[torchx/lintrunner] Propagate pyre cmd errors in pyre_linter.py

[kiukchung]

The PYRE check for lintrunner doesn't check the return code of running subprocess.run(["pyre", "--output=json", "incremental"]) therefore will try to parse the content of the command's console output as json even if the command fails.

This results in the propagated error message being non-descriptive.

Before

Before the change pyre --output=json incremental failing would result in:

$ lintrunner
Warning: Could not find a lintrunner config at: '.lintrunner.private.toml'. Continuing without using configuration file.


>>> General linter failure:

  Advice (pyre) command-failed
    Failed due to JSONDecodeError:
    Expecting value: line 1 column 1 (char 0)

After

After this fix, we get the proper root cause

$ lintrunner

>>> General linter failure:

  Advice (pyre) command-failed
    Failed due to CalledProcessError:
    Command '['pyre', '--output=json', 'incremental']' returned non-zero exit
    status 2.

[kiukchung]

hmmm nvm this doesn't work since pyre will exit with a non-zero exit code when there are type checking errors.