On Mon, Jun 15, 2020 at 05:45:28PM +0000, Bird, Tim wrote: > Personally, as a human I find the space prefix a bit easier to read. > However, I think that in "normal" kernel log output it is unusual for > a line to be prefixed with a hash (#), so this might be easier to > both visually distinguish and for automated parsing. > So I'm torn. I don't have a strong opinion on space vs. hash prefix > for indicating sub-test. I think the KUnit convention of matching > whatever was the prefix of the "TAP version 14" line is clever, but > it would be nice to just pick a prefix and be done with it. Are there plans in kernelci for doing any parsing of subtests? (As in, what do we break if we move from "# " to " "?) I'm really thinking " " makes sense now. :) -- Kees Cook
