diff options
author | Simon Marlow <marlowsd@gmail.com> | 2012-10-03 09:30:56 +0100 |
---|---|---|
committer | Simon Marlow <marlowsd@gmail.com> | 2012-10-08 09:04:40 +0100 |
commit | a7c0387d20c1c9994d1100b14fbb8fb4e28a259e (patch) | |
tree | b95d0a512f951a4a463f1aa5178b0cd5c4fdb410 /compiler/codeGen/StgCmmBind.hs | |
parent | aed37acd4d157791381800d5de960a2461bcbef3 (diff) | |
download | haskell-a7c0387d20c1c9994d1100b14fbb8fb4e28a259e.tar.gz |
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.
Diffstat (limited to 'compiler/codeGen/StgCmmBind.hs')
-rw-r--r-- | compiler/codeGen/StgCmmBind.hs | 41 |
1 files changed, 25 insertions, 16 deletions
diff --git a/compiler/codeGen/StgCmmBind.hs b/compiler/codeGen/StgCmmBind.hs index 89d27dd161..5e46dcfd65 100644 --- a/compiler/codeGen/StgCmmBind.hs +++ b/compiler/codeGen/StgCmmBind.hs @@ -10,7 +10,7 @@ module StgCmmBind ( cgTopRhsClosure, cgBind, emitBlackHoleCode, - pushUpdateFrame + pushUpdateFrame, emitUpdateFrame ) where #include "HsVersions.h" @@ -37,7 +37,6 @@ import CLabel import StgSyn import CostCentre import Id -import Control.Monad import Name import Module import ListSetOps @@ -48,6 +47,8 @@ import FastString import Maybes import DynFlags +import Control.Monad + ------------------------------------------------------------------------ -- Top-level bindings ------------------------------------------------------------------------ @@ -460,7 +461,7 @@ closureCodeBody top_lvl bndr cl_info cc args arity body fv_details (CmmMachOp (mo_wordSub dflags) [ CmmReg nodeReg , mkIntExpr dflags (funTag dflags cl_info) ]) - ; whenC node_points (ldvEnterClosure cl_info) + ; when node_points (ldvEnterClosure cl_info) ; granYield arg_regs node_points -- Main payload @@ -525,8 +526,8 @@ thunkCode cl_info fv_details _cc node arity body ; entryHeapCheck cl_info node' arity [] $ do { -- Overwrite with black hole if necessary -- but *after* the heap-overflow check - ; whenC (blackHoleOnEntry cl_info && node_points) - (blackHoleIt cl_info) + ; when (blackHoleOnEntry cl_info && node_points) + (blackHoleIt cl_info node) -- Push update frame ; setupUpdate cl_info node $ @@ -545,13 +546,14 @@ thunkCode cl_info fv_details _cc node arity body -- Update and black-hole wrappers ------------------------------------------------------------------------ -blackHoleIt :: ClosureInfo -> FCode () +blackHoleIt :: ClosureInfo -> LocalReg -> FCode () -- Only called for closures with no args -- Node points to the closure -blackHoleIt closure_info = emitBlackHoleCode (closureSingleEntry closure_info) +blackHoleIt closure_info node + = emitBlackHoleCode (closureSingleEntry closure_info) (CmmReg (CmmLocal node)) -emitBlackHoleCode :: Bool -> FCode () -emitBlackHoleCode is_single_entry = do +emitBlackHoleCode :: Bool -> CmmExpr -> FCode () +emitBlackHoleCode is_single_entry node = do dflags <- getDynFlags -- Eager blackholing is normally disabled, but can be turned on with @@ -578,12 +580,12 @@ emitBlackHoleCode is_single_entry = do -- profiling), so currently eager blackholing doesn't -- work with profiling. - whenC eager_blackholing $ do + when eager_blackholing $ do tickyBlackHole (not is_single_entry) - emitStore (cmmOffsetW dflags (CmmReg nodeReg) (fixedHdrSize dflags)) + emitStore (cmmOffsetW dflags node (fixedHdrSize dflags)) (CmmReg (CmmGlobal CurrentTSO)) emitPrimCall [] MO_WriteBarrier [] - emitStore (CmmReg nodeReg) (CmmReg (CmmGlobal EagerBlackholeInfo)) + emitStore node (CmmReg (CmmGlobal EagerBlackholeInfo)) setupUpdate :: ClosureInfo -> LocalReg -> FCode () -> FCode () -- Nota Bene: this function does not change Node (even if it's a CAF), @@ -634,13 +636,20 @@ pushUpdateFrame lbl updatee body let hdr = fixedHdrSize dflags * wORD_SIZE dflags frame = updfr + hdr + sIZEOF_StgUpdateFrame_NoHdr dflags - off_updatee = hdr + oFFSET_StgUpdateFrame_updatee dflags -- - emitStore (CmmStackSlot Old frame) (mkLblExpr lbl) - emitStore (CmmStackSlot Old (frame - off_updatee)) updatee - initUpdFrameProf frame + emitUpdateFrame dflags (CmmStackSlot Old frame) lbl updatee withUpdFrameOff frame body +emitUpdateFrame :: DynFlags -> CmmExpr -> CLabel -> CmmExpr -> FCode () +emitUpdateFrame dflags frame lbl updatee = do + let + hdr = fixedHdrSize dflags * wORD_SIZE dflags + off_updatee = hdr + oFFSET_StgUpdateFrame_updatee dflags + -- + emitStore frame (mkLblExpr lbl) + emitStore (cmmOffset dflags frame off_updatee) updatee + initUpdFrameProf frame + ----------------------------------------------------------------------------- -- Entering a CAF -- |