CmmOpt.hs 23.9 KB
Newer Older
1 2 3 4 5 6 7
{-# OPTIONS -w #-}
-- The above warning supression flag is a temporary kludge.
-- While working on this module you are encouraged to remove it and fix
-- any warnings in the module. See
--     http://hackage.haskell.org/trac/ghc/wiki/Commentary/CodingStyle#Warnings
-- for details

8 9 10 11 12 13 14 15 16
-----------------------------------------------------------------------------
--
-- Cmm optimisation
--
-- (c) The University of Glasgow 2006
--
-----------------------------------------------------------------------------

module CmmOpt (
17
	cmmEliminateDeadBlocks,
18 19 20 21 22 23 24
	cmmMiniInline,
	cmmMachOpFold,
	cmmLoopifyForC,
 ) where

#include "HsVersions.h"

25
import OldCmm
Simon Marlow's avatar
Simon Marlow committed
26 27
import CmmUtils
import CLabel
28
import StaticFlags
29 30

import UniqFM
Simon Marlow's avatar
Simon Marlow committed
31
import Unique
32
import FastTypes
33
import Outputable
34
import BlockId
35

Simon Marlow's avatar
Simon Marlow committed
36 37 38
import Data.Bits
import Data.Word
import Data.Int
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
import Data.Maybe

import Compiler.Hoopl hiding (Unique)

-- -----------------------------------------------------------------------------
-- Eliminates dead blocks

{-
We repeatedly expand the set of reachable blocks until we hit a
fixpoint, and then prune any blocks that were not in this set.  This is
actually a required optimization, as dead blocks can cause problems
for invariants in the linear register allocator (and possibly other
places.)
-}

-- Deep fold over statements could probably be abstracted out, but it
-- might not be worth the effort since OldCmm is moribund
cmmEliminateDeadBlocks :: [CmmBasicBlock] -> [CmmBasicBlock]
cmmEliminateDeadBlocks [] = []
cmmEliminateDeadBlocks blocks@(BasicBlock base_id _:_) =
    let -- Calculate what's reachable from what block
        -- We have to do a deep fold into CmmExpr because
        -- there may be a BlockId in the CmmBlock literal.
        reachableMap = foldl f emptyBlockMap blocks
            where f m (BasicBlock block_id stmts) = mapInsert block_id (reachableFrom stmts) m
        reachableFrom stmts = foldl stmt emptyBlockSet stmts
            where
                stmt m CmmNop = m
                stmt m (CmmComment _) = m
                stmt m (CmmAssign _ e) = expr m e
                stmt m (CmmStore e1 e2) = expr (expr m e1) e2
                stmt m (CmmCall c _ as _ _) = f (actuals m as) c
                    where f m (CmmCallee e _) = expr m e
                          f m (CmmPrim _) = m
                stmt m (CmmBranch b) = setInsert b m
                stmt m (CmmCondBranch e b) = setInsert b (expr m e)
                stmt m (CmmSwitch e bs) = foldl (flip setInsert) (expr m e) (catMaybes bs)
                stmt m (CmmJump e as) = expr (actuals m as) e
                stmt m (CmmReturn as) = actuals m as
                actuals m as = foldl (\m h -> expr m (hintlessCmm h)) m as
                expr m (CmmLit l) = lit m l
                expr m (CmmLoad e _) = expr m e
                expr m (CmmReg _) = m
                expr m (CmmMachOp _ es) = foldl expr m es
                expr m (CmmStackSlot _ _) = m
                expr m (CmmRegOff _ _) = m
                lit m (CmmBlock b) = setInsert b m
                lit m _ = m
        -- Expand reachable set until you hit fixpoint
        initReachable = setSingleton base_id :: BlockSet
        expandReachable old_set new_set =
            if setSize new_set > setSize old_set
                then expandReachable new_set $ setFold
                        (\x s -> maybe setEmpty id (mapLookup x reachableMap) `setUnion` s)
                        new_set
                        (setDifference new_set old_set)
                else new_set -- fixpoint achieved
        reachable = expandReachable setEmpty initReachable
    in filter (\(BasicBlock block_id _) -> setMember block_id reachable) blocks
98 99 100 101

-- -----------------------------------------------------------------------------
-- The mini-inliner

Simon Marlow's avatar
Simon Marlow committed
102 103 104 105 106 107 108 109 110 111 112 113
{-
This pass inlines assignments to temporaries that are used just
once.  It works as follows:

  - count uses of each temporary
  - for each temporary that occurs just once:
	- attempt to push it forward to the statement that uses it
        - only push forward past assignments to other temporaries
	  (assumes that temporaries are single-assignment)
	- if we reach the statement that uses it, inline the rhs
	  and delete the original assignment.

114 115 116 117
[N.B. In the Quick C-- compiler, this optimization is achieved by a
 combination of two dataflow passes: forward substitution (peephole
 optimization) and dead-assignment elimination.  ---NR]

Simon Marlow's avatar
Simon Marlow committed
118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
Possible generalisations: here is an example from factorial

Fac_zdwfac_entry:
    cmG:
        _smi = R2;
        if (_smi != 0) goto cmK;
        R1 = R3;
        jump I64[Sp];
    cmK:
        _smn = _smi * R3;
        R2 = _smi + (-1);
        R3 = _smn;
        jump Fac_zdwfac_info;

We want to inline _smi and _smn.  To inline _smn:

   - we must be able to push forward past assignments to global regs.
     We can do this if the rhs of the assignment we are pushing
     forward doesn't refer to the global reg being assigned to; easy
     to test.

To inline _smi:

   - It is a trivial replacement, reg for reg, but it occurs more than
     once.
   - We can inline trivial assignments even if the temporary occurs
     more than once, as long as we don't eliminate the original assignment
     (this doesn't help much on its own).
   - We need to be able to propagate the assignment forward through jumps;
     if we did this, we would find that it can be inlined safely in all
     its occurrences.
-}

151 152 153 154
countUses :: UserOfLocalRegs a => a -> UniqFM Int
countUses a = foldRegsUsed (\m r -> addToUFM m r (count m r + 1)) emptyUFM a
  where count m r = lookupWithDefaultUFM m (0::Int) r

155 156
cmmMiniInline :: [CmmBasicBlock] -> [CmmBasicBlock]
cmmMiniInline blocks = map do_inline blocks 
157 158
  where do_inline (BasicBlock id stmts)
          = BasicBlock id (cmmMiniInlineStmts (countUses blocks) stmts)
159 160 161

cmmMiniInlineStmts :: UniqFM Int -> [CmmStmt] -> [CmmStmt]
cmmMiniInlineStmts uses [] = []
162
cmmMiniInlineStmts uses (stmt@(CmmAssign (CmmLocal (LocalReg u _)) expr) : stmts)
163 164 165 166 167
        -- not used at all: just discard this assignment
  | Nothing <- lookupUFM uses u
  = cmmMiniInlineStmts uses stmts

        -- used once: try to inline at the use site
168 169 170 171 172 173 174 175 176 177 178
  | Just 1 <- lookupUFM uses u,
    Just stmts' <- lookForInline u expr stmts
  = 
#ifdef NCG_DEBUG
     trace ("nativeGen: inlining " ++ showSDoc (pprStmt stmt)) $
#endif
     cmmMiniInlineStmts uses stmts'

cmmMiniInlineStmts uses (stmt:stmts)
  = stmt : cmmMiniInlineStmts uses stmts

179 180 181 182
lookForInline u expr stmts = lookForInline' u expr regset stmts
    where regset = foldRegsUsed extendRegSet emptyRegSet expr

lookForInline' u expr regset (stmt : rest)
183 184 185 186
  | Just 1 <- lookupUFM (countUses stmt) u, ok_to_inline
  = Just (inlineStmt u expr stmt : rest)

  | ok_to_skip
187
  = case lookForInline' u expr regset rest of
188 189 190 191 192
           Nothing    -> Nothing
           Just stmts -> Just (stmt:stmts)

  | otherwise 
  = Nothing
193

194 195 196 197 198 199 200 201 202
  where
	-- we don't inline into CmmCall if the expression refers to global
	-- registers.  This is a HACK to avoid global registers clashing with
	-- C argument-passing registers, really the back-end ought to be able
	-- to handle it properly, but currently neither PprC nor the NCG can
	-- do it.  See also CgForeignCall:load_args_into_temps.
    ok_to_inline = case stmt of
		     CmmCall{} -> hasNoGlobalRegs expr
		     _ -> True
203

204 205 206 207 208 209 210 211
   -- Expressions aren't side-effecting.  Temporaries may or may not
   -- be single-assignment depending on the source (the old code
   -- generator creates single-assignment code, but hand-written Cmm
   -- and Cmm from the new code generator is not single-assignment.)
   -- So we do an extra check to make sure that the register being
   -- changed is not one we were relying on.  I don't know how much of a
   -- performance hit this is (we have to create a regset for every
   -- instruction.) -- EZY
212 213
    ok_to_skip = case stmt of
                 CmmNop -> True
214
                 CmmComment{} -> True
215
                 CmmAssign (CmmLocal r@(LocalReg u' _)) rhs | u' /= u && not (r `elemRegSet` regset) -> True
216 217 218 219
                 CmmAssign g@(CmmGlobal _) rhs -> not (g `regUsedIn` expr)
                 _other -> False


220 221 222
inlineStmt :: Unique -> CmmExpr -> CmmStmt -> CmmStmt
inlineStmt u a (CmmAssign r e) = CmmAssign r (inlineExpr u a e)
inlineStmt u a (CmmStore e1 e2) = CmmStore (inlineExpr u a e1) (inlineExpr u a e2)
223 224
inlineStmt u a (CmmCall target regs es srt ret)
   = CmmCall (infn target) regs es' srt ret
225
   where infn (CmmCallee fn cconv) = CmmCallee (inlineExpr u a fn) cconv
226
	 infn (CmmPrim p) = CmmPrim p
227
	 es' = [ (CmmHinted (inlineExpr u a e) hint) | (CmmHinted e hint) <- es ]
228 229 230 231 232 233
inlineStmt u a (CmmCondBranch e d) = CmmCondBranch (inlineExpr u a e) d
inlineStmt u a (CmmSwitch e d) = CmmSwitch (inlineExpr u a e) d
inlineStmt u a (CmmJump e d) = CmmJump (inlineExpr u a e) d
inlineStmt u a other_stmt = other_stmt

inlineExpr :: Unique -> CmmExpr -> CmmExpr -> CmmExpr
234
inlineExpr u a e@(CmmReg (CmmLocal (LocalReg u' _)))
235 236
  | u == u' = a
  | otherwise = e
237 238
inlineExpr u a e@(CmmRegOff (CmmLocal (LocalReg u' rep)) off)
  | u == u' = CmmMachOp (MO_Add width) [a, CmmLit (CmmInt (fromIntegral off) width)]
239
  | otherwise = e
240 241
  where
    width = typeWidth rep
242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
inlineExpr u a (CmmLoad e rep) = CmmLoad (inlineExpr u a e) rep
inlineExpr u a (CmmMachOp op es) = CmmMachOp op (map (inlineExpr u a) es)
inlineExpr u a other_expr = other_expr

-- -----------------------------------------------------------------------------
-- MachOp constant folder

-- Now, try to constant-fold the MachOps.  The arguments have already
-- been optimized and folded.

cmmMachOpFold
    :: MachOp	    	-- The operation from an CmmMachOp
    -> [CmmExpr]   	-- The optimized arguments
    -> CmmExpr

cmmMachOpFold op arg@[CmmLit (CmmInt x rep)]
  = case op of
      MO_S_Neg r -> CmmLit (CmmInt (-x) rep)
      MO_Not r   -> CmmLit (CmmInt (complement x) rep)

	-- these are interesting: we must first narrow to the 
	-- "from" type, in order to truncate to the correct size.
	-- The final narrow/widen to the destination type
	-- is implicit in the CmmLit.
266 267 268
      MO_SF_Conv from to -> CmmLit (CmmFloat (fromInteger x) to)
      MO_SS_Conv from to -> CmmLit (CmmInt (narrowS from x) to)
      MO_UU_Conv from to -> CmmLit (CmmInt (narrowU from x) to)
269 270 271 272 273

      _ -> panic "cmmMachOpFold: unknown unary op"


-- Eliminate conversion NOPs
274 275
cmmMachOpFold (MO_SS_Conv rep1 rep2) [x] | rep1 == rep2 = x
cmmMachOpFold (MO_UU_Conv rep1 rep2) [x] | rep1 == rep2 = x
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293

-- Eliminate nested conversions where possible
cmmMachOpFold conv_outer args@[CmmMachOp conv_inner [x]]
  | Just (rep1,rep2,signed1) <- isIntConversion conv_inner,
    Just (_,   rep3,signed2) <- isIntConversion conv_outer
  = case () of
	-- widen then narrow to the same size is a nop
      _ | rep1 < rep2 && rep1 == rep3 -> x
	-- Widen then narrow to different size: collapse to single conversion
	-- but remember to use the signedness from the widening, just in case
	-- the final conversion is a widen.
	| rep1 < rep2 && rep2 > rep3 ->
	    cmmMachOpFold (intconv signed1 rep1 rep3) [x]
	-- Nested widenings: collapse if the signedness is the same
	| rep1 < rep2 && rep2 < rep3 && signed1 == signed2 ->
	    cmmMachOpFold (intconv signed1 rep1 rep3) [x]
	-- Nested narrowings: collapse
	| rep1 > rep2 && rep2 > rep3 ->
294
	    cmmMachOpFold (MO_UU_Conv rep1 rep3) [x]
295 296 297
	| otherwise ->
	    CmmMachOp conv_outer args
  where
298
	isIntConversion (MO_UU_Conv rep1 rep2) 
299
	  = Just (rep1,rep2,False)
300
	isIntConversion (MO_SS_Conv rep1 rep2)
301 302 303
	  = Just (rep1,rep2,True)
	isIntConversion _ = Nothing

304 305
	intconv True  = MO_SS_Conv
	intconv False = MO_UU_Conv
306 307 308 309 310 311 312 313 314

-- ToDo: a narrow of a load can be collapsed into a narrow load, right?
-- but what if the architecture only supports word-sized loads, should
-- we do the transformation anyway?

cmmMachOpFold mop args@[CmmLit (CmmInt x xrep), CmmLit (CmmInt y _)]
  = case mop of
	-- for comparisons: don't forget to narrow the arguments before
	-- comparing, since they might be out of range.
315 316
    	MO_Eq r   -> CmmLit (CmmInt (if x_u == y_u then 1 else 0) wordWidth)
    	MO_Ne r   -> CmmLit (CmmInt (if x_u /= y_u then 1 else 0) wordWidth)
317

318 319 320 321
    	MO_U_Gt r -> CmmLit (CmmInt (if x_u >  y_u then 1 else 0) wordWidth)
    	MO_U_Ge r -> CmmLit (CmmInt (if x_u >= y_u then 1 else 0) wordWidth)
    	MO_U_Lt r -> CmmLit (CmmInt (if x_u <  y_u then 1 else 0) wordWidth)
    	MO_U_Le r -> CmmLit (CmmInt (if x_u <= y_u then 1 else 0) wordWidth)
322

323 324 325 326
    	MO_S_Gt r -> CmmLit (CmmInt (if x_s >  y_s then 1 else 0) wordWidth) 
    	MO_S_Ge r -> CmmLit (CmmInt (if x_s >= y_s then 1 else 0) wordWidth)
    	MO_S_Lt r -> CmmLit (CmmInt (if x_s <  y_s then 1 else 0) wordWidth)
    	MO_S_Le r -> CmmLit (CmmInt (if x_s <= y_s then 1 else 0) wordWidth)
327 328 329 330

    	MO_Add r -> CmmLit (CmmInt (x + y) r)
    	MO_Sub r -> CmmLit (CmmInt (x - y) r)
    	MO_Mul r -> CmmLit (CmmInt (x * y) r)
331 332
    	MO_U_Quot r | y /= 0 -> CmmLit (CmmInt (x_u `quot` y_u) r)
    	MO_U_Rem  r | y /= 0 -> CmmLit (CmmInt (x_u `rem`  y_u) r)
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
    	MO_S_Quot r | y /= 0 -> CmmLit (CmmInt (x `quot` y) r)
    	MO_S_Rem  r | y /= 0 -> CmmLit (CmmInt (x `rem` y) r)

	MO_And   r -> CmmLit (CmmInt (x .&. y) r)
	MO_Or    r -> CmmLit (CmmInt (x .|. y) r)
	MO_Xor   r -> CmmLit (CmmInt (x `xor` y) r)

        MO_Shl   r -> CmmLit (CmmInt (x `shiftL` fromIntegral y) r)
        MO_U_Shr r -> CmmLit (CmmInt (x_u `shiftR` fromIntegral y) r)
        MO_S_Shr r -> CmmLit (CmmInt (x `shiftR` fromIntegral y) r)

	other      -> CmmMachOp mop args

   where
	x_u = narrowU xrep x
	y_u = narrowU xrep y
	x_s = narrowS xrep x
	y_s = narrowS xrep y
	

-- When possible, shift the constants to the right-hand side, so that we
-- can match for strength reductions.  Note that the code generator will
-- also assume that constants have been shifted to the right when
-- possible.

cmmMachOpFold op [x@(CmmLit _), y]
   | not (isLit y) && isCommutableMachOp op 
   = cmmMachOpFold op [y, x]

-- Turn (a+b)+c into a+(b+c) where possible.  Because literals are
-- moved to the right, it is more likely that we will find
-- opportunities for constant folding when the expression is
-- right-associated.
--
-- ToDo: this appears to introduce a quadratic behaviour due to the
-- nested cmmMachOpFold.  Can we fix this?
--
-- Why do we check isLit arg1?  If arg1 is a lit, it means that arg2
-- is also a lit (otherwise arg1 would be on the right).  If we
-- put arg1 on the left of the rearranged expression, we'll get into a
-- loop:  (x1+x2)+x3 => x1+(x2+x3)  => (x2+x3)+x1 => x2+(x3+x1) ...
--
375 376 377
-- Also don't do it if arg1 is PicBaseReg, so that we don't separate the
-- PicBaseReg from the corresponding label (or label difference).
--
378
cmmMachOpFold mop1 [CmmMachOp mop2 [arg1,arg2], arg3]
379
   | mop2 `associates_with` mop1
380
     && not (isLit arg1) && not (isPicReg arg1)
381 382 383 384 385 386 387 388 389 390
   = cmmMachOpFold mop2 [arg1, cmmMachOpFold mop1 [arg2,arg3]]
   where
     MO_Add{} `associates_with` MO_Sub{} = True
     mop1 `associates_with` mop2 =
        mop1 == mop2 && isAssociativeMachOp mop1

-- special case: (a - b) + c  ==>  a + (c - b)
cmmMachOpFold mop1@(MO_Add{}) [CmmMachOp mop2@(MO_Sub{}) [arg1,arg2], arg3]
   | not (isLit arg1) && not (isPicReg arg1)
   = cmmMachOpFold mop1 [arg1, cmmMachOpFold mop2 [arg3,arg2]]
391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410

-- Make a RegOff if we can
cmmMachOpFold (MO_Add _) [CmmReg reg, CmmLit (CmmInt n rep)]
  = CmmRegOff reg (fromIntegral (narrowS rep n))
cmmMachOpFold (MO_Add _) [CmmRegOff reg off, CmmLit (CmmInt n rep)]
  = CmmRegOff reg (off + fromIntegral (narrowS rep n))
cmmMachOpFold (MO_Sub _) [CmmReg reg, CmmLit (CmmInt n rep)]
  = CmmRegOff reg (- fromIntegral (narrowS rep n))
cmmMachOpFold (MO_Sub _) [CmmRegOff reg off, CmmLit (CmmInt n rep)]
  = CmmRegOff reg (off - fromIntegral (narrowS rep n))

-- Fold label(+/-)offset into a CmmLit where possible

cmmMachOpFold (MO_Add _) [CmmLit (CmmLabel lbl), CmmLit (CmmInt i rep)]
  = CmmLit (CmmLabelOff lbl (fromIntegral (narrowU rep i)))
cmmMachOpFold (MO_Add _) [CmmLit (CmmInt i rep), CmmLit (CmmLabel lbl)]
  = CmmLit (CmmLabelOff lbl (fromIntegral (narrowU rep i)))
cmmMachOpFold (MO_Sub _) [CmmLit (CmmLabel lbl), CmmLit (CmmInt i rep)]
  = CmmLit (CmmLabelOff lbl (fromIntegral (negate (narrowU rep i))))

411

412 413 414 415 416 417
-- Comparison of literal with widened operand: perform the comparison
-- at the smaller width, as long as the literal is within range.

-- We can't do the reverse trick, when the operand is narrowed:
-- narrowing throws away bits from the operand, there's no way to do
-- the same comparison at the larger size.
418 419 420 421 422

#if i386_TARGET_ARCH || x86_64_TARGET_ARCH
-- powerPC NCG has a TODO for I8/I16 comparisons, so don't try

cmmMachOpFold cmp [CmmMachOp conv [x], CmmLit (CmmInt i _)]
423 424 425 426 427 428 429 430
  |     -- if the operand is widened:
    Just (rep, signed, narrow_fn) <- maybe_conversion conv,
        -- and this is a comparison operation:
    Just narrow_cmp <- maybe_comparison cmp rep signed,
        -- and the literal fits in the smaller size:
    i == narrow_fn rep i
        -- then we can do the comparison at the smaller size
  = cmmMachOpFold narrow_cmp [x, CmmLit (CmmInt i rep)]
431
 where
432
    maybe_conversion (MO_UU_Conv from to)
433 434
        | to > from
        = Just (from, False, narrowU)
435 436
    maybe_conversion (MO_SS_Conv from to)
        | to > from
437
        = Just (from, True, narrowS)
438

Simon Marlow's avatar
Simon Marlow committed
439 440
        -- don't attempt to apply this optimisation when the source
        -- is a float; see #1916
441 442
    maybe_conversion _ = Nothing
    
443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459
        -- careful (#2080): if the original comparison was signed, but
        -- we were doing an unsigned widen, then we must do an
        -- unsigned comparison at the smaller size.
    maybe_comparison (MO_U_Gt _) rep _     = Just (MO_U_Gt rep)
    maybe_comparison (MO_U_Ge _) rep _     = Just (MO_U_Ge rep)
    maybe_comparison (MO_U_Lt _) rep _     = Just (MO_U_Lt rep)
    maybe_comparison (MO_U_Le _) rep _     = Just (MO_U_Le rep)
    maybe_comparison (MO_Eq   _) rep _     = Just (MO_Eq   rep)
    maybe_comparison (MO_S_Gt _) rep True  = Just (MO_S_Gt rep)
    maybe_comparison (MO_S_Ge _) rep True  = Just (MO_S_Ge rep)
    maybe_comparison (MO_S_Lt _) rep True  = Just (MO_S_Lt rep)
    maybe_comparison (MO_S_Le _) rep True  = Just (MO_S_Le rep)
    maybe_comparison (MO_S_Gt _) rep False = Just (MO_U_Gt rep)
    maybe_comparison (MO_S_Ge _) rep False = Just (MO_U_Ge rep)
    maybe_comparison (MO_S_Lt _) rep False = Just (MO_U_Lt rep)
    maybe_comparison (MO_S_Le _) rep False = Just (MO_U_Le rep)
    maybe_comparison _ _ _ = Nothing
460 461 462

#endif

463 464 465 466 467 468 469 470 471 472 473 474 475 476
-- We can often do something with constants of 0 and 1 ...

cmmMachOpFold mop args@[x, y@(CmmLit (CmmInt 0 _))]
  = case mop of
    	MO_Add   r -> x
    	MO_Sub   r -> x
    	MO_Mul   r -> y
    	MO_And   r -> y
    	MO_Or    r -> x
    	MO_Xor   r -> x
    	MO_Shl   r -> x
    	MO_S_Shr r -> x
    	MO_U_Shr r -> x
        MO_Ne    r | isComparisonExpr x -> x
477
	MO_Eq    r | Just x' <- maybeInvertCmmExpr x -> x'
478 479
	MO_U_Gt  r | isComparisonExpr x -> x
	MO_S_Gt  r | isComparisonExpr x -> x
480 481 482 483
	MO_U_Lt  r | isComparisonExpr x -> CmmLit (CmmInt 0 wordWidth)
	MO_S_Lt  r | isComparisonExpr x -> CmmLit (CmmInt 0 wordWidth)
	MO_U_Ge  r | isComparisonExpr x -> CmmLit (CmmInt 1 wordWidth)
	MO_S_Ge  r | isComparisonExpr x -> CmmLit (CmmInt 1 wordWidth)
484 485
	MO_U_Le  r | Just x' <- maybeInvertCmmExpr x -> x'
	MO_S_Le  r | Just x' <- maybeInvertCmmExpr x -> x'
486 487 488 489 490 491 492 493 494
    	other    -> CmmMachOp mop args

cmmMachOpFold mop args@[x, y@(CmmLit (CmmInt 1 rep))]
  = case mop of
    	MO_Mul    r -> x
    	MO_S_Quot r -> x
    	MO_U_Quot r -> x
    	MO_S_Rem  r -> CmmLit (CmmInt 0 rep)
    	MO_U_Rem  r -> CmmLit (CmmInt 0 rep)
495
        MO_Ne    r | Just x' <- maybeInvertCmmExpr x -> x'
496
	MO_Eq    r | isComparisonExpr x -> x
497 498
	MO_U_Lt  r | Just x' <- maybeInvertCmmExpr x -> x'
	MO_S_Lt  r | Just x' <- maybeInvertCmmExpr x -> x'
499 500 501 502
	MO_U_Gt  r | isComparisonExpr x -> CmmLit (CmmInt 0 wordWidth)
	MO_S_Gt  r | isComparisonExpr x -> CmmLit (CmmInt 0 wordWidth)
	MO_U_Le  r | isComparisonExpr x -> CmmLit (CmmInt 1 wordWidth)
	MO_S_Le  r | isComparisonExpr x -> CmmLit (CmmInt 1 wordWidth)
503 504 505 506 507 508 509 510 511
	MO_U_Ge  r | isComparisonExpr x -> x
	MO_S_Ge  r | isComparisonExpr x -> x
    	other       -> CmmMachOp mop args

-- Now look for multiplication/division by powers of 2 (integers).

cmmMachOpFold mop args@[x, y@(CmmLit (CmmInt n _))]
  = case mop of
    	MO_Mul rep
512
	   | Just p <- exactLog2 n ->
513
                 cmmMachOpFold (MO_Shl rep) [x, CmmLit (CmmInt p rep)]
514 515
    	MO_U_Quot rep
	   | Just p <- exactLog2 n ->
516
                 cmmMachOpFold (MO_U_Shr rep) [x, CmmLit (CmmInt p rep)]
517
    	MO_S_Quot rep
518 519 520 521
	   | Just p <- exactLog2 n, 
	     CmmReg _ <- x ->	-- We duplicate x below, hence require
				-- it is a reg.  FIXME: remove this restriction.
		-- shift right is not the same as quot, because it rounds
Simon Marlow's avatar
Simon Marlow committed
522
		-- to minus infinity, whereasq quot rounds toward zero.
523 524 525 526 527 528 529 530 531 532 533 534 535 536
		-- To fix this up, we add one less than the divisor to the
		-- dividend if it is a negative number.
		--
		-- to avoid a test/jump, we use the following sequence:
		-- 	x1 = x >> word_size-1  (all 1s if -ve, all 0s if +ve)
		--      x2 = y & (divisor-1)
		--      result = (x+x2) >>= log2(divisor)
		-- this could be done a bit more simply using conditional moves,
		-- but we're processor independent here.
		--
		-- we optimise the divide by 2 case slightly, generating
		--      x1 = x >> word_size-1  (unsigned)
		--      return = (x + x1) >>= log2(divisor)
		let 
537
		    bits = fromIntegral (widthInBits rep) - 1
538 539 540 541 542 543
		    shr = if p == 1 then MO_U_Shr rep else MO_S_Shr rep
		    x1 = CmmMachOp shr [x, CmmLit (CmmInt bits rep)]
		    x2 = if p == 1 then x1 else
			 CmmMachOp (MO_And rep) [x1, CmmLit (CmmInt (n-1) rep)]
		    x3 = CmmMachOp (MO_Add rep) [x, x2]
		in
544
                cmmMachOpFold (MO_S_Shr rep) [x3, CmmLit (CmmInt p rep)]
545
    	other
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563
           -> unchanged
    where
       unchanged = CmmMachOp mop args

-- Anything else is just too hard.

cmmMachOpFold mop args = CmmMachOp mop args

-- -----------------------------------------------------------------------------
-- exactLog2

-- This algorithm for determining the $\log_2$ of exact powers of 2 comes
-- from GCC.  It requires bit manipulation primitives, and we use GHC
-- extensions.  Tough.
-- 
-- Used to be in MachInstrs --SDM.
-- ToDo: remove use of unboxery --SDM.

564 565 566 567 568
-- Unboxery removed in favor of FastInt; but is the function supposed to fail
-- on inputs >= 2147483648, or was that just an implementation artifact?
-- And is this speed-critical, or can we just use Integer operations
-- (including Data.Bits)?
--  --Isaac Dupree
569 570

exactLog2 :: Integer -> Maybe Integer
571 572
exactLog2 x_
  = if (x_ <= 0 || x_ >= 2147483648) then
573 574
       Nothing
    else
575 576
       case iUnbox (fromInteger x_) of { x ->
       if (x `bitAndFastInt` negateFastInt x) /=# x then
577 578
	  Nothing
       else
579
	  Just (toInteger (iBox (pow2 x)))
580 581
       }
  where
582 583
    pow2 x | x ==# _ILIT(1) = _ILIT(0)
           | otherwise = _ILIT(1) +# pow2 (x `shiftR_FastInt` _ILIT(1))
584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611


-- -----------------------------------------------------------------------------
-- Loopify for C

{-
 This is a simple pass that replaces tail-recursive functions like this:

   fac() {
     ...
     jump fac();
   }

 with this:

  fac() {
   L:
     ...
     goto L;
  }

  the latter generates better C code, because the C compiler treats it
  like a loop, and brings full loop optimisation to bear.

  In my measurements this makes little or no difference to anything
  except factorial, but what the hell.
-}

612
cmmLoopifyForC :: RawCmmTop -> RawCmmTop
613
cmmLoopifyForC p@(CmmProc info entry_lbl
614
                 (ListGraph blocks@(BasicBlock top_id _ : _)))
615 616 617
  | null info = p  -- only if there's an info table, ignore case alts
  | otherwise =  
--  pprTrace "jump_lbl" (ppr jump_lbl <+> ppr entry_lbl) $
618
  CmmProc info entry_lbl (ListGraph blocks')
619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
  where blocks' = [ BasicBlock id (map do_stmt stmts)
		  | BasicBlock id stmts <- blocks ]

        do_stmt (CmmJump (CmmLit (CmmLabel lbl)) _) | lbl == jump_lbl
		= CmmBranch top_id
	do_stmt stmt = stmt

	jump_lbl | tablesNextToCode = entryLblToInfoLbl entry_lbl
		 | otherwise        = entry_lbl

cmmLoopifyForC top = top

-- -----------------------------------------------------------------------------
-- Utils

isLit (CmmLit _) = True
isLit _          = False

isComparisonExpr :: CmmExpr -> Bool
isComparisonExpr (CmmMachOp op _) = isComparisonMachOp op
isComparisonExpr _other 	    = False

641
isPicReg (CmmReg (CmmGlobal PicBaseReg)) = True
642
isPicReg _ = False
643