Document USE_RWLOCK macro

(fix of commit 3bfb499)

Issue #473 (bdwgc).

* docs/README.macros (USE_RWLOCK): Document.
* docs/scale.md (Options for enhanced scalability): Document USE_RWLOCK
macro.

docs/README.macros

@@ -411,6 +411,10 @@ GC_USE_DLOPEN_WRAP      Causes the collector to redefine malloc and
   dlopen and dlsym to refer to the original versions.  This makes it possible
   to build an LD_PRELOADable malloc replacement library.
 
+USE_RWLOCK      Use rwlock for the allocator lock instead of mutex.
+  Thus enable usage of the reader (shared) mode of the allocator lock where
+  possible.
+
 THREAD_LOCAL_ALLOC      Defines GC_malloc(), GC_malloc_atomic() and
   GC_gcj_malloc() to use a per-thread set of free-lists. These then allocate
   in a way that usually does not involve acquisition of the allocator (global)

docs/scale.md

@@ -61,6 +61,13 @@ This _reduces performance_ for the standard allocation functions, though
 it usually improves performance when thread-local allocation is used heavily,
 and, thus, the number of short-duration lock acquisitions is greatly reduced.
 
+Also, `USE_RWLOCK` macro (experimental) should be noted which changes the
+allocator lock implementation base from a mutex (`CRITICAL_SECTION` in case
+of Win32) to `pthread_rwlock_t` (`SRWLOCK`, respectively), thus enabling
+acquisition of a slim lock in the reader (shared) mode where possible.  See
+the description of `GC_call_with_reader_lock` and `GC_REVEAL_POINTER` entities
+in `gc.h` for more details.
+
 ## The Parallel Marking Algorithm
 
 We use an algorithm similar to