Bug 85221 – [chromium] DomStorage events handling needs TLC (2)

• Status: RESOLVED
• Resolution: FIXED
• Component: WebCore Misc.

[Michael Nordman]

See https://bugs.webkit.org/show_bug.cgi?id=85215 for more info.

Alter the WebKit::WebStorageArea and WebCore::StorageArea virtual interfaces
such that the mutators no longer return old values. This is to allow implementations
of the interface to operate more asynchronously.

Also clean up from the last patch, remove support for the DEPRECATED event
dispatching API.

[Michael Nordman]

Created attachment 139494 [details]
events3

[WebKit Review Bot]

Please wait for approval from [e-mail], [e-mail], [e-mail], [e-mail] or [e-mail] before submitting, as this patch contains changes to the Chromium public API. See also https://trac.webkit.org/wiki/ChromiumWebKitAPI.

[Michael Nordman]

This is the third in the multi-sided sequence of changes.
wkpatch1 - https://bugs.webkit.org/show_bug.cgi?id=84387
crpatch2 - http://codereview.chromium.org/10201010/

[WebKit Review Bot]

Comment on attachment 139494 [details]
events3

Attachment 139494 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/12586367

New failing tests:
storage/domstorage/events/documentURI.html
storage/domstorage/events/basic-setattribute.html
storage/domstorage/events/basic-body-attribute.html
storage/domstorage/events/basic.html
storage/domstorage/events/case-sensitive.html

[WebKit Review Bot]

Created attachment 139510 [details]
Archive of layout-test-results from ec2-cr-linux-01

The attached test failures were seen while running run-webkit-tests on the chromium-ews.
Bot: ec2-cr-linux-01  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

[Michael Nordman]

(In reply to comment #5)
> Created an attachment (id=139510) [details]
> Archive of layout-test-results from ec2-cr-linux-01
> 
> The attached test failures were seen while running run-webkit-tests on the chromium-ews.
> Bot: ec2-cr-linux-01  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

crpatch two is r134037
https://src.chromium.org/viewvc/chrome?view=rev&revision=134037

From the bot build log...
48>________ running 'svn update /mnt/git/webkit-chromium-ews/Source/WebKit/chromium/webkit --revision 133673 --non-interactive --force --accept theirs-conflict --ignore-externals' in '/mnt/git/webkit-chromium-ews/Source/WebKit/chromium'
48>At revision 133673.

133673 < 134037... looks like i have to inverse rolldeps

[Michael Nordman]

> 133673 < 134037... looks like i have to inverse rolldeps

ah... gavin already did it today

[Michael Nordman]

Created attachment 139530 [details]
events3

re-upping the same patch to get new crlinux try job to run (idk, maybe there's a better way to do that)

[WebKit Review Bot]

Comment on attachment 139530 [details]
events3

Attachment 139530 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/12598222

[Michael Nordman]

Created attachment 139551 [details]
events3

fixed the gyp file to delete the 3 files listed in the changelog instead of 2 out of those 3 and then one that wasn't supposed to be deleted :/

[Michael Nordman]

hooray for green bots!

[Adam Barth]

Comment on attachment 139551 [details]
events3

View in context: https://bugs.webkit.org/attachment.cgi?id=139551&action=review

> Source/WebKit/chromium/src/WebViewImpl.h:304
> +    // A pageGroup identifies a namespace of pages. Page groups are used on OSX

OSX -> PLATFORM(MAC)

Specifically, it's used by the Apple port on OSX, not by other ports (like Chromium) on OSX.

[Michael Nordman]

thank you!

i'll fixup the comment prior to committing... like we chatted about, i'll be waiting till after the chromium branch date to land

[Michael Nordman]

> i'll fixup the comment prior to committing... like we chatted about, i'll be waiting till after the chromium branch date to land

just a heads up, i'm still waiting to land this till branch is safely behind us (which apparently is not quite yet)

[Michael Nordman]

Committed r116712: <http://trac.webkit.org/changeset/116712>

[WebKit Review Bot]

Re-opened since this is blocked by 86484

[Michael Nordman]

Created attachment 142825 [details]
events3-again

Let's try this one again. The original patch got reverted because it made for spectacular crash fest, see http://code.google.com/p/chromium/issues/detail?id=127919 for those details.

This patch very close to the original with minor of differences to avoid crashing the crash described in the crbug. The critical difference is to not create a page's sessionStorage object while iterating in the 'find' method.

[Michael Nordman]

Created attachment 142828 [details]
events3-again

added the ChangeLog to the patch

[Adam Barth]

Comment on attachment 142828 [details]
events3-again

View in context: https://bugs.webkit.org/attachment.cgi?id=142828&action=review

> Source/WebKit/chromium/src/WebViewImpl.h:323
> +    // A pageGroup identifies a namespace of pages. Page groups are used on OSX

OSX -> PLATFORM(MAC)

[Michael Nordman]

Committed r117797: <http://trac.webkit.org/changeset/117797>