Revert "do not intentionally suppress errors in precompile script from being reported or failing the result"

[KristofferC]

Reverts #55909

Fixes #56063. Caused a 3x startup regression in 1.11.0.

[giordano]

Fixes #56063.

That's still an issue on master for v1.12, isn't it?

[KristofferC]

That's still an issue on master for v1.12, isn't it?

I don't get it. This PR reverts the PR that caused the regression?

[nsajko]

The regression doesn't seem to exist on master, though, just on v1.11:

$ hyperfine './nightly_normal/julia-ecf41b18cc/bin/julia --startup-file=no -e "1+1"'
Benchmark 1: ./nightly_normal/julia-ecf41b18cc/bin/julia --startup-file=no -e "1+1"
  Time (mean ± σ):     139.6 ms ±   0.8 ms    [User: 71.1 ms, System: 66.8 ms]
  Range (min … max):   138.2 ms … 141.6 ms    21 runs

--trace-compile=stderr also doesn't produce any output for startup on nightly.

[IanButterworth]

Does anyone happen to see the issue with that PR?

[vtjnash]

This should be against the backport branch, not master, since it doesn't affect master

[KristofferC]

I just pushed the commit to the backport branch.

[topolarity]

Did we ever figure out why this was causing issues on 1.11? Seems pretty mysterious