Summary: | new-run-webkit-tests: print baseline search path as part of config output | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Dirk Pranke <dpranke> | ||||
Component: | New Bugs | Assignee: | Dirk Pranke <dpranke> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | abarth, dglazkov, eric, ojan, tony | ||||
Priority: | P2 | ||||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Dirk Pranke
2011-07-13 19:13:57 PDT
Created attachment 100756 [details]
Patch
Comment on attachment 100756 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=100756&action=review I would have just printed it one-per line. Some of these might get very long. But it's definitely better than what we have. > Tools/Scripts/webkitpy/layout_tests/controllers/manager.py:1103 > + fallback_path = [x.split(self._port.filesystem.sep)[-1] for x in self._port.baseline_search_path()] Why not use filesystem.split? I would probably have written this map(lambda path: self._port.filesystem.split(path)[-1], self._port.baseline_search_path()) (In reply to comment #2) > (From update of attachment 100756 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=100756&action=review > > I would have just printed it one-per line. Some of these might get very long. But it's definitely better than what we have. > I intentionally didn't do this because one-per line is much more verbose and harder to read (IMO). The longest each component might get is something like 'chromium-gpu-mac-snowleopard', which prints: Baseline search path: chromium-gpu-mac -> chromium-gpu -> chromium-mac -> chromium -> mac-snowleopard -> mac -> generic which is still significantly shorter than, say, the command line for DRT from the same output: /Volumes/Src/src/dev2/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/port/mock_drt.pyc --platform chromium-gpu-mac-snowleopard --test-shell --pixel-tests=/Volumes/Src/src/dev2/src/webkit/Release/layout-test-results/png_result0.png --enable-accelerated-compositing --enable-accelerated-2d-canvas > > Tools/Scripts/webkitpy/layout_tests/controllers/manager.py:1103 > > + fallback_path = [x.split(self._port.filesystem.sep)[-1] for x in self._port.baseline_search_path()] > > Why not use filesystem.split? I would probably have written this map(lambda path: self._port.filesystem.split(path)[-1], self._port.baseline_search_path()) Hm. I was thinking filesystem.split() would return the full dirname and an empty basename, but I guess in this case the full dirname is to .../LayoutTests/platform, and the basename is what you want. Will change. Note that you can use this command to see the baseline search path of any port, even one for an operating system different than the one you are running on: % new-run-webkit-tests --platform XX --no-build --nocheck-sys-deps --print config -n Seems useful, thanks! Committed r90973: <http://trac.webkit.org/changeset/90973> sample output: $ new-run-webkit-tests --platform chromium-gpu-mac-leopard --no-build --nocheck-sys-deps --print config -n Using port 'chromium-gpu-mac-leopard' Test configuration: <leopard, x86, release, gpu> Placing test results in /Volumes/Src/src/master/src/webkit/Release/layout-test-results Baseline search path: chromium-gpu-mac -> chromium-gpu -> chromium-mac-leopard -> chromium-mac -> chromium -> mac-leopard -> mac-snowleopard -> mac -> generic Using Release build Pixel tests enabled Regular timeout: 6000, slow test timeout: 30000 Command line: /Volumes/Src/src/master/src/xcodebuild/Release/DumpRenderTree.app/Contents/MacOS/DumpRenderTree --test-shell --pixel-tests=/Volumes/Src/src/master/src/webkit/Release/layout-test-results/png_result0.png --enable-accelerated-compositing --enable-accelerated-2d-canvas Worker model: processes Running 1 DumpRenderTree over 1 shard $ |