Bug 43710

Summary:

[Qt] Fix layering violations in qt platform code that crash WebKit2

Product:

WebKit

Reporter:

Antti Koivisto <koivisto>

Component:

New Bugs

Assignee:

Nobody <webkit-unassigned>

Status:

RESOLVED DUPLICATE

Severity:

Normal

CC:

darin, jesus, kenneth, luiz, tonikitoo, webkit.review.bot

Priority:

Keywords:

Version:

528+ (Nightly build)

Hardware:

OS:

OS X 10.5

Attachments:

Description	Flags
patch	none

Antti Koivisto

Reported 2010-08-09 02:03:13 PDT

Qt networking code in WebCore casts to WebKit types in a few places. These don't exist with WebKit2 -> crash.

Attachments
patch (12.67 KB, patch) 2010-08-09 02:08 PDT, Antti Koivisto	no flags	Details Formatted Diff Diff
View All Add attachment proposed patch, testcase, etc.

Antti Koivisto

Comment 1 2010-08-09 02:08:48 PDT

Created attachment 63872 [details] patch

WebKit Review Bot

Comment 2 2010-08-09 02:11:06 PDT

Attachment 63872 [details] did not pass style-queue: Failed to run "['WebKitTools/Scripts/check-webkit-style']" exit_code: 1 WebCore/platform/network/qt/ResourceRequest.h:39: Code inside a namespace should not be indented. [whitespace/indent] [4] Total errors found: 1 in 12 files If any of these errors are false positives, please file a bug against check-webkit-style.

Kenneth Rohde Christiansen

Comment 3 2010-08-09 08:20:10 PDT

Comment on attachment 63872 [details] patch Jesus has been looking into fixing this in a more generic way together with Darin Adler. Can you please consult with him?

Antti Koivisto

Comment 4 2010-08-09 09:05:13 PDT

What are the benefits of this more generic fix? This is not complicated stuff.

Jesus Sanchez-Palencia

Comment 5 2010-08-09 10:16:33 PDT

(In reply to comment #4) > What are the benefits of this more generic fix? This is not complicated stuff. The generic solution is being developed at https://bugs.webkit.org/show_bug.cgi?id=42292 . The benefits of it are solving all the problems we have now in a way that we stay clear of layering violations and that can be easily expanded whenever needed without having to add anything else to the FrameLoaderClient. Moreover, we solve the problem for all WebKit's ports.

Antonio Gomes

Comment 6 2010-08-09 10:20:11 PDT

Imho, I think the general approach is the way to go too. Darin has justified the general approach pretty much.

Antti Koivisto

Comment 7 2010-08-10 02:41:08 PDT

ok, ok :) *** This bug has been marked as a duplicate of bug 42292 ***

Note You need to log in before you can comment on or make changes to this bug.