Refactor: BLF Use Vector for Glyph Cache List

Replace FontBLF's ListBase with a GlyphCacheListBLF class that uses a
vector of GlyphCacheBLF, holds the ThreadMutex, has methods for
acquire, release, clear. Seems like a simple first step.

@@ -171,0 +148,4 @@
+
+  GlyphCacheBLF *gc = nullptr;
+
+  for (auto &entry : list) {

@@ -107,2 +109,4 @@
 } KerningCacheBLF;
 
+class GlyphCacheListBLF {
+ private:

@@ -108,1 +110,4 @@
 
+class GlyphCacheListBLF {
+ private:
+  std::vector<GlyphCacheBLF *> list;

@@ -109,0 +115,4 @@
+  FontBLF *font;
+
+  /** Mutex lock for glyph cache. */
+  ThreadMutex glyph_cache_mutex;

@@ -108,1 +112,4 @@
 
+class GlyphCacheListBLF {
+  blender::Vector<GlyphCacheBLF *> list;
+  FontBLF *font;