Lennart Poettering
Lennart Poettering
Hmm, this will break cases where we have no rootfs initially, but we do have a /usr/ partition and then repart creates/encrypts/formats a root partition for us. Not sure what...
So I wonder. Running systemd-repart in the initrd is necessary when the root fs doesn't exist yet. But otherwise, why would we run it in the initrd? We don't make...
So, we should figure out how local-fs.target, local-fs-pre.target, rootfs-ready.target are supposed to operate regarding active/passive. i.e. which side pulls them in, and if you start them manually what do you...
So I am OK with setting LoaderTimeInitUSec if it wasn't set so far. I am not fine with overriding LoaderTimeExecUSec I must say. I would be OK to setting it...
I still think that LoaderTimeExecUSec when unset, should be set to the *start* time of sd-stub, not to the handoff time as you do right now. i.e. so that LoaderTimeExecUSec...
well, it will pass data that is misleading, since if sd-boot is used LoaderTimeInit…LoaderTimeExec will cover the whole boot menu time, but not more, while with your patch if sd-boot...
Please provide a full backtrace of the generated coredump. i.e. open the coredump in gdb and then do "bt full", with debug symbols installed.
I am a bit puzzled about this? This controls which client can *resolve* a specific host? i.e. it doesn't control which IP addresses it can connect to, hence it can...
exposing a dbus and/or varlink interface might be OK, but call-outs to shell scripts? nah, we explicitly don#t want to be the glue that glues shell scripts together, so no,...