Account for device scale factor in AppWindow.setShape(), under Windows.

Account for device scale factor in AppWindow.setShape(), under Windows.

BUG=408728, 410593
Committed: https://crrev.com/a60ad8f6f93b05c237928ccd686a0c0dd8e3beaa
Cr-Commit-Position: refs/heads/master@{#293361}

[Wez, 2014-09-03 23:45:50]

wez@chromium.org changed reviewers:
+ sky@chromium.org

[Wez, 2014-09-03 23:45:50]

PTAL

[sky, 2014-09-04 13:23:05]

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
File ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc (right):

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:283: 
nit: no newline here.

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:290: const float&
scale = gfx::win::GetDeviceScaleFactor();
I think you want GetDPIScale() and no &, just const float scale. If you don't
think you want GetDPI, how come you want the scale factor?

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:301: if
(!rects.empty()) {
nit: no {}

[sky, 2014-09-04 17:40:09]

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
File ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc (right):

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:290: const float&
scale = gfx::win::GetDeviceScaleFactor();
On 2014/09/04 13:23:05, sky wrote:
> I think you want GetDPIScale() and no &, just const float scale. If you don't
> think you want GetDPI, how come you want the scale factor?

Actually, both are the same, so it shouldn't matter.

[Wez, 2014-09-04 17:50:50]

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
File ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc (right):

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:283: 
On 2014/09/04 13:23:05, sky wrote:
> nit: no newline here.

Done.

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:290: const float&
scale = gfx::win::GetDeviceScaleFactor();
On 2014/09/04 13:23:05, sky wrote:
> I think you want GetDPIScale() and no &, just const float scale. If you don't
> think you want GetDPI, how come you want the scale factor?

Re GetDPIScale: Done.

Re const float&: We want to use const& when complex types are returned, IIUC, to
help the compiler optimize out a potential copy, so it seems preferable to use
it consistently for both complex and POD types - that OK?

https://codereview.chromium.org/532913003/diff/1/ui/views/widget/desktop_aura...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:301: if
(!rects.empty()) {
On 2014/09/04 13:23:05, sky wrote:
> nit: no {}

Done.

[sky, 2014-09-04 18:00:56]

https://codereview.chromium.org/532913003/diff/20001/ui/views/widget/desktop_...
File ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc (right):

https://codereview.chromium.org/532913003/diff/20001/ui/views/widget/desktop_...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:286: SkRegion
device_region;
Does Skia not have a way to scale a SkRegion? Seems like that would be the ideal
way to go about this.

[Wez, 2014-09-04 18:19:08]

https://codereview.chromium.org/532913003/diff/20001/ui/views/widget/desktop_...
File ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc (right):

https://codereview.chromium.org/532913003/diff/20001/ui/views/widget/desktop_...
ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc:286: SkRegion
device_region;
On 2014/09/04 18:00:56, sky wrote:
> Does Skia not have a way to scale a SkRegion? Seems like that would be the
ideal
> way to go about this.

Sadly no. :(

My guess is that you're really intended to do this with SkPath (hence my comment
above), which has float coordinates, so can be scaled efficiently, and then
convert to SkRegion at the last possible moment, otherwise SkRegion will need
(inefficient) code approximating what I have below.

You can [usually] convert an SkRegion to an SkPath, which has float coordinates
and can be efficiently scaled to a new size, and then convertd back to SkRegion,
or in this case directly to HRGN, but of course that relies on the (non-trivial)
conversion back and forth.

[sky, 2014-09-04 18:56:03]

LGTM

[Wez, 2014-09-04 19:01:53]

The CQ bit was checked by wez@chromium.org

[commit-bot: I haz the power, 2014-09-04 19:03:19]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/wez@chromium.org/532913003/20001

[commit-bot: I haz the power, 2014-09-04 22:36:15]

Committed patchset #2 (id:20001) as 5b4a16d3d6a08cdc2f97e52992a8590f6089dd1b

[commit-bot: I haz the power, 2014-09-10 03:33:25]

Patchset 2 (id:??) landed as
https://crrev.com/a60ad8f6f93b05c237928ccd686a0c0dd8e3beaa
Cr-Commit-Position: refs/heads/master@{#293361}