Produce new-style Cmm from the Cmm parser

The main change here is that the Cmm parser now allows high-level cmm
code with argument-passing and function calls.  For example:

foo ( gcptr a, bits32 b )
{
  if (b > 0) {
     // we can make tail calls passing arguments:
     jump stg_ap_0_fast(a);
  }

  return (x,y);
}

More details on the new cmm syntax are in Note [Syntax of .cmm files]
in CmmParse.y.

The old syntax is still more-or-less supported for those occasional
code fragments that really need to explicitly manipulate the stack.
However there are a couple of differences: it is now obligatory to
give a list of live GlobalRegs on every jump, e.g.

  jump %ENTRY_CODE(Sp(0)) [R1];

Again, more details in Note [Syntax of .cmm files].

I have rewritten most of the .cmm files in the RTS into the new
syntax, except for AutoApply.cmm which is generated by the genapply
program: this file could be generated in the new syntax instead and
would probably be better off for it, but I ran out of enthusiasm.

Some other changes in this batch:

 - The PrimOp calling convention is gone, primops now use the ordinary
   NativeNodeCall convention.  This means that primops and "foreign
   import prim" code must be written in high-level cmm, but they can
   now take more than 10 arguments.

 - CmmSink now does constant-folding (should fix #7219)

 - .cmm files now go through the cmmPipeline, and as a result we
   generate better code in many cases.  All the object files generated
   for the RTS .cmm files are now smaller.  Performance should be
   better too, but I haven't measured it yet.

 - RET_DYN frames are removed from the RTS, lots of code goes away

 - we now have some more canned GC points to cover unboxed-tuples with
   2-4 pointers, which will reduce code size a little.

1 files changed, 32 insertions, 24 deletions

diff --git a/rts/StgStartup.cmm b/rts/StgStartup.cmm
index 4aace82deb..6793913464 100644
--- a/rts/StgStartup.cmm
+++ b/rts/StgStartup.cmm
@@ -35,11 +35,9 @@
    -------------------------------------------------------------------------- */
 
 INFO_TABLE_RET(stg_stop_thread, STOP_FRAME,
-#if defined(PROFILING)
-  W_ unused,
-  W_ unused
-#endif
-)
+               W_ info_ptr,
+               PROF_HDR_FIELDS(W_))
+/* no return list: explicit stack layout */
 {
     /* 
        The final exit.
@@ -75,7 +73,7 @@ INFO_TABLE_RET(stg_stop_thread, STOP_FRAME,
     StgRegTable_rRet(BaseReg) = ThreadFinished;
     R1 = BaseReg;
 
-    jump StgReturn;
+    jump StgReturn [R1];
 }
 
 /* -----------------------------------------------------------------------------
@@ -87,46 +85,57 @@ INFO_TABLE_RET(stg_stop_thread, STOP_FRAME,
    the thread's state away nicely.
    -------------------------------------------------------------------------- */
 
-stg_returnToStackTop
+stg_returnToStackTop /* no args: explicit stack layout */
 {
   LOAD_THREAD_STATE();
   CHECK_SENSIBLE_REGS();
-  jump %ENTRY_CODE(Sp(0));
+  jump %ENTRY_CODE(Sp(0)) [];
 }
 
-stg_returnToSched
+stg_returnToSched /* no args: explicit stack layout */
 {
+  W_ r1;
+  r1 = R1; // foreign calls may clobber R1
   SAVE_THREAD_STATE();
   foreign "C" threadPaused(MyCapability() "ptr", CurrentTSO);
-  jump StgReturn;
+  R1 = r1;
+  jump StgReturn [R1];
 }
 
 // A variant of stg_returnToSched that doesn't call threadPaused() on the
 // current thread.  This is used for switching from compiled execution to the
 // interpreter, where calling threadPaused() on every switch would be too
 // expensive.
-stg_returnToSchedNotPaused
+stg_returnToSchedNotPaused /* no args: explicit stack layout */
 {
   SAVE_THREAD_STATE();
-  jump StgReturn;
+  jump StgReturn [R1];
 }
 
 // A variant of stg_returnToSched, but instead of returning directly to the
 // scheduler, we jump to the code fragment pointed to by R2.  This lets us
 // perform some final actions after making the thread safe, such as unlocking
 // the MVar on which we are about to block in SMP mode.
-stg_returnToSchedButFirst
+stg_returnToSchedButFirst /* no args: explicit stack layout */
 {
+  W_ r1, r2, r3;
+  r1 = R1;
+  r2 = R2;
+  r3 = R3;
   SAVE_THREAD_STATE();
+  // foreign calls may clobber R1/R2/.., so we save them above
   foreign "C" threadPaused(MyCapability() "ptr", CurrentTSO);
-  jump R2;
+  R1 = r1;
+  R2 = r2;
+  R3 = r3;
+  jump R2 [R1,R3];
 }
 
-stg_threadFinished
+stg_threadFinished /* no args: explicit stack layout */
 {
   StgRegTable_rRet(BaseReg) = ThreadFinished;
   R1 = BaseReg;
-  jump StgReturn;
+  jump StgReturn [R1];
 }  
 
 /* -----------------------------------------------------------------------------
@@ -143,31 +152,30 @@ stg_threadFinished
 
     ------------------------------------------------------------------------- */
 
-INFO_TABLE_RET(stg_forceIO, RET_SMALL)
-
+INFO_TABLE_RET(stg_forceIO, RET_SMALL, P_ info_ptr)
+    return (P_ ret)
 {
-  Sp_adj(1);
-  ENTER();
+    ENTER(ret);
 }
 
 /* -----------------------------------------------------------------------------
    Special STG entry points for module registration.
    -------------------------------------------------------------------------- */
 
-stg_init_finish
+stg_init_finish /* no args: explicit stack layout */
 {
-  jump StgReturn;
+  jump StgReturn [];
 }
 
 /* On entry to stg_init:
  *    init_stack[0] = &stg_init_ret;
  *    init_stack[1] = __stginit_Something;
  */
-stg_init
+stg_init /* no args: explicit stack layout */
 {
   W_ next;
   Sp = W_[BaseReg + OFFSET_StgRegTable_rSp];
   next = W_[Sp];
   Sp_adj(1);
-  jump next;
+  jump next [];
 }