Summary: | support -wk2 port names properly in webkitpy.layout_tests.port | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Dirk Pranke <dpranke> | ||||
Component: | Tools / Tests | Assignee: | Dirk Pranke <dpranke> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | abarth, eric, ojan, rniwa, tony | ||||
Priority: | P2 | Keywords: | NRWT | ||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Dirk Pranke
2012-12-11 19:26:13 PST
Created attachment 179137 [details]
Patch
The uploaded patch at least makes port names ending in '-wk2' work properly. I'm still not sure what to do about release/debug, but I'm not sure if there's any cases where that information is needed from the port name, so this should fix most if not all of the current issues. (retitling the bug subject as well). Comment on attachment 179137 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=179137&action=review Bleh. This code all stinks. (not your fault). But if this moves us closer to expected behavior, and you feel we've sufficiently tested this, then this is OK. And we'll just keep making it better... eventually. In my view some sort of PortFactory should be responsible for computing the full "name" (or config dictionary) for the Port and then creating the port with that. Our current behavior of having the Port constructors do that doesn't work well. :( > Tools/Scripts/webkitpy/layout_tests/port/base.py:115 > + if self._name and '-wk2' in self._name: > + self._options.webkit_test_runner = True Do you need to use your fancy set_default instead? (In reply to comment #3) > (From update of attachment 179137 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=179137&action=review > > Bleh. This code all stinks. (not your fault). But if this moves us closer to expected behavior, and you feel we've sufficiently tested this, then this is OK. And we'll just keep making it better... eventually. > > In my view some sort of PortFactory should be responsible for computing the full "name" (or config dictionary) for the Port and then creating the port with that. Our current behavior of having the Port constructors do that doesn't work well. :( > Agreed. The PortFactory is trying to do that now, but we're not quite there yet (this patch does get us closer). I think the testing is sufficient. I guess we'll find out :) > > Tools/Scripts/webkitpy/layout_tests/port/base.py:115 > > + if self._name and '-wk2' in self._name: > > + self._options.webkit_test_runner = True > > Do you need to use your fancy set_default instead? No, we don't want this to be a default option; it should override whatever the default was, since -wk2 is being explicitly requested. Committed r137650: <http://trac.webkit.org/changeset/137650> |