Merging in the new codegen branch

This merge does not turn on the new codegen (which only compiles
a select few programs at this point),
but it does introduce some changes to the old code generator.

The high bits:
1. The Rep Swamp patch is finally here.
   The highlight is that the representation of types at the
   machine level has changed.
   Consequently, this patch contains updates across several back ends.
2. The new Stg -> Cmm path is here, although it appears to have a
   fair number of bugs lurking.
3. Many improvements along the CmmCPSZ path, including:
   o stack layout
   o some code for infotables, half of which is right and half wrong
   o proc-point splitting

1 files changed, 5 insertions, 2 deletions

diff --git a/compiler/codeGen/CgTailCall.lhs b/compiler/codeGen/CgTailCall.lhs
index 475196abba..4f890998ae 100644
--- a/compiler/codeGen/CgTailCall.lhs
+++ b/compiler/codeGen/CgTailCall.lhs
@@ -41,6 +41,7 @@ import Type
 import Id
 import StgSyn
 import PrimOp
+import FastString
 import Outputable
 
 import Control.Monad
@@ -116,7 +117,7 @@ performTailCall fun_info arg_amodes pending_assts
 	; EndOfBlockInfo sp _ <- getEndOfBlockInfo
 	; this_pkg <- getThisPackage
 
-	; case (getCallMethod fun_name lf_info (length arg_amodes)) of
+	; case (getCallMethod fun_name fun_has_cafs lf_info (length arg_amodes)) of
 
 	    -- Node must always point to things we enter
 	    EnterIt -> do
@@ -183,8 +184,10 @@ performTailCall fun_info arg_amodes pending_assts
 	        }
 	}
   where
-    fun_name  = idName (cgIdInfoId fun_info)
+    fun_id    = cgIdInfoId fun_info
+    fun_name  = idName fun_id
     lf_info   = cgIdInfoLF fun_info
+    fun_has_cafs = idCafInfo fun_id
     untag_node = CmmAssign nodeReg (cmmUntag (CmmReg nodeReg))
     -- Test if closure is a constructor
     maybeSwitchOnCons enterClosure eob