Don't use pageLogicalHeight() to resolve percentage values in subframes

Don't use pageLogicalHeight() to resolve percentage values in subframes

In printing, pageLogicalHeight() is set to zero for all subframes
which makes is a bad candidate for resolving percentages against. For
subframes a better candidate would be the view height, as for
non-printing documents.

The RenderView knows when to layout in printing mode (top-level
document) and when not (subframes), so move
viewLogicalHeightForPercentages() to the RenderView and use its
knowledge to return the view height for subframes in printing to
resolve percentages against.

BUG=370948
R=ojan

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=174051

[davve, 2014-05-11 08:37:40]

Ojan, do you have any clue what's up with the off-by-one in
https://storage.googleapis.com/chromium-layout-test-archives/mac_blink_rel/63...

Seems to only happen on mac for some reason. Anything special considerations for
printing tests?

[davve, 2014-05-14 14:51:32]

Mac reftest issue taken care of now, by a slightly ugly fixed height on the body
element. PTAL.

[ojan, 2014-05-14 22:42:32]

lgtm

I don't have any good guesses for where the off-by-one bug comes from.
Unfortunately, we have close to 0 expertise in the printing code. It's entirely
inherited from WebKit and we've mostly not touched it at all. 

Your fix for that test-case seems fine to me. It's a bit sad. If you are
willing, could you make upload the old version of the test case and file a bug?
At least then we have a place to point to and have the discussion when we
encounter this bug in future tests.

[commit-bot: I haz the power, 2014-05-14 22:43:10]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/davve@opera.com/273043002/20001

[commit-bot: I haz the power, 2014-05-14 22:57:53]

Change committed as 174051

[davve, 2014-05-15 10:42:36]

On 2014/05/14 22:42:32, ojan wrote:

> Your fix for that test-case seems fine to me. It's a bit sad. If you are
> willing, could you make upload the old version of the test case and file a
bug?
> At least then we have a place to point to and have the discussion when we
> encounter this bug in future tests.

https://code.google.com/p/chromium/issues/detail?id=373693 filed.

I'm not sure what you mean by upload, do you think it's worthwhile having in
LayoutTests/ too (with corresponding TestExpectation)?

[ojan, 2014-05-15 19:19:13]

On 2014/05/15 10:42:36, David Vest wrote:
> On 2014/05/14 22:42:32, ojan wrote:
> 
> > Your fix for that test-case seems fine to me. It's a bit sad. If you are
> > willing, could you make upload the old version of the test case and file a
> bug?
> > At least then we have a place to point to and have the discussion when we
> > encounter this bug in future tests.
> 
> https://code.google.com/p/chromium/issues/detail?id=373693 filed.
> 
> I'm not sure what you mean by upload, do you think it's worthwhile having in
> LayoutTests/ too (with corresponding TestExpectation)?

I just meant attaching the test case to the bug as you did. Thanks!