Port SkLazyPtr to new SkAtomics.h

Port SkLazyPtr to new SkAtomics.h

No algorithmic changes.  The new APIs let us avoid a few ugly trips through void*,
and I've made the consume/acquire/release decision explicitly conditioned on TSAN.

This should fix the attached bug, which is TSAN seeing us implementing the
sk_consume_load() with a relaxed load, where we used to pass __ATOMIC_CONSUME
to TSAN.  This restores us to the status quo of a couple weeks ago, where we
use relaxed loads (to avoid an extra dmb on ARM) for all setups except TSAN,
who gets the logically correct memory order, consume.

No public API changes.
TBR=reed@google.com

BUG=chromium:455606
Committed: https://skia.googlesource.com/skia/+/e72a80db3a05c78e65caf4c7cd24b73ebae74868

[mtklein_C, 2015-02-09 22:17:03]

mtklein@chromium.org changed reviewers:
+ bungeman@google.com

[mtklein_C, 2015-02-09 22:17:03]

This one's not exactly a rush, but it is a fix for a few TSAN-related Chrome
bugs.

[mtklein, 2015-02-09 22:30:16]

mtklein@google.com changed required reviewers:
+ bungeman@google.com

[mtklein, 2015-02-09 22:30:20]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2015-02-09 22:30:49]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/908943002/40001

[mtklein, 2015-02-09 22:31:51]

The CQ bit was unchecked by mtklein@google.com

[bungeman-skia, 2015-02-09 22:39:52]

lgtm

[mtklein, 2015-02-09 22:46:35]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2015-02-09 22:46:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/908943002/40001

[commit-bot: I haz the power, 2015-02-09 22:47:09]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/e72a80db3a05c78e65caf4c7cd24b73ebae74868