llvm/lib/CodeGen/LiveInterval.cpp

870 lines
28 KiB
C++
Raw Normal View History

//===-- LiveInterval.cpp - Live Interval Representation -------------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the LiveRange and LiveInterval classes. Given some
// numbering of each the machine instructions an interval [i, j) is said to be a
// live interval for register v if there is no instruction with number j' > j
// such that v is live at j' abd there is no instruction with number i' < i such
// that v is live at i'. In this implementation intervals can have holes,
// i.e. an interval might look like [1,20), [50,65), [1000,1001). Each
// individual range is represented as an instance of LiveRange, and the whole
// interval is represented as an instance of LiveInterval.
//
//===----------------------------------------------------------------------===//
#include "llvm/CodeGen/LiveInterval.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/Support/Streams.h"
#include "llvm/Target/TargetRegisterInfo.h"
#include <algorithm>
#include <ostream>
using namespace llvm;
// An example for liveAt():
//
// this = [1,4), liveAt(0) will return false. The instruction defining this
// spans slots [0,3]. The interval belongs to an spilled definition of the
// variable it represents. This is because slot 1 is used (def slot) and spans
// up to slot 3 (store slot).
//
bool LiveInterval::liveAt(unsigned I) const {
Ranges::const_iterator r = std::upper_bound(ranges.begin(), ranges.end(), I);
if (r == ranges.begin())
return false;
--r;
return r->contains(I);
}
// liveBeforeAndAt - Check if the interval is live at the index and the index
// just before it. If index is liveAt, check if it starts a new live range.
// If it does, then check if the previous live range ends at index-1.
bool LiveInterval::liveBeforeAndAt(unsigned I) const {
Ranges::const_iterator r = std::upper_bound(ranges.begin(), ranges.end(), I);
if (r == ranges.begin())
return false;
--r;
if (!r->contains(I))
return false;
if (I != r->start)
return true;
// I is the start of a live range. Check if the previous live range ends
// at I-1.
if (r == ranges.begin())
return false;
return r->end == I;
}
// overlaps - Return true if the intersection of the two live intervals is
// not empty.
//
// An example for overlaps():
//
// 0: A = ...
// 4: B = ...
// 8: C = A + B ;; last use of A
//
// The live intervals should look like:
//
// A = [3, 11)
// B = [7, x)
// C = [11, y)
//
// A->overlaps(C) should return false since we want to be able to join
// A and C.
//
bool LiveInterval::overlapsFrom(const LiveInterval& other,
const_iterator StartPos) const {
const_iterator i = begin();
const_iterator ie = end();
const_iterator j = StartPos;
const_iterator je = other.end();
assert((StartPos->start <= i->start || StartPos == other.begin()) &&
StartPos != other.end() && "Bogus start position hint!");
This patch makes use of the infrastructure implemented before to safely and aggressively coallesce live ranges even if they overlap. Consider this LLVM code for example: int %test(int %X) { %Y = mul int %X, 1 ;; Codegens to Y = X %Z = add int %X, %Y ret int %Z } The mul is just there to get a copy into the code stream. This produces this machine code: (0x869e5a8, LLVM BB @0x869b9a0): %reg1024 = mov <fi#-2>, 1, %NOREG, 0 ;; "X" %reg1025 = mov %reg1024 ;; "Y" (subsumed by X) %reg1026 = add %reg1024, %reg1025 %EAX = mov %reg1026 ret Note that the life times of reg1024 and reg1025 overlap, even though they contain the same value. This results in this machine code: test: mov %EAX, DWORD PTR [%ESP + 4] mov %ECX, %EAX add %EAX, %ECX ret Another, worse case involves loops and PHI nodes. Consider this trivial loop: testcase: int %test2(int %X) { entry: br label %Loop Loop: %Y = phi int [%X, %entry], [%Z, %Loop] %Z = add int %Y, 1 %cond = seteq int %Z, 100 br bool %cond, label %Out, label %Loop Out: ret int %Z } Because of interactions between the PHI elimination pass and the register allocator, this got compiled to this code: test2: mov %ECX, DWORD PTR [%ESP + 4] .LBBtest2_1: *** mov %EAX, %ECX inc %EAX cmp %EAX, 100 *** mov %ECX, %EAX jne .LBBtest2_1 ret Or on powerpc, this code: _test2: mflr r0 stw r0, 8(r1) stwu r1, -60(r1) .LBB_test2_1: addi r2, r3, 1 cmpwi cr0, r2, 100 *** or r3, r2, r2 bne cr0, .LBB_test2_1 *** or r3, r2, r2 lwz r0, 68(r1) mtlr r0 addi r1, r1, 60 blr 0 With this improvement in place, we now generate this code for these two testcases, which is what we want: test: mov %EAX, DWORD PTR [%ESP + 4] add %EAX, %EAX ret test2: mov %EAX, DWORD PTR [%ESP + 4] .LBBtest2_1: inc %EAX cmp %EAX, 100 jne .LBBtest2_1 # Loop ret Or on PPC: _test2: mflr r0 stw r0, 8(r1) stwu r1, -60(r1) .LBB_test2_1: addi r3, r3, 1 cmpwi cr0, r3, 100 bne cr0, .LBB_test2_1 lwz r0, 68(r1) mtlr r0 addi r1, r1, 60 blr 0 Static numbers for spill code loads/stores/reg-reg copies (smaller is better): em3d: before: 47/25/26 after: 44/22/24 164.gzip: before: 433/245/310 after: 403/231/278 175.vpr: before: 3721/2189/1581 after: 4144/2081/1423 176.gcc: before: 26195/8866/9235 after: 25942/8082/8275 186.crafty: before: 4295/2587/3079 after: 4119/2519/2916 252.eon: before: 12754/7585/5803 after: 12508/7425/5643 256.bzip2: before: 463/226/315 after: 482:241/309 Runtime perf number samples on X86: gzip: before: 41.09 after: 39.86 bzip2: runtime: before: 56.71s after: 57.07s gcc: before: 6.16 after: 6.12 eon: before: 2.03s after: 2.00s git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@15194 91177308-0d34-0410-b5e6-96231b3b80d8
2004-07-25 07:11:19 +00:00
if (i->start < j->start) {
i = std::upper_bound(i, ie, j->start);
if (i != ranges.begin()) --i;
} else if (j->start < i->start) {
++StartPos;
if (StartPos != other.end() && StartPos->start <= i->start) {
assert(StartPos < other.end() && i < end());
j = std::upper_bound(j, je, i->start);
if (j != other.ranges.begin()) --j;
}
} else {
return true;
}
if (j == je) return false;
while (i != ie) {
if (i->start > j->start) {
std::swap(i, j);
std::swap(ie, je);
}
if (i->end > j->start)
return true;
++i;
}
return false;
}
/// overlaps - Return true if the live interval overlaps a range specified
/// by [Start, End).
bool LiveInterval::overlaps(unsigned Start, unsigned End) const {
assert(Start < End && "Invalid range");
const_iterator I = begin();
const_iterator E = end();
const_iterator si = std::upper_bound(I, E, Start);
const_iterator ei = std::upper_bound(I, E, End);
if (si != ei)
return true;
if (si == I)
return false;
--si;
return si->contains(Start);
}
/// extendIntervalEndTo - This method is used when we want to extend the range
/// specified by I to end at the specified endpoint. To do this, we should
/// merge and eliminate all ranges that this will overlap with. The iterator is
/// not invalidated.
void LiveInterval::extendIntervalEndTo(Ranges::iterator I, unsigned NewEnd) {
assert(I != ranges.end() && "Not a valid interval!");
VNInfo *ValNo = I->valno;
unsigned OldEnd = I->end;
// Search for the first interval that we can't merge with.
Ranges::iterator MergeTo = next(I);
for (; MergeTo != ranges.end() && NewEnd >= MergeTo->end; ++MergeTo) {
assert(MergeTo->valno == ValNo && "Cannot merge with differing values!");
}
// If NewEnd was in the middle of an interval, make sure to get its endpoint.
I->end = std::max(NewEnd, prior(MergeTo)->end);
// Erase any dead ranges.
ranges.erase(next(I), MergeTo);
// Update kill info.
removeKills(ValNo, OldEnd, I->end-1);
// If the newly formed range now touches the range after it and if they have
// the same value number, merge the two ranges into one range.
Ranges::iterator Next = next(I);
if (Next != ranges.end() && Next->start <= I->end && Next->valno == ValNo) {
I->end = Next->end;
ranges.erase(Next);
}
}
/// extendIntervalStartTo - This method is used when we want to extend the range
/// specified by I to start at the specified endpoint. To do this, we should
/// merge and eliminate all ranges that this will overlap with.
LiveInterval::Ranges::iterator
LiveInterval::extendIntervalStartTo(Ranges::iterator I, unsigned NewStart) {
assert(I != ranges.end() && "Not a valid interval!");
VNInfo *ValNo = I->valno;
// Search for the first interval that we can't merge with.
Ranges::iterator MergeTo = I;
do {
if (MergeTo == ranges.begin()) {
I->start = NewStart;
ranges.erase(MergeTo, I);
return I;
}
assert(MergeTo->valno == ValNo && "Cannot merge with differing values!");
--MergeTo;
} while (NewStart <= MergeTo->start);
// If we start in the middle of another interval, just delete a range and
// extend that interval.
if (MergeTo->end >= NewStart && MergeTo->valno == ValNo) {
MergeTo->end = I->end;
} else {
// Otherwise, extend the interval right after.
++MergeTo;
MergeTo->start = NewStart;
MergeTo->end = I->end;
}
ranges.erase(next(MergeTo), next(I));
return MergeTo;
}
LiveInterval::iterator
LiveInterval::addRangeFrom(LiveRange LR, iterator From) {
unsigned Start = LR.start, End = LR.end;
iterator it = std::upper_bound(From, ranges.end(), Start);
// If the inserted interval starts in the middle or right at the end of
// another interval, just extend that interval to contain the range of LR.
if (it != ranges.begin()) {
iterator B = prior(it);
if (LR.valno == B->valno) {
if (B->start <= Start && B->end >= Start) {
extendIntervalEndTo(B, End);
return B;
}
} else {
// Check to make sure that we are not overlapping two live ranges with
// different valno's.
assert(B->end <= Start &&
"Cannot overlap two LiveRanges with differing ValID's"
" (did you def the same reg twice in a MachineInstr?)");
}
}
// Otherwise, if this range ends in the middle of, or right next to, another
// interval, merge it into that interval.
if (it != ranges.end()) {
if (LR.valno == it->valno) {
if (it->start <= End) {
it = extendIntervalStartTo(it, Start);
// If LR is a complete superset of an interval, we may need to grow its
// endpoint as well.
if (End > it->end)
extendIntervalEndTo(it, End);
else if (End < it->end)
// Overlapping intervals, there might have been a kill here.
removeKill(it->valno, End);
return it;
}
} else {
// Check to make sure that we are not overlapping two live ranges with
// different valno's.
assert(it->start >= End &&
"Cannot overlap two LiveRanges with differing ValID's");
}
}
// Otherwise, this is just a new range that doesn't interact with anything.
// Insert it.
return ranges.insert(it, LR);
}
/// isInOneLiveRange - Return true if the range specified is entirely in the
/// a single LiveRange of the live interval.
bool LiveInterval::isInOneLiveRange(unsigned Start, unsigned End) {
Ranges::iterator I = std::upper_bound(ranges.begin(), ranges.end(), Start);
if (I == ranges.begin())
return false;
--I;
return I->contains(Start) && I->contains(End-1);
}
/// removeRange - Remove the specified range from this interval. Note that
/// the range must be in a single LiveRange in its entirety.
void LiveInterval::removeRange(unsigned Start, unsigned End,
bool RemoveDeadValNo) {
// Find the LiveRange containing this span.
Ranges::iterator I = std::upper_bound(ranges.begin(), ranges.end(), Start);
assert(I != ranges.begin() && "Range is not in interval!");
--I;
assert(I->contains(Start) && I->contains(End-1) &&
"Range is not entirely in interval!");
// If the span we are removing is at the start of the LiveRange, adjust it.
VNInfo *ValNo = I->valno;
if (I->start == Start) {
if (I->end == End) {
removeKills(I->valno, Start, End);
if (RemoveDeadValNo) {
// Check if val# is dead.
bool isDead = true;
for (const_iterator II = begin(), EE = end(); II != EE; ++II)
if (II != I && II->valno == ValNo) {
isDead = false;
break;
}
if (isDead) {
// Now that ValNo is dead, remove it. If it is the largest value
// number, just nuke it (and any other deleted values neighboring it),
// otherwise mark it as ~1U so it can be nuked later.
if (ValNo->id == getNumValNums()-1) {
do {
VNInfo *VNI = valnos.back();
valnos.pop_back();
VNI->~VNInfo();
} while (!valnos.empty() && valnos.back()->def == ~1U);
} else {
ValNo->def = ~1U;
}
}
}
ranges.erase(I); // Removed the whole LiveRange.
} else
I->start = End;
return;
}
// Otherwise if the span we are removing is at the end of the LiveRange,
// adjust the other way.
if (I->end == End) {
removeKills(ValNo, Start, End);
I->end = Start;
return;
}
// Otherwise, we are splitting the LiveRange into two pieces.
unsigned OldEnd = I->end;
I->end = Start; // Trim the old interval.
// Insert the new one.
ranges.insert(next(I), LiveRange(End, OldEnd, ValNo));
}
/// removeValNo - Remove all the ranges defined by the specified value#.
/// Also remove the value# from value# list.
void LiveInterval::removeValNo(VNInfo *ValNo) {
if (empty()) return;
Ranges::iterator I = ranges.end();
Ranges::iterator E = ranges.begin();
do {
--I;
if (I->valno == ValNo)
ranges.erase(I);
} while (I != E);
// Now that ValNo is dead, remove it. If it is the largest value
// number, just nuke it (and any other deleted values neighboring it),
// otherwise mark it as ~1U so it can be nuked later.
if (ValNo->id == getNumValNums()-1) {
do {
VNInfo *VNI = valnos.back();
valnos.pop_back();
VNI->~VNInfo();
} while (!valnos.empty() && valnos.back()->def == ~1U);
} else {
ValNo->def = ~1U;
}
}
/// scaleNumbering - Renumber VNI and ranges to provide gaps for new
/// instructions.
void LiveInterval::scaleNumbering(unsigned factor) {
// Scale ranges.
for (iterator RI = begin(), RE = end(); RI != RE; ++RI) {
RI->start = InstrSlots::scale(RI->start, factor);
RI->end = InstrSlots::scale(RI->end, factor);
}
// Scale VNI info.
for (vni_iterator VNI = vni_begin(), VNIE = vni_end(); VNI != VNIE; ++VNI) {
VNInfo *vni = *VNI;
if (vni->def != ~0U && vni->def != ~1U) {
vni->def = InstrSlots::scale(vni->def, factor);
}
for (unsigned i = 0; i < vni->kills.size(); ++i) {
if (vni->kills[i] != 0)
vni->kills[i] = InstrSlots::scale(vni->kills[i], factor);
}
}
}
/// getLiveRangeContaining - Return the live range that contains the
/// specified index, or null if there is none.
LiveInterval::const_iterator
LiveInterval::FindLiveRangeContaining(unsigned Idx) const {
const_iterator It = std::upper_bound(begin(), end(), Idx);
if (It != ranges.begin()) {
--It;
if (It->contains(Idx))
return It;
}
return end();
}
LiveInterval::iterator
LiveInterval::FindLiveRangeContaining(unsigned Idx) {
iterator It = std::upper_bound(begin(), end(), Idx);
if (It != begin()) {
--It;
if (It->contains(Idx))
return It;
}
return end();
}
/// findDefinedVNInfo - Find the VNInfo that's defined at the specified index
/// (register interval) or defined by the specified register (stack inteval).
VNInfo *LiveInterval::findDefinedVNInfo(unsigned DefIdxOrReg) const {
VNInfo *VNI = NULL;
for (LiveInterval::const_vni_iterator i = vni_begin(), e = vni_end();
i != e; ++i)
if ((*i)->def == DefIdxOrReg) {
VNI = *i;
break;
}
return VNI;
}
/// join - Join two live intervals (this, and other) together. This applies
/// mappings to the value numbers in the LHS/RHS intervals as specified. If
/// the intervals are not joinable, this aborts.
void LiveInterval::join(LiveInterval &Other, const int *LHSValNoAssignments,
const int *RHSValNoAssignments,
SmallVector<VNInfo*, 16> &NewVNInfo,
MachineRegisterInfo *MRI) {
// Determine if any of our live range values are mapped. This is uncommon, so
// we want to avoid the interval scan if not.
bool MustMapCurValNos = false;
unsigned NumVals = getNumValNums();
unsigned NumNewVals = NewVNInfo.size();
for (unsigned i = 0; i != NumVals; ++i) {
unsigned LHSValID = LHSValNoAssignments[i];
if (i != LHSValID ||
(NewVNInfo[LHSValID] && NewVNInfo[LHSValID] != getValNumInfo(i)))
MustMapCurValNos = true;
}
// If we have to apply a mapping to our base interval assignment, rewrite it
// now.
if (MustMapCurValNos) {
// Map the first live range.
iterator OutIt = begin();
OutIt->valno = NewVNInfo[LHSValNoAssignments[OutIt->valno->id]];
++OutIt;
for (iterator I = OutIt, E = end(); I != E; ++I) {
OutIt->valno = NewVNInfo[LHSValNoAssignments[I->valno->id]];
// If this live range has the same value # as its immediate predecessor,
// and if they are neighbors, remove one LiveRange. This happens when we
// have [0,3:0)[4,7:1) and map 0/1 onto the same value #.
if (OutIt->valno == (OutIt-1)->valno && (OutIt-1)->end == OutIt->start) {
(OutIt-1)->end = OutIt->end;
} else {
if (I != OutIt) {
OutIt->start = I->start;
OutIt->end = I->end;
}
// Didn't merge, on to the next one.
++OutIt;
}
}
// If we merge some live ranges, chop off the end.
ranges.erase(OutIt, end());
}
// Remember assignements because val# ids are changing.
SmallVector<unsigned, 16> OtherAssignments;
for (iterator I = Other.begin(), E = Other.end(); I != E; ++I)
OtherAssignments.push_back(RHSValNoAssignments[I->valno->id]);
// Update val# info. Renumber them and make sure they all belong to this
// LiveInterval now. Also remove dead val#'s.
unsigned NumValNos = 0;
for (unsigned i = 0; i < NumNewVals; ++i) {
VNInfo *VNI = NewVNInfo[i];
if (VNI) {
if (NumValNos >= NumVals)
valnos.push_back(VNI);
else
valnos[NumValNos] = VNI;
VNI->id = NumValNos++; // Renumber val#.
}
}
if (NumNewVals < NumVals)
valnos.resize(NumNewVals); // shrinkify
// Okay, now insert the RHS live ranges into the LHS.
iterator InsertPos = begin();
unsigned RangeNo = 0;
for (iterator I = Other.begin(), E = Other.end(); I != E; ++I, ++RangeNo) {
// Map the valno in the other live range to the current live range.
I->valno = NewVNInfo[OtherAssignments[RangeNo]];
assert(I->valno && "Adding a dead range?");
InsertPos = addRangeFrom(*I, InsertPos);
}
weight += Other.weight;
// Update regalloc hint if currently there isn't one.
if (TargetRegisterInfo::isVirtualRegister(reg) &&
TargetRegisterInfo::isVirtualRegister(Other.reg)) {
std::pair<MachineRegisterInfo::RegAllocHintType, unsigned> Hint =
MRI->getRegAllocationHint(reg);
if (Hint.first == MachineRegisterInfo::RA_None) {
std::pair<MachineRegisterInfo::RegAllocHintType, unsigned> OtherHint =
MRI->getRegAllocationHint(Other.reg);
if (OtherHint.first != MachineRegisterInfo::RA_None)
MRI->setRegAllocationHint(reg, OtherHint.first, OtherHint.second);
}
}
}
/// MergeRangesInAsValue - Merge all of the intervals in RHS into this live
/// interval as the specified value number. The LiveRanges in RHS are
/// allowed to overlap with LiveRanges in the current interval, but only if
/// the overlapping LiveRanges have the specified value number.
void LiveInterval::MergeRangesInAsValue(const LiveInterval &RHS,
VNInfo *LHSValNo) {
// TODO: Make this more efficient.
iterator InsertPos = begin();
for (const_iterator I = RHS.begin(), E = RHS.end(); I != E; ++I) {
// Map the valno in the other live range to the current live range.
LiveRange Tmp = *I;
Tmp.valno = LHSValNo;
InsertPos = addRangeFrom(Tmp, InsertPos);
}
}
/// MergeValueInAsValue - Merge all of the live ranges of a specific val#
/// in RHS into this live interval as the specified value number.
/// The LiveRanges in RHS are allowed to overlap with LiveRanges in the
/// current interval, it will replace the value numbers of the overlaped
/// live ranges with the specified value number.
void LiveInterval::MergeValueInAsValue(const LiveInterval &RHS,
const VNInfo *RHSValNo, VNInfo *LHSValNo) {
SmallVector<VNInfo*, 4> ReplacedValNos;
iterator IP = begin();
for (const_iterator I = RHS.begin(), E = RHS.end(); I != E; ++I) {
if (I->valno != RHSValNo)
continue;
unsigned Start = I->start, End = I->end;
IP = std::upper_bound(IP, end(), Start);
// If the start of this range overlaps with an existing liverange, trim it.
if (IP != begin() && IP[-1].end > Start) {
if (IP[-1].valno != LHSValNo) {
ReplacedValNos.push_back(IP[-1].valno);
IP[-1].valno = LHSValNo; // Update val#.
}
Start = IP[-1].end;
// Trimmed away the whole range?
if (Start >= End) continue;
}
// If the end of this range overlaps with an existing liverange, trim it.
if (IP != end() && End > IP->start) {
if (IP->valno != LHSValNo) {
ReplacedValNos.push_back(IP->valno);
IP->valno = LHSValNo; // Update val#.
}
End = IP->start;
// If this trimmed away the whole range, ignore it.
if (Start == End) continue;
}
// Map the valno in the other live range to the current live range.
IP = addRangeFrom(LiveRange(Start, End, LHSValNo), IP);
}
SmallSet<VNInfo*, 4> Seen;
for (unsigned i = 0, e = ReplacedValNos.size(); i != e; ++i) {
VNInfo *V1 = ReplacedValNos[i];
if (Seen.insert(V1)) {
bool isDead = true;
for (const_iterator I = begin(), E = end(); I != E; ++I)
if (I->valno == V1) {
isDead = false;
break;
}
if (isDead) {
// Now that V1 is dead, remove it. If it is the largest value number,
// just nuke it (and any other deleted values neighboring it), otherwise
// mark it as ~1U so it can be nuked later.
if (V1->id == getNumValNums()-1) {
do {
VNInfo *VNI = valnos.back();
valnos.pop_back();
VNI->~VNInfo();
} while (!valnos.empty() && valnos.back()->def == ~1U);
} else {
V1->def = ~1U;
}
}
}
}
}
/// MergeInClobberRanges - For any live ranges that are not defined in the
/// current interval, but are defined in the Clobbers interval, mark them
/// used with an unknown definition value.
void LiveInterval::MergeInClobberRanges(const LiveInterval &Clobbers,
BumpPtrAllocator &VNInfoAllocator) {
if (Clobbers.empty()) return;
DenseMap<VNInfo*, VNInfo*> ValNoMaps;
VNInfo *UnusedValNo = 0;
iterator IP = begin();
for (const_iterator I = Clobbers.begin(), E = Clobbers.end(); I != E; ++I) {
// For every val# in the Clobbers interval, create a new "unknown" val#.
VNInfo *ClobberValNo = 0;
DenseMap<VNInfo*, VNInfo*>::iterator VI = ValNoMaps.find(I->valno);
if (VI != ValNoMaps.end())
ClobberValNo = VI->second;
else if (UnusedValNo)
ClobberValNo = UnusedValNo;
else {
UnusedValNo = ClobberValNo = getNextValue(~0U, 0, VNInfoAllocator);
ValNoMaps.insert(std::make_pair(I->valno, ClobberValNo));
}
Implement support for using modeling implicit-zero-extension on x86-64 with SUBREG_TO_REG, teach SimpleRegisterCoalescing to coalesce SUBREG_TO_REG instructions (which are similar to INSERT_SUBREG instructions), and teach the DAGCombiner to take advantage of this on targets which support it. This eliminates many redundant zero-extension operations on x86-64. This adds a new TargetLowering hook, isZExtFree. It's similar to isTruncateFree, except it only applies to actual definitions, and not no-op truncates which may not zero the high bits. Also, this adds a new optimization to SimplifyDemandedBits: transform operations like x+y into (zext (add (trunc x), (trunc y))) on targets where all the casts are no-ops. In contexts where the high part of the add is explicitly masked off, this allows the mask operation to be eliminated. Fix the DAGCombiner to avoid undoing these transformations to eliminate casts on targets where the casts are no-ops. Also, this adds a new two-address lowering heuristic. Since two-address lowering runs before coalescing, it helps to be able to look through copies when deciding whether commuting and/or three-address conversion are profitable. Also, fix a bug in LiveInterval::MergeInClobberRanges. It didn't handle the case that a clobber range extended both before and beyond an existing live range. In that case, multiple live ranges need to be added. This was exposed by the new subreg coalescing code. Remove 2008-05-06-SpillerBug.ll. It was bugpoint-reduced, and the spiller behavior it was looking for no longer occurrs with the new instruction selection. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@68576 91177308-0d34-0410-b5e6-96231b3b80d8
2009-04-08 00:15:30 +00:00
bool Done = false;
unsigned Start = I->start, End = I->end;
Implement support for using modeling implicit-zero-extension on x86-64 with SUBREG_TO_REG, teach SimpleRegisterCoalescing to coalesce SUBREG_TO_REG instructions (which are similar to INSERT_SUBREG instructions), and teach the DAGCombiner to take advantage of this on targets which support it. This eliminates many redundant zero-extension operations on x86-64. This adds a new TargetLowering hook, isZExtFree. It's similar to isTruncateFree, except it only applies to actual definitions, and not no-op truncates which may not zero the high bits. Also, this adds a new optimization to SimplifyDemandedBits: transform operations like x+y into (zext (add (trunc x), (trunc y))) on targets where all the casts are no-ops. In contexts where the high part of the add is explicitly masked off, this allows the mask operation to be eliminated. Fix the DAGCombiner to avoid undoing these transformations to eliminate casts on targets where the casts are no-ops. Also, this adds a new two-address lowering heuristic. Since two-address lowering runs before coalescing, it helps to be able to look through copies when deciding whether commuting and/or three-address conversion are profitable. Also, fix a bug in LiveInterval::MergeInClobberRanges. It didn't handle the case that a clobber range extended both before and beyond an existing live range. In that case, multiple live ranges need to be added. This was exposed by the new subreg coalescing code. Remove 2008-05-06-SpillerBug.ll. It was bugpoint-reduced, and the spiller behavior it was looking for no longer occurrs with the new instruction selection. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@68576 91177308-0d34-0410-b5e6-96231b3b80d8
2009-04-08 00:15:30 +00:00
// If a clobber range starts before an existing range and ends after
// it, the clobber range will need to be split into multiple ranges.
// Loop until the entire clobber range is handled.
while (!Done) {
Done = true;
IP = std::upper_bound(IP, end(), Start);
unsigned SubRangeStart = Start;
unsigned SubRangeEnd = End;
// If the start of this range overlaps with an existing liverange, trim it.
if (IP != begin() && IP[-1].end > SubRangeStart) {
SubRangeStart = IP[-1].end;
// Trimmed away the whole range?
if (SubRangeStart >= SubRangeEnd) continue;
}
// If the end of this range overlaps with an existing liverange, trim it.
if (IP != end() && SubRangeEnd > IP->start) {
// If the clobber live range extends beyond the existing live range,
// it'll need at least another live range, so set the flag to keep
// iterating.
if (SubRangeEnd > IP->end) {
Start = IP->end;
Done = false;
}
SubRangeEnd = IP->start;
// If this trimmed away the whole range, ignore it.
if (SubRangeStart == SubRangeEnd) continue;
}
// Insert the clobber interval.
IP = addRangeFrom(LiveRange(SubRangeStart, SubRangeEnd, ClobberValNo),
IP);
UnusedValNo = 0;
}
}
if (UnusedValNo) {
// Delete the last unused val#.
valnos.pop_back();
UnusedValNo->~VNInfo();
}
}
void LiveInterval::MergeInClobberRange(unsigned Start, unsigned End,
BumpPtrAllocator &VNInfoAllocator) {
// Find a value # to use for the clobber ranges. If there is already a value#
// for unknown values, use it.
VNInfo *ClobberValNo = getNextValue(~0U, 0, VNInfoAllocator);
iterator IP = begin();
IP = std::upper_bound(IP, end(), Start);
// If the start of this range overlaps with an existing liverange, trim it.
if (IP != begin() && IP[-1].end > Start) {
Start = IP[-1].end;
// Trimmed away the whole range?
if (Start >= End) return;
}
// If the end of this range overlaps with an existing liverange, trim it.
if (IP != end() && End > IP->start) {
End = IP->start;
// If this trimmed away the whole range, ignore it.
if (Start == End) return;
}
// Insert the clobber interval.
addRangeFrom(LiveRange(Start, End, ClobberValNo), IP);
}
/// MergeValueNumberInto - This method is called when two value nubmers
/// are found to be equivalent. This eliminates V1, replacing all
/// LiveRanges with the V1 value number with the V2 value number. This can
/// cause merging of V1/V2 values numbers and compaction of the value space.
VNInfo* LiveInterval::MergeValueNumberInto(VNInfo *V1, VNInfo *V2) {
assert(V1 != V2 && "Identical value#'s are always equivalent!");
// This code actually merges the (numerically) larger value number into the
// smaller value number, which is likely to allow us to compactify the value
// space. The only thing we have to be careful of is to preserve the
// instruction that defines the result value.
// Make sure V2 is smaller than V1.
if (V1->id < V2->id) {
copyValNumInfo(V1, V2);
std::swap(V1, V2);
}
// Merge V1 live ranges into V2.
for (iterator I = begin(); I != end(); ) {
iterator LR = I++;
if (LR->valno != V1) continue; // Not a V1 LiveRange.
// Okay, we found a V1 live range. If it had a previous, touching, V2 live
// range, extend it.
if (LR != begin()) {
iterator Prev = LR-1;
if (Prev->valno == V2 && Prev->end == LR->start) {
Prev->end = LR->end;
// Erase this live-range.
ranges.erase(LR);
I = Prev+1;
LR = Prev;
}
}
// Okay, now we have a V1 or V2 live range that is maximally merged forward.
// Ensure that it is a V2 live-range.
LR->valno = V2;
// If we can merge it into later V2 live ranges, do so now. We ignore any
// following V1 live ranges, as they will be merged in subsequent iterations
// of the loop.
if (I != end()) {
if (I->start == LR->end && I->valno == V2) {
LR->end = I->end;
ranges.erase(I);
I = LR+1;
}
}
}
// Now that V1 is dead, remove it. If it is the largest value number, just
// nuke it (and any other deleted values neighboring it), otherwise mark it as
// ~1U so it can be nuked later.
if (V1->id == getNumValNums()-1) {
do {
VNInfo *VNI = valnos.back();
valnos.pop_back();
VNI->~VNInfo();
} while (valnos.back()->def == ~1U);
} else {
V1->def = ~1U;
}
return V2;
}
void LiveInterval::Copy(const LiveInterval &RHS,
MachineRegisterInfo *MRI,
BumpPtrAllocator &VNInfoAllocator) {
ranges.clear();
valnos.clear();
std::pair<MachineRegisterInfo::RegAllocHintType, unsigned> Hint =
MRI->getRegAllocationHint(RHS.reg);
MRI->setRegAllocationHint(reg, Hint.first, Hint.second);
weight = RHS.weight;
for (unsigned i = 0, e = RHS.getNumValNums(); i != e; ++i) {
const VNInfo *VNI = RHS.getValNumInfo(i);
VNInfo *NewVNI = getNextValue(~0U, 0, VNInfoAllocator);
copyValNumInfo(NewVNI, VNI);
}
for (unsigned i = 0, e = RHS.ranges.size(); i != e; ++i) {
const LiveRange &LR = RHS.ranges[i];
addRange(LiveRange(LR.start, LR.end, getValNumInfo(LR.valno->id)));
}
}
unsigned LiveInterval::getSize() const {
unsigned Sum = 0;
for (const_iterator I = begin(), E = end(); I != E; ++I)
Sum += I->end - I->start;
return Sum;
}
std::ostream& llvm::operator<<(std::ostream& os, const LiveRange &LR) {
return os << '[' << LR.start << ',' << LR.end << ':' << LR.valno->id << ")";
}
void LiveRange::dump() const {
cerr << *this << "\n";
}
void LiveInterval::print(std::ostream &OS,
const TargetRegisterInfo *TRI) const {
if (isStackSlot())
OS << "SS#" << getStackSlotIndex();
else if (TRI && TargetRegisterInfo::isPhysicalRegister(reg))
OS << TRI->getName(reg);
else
OS << "%reg" << reg;
OS << ',' << weight;
if (empty())
OS << " EMPTY";
else {
OS << " = ";
for (LiveInterval::Ranges::const_iterator I = ranges.begin(),
E = ranges.end(); I != E; ++I)
OS << *I;
}
// Print value number info.
if (getNumValNums()) {
OS << " ";
unsigned vnum = 0;
for (const_vni_iterator i = vni_begin(), e = vni_end(); i != e;
++i, ++vnum) {
const VNInfo *vni = *i;
if (vnum) OS << " ";
OS << vnum << "@";
if (vni->def == ~1U) {
OS << "x";
} else {
if (vni->def == ~0U)
OS << "?";
else
OS << vni->def;
unsigned ee = vni->kills.size();
if (ee || vni->hasPHIKill) {
OS << "-(";
for (unsigned j = 0; j != ee; ++j) {
OS << vni->kills[j];
if (j != ee-1)
OS << " ";
}
if (vni->hasPHIKill) {
if (ee)
OS << " ";
OS << "phi";
}
OS << ")";
}
}
}
}
}
void LiveInterval::dump() const {
cerr << *this << "\n";
}
void LiveRange::print(std::ostream &os) const {
os << *this;
}