David Chisnall

I would like to cleanly separate the provided config format from the EEID format. The former should ideally be [UCL](https://github.com/vstakhov/libucl), the latter should be sorted, minified, JSON. This would also...

You gain a huge amount from using libUCL on host side. It will output compact JSON, so does not need to be present in the enclave.

@wintersteiger , FYI: please keep this in mind when replacing the config data. I hope that we'll get a lot of the fixes in `sgxlkl_app_config.c` for free when we move...

I agree, having a separate attested config makes a lot of sense, I don't have strong opinions as to whether we want attested vs nonattested config or a load of...

OE core provides a number of these functions that we can use: Functions expected in any freestanding C implementation: `memcmp`, `memcpy`, `memmove` `memset`. `oe_`-prefixed versions of some useful things: `oe_strlen`,...

@wintersteiger, I believe this is not an issue with the recent OE, which ensures that FS base is preserved across ecalls.

Yay, ticks!

This getenv calls are a bit unfortunate. They shouldn’t actually be getenv calls at all, they should be querying flags in the confit structure, which should be set from environment...

Adding to the relayering project. This may or may not be required to avoid the `lkl_run_in_kernel_stack` code once we have replaced cryptsetup.

@vtikoo, I think the `syscall_cp` assembly was not yet ported over to LKL, which may be related here?