Valentin David

@olivercalder Could you have a look at the commits that are on top of your two commits and see if that makes sense to you. The main one is https://github.com/snapcore/snapd/pull/13511/commits/7be3e78303fa4273cba1ee3037db16ea01f45a60...

> Back then we wanted to make sure the setting is not ephemeral, and after reboot you still have the timezone set correctly. Having said that this looks good to...

@pedronis this is the minimal implementation. I have only done manual tests. I will write tests now. But it would be nice if you could go through this minimal change...

I have managed to get it to work on classic. However there are some manual steps required: * Boot with recovery keys * Remove `/var/lib/snapd/save/device/fde/tpm-lockout-auth` * Write `5` to `/sys/class/tpm/tpm0/ppi/request`....

It feels to me that the locking of tpm when booting with recovery keys is a bug. We should not try to call `MarkBootSuccess` when we have used recovery keys.

Marking as blocked because of lockout behavior on fresh TPM.

@bboozzoo > That's orthogonal to this branch, but perhaps we should think about reorganizing tests a little now. We have a workflow which builds the snapd snap from the branch....

> How is that going to work on, say, Fedora, where /snap is not present? Do we always invoke them with the dynamic linker? You need to have at least...

> ``` > zyga@ciri:~/Canonical/snapd/squashfs-root/lib$ ll > [...] > lrwxrwxrwx 1 zyga zyga 39 sty 2 13:22 ld-linux-aarch64.so.1 -> aarch64-linux-gnu/ld-linux-aarch64.so.1 > ``` That link should not exist at all. It just...

> > ``` > > zyga@ciri:~/Canonical/snapd/squashfs-root/lib$ ll > > [...] > > lrwxrwxrwx 1 zyga zyga 39 sty 2 13:22 ld-linux-aarch64.so.1 -> aarch64-linux-gnu/ld-linux-aarch64.so.1 > > ``` > > That link...