Bug 11345 - 1 fewer shouldChangeSelection delegate call for fast/forms/focus-control-to-page
Summary: 1 fewer shouldChangeSelection delegate call for fast/forms/focus-control-to-page
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Forms (show other bugs)
Version: 420+
Hardware: Mac OS X 10.4
: P2 Normal
Assignee: Geoffrey Garen
URL: http://build.webkit.org/results/post-...
Keywords: LayoutTestFailure, Regression
Depends on:
Blocks:
 
Reported: 2006-10-18 01:50 PDT by Adam Roben (:aroben)
Modified: 2006-10-20 12:26 PDT (History)
2 users (show)

See Also:


Attachments
fix (3.85 KB, patch)
2006-10-20 12:18 PDT, Geoffrey Garen
ggaren: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Adam Roben (:aroben) 2006-10-18 01:50:09 PDT
The new listbox control (landed in r16718) broke fast/forms/display-none-in-onchange-keyboard
Comment 1 Adam Roben (:aroben) 2006-10-18 01:51:03 PDT
Looking at the test page and tabbing among the controls manually, it seems that the "two" item in the second <select> doesn't get focused. Perhaps this is the problem?
Comment 2 Adam Roben (:aroben) 2006-10-18 12:57:39 PDT
I was clearly too tired when filing these bugs. The test broken by this change is actually fast/forms/focus-control-to-page
Comment 3 Adam Roben (:aroben) 2006-10-18 12:58:29 PDT
To be clear, r16718 (new RenderListBox) broke fast/forms/focus-control-to-page.
Comment 4 Adele Peterson 2006-10-18 14:19:11 PDT
I rolled out the list box change, and the test still fails
Comment 5 Geoffrey Garen 2006-10-19 15:01:05 PDT
I'll take this one.
Comment 6 Geoffrey Garen 2006-10-19 15:09:52 PDT
Nightlies confirm that this regression happened from r16710 to 16729.
Comment 7 Geoffrey Garen 2006-10-19 15:43:30 PDT
Surprisingly, rolling out 16718 and 16719 doesn't fix the bug.
Comment 8 Geoffrey Garen 2006-10-19 16:57:46 PDT
Narrowed down to 16710:16720.
Comment 9 Geoffrey Garen 2006-10-19 17:15:05 PDT
Narrowed down to 16714:16717.
Comment 10 Geoffrey Garen 2006-10-19 17:27:42 PDT
Narrowed down to 16715:16717, which by the way is impossible, so... ?
Comment 11 Adam Roben (:aroben) 2006-10-19 17:59:38 PDT
Did you narrow this down by rolling out changes, or actually going back to those revisions and testing?
Comment 12 Geoffrey Garen 2006-10-20 09:57:52 PDT
Some of each. New data:

r16718 has the regression. Rolling back to r16717 fixes the regression. However, Rolling r16718 out of TOT does not fix the regression.
Comment 13 Geoffrey Garen 2006-10-20 12:18:26 PDT
Created attachment 11169 [details]
fix

Beth reviewed this.