Unify the three places that patch font loading for PDFium on Windows.

Unify the three places that patch font loading for PDFium on Windows.

This also removed the code from content, which doesn't belong there because it's for a feature in Chrome.

In order to implement this in a unified way, I needed to create a public interface around ChildThread. A lot of the change is updating callers of ChildThread to use ChildThreadImpl instead.

BUG=448473
Committed: https://crrev.com/8a021512a24f87d3b1723f260e0380b874f3b9eb
Cr-Commit-Position: refs/heads/master@{#314367}

[jam, 2015-02-03 05:06:54]

Patchset #1 (id:1) has been deleted

[jam, 2015-02-03 06:52:55]

jam@chromium.org changed reviewers:
+ scottmg@chromium.org

[scottmg, 2015-02-03 17:52:06]

It's too bad it's such a big change for something we can hopefully remove once
we change Skia.

But thanks for fixing this! LGTM.

+raymes fyi

https://codereview.chromium.org/897563002/diff/120001/chrome/utility/printing...
File chrome/utility/printing_handler.cc (left):

https://codereview.chromium.org/897563002/diff/120001/chrome/utility/printing...
chrome/utility/printing_handler.cc:183: // Patch the IAT for handling specific
APIs known to fail in the sandbox.
There's no (useful) test for any of this stuff because they're all mock drivers
at the print level... so will have to keep an eye on this.

https://codereview.chromium.org/897563002/diff/120001/content/child/child_thr...
File content/child/child_thread_impl.h (right):

https://codereview.chromium.org/897563002/diff/120001/content/child/child_thr...
content/child/child_thread_impl.h:63: virtual public ChildThread {
Why does this need to be virtual?

[jam, 2015-02-03 18:11:42]

The CQ bit was checked by jam@chromium.org

[commit-bot: I haz the power, 2015-02-03 18:11:53]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/897563002/120001

[jam, 2015-02-03 18:13:04]

https://codereview.chromium.org/897563002/diff/120001/content/child/child_thr...
File content/child/child_thread_impl.h (right):

https://codereview.chromium.org/897563002/diff/120001/content/child/child_thr...
content/child/child_thread_impl.h:63: virtual public ChildThread {
On 2015/02/03 17:52:06, scottmg wrote:
> Why does this need to be virtual?

It's because of the diamond inheritance: 

ChildThreadImpl derives from ChildThread
RenderThread derives from ChildThread
RenderThreadImpl derives from RenderThread & ChildThreadImpl

[commit-bot: I haz the power, 2015-02-03 18:16:40]

Committed patchset #6 (id:120001)

[commit-bot: I haz the power, 2015-02-03 18:17:32]

Patchset 6 (id:??) landed as
https://crrev.com/8a021512a24f87d3b1723f260e0380b874f3b9eb
Cr-Commit-Position: refs/heads/master@{#314367}