Describe spesh plugin mechanism in ChangeLog

All the spesh changes were part of implementing that.

docs/ChangeLog

@@ -43,34 +43,12 @@ Platform:
 + Only use the 64 bit memory access code in memmem32 for cpu's that support
   unaligned accesses.
 
-?Spesh:
-+ Set type and value facts on getattr guard temps
-+ Properly steal deopt annotation
-+ Implement code-gen for attr spesh plugin guards
-+ Fix a memory leak
-+ Allow recursive use of nqp::speshresolve
-+ Only use a deopt table index once in the graph
-+ Set facts on spesh plugin resolvee
-+ Make spesh plugins work correctly with the JIT
-+ Implement spesh plugin concrete/type object deopt
-+ Clear up plugin state on resolver exception
-+ Implement spesh plugin type guard deopt
-+ Correct instruction size in deopt target calc
-+ Emit exact object match guard for spesh plugins
-+ Transform speshresolve into spesh slot lookup
-+ Correct inverted condition
-+ Analyze spesh plugin guard indices for stability
-+ Aggregate spesh plugin resolution statistics
-+ Use unsigned ints in binary search
-+ Fix a silly thinko in guard marking
-+ Write plugin resolution outcomes into spesh log
-+ Missing null check on guard marking
-+ Implement plugin guards on fetched attributes
-+ Concreteness and type guards for spesh plugins
-+ Implement spesh plugin guard recording obj case
-+ Prepare data structures for plugin guards
-+ Implement spesh plugin resolution by position
-+ Implement spesh plugin registration/resolution
+Spesh:
++ Implement a new spesh plugin mechanism, which allows for a mapping of guards
+  to results. The setup of the guards is implemented in code running on MoarVM,
+  and the interpreter then walks the guard table to pick the applicable result.
+  Spesh itself can then look at this table and compile the guards in as needed,
+  together with JIT-compiling them.
 
 Strings:
 + Make MVM_string_utf8_decode 8% faster with vectorized 8bit check