[Qt][WK2] fast/viewport/viewport-91.html still fails after r121555 and r121661

[János Badics]

This test:
fast/viewport/viewport-91.html fails on qt-5.0-wk2 after r121555 and r121661

See original bug at
https://bugs.webkit.org/show_bug.cgi?id=88114

[Csaba Osztrogonác]

https://trac.webkit.org/changeset/121661 fixed it in general, but it didn't fix it on Qt-WK2.

[Balazs Kelemen]

I think the WebKit2 part of the original patch is not enough:

diff --git a/Source/WebKit2/WebProcess/WebPage/WebPage.cpp b/Source/WebKit2/WebProcess/WebPage/WebPage.cpp
-    ViewportAttributes attr = computeViewportAttributes(m_page->viewportArguments(), minimumLayoutFallbackWidth, deviceWidth, deviceHeight, static_cast<int>(160 * m_page->deviceScaleFactor()), m_viewportSize);
+    ViewportAttributes attr = computeViewportAttributes(m_page->viewportArguments(), minimumLayoutFallbackWidth, deviceWidth, deviceHeight, m_page->deviceScaleFactor(), m_viewportSize);

Only the Mac port set up the deviceScaleFactor for the page, other ports leave it to the embedder. I think we should either set this up for every port or just allow ports to use the old (deprecated?) behavior.

[Balazs Kelemen]

(In reply to comment #2)
> I think the WebKit2 part of the original patch is not enough:
> 
> diff --git a/Source/WebKit2/WebProcess/WebPage/WebPage.cpp b/Source/WebKit2/WebProcess/WebPage/WebPage.cpp
> -    ViewportAttributes attr = computeViewportAttributes(m_page->viewportArguments(), minimumLayoutFallbackWidth, deviceWidth, deviceHeight, static_cast<int>(160 * m_page->deviceScaleFactor()), m_viewportSize);
> +    ViewportAttributes attr = computeViewportAttributes(m_page->viewportArguments(), minimumLayoutFallbackWidth, deviceWidth, deviceHeight, m_page->deviceScaleFactor(), m_viewportSize);
> 
> Only the Mac port set up the deviceScaleFactor for the page, other ports leave it to the embedder. I think we should either set this up for every port or just allow ports to use the old (deprecated?) behavior.

Ehh, I'm wrong, this should not change anything. But than how was it good, and how went it wrong???

[Balazs Kelemen]

Created attachment 150419 [details]
Patch

[Balazs Kelemen]

(In reply to comment #4)
> Created an attachment (id=150419) [details]
> Patch

We need to use the constant for layoutTestController.dumpConfigurationForViewport because the device dpi value has no chance to go through the UI process that calculates the deviceScaleFactor value.

[Build Bot]

Comment on attachment 150419 [details]
Patch

Attachment 150419 [details] did not pass mac-ews (mac):
Output: http://queues.webkit.org/results/13124196

[Csaba Osztrogonác]

Skipped on qt-5.0-wk2 by r121680. Please unskip it with the proper fix.

[WebKit Review Bot]

Comment on attachment 150419 [details]
Patch

Attachment 150419 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/13131184

New failing tests:
fast/loader/loadInProgress.html

[WebKit Review Bot]

Created attachment 150433 [details]
Archive of layout-test-results from gce-cr-linux-06

The attached test failures were seen while running run-webkit-tests on the chromium-ews.
Bot: gce-cr-linux-06  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.39-gcg-201203291735-x86_64-with-Ubuntu-10.04-lucid

[Balazs Kelemen]

Created attachment 150435 [details]
Patch

[Csaba Osztrogonác]

Comment on attachment 150435 [details]
Patch

rs=me

[Balazs Kelemen]

Comment on attachment 150435 [details]
Patch

Clearing flags on attachment: 150435

Committed r121740: <http://trac.webkit.org/changeset/121740>

[Balazs Kelemen]

All reviewed patches have been landed.  Closing bug.