Bypass containing block size special-code for fixed-positioneds when printing.

Bypass containing block size special-code for fixed-positioneds when printing.

Consulting FrameView during printing doesn't really provide reliable results,
since class might just know the size of what we have on the screen.

While the size of their containing block (i.e. the page) is now calculated
correctly when printing, note that fixed positioned objects are still not
repeated on every page as they should according to the spec; they are still
only printed on the first page. But at least with the correct size and
position now. :)

BUG=544797
R=leviw@chromium.org

Committed: https://crrev.com/d68346d7b83e31397b1c8e7b52ad7642bbaa3579
Cr-Commit-Position: refs/heads/master@{#355893}

[mstensho (USE GERRIT), 2015-10-23 18:38:49]

Printing is a mess. :( But I think this change makes sense. Not quite sure why
it's a good idea to have this special code at all, but some tests that use
scaling are going to fail without it, so I just decided to stop digging and keep
it for now (but avoid it when printing).

[leviw_travelin_and_unemployed, 2015-10-23 19:42:58]

Yeah. Printing hasn't gotten much love. LGTM.

https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/printing/fixed-positioned.html (right):

https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/printing/fixed-positioned.html:6: <div
style="position:fixed; top:0; left:0; right:0; bottom:0; border:5px solid
yellow;"></div>
Without this, you'll get the viewport values for sizing instead of the pages,
no?

[mstensho (USE GERRIT), 2015-10-23 19:49:44]

https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
File third_party/WebKit/LayoutTests/printing/fixed-positioned.html (right):

https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/printing/fixed-positioned.html:6: <div
style="position:fixed; top:0; left:0; right:0; bottom:0; border:5px solid
yellow;"></div>
On 2015/10/23 19:42:58, leviw wrote:
> Without this, you'll get the viewport values for sizing instead of the pages,
> no?

I don't think I understand the question. Without what?

[leviw_travelin_and_unemployed, 2015-10-23 19:52:55]

On 2015/10/23 at 19:49:44, mstensho wrote:
>
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> File third_party/WebKit/LayoutTests/printing/fixed-positioned.html (right):
> 
>
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> third_party/WebKit/LayoutTests/printing/fixed-positioned.html:6: <div
style="position:fixed; top:0; left:0; right:0; bottom:0; border:5px solid
yellow;"></div>
> On 2015/10/23 19:42:58, leviw wrote:
> > Without this, you'll get the viewport values for sizing instead of the
pages,
> > no?
> 
> I don't think I understand the question. Without what?

The changes to LayoutBox I meant. That's how we'll end up sizing without your
change to the conditions in containingBlockBlahBlahBlahWidthHeight

[mstensho (USE GERRIT), 2015-10-23 20:11:14]

On 2015/10/23 19:52:55, leviw wrote:
> On 2015/10/23 at 19:49:44, mstensho wrote:
> >
>
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> > File third_party/WebKit/LayoutTests/printing/fixed-positioned.html (right):
> > 
> >
>
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> > third_party/WebKit/LayoutTests/printing/fixed-positioned.html:6: <div
> style="position:fixed; top:0; left:0; right:0; bottom:0; border:5px solid
> yellow;"></div>
> > On 2015/10/23 19:42:58, leviw wrote:
> > > Without this, you'll get the viewport values for sizing instead of the
> pages,
> > > no?
> > 
> > I don't think I understand the question. Without what?
> 
> The changes to LayoutBox I meant. That's how we'll end up sizing without your
> change to the conditions in containingBlockBlahBlahBlahWidthHeight

Yes. Yeah. Sometimes. At least when running layout tests, I think. When printing
in e.g. Chrome, you usually get a mix; the height being correct-ish (yes,
correct-ish, not correct - a pixel or so off), while the width just becomes
wrong. FrameView::layoutSize() starts to act as a random generator when entering
printing. :( There's what you have on the screen, there's the page size reported
from the printer driver (I guess), then there's at least two ways of scaling
that value by a factor of 1.25.

There is still low-hanging weird to fix in printing land for those who can be
bothered. I'm not sure if that someone is going to me, unless it turns out that
I've broken something recently.

[leviw_travelin_and_unemployed, 2015-10-23 20:12:21]

On 2015/10/23 at 20:11:14, mstensho wrote:
> On 2015/10/23 19:52:55, leviw wrote:
> > On 2015/10/23 at 19:49:44, mstensho wrote:
> > >
> >
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> > > File third_party/WebKit/LayoutTests/printing/fixed-positioned.html
(right):
> > > 
> > >
> >
https://codereview.chromium.org/1411313004/diff/1/third_party/WebKit/LayoutTe...
> > > third_party/WebKit/LayoutTests/printing/fixed-positioned.html:6: <div
> > style="position:fixed; top:0; left:0; right:0; bottom:0; border:5px solid
> > yellow;"></div>
> > > On 2015/10/23 19:42:58, leviw wrote:
> > > > Without this, you'll get the viewport values for sizing instead of the
> > pages,
> > > > no?
> > > 
> > > I don't think I understand the question. Without what?
> > 
> > The changes to LayoutBox I meant. That's how we'll end up sizing without
your
> > change to the conditions in containingBlockBlahBlahBlahWidthHeight
> 
> Yes. Yeah. Sometimes. At least when running layout tests, I think. When
printing in e.g. Chrome, you usually get a mix; the height being correct-ish
(yes, correct-ish, not correct - a pixel or so off), while the width just
becomes wrong. FrameView::layoutSize() starts to act as a random generator when
entering printing. :( There's what you have on the screen, there's the page size
reported from the printer driver (I guess), then there's at least two ways of
scaling that value by a factor of 1.25.
> 
> There is still low-hanging weird to fix in printing land for those who can be
bothered. I'm not sure if that someone is going to me, unless it turns out that
I've broken something recently.

Cool. Thanks :)

[mstensho (USE GERRIT), 2015-10-23 20:20:54]

The CQ bit was checked by mstensho@opera.com

[commit-bot: I haz the power, 2015-10-23 20:21:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1411313004/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1411313004/1

[commit-bot: I haz the power, 2015-10-23 21:37:29]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-10-23 21:38:25]

Patchset 1 (id:??) landed as
https://crrev.com/d68346d7b83e31397b1c8e7b52ad7642bbaa3579
Cr-Commit-Position: refs/heads/master@{#355893}