Make the GPU feature info code more compact.

Make the GPU feature info code more compact.

AddExtensionString is called quite a lot, always with a constant
string literal, which makes it a bit unnecessary to convert the
argument to a std::string before every function call.

This small change cuts away 7 KB of the machine code:

Total change: -7208 bytes
=========================
  1 added, totalling +559 bytes across 1 sources
  1 removed, totalling -471 bytes across 1 sources
  1 shrunk, for a net change of -7296 bytes (25422 bytes before, 18126 bytes after) across 1 sources
  267984 unchanged, totalling 50299073 bytes
Per-source Analysis:

--------------------------------------------------------------------------------------------------------------------
 -7208 - Source: /home/bratell/src/chromium/src/gpu/command_buffer/service/feature_info.cc - (gained 559, lost 7767)
--------------------------------------------------------------------------------------------------------------------
  New symbols:
       +559: gpu::gles2::FeatureInfo::AddExtensionString(char const*) type=t, size=559 bytes
  Removed symbols:
       -471: gpu::gles2::FeatureInfo::AddExtensionString(std::string const&) type=t, size=471 bytes
  Shrunk symbols:
      -7296: gpu::gles2::FeatureInfo::InitializeFeatures() type=t, (was 25422 bytes, now 18126 bytes)

BUG=

Committed: https://crrev.com/6a4eb0c42d3f1b29f82b7e49f33b89ea6eadc6a9
Cr-Commit-Position: refs/heads/master@{#296785}

[Daniel Bratell, 2014-09-15 13:20:16]

bratell@opera.com changed reviewers:
+ bajones@chromium.org

[Daniel Bratell, 2014-09-18 15:42:28]

bajones, can you take a look at this? Just something I tripped over when looking
for suspiciously large code.

[bajones, 2014-09-18 15:51:18]

On 2014/09/18 at 15:42:28, bratell wrote:
> bajones, can you take a look at this? Just something I tripped over when
looking for suspiciously large code.

I'm surprised this makes that much of a difference! Didn't realize we called the
function that frequently.

I don't see any reason why this would ever cause us problem, so LGTM.

[Daniel Bratell, 2014-09-25 17:16:48]

The CQ bit was checked by bratell@opera.com

[Daniel Bratell, 2014-09-25 17:16:56]

The CQ bit was checked by bratell@opera.com

[commit-bot: I haz the power, 2014-09-25 17:18:31]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/571073003/1

[commit-bot: I haz the power, 2014-09-25 20:45:47]

Committed patchset #1 (id:1) as e9eea0f241f2e9801beb4691ccedaa8a43161967

[commit-bot: I haz the power, 2014-09-25 20:46:31]

Patchset 1 (id:??) landed as
https://crrev.com/6a4eb0c42d3f1b29f82b7e49f33b89ea6eadc6a9
Cr-Commit-Position: refs/heads/master@{#296785}