|
Description
Mark Lam
2013-09-06 15:12:34 PDT
Please make sure to also edit all scripts that rely on fast/js/regress being in fast. Will leave fast/js/resources where it is so that I don't have to touch every non-js test that uses it. (In reply to comment #2) > Will leave fast/js/resources where it is so that I don't have to touch every non-js test that uses it. Search-and-replace? I changed my mind. I will search and replace all the paths automatically in the affected test directories. Could you move js-test-pre.js & js-test-post.js to LayoutTests/resources first? (You probably need to svn cp and make the change in several separate commits). There are many other JS files that are dependent by the rest of layout tests. Also, please be sure to update all TestExpectations files when you make the move. Otherwise, hundreds of tests that have been marked failing, etc... are going to start failing and NRWT will spits out, again, hundreds of lint errors, and all kinds of the hell ensues. Created attachment 210951 [details]
Change LayoutTests' f*-j* files to use pre and post js files in LayoutTests/resources
(In reply to comment #7) > Created an attachment (id=210951) [details] > Change LayoutTests' f*-j* files to use pre and post js files in LayoutTests/resources rs=Filip Pizlo. Landed in r155279: <http://trac.webkit.org/r155279> Created attachment 210959 [details]
Change LayoutTests' m*-p* files to use pre and post js files in LayoutTests/resources.
(In reply to comment #9) > Created an attachment (id=210959) [details] > Change LayoutTests' m*-p* files to use pre and post js files in LayoutTests/resources. rs=Filip Pizlo. Landed in r155282: <http://trac.webkit.org/r155282>. Created attachment 210961 [details]
Change LayoutTests' s* files to use pre and post js files in LayoutTests/resources.
(In reply to comment #11) > Created an attachment (id=210961) [details] > Change LayoutTests' s* files to use pre and post js files in LayoutTests/resources. rs=Filip Pizlo. Landed in r155284: <http://trac.webkit.org/r155284>. Created attachment 210964 [details]
Change LayoutTests' t* files to use pre and post js files in LayoutTests/resources.
(In reply to comment #13) > Created an attachment (id=210964) [details] > Change LayoutTests' t* files to use pre and post js files in LayoutTests/resources. rs=Filip Pizlo. Landed in r155285: <http://trac.webkit.org/r155285>. Created attachment 210965 [details]
Change LayoutTests' u*-w* files to use pre and post js files in LayoutTests/resources.
(In reply to comment #15) > Created an attachment (id=210965) [details] > Change LayoutTests' u*-w* files to use pre and post js files in LayoutTests/resources. rs=Filip Pizlo. Landed in r155286: <http://trac.webkit.org/r155286>. http/tests/security/shape-inside-image-origin.html is now "failing": <http://build.webkit.org/results/Apple%20MountainLion%20Release%20WK1%20(Tests)/r155306%20(12666)/results.html> Mark, can you update test results accordingly? > Mark, can you update test results accordingly? Mark doesn't seem to be around, so I did that in r155312. Created attachment 211063 [details]
Change some remaining fast/* files to use pre and post js files in LayoutTests/resources.
(In reply to comment #19) > Created an attachment (id=211063) [details] > Change some remaining fast/* files to use pre and post js files in LayoutTests/resources. rs=Filip Pizlo. Landed in r155369: <http://trac.webkit.org/r155369>. Created attachment 211101 [details]
Point Tools scripts to LayoutTests/resources, and remove old fast/js/resources pre and post test files.
(In reply to comment #21) > Created an attachment (id=211101) [details] > Point Tools scripts to LayoutTests/resources, and remove old fast/js/resources pre and post test files. Thanks to rniwa for the review. Landed in r155392: <http://trac.webkit.org/r155392>. Created attachment 211110 [details]
Fix path to post file in a fwe tests, and adjust the expected results to match.
Comment on attachment 211110 [details] Fix path to post file in a fwe tests, and adjust the expected results to match. (In reply to comment #23) > Created an attachment (id=211110) [details] > Fix path to post file in a fwe tests, and adjust the expected results to match. rs=Filip Pizlo. Landed in r155401: <http://trac.webkit.org/r155401>. The webkitpy tests started failing after this: http://build.webkit.org/builders/Apple%20MountainLion%20Debug%20WK1%20%28Tests%29/builds/9980 (In reply to comment #25) > The webkitpy tests started failing after this: http://build.webkit.org/builders/Apple%20MountainLion%20Debug%20WK1%20%28Tests%29/builds/9980 r155392 has been rolled out in r155404: <http://trac.webkit.org/r155404>. Created attachment 211130 [details]
Remove un-needed js-test-style.css.
(In reply to comment #27) > Created an attachment (id=211130) [details] > Remove un-needed js-test-style.css. Thanks to rniwa for the review. Landed in r155410: <http://trac.webkit.org/r155410>. (In reply to comment #26) > (In reply to comment #25) > > The webkitpy tests started failing after this: http://build.webkit.org/builders/Apple%20MountainLion%20Debug%20WK1%20%28Tests%29/builds/9980 > > r155392 has been rolled out in r155404: <http://trac.webkit.org/r155404>. r155392 has been re-landed in r155411: <http://trac.webkit.org/r155411>. The issue was that update_webgl_conformance_tests_unittest.py was referencing js-test-style.css in the wrong path. However, js-test-style.css was deleted in r155410. Hence, the fix here is simply to delete the line that references js-test-style.css in update_webgl_conformance_tests_unittest.py. (In reply to comment #29) > (In reply to comment #26) > > (In reply to comment #25) > > > The webkitpy tests started failing after this: http://build.webkit.org/builders/Apple%20MountainLion%20Debug%20WK1%20%28Tests%29/builds/9980 > > > > r155392 has been rolled out in r155404: <http://trac.webkit.org/r155404>. > > r155392 has been re-landed in r155411: <http://trac.webkit.org/r155411>. > > The issue was that update_webgl_conformance_tests_unittest.py was referencing js-test-style.css in the wrong path. However, js-test-style.css was deleted in r155410. Hence, the fix here is simply to delete the line that references js-test-style.css in update_webgl_conformance_tests_unittest.py. Makes sense. Thanks! Created attachment 211164 [details]
Remove more references to the deleted js-test-style.css
Comment on attachment 211164 [details] Remove more references to the deleted js-test-style.css View in context: https://bugs.webkit.org/attachment.cgi?id=211164&action=review > LayoutTests/fast/js/line-column-numbers.html:5 > + This blank line was added to compensate for the line I delete above. This way I don't have to change all the result line numbers in the expected file. > LayoutTests/fast/js/stack-trace.html:6 > + This blank line was added to compensate for the line I delete above. This way I don't have to change all the result line numbers in the expected file. Comment on attachment 211164 [details] Remove more references to the deleted js-test-style.css View in context: https://bugs.webkit.org/attachment.cgi?id=211164&action=review > LayoutTests/platform/mac/accessibility/supports-focus-setting.html:6 > <script> > var successfullyParsed = false; > </script> successfullyParsed is completely useless in all these accessibility tests because of the fact it's defined before the actual test code :( (In reply to comment #31) > Created an attachment (id=211164) [details] > Remove more references to the deleted js-test-style.css Thanks to rniwa for the review. Landed in r155423: <http://trac.webkit.org/r155423>. Comment on attachment 211164 [details] Remove more references to the deleted js-test-style.css Attachment 211164 [details] did not pass efl-wk2-ews (efl-wk2): Output: http://webkit-queues.appspot.com/results/1732627 (In reply to comment #35) > (From update of attachment 211164 [details]) > Attachment 211164 [details] did not pass efl-wk2-ews (efl-wk2): > Output: http://webkit-queues.appspot.com/results/1732627 This ews failure is bogus. It complains of a C++ build error while the patch does not change any C++ code. Created attachment 211207 [details]
Move fast/js to js (Just the Tools/Scripts + LayoutTests/ChangeLog diff. The LayoutTests file move diff is too large and not as interesting).
In this last patch, I also moved LayoutTests/platform/*/fast/js to LayoutTests/platform/*/js. Relative paths to LayoutTests/resources needed to be updated in fast/js and platform/fast/js files. Also updated all TestExpectation files.
(In reply to comment #37) > Created an attachment (id=211207) [details] > Move fast/js to js (Just the Tools/Scripts + LayoutTests/ChangeLog diff. The LayoutTests file move diff is too large and not as interesting). > > In this last patch, I also moved LayoutTests/platform/*/fast/js to LayoutTests/platform/*/js. Relative paths to LayoutTests/resources needed to be updated in fast/js and platform/fast/js files. Also updated all TestExpectation files. rs=Filip Pizlo. Landed in r155452: <http://trac.webkit.org/r155452>. Note: though this patch is big, it had to be landed in one shot or things will break. Now, I'll keep an eye on the bots. If all goes well as expected, I'll close this bug. Created attachment 211219 [details] Move image-preload-helper.js from js/resources to resources. This patch was intended to go with the fast/js to js move, but was missing. It will resolve the fast/events/mouse-cursor*.html failures that showed up with r155452. (In reply to comment #39) > Created an attachment (id=211219) [details] > Move image-preload-helper.js from js/resources to resources. rs=Filip Pizlo. Landed in r155462: <http://trac.webkit.org/r155462>. Created attachment 211247 [details]
Replace remaining "fast js" names in scripts.
(In reply to comment #41) > Created an attachment (id=211247) [details] > Replace remaining "fast js" names in scripts. rs-Filip Pizlo. Landed in r155479: <http://trac.webkit.org/r155479>. (In reply to comment #38) > rs=Filip Pizlo. Landed in r155452: <http://trac.webkit.org/r155452>. > > ... I'll keep an eye on the bots. If all goes well as expected, I'll close this bug. The one issue with mouse-cursor tests have been fixed. The bots do not show regressions due to patches in this bug. Closing the bug as fixed. |