Create ScreenWin::GetSystemMetricsInDIP and Correctly Deprecate display::win::GetSystemMetricsInDIP

Create ScreenWin::GetSystemMetricsInDIP and Correctly Deprecate display::win::GetSystemMetricsInDIP

This allows GetSystemMetricsInDIP to use the primary display for DIP
calculations consistent with MSDN.

BUG=426656
Committed: https://crrev.com/f7e4003f7b3c44cdf2e31c8356cdd57ff88caa1f
Cr-Commit-Position: refs/heads/master@{#409021}

[robliao, 2016-07-27 21:23:34]

The CQ bit was checked by robliao@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-27 21:23:45]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[robliao, 2016-07-27 21:23:59]

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

[robliao, 2016-07-27 21:24:00]

sky: Please review this CL. Thanks!

[sky, 2016-07-27 21:44:11]

https://codereview.chromium.org/2186163002/diff/1/ui/display/win/screen_win.h
File ui/display/win/screen_win.h (right):

https://codereview.chromium.org/2186163002/diff/1/ui/display/win/screen_win.h...
ui/display/win/screen_win.h:95: // Returns the result of GetSystemMetrics for
|metric| in DIP.
When would you use this rather than supplying an HWND and using it's scale?

[robliao, 2016-07-27 22:04:01]

https://codereview.chromium.org/2186163002/diff/1/ui/display/win/screen_win.h
File ui/display/win/screen_win.h (right):

https://codereview.chromium.org/2186163002/diff/1/ui/display/win/screen_win.h...
ui/display/win/screen_win.h:95: // Returns the result of GetSystemMetrics for
|metric| in DIP.
On 2016/07/27 21:44:11, sky wrote:
> When would you use this rather than supplying an HWND and using it's scale?

You'd use this when you actually want to just do things in DIP. That turns out
to be HWND invariant.

GetSystemMetrics always returns results in physical pixels based off of the
primary monitor's DPI. If we removed GetSystemMetricsInDIP, to obtain the
equivalent result, you'd perform this calculation:
GetSystemMetricsForHwnd(hwnd, ...) / GetScaleFactorForHwnd(hwnd). This is
equivalent to

(GetSystemMetrics(metric) * HwndScaleFactor / PrimaryMonitorScaleFactor) /
HwndScaleFactor, leaving us with
GetSystemMetrics(metric) / PrimaryMonitorScaleFactor

Developers who need a physical metric for an HWND (like
BrowserDesktopWindowTreeHostWin's physical frame size) will want use
GetSystemMetricsForHwnd and avoid any rounding errors that arise from
GetSystemMetrics.

I've amended the comments to capture some of this.

[sky, 2016-07-27 22:13:05]

https://codereview.chromium.org/2186163002/diff/20001/ui/display/win/screen_w...
File ui/display/win/screen_win.cc (right):

https://codereview.chromium.org/2186163002/diff/20001/ui/display/win/screen_w...
ui/display/win/screen_win.cc:326: return static_cast<int>(std::round(
I think you should ceil this, as that what we generally do.
Also, this isn't at all obvious. Please add a comment as to why this uses the
primary_display device scale factor.

[robliao, 2016-07-29 18:13:40]

https://codereview.chromium.org/2186163002/diff/20001/ui/display/win/screen_w...
File ui/display/win/screen_win.cc (right):

https://codereview.chromium.org/2186163002/diff/20001/ui/display/win/screen_w...
ui/display/win/screen_win.cc:326: return static_cast<int>(std::round(
On 2016/07/27 22:13:05, sky wrote:
> I think you should ceil this, as that what we generally do.
> Also, this isn't at all obvious. Please add a comment as to why this uses the
> primary_display device scale factor.

Added a comment. 
My inclination is that we should stick with rounding as this is what Windows
seems to do with most of the metrics (some exceptions apply). Using rounding
will keep us closer to the actual value used by Windows.

If there's a compelling reason to use ceil, happy to change it.

[sky, 2016-08-01 15:39:41]

I prefer that we stay consistent, which means ceiling, but you may be right that
in this case it won't matter as this isn't necessarily used with painting. Ok,
LGTM

[robliao, 2016-08-01 17:11:20]

The CQ bit was checked by robliao@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-01 17:11:45]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-01 18:21:44]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-01 18:21:45]

Dry run: This issue passed the CQ dry run.

[robliao, 2016-08-01 19:19:44]

The CQ bit was checked by robliao@chromium.org

[commit-bot: I haz the power, 2016-08-01 19:20:25]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-01 19:25:12]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-08-01 19:29:44]

Description was changed from

==========
Create ScreenWin::GetSystemMetricsInDIP and Correctly Deprecate
display::win::GetSystemMetricsInDIP

This allows GetSystemMetricsInDIP to use the primary display for DIP
calculations consistent with MSDN.

BUG=426656
==========

to

==========
Create ScreenWin::GetSystemMetricsInDIP and Correctly Deprecate
display::win::GetSystemMetricsInDIP

This allows GetSystemMetricsInDIP to use the primary display for DIP
calculations consistent with MSDN.

BUG=426656

Committed: https://crrev.com/f7e4003f7b3c44cdf2e31c8356cdd57ff88caa1f
Cr-Commit-Position: refs/heads/master@{#409021}
==========

[commit-bot: I haz the power, 2016-08-01 19:29:45]

Patchset 3 (id:??) landed as
https://crrev.com/f7e4003f7b3c44cdf2e31c8356cdd57ff88caa1f
Cr-Commit-Position: refs/heads/master@{#409021}