Support pixel antialising in DirectWrite.

Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

Committed: https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2

[bungeman-skia, 2016-06-13 22:25:04]

Description was changed from

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
==========

to

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002
==========

[bungeman-skia, 2016-06-14 14:40:52]

bungeman@google.com changed reviewers:
+ mtklein@google.com, reed@google.com

[bungeman-skia, 2016-06-14 14:40:53]

I think this handles all the various cases correctly after looking at almost all
possible combinations of the 'hPlbs' keys in SampleApp. This sort of change will
cause a very large set of rebaselines in Blink, but only on newer Windows bots.
It may be necessary to provide a way to hold this out of Blink until the
rebaselines can happen.

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkFontMgr_win_dw.cpp
File src/ports/SkFontMgr_win_dw.cpp (right):

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkFontMgr_win_dw....
src/ports/SkFontMgr_win_dw.cpp:20: #include "SkTypeface_win.h"
This is a fix for an existing issue when SK_HAS_DWRITE_2_H is false.

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkScalerContext_w...
File src/ports/SkScalerContext_win_dw.cpp (right):

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkScalerContext_w...
src/ports/SkScalerContext_win_dw.cpp:331: // DWRITE_TEXTURE_ALIASED_1x1 is not
misnamed, it must also be used with grayscale.
s/not/now, will fix

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkTypeface_win_dw...
File src/ports/SkTypeface_win_dw.cpp (right):

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkTypeface_win_dw...
src/ports/SkTypeface_win_dw.cpp:255: rec->fFlags |=
SkScalerContext::kGenA8FromLCD_Flag;
Previously this flag was logically implicitly always on. Keep it on for this
case.

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkTypeface_win_dw.h
File src/ports/SkTypeface_win_dw.h (right):

https://codereview.chromium.org/2065833002/diff/1/src/ports/SkTypeface_win_dw...
src/ports/SkTypeface_win_dw.h:82: void* fFactory2 = nullptr;
This makes the implementation code much easier to follow due to the nature of
the '#if' blocks. If we can require Windows 8.1 SDK or higher, we can remove
most of this ugly.

[andrei.bucur, 2016-06-14 15:14:52]

lgtm

LGTM, thank you for moving this so fast!

There was an extra question in the original bug report, about a function called
IDWriteGlyphRunAnalysis::GetAlphaBlendParams.

In a way, that function seems to be at the border between DWrite and D2D. Have
you thought about using it in Skia? I'm not personally familiar with the math
that should be applied to use those coefficients but I'm wondering if it can
bring improvements to the quality of glyph rendering with DWrite.

[bungeman-skia, 2016-06-15 20:48:29]

The CQ bit was checked by bungeman@google.com to run a CQ dry run

[commit-bot: I haz the power, 2016-06-15 20:48:38]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2065833002/20001

[commit-bot: I haz the power, 2016-06-15 21:08:06]

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

[commit-bot: I haz the power, 2016-06-15 21:08:07]

Dry run: This issue passed the CQ dry run.

[bungeman-skia, 2016-06-15 21:19:58]

Patch Set 2 adds SK_IGNORE_DW_GRAY_FIX so that the change to pixel anti-aliased
glyphs will be held out of Chromium for now.

[bungeman-skia, 2016-06-15 21:43:18]

The CQ bit was checked by bungeman@google.com to run a CQ dry run

[commit-bot: I haz the power, 2016-06-15 21:43:30]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2065833002/40001

[commit-bot: I haz the power, 2016-06-15 22:01:47]

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

[commit-bot: I haz the power, 2016-06-15 22:01:48]

Dry run: This issue passed the CQ dry run.

[mtklein, 2016-06-16 13:42:18]

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
File src/ports/SkScalerContext_win_dw.h (right):

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
src/ports/SkScalerContext_win_dw.h:79: DWRITE_TEXT_ANTIALIAS_MODE
fAntiAliasMode;
Looks like DWRITE_TEXT_ANTIALIAS_MODE (and _CLEARTYPE and _GRAYSCALE) is part of
dwrite_1.h.  Is the grayscale antialiasing really dwrite_2 dependent?

[bungeman-skia, 2016-06-16 14:31:13]

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
File src/ports/SkScalerContext_win_dw.h (right):

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
src/ports/SkScalerContext_win_dw.h:79: DWRITE_TEXT_ANTIALIAS_MODE
fAntiAliasMode;
On 2016/06/16 13:42:17, mtklein wrote:
> Looks like DWRITE_TEXT_ANTIALIAS_MODE (and _CLEARTYPE and _GRAYSCALE) is part
of
> dwrite_1.h.  Is the grayscale antialiasing really dwrite_2 dependent?

DWRITE_TEXT_ANTIALIAS_MODE is part of dwrite_1.h. Unfortunately it is only used
in dwrite_1.h by IDWriteBitmapRenderTarget1 which is a weird GDI interop thing
(you need an HDC to create one apparently). It wasn't until dwrite_2.h that
there was a CreateGlyphRunAnalysis that takes DWRITE_TEXT_ANTIALIAS_MODE. In
other words, DWRITE_TEXT_ANTIALIAS_MODE isn't really usable until dwrite_2.h.
I'm not including dwrite_1.h here explicitly I suppose, but DirectWrite says
that any later header contains all the earlier ones (it pretty much has to the
way its designed).

[mtklein, 2016-06-16 15:19:29]

lgtm, with a suggestion to hoist all the pragmas into one

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
File src/ports/SkScalerContext_win_dw.cpp (right):

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
src/ports/SkScalerContext_win_dw.cpp:469: "Could not create DW2 glyph run
analysis.");
<top of file>
#if !SK_HAS_DWRITE_2_H
    #pragma ...
#end

#if SK_HAS_DWRITE_2_H
    HRNM(...)
#endif

?

[bungeman-skia, 2016-06-16 17:04:58]

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
File src/ports/SkScalerContext_win_dw.cpp (right):

https://codereview.chromium.org/2065833002/diff/40001/src/ports/SkScalerConte...
src/ports/SkScalerContext_win_dw.cpp:469: "Could not create DW2 glyph run
analysis.");
On 2016/06/16 15:19:29, mtklein wrote:
> <top of file>
> #if !SK_HAS_DWRITE_2_H
>     #pragma ...
> #end
> 
> #if SK_HAS_DWRITE_2_H
>     HRNM(...)
> #endif
> 
> ?

Done.

[bungeman-skia, 2016-06-16 17:08:46]

Description was changed from

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002
==========

to

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.
==========

[bungeman-skia, 2016-06-16 17:27:51]

The CQ bit was checked by bungeman@google.com

[bungeman-skia, 2016-06-16 17:27:51]

The patchset sent to the CQ was uploaded after l-g-t-m from
andrei.bucur@gmail.com, mtklein@google.com
Link to the patchset: https://codereview.chromium.org/2065833002/#ps60001
(title: "Move compile time warnings to top.")

[commit-bot: I haz the power, 2016-06-16 17:27:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2065833002/60001

[commit-bot: I haz the power, 2016-06-16 17:42:25]

Description was changed from

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.
==========

to

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.

Committed:
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2
==========

[commit-bot: I haz the power, 2016-06-16 17:42:26]

Committed patchset #4 (id:60001) as
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2

[bungeman-skia, 2016-06-16 21:41:41]

A revert of this CL (patchset #4 id:60001) has been created in
https://codereview.chromium.org/2075913002/ by bungeman@google.com.

The reason for reverting is: Moving to the new glyph run analysis changes things
anyway, and the Chromium Win 10 bots are having issues with this. Revert until I
have time to make the suppression wider..

[bungeman-skia, 2016-06-16 21:59:05]

Description was changed from

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.

Committed:
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2
==========

to

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.

Committed:
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2
==========

[bungeman-skia, 2016-07-05 20:24:09]

Description was changed from

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

TBR=reed
Will move SkTScopedComPtr into src.

Committed:
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2
==========

to

==========
Support pixel antialising in DirectWrite.

DirectWrite2 supports pixel antialiasing and rendering without hinting.

BUG=skia:5416
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2065833002

Committed:
https://skia.googlesource.com/skia/+/bd770d619553a88eeaa64ff29082f62db5c9b4d2
==========

[bungeman-skia, 2016-07-06 22:59:03]

The CQ bit was checked by bungeman@google.com to run a CQ dry run

[commit-bot: I haz the power, 2016-07-06 22:59:07]

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, 2016-07-06 23:20:30]

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

[commit-bot: I haz the power, 2016-07-06 23:20:31]

Dry run: This issue passed the CQ dry run.

[bungeman-skia, 2017-03-02 16:04:25]

This CL was resurrected and continued at
https://skia-review.googlesource.com/c/9145/ .