Investigate problem after #200

[usbalbin]

Seems to be some sporadic CI failures after #200 got merged.

Error seems to be this:

Install Rust for Xtensa

Run esp-rs/xtensa-toolchain@main
[...]
Run curl -L https://github.com/esp-rs/embuild/releases/latest/download/ldproxy-x86_64-unknown-linux-gnu.zip 
[...]
Run curl -L https://github.com/esp-rs/espup/releases/latest/download/espup-x86_64-unknown-linux-gnu  -o 
[...]
Run bash /home/runner/work/_actions/esp-rs/xtensa-toolchain/main/install.sh all latest
  bash /home/runner/work/_actions/esp-rs/xtensa-toolchain/main/install.sh all latest
  [[ "true" = true ]] && rustup default esp || true
  [[ "true" = true ]] && rustup override unset || true
  shell: /usr/bin/bash --noprofile --norc -e -o pipefail {0}
  env:
    rust_toolchain: nightly
jq: error (at <stdin>:4): Cannot index string with string "prerelease"
ERROR: version number is not correctly formatted: 
Error: Process completed with exit code 1.

[usbalbin]

@ivmarkov @MabezDev this seems to be very random, this time everything seems to work fine... I believe during #200 at some point I got the same error, however it disappeared just as quick. I assume this is the line in the failing script. Could this be the curl github API failing?

[MabezDev]

This is a transient failure we've being seeing from the xtensa-toolchain action, the failing jobs just need to be rerun. I think @jessebraham or @SergioGasquez is looking into it.