New SkRTConf macro SK_CONF_TRY_SET: no complaint on missing configuration

New SkRTConf macro SK_CONF_TRY_SET: no complaint on missing configuration

SK_CONF_TRY_SET() is like SK_CONF_SET(), but doesn't complain if
confname can't be found.  This is useful if the SK_CONF_DECLARE is
inside a source file whose linkage is dependent on the system.

Internally to the SkRTConf system, SkRTConfRegistry::set() was given
an additional parameter controling wanrings.

A new RuntimeConfig unit test was introduced.  It should run silently.
In the future, it should be expanded to cover all of the SkRTConf
functionality.

(For example, the images.jpeg.suppressDecoderWarnings variable is
defined and used only in SkImageDecoder_libjpeg.cpp, but on MacOS, we
use Core Graphics via SkImageDecoder_CG.cpp - SkImageDecoder_libjpeg
is never linked in.  The same is true of the Windows Imaging Component
on Windows.)

BUG=
R=reed@google.com

Committed: https://code.google.com/p/skia/source/detail?r=12155

[hal.canary, 2013-11-01 15:35:28]

Cary, this is a fix for the bug you reported to me with SK_CONF_SET.

[caryclark, 2013-11-01 16:57:36]

lgtm (with humper's approval)

https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp
File tests/RuntimeConfigTest.cpp (right):

https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp#n...
tests/RuntimeConfigTest.cpp:19: #ifdef SK_DEVELOPER
I find this difficult to parse. Maybe outdent the #if lines?

[hal.canary, 2013-11-01 17:01:18]

https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp
File tests/RuntimeConfigTest.cpp (right):

https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp#n...
tests/RuntimeConfigTest.cpp:19: #ifdef SK_DEVELOPER
On 2013/11/01 16:57:36, caryclark wrote:
> I find this difficult to parse. Maybe outdent the #if lines?

Done.

[humper, 2013-11-01 19:19:35]

On 2013/11/01 17:01:18, halcanary wrote:
> https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp
> File tests/RuntimeConfigTest.cpp (right):
> 
>
https://codereview.chromium.org/54503007/diff/1/tests/RuntimeConfigTest.cpp#n...
> tests/RuntimeConfigTest.cpp:19: #ifdef SK_DEVELOPER
> On 2013/11/01 16:57:36, caryclark wrote:
> > I find this difficult to parse. Maybe outdent the #if lines?
> 
> Done.

Looks really valuable, LGTM

[hal.canary, 2013-11-06 14:22:41]

> ** Presubmit ERRORS **
> Since the CL is editing public API, you must have an LGTM
> from one of: ('reed@chromium.org', 'reed@google.com',
> 'bsalomon@chromium.org', 'bsalomon@google.com')

Brian or Mike, can you please take a look; thanks!

[reed1, 2013-11-06 15:02:20]

lgtm

https://codereview.chromium.org/54503007/diff/80001/include/utils/SkRTConf.h
File include/utils/SkRTConf.h (right):

https://codereview.chromium.org/54503007/diff/80001/include/utils/SkRTConf.h#...
include/utils/SkRTConf.h:55: #define SK_CONF_SET(confname, value) \
at some point we should clean-up the readability of this block of macros. add
whitespace? more std formatted comments?

[hal.canary, 2013-11-06 15:07:47]

Committed patchset #2 manually as r12155 (presubmit successful).