Is there any way to restart broken Enclave?

When an exception occurs in Enclave, exception handler will be called. However, we want to close the broken enclave, and start a new enclave, it seems has no valid method. 1) If the exception make process killed, then we can do nothing. 2) If let exception handler return, then control flow go back to error position.

I know we can adjust rip address in in-Enclave handler registered in `sgx_register_exception_handler`, but it only available on SGXv2 when MiscSelect is open. I want a method also avaliable for SGXv1.

Start by reviewing the exception flow around sgx_register_exception_handler and the distinction between SGXv1 and SGXv2 described in the issue. Determine whether a supported way exists to close a broken enclave and start a replacement after an exception on SGXv1; the issue is done when that behavior or limitation is clearly established.

Written by the indexing model from the issue text.