llvm-project/llvm/lib/Target/PowerPC/PPCAsmPrinter.cpp

1122 lines
42 KiB
C++
Raw Normal View History

//===-- PPCAsmPrinter.cpp - Print machine instrs to PowerPC assembly ------===//
//
2004-06-22 00:55:25 +08:00
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
2004-06-22 00:55:25 +08:00
//===----------------------------------------------------------------------===//
//
// This file contains a printer that converts from our internal representation
// of machine-dependent LLVM code to PowerPC assembly language. This printer is
2004-07-29 04:18:53 +08:00
// the output mechanism used by `llc'.
2004-06-22 00:55:25 +08:00
//
// Documentation at http://developer.apple.com/documentation/DeveloperTools/
// Reference/Assembler/ASMIntroduction/chapter_1_section_1.html
//
2004-06-22 00:55:25 +08:00
//===----------------------------------------------------------------------===//
#define DEBUG_TYPE "asmprinter"
#include "PPC.h"
#include "InstPrinter/PPCInstPrinter.h"
#include "MCTargetDesc/PPCPredicates.h"
#include "PPCSubtarget.h"
#include "PPCTargetMachine.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/StringExtras.h"
2004-06-22 00:55:25 +08:00
#include "llvm/Assembly/Writer.h"
#include "llvm/CodeGen/AsmPrinter.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
2004-06-22 00:55:25 +08:00
#include "llvm/CodeGen/MachineInstr.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineModuleInfoImpls.h"
#include "llvm/CodeGen/TargetLoweringObjectFileImpl.h"
#include "llvm/DebugInfo.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Module.h"
#include "llvm/MC/MCAsmInfo.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/MCExpr.h"
#include "llvm/MC/MCInst.h"
#include "llvm/MC/MCInstBuilder.h"
#include "llvm/MC/MCSectionELF.h"
#include "llvm/MC/MCSectionMachO.h"
#include "llvm/MC/MCStreamer.h"
#include "llvm/MC/MCSymbol.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ELF.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/TargetRegistry.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/Mangler.h"
#include "llvm/Target/TargetInstrInfo.h"
#include "llvm/Target/TargetOptions.h"
#include "llvm/Target/TargetRegisterInfo.h"
using namespace llvm;
2004-06-22 00:55:25 +08:00
namespace {
class PPCAsmPrinter : public AsmPrinter {
protected:
MapVector<MCSymbol*, MCSymbol*> TOC;
const PPCSubtarget &Subtarget;
uint64_t TOCLabelID;
public:
explicit PPCAsmPrinter(TargetMachine &TM, MCStreamer &Streamer)
: AsmPrinter(TM, Streamer),
Subtarget(TM.getSubtarget<PPCSubtarget>()), TOCLabelID(0) {}
2004-06-22 00:55:25 +08:00
virtual const char *getPassName() const {
return "PowerPC Assembly Printer";
2004-06-22 00:55:25 +08:00
}
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
MCSymbol *lookUpOrCreateTOCEntry(MCSymbol *Sym);
virtual void EmitInstruction(const MachineInstr *MI);
void printOperand(const MachineInstr *MI, unsigned OpNo, raw_ostream &O);
2008-08-09 02:22:59 +08:00
2006-02-02 06:38:46 +08:00
bool PrintAsmOperand(const MachineInstr *MI, unsigned OpNo,
unsigned AsmVariant, const char *ExtraCode,
raw_ostream &O);
bool PrintAsmMemoryOperand(const MachineInstr *MI, unsigned OpNo,
unsigned AsmVariant, const char *ExtraCode,
raw_ostream &O);
2008-08-09 02:22:59 +08:00
MachineLocation getDebugValueLocation(const MachineInstr *MI) const {
MachineLocation Location;
assert(MI->getNumOperands() == 4 && "Invalid no. of machine operands!");
// Frame address. Currently handles register +- offset only.
if (MI->getOperand(0).isReg() && MI->getOperand(2).isImm())
Location.set(MI->getOperand(0).getReg(), MI->getOperand(2).getImm());
else {
DEBUG(dbgs() << "DBG_VALUE instruction ignored! " << *MI << "\n");
}
return Location;
}
};
2008-08-09 02:22:59 +08:00
/// PPCLinuxAsmPrinter - PowerPC assembly printer, customized for Linux
class PPCLinuxAsmPrinter : public PPCAsmPrinter {
public:
explicit PPCLinuxAsmPrinter(TargetMachine &TM, MCStreamer &Streamer)
: PPCAsmPrinter(TM, Streamer) {}
virtual const char *getPassName() const {
return "Linux PPC Assembly Printer";
}
bool doFinalization(Module &M);
2008-08-09 02:22:59 +08:00
virtual void EmitFunctionEntryLabel();
void EmitFunctionBodyEnd();
};
2008-08-09 02:22:59 +08:00
/// PPCDarwinAsmPrinter - PowerPC assembly printer, customized for Darwin/Mac
/// OS X
class PPCDarwinAsmPrinter : public PPCAsmPrinter {
public:
explicit PPCDarwinAsmPrinter(TargetMachine &TM, MCStreamer &Streamer)
: PPCAsmPrinter(TM, Streamer) {}
virtual const char *getPassName() const {
return "Darwin PPC Assembly Printer";
}
2008-08-09 02:22:59 +08:00
2004-06-22 00:55:25 +08:00
bool doFinalization(Module &M);
void EmitStartOfAsmFile(Module &M);
2008-08-09 02:22:59 +08:00
void EmitFunctionStubs(const MachineModuleInfoMachO::SymbolListTy &Stubs);
2004-06-22 00:55:25 +08:00
};
} // end of anonymous namespace
/// stripRegisterPrefix - This method strips the character prefix from a
/// register name so that only the number is left. Used by for linux asm.
static const char *stripRegisterPrefix(const char *RegName) {
switch (RegName[0]) {
case 'r':
case 'f':
case 'v': return RegName + 1;
case 'c': if (RegName[1] == 'r') return RegName + 2;
}
return RegName;
}
void PPCAsmPrinter::printOperand(const MachineInstr *MI, unsigned OpNo,
raw_ostream &O) {
const MachineOperand &MO = MI->getOperand(OpNo);
2004-06-22 00:55:25 +08:00
switch (MO.getType()) {
case MachineOperand::MO_Register: {
const char *RegName = PPCInstPrinter::getRegisterName(MO.getReg());
// Linux assembler (Others?) does not take register mnemonics.
// FIXME - What about special registers used in mfspr/mtspr?
if (!Subtarget.isDarwin()) RegName = stripRegisterPrefix(RegName);
O << RegName;
return;
}
case MachineOperand::MO_Immediate:
O << MO.getImm();
return;
case MachineOperand::MO_MachineBasicBlock:
O << *MO.getMBB()->getSymbol();
return;
case MachineOperand::MO_JumpTableIndex:
O << MAI->getPrivateGlobalPrefix() << "JTI" << getFunctionNumber()
<< '_' << MO.getIndex();
// FIXME: PIC relocation model
2004-06-22 00:55:25 +08:00
return;
case MachineOperand::MO_ConstantPoolIndex:
O << MAI->getPrivateGlobalPrefix() << "CPI" << getFunctionNumber()
<< '_' << MO.getIndex();
2004-06-22 00:55:25 +08:00
return;
case MachineOperand::MO_BlockAddress:
O << *GetBlockAddressSymbol(MO.getBlockAddress());
return;
case MachineOperand::MO_ExternalSymbol: {
// Computing the address of an external symbol, not calling it.
if (TM.getRelocationModel() == Reloc::Static) {
O << *GetExternalSymbolSymbol(MO.getSymbolName());
return;
}
MCSymbol *NLPSym =
OutContext.GetOrCreateSymbol(StringRef(MAI->getGlobalPrefix())+
MO.getSymbolName()+"$non_lazy_ptr");
MachineModuleInfoImpl::StubValueTy &StubSym =
MMI->getObjFileInfo<MachineModuleInfoMachO>().getGVStubEntry(NLPSym);
if (StubSym.getPointer() == 0)
StubSym = MachineModuleInfoImpl::
StubValueTy(GetExternalSymbolSymbol(MO.getSymbolName()), true);
O << *NLPSym;
return;
}
case MachineOperand::MO_GlobalAddress: {
// Computing the address of a global symbol, not calling it.
const GlobalValue *GV = MO.getGlobal();
MCSymbol *SymToPrint;
// External or weakly linked global variables need non-lazily-resolved stubs
if (TM.getRelocationModel() != Reloc::Static &&
(GV->isDeclaration() || GV->isWeakForLinker())) {
if (!GV->hasHiddenVisibility()) {
SymToPrint = GetSymbolWithGlobalValueBase(GV, "$non_lazy_ptr");
MachineModuleInfoImpl::StubValueTy &StubSym =
MMI->getObjFileInfo<MachineModuleInfoMachO>()
.getGVStubEntry(SymToPrint);
if (StubSym.getPointer() == 0)
StubSym = MachineModuleInfoImpl::
StubValueTy(Mang->getSymbol(GV), !GV->hasInternalLinkage());
} else if (GV->isDeclaration() || GV->hasCommonLinkage() ||
GV->hasAvailableExternallyLinkage()) {
SymToPrint = GetSymbolWithGlobalValueBase(GV, "$non_lazy_ptr");
MachineModuleInfoImpl::StubValueTy &StubSym =
MMI->getObjFileInfo<MachineModuleInfoMachO>().
getHiddenGVStubEntry(SymToPrint);
if (StubSym.getPointer() == 0)
StubSym = MachineModuleInfoImpl::
StubValueTy(Mang->getSymbol(GV), !GV->hasInternalLinkage());
} else {
SymToPrint = Mang->getSymbol(GV);
}
} else {
SymToPrint = Mang->getSymbol(GV);
2004-06-22 00:55:25 +08:00
}
O << *SymToPrint;
2008-08-09 02:22:59 +08:00
printOffset(MO.getOffset(), O);
2004-06-22 00:55:25 +08:00
return;
}
2004-06-22 00:55:25 +08:00
default:
O << "<unknown operand type: " << MO.getType() << ">";
return;
2004-06-22 00:55:25 +08:00
}
}
/// PrintAsmOperand - Print out an operand for an inline asm expression.
///
bool PPCAsmPrinter::PrintAsmOperand(const MachineInstr *MI, unsigned OpNo,
2008-08-09 02:22:59 +08:00
unsigned AsmVariant,
const char *ExtraCode, raw_ostream &O) {
// Does this asm operand have a single letter operand modifier?
if (ExtraCode && ExtraCode[0]) {
if (ExtraCode[1] != 0) return true; // Unknown modifier.
2008-08-09 02:22:59 +08:00
switch (ExtraCode[0]) {
default:
// See if this is a generic print operand
return AsmPrinter::PrintAsmOperand(MI, OpNo, AsmVariant, ExtraCode, O);
case 'c': // Don't print "$" before a global var name or constant.
break; // PPC never has a prefix.
2008-08-09 02:22:59 +08:00
case 'L': // Write second word of DImode reference.
// Verify that this operand has two consecutive registers.
if (!MI->getOperand(OpNo).isReg() ||
OpNo+1 == MI->getNumOperands() ||
!MI->getOperand(OpNo+1).isReg())
return true;
++OpNo; // Return the high-part.
break;
case 'I':
// Write 'i' if an integer constant, otherwise nothing. Used to print
// addi vs add, etc.
if (MI->getOperand(OpNo).isImm())
O << "i";
return false;
}
}
2008-08-09 02:22:59 +08:00
printOperand(MI, OpNo, O);
return false;
}
// At the moment, all inline asm memory operands are a single register.
// In any case, the output of this routine should always be just one
// assembler operand.
bool PPCAsmPrinter::PrintAsmMemoryOperand(const MachineInstr *MI, unsigned OpNo,
2008-08-09 02:22:59 +08:00
unsigned AsmVariant,
const char *ExtraCode,
raw_ostream &O) {
if (ExtraCode && ExtraCode[0]) {
if (ExtraCode[1] != 0) return true; // Unknown modifier.
switch (ExtraCode[0]) {
default: return true; // Unknown modifier.
case 'y': // A memory reference for an X-form instruction
{
const char *RegName = "r0";
if (!Subtarget.isDarwin()) RegName = stripRegisterPrefix(RegName);
O << RegName << ", ";
printOperand(MI, OpNo, O);
return false;
}
}
}
assert(MI->getOperand(OpNo).isReg());
O << "0(";
printOperand(MI, OpNo, O);
O << ")";
return false;
}
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
/// lookUpOrCreateTOCEntry -- Given a symbol, look up whether a TOC entry
/// exists for it. If not, create one. Then return a symbol that references
/// the TOC entry.
MCSymbol *PPCAsmPrinter::lookUpOrCreateTOCEntry(MCSymbol *Sym) {
MCSymbol *&TOCEntry = TOC[Sym];
// To avoid name clash check if the name already exists.
while (TOCEntry == 0) {
if (OutContext.LookupSymbol(Twine(MAI->getPrivateGlobalPrefix()) +
"C" + Twine(TOCLabelID++)) == 0) {
TOCEntry = GetTempSymbol("C", TOCLabelID);
}
}
return TOCEntry;
}
/// EmitInstruction -- Print out a single PowerPC MI in Darwin syntax to
/// the current output stream.
2004-06-22 00:55:25 +08:00
///
void PPCAsmPrinter::EmitInstruction(const MachineInstr *MI) {
MCInst TmpInst;
// Lower multi-instruction pseudo operations.
switch (MI->getOpcode()) {
default: break;
case TargetOpcode::DBG_VALUE: {
if (!isVerbose() || !OutStreamer.hasRawTextSupport()) return;
SmallString<32> Str;
raw_svector_ostream O(Str);
unsigned NOps = MI->getNumOperands();
assert(NOps==4);
O << '\t' << MAI->getCommentString() << "DEBUG_VALUE: ";
// cast away const; DIetc do not take const operands for some reason.
DIVariable V(const_cast<MDNode *>(MI->getOperand(NOps-1).getMetadata()));
O << V.getName();
O << " <- ";
// Frame address. Currently handles register +- offset only.
assert(MI->getOperand(0).isReg() && MI->getOperand(1).isImm());
O << '['; printOperand(MI, 0, O); O << '+'; printOperand(MI, 1, O);
O << ']';
O << "+";
printOperand(MI, NOps-2, O);
OutStreamer.EmitRawText(O.str());
return;
}
case PPC::MovePCtoLR:
case PPC::MovePCtoLR8: {
// Transform %LR = MovePCtoLR
// Into this, where the label is the PIC base:
// bl L1$pb
// L1$pb:
MCSymbol *PICBase = MF->getPICBaseSymbol();
// Emit the 'bl'.
OutStreamer.EmitInstruction(MCInstBuilder(PPC::BL_Darwin) // Darwin vs SVR4 doesn't matter here.
// FIXME: We would like an efficient form for this, so we don't have to do
// a lot of extra uniquing.
.addExpr(MCSymbolRefExpr::Create(PICBase, OutContext)));
// Emit the label.
OutStreamer.EmitLabel(PICBase);
return;
}
case PPC::LDtocJTI:
case PPC::LDtocCPT:
case PPC::LDtoc: {
// Transform %X3 = LDtoc <ga:@min1>, %X2
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
// Change the opcode to LD, and the global address operand to be a
// reference to the TOC entry we will synthesize later.
TmpInst.setOpcode(PPC::LD);
const MachineOperand &MO = MI->getOperand(1);
// Map symbol -> label of TOC entry
assert(MO.isGlobal() || MO.isCPI() || MO.isJTI());
MCSymbol *MOSymbol = 0;
if (MO.isGlobal())
MOSymbol = Mang->getSymbol(MO.getGlobal());
else if (MO.isCPI())
MOSymbol = GetCPISymbol(MO.getIndex());
else if (MO.isJTI())
MOSymbol = GetJTISymbol(MO.getIndex());
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
MCSymbol *TOCEntry = lookUpOrCreateTOCEntry(MOSymbol);
const MCExpr *Exp =
MCSymbolRefExpr::Create(TOCEntry, MCSymbolRefExpr::VK_PPC_TOC_ENTRY,
OutContext);
TmpInst.getOperand(1) = MCOperand::CreateExpr(Exp);
OutStreamer.EmitInstruction(TmpInst);
return;
}
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
case PPC::ADDIStocHA: {
// Transform %Xd = ADDIStocHA %X2, <ga:@sym>
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
// Change the opcode to ADDIS8. If the global address is external,
// has common linkage, is a function address, or is a jump table
// address, then generate a TOC entry and reference that. Otherwise
// reference the symbol directly.
TmpInst.setOpcode(PPC::ADDIS8);
const MachineOperand &MO = MI->getOperand(2);
assert((MO.isGlobal() || MO.isCPI() || MO.isJTI()) &&
"Invalid operand for ADDIStocHA!");
MCSymbol *MOSymbol = 0;
bool IsExternal = false;
bool IsFunction = false;
bool IsCommon = false;
bool IsAvailExt = false;
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
if (MO.isGlobal()) {
const GlobalValue *GValue = MO.getGlobal();
const GlobalAlias *GAlias = dyn_cast<GlobalAlias>(GValue);
const GlobalValue *RealGValue = GAlias ?
GAlias->resolveAliasedGlobal(false) : GValue;
MOSymbol = Mang->getSymbol(RealGValue);
const GlobalVariable *GVar = dyn_cast<GlobalVariable>(RealGValue);
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
IsExternal = GVar && !GVar->hasInitializer();
IsCommon = GVar && RealGValue->hasCommonLinkage();
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
IsFunction = !GVar;
IsAvailExt = GVar && RealGValue->hasAvailableExternallyLinkage();
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
} else if (MO.isCPI())
MOSymbol = GetCPISymbol(MO.getIndex());
else if (MO.isJTI())
MOSymbol = GetJTISymbol(MO.getIndex());
if (IsExternal || IsFunction || IsCommon || IsAvailExt || MO.isJTI())
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
MOSymbol = lookUpOrCreateTOCEntry(MOSymbol);
const MCExpr *Exp =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_TOC16_HA,
OutContext);
TmpInst.getOperand(2) = MCOperand::CreateExpr(Exp);
OutStreamer.EmitInstruction(TmpInst);
return;
}
case PPC::LDtocL: {
// Transform %Xd = LDtocL <ga:@sym>, %Xs
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
// Change the opcode to LDrs, which is a form of LD with the offset
// specified by a SymbolLo. If the global address is external, has
// common linkage, or is a jump table address, then reference the
// associated TOC entry. Otherwise reference the symbol directly.
TmpInst.setOpcode(PPC::LDrs);
const MachineOperand &MO = MI->getOperand(1);
assert((MO.isGlobal() || MO.isJTI()) && "Invalid operand for LDtocL!");
MCSymbol *MOSymbol = 0;
if (MO.isJTI())
MOSymbol = lookUpOrCreateTOCEntry(GetJTISymbol(MO.getIndex()));
else {
const GlobalValue *GValue = MO.getGlobal();
const GlobalAlias *GAlias = dyn_cast<GlobalAlias>(GValue);
const GlobalValue *RealGValue = GAlias ?
GAlias->resolveAliasedGlobal(false) : GValue;
MOSymbol = Mang->getSymbol(RealGValue);
const GlobalVariable *GVar = dyn_cast<GlobalVariable>(RealGValue);
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
if (!GVar || !GVar->hasInitializer() || RealGValue->hasCommonLinkage() ||
RealGValue->hasAvailableExternallyLinkage())
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
MOSymbol = lookUpOrCreateTOCEntry(MOSymbol);
}
const MCExpr *Exp =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_TOC16_LO,
OutContext);
TmpInst.getOperand(1) = MCOperand::CreateExpr(Exp);
OutStreamer.EmitInstruction(TmpInst);
return;
}
case PPC::ADDItocL: {
// Transform %Xd = ADDItocL %Xs, <ga:@sym>
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
// Change the opcode to ADDI8L. If the global address is external, then
// generate a TOC entry and reference that. Otherwise reference the
// symbol directly.
TmpInst.setOpcode(PPC::ADDI8L);
const MachineOperand &MO = MI->getOperand(2);
assert((MO.isGlobal() || MO.isCPI()) && "Invalid operand for ADDItocL");
MCSymbol *MOSymbol = 0;
bool IsExternal = false;
bool IsFunction = false;
if (MO.isGlobal()) {
const GlobalValue *GValue = MO.getGlobal();
const GlobalAlias *GAlias = dyn_cast<GlobalAlias>(GValue);
const GlobalValue *RealGValue = GAlias ?
GAlias->resolveAliasedGlobal(false) : GValue;
MOSymbol = Mang->getSymbol(RealGValue);
const GlobalVariable *GVar = dyn_cast<GlobalVariable>(RealGValue);
This patch implements medium code model support for 64-bit PowerPC. The default for 64-bit PowerPC is small code model, in which TOC entries must be addressable using a 16-bit offset from the TOC pointer. Additionally, only TOC entries are addressed via the TOC pointer. With medium code model, TOC entries and data sections can all be addressed via the TOC pointer using a 32-bit offset. Cooperation with the linker allows 16-bit offsets to be used when these are sufficient, reducing the number of extra instructions that need to be executed. Medium code model also does not generate explicit TOC entries in ".section toc" for variables that are wholly internal to the compilation unit. Consider a load of an external 4-byte integer. With small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei With medium model, it instead generates: addis 3, 2, .LC1@toc@ha ld 3, .LC1@toc@l(3) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc ei[TC],ei Here .LC1@toc@ha is a relocation requesting the upper 16 bits of the 32-bit offset of ei's TOC entry from the TOC base pointer. Similarly, .LC1@toc@l is a relocation requesting the lower 16 bits. Note that if the linker determines that ei's TOC entry is within a 16-bit offset of the TOC base pointer, it will replace the "addis" with a "nop", and replace the "ld" with the identical "ld" instruction from the small code model example. Consider next a load of a function-scope static integer. For small code model, the compiler generates: ld 3, .LC1@toc(2) lwz 4, 0(3) .section .toc,"aw",@progbits .LC1: .tc test_fn_static.si[TC],test_fn_static.si .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 For medium code model, the compiler generates: addis 3, 2, test_fn_static.si@toc@ha addi 3, 3, test_fn_static.si@toc@l lwz 4, 0(3) .type test_fn_static.si,@object .local test_fn_static.si .comm test_fn_static.si,4,4 Again, the linker may replace the "addis" with a "nop", calculating only a 16-bit offset when this is sufficient. Note that it would be more efficient for the compiler to generate: addis 3, 2, test_fn_static.si@toc@ha lwz 4, test_fn_static.si@toc@l(3) The current patch does not perform this optimization yet. This will be addressed as a peephole optimization in a later patch. For the moment, the default code model for 64-bit PowerPC will remain the small code model. We plan to eventually change the default to medium code model, which matches current upstream GCC behavior. Note that the different code models are ABI-compatible, so code compiled with different models will be linked and execute correctly. I've tested the regression suite and the application/benchmark test suite in two ways: Once with the patch as submitted here, and once with additional logic to force medium code model as the default. The tests all compile cleanly, with one exception. The mandel-2 application test fails due to an unrelated ABI compatibility with passing complex numbers. It just so happens that small code model was incredibly lucky, in that temporary values in floating-point registers held the expected values needed by the external library routine that was called incorrectly. My current thought is to correct the ABI problems with _Complex before making medium code model the default, to avoid introducing this "regression." Here are a few comments on how the patch works, since the selection code can be difficult to follow: The existing logic for small code model defines three pseudo-instructions: LDtoc for most uses, LDtocJTI for jump table addresses, and LDtocCPT for constant pool addresses. These are expanded by SelectCodeCommon(). The pseudo-instruction approach doesn't work for medium code model, because we need to generate two instructions when we match the same pattern. Instead, new logic in PPCDAGToDAGISel::Select() intercepts the TOC_ENTRY node for medium code model, and generates an ADDIStocHA followed by either a LDtocL or an ADDItocL. These new node types correspond naturally to the sequences described above. The addis/ld sequence is generated for the following cases: * Jump table addresses * Function addresses * External global variables * Tentative definitions of global variables (common linkage) The addis/addi sequence is generated for the following cases: * Constant pool entries * File-scope static global variables * Function-scope static variables Expanding to the two-instruction sequences at select time exposes the instructions to subsequent optimization, particularly scheduling. The rest of the processing occurs at assembly time, in PPCAsmPrinter::EmitInstruction. Each of the instructions is converted to a "real" PowerPC instruction. When a TOC entry needs to be created, this is done here in the same manner as for the existing LDtoc, LDtocJTI, and LDtocCPT pseudo-instructions (I factored out a new routine to handle this). I had originally thought that if a TOC entry was needed for LDtocL or ADDItocL, it would already have been generated for the previous ADDIStocHA. However, at higher optimization levels, the ADDIStocHA may appear in a different block, which may be assembled textually following the block containing the LDtocL or ADDItocL. So it is necessary to include the possibility of creating a new TOC entry for those two instructions. Note that for LDtocL, we generate a new form of LD called LDrs. This allows specifying the @toc@l relocation for the offset field of the LD instruction (i.e., the offset is replaced by a SymbolLo relocation). When the peephole optimization described above is added, we will need to do similar things for all immediate-form load and store operations. The seven "mcm-n.ll" test cases are kept separate because otherwise the intermingling of various TOC entries and so forth makes the tests fragile and hard to understand. The above assumes use of an external assembler. For use of the integrated assembler, new relocations are added and used by PPCELFObjectWriter. Testing is done with "mcm-obj.ll", which tests for proper generation of the various relocations for the same sequences tested with the external assembler. llvm-svn: 168708
2012-11-28 01:35:46 +08:00
IsExternal = GVar && !GVar->hasInitializer();
IsFunction = !GVar;
} else if (MO.isCPI())
MOSymbol = GetCPISymbol(MO.getIndex());
if (IsFunction || IsExternal)
MOSymbol = lookUpOrCreateTOCEntry(MOSymbol);
const MCExpr *Exp =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_TOC16_LO,
OutContext);
TmpInst.getOperand(2) = MCOperand::CreateExpr(Exp);
OutStreamer.EmitInstruction(TmpInst);
return;
}
case PPC::ADDISgotTprelHA: {
// Transform: %Xd = ADDISgotTprelHA %X2, <ga:@sym>
// Into: %Xd = ADDIS8 %X2, sym@got@tlsgd@ha
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymGotTprel =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TPREL16_HA,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDIS8)
.addReg(MI->getOperand(0).getReg())
.addReg(PPC::X2)
.addExpr(SymGotTprel));
return;
}
case PPC::LDgotTprelL: {
// Transform %Xd = LDgotTprelL <ga:@sym>, %Xs
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
// Change the opcode to LDrs, which is a form of LD with the offset
// specified by a SymbolLo.
TmpInst.setOpcode(PPC::LDrs);
const MachineOperand &MO = MI->getOperand(1);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *Exp =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TPREL16_LO,
OutContext);
TmpInst.getOperand(1) = MCOperand::CreateExpr(Exp);
OutStreamer.EmitInstruction(TmpInst);
return;
}
This patch implements the general dynamic TLS model for 64-bit PowerPC. Given a thread-local symbol x with global-dynamic access, the generated code to obtain x's address is: Instruction Relocation Symbol addis ra,r2,x@got@tlsgd@ha R_PPC64_GOT_TLSGD16_HA x addi r3,ra,x@got@tlsgd@l R_PPC64_GOT_TLSGD16_L x bl __tls_get_addr(x@tlsgd) R_PPC64_TLSGD x R_PPC64_REL24 __tls_get_addr nop <use address in r3> The implementation borrows from the medium code model work for introducing special forms of ADDIS and ADDI into the DAG representation. This is made slightly more complicated by having to introduce a call to the external function __tls_get_addr. Using the full call machinery is overkill and, more importantly, makes it difficult to add a special relocation. So I've introduced another opcode GET_TLS_ADDR to represent the function call, and surrounded it with register copies to set up the parameter and return value. Most of the code is pretty straightforward. I ran into one peculiarity when I introduced a new PPC opcode BL8_NOP_ELF_TLSGD, which is just like BL8_NOP_ELF except that it takes another parameter to represent the symbol ("x" above) that requires a relocation on the call. Something in the TblGen machinery causes BL8_NOP_ELF and BL8_NOP_ELF_TLSGD to be treated identically during the emit phase, so this second operand was never visited to generate relocations. This is the reason for the slightly messy workaround in PPCMCCodeEmitter.cpp:getDirectBrEncoding(). Two new tests are included to demonstrate correct external assembly and correct generation of relocations using the integrated assembler. Comments welcome! Thanks, Bill llvm-svn: 169910
2012-12-12 04:30:11 +08:00
case PPC::ADDIStlsgdHA: {
// Transform: %Xd = ADDIStlsgdHA %X2, <ga:@sym>
// Into: %Xd = ADDIS8 %X2, sym@got@tlsgd@ha
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymGotTlsGD =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TLSGD16_HA,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDIS8)
.addReg(MI->getOperand(0).getReg())
.addReg(PPC::X2)
.addExpr(SymGotTlsGD));
return;
}
case PPC::ADDItlsgdL: {
// Transform: %Xd = ADDItlsgdL %Xs, <ga:@sym>
// Into: %Xd = ADDI8L %Xs, sym@got@tlsgd@l
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymGotTlsGD =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TLSGD16_LO,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDI8L)
.addReg(MI->getOperand(0).getReg())
.addReg(MI->getOperand(1).getReg())
.addExpr(SymGotTlsGD));
return;
}
case PPC::GETtlsADDR: {
// Transform: %X3 = GETtlsADDR %X3, <ga:@sym>
// Into: BL8_NOP_ELF_TLSGD __tls_get_addr(sym@tlsgd)
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
StringRef Name = "__tls_get_addr";
MCSymbol *TlsGetAddr = OutContext.GetOrCreateSymbol(Name);
const MCSymbolRefExpr *TlsRef =
MCSymbolRefExpr::Create(TlsGetAddr, MCSymbolRefExpr::VK_None, OutContext);
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymVar =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_TLSGD,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::BL8_NOP_ELF_TLSGD)
.addExpr(TlsRef)
.addExpr(SymVar));
return;
}
case PPC::ADDIStlsldHA: {
// Transform: %Xd = ADDIStlsldHA %X2, <ga:@sym>
// Into: %Xd = ADDIS8 %X2, sym@got@tlsld@ha
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymGotTlsLD =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TLSLD16_HA,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDIS8)
.addReg(MI->getOperand(0).getReg())
.addReg(PPC::X2)
.addExpr(SymGotTlsLD));
return;
}
case PPC::ADDItlsldL: {
// Transform: %Xd = ADDItlsldL %Xs, <ga:@sym>
// Into: %Xd = ADDI8L %Xs, sym@got@tlsld@l
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymGotTlsLD =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_GOT_TLSLD16_LO,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDI8L)
.addReg(MI->getOperand(0).getReg())
.addReg(MI->getOperand(1).getReg())
.addExpr(SymGotTlsLD));
return;
}
case PPC::GETtlsldADDR: {
// Transform: %X3 = GETtlsldADDR %X3, <ga:@sym>
// Into: BL8_NOP_ELF_TLSLD __tls_get_addr(sym@tlsld)
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
StringRef Name = "__tls_get_addr";
MCSymbol *TlsGetAddr = OutContext.GetOrCreateSymbol(Name);
const MCSymbolRefExpr *TlsRef =
MCSymbolRefExpr::Create(TlsGetAddr, MCSymbolRefExpr::VK_None, OutContext);
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymVar =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_TLSLD,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::BL8_NOP_ELF_TLSLD)
.addExpr(TlsRef)
.addExpr(SymVar));
return;
}
case PPC::ADDISdtprelHA: {
// Transform: %Xd = ADDISdtprelHA %X3, <ga:@sym>
// Into: %Xd = ADDIS8 %X3, sym@dtprel@ha
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymDtprel =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_DTPREL16_HA,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDIS8)
.addReg(MI->getOperand(0).getReg())
.addReg(PPC::X3)
.addExpr(SymDtprel));
return;
}
case PPC::ADDIdtprelL: {
// Transform: %Xd = ADDIdtprelL %Xs, <ga:@sym>
// Into: %Xd = ADDI8L %Xs, sym@dtprel@l
assert(Subtarget.isPPC64() && "Not supported for 32-bit PowerPC");
const MachineOperand &MO = MI->getOperand(2);
const GlobalValue *GValue = MO.getGlobal();
MCSymbol *MOSymbol = Mang->getSymbol(GValue);
const MCExpr *SymDtprel =
MCSymbolRefExpr::Create(MOSymbol, MCSymbolRefExpr::VK_PPC_DTPREL16_LO,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDI8L)
.addReg(MI->getOperand(0).getReg())
.addReg(MI->getOperand(1).getReg())
.addExpr(SymDtprel));
return;
}
case PPC::MFCRpseud:
case PPC::MFCR8pseud:
// Transform: %R3 = MFCRpseud %CR7
// Into: %R3 = MFCR ;; cr7
OutStreamer.AddComment(PPCInstPrinter::
getRegisterName(MI->getOperand(1).getReg()));
OutStreamer.EmitInstruction(MCInstBuilder(Subtarget.isPPC64() ? PPC::MFCR8 : PPC::MFCR)
.addReg(MI->getOperand(0).getReg()));
return;
case PPC::SYNC:
// In Book E sync is called msync, handle this special case here...
if (Subtarget.isBookE()) {
OutStreamer.EmitRawText(StringRef("\tmsync"));
return;
}
}
LowerPPCMachineInstrToMCInst(MI, TmpInst, *this, Subtarget.isDarwin());
OutStreamer.EmitInstruction(TmpInst);
}
void PPCLinuxAsmPrinter::EmitFunctionEntryLabel() {
if (!Subtarget.isPPC64()) // linux/ppc32 - Normal entry label.
return AsmPrinter::EmitFunctionEntryLabel();
// Emit an official procedure descriptor.
const MCSection *Current = OutStreamer.getCurrentSection();
const MCSectionELF *Section = OutStreamer.getContext().getELFSection(".opd",
ELF::SHT_PROGBITS, ELF::SHF_WRITE | ELF::SHF_ALLOC,
SectionKind::getReadOnly());
OutStreamer.SwitchSection(Section);
OutStreamer.EmitLabel(CurrentFnSym);
OutStreamer.EmitValueToAlignment(8);
MCSymbol *Symbol1 =
OutContext.GetOrCreateSymbol(".L." + Twine(CurrentFnSym->getName()));
// Generates a R_PPC64_ADDR64 (from FK_DATA_8) relocation for the function
// entry point.
OutStreamer.EmitValue(MCSymbolRefExpr::Create(Symbol1, OutContext),
8 /*size*/);
MCSymbol *Symbol2 = OutContext.GetOrCreateSymbol(StringRef(".TOC."));
// Generates a R_PPC64_TOC relocation for TOC base insertion.
OutStreamer.EmitValue(MCSymbolRefExpr::Create(Symbol2,
MCSymbolRefExpr::VK_PPC_TOC, OutContext),
8/*size*/);
// Emit a null environment pointer.
OutStreamer.EmitIntValue(0, 8 /* size */);
OutStreamer.SwitchSection(Current);
MCSymbol *RealFnSym = OutContext.GetOrCreateSymbol(
".L." + Twine(CurrentFnSym->getName()));
OutStreamer.EmitLabel(RealFnSym);
CurrentFnSymForSize = RealFnSym;
}
bool PPCLinuxAsmPrinter::doFinalization(Module &M) {
const DataLayout *TD = TM.getDataLayout();
Revert the majority of the next patch in the address space series: r165941: Resubmit the changes to llvm core to update the functions to support different pointer sizes on a per address space basis. Despite this commit log, this change primarily changed stuff outside of VMCore, and those changes do not carry any tests for correctness (or even plausibility), and we have consistently found questionable or flat out incorrect cases in these changes. Most of them are probably correct, but we need to devise a system that makes it more clear when we have handled the address space concerns correctly, and ideally each pass that gets updated would receive an accompanying test case that exercises that pass specificaly w.r.t. alternate address spaces. However, from this commit, I have retained the new C API entry points. Those were an orthogonal change that probably should have been split apart, but they seem entirely good. In several places the changes were very obvious cleanups with no actual multiple address space code added; these I have not reverted when I spotted them. In a few other places there were merge conflicts due to a cleaner solution being implemented later, often not using address spaces at all. In those cases, I've preserved the new code which isn't address space dependent. This is part of my ongoing effort to clean out the partial address space code which carries high risk and low test coverage, and not likely to be finished before the 3.2 release looms closer. Duncan and I would both like to see the above issues addressed before we return to these changes. llvm-svn: 167222
2012-11-01 17:14:31 +08:00
bool isPPC64 = TD->getPointerSizeInBits() == 64;
if (isPPC64 && !TOC.empty()) {
const MCSectionELF *Section = OutStreamer.getContext().getELFSection(".toc",
ELF::SHT_PROGBITS, ELF::SHF_WRITE | ELF::SHF_ALLOC,
SectionKind::getReadOnly());
OutStreamer.SwitchSection(Section);
for (MapVector<MCSymbol*, MCSymbol*>::iterator I = TOC.begin(),
E = TOC.end(); I != E; ++I) {
OutStreamer.EmitLabel(I->second);
MCSymbol *S = OutContext.GetOrCreateSymbol(I->first->getName());
OutStreamer.EmitTCEntry(*S);
}
}
MachineModuleInfoELF &MMIELF =
MMI->getObjFileInfo<MachineModuleInfoELF>();
MachineModuleInfoELF::SymbolListTy Stubs = MMIELF.GetGVStubList();
if (!Stubs.empty()) {
OutStreamer.SwitchSection(getObjFileLowering().getDataSection());
for (unsigned i = 0, e = Stubs.size(); i != e; ++i) {
// L_foo$stub:
OutStreamer.EmitLabel(Stubs[i].first);
// .long _foo
OutStreamer.EmitValue(MCSymbolRefExpr::Create(Stubs[i].second.getPointer(),
OutContext),
isPPC64 ? 8 : 4/*size*/, 0/*addrspace*/);
}
Stubs.clear();
OutStreamer.AddBlankLine();
}
return AsmPrinter::doFinalization(M);
}
/// EmitFunctionBodyEnd - Print the traceback table before the .size
/// directive.
///
void PPCLinuxAsmPrinter::EmitFunctionBodyEnd() {
// Only the 64-bit target requires a traceback table. For now,
// we only emit the word of zeroes that GDB requires to find
// the end of the function, and zeroes for the eight-byte
// mandatory fields.
// FIXME: We should fill in the eight-byte mandatory fields as described in
// the PPC64 ELF ABI (this is a low-priority item because GDB does not
// currently make use of these fields).
if (Subtarget.isPPC64()) {
OutStreamer.EmitIntValue(0, 4/*size*/);
OutStreamer.EmitIntValue(0, 8/*size*/);
}
}
void PPCDarwinAsmPrinter::EmitStartOfAsmFile(Module &M) {
static const char *const CPUDirectives[] = {
"",
"ppc",
"ppc440",
"ppc601",
"ppc602",
"ppc603",
"ppc7400",
"ppc750",
"ppc970",
"ppcA2",
"ppce500mc",
"ppce5500",
"power3",
"power4",
"power5",
"power5x",
"power6",
"power6x",
"power7",
"ppc64"
};
unsigned Directive = Subtarget.getDarwinDirective();
if (Subtarget.hasMFOCRF() && Directive < PPC::DIR_970)
Directive = PPC::DIR_970;
if (Subtarget.hasAltivec() && Directive < PPC::DIR_7400)
Directive = PPC::DIR_7400;
if (Subtarget.isPPC64() && Directive < PPC::DIR_64)
Directive = PPC::DIR_64;
assert(Directive <= PPC::DIR_64 && "Directive out of range.");
// FIXME: This is a total hack, finish mc'izing the PPC backend.
if (OutStreamer.hasRawTextSupport())
OutStreamer.EmitRawText("\t.machine " + Twine(CPUDirectives[Directive]));
2008-08-09 02:22:59 +08:00
// Prime text sections so they are adjacent. This reduces the likelihood a
// large data or debug section causes a branch to exceed 16M limit.
const TargetLoweringObjectFileMachO &TLOFMacho =
static_cast<const TargetLoweringObjectFileMachO &>(getObjFileLowering());
OutStreamer.SwitchSection(TLOFMacho.getTextCoalSection());
if (TM.getRelocationModel() == Reloc::PIC_) {
OutStreamer.SwitchSection(
OutContext.getMachOSection("__TEXT", "__picsymbolstub1",
MCSectionMachO::S_SYMBOL_STUBS |
MCSectionMachO::S_ATTR_PURE_INSTRUCTIONS,
32, SectionKind::getText()));
} else if (TM.getRelocationModel() == Reloc::DynamicNoPIC) {
OutStreamer.SwitchSection(
OutContext.getMachOSection("__TEXT","__symbol_stub1",
MCSectionMachO::S_SYMBOL_STUBS |
MCSectionMachO::S_ATTR_PURE_INSTRUCTIONS,
16, SectionKind::getText()));
}
OutStreamer.SwitchSection(getObjFileLowering().getTextSection());
}
static MCSymbol *GetLazyPtr(MCSymbol *Sym, MCContext &Ctx) {
// Remove $stub suffix, add $lazy_ptr.
StringRef NoStub = Sym->getName().substr(0, Sym->getName().size()-5);
return Ctx.GetOrCreateSymbol(NoStub + "$lazy_ptr");
}
static MCSymbol *GetAnonSym(MCSymbol *Sym, MCContext &Ctx) {
// Add $tmp suffix to $stub, yielding $stub$tmp.
return Ctx.GetOrCreateSymbol(Sym->getName() + "$tmp");
}
void PPCDarwinAsmPrinter::
EmitFunctionStubs(const MachineModuleInfoMachO::SymbolListTy &Stubs) {
Revert the majority of the next patch in the address space series: r165941: Resubmit the changes to llvm core to update the functions to support different pointer sizes on a per address space basis. Despite this commit log, this change primarily changed stuff outside of VMCore, and those changes do not carry any tests for correctness (or even plausibility), and we have consistently found questionable or flat out incorrect cases in these changes. Most of them are probably correct, but we need to devise a system that makes it more clear when we have handled the address space concerns correctly, and ideally each pass that gets updated would receive an accompanying test case that exercises that pass specificaly w.r.t. alternate address spaces. However, from this commit, I have retained the new C API entry points. Those were an orthogonal change that probably should have been split apart, but they seem entirely good. In several places the changes were very obvious cleanups with no actual multiple address space code added; these I have not reverted when I spotted them. In a few other places there were merge conflicts due to a cleaner solution being implemented later, often not using address spaces at all. In those cases, I've preserved the new code which isn't address space dependent. This is part of my ongoing effort to clean out the partial address space code which carries high risk and low test coverage, and not likely to be finished before the 3.2 release looms closer. Duncan and I would both like to see the above issues addressed before we return to these changes. llvm-svn: 167222
2012-11-01 17:14:31 +08:00
bool isPPC64 = TM.getDataLayout()->getPointerSizeInBits() == 64;
const TargetLoweringObjectFileMachO &TLOFMacho =
static_cast<const TargetLoweringObjectFileMachO &>(getObjFileLowering());
// .lazy_symbol_pointer
const MCSection *LSPSection = TLOFMacho.getLazySymbolPointerSection();
// Output stubs for dynamically-linked functions
if (TM.getRelocationModel() == Reloc::PIC_) {
const MCSection *StubSection =
OutContext.getMachOSection("__TEXT", "__picsymbolstub1",
MCSectionMachO::S_SYMBOL_STUBS |
MCSectionMachO::S_ATTR_PURE_INSTRUCTIONS,
32, SectionKind::getText());
for (unsigned i = 0, e = Stubs.size(); i != e; ++i) {
OutStreamer.SwitchSection(StubSection);
EmitAlignment(4);
MCSymbol *Stub = Stubs[i].first;
MCSymbol *RawSym = Stubs[i].second.getPointer();
MCSymbol *LazyPtr = GetLazyPtr(Stub, OutContext);
MCSymbol *AnonSymbol = GetAnonSym(Stub, OutContext);
OutStreamer.EmitLabel(Stub);
OutStreamer.EmitSymbolAttribute(RawSym, MCSA_IndirectSymbol);
// mflr r0
OutStreamer.EmitInstruction(MCInstBuilder(PPC::MFLR).addReg(PPC::R0));
// FIXME: MCize this.
OutStreamer.EmitRawText("\tbcl 20, 31, " + Twine(AnonSymbol->getName()));
OutStreamer.EmitLabel(AnonSymbol);
// mflr r11
OutStreamer.EmitInstruction(MCInstBuilder(PPC::MFLR).addReg(PPC::R11));
// addis r11, r11, ha16(LazyPtr - AnonSymbol)
const MCExpr *Sub =
MCBinaryExpr::CreateSub(MCSymbolRefExpr::Create(LazyPtr, OutContext),
MCSymbolRefExpr::Create(AnonSymbol, OutContext),
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::ADDIS)
.addReg(PPC::R11)
.addReg(PPC::R11)
.addExpr(Sub));
// mtlr r0
OutStreamer.EmitInstruction(MCInstBuilder(PPC::MTLR).addReg(PPC::R0));
// ldu r12, lo16(LazyPtr - AnonSymbol)(r11)
// lwzu r12, lo16(LazyPtr - AnonSymbol)(r11)
OutStreamer.EmitInstruction(MCInstBuilder(isPPC64 ? PPC::LDU : PPC::LWZU)
.addReg(PPC::R12)
.addExpr(Sub).addExpr(Sub)
.addReg(PPC::R11));
// mtctr r12
OutStreamer.EmitInstruction(MCInstBuilder(PPC::MTCTR).addReg(PPC::R12));
// bctr
OutStreamer.EmitInstruction(MCInstBuilder(PPC::BCTR));
OutStreamer.SwitchSection(LSPSection);
OutStreamer.EmitLabel(LazyPtr);
OutStreamer.EmitSymbolAttribute(RawSym, MCSA_IndirectSymbol);
MCSymbol *DyldStubBindingHelper =
OutContext.GetOrCreateSymbol(StringRef("dyld_stub_binding_helper"));
if (isPPC64) {
// .quad dyld_stub_binding_helper
OutStreamer.EmitSymbolValue(DyldStubBindingHelper, 8);
} else {
// .long dyld_stub_binding_helper
OutStreamer.EmitSymbolValue(DyldStubBindingHelper, 4);
}
}
OutStreamer.AddBlankLine();
return;
}
const MCSection *StubSection =
OutContext.getMachOSection("__TEXT","__symbol_stub1",
MCSectionMachO::S_SYMBOL_STUBS |
MCSectionMachO::S_ATTR_PURE_INSTRUCTIONS,
16, SectionKind::getText());
for (unsigned i = 0, e = Stubs.size(); i != e; ++i) {
MCSymbol *Stub = Stubs[i].first;
MCSymbol *RawSym = Stubs[i].second.getPointer();
MCSymbol *LazyPtr = GetLazyPtr(Stub, OutContext);
OutStreamer.SwitchSection(StubSection);
EmitAlignment(4);
OutStreamer.EmitLabel(Stub);
OutStreamer.EmitSymbolAttribute(RawSym, MCSA_IndirectSymbol);
// lis r11, ha16(LazyPtr)
const MCExpr *LazyPtrHa16 =
MCSymbolRefExpr::Create(LazyPtr, MCSymbolRefExpr::VK_PPC_DARWIN_HA16,
OutContext);
OutStreamer.EmitInstruction(MCInstBuilder(PPC::LIS)
.addReg(PPC::R11)
.addExpr(LazyPtrHa16));
const MCExpr *LazyPtrLo16 =
MCSymbolRefExpr::Create(LazyPtr, MCSymbolRefExpr::VK_PPC_DARWIN_LO16,
OutContext);
// ldu r12, lo16(LazyPtr)(r11)
// lwzu r12, lo16(LazyPtr)(r11)
OutStreamer.EmitInstruction(MCInstBuilder(isPPC64 ? PPC::LDU : PPC::LWZU)
.addReg(PPC::R12)
.addExpr(LazyPtrLo16).addExpr(LazyPtrLo16)
.addReg(PPC::R11));
// mtctr r12
OutStreamer.EmitInstruction(MCInstBuilder(PPC::MTCTR).addReg(PPC::R12));
// bctr
OutStreamer.EmitInstruction(MCInstBuilder(PPC::BCTR));
OutStreamer.SwitchSection(LSPSection);
OutStreamer.EmitLabel(LazyPtr);
OutStreamer.EmitSymbolAttribute(RawSym, MCSA_IndirectSymbol);
MCSymbol *DyldStubBindingHelper =
OutContext.GetOrCreateSymbol(StringRef("dyld_stub_binding_helper"));
if (isPPC64) {
// .quad dyld_stub_binding_helper
OutStreamer.EmitSymbolValue(DyldStubBindingHelper, 8);
} else {
// .long dyld_stub_binding_helper
OutStreamer.EmitSymbolValue(DyldStubBindingHelper, 4);
}
}
OutStreamer.AddBlankLine();
}
bool PPCDarwinAsmPrinter::doFinalization(Module &M) {
Revert the majority of the next patch in the address space series: r165941: Resubmit the changes to llvm core to update the functions to support different pointer sizes on a per address space basis. Despite this commit log, this change primarily changed stuff outside of VMCore, and those changes do not carry any tests for correctness (or even plausibility), and we have consistently found questionable or flat out incorrect cases in these changes. Most of them are probably correct, but we need to devise a system that makes it more clear when we have handled the address space concerns correctly, and ideally each pass that gets updated would receive an accompanying test case that exercises that pass specificaly w.r.t. alternate address spaces. However, from this commit, I have retained the new C API entry points. Those were an orthogonal change that probably should have been split apart, but they seem entirely good. In several places the changes were very obvious cleanups with no actual multiple address space code added; these I have not reverted when I spotted them. In a few other places there were merge conflicts due to a cleaner solution being implemented later, often not using address spaces at all. In those cases, I've preserved the new code which isn't address space dependent. This is part of my ongoing effort to clean out the partial address space code which carries high risk and low test coverage, and not likely to be finished before the 3.2 release looms closer. Duncan and I would both like to see the above issues addressed before we return to these changes. llvm-svn: 167222
2012-11-01 17:14:31 +08:00
bool isPPC64 = TM.getDataLayout()->getPointerSizeInBits() == 64;
// Darwin/PPC always uses mach-o.
const TargetLoweringObjectFileMachO &TLOFMacho =
static_cast<const TargetLoweringObjectFileMachO &>(getObjFileLowering());
MachineModuleInfoMachO &MMIMacho =
MMI->getObjFileInfo<MachineModuleInfoMachO>();
MachineModuleInfoMachO::SymbolListTy Stubs = MMIMacho.GetFnStubList();
if (!Stubs.empty())
EmitFunctionStubs(Stubs);
if (MAI->doesSupportExceptionHandling() && MMI) {
// Add the (possibly multiple) personalities to the set of global values.
// Only referenced functions get into the Personalities list.
const std::vector<const Function*> &Personalities = MMI->getPersonalities();
for (std::vector<const Function*>::const_iterator I = Personalities.begin(),
E = Personalities.end(); I != E; ++I) {
if (*I) {
MCSymbol *NLPSym = GetSymbolWithGlobalValueBase(*I, "$non_lazy_ptr");
MachineModuleInfoImpl::StubValueTy &StubSym =
MMIMacho.getGVStubEntry(NLPSym);
StubSym = MachineModuleInfoImpl::StubValueTy(Mang->getSymbol(*I), true);
}
}
}
// Output stubs for dynamically-linked functions.
Stubs = MMIMacho.GetGVStubList();
// Output macho stubs for external and common global variables.
if (!Stubs.empty()) {
// Switch with ".non_lazy_symbol_pointer" directive.
OutStreamer.SwitchSection(TLOFMacho.getNonLazySymbolPointerSection());
EmitAlignment(isPPC64 ? 3 : 2);
for (unsigned i = 0, e = Stubs.size(); i != e; ++i) {
// L_foo$stub:
OutStreamer.EmitLabel(Stubs[i].first);
// .indirect_symbol _foo
MachineModuleInfoImpl::StubValueTy &MCSym = Stubs[i].second;
OutStreamer.EmitSymbolAttribute(MCSym.getPointer(), MCSA_IndirectSymbol);
if (MCSym.getInt())
// External to current translation unit.
OutStreamer.EmitIntValue(0, isPPC64 ? 8 : 4/*size*/);
else
// Internal to current translation unit.
//
// When we place the LSDA into the TEXT section, the type info pointers
// need to be indirect and pc-rel. We accomplish this by using NLPs.
// However, sometimes the types are local to the file. So we need to
// fill in the value for the NLP in those cases.
OutStreamer.EmitValue(MCSymbolRefExpr::Create(MCSym.getPointer(),
OutContext),
isPPC64 ? 8 : 4/*size*/);
}
Stubs.clear();
OutStreamer.AddBlankLine();
}
Stubs = MMIMacho.GetHiddenGVStubList();
if (!Stubs.empty()) {
OutStreamer.SwitchSection(getObjFileLowering().getDataSection());
EmitAlignment(isPPC64 ? 3 : 2);
for (unsigned i = 0, e = Stubs.size(); i != e; ++i) {
// L_foo$stub:
OutStreamer.EmitLabel(Stubs[i].first);
// .long _foo
OutStreamer.EmitValue(MCSymbolRefExpr::
Create(Stubs[i].second.getPointer(),
OutContext),
isPPC64 ? 8 : 4/*size*/);
}
Stubs.clear();
OutStreamer.AddBlankLine();
}
// Funny Darwin hack: This flag tells the linker that no global symbols
// contain code that falls through to other global symbols (e.g. the obvious
// implementation of multiple entry points). If this doesn't occur, the
// linker can safely perform dead code stripping. Since LLVM never generates
// code that does this, it is always safe to set.
OutStreamer.EmitAssemblerFlag(MCAF_SubsectionsViaSymbols);
return AsmPrinter::doFinalization(M);
2004-06-22 00:55:25 +08:00
}
/// createPPCAsmPrinterPass - Returns a pass that prints the PPC assembly code
/// for a MachineFunction to the given output stream, in a format that the
/// Darwin assembler can deal with.
///
static AsmPrinter *createPPCAsmPrinterPass(TargetMachine &tm,
MCStreamer &Streamer) {
const PPCSubtarget *Subtarget = &tm.getSubtarget<PPCSubtarget>();
if (Subtarget->isDarwin())
return new PPCDarwinAsmPrinter(tm, Streamer);
return new PPCLinuxAsmPrinter(tm, Streamer);
}
// Force static initialization.
Reapply TargetRegistry refactoring commits. --- Reverse-merging r75799 into '.': U test/Analysis/PointerTracking U include/llvm/Target/TargetMachineRegistry.h U include/llvm/Target/TargetMachine.h U include/llvm/Target/TargetRegistry.h U include/llvm/Target/TargetSelect.h U tools/lto/LTOCodeGenerator.cpp U tools/lto/LTOModule.cpp U tools/llc/llc.cpp U lib/Target/PowerPC/PPCTargetMachine.h U lib/Target/PowerPC/AsmPrinter/PPCAsmPrinter.cpp U lib/Target/PowerPC/PPCTargetMachine.cpp U lib/Target/PowerPC/PPC.h U lib/Target/ARM/ARMTargetMachine.cpp U lib/Target/ARM/AsmPrinter/ARMAsmPrinter.cpp U lib/Target/ARM/ARMTargetMachine.h U lib/Target/ARM/ARM.h U lib/Target/XCore/XCoreTargetMachine.cpp U lib/Target/XCore/XCoreTargetMachine.h U lib/Target/PIC16/PIC16TargetMachine.cpp U lib/Target/PIC16/PIC16TargetMachine.h U lib/Target/Alpha/AsmPrinter/AlphaAsmPrinter.cpp U lib/Target/Alpha/AlphaTargetMachine.cpp U lib/Target/Alpha/AlphaTargetMachine.h U lib/Target/X86/X86TargetMachine.h U lib/Target/X86/X86.h U lib/Target/X86/AsmPrinter/X86ATTAsmPrinter.h U lib/Target/X86/AsmPrinter/X86AsmPrinter.cpp U lib/Target/X86/AsmPrinter/X86IntelAsmPrinter.h U lib/Target/X86/X86TargetMachine.cpp U lib/Target/MSP430/MSP430TargetMachine.cpp U lib/Target/MSP430/MSP430TargetMachine.h U lib/Target/CppBackend/CPPTargetMachine.h U lib/Target/CppBackend/CPPBackend.cpp U lib/Target/CBackend/CTargetMachine.h U lib/Target/CBackend/CBackend.cpp U lib/Target/TargetMachine.cpp U lib/Target/IA64/IA64TargetMachine.cpp U lib/Target/IA64/AsmPrinter/IA64AsmPrinter.cpp U lib/Target/IA64/IA64TargetMachine.h U lib/Target/IA64/IA64.h U lib/Target/MSIL/MSILWriter.cpp U lib/Target/CellSPU/SPUTargetMachine.h U lib/Target/CellSPU/SPU.h U lib/Target/CellSPU/AsmPrinter/SPUAsmPrinter.cpp U lib/Target/CellSPU/SPUTargetMachine.cpp U lib/Target/Mips/AsmPrinter/MipsAsmPrinter.cpp U lib/Target/Mips/MipsTargetMachine.cpp U lib/Target/Mips/MipsTargetMachine.h U lib/Target/Mips/Mips.h U lib/Target/Sparc/AsmPrinter/SparcAsmPrinter.cpp U lib/Target/Sparc/SparcTargetMachine.cpp U lib/Target/Sparc/SparcTargetMachine.h U lib/ExecutionEngine/JIT/TargetSelect.cpp U lib/Support/TargetRegistry.cpp llvm-svn: 75820
2009-07-16 04:24:03 +08:00
extern "C" void LLVMInitializePowerPCAsmPrinter() {
TargetRegistry::RegisterAsmPrinter(ThePPC32Target, createPPCAsmPrinterPass);
TargetRegistry::RegisterAsmPrinter(ThePPC64Target, createPPCAsmPrinterPass);
}