Never paginate fixed-positioned objects.

Never paginate fixed-positioned objects.

They are supposed to be repeated on every page anyway. Not that we currently do
that, but there's still no need to paginate them. It can only do harm.

Added a flexbox test, which is what the bug report was about. This regressed
for flexbox (and other types of non-block-container blocks as well) with
https://codereview.chromium.org/1360753002 , but this has actually been broken
for ages for regular block containers, so I threw in a bonus test for that too.

BUG=535163
R=leviw@chromium.org

Committed: https://crrev.com/344e69ca2b1f2e20346eeb1f6d3d7d9fad38fd82
Cr-Commit-Position: refs/heads/master@{#356041}

[mstensho (USE GERRIT), 2015-10-23 18:57:52]

Okay, I hope you've been enjoying all the printing CLs I've sent recently. :-P

Time for dessert. Here's the thing that I actually wanted to fix. The rest was
just to be able to write layout tests for the code changes (yes, it takes 3
other CLs to be able to write a layout test for this glorified one-liner).

Note that we should ideally get https://codereview.chromium.org/1411313004/
landed before this one, but it's not such a big deal, except that the tests are
kind of useless without that other CL (they'd pass even without the code change
here, thanks to general borkification in printing land).

When printing (or print-previewing) for realz (and not just using the layout
tests framework), the actual fix in the code will work with or without that
other CL.

[leviw_travelin_and_unemployed, 2015-10-23 19:51:42]

Nothing I enjoy more than looking at print code!

Reasonable fix. Seems like we could also tickle this with an element that
contains fixed positions and is paginated? May be worth adding yet *another*
bonus test for that that doesn't require printing?

Speaking of printing, final question: do we have a bug for the fact that we
don't print fixed position elements on every page? Not that I think it's super
high priority or anything, just hoping it's trackable.

LGTM. Thanks, as usual, for the nice descriptions for those of us following
along at home ;)

[mstensho (USE GERRIT), 2015-10-23 20:02:44]

On 2015/10/23 19:51:42, leviw wrote:
> Reasonable fix. Seems like we could also tickle this with an element that
> contains fixed positions and is paginated? May be worth adding yet *another*
> bonus test for that that doesn't require printing?

Do you have any suggestions? I cannot think of any way to achieve this without
printing. If it's fixed-positioned, it kind of escapes all ancestors and
pagination contexts. But at least we can trap it on paper! Muhahaha!

> Speaking of printing, final question: do we have a bug for the fact that we
> don't print fixed position elements on every page? Not that I think it's super
> high priority or anything, just hoping it's trackable.

Actually, we should just keep crbug.com/544797 open until that gets fixed.

[leviw_travelin_and_unemployed, 2015-10-23 20:04:35]

On 2015/10/23 at 20:02:44, mstensho wrote:
> On 2015/10/23 19:51:42, leviw wrote:
> > Reasonable fix. Seems like we could also tickle this with an element that
> > contains fixed positions and is paginated? May be worth adding yet *another*
> > bonus test for that that doesn't require printing?
> 
> Do you have any suggestions? I cannot think of any way to achieve this without
printing. If it's fixed-positioned, it kind of escapes all ancestors and
pagination contexts. But at least we can trap it on paper! Muhahaha!

Transformed elements can contain fixed position descendants.
http://www.w3.org/TR/css3-2d-transforms/#issue-ca2c412c

> 
> > Speaking of printing, final question: do we have a bug for the fact that we
> > don't print fixed position elements on every page? Not that I think it's
super
> > high priority or anything, just hoping it's trackable.
> 
> Actually, we should just keep crbug.com/544797 open until that gets fixed.

[mstensho (USE GERRIT), 2015-10-26 11:59:22]

On 2015/10/23 20:04:35, leviw wrote:
> On 2015/10/23 at 20:02:44, mstensho wrote:
> > On 2015/10/23 19:51:42, leviw wrote:
> > > Reasonable fix. Seems like we could also tickle this with an element that
> > > contains fixed positions and is paginated? May be worth adding yet
*another*
> > > bonus test for that that doesn't require printing?
> > 
> > Do you have any suggestions? I cannot think of any way to achieve this
without
> printing. If it's fixed-positioned, it kind of escapes all ancestors and
> pagination contexts. But at least we can trap it on paper! Muhahaha!
> 
> Transformed elements can contain fixed position descendants.
> http://www.w3.org/TR/css3-2d-transforms/#issue-ca2c412c

Right! And this CL also handles those. It's not obvious that pagination should
be disabled for those (should they perhaps also be repeated in each column, too,
like regular fixed positioned objects should be repeated on each page?), but
let's do it for now. I suppose that the reason why fixed positioned objects
cease to be fixed positioned when inside transforms is that nobody wants to
implement that, and that there's nothing more to it than that.

Added a test for this. It's a bit of a weird one, due to buggy code in
LayoutState, that pretty much caused us to disable pagination for such objects
already. Unless we use an evil test. :)

The buggy code in LayoutState doesn't matter anymore, but it should either be
fixed or removed (it's only really used by pagination for now). The code is both
useless and harmless when this CL has landed.

[mstensho (USE GERRIT), 2015-10-26 11:59:44]

The CQ bit was checked by mstensho@opera.com

[mstensho (USE GERRIT), 2015-10-26 11:59:44]

The patchset sent to the CQ was uploaded after l-g-t-m from leviw@chromium.org
Link to the patchset: https://codereview.chromium.org/1426443003/#ps20001
(title: "Test for fixed positioned inside transform inside multicol.")

[commit-bot: I haz the power, 2015-10-26 11:59:53]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1426443003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1426443003/20001

[commit-bot: I haz the power, 2015-10-26 13:17:21]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-10-26 13:18:05]

Patchset 2 (id:??) landed as
https://crrev.com/344e69ca2b1f2e20346eeb1f6d3d7d9fad38fd82
Cr-Commit-Position: refs/heads/master@{#356041}