simple patch to always init SkTextBlob uniqueID

simple patch to always init SkTextBlob uniqueID

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/2af858354d913397a6c316ef46a5d52d686e10ab

[joshualitt, 2015-03-24 20:28:54]

joshualitt@google.com changed reviewers:
+ fmalita@chromium.org, joshualitt@google.com, mtklein@google.com,
reed@google.com

[joshualitt, 2015-03-24 20:28:55]

Simple patch to always init SkTextBlob's uniqueID.  It means everyone has to pay
the price for the atomic_inc, but we can also inline the getter which I think
will more than offset the cost overtime.

[mtklein, 2015-03-24 20:45:08]

https://codereview.chromium.org/1036613002/diff/1/include/core/SkTextBlob.h
File include/core/SkTextBlob.h (right):

https://codereview.chromium.org/1036613002/diff/1/include/core/SkTextBlob.h#n...
include/core/SkTextBlob.h:101: mutable uint32_t fUniqueID;
mutable -> const ?

https://codereview.chromium.org/1036613002/diff/1/src/core/SkTextBlob.cpp
File src/core/SkTextBlob.cpp (right):

https://codereview.chromium.org/1036613002/diff/1/src/core/SkTextBlob.cpp#new...
src/core/SkTextBlob.cpp:114: , fUniqueID(SK_InvalidGenID) {
let's do

static int32_t gNextID = 1;
static int32_t next_id() {
    int32_t id;
    do { 
        id = sk_atomic_inc(&gNextID);
    } while (id == SK_InvalidGenID);
    return id;
}

....

    , fUniqueID(next_id())  {}

That lets fUniqueID be const.

[joshualitt, 2015-03-24 21:00:21]

On 2015/03/24 20:45:08, mtklein wrote:
> https://codereview.chromium.org/1036613002/diff/1/include/core/SkTextBlob.h
> File include/core/SkTextBlob.h (right):
> 
>
https://codereview.chromium.org/1036613002/diff/1/include/core/SkTextBlob.h#n...
> include/core/SkTextBlob.h:101: mutable uint32_t fUniqueID;
> mutable -> const ?
> 
> https://codereview.chromium.org/1036613002/diff/1/src/core/SkTextBlob.cpp
> File src/core/SkTextBlob.cpp (right):
> 
>
https://codereview.chromium.org/1036613002/diff/1/src/core/SkTextBlob.cpp#new...
> src/core/SkTextBlob.cpp:114: , fUniqueID(SK_InvalidGenID) {
> let's do
> 
> static int32_t gNextID = 1;
> static int32_t next_id() {
>     int32_t id;
>     do { 
>         id = sk_atomic_inc(&gNextID);
>     } while (id == SK_InvalidGenID);
>     return id;
> }
> 
> ....
> 
>     , fUniqueID(next_id())  {}
> 
> That lets fUniqueID be const.

feedback incorporated

[mtklein, 2015-03-24 21:07:50]

lgtm

[f(malita), 2015-03-24 21:30:31]

The CQ bit was checked by fmalita@chromium.org

[f(malita), 2015-03-24 21:30:31]

lgtm

[f(malita), 2015-03-24 21:30:34]

The CQ bit was unchecked by fmalita@chromium.org

[joshualitt, 2015-03-25 14:15:12]

On 2015/03/24 21:30:31, f(malita) wrote:
> lgtm

@reed, can I get an API l-g-t-m?

[reed1, 2015-03-25 20:23:30]

lgtm

[joshualitt, 2015-03-25 20:25:24]

The CQ bit was checked by joshualitt@google.com

[commit-bot: I haz the power, 2015-03-25 20:25:33]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1036613002/20001

[commit-bot: I haz the power, 2015-03-25 20:40:20]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/2af858354d913397a6c316ef46a5d52d686e10ab