| 2 Mar 2025 |
 hexa | whatever is on 24.11 | 17:54:26 |
| hexa | 2.91.1 | 17:54:35 |
 raitobezarius | so 2.91.1 I guess | 17:54:38 |
| raitobezarius | i will investigate | 17:54:51 |
| hexa | thx | 17:55:00 |
| hexa | capstone update | 17:55:49 |
| hexa | 5.0.3 -> 50.05 | 17:55:51 |
| hexa | * 5.0.3 -> 5.0.5 | 17:55:53 |
| hexa | https://termbin.com/1pam | 17:56:30 |
| hexa | OVMF-202411 -> OVMF-202502 | 17:56:44 |
| raitobezarius | OVMF looks more plausible to me | 17:56:58 |
| raitobezarius | though this is firmware… | 17:57:03 |
| raitobezarius | capstone is just there for diassembling, surely they're not starting weird cgroup while diassembling in qemu | 17:57:20 |
 adamcstephens | If OVMF is the culprit, and we have better tests we should run it against, please let me know or open a PR. | 19:13:28 |
| raitobezarius | it doesn't reproduce in an obvious way on x86_64 | 21:25:33 |
| hexa | 59d6c1268a1ec26d23841fbe25f905d72f25b05b is the first bad commit
commit 59d6c1268a1ec26d23841fbe25f905d72f25b05b (HEAD)
Author: Adam C. Stephens <adam@valkor.net>
Date: Tue Feb 25 10:26:48 2025 -0500
edk2: 202411 -> 202502
https://github.com/tianocore/edk2/releases/tag/edk2-stable202502
pkgs/by-name/ed/edk2/package.nix | 12 +++++-------
1 file changed, 5 insertions(+), 7 deletions(-)
| 21:26:43 |
| raitobezarius | well i guess i had a good intuition but how did you reproduced it? | 21:27:13 |
| raitobezarius | * well i guess i had a good intuition but how did you reproduce it? | 21:27:17 |
| hexa | failing builds got stuck here | 21:28:31 |
| hexa |
nixos-disk-image-aarch64-linux> Press ESC in 5 seconds to skip startup.nsh or any other key to continue.Press ESC in 4 seconds to skip startup.nsh or any other key to continue.Press ESC in 3 seconds to skip startup.nsh or any other key to continue.Press ESC in 2 seconds to skip startup.nsh or any other key to continue.Press ESC in 1 seconds to skip startup.nsh or any other key to continue.
| 21:28:34 |
| hexa | for much longer than 5s | 21:29:03 |
| raitobezarius | so there's kinda two problems | 21:29:12 |
| raitobezarius | first is that terminating builds seems to be misbehaving (which is expected in Lix tbh) | 21:29:23 |
| raitobezarius | second is that booting on AA64 ceased to work on that OVMF | 21:29:42 |
| raitobezarius | one guess would be that openssl is desync | 21:30:24 |
| raitobezarius | that OVMF upgraded to OpenSSL 3.4 | 21:30:34 |
| raitobezarius | but now someone would need to git bisect over AA64 the boot failure | 21:30:55 |
| raitobezarius | honestly if mbedTLS is now available on public EDK2, people should move to it | 21:31:29 |
| raitobezarius | the other weird change is all the confidential computing stuff | 21:31:50 |
| adamcstephens | I upgraded to openssl during the bump since upstream had done so as mentioned in the release notes, both to 3.4 | 21:36:47 |
