Add sk_parallel_for()

This should be a drop-in replacement for most for-loops to make them run in parallel:
   for (int i = 0; i < N; i++) { code... }
   ~~~>
   sk_parallel_for(N, [&](int i) { code... });

This is just syntax sugar over SkTaskGroup to make this use case really easy to write.
There's no more overhead that we weren't already forced to add using an interface like batch(),
and no extra heap allocations.

I've replaced 3 uses of SkTaskGroup with sk_parallel_for:
  1) My unit tests for SkOnce.
  2) Cary's path fuzzer.
  3) SkMultiPictureDraw.
Performance should be the same.  Please compare left and right for readability. :)

BUG=skia:

No public API changes.
TBR=reed@google.com

Review URL: https://codereview.chromium.org/1184373003

1 files changed, 15 insertions, 5 deletions

diff --git a/src/core/SkTaskGroup.cpp b/src/core/SkTaskGroup.cpp
index 59319c148d..4504defb1b 100644
--- a/src/core/SkTaskGroup.cpp
+++ b/src/core/SkTaskGroup.cpp
@@ -5,6 +5,7 @@
  * found in the LICENSE file.
  */
 
+#include "SkOnce.h"
 #include "SkRunnable.h"
 #include "SkSemaphore.h"
 #include "SkSpinlock.h"
@@ -13,18 +14,27 @@
 #include "SkThreadUtils.h"
 
 #if defined(SK_BUILD_FOR_WIN32)
-    static inline int num_cores() {
+    static void query_num_cores(int* num_cores) {
         SYSTEM_INFO sysinfo;
         GetSystemInfo(&sysinfo);
-        return sysinfo.dwNumberOfProcessors;
+        *num_cores = sysinfo.dwNumberOfProcessors;
     }
 #else
     #include <unistd.h>
-    static inline int num_cores() {
-        return (int) sysconf(_SC_NPROCESSORS_ONLN);
+    static void query_num_cores(int* num_cores) {
+        *num_cores = (int)sysconf(_SC_NPROCESSORS_ONLN);
     }
 #endif
 
+// We cache sk_num_cores() so we only query the OS once.
+SK_DECLARE_STATIC_ONCE(g_query_num_cores_once);
+int sk_num_cores() {
+    static int num_cores = 0;
+    SkOnce(&g_query_num_cores_once, query_num_cores, &num_cores);
+    SkASSERT(num_cores > 0);
+    return num_cores;
+}
+
 namespace {
 
 class ThreadPool : SkNoncopyable {
@@ -98,7 +108,7 @@ private:
 
     explicit ThreadPool(int threads) {
         if (threads == -1) {
-            threads = num_cores();
+            threads = sk_num_cores();
         }
         for (int i = 0; i < threads; i++) {
             fThreads.push(SkNEW_ARGS(SkThread, (&ThreadPool::Loop, this)));