Files
Nim/compiler/semfold.nim
zah ca4b971bc8 Initial version of the hot-code reloading support for native targets (#10729)
* squashed work by Zahary

* squashing a ton of useful history... otherwise rebasing on top of upstream Nim after commit 82c009a2cb would be impossible.

* Code review changes; Working test suite (without code reloading enabled)

* - documentation
- implemented the HCR test - almost works...
- fix the issue on Unix where for executable targets the source file for the main module of a project in nimcache was being overwritten with the binary itself (and thus the actual source code was lost)
- fixing embedded paths to shared objects on unix (the "lib" prefix was being prepended to the entire path instead of just the filename)
- other fixes
- removing unnecessary includes since that file is already included in chcks.nim which is in turn included in system.nim (and previously was getting imported in chcks.nim but then system.nim improts something... and that breaks HCR (perhaps it could be fixed but it would be nice not to import anything in system))

* fix for clang & C++ - explicitly casting a function pointer to void*
more stable mangling of parameter names when HCR is on
the length of the static arrays in the DatInit functions is now part of the name of the variables, so when they get resized they get also recreated
more stable mangling for inline functions - no longer depends on the module which first used them
work on the new complicated HCR test - turned surprisingly complex - WIP
test now successfully passes even when re-running `koch test` (previously when the nimcache wasn't cold that lead to errors)
better documentation
calling setStackBottomWith for PreMain
passes over the HcrInit/DatInit/Init calls of all modules are now in the proper order (first all of one type, then all of the next). Also typeinfo globals are registered (created) in a single pass before the DatInit pass (because of the way generic instantiations are handled)
Fix the test suite execution on macOs
fix for being able to query the program arguments when using HCR on posix!
other fixes

* Bugfix: Fix a compilation error in C++ mode when a function pointer
is converted to a raw pointer

* basic documentation for the new hot code reloading semantics

* Add change log entry

* Don't re-execute the top-level statements while reloading JS code

* fix a number of tests broken in a recent bugfix

* Review changes

* Added {.executeOnReload.} pragma that indicates top-level statements
  that should be executed on each reload. To make this work, I've modified
  the way the `if (hcr_init_) {...}` guards are produced in the init code.
  This still needs more work as the new guards seem to be inserted within
  the previously generated guards.

  This change also removes the need for `lastRegistedGlobal` in nimhcr.

* Implemented the `signatureHash` magic and the `hasModuleChanged` API
  depending on it (the actual logic is not imlemented yet).

* Add the "hcr" prefix to all HCR-related symbols in the system module.
  Added a new `hotcodereloading` module exporting the high-level API to
  the user.

  Besides being more hygienic, this was also required in order to make
  it possible to use macros in the high-level API. Without the split,
  `system` would have to import `macros`, which was going to produce
  the well-known init problems.

* Attempted to solve the "GC markers problem".

  Crashes were expected with the previous code, because the GC markers
  were compiled as normal procs are registered in the GC. When their
  module is unloaded, dangling pointers will remain in the GC tables.
  To solve this issue, I don't register any GC markers when HCR is on,
  but I add them to the HCR globals metadata and I use a single marker
  registed in nimhcr during the initialization of the system module that
  will be responsible for marking all globals.

* fix a compilation error

* - implemented the hasModuleChanged functionality
- tuples can be returned and broken into different vars in global scope
- added comments for the closnig scopes of the if statements in the init proc
- the new executeOnReload pragma works now!
- other fixes

* finally! fixing this hack in a proper way - declaring the destructor out of line (out of the class body) - we no longer need to forward-declare popCurrentExceptionEx

* Force full module parsing

This is a temporary hack that breaks some tests. I'll investigate
later how these can be fixed.

* tuples are now properly handled when global!

* these comments mess up the codegen in debug mode when $n is not actually a new line (or something like that) - these labels are intended only for GOTO labels anyway...

* "solved" the issue with the .pdb locks on windows when a binary is being debugged and hot code reloading is used at the same time

* fixes after rebasing...

* small fixes for the test

* better handling of globals! no more compiler crashes for locals with the global pragma, also simplified code around loops in global scope which have local vars (actually globals)

* we can now use the global pragma even for ... globals!

* the right output

* lets try those boehm GC tests

* after the test is ran it will be at its starting state - no git modifications

* clarification in the docs

* removed unnecessary line directives for forward declarations of functions - they were causing trouble with hot code reloading when no semantic change propagates to the main module but a line directive got changed and thus the main module had to be recompiled since the .c code had changed

* fixed bug! was inserting duplicate keys into the table and later was removing only 1 copy of all the duplicates (after a few reloads)

* no longer breaking into DatInit code when not supposed to

* fixes after rebasing

* yet more fixes after rebasing

* Update jssys.nim

* Rework the HCR path-handling logic

After reviewing the code more carefully, I've noticed that the old logic
will be broken when the user overrides the '--out:f' compiler option.

Besides fixing this issues, I took the opportunity to implement the
missing '--outdir:d' option.

Other changes:

* ./koch test won't overwrite any HCR and RTL builds located in nim/lib
* HCR and RTL are compiled with --threads:on by default

* Clean up the globals registration logic

* Handle non-flattened top-level stmtlists in JS as well

* The HCR is not supported with the Boehm GC yet

Also fixes some typos and the expected output of the HCR integration test

* The GC marker procs are now properly used as trampolines

* Fix the HCR integration test in release builds

* Fix ./koch tools

* this forward declaration doesn't seem to be necessary, and in fact breaks HCR because a 2nd function pointer is emitted for this externed/rtl func

* the forward declaration I removed in the last commit was actually necessary

* Attempt to make all tests green

* Fix tgenscript

* BAT file for running the HCR integration test on Windows [skip ci]

* Fix the docgen tests

* A final fix for Travis (hopefully)
2019-02-26 15:48:55 +01:00

745 lines
28 KiB
Nim

#
#
# The Nim Compiler
# (c) Copyright 2015 Andreas Rumpf
#
# See the file "copying.txt", included in this
# distribution, for details about the copyright.
#
# this module folds constants; used by semantic checking phase
# and evaluation phase
import
strutils, options, ast, astalgo, trees, treetab, nimsets,
nversion, platform, math, msgs, os, condsyms, idents, renderer, types,
commands, magicsys, modulegraphs, strtabs, lineinfos
proc newIntNodeT*(intVal: BiggestInt, n: PNode; g: ModuleGraph): PNode =
case skipTypes(n.typ, abstractVarRange).kind
of tyInt:
result = newIntNode(nkIntLit, intVal)
# See bug #6989. 'pred' et al only produce an int literal type if the
# original type was 'int', not a distinct int etc.
if n.typ.kind == tyInt:
result.typ = getIntLitType(g, result)
else:
result.typ = n.typ
# hrm, this is not correct: 1 + high(int) shouldn't produce tyInt64 ...
#setIntLitType(result)
of tyChar:
result = newIntNode(nkCharLit, intVal)
result.typ = n.typ
else:
result = newIntNode(nkIntLit, intVal)
result.typ = n.typ
result.info = n.info
proc newFloatNodeT*(floatVal: BiggestFloat, n: PNode; g: ModuleGraph): PNode =
result = newFloatNode(nkFloatLit, floatVal)
result.typ = n.typ
result.info = n.info
proc newStrNodeT*(strVal: string, n: PNode; g: ModuleGraph): PNode =
result = newStrNode(nkStrLit, strVal)
result.typ = n.typ
result.info = n.info
proc getConstExpr*(m: PSym, n: PNode; g: ModuleGraph): PNode
# evaluates the constant expression or returns nil if it is no constant
# expression
proc evalOp*(m: TMagic, n, a, b, c: PNode; g: ModuleGraph): PNode
proc checkInRange(conf: ConfigRef; n: PNode, res: BiggestInt): bool =
if res in firstOrd(conf, n.typ)..lastOrd(conf, n.typ):
result = true
proc foldAdd(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
let res = a +% b
if ((res xor a) >= 0'i64 or (res xor b) >= 0'i64) and
checkInRange(g.config, n, res):
result = newIntNodeT(res, n, g)
proc foldSub*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
let res = a -% b
if ((res xor a) >= 0'i64 or (res xor not b) >= 0'i64) and
checkInRange(g.config, n, res):
result = newIntNodeT(res, n, g)
proc foldUnarySub(a: BiggestInt, n: PNode, g: ModuleGraph): PNode =
if a != firstOrd(g.config, n.typ):
result = newIntNodeT(-a, n, g)
proc foldAbs*(a: BiggestInt, n: PNode; g: ModuleGraph): PNode =
if a != firstOrd(g.config, n.typ):
result = newIntNodeT(abs(a), n, g)
proc foldMod*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
if b != 0'i64:
result = newIntNodeT(a mod b, n, g)
proc foldModU*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
if b != 0'i64:
result = newIntNodeT(a %% b, n, g)
proc foldDiv*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
if b != 0'i64 and (a != firstOrd(g.config, n.typ) or b != -1'i64):
result = newIntNodeT(a div b, n, g)
proc foldDivU*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
if b != 0'i64:
result = newIntNodeT(a /% b, n, g)
proc foldMul*(a, b: BiggestInt, n: PNode; g: ModuleGraph): PNode =
let res = a *% b
let floatProd = toBiggestFloat(a) * toBiggestFloat(b)
let resAsFloat = toBiggestFloat(res)
# Fast path for normal case: small multiplicands, and no info
# is lost in either method.
if resAsFloat == floatProd and checkInRange(g.config, n, res):
return newIntNodeT(res, n, g)
# Somebody somewhere lost info. Close enough, or way off? Note
# that a != 0 and b != 0 (else resAsFloat == floatProd == 0).
# The difference either is or isn't significant compared to the
# true value (of which floatProd is a good approximation).
# abs(diff)/abs(prod) <= 1/32 iff
# 32 * abs(diff) <= abs(prod) -- 5 good bits is "close enough"
if 32.0 * abs(resAsFloat - floatProd) <= abs(floatProd) and
checkInRange(g.config, n, res):
return newIntNodeT(res, n, g)
proc ordinalValToString*(a: PNode; g: ModuleGraph): string =
# because $ has the param ordinal[T], `a` is not necessarily an enum, but an
# ordinal
var x = getInt(a)
var t = skipTypes(a.typ, abstractRange)
case t.kind
of tyChar:
result = $chr(int(x) and 0xff)
of tyEnum:
var n = t.n
for i in countup(0, sonsLen(n) - 1):
if n.sons[i].kind != nkSym: internalError(g.config, a.info, "ordinalValToString")
var field = n.sons[i].sym
if field.position == x:
if field.ast == nil:
return field.name.s
else:
return field.ast.strVal
localError(g.config, a.info,
"Cannot convert int literal to $1. The value is invalid." %
[typeToString(t)])
else:
result = $x
proc isFloatRange(t: PType): bool {.inline.} =
result = t.kind == tyRange and t.sons[0].kind in {tyFloat..tyFloat128}
proc isIntRange(t: PType): bool {.inline.} =
result = t.kind == tyRange and t.sons[0].kind in {
tyInt..tyInt64, tyUInt8..tyUInt32}
proc pickIntRange(a, b: PType): PType =
if isIntRange(a): result = a
elif isIntRange(b): result = b
else: result = a
proc isIntRangeOrLit(t: PType): bool =
result = isIntRange(t) or isIntLit(t)
proc makeRange(typ: PType, first, last: BiggestInt; g: ModuleGraph): PType =
let minA = min(first, last)
let maxA = max(first, last)
let lowerNode = newIntNode(nkIntLit, minA)
if typ.kind == tyInt and minA == maxA:
result = getIntLitType(g, lowerNode)
elif typ.kind in {tyUint, tyUInt64}:
# these are not ordinal types, so you get no subrange type for these:
result = typ
else:
var n = newNode(nkRange)
addSon(n, lowerNode)
addSon(n, newIntNode(nkIntLit, maxA))
result = newType(tyRange, typ.owner)
result.n = n
addSonSkipIntLit(result, skipTypes(typ, {tyRange}))
proc makeRangeF(typ: PType, first, last: BiggestFloat; g: ModuleGraph): PType =
var n = newNode(nkRange)
addSon(n, newFloatNode(nkFloatLit, min(first.float, last.float)))
addSon(n, newFloatNode(nkFloatLit, max(first.float, last.float)))
result = newType(tyRange, typ.owner)
result.n = n
addSonSkipIntLit(result, skipTypes(typ, {tyRange}))
proc fitLiteral(c: ConfigRef, n: PNode): PNode =
# Trim the literal value in order to make it fit in the destination type
if n == nil:
# `n` may be nil if the overflow check kicks in
return
doAssert n.kind in {nkIntLit, nkCharLit}
result = n
let typ = n.typ.skipTypes(abstractRange)
if typ.kind in tyUInt..tyUint32:
result.intVal = result.intVal and lastOrd(c, typ, fixedUnsigned=true)
proc evalOp(m: TMagic, n, a, b, c: PNode; g: ModuleGraph): PNode =
template doAndFit(op: untyped): untyped =
# Implements wrap-around behaviour for unsigned types
fitLiteral(g.config, op)
# b and c may be nil
result = nil
case m
of mOrd: result = newIntNodeT(getOrdValue(a), n, g)
of mChr: result = newIntNodeT(getInt(a), n, g)
of mUnaryMinusI, mUnaryMinusI64: result = foldUnarySub(getInt(a), n, g)
of mUnaryMinusF64: result = newFloatNodeT(- getFloat(a), n, g)
of mNot: result = newIntNodeT(1 - getInt(a), n, g)
of mCard: result = newIntNodeT(nimsets.cardSet(g.config, a), n, g)
of mBitnotI: result = doAndFit(newIntNodeT(not getInt(a), n, g))
of mLengthArray: result = newIntNodeT(lengthOrd(g.config, a.typ), n, g)
of mLengthSeq, mLengthOpenArray, mXLenSeq, mLengthStr, mXLenStr:
if a.kind == nkNilLit:
result = newIntNodeT(0, n, g)
elif a.kind in {nkStrLit..nkTripleStrLit}:
result = newIntNodeT(len a.strVal, n, g)
else:
result = newIntNodeT(sonsLen(a), n, g)
of mUnaryPlusI, mUnaryPlusF64: result = a # throw `+` away
of mToFloat, mToBiggestFloat:
result = newFloatNodeT(toFloat(int(getInt(a))), n, g)
# XXX: Hides overflow/underflow
of mToInt, mToBiggestInt: result = newIntNodeT(system.toInt(getFloat(a)), n, g)
of mAbsF64: result = newFloatNodeT(abs(getFloat(a)), n, g)
of mAbsI: result = foldAbs(getInt(a), n, g)
of mZe8ToI, mZe8ToI64, mZe16ToI, mZe16ToI64, mZe32ToI64, mZeIToI64:
# byte(-128) = 1...1..1000_0000'64 --> 0...0..1000_0000'64
result = newIntNodeT(getInt(a) and (`shl`(1, getSize(g.config, a.typ) * 8) - 1), n, g)
of mToU8: result = newIntNodeT(getInt(a) and 0x000000FF, n, g)
of mToU16: result = newIntNodeT(getInt(a) and 0x0000FFFF, n, g)
of mToU32: result = newIntNodeT(getInt(a) and 0x00000000FFFFFFFF'i64, n, g)
of mUnaryLt: result = doAndFit(foldSub(getOrdValue(a), 1, n, g))
of mSucc: result = doAndFit(foldAdd(getOrdValue(a), getInt(b), n, g))
of mPred: result = doAndFit(foldSub(getOrdValue(a), getInt(b), n, g))
of mAddI: result = foldAdd(getInt(a), getInt(b), n, g)
of mSubI: result = foldSub(getInt(a), getInt(b), n, g)
of mMulI: result = foldMul(getInt(a), getInt(b), n, g)
of mMinI:
if getInt(a) > getInt(b): result = newIntNodeT(getInt(b), n, g)
else: result = newIntNodeT(getInt(a), n, g)
of mMaxI:
if getInt(a) > getInt(b): result = newIntNodeT(getInt(a), n, g)
else: result = newIntNodeT(getInt(b), n, g)
of mShlI:
case skipTypes(n.typ, abstractRange).kind
of tyInt8: result = newIntNodeT(int8(getInt(a)) shl int8(getInt(b)), n, g)
of tyInt16: result = newIntNodeT(int16(getInt(a)) shl int16(getInt(b)), n, g)
of tyInt32: result = newIntNodeT(int32(getInt(a)) shl int32(getInt(b)), n, g)
of tyInt64, tyInt:
result = newIntNodeT(`shl`(getInt(a), getInt(b)), n, g)
of tyUInt..tyUInt64:
result = doAndFit(newIntNodeT(`shl`(getInt(a), getInt(b)), n, g))
else: internalError(g.config, n.info, "constant folding for shl")
of mShrI:
case skipTypes(n.typ, abstractRange).kind
of tyInt8: result = newIntNodeT(int8(getInt(a)) shr int8(getInt(b)), n, g)
of tyInt16: result = newIntNodeT(int16(getInt(a)) shr int16(getInt(b)), n, g)
of tyInt32: result = newIntNodeT(int32(getInt(a)) shr int32(getInt(b)), n, g)
of tyInt64, tyInt, tyUInt..tyUInt64:
result = newIntNodeT(`shr`(getInt(a), getInt(b)), n, g)
else: internalError(g.config, n.info, "constant folding for shr")
of mAshrI:
case skipTypes(n.typ, abstractRange).kind
of tyInt8: result = newIntNodeT(ashr(int8(getInt(a)), int8(getInt(b))), n, g)
of tyInt16: result = newIntNodeT(ashr(int16(getInt(a)), int16(getInt(b))), n, g)
of tyInt32: result = newIntNodeT(ashr(int32(getInt(a)), int32(getInt(b))), n, g)
of tyInt64, tyInt:
result = newIntNodeT(ashr(getInt(a), getInt(b)), n, g)
else: internalError(g.config, n.info, "constant folding for ashr")
of mDivI: result = foldDiv(getInt(a), getInt(b), n, g)
of mModI: result = foldMod(getInt(a), getInt(b), n, g)
of mAddF64: result = newFloatNodeT(getFloat(a) + getFloat(b), n, g)
of mSubF64: result = newFloatNodeT(getFloat(a) - getFloat(b), n, g)
of mMulF64: result = newFloatNodeT(getFloat(a) * getFloat(b), n, g)
of mDivF64:
result = newFloatNodeT(getFloat(a) / getFloat(b), n, g)
of mMaxF64:
if getFloat(a) > getFloat(b): result = newFloatNodeT(getFloat(a), n, g)
else: result = newFloatNodeT(getFloat(b), n, g)
of mMinF64:
if getFloat(a) > getFloat(b): result = newFloatNodeT(getFloat(b), n, g)
else: result = newFloatNodeT(getFloat(a), n, g)
of mIsNil: result = newIntNodeT(ord(a.kind == nkNilLit), n, g)
of mLtI, mLtB, mLtEnum, mLtCh:
result = newIntNodeT(ord(getOrdValue(a) < getOrdValue(b)), n, g)
of mLeI, mLeB, mLeEnum, mLeCh:
result = newIntNodeT(ord(getOrdValue(a) <= getOrdValue(b)), n, g)
of mEqI, mEqB, mEqEnum, mEqCh:
result = newIntNodeT(ord(getOrdValue(a) == getOrdValue(b)), n, g)
of mLtF64: result = newIntNodeT(ord(getFloat(a) < getFloat(b)), n, g)
of mLeF64: result = newIntNodeT(ord(getFloat(a) <= getFloat(b)), n, g)
of mEqF64: result = newIntNodeT(ord(getFloat(a) == getFloat(b)), n, g)
of mLtStr: result = newIntNodeT(ord(getStr(a) < getStr(b)), n, g)
of mLeStr: result = newIntNodeT(ord(getStr(a) <= getStr(b)), n, g)
of mEqStr: result = newIntNodeT(ord(getStr(a) == getStr(b)), n, g)
of mLtU, mLtU64:
result = newIntNodeT(ord(`<%`(getOrdValue(a), getOrdValue(b))), n, g)
of mLeU, mLeU64:
result = newIntNodeT(ord(`<=%`(getOrdValue(a), getOrdValue(b))), n, g)
of mBitandI, mAnd: result = doAndFit(newIntNodeT(a.getInt and b.getInt, n, g))
of mBitorI, mOr: result = doAndFit(newIntNodeT(getInt(a) or getInt(b), n, g))
of mBitxorI, mXor: result = doAndFit(newIntNodeT(a.getInt xor b.getInt, n, g))
of mAddU: result = doAndFit(newIntNodeT(`+%`(getInt(a), getInt(b)), n, g))
of mSubU: result = doAndFit(newIntNodeT(`-%`(getInt(a), getInt(b)), n, g))
of mMulU: result = doAndFit(newIntNodeT(`*%`(getInt(a), getInt(b)), n, g))
of mModU: result = doAndFit(foldModU(getInt(a), getInt(b), n, g))
of mDivU: result = doAndFit(foldDivU(getInt(a), getInt(b), n, g))
of mLeSet: result = newIntNodeT(ord(containsSets(g.config, a, b)), n, g)
of mEqSet: result = newIntNodeT(ord(equalSets(g.config, a, b)), n, g)
of mLtSet:
result = newIntNodeT(ord(containsSets(g.config, a, b) and not equalSets(g.config, a, b)), n, g)
of mMulSet:
result = nimsets.intersectSets(g.config, a, b)
result.info = n.info
of mPlusSet:
result = nimsets.unionSets(g.config, a, b)
result.info = n.info
of mMinusSet:
result = nimsets.diffSets(g.config, a, b)
result.info = n.info
of mSymDiffSet:
result = nimsets.symdiffSets(g.config, a, b)
result.info = n.info
of mConStrStr: result = newStrNodeT(getStrOrChar(a) & getStrOrChar(b), n, g)
of mInSet: result = newIntNodeT(ord(inSet(a, b)), n, g)
of mRepr:
# BUGFIX: we cannot eval mRepr here for reasons that I forgot.
discard
of mIntToStr, mInt64ToStr: result = newStrNodeT($(getOrdValue(a)), n, g)
of mBoolToStr:
if getOrdValue(a) == 0: result = newStrNodeT("false", n, g)
else: result = newStrNodeT("true", n, g)
of mCopyStr: result = newStrNodeT(substr(getStr(a), int(getOrdValue(b))), n, g)
of mCopyStrLast:
result = newStrNodeT(substr(getStr(a), int(getOrdValue(b)),
int(getOrdValue(c))), n, g)
of mFloatToStr: result = newStrNodeT($getFloat(a), n, g)
of mCStrToStr, mCharToStr:
if a.kind == nkBracket:
var s = ""
for b in a.sons:
s.add b.getStrOrChar
result = newStrNodeT(s, n, g)
else:
result = newStrNodeT(getStrOrChar(a), n, g)
of mStrToStr: result = newStrNodeT(getStrOrChar(a), n, g)
of mEnumToStr: result = newStrNodeT(ordinalValToString(a, g), n, g)
of mArrToSeq:
result = copyTree(a)
result.typ = n.typ
of mCompileOption:
result = newIntNodeT(ord(commands.testCompileOption(g.config, a.getStr, n.info)), n, g)
of mCompileOptionArg:
result = newIntNodeT(ord(
testCompileOptionArg(g.config, getStr(a), getStr(b), n.info)), n, g)
of mEqProc:
result = newIntNodeT(ord(
exprStructuralEquivalent(a, b, strictSymEquality=true)), n, g)
else: discard
proc getConstIfExpr(c: PSym, n: PNode; g: ModuleGraph): PNode =
result = nil
for i in countup(0, sonsLen(n) - 1):
var it = n.sons[i]
if it.len == 2:
var e = getConstExpr(c, it.sons[0], g)
if e == nil: return nil
if getOrdValue(e) != 0:
if result == nil:
result = getConstExpr(c, it.sons[1], g)
if result == nil: return
elif it.len == 1:
if result == nil: result = getConstExpr(c, it.sons[0], g)
else: internalError(g.config, it.info, "getConstIfExpr()")
proc leValueConv*(a, b: PNode): bool =
result = false
case a.kind
of nkCharLit..nkUInt64Lit:
case b.kind
of nkCharLit..nkUInt64Lit: result = a.intVal <= b.intVal
of nkFloatLit..nkFloat128Lit: result = a.intVal <= round(b.floatVal).int
else: result = false #internalError(a.info, "leValueConv")
of nkFloatLit..nkFloat128Lit:
case b.kind
of nkFloatLit..nkFloat128Lit: result = a.floatVal <= b.floatVal
of nkCharLit..nkUInt64Lit: result = a.floatVal <= toFloat(int(b.intVal))
else: result = false # internalError(a.info, "leValueConv")
else: result = false # internalError(a.info, "leValueConv")
proc magicCall(m: PSym, n: PNode; g: ModuleGraph): PNode =
if sonsLen(n) <= 1: return
var s = n.sons[0].sym
var a = getConstExpr(m, n.sons[1], g)
var b, c: PNode
if a == nil: return
if sonsLen(n) > 2:
b = getConstExpr(m, n.sons[2], g)
if b == nil: return
if sonsLen(n) > 3:
c = getConstExpr(m, n.sons[3], g)
if c == nil: return
result = evalOp(s.magic, n, a, b, c, g)
proc getAppType(n: PNode; g: ModuleGraph): PNode =
if g.config.globalOptions.contains(optGenDynLib):
result = newStrNodeT("lib", n, g)
elif g.config.globalOptions.contains(optGenStaticLib):
result = newStrNodeT("staticlib", n, g)
elif g.config.globalOptions.contains(optGenGuiApp):
result = newStrNodeT("gui", n, g)
else:
result = newStrNodeT("console", n, g)
proc rangeCheck(n: PNode, value: BiggestInt; g: ModuleGraph) =
var err = false
if n.typ.skipTypes({tyRange}).kind in {tyUInt..tyUInt64}:
err = value <% firstOrd(g.config, n.typ) or value >% lastOrd(g.config, n.typ, fixedUnsigned=true)
else:
err = value < firstOrd(g.config, n.typ) or value > lastOrd(g.config, n.typ)
if err:
localError(g.config, n.info, "cannot convert " & $value &
" to " & typeToString(n.typ))
proc foldConv(n, a: PNode; g: ModuleGraph; check = false): PNode =
let dstTyp = skipTypes(n.typ, abstractRange)
let srcTyp = skipTypes(a.typ, abstractRange)
# XXX range checks?
case dstTyp.kind
of tyInt..tyInt64, tyUint..tyUInt64:
case srcTyp.kind
of tyFloat..tyFloat64:
result = newIntNodeT(int(getFloat(a)), n, g)
of tyChar:
result = newIntNodeT(getOrdValue(a), n, g)
of tyUInt..tyUInt64, tyInt..tyInt64:
let toSigned = dstTyp.kind in tyInt..tyInt64
var val = a.getOrdValue
if dstTyp.kind in {tyInt, tyInt64, tyUint, tyUInt64}:
# No narrowing needed
discard
elif dstTyp.kind in {tyInt..tyInt64}:
# Signed type: Overflow check (if requested) and conversion
if check: rangeCheck(n, val, g)
let mask = (`shl`(1, getSize(g.config, dstTyp) * 8) - 1)
let valSign = val < 0
val = abs(val) and mask
if valSign: val = -val
else:
# Unsigned type: Conversion
let mask = (`shl`(1, getSize(g.config, dstTyp) * 8) - 1)
val = val and mask
result = newIntNodeT(val, n, g)
else:
result = a
result.typ = n.typ
if check and result.kind in {nkCharLit..nkUInt64Lit}:
rangeCheck(n, result.intVal, g)
of tyFloat..tyFloat64:
case srcTyp.kind
of tyInt..tyInt64, tyEnum, tyBool, tyChar:
result = newFloatNodeT(toBiggestFloat(getOrdValue(a)), n, g)
else:
result = a
result.typ = n.typ
of tyOpenArray, tyVarargs, tyProc, tyPointer:
discard
else:
result = a
result.typ = n.typ
proc getArrayConstr(m: PSym, n: PNode; g: ModuleGraph): PNode =
if n.kind == nkBracket:
result = n
else:
result = getConstExpr(m, n, g)
if result == nil: result = n
proc foldArrayAccess(m: PSym, n: PNode; g: ModuleGraph): PNode =
var x = getConstExpr(m, n.sons[0], g)
if x == nil or x.typ.skipTypes({tyGenericInst, tyAlias, tySink}).kind == tyTypeDesc:
return
var y = getConstExpr(m, n.sons[1], g)
if y == nil: return
var idx = getOrdValue(y)
case x.kind
of nkPar, nkTupleConstr:
if idx >= 0 and idx < sonsLen(x):
result = x.sons[int(idx)]
if result.kind == nkExprColonExpr: result = result.sons[1]
else:
localError(g.config, n.info, formatErrorIndexBound(idx, sonsLen(x)-1) & $n)
of nkBracket:
idx = idx - firstOrd(g.config, x.typ)
if idx >= 0 and idx < x.len: result = x.sons[int(idx)]
else: localError(g.config, n.info, formatErrorIndexBound(idx, x.len-1) & $n)
of nkStrLit..nkTripleStrLit:
result = newNodeIT(nkCharLit, x.info, n.typ)
if idx >= 0 and idx < len(x.strVal):
result.intVal = ord(x.strVal[int(idx)])
elif idx == len(x.strVal) and optLaxStrings in g.config.options:
discard
else:
localError(g.config, n.info, formatErrorIndexBound(idx, len(x.strVal)-1) & $n)
else: discard
proc foldFieldAccess(m: PSym, n: PNode; g: ModuleGraph): PNode =
# a real field access; proc calls have already been transformed
var x = getConstExpr(m, n.sons[0], g)
if x == nil or x.kind notin {nkObjConstr, nkPar, nkTupleConstr}: return
var field = n.sons[1].sym
for i in countup(ord(x.kind == nkObjConstr), sonsLen(x) - 1):
var it = x.sons[i]
if it.kind != nkExprColonExpr:
# lookup per index:
result = x.sons[field.position]
if result.kind == nkExprColonExpr: result = result.sons[1]
return
if it.sons[0].sym.name.id == field.name.id:
result = x.sons[i].sons[1]
return
localError(g.config, n.info, "field not found: " & field.name.s)
proc foldConStrStr(m: PSym, n: PNode; g: ModuleGraph): PNode =
result = newNodeIT(nkStrLit, n.info, n.typ)
result.strVal = ""
for i in countup(1, sonsLen(n) - 1):
let a = getConstExpr(m, n.sons[i], g)
if a == nil: return nil
result.strVal.add(getStrOrChar(a))
proc newSymNodeTypeDesc*(s: PSym; info: TLineInfo): PNode =
result = newSymNode(s, info)
if s.typ.kind != tyTypeDesc:
result.typ = newType(tyTypeDesc, s.owner)
result.typ.addSonSkipIntLit(s.typ)
else:
result.typ = s.typ
proc getConstExpr(m: PSym, n: PNode; g: ModuleGraph): PNode =
result = nil
case n.kind
of nkSym:
var s = n.sym
case s.kind
of skEnumField:
result = newIntNodeT(s.position, n, g)
of skConst:
case s.magic
of mIsMainModule: result = newIntNodeT(ord(sfMainModule in m.flags), n, g)
of mCompileDate: result = newStrNodeT(getDateStr(), n, g)
of mCompileTime: result = newStrNodeT(getClockStr(), n, g)
of mCpuEndian: result = newIntNodeT(ord(CPU[g.config.target.targetCPU].endian), n, g)
of mHostOS: result = newStrNodeT(toLowerAscii(platform.OS[g.config.target.targetOS].name), n, g)
of mHostCPU: result = newStrNodeT(platform.CPU[g.config.target.targetCPU].name.toLowerAscii, n, g)
of mBuildOS: result = newStrNodeT(toLowerAscii(platform.OS[g.config.target.hostOS].name), n, g)
of mBuildCPU: result = newStrNodeT(platform.CPU[g.config.target.hostCPU].name.toLowerAscii, n, g)
of mAppType: result = getAppType(n, g)
of mNaN: result = newFloatNodeT(NaN, n, g)
of mInf: result = newFloatNodeT(Inf, n, g)
of mNegInf: result = newFloatNodeT(NegInf, n, g)
of mIntDefine:
if isDefined(g.config, s.name.s):
try:
result = newIntNodeT(g.config.symbols[s.name.s].parseInt, n, g)
except ValueError:
localError(g.config, s.info,
"{.intdefine.} const was set to an invalid integer: '" &
g.config.symbols[s.name.s] & "'")
of mStrDefine:
if isDefined(g.config, s.name.s):
result = newStrNodeT(g.config.symbols[s.name.s], n, g)
of mBoolDefine:
if isDefined(g.config, s.name.s):
try:
result = newIntNodeT(g.config.symbols[s.name.s].parseBool.int, n, g)
except ValueError:
localError(g.config, s.info,
"{.booldefine.} const was set to an invalid bool: '" &
g.config.symbols[s.name.s] & "'")
else:
result = copyTree(s.ast)
of skProc, skFunc, skMethod:
result = n
of skParam:
if s.typ != nil and s.typ.kind == tyTypeDesc:
result = newSymNodeTypeDesc(s, n.info)
of skType:
# XXX gensym'ed symbols can come here and cannot be resolved. This is
# dirty, but correct.
if s.typ != nil:
result = newSymNodeTypeDesc(s, n.info)
of skGenericParam:
if s.typ.kind == tyStatic:
if s.typ.n != nil and tfUnresolved notin s.typ.flags:
result = s.typ.n
result.typ = s.typ.base
elif s.typ.isIntLit:
result = s.typ.n
else:
result = newSymNodeTypeDesc(s, n.info)
else: discard
of nkCharLit..nkNilLit:
result = copyNode(n)
of nkIfExpr:
result = getConstIfExpr(m, n, g)
of nkCallKinds:
if n.sons[0].kind != nkSym: return
var s = n.sons[0].sym
if s.kind != skProc and s.kind != skFunc: return
try:
case s.magic
of mNone:
# If it has no sideEffect, it should be evaluated. But not here.
return
of mLow:
result = newIntNodeT(firstOrd(g.config, n.sons[1].typ), n, g)
of mHigh:
if skipTypes(n.sons[1].typ, abstractVar+{tyUserTypeClassInst}).kind notin
{tySequence, tyString, tyCString, tyOpenArray, tyVarargs}:
result = newIntNodeT(lastOrd(g.config, skipTypes(n[1].typ, abstractVar)), n, g)
else:
var a = getArrayConstr(m, n.sons[1], g)
if a.kind == nkBracket:
# we can optimize it away:
result = newIntNodeT(sonsLen(a)-1, n, g)
of mLengthOpenArray:
var a = getArrayConstr(m, n.sons[1], g)
if a.kind == nkBracket:
# we can optimize it away! This fixes the bug ``len(134)``.
result = newIntNodeT(sonsLen(a), n, g)
else:
result = magicCall(m, n, g)
of mLengthArray:
# It doesn't matter if the argument is const or not for mLengthArray.
# This fixes bug #544.
result = newIntNodeT(lengthOrd(g.config, n.sons[1].typ), n, g)
of mSizeOf:
let size = getSize(g.config, n[1].typ)
if size >= 0:
result = newIntNode(nkIntLit, size)
result.info = n.info
result.typ = getSysType(g, n.info, tyInt)
else:
result = nil
of mAstToStr:
result = newStrNodeT(renderTree(n[1], {renderNoComments}), n, g)
of mConStrStr:
result = foldConStrStr(m, n, g)
of mIs:
# The only kind of mIs node that comes here is one depending on some
# generic parameter and that's (hopefully) handled at instantiation time
discard
else:
result = magicCall(m, n, g)
except OverflowError:
localError(g.config, n.info, "over- or underflow")
except DivByZeroError:
localError(g.config, n.info, "division by zero")
of nkAddr:
var a = getConstExpr(m, n.sons[0], g)
if a != nil:
result = n
n.sons[0] = a
of nkBracket:
result = copyTree(n)
for i in countup(0, sonsLen(n) - 1):
var a = getConstExpr(m, n.sons[i], g)
if a == nil: return nil
result.sons[i] = a
incl(result.flags, nfAllConst)
of nkRange:
var a = getConstExpr(m, n.sons[0], g)
if a == nil: return
var b = getConstExpr(m, n.sons[1], g)
if b == nil: return
result = copyNode(n)
addSon(result, a)
addSon(result, b)
of nkCurly:
result = copyTree(n)
for i in countup(0, sonsLen(n) - 1):
var a = getConstExpr(m, n.sons[i], g)
if a == nil: return nil
result.sons[i] = a
incl(result.flags, nfAllConst)
#of nkObjConstr:
# result = copyTree(n)
# for i in countup(1, sonsLen(n) - 1):
# var a = getConstExpr(m, n.sons[i].sons[1])
# if a == nil: return nil
# result.sons[i].sons[1] = a
# incl(result.flags, nfAllConst)
of nkPar, nkTupleConstr:
# tuple constructor
result = copyTree(n)
if (sonsLen(n) > 0) and (n.sons[0].kind == nkExprColonExpr):
for i in countup(0, sonsLen(n) - 1):
var a = getConstExpr(m, n.sons[i].sons[1], g)
if a == nil: return nil
result.sons[i].sons[1] = a
else:
for i in countup(0, sonsLen(n) - 1):
var a = getConstExpr(m, n.sons[i], g)
if a == nil: return nil
result.sons[i] = a
incl(result.flags, nfAllConst)
of nkChckRangeF, nkChckRange64, nkChckRange:
var a = getConstExpr(m, n.sons[0], g)
if a == nil: return
if leValueConv(n.sons[1], a) and leValueConv(a, n.sons[2]):
result = a # a <= x and x <= b
result.typ = n.typ
else:
localError(g.config, n.info,
"conversion from $1 to $2 is invalid" %
[typeToString(n.sons[0].typ), typeToString(n.typ)])
of nkStringToCString, nkCStringToString:
var a = getConstExpr(m, n.sons[0], g)
if a == nil: return
result = a
result.typ = n.typ
of nkHiddenStdConv, nkHiddenSubConv, nkConv:
var a = getConstExpr(m, n.sons[1], g)
if a == nil: return
# XXX: we should enable `check` for other conversion types too
result = foldConv(n, a, g, check=n.kind == nkHiddenSubConv)
of nkCast:
var a = getConstExpr(m, n.sons[1], g)
if a == nil: return
if n.typ != nil and n.typ.kind in NilableTypes:
# we allow compile-time 'cast' for pointer types:
result = a
result.typ = n.typ
of nkBracketExpr: result = foldArrayAccess(m, n, g)
of nkDotExpr: result = foldFieldAccess(m, n, g)
of nkStmtListExpr:
if n.len == 2 and n[0].kind == nkComesFrom:
result = getConstExpr(m, n[1], g)
else:
discard