Currently I have benchmark tests that I run with LIT that have meaningful output even when the test passes. The aim of this patch is to modify the existing LIT output options in such a way that they can be used to always print the output of a test with minimal changes to how output is currently handled.
The main change in this patch is that -v can be specified multiple times to increase the amount of information that is output. Using lit -vv will cause LIT to report a passing tests output (if any is present).
This patch has a secondary drive-by change that changes when metrics are output. Currently metrics are reported when:
- The test failed.
- The test passed and neither '-s' nor '-q' are specified.
With this patch metrics are reported ONLY when '-s' is not specified and one of the following is true:
- The test failed.
- The test passed and neither '-s' nor '-q' are specified and '-vv' is not specified.
- The test passed and '-vv' is specified.