Okay, it's been two weeks on this RFC, so I'm going to prepare a pros/cons summary of the discussion so far in preparation for further action.
Regarding the proposal to **replace SGX support in TVM with Fortanix**: **Pros** * tightly integrated into Rust ecosystem, which gives support for more crates as well as a significantly simpler build process * better performance due to low overhead enclave entry/exit * better auditing of usercalls than [eo]calls; better auditing of Rust `std` * easier to build new functionality like RPCs * easier to debug enclaves (e.g., working backtraces, `println!`, `panic!`) **Cons** * malicious operator could deny availability by blocking network access * end user needs to know that data sent outside of the enclave must go through a secure channel I'd be glad to leave our current SGX infrastructure in place except that it adds maintenance burden as we upgrade our `tvm` crate. Even if the toolchain were robust enough to be tested in CI, we would still want to use Rust's own toolchain as the primary means for SGX support. Thus, unless anyone has any strong objections, we will deprecate `rust-sgx-sdk` in favor of `x86_64-fortanix-unknown-sgx` as proposed in #2885. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/dmlc/tvm/issues/2887#issuecomment-481865630
