Fix more clang errors.

Fix more clang errors.

Fixed -Wunused-private-field errors.

R=reed@google.com,bsalomon@google.com

Committed: https://code.google.com/p/skia/source/detail?r=8547

[tfarina, 2013-04-03 00:11:05]

You will notice that the changes in bench/MatrixBench.cpp are controversial, but
it seems clang is not smart enough to expand the macro and see that fCount is
actually used?

[tfarina, 2013-04-03 00:15:01]

This is the clang report:

../../bench/MatrixBench.cpp:308:79: error: member initializer 'fCount' does not
name a non-static data member or base class
    ScaleTransDoubleMatrixBench(void* p) : INHERITED(p, "scaletrans_double"),
fCount (16) {
                                                                             
^~~~~~~~~~~
../../bench/MatrixBench.cpp:301:9: error: private field 'fCount' is not used
[-Werror,-Wunused-private-field]

[bsalomon, 2013-04-03 00:31:24]

On 2013/04/03 00:15:01, tfarina wrote:
> This is the clang report:
> 
> ../../bench/MatrixBench.cpp:308:79: error: member initializer 'fCount' does
not
> name a non-static data member or base class
>     ScaleTransDoubleMatrixBench(void* p) : INHERITED(p, "scaletrans_double"),
> fCount (16) {
>                                                                              
> ^~~~~~~~~~~
> ../../bench/MatrixBench.cpp:301:9: error: private field 'fCount' is not used
> [-Werror,-Wunused-private-field]

Can we replace the declarations of fCount in the first two files with:

enum {
    kCount = SkBENCHLOOP(fCount)
};

and the use kCount in the loops?

[tfarina, 2013-04-03 00:41:21]

On 2013/04/03 00:31:24, bsalomon wrote:
> Can we replace the declarations of fCount in the first two files with:
> 
> enum {
>     kCount = SkBENCHLOOP(fCount)
> };
> 
> and the use kCount in the loops?

Almost, still using fCount generates this error:
../../bench/MatrixBench.cpp:304:9: error: private field 'fCount' is not used
[-Werror,-Wunused-private-field]
    int fCount;

Patch set 2 is slight adaptation of your version that passes on clang.

[tfarina, 2013-04-04 00:54:49]

Guys, does patch set 2 looks good??

[bsalomon, 2013-04-04 00:56:31]

On 2013/04/04 00:54:49, tfarina wrote:
> Guys, does patch set 2 looks good??

lgtm

[tfarina, 2013-04-05 21:13:27]

Committed patchset #2 manually as r8547 (presubmit successful).