With old run-webkit-tests, I could do e.g. "run-webkit-tests -i inspector,http/inspector" to skip some tests. This is very useful when testing locally. For example, all inspector tests time out on my machine now for whatever mysterious reason, but I'd like to work on other issues first, running all tests except for these.
As mentioned on IRC, you can work around this of course by adding lines to the Skipped file or test_expectations.txt, but that's not a great workaround. This is a fine feature request.
Created attachment 131768 [details] Patch
Note that '-i inspector' won't work with just this patch because we're not handling directories properly; that's a separate patch since the code is in a different place ...
See bug 81067 as well.
(In reply to comment #3) > Note that '-i inspector' won't work with just this patch because we're not handling directories properly; that's a separate patch since the code is in a different place ... Never mind, this looks like it's user error on my part. Skipping directories will also work. However, if you are using test_expectations and you attempt to skip a directory on the command line and there is a more-specific expectation in the expectations file, that will override the Skip. This is probably also a bug, but much lower priority.
Comment on attachment 131768 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=131768&action=review > Tools/ChangeLog:14 > + Note that there is a separate bug that still needs fixing where > + directories are not handled properly either by the command line > + flag or in Skipped files. Is this still accurate? Can you point to the bug?
(In reply to comment #6) > (From update of attachment 131768 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=131768&action=review > > > Tools/ChangeLog:14 > > + Note that there is a separate bug that still needs fixing where > > + directories are not handled properly either by the command line > > + flag or in Skipped files. > > Is this still accurate? Can you point to the bug? No, this is no longer accurate; I will remove it from the ChangeLog.
Committed r110724: <http://trac.webkit.org/changeset/110724>