Use a BitArray in RuntimeCSSEnabled instead of a Vector<bool>

Use a BitArray in RuntimeCSSEnabled instead of a Vector<bool>

Use a BitArray in RuntimeCSSEnabled instead of a Vector<bool> to keep track
of which features are enabled at runtime in order to use a little less memory.

This CL also adds a value argument to the WTF::BitArray constructor so that
bits can be initialized to 1 instead of 0,  as RuntimeCSSEnabled expects all
bits to be initially 1. A clear() method is also added so that bits can be
unset and not simply set.

R=eseidel@chromium.org, dstockwell@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=174531

[eseidel, 2014-05-21 00:42:44]

lgtm

I feel like I tried this once and didn't for some reason.

In any case, the perf bots will tell us if this is trouble for some reason.

[commit-bot: I haz the power, 2014-05-21 00:43:12]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/ch.dumez@samsung.com/290793004/1

[dstockwell, 2014-05-21 00:45:40]

Last minute nits, feel free to ignore :P

https://codereview.chromium.org/290793004/diff/1/Source/core/css/RuntimeCSSEn...
File Source/core/css/RuntimeCSSEnabled.cpp (right):

https://codereview.chromium.org/290793004/diff/1/Source/core/css/RuntimeCSSEn...
Source/core/css/RuntimeCSSEnabled.cpp:133: static BitVector* switches = 0;
A BitArray might be more appropriate.

https://codereview.chromium.org/290793004/diff/1/Source/core/css/RuntimeCSSEn...
Source/core/css/RuntimeCSSEnabled.cpp:145: // Values all start at 0. Vector
RELEASE_ASSERTS will catch if we're ever wrong.
This is no longer true, shouldn't be necessary though.

[Inactive, 2014-05-21 00:48:08]

The CQ bit was unchecked by ch.dumez@samsung.com

[Inactive, 2014-05-21 01:04:48]

https://codereview.chromium.org/290793004/diff/1/Source/core/css/RuntimeCSSEn...
File Source/core/css/RuntimeCSSEnabled.cpp (right):

https://codereview.chromium.org/290793004/diff/1/Source/core/css/RuntimeCSSEn...
Source/core/css/RuntimeCSSEnabled.cpp:133: static BitVector* switches = 0;
On 2014/05/21 00:45:41, dstockwell wrote:
> A BitArray might be more appropriate.

Ah, I didn't know we had one. This is more suitable indeed, thanks. I'll update
the CL shortly.

[Inactive, 2014-05-21 02:43:57]

I updated the CL to use a BitArray instead of a BitVector. Please take another
look.

[dstockwell, 2014-05-21 02:46:35]

lgtm

[Inactive, 2014-05-21 02:57:33]

Eric, could you please take another look at wtf/ as well? The code there is
entirely new since I am now extending BitArray instead of BitVector.

[eseidel, 2014-05-21 22:35:30]

lgtm

[commit-bot: I haz the power, 2014-05-22 08:09:48]

Change committed as 174531