7952b7a000+2be000] likely on CPU 0 (core 0, socket 0)
[ 11.930098] iniparse[588]: segfault at 7f2d13575413 ip 00007f2d13575413 sp 00007ffed9ee28c0 error 15 in iniparse[2bd413,7f2d132b8000+2be000] likely on CPU 0 (core 0, socket 0)
[ 11.930776] iniparse[589]: segfault at 7f2b09508413 ip 00007f2b09508413 sp 00007fffad4d29f0 error 15 in iniparse[2bd413,7f2b0924b000+2be000] likely on CPU 2 (core 2, socket 0)
[ 11.930996] iniparse[590]: segfault at 7fe97a3b7413 ip 00007fe97a3b7413 sp 00007ffc58c686c0 error 15 in iniparse[2bd413,7fe97a0fa000+2be000] likely on CPU 0 (core 0, socket 0)
[ 11.931652] iniparse[591]: segfault at 7f58aa714413 ip 00007f58aa714413 sp 00007fffb63cebd0 error 15 in iniparse[2bd413,7f58aa457000+2be000] likely on CPU 0 (core 0, socket 0)
[ 11.932167] iniparse[593]: segfault at 7f99d160a413 ip 00007f99d160a413 sp 00007ffec1763850 error 15 in iniparse[2bd413,7f99d134d000+2be000] likely on CPU 2 (core 2, socket 0)
Both units produce the same crash pattern with the same code bytes. The binary needs to be recompiled with proper IBT-compatible indirect calls, or CET compatibility needs to be checked before release.
---
Bug 2 - md9 (OS array) always starts degraded at boot
On both units, md9 boots with 1 out of 2 mirrors and immediately starts a recover. It completes in a few seconds so the array ends up fine, but it shouldn't be doing this on every clean boot.
Code: Select all
[ 3.404811] md: md9 stopped.
[ 3.451769] md/raid1:md9: active with 1 out of 2 mirrors
[ 3.452809] md9: detected capacity change from 0 to 15990784
[ 5.013596] md: recover of RAID array md9
[ 8.132323] md: md9: recover done.
Current array state after boot (both units clean):
Code: Select all
/dev/md9:
Version : 1.2
Raid Level : raid1
Array Size : 7995392 (7.63 GiB 8.19 GB)
Raid Devices : 2
Total Devices : 2
State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0
Looks like the initramfs is assembling the array before the second NVMe is ready. The udev settle timeout might need adjusting for this platform?
---
Bug 3 - journald.conf has invalid size values, logging limits ignored
Both units log this on every boot:
Code: Select all
[ 11.401837] systemd-journald[404]: /etc/systemd/journald.conf:26: Failed to parse size value, ignoring: 90%
[ 11.401837] systemd-journald[404]: /etc/systemd/journald.conf:30: Failed to parse size value, ignoring: 90%
The config is using percentage values for SystemMaxUse and RuntimeMaxUse but systemd 249 (which TOS7 ships) does not support percentage syntax for these settings. The values are silently ignored and journal sizing falls back to defaults. Should be fixed to use absolute values like 1G or 2G.
---
All three reproduce on every boot across both units. Let me know if you need full dmesg dumps or any other info.
lets wait for their approach now.