RAC Support for RISC-V

[microhobby]

My dear maintainers, as maybe you know, I'm working on builds for the Common Torizon: https://github.com/commontorizon/meta-common-torizon

One of the targets has riscv64 architecture, I'm having issues when trying to build RAC for this target. I have not investigated the root cause in depth, first I want to ask if I would expect some dependency not supported for RISC-V. Thought I'd ask first before investigating, maybe the maintainers know of some limitation for this architecture 🤔.

Maybe @simao Let me know.

[simao]

Which errors are you getting? I haven't tried to build for that arch...

[microhobby]

Follow the logs:

error: failed to run custom build command for `ring v0.16.20`
| note: To improve backtraces for build dependencies, set the CARGO_PROFILE_RELEASE_BUILD_OVERRIDE_DEBUG=true environment variable to enable debug information generation.
| 
| Caused by:
|   process didn't exit successfully: `/workdir/torizon/build-torizon/tmp/work/riscv64-tdx-linux/rac/0.0+gitAUTOINC+719afe6b82-r0/build/target/release/build/ring-5e4ed97de4206132/build-script-build` (exit status: 101)
|   --- stderr
|   thread 'main' panicked at 'called `Option::unwrap()` on a `None` value', /workdir/torizon/build-torizon/tmp/work/riscv64-tdx-linux/rac/0.0+gitAUTOINC+719afe6b82-r0/cargo_home/bitbake/ring-0.16.20/build.rs:358:10
|   stack backtrace:
|      0: rust_begin_unwind
|      1: core::panicking::panic_fmt
|      2: core::panicking::panic
|      3: core::option::Option<T>::unwrap
|      4: build_script_build::build_c_code
|      5: build_script_build::ring_build_rs_main
|      6: build_script_build::main
|      7: core::ops::function::FnOnce::call_once
|   note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
| warning: build failed, waiting for other jobs to finish...
|     Building [============>            ] 184/335: sysinfo, memoffset(build)...
|     Building [============>            ] 185/335: sysinfo, memoffset(build)...
|     Building [============>            ] 186/335: sysinfo, memoffset(build)...
|     Building [============>            ] 187/335: sysinfo, memoffset(build)...
|     Building [=============>           ] 188/335: sysinfo, proc-macro-error...
|     Building [=============>           ] 189/335: sysinfo, proc-macro-error...
|     Building [=============>           ] 190/335: sysinfo, proc-macro-error...
|     Building [=============>           ] 191/335: sysinfo, proc-macro-error...
|     Building [=============>           ] 192/335: sysinfo, proc-macro-error...
|     Building [=============>           ] 193/335: sysinfo, futures-macro, n...
|     Building [=============>           ] 194/335: sysinfo, futures-macro, n...
|     Building [=============>           ] 195/335: sysinfo, futures-macro, n...
|     Building [=============>           ] 196/335: sysinfo, futures-macro, n...
|     Building [=============>           ] 197/335: futures-macro, num-bigint...
|     Building [=============>           ] 198/335: futures-macro, num-bigint...
|     Building [=============>           ] 199/335: num-bigint, serde_derive,...
|     Building [=============>           ] 200/335: num-bigint, serde_derive,...
|     Building [==============>          ] 201/335: num-bigint, serde_derive
|     Building [==============>          ] 202/335: serde_derive

[microhobby]

@simao yeah, appears that ring does not support RISC-V: briansmith/ring#1182

[microhobby]

@simao but no worries, this is very low priority for now.

[simao]

Unfortunately ring is quite important for us :( Maybe riscv support will be merged eventually in ring.

[tkfu]

Just pinging on this one @microhobby: it seems ring 0.17.0, released Oct. 6, now supports RISC-V. Do you mind trying out a build with that dependency bumped to check if it works out of the box, since you've already got something set up?

[simao]

upstream tough still depends on ring 16, tried to compile it with 17 but does not compile. All other dependencies seem to have updated to 0.17 though so we could potentially update them.

[tkfu]

Looks like tough upstream updated to ring 17: awslabs/tough#717

[microhobby]

I will tested it for Common Torizon v6.7.0

[simao]

Unfortunately tough also changed his:

    tough: migrate to async

    Depend on tokio for all IO access and convert all related interfaces to
    async.
    Replace `std::io::Read` with `futures_core::stream::Stream` for
    hopefully seamless migration once `std::async_iter` stabilized.
    Use async-compatible locks from tokio and coincidentally fix a locking issue in
    datastore.

    This also affects the crates tough-kms, tough-ssm and tuftool.

so I will need some time to make RAC compile with the latest upstream changes from tough.

[simao]

tough upgraded ring but ring was still being used by other deps.

You can try #9 this branch is using the latest version of ring.