Refactor the story around switches (#10137)

This re-implements the code generation for case expressions at the Stg →
Cmm level, both for data type cases as well as for integral literal
cases. (Cases on float are still treated as before).

The goal is to allow for fancier strategies in implementing them, for a
cleaner separation of the strategy from the gritty details of Cmm, and
to run this later than the Common Block Optimization, allowing for one
way to attack #10124. The new module CmmSwitch contains a number of
notes explaining this changes. For example, it creates larger
consecutive jump tables than the previous code, if possible.

nofib shows little significant overall improvement of runtime. The
rather large wobbling comes from changes in the code block order
(see #8082, not much we can do about it). But the decrease in code size
alone makes this worthwhile.

```
        Program           Size    Allocs   Runtime   Elapsed  TotalMem
            Min          -1.8%      0.0%     -6.1%     -6.1%     -2.9%
            Max          -0.7%     +0.0%     +5.6%     +5.7%     +7.8%
 Geometric Mean          -1.4%     -0.0%     -0.3%     -0.3%     +0.0%
```

Compilation time increases slightly:
```
        -1 s.d.                -----            -2.0%
        +1 s.d.                -----            +2.5%
        Average                -----            +0.3%
```

The test case T783 regresses a lot, but it is the only one exhibiting
any regression. The cause is the changed order of branches in an
if-then-else tree, which makes the hoople data flow analysis traverse
the blocks in a suboptimal order. Reverting that gets rid of this
regression, but has a consistent, if only very small (+0.2%), negative
effect on runtime. So I conclude that this test is an extreme outlier
and no reason to change the code.

Differential Revision: https://phabricator.haskell.org/D720

1 files changed, 10 insertions, 15 deletions

diff --git a/compiler/cmm/PprC.hs b/compiler/cmm/PprC.hs
index a2c3abf320..92c818242d 100644
--- a/compiler/cmm/PprC.hs
+++ b/compiler/cmm/PprC.hs
@@ -33,6 +33,7 @@ import Cmm hiding (pprBBlock)
 import PprCmm ()
 import Hoopl
 import CmmUtils
+import CmmSwitch
 
 -- Utils
 import CPrim
@@ -299,21 +300,12 @@ pprCondBranch expr yes no
 --
 -- we find the fall-through cases
 --
--- N.B. we remove Nothing's from the list of branches, as they are
--- 'undefined'. However, they may be defined one day, so we better
--- document this behaviour.
---
-pprSwitch :: DynFlags -> CmmExpr -> [ Maybe BlockId ] -> SDoc
-pprSwitch dflags e maybe_ids
-  = let pairs  = [ (ix, ident) | (ix,Just ident) <- zip [0..] maybe_ids ]
-        pairs2 = [ (map fst as, snd (head as)) | as <- groupBy sndEq pairs ]
-    in
-        (hang (ptext (sLit "switch") <+> parens ( pprExpr e ) <+> lbrace)
-                4 (vcat ( map caseify pairs2 )))
-        $$ rbrace
-
+pprSwitch :: DynFlags -> CmmExpr -> SwitchTargets -> SDoc
+pprSwitch dflags e ids
+  = (hang (ptext (sLit "switch") <+> parens ( pprExpr e ) <+> lbrace)
+                4 (vcat ( map caseify pairs ) $$ def)) $$ rbrace
   where
-    sndEq (_,x) (_,y) = x == y
+    (pairs, mbdef) = switchTargetsFallThrough ids
 
     -- fall through case
     caseify (ix:ixs, ident) = vcat (map do_fallthrough ixs) $$ final_branch ix
@@ -326,7 +318,10 @@ pprSwitch dflags e maybe_ids
                 hsep [ ptext (sLit "case") , pprHexVal ix (wordWidth dflags) <> colon ,
                        ptext (sLit "goto") , (pprBlockId ident) <> semi ]
 
-    caseify (_     , _    ) = panic "pprSwtich: swtich with no cases!"
+    caseify (_     , _    ) = panic "pprSwitch: switch with no cases!"
+
+    def | Just l <- mbdef = ptext (sLit "default: goto") <+> pprBlockId l <> semi
+        | otherwise       = empty
 
 -- ---------------------------------------------------------------------
 -- Expressions.