Make a flipped fDirtyBit always mean "this field is not the default".

Make a flipped fDirtyBit always mean "this field is not the default".

Haven't been able to measure any perf effect, so I figure I'll let the bots have at it.

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/ee902cd4f4b3c8f355c0f4de35152a14904ec54a

[mtklein, 2014-09-22 18:18:18]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2014-09-22 18:18:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/595583003/1

[commit-bot: I haz the power, 2014-09-22 18:18:52]

Note for Reviewers:
The CQ is waiting for an approval.
If you believe that the CL is not ready yet, or if you would like to L-G-T-M
with comments then please uncheck the CQ checkbox.

Waiting for LGTM from a valid reviewer till 2014-09-23 00:18 UTC

[mtklein, 2014-09-22 18:31:07]

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

[reed1, 2014-09-22 18:40:06]

This change looks correct.

That said, it does seem like a code/perf burden to track the dirty bits at all.
i.e. this is information that we seem to only ever use during recording (is that
correct?)

I just am wondering out-loud if its worth paying this cost in all our setters
instead of applying the cost more to just recording.

... and I have no numbers at all to back up my concern :)

lgtm

[commit-bot: I haz the power, 2014-09-22 18:40:39]

Committed patchset #1 (id:1) as ee902cd4f4b3c8f355c0f4de35152a14904ec54a

[mtklein, 2014-09-22 18:45:19]

On 2014/09/22 18:40:39, I haz the power (commit-bot) wrote:
> Committed patchset #1 (id:1) as ee902cd4f4b3c8f355c0f4de35152a14904ec54a

Super agreed.  I also don't like having so much recording logic in SkPaint.  I'm
working on variable-length paint encoding for recording now.  I'll start with a
version that doesn't use dirty bits (working from SkPaint's public API) and see
what I can get out of that.  If there's a win or it's no worse, we can drop the
dirty bits entirely.