Inform user that partitions are not to be executed alone

[wucke13]

https://github.com/aeronautical-informatics/apex-linux/blob/3a503936950c6e47ac293481f1ff78a87a80b89a/partition/src/lib.rs#L28

panics if one tries to execute a partition freestanding, without the hypervisor:

thread '<unnamed>' panicked at 'called `Result::unwrap()` on an `Err` value: TypedError { err: PartitionInit, source: environment variable not found }', /sources/apex-rs-linux-0.1.0-main/src/lib.rs:28:45
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
fatal runtime error: failed to initiate panic, error 5
zsh: IOT instruction (core dumped)  result/bin/taws

Maybe it would be clever to print a message to the user, explaining briefly what to do instead (run the partition through the hypervisor).

[dadada]

We should make sure the environment is checked as early as possible. Doing this lazily might not be the right way, since it's harder to find a condition in which it is save to abort the execution of the partition later, than earlier.

[cvengler]

How does one execute a parition freestanding in the first place? The partition crate is a library crate, isn't it? And running the partition binary stand-alone results in an infinite running process but not in an error (on my machine).

[dadada]

I also can not reproduce the crash, same as @emilengler. Tested with the hello example and another partition. I'd like to close the issue as non-relevant.

[wucke13]

How does one execute a parition freestanding in the first place? The partition crate is a library crate, isn't it? And running the partition binary stand-alone results in an infinite running process but not in an error (on my machine).

The partition images are just normal executables, statically linked. You can totally just execute them like ./partition-image.

[wucke13]

I also can not reproduce the crash, same as @emilengler. Tested with the hello example and another partition. I'd like to close the issue as non-relevant.

I'm wondering then, what behavior do you see when directly executing a partition?

[dadada]

I'm wondering then, what behavior do you see when directly executing a partition?

The partition waits for a file-mutex that it never manages to aquire, but I don't see any crashes.

[dadada]

Could you maybe provide a backtrace?

[cvengler]

Any updates?

[sevenautumns]

Not really.
Priority lies with #47 rn.
Because we are going to use ptrace(2) for hypervisor/partition communication, we will probably remove the Lazy statics in partitions as well.
This Issue will be addressed after #47 resolved.