--- gforth/profile.fs	2004/08/23 14:03:52	1.2
+++ gforth/profile.fs	2004/09/03 14:04:47	1.3
@@ -28,13 +28,30 @@
 
 true constant count-calls? \ do some profiling of colon definitions etc.
 
+\ for true COUNT-CALLS?:
+
+\ What data do I need for evaluating the effectiveness of (partial) inlining?
+
+\ static and dynamic counts of everything:
+
+\ original BB length (histogram and average)
+\ BB length with partial inlining (histogram and average)
+\   since we cannot partially inline library calls, we use a parameter
+\   that represents the amount of partial inlining we can expect there.
+\ number of tail calls (original and after partial inlining)
+\ number of calls (original and after partial inlining)
+\ reason for BB end: call, return, execute, branch
+
+\ how many static calls are there to a word?  How many of the dynamic
+\ calls call just a single word?
+
 struct
     cell%    field profile-next
     cell% 2* field profile-count
     cell% 2* field profile-sourcepos
     cell%    field profile-char \ character position in line
     count-calls? [if]
-	cell% field profile-colondef? \ is this a colon definition
+	cell% field profile-colondef? \ is this a colon definition start
 	cell% field profile-calls \ static calls to the colon def
 	cell% field profile-straight-line \ may contain calls, but no other CF
 	cell% field profile-calls-from \ static calls in the colon def
@@ -155,8 +172,8 @@ Defer before-word-profile ( -- )
     r> ! ; \ change hook behaviour
 
 hook-profiling-into then-like
-\ hook-profiling-into if-like
-\ hook-profiling-into ahead-like
+\ hook-profiling-into if-like    \ subsumed by other-control-flow
+\ hook-profiling-into ahead-like \ subsumed by other-control-flow
 hook-profiling-into other-control-flow
 hook-profiling-into begin-like
 hook-profiling-into again-like