Cache hb_font_t and the user data container instead of hb_face_t

Cache hb_font_t and the user data container instead of hb_face_t

Previously, we did cache the hb_face_t but not the hb_font_t, and
instead generated an hb_font_t for each HarfBuzzFace object,
outside of the cache. In the prepareHarfBuzzFontData() function,
we also reset the font functions on the newly created
hb_font_t. So in the HarfBuzzFace constructor and in each created
HarfBuzzFace object, we would create hb_font_t and reset it,
instead of reusing this from the cache, too.

This lead to HarfBuzz reloading font tables for each
FontPlatformData instance, instead of using the
per-unique-typeface caching mechanism.

This should address the memory regression we have been observing
on Android.

BUG=577306
R=behdad,eae

[eae, 2016-05-04 12:46:03]

https://codereview.chromium.org/1943993003/diff/1/third_party/WebKit/Source/p...
File third_party/WebKit/Source/platform/fonts/shaping/HarfBuzzFace.cpp (right):

https://codereview.chromium.org/1943993003/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/fonts/shaping/HarfBuzzFace.cpp:80: //
independet of size, then consider using this here.
If (when) we fix that, would we even need this cache? Long term it would be nice
to have one true font cache instead of the multiple layers of caches we have
today...

[behdad, 2016-05-04 12:47:13]

lgtm

Great!  I tested this in Content Shell on Android and indeed it reduces the font
instantiations and table loadings.

[drott, 2016-05-04 12:54:00]

Bug filed for FIXME

[eae, 2016-05-04 12:54:48]

LGTM, thank you!

[drott, 2016-05-04 12:54:54]

The CQ bit was checked by drott@chromium.org

[drott, 2016-05-04 12:54:54]

The patchset sent to the CQ was uploaded after l-g-t-m from behdad@chromium.org
Link to the patchset: https://codereview.chromium.org/1943993003/#ps20001
(title: "Bug filed for FIXME")

[eae, 2016-05-04 12:54:54]

The CQ bit was checked by eae@chromium.org

[commit-bot: I haz the power, 2016-05-04 12:55:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/20001

[eae, 2016-05-04 14:58:16]

The CQ bit was unchecked by eae@chromium.org

[eae, 2016-05-04 14:59:00]

The CQ bit was checked by eae@chromium.org

[commit-bot: I haz the power, 2016-05-04 14:59:10]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/20001

[commit-bot: I haz the power, 2016-05-04 16:38:12]

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

[commit-bot: I haz the power, 2016-05-04 16:38:13]

Try jobs failed on following builders:
  win_chromium_rel_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_rel_...)

[drott, 2016-05-04 17:57:47]

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

[drott, 2016-05-04 17:57:50]

Attempt to fix windows by rounding scale factor

[commit-bot: I haz the power, 2016-05-04 17:58:29]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/40001

[drott, 2016-05-04 17:58:30]

The CQ bit was unchecked by drott@chromium.org

[drott, 2016-05-04 17:58:32]

The CQ bit was checked by drott@chromium.org

[drott, 2016-05-04 17:58:32]

The patchset sent to the CQ was uploaded after l-g-t-m from eae@chromium.org,
behdad@chromium.org
Link to the patchset: https://codereview.chromium.org/1943993003/#ps40001
(title: "Attempt to fix windows by rounding scale factor")

[commit-bot: I haz the power, 2016-05-04 17:59:03]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/40001

[commit-bot: I haz the power, 2016-05-04 22:39:38]

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

[commit-bot: I haz the power, 2016-05-04 22:39:39]

Try jobs failed on following builders:
  win_chromium_rel_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_rel_...)

[drott, 2016-05-05 10:40:45]

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

[drott, 2016-05-05 10:40:47]

Additional paint setup instead of rounding

[commit-bot: I haz the power, 2016-05-05 10:40:52]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/60001

[commit-bot: I haz the power, 2016-05-05 12:24:35]

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

[commit-bot: I haz the power, 2016-05-05 12:24:36]

Dry run: This issue passed the CQ dry run.

[drott, 2016-05-05 14:38:56]

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

[drott, 2016-05-05 14:38:58]

Reset paint flegs properly

[commit-bot: I haz the power, 2016-05-05 14:39:12]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1943993003/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1943993003/80001

[commit-bot: I haz the power, 2016-05-05 15:44:49]

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

[commit-bot: I haz the power, 2016-05-05 15:44:51]

Dry run: This issue passed the CQ dry run.

[drott, 2016-05-06 11:59:26]

We landed the other version, with the Windows change separately.