llvm-6502/lib/MC/MCExpr.cpp

662 lines
23 KiB
C++
Raw Normal View History

//===- MCExpr.cpp - Assembly Level Expression Implementation --------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
#define DEBUG_TYPE "mcexpr"
#include "llvm/MC/MCExpr.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/MC/MCAsmLayout.h"
#include "llvm/MC/MCAssembler.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/MCObjectWriter.h"
#include "llvm/MC/MCSymbol.h"
#include "llvm/MC/MCValue.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/raw_ostream.h"
using namespace llvm;
namespace {
namespace stats {
STATISTIC(MCExprEvaluate, "Number of MCExpr evaluations");
}
}
void MCExpr::print(raw_ostream &OS) const {
switch (getKind()) {
case MCExpr::Target:
return cast<MCTargetExpr>(this)->PrintImpl(OS);
case MCExpr::Constant:
OS << cast<MCConstantExpr>(*this).getValue();
return;
case MCExpr::SymbolRef: {
const MCSymbolRefExpr &SRE = cast<MCSymbolRefExpr>(*this);
const MCSymbol &Sym = SRE.getSymbol();
// Parenthesize names that start with $ so that they don't look like
// absolute names.
bool UseParens = Sym.getName()[0] == '$';
if (SRE.getKind() == MCSymbolRefExpr::VK_PPC_DARWIN_HA16 ||
SRE.getKind() == MCSymbolRefExpr::VK_PPC_DARWIN_LO16) {
OS << MCSymbolRefExpr::getVariantKindName(SRE.getKind());
UseParens = true;
}
if (UseParens)
OS << '(' << Sym << ')';
else
OS << Sym;
if (SRE.getKind() == MCSymbolRefExpr::VK_ARM_NONE ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_PLT ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_TLSGD ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_GOT ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_GOTOFF ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_TPOFF ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_GOTTPOFF ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_TARGET1 ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_TARGET2 ||
SRE.getKind() == MCSymbolRefExpr::VK_ARM_PREL31)
OS << MCSymbolRefExpr::getVariantKindName(SRE.getKind());
else if (SRE.getKind() != MCSymbolRefExpr::VK_None &&
SRE.getKind() != MCSymbolRefExpr::VK_PPC_DARWIN_HA16 &&
SRE.getKind() != MCSymbolRefExpr::VK_PPC_DARWIN_LO16)
OS << '@' << MCSymbolRefExpr::getVariantKindName(SRE.getKind());
return;
}
case MCExpr::Unary: {
const MCUnaryExpr &UE = cast<MCUnaryExpr>(*this);
switch (UE.getOpcode()) {
case MCUnaryExpr::LNot: OS << '!'; break;
case MCUnaryExpr::Minus: OS << '-'; break;
case MCUnaryExpr::Not: OS << '~'; break;
case MCUnaryExpr::Plus: OS << '+'; break;
}
OS << *UE.getSubExpr();
return;
}
case MCExpr::Binary: {
const MCBinaryExpr &BE = cast<MCBinaryExpr>(*this);
// Only print parens around the LHS if it is non-trivial.
if (isa<MCConstantExpr>(BE.getLHS()) || isa<MCSymbolRefExpr>(BE.getLHS())) {
OS << *BE.getLHS();
} else {
OS << '(' << *BE.getLHS() << ')';
}
switch (BE.getOpcode()) {
case MCBinaryExpr::Add:
// Print "X-42" instead of "X+-42".
if (const MCConstantExpr *RHSC = dyn_cast<MCConstantExpr>(BE.getRHS())) {
if (RHSC->getValue() < 0) {
OS << RHSC->getValue();
return;
}
}
OS << '+';
break;
case MCBinaryExpr::And: OS << '&'; break;
case MCBinaryExpr::Div: OS << '/'; break;
case MCBinaryExpr::EQ: OS << "=="; break;
case MCBinaryExpr::GT: OS << '>'; break;
case MCBinaryExpr::GTE: OS << ">="; break;
case MCBinaryExpr::LAnd: OS << "&&"; break;
case MCBinaryExpr::LOr: OS << "||"; break;
case MCBinaryExpr::LT: OS << '<'; break;
case MCBinaryExpr::LTE: OS << "<="; break;
case MCBinaryExpr::Mod: OS << '%'; break;
case MCBinaryExpr::Mul: OS << '*'; break;
case MCBinaryExpr::NE: OS << "!="; break;
case MCBinaryExpr::Or: OS << '|'; break;
case MCBinaryExpr::Shl: OS << "<<"; break;
case MCBinaryExpr::Shr: OS << ">>"; break;
case MCBinaryExpr::Sub: OS << '-'; break;
case MCBinaryExpr::Xor: OS << '^'; break;
}
// Only print parens around the LHS if it is non-trivial.
if (isa<MCConstantExpr>(BE.getRHS()) || isa<MCSymbolRefExpr>(BE.getRHS())) {
OS << *BE.getRHS();
} else {
OS << '(' << *BE.getRHS() << ')';
}
return;
}
}
llvm_unreachable("Invalid expression kind!");
}
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
void MCExpr::dump() const {
print(dbgs());
dbgs() << '\n';
}
#endif
/* *** */
const MCBinaryExpr *MCBinaryExpr::Create(Opcode Opc, const MCExpr *LHS,
const MCExpr *RHS, MCContext &Ctx) {
return new (Ctx) MCBinaryExpr(Opc, LHS, RHS);
}
const MCUnaryExpr *MCUnaryExpr::Create(Opcode Opc, const MCExpr *Expr,
MCContext &Ctx) {
return new (Ctx) MCUnaryExpr(Opc, Expr);
}
const MCConstantExpr *MCConstantExpr::Create(int64_t Value, MCContext &Ctx) {
return new (Ctx) MCConstantExpr(Value);
}
/* *** */
const MCSymbolRefExpr *MCSymbolRefExpr::Create(const MCSymbol *Sym,
VariantKind Kind,
MCContext &Ctx) {
return new (Ctx) MCSymbolRefExpr(Sym, Kind);
}
const MCSymbolRefExpr *MCSymbolRefExpr::Create(StringRef Name, VariantKind Kind,
MCContext &Ctx) {
return Create(Ctx.GetOrCreateSymbol(Name), Kind, Ctx);
}
StringRef MCSymbolRefExpr::getVariantKindName(VariantKind Kind) {
switch (Kind) {
case VK_Invalid: return "<<invalid>>";
case VK_None: return "<<none>>";
case VK_GOT: return "GOT";
case VK_GOTOFF: return "GOTOFF";
case VK_GOTPCREL: return "GOTPCREL";
case VK_GOTTPOFF: return "GOTTPOFF";
case VK_INDNTPOFF: return "INDNTPOFF";
case VK_NTPOFF: return "NTPOFF";
case VK_GOTNTPOFF: return "GOTNTPOFF";
case VK_PLT: return "PLT";
case VK_TLSGD: return "TLSGD";
case VK_TLSLD: return "TLSLD";
case VK_TLSLDM: return "TLSLDM";
case VK_TPOFF: return "TPOFF";
case VK_DTPOFF: return "DTPOFF";
case VK_TLVP: return "TLVP";
case VK_SECREL: return "SECREL32";
case VK_ARM_NONE: return "(NONE)";
case VK_ARM_PLT: return "(PLT)";
case VK_ARM_GOT: return "(GOT)";
case VK_ARM_GOTOFF: return "(GOTOFF)";
case VK_ARM_TPOFF: return "(tpoff)";
case VK_ARM_GOTTPOFF: return "(gottpoff)";
case VK_ARM_TLSGD: return "(tlsgd)";
case VK_ARM_TARGET1: return "(target1)";
case VK_ARM_TARGET2: return "(target2)";
case VK_ARM_PREL31: return "(prel31)";
case VK_PPC_TOC: return "tocbase";
case VK_PPC_TOC_ENTRY: return "toc";
case VK_PPC_DARWIN_HA16: return "ha16";
case VK_PPC_DARWIN_LO16: return "lo16";
case VK_PPC_GAS_HA16: return "ha";
case VK_PPC_GAS_LO16: return "l";
case VK_PPC_TPREL16_HA: return "tprel@ha";
case VK_PPC_TPREL16_LO: return "tprel@l";
case VK_PPC_DTPREL16_HA: return "dtprel@ha";
case VK_PPC_DTPREL16_LO: return "dtprel@l";
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. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@168708 91177308-0d34-0410-b5e6-96231b3b80d8
2012-11-27 17:35:46 +00:00
case VK_PPC_TOC16_HA: return "toc@ha";
case VK_PPC_TOC16_LO: return "toc@l";
case VK_PPC_GOT_TPREL16_HA: return "got@tprel@ha";
case VK_PPC_GOT_TPREL16_LO: return "got@tprel@l";
case VK_PPC_TLS: return "tls";
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 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@169910 91177308-0d34-0410-b5e6-96231b3b80d8
2012-12-11 20:30:11 +00:00
case VK_PPC_GOT_TLSGD16_HA: return "got@tlsgd@ha";
case VK_PPC_GOT_TLSGD16_LO: return "got@tlsgd@l";
case VK_PPC_GOT_TLSLD16_HA: return "got@tlsld@ha";
case VK_PPC_GOT_TLSLD16_LO: return "got@tlsld@l";
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 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@169910 91177308-0d34-0410-b5e6-96231b3b80d8
2012-12-11 20:30:11 +00:00
case VK_PPC_TLSGD: return "tlsgd";
case VK_PPC_TLSLD: return "tlsld";
case VK_Mips_GPREL: return "GPREL";
case VK_Mips_GOT_CALL: return "GOT_CALL";
case VK_Mips_GOT16: return "GOT16";
case VK_Mips_GOT: return "GOT";
case VK_Mips_ABS_HI: return "ABS_HI";
case VK_Mips_ABS_LO: return "ABS_LO";
case VK_Mips_TLSGD: return "TLSGD";
case VK_Mips_TLSLDM: return "TLSLDM";
case VK_Mips_DTPREL_HI: return "DTPREL_HI";
case VK_Mips_DTPREL_LO: return "DTPREL_LO";
case VK_Mips_GOTTPREL: return "GOTTPREL";
case VK_Mips_TPREL_HI: return "TPREL_HI";
case VK_Mips_TPREL_LO: return "TPREL_LO";
case VK_Mips_GPOFF_HI: return "GPOFF_HI";
case VK_Mips_GPOFF_LO: return "GPOFF_LO";
case VK_Mips_GOT_DISP: return "GOT_DISP";
case VK_Mips_GOT_PAGE: return "GOT_PAGE";
case VK_Mips_GOT_OFST: return "GOT_OFST";
case VK_Mips_HIGHER: return "HIGHER";
case VK_Mips_HIGHEST: return "HIGHEST";
case VK_Mips_GOT_HI16: return "GOT_HI16";
case VK_Mips_GOT_LO16: return "GOT_LO16";
case VK_Mips_CALL_HI16: return "CALL_HI16";
case VK_Mips_CALL_LO16: return "CALL_LO16";
case VK_COFF_IMGREL32: return "IMGREL32";
}
llvm_unreachable("Invalid variant kind");
}
MCSymbolRefExpr::VariantKind
MCSymbolRefExpr::getVariantKindForName(StringRef Name) {
return StringSwitch<VariantKind>(Name)
.Case("GOT", VK_GOT)
.Case("got", VK_GOT)
.Case("GOTOFF", VK_GOTOFF)
.Case("gotoff", VK_GOTOFF)
.Case("GOTPCREL", VK_GOTPCREL)
.Case("gotpcrel", VK_GOTPCREL)
.Case("GOTTPOFF", VK_GOTTPOFF)
.Case("gottpoff", VK_GOTTPOFF)
.Case("INDNTPOFF", VK_INDNTPOFF)
.Case("indntpoff", VK_INDNTPOFF)
.Case("NTPOFF", VK_NTPOFF)
.Case("ntpoff", VK_NTPOFF)
.Case("GOTNTPOFF", VK_GOTNTPOFF)
.Case("gotntpoff", VK_GOTNTPOFF)
.Case("PLT", VK_PLT)
.Case("plt", VK_PLT)
.Case("TLSGD", VK_TLSGD)
.Case("tlsgd", VK_TLSGD)
.Case("TLSLD", VK_TLSLD)
.Case("tlsld", VK_TLSLD)
.Case("TLSLDM", VK_TLSLDM)
.Case("tlsldm", VK_TLSLDM)
.Case("TPOFF", VK_TPOFF)
.Case("tpoff", VK_TPOFF)
.Case("DTPOFF", VK_DTPOFF)
.Case("dtpoff", VK_DTPOFF)
.Case("TLVP", VK_TLVP)
.Case("tlvp", VK_TLVP)
.Case("IMGREL", VK_COFF_IMGREL32)
.Case("imgrel", VK_COFF_IMGREL32)
.Default(VK_Invalid);
}
/* *** */
void MCTargetExpr::anchor() {}
/* *** */
bool MCExpr::EvaluateAsAbsolute(int64_t &Res) const {
return EvaluateAsAbsolute(Res, 0, 0, 0);
}
bool MCExpr::EvaluateAsAbsolute(int64_t &Res,
const MCAsmLayout &Layout) const {
return EvaluateAsAbsolute(Res, &Layout.getAssembler(), &Layout, 0);
}
bool MCExpr::EvaluateAsAbsolute(int64_t &Res,
const MCAsmLayout &Layout,
const SectionAddrMap &Addrs) const {
return EvaluateAsAbsolute(Res, &Layout.getAssembler(), &Layout, &Addrs);
}
bool MCExpr::EvaluateAsAbsolute(int64_t &Res, const MCAssembler &Asm) const {
return EvaluateAsAbsolute(Res, &Asm, 0, 0);
}
bool MCExpr::EvaluateAsAbsolute(int64_t &Res, const MCAssembler *Asm,
const MCAsmLayout *Layout,
const SectionAddrMap *Addrs) const {
MCValue Value;
// Fast path constants.
if (const MCConstantExpr *CE = dyn_cast<MCConstantExpr>(this)) {
Res = CE->getValue();
return true;
}
// FIXME: The use if InSet = Addrs is a hack. Setting InSet causes us
// absolutize differences across sections and that is what the MachO writer
// uses Addrs for.
bool IsRelocatable =
EvaluateAsRelocatableImpl(Value, Asm, Layout, Addrs, /*InSet*/ Addrs);
// Record the current value.
Res = Value.getConstant();
return IsRelocatable && Value.isAbsolute();
}
/// \brief Helper method for \see EvaluateSymbolAdd().
static void AttemptToFoldSymbolOffsetDifference(const MCAssembler *Asm,
const MCAsmLayout *Layout,
const SectionAddrMap *Addrs,
bool InSet,
const MCSymbolRefExpr *&A,
const MCSymbolRefExpr *&B,
int64_t &Addend) {
if (!A || !B)
return;
const MCSymbol &SA = A->getSymbol();
const MCSymbol &SB = B->getSymbol();
if (SA.isUndefined() || SB.isUndefined())
return;
if (!Asm->getWriter().IsSymbolRefDifferenceFullyResolved(*Asm, A, B, InSet))
return;
MCSymbolData &AD = Asm->getSymbolData(SA);
MCSymbolData &BD = Asm->getSymbolData(SB);
if (AD.getFragment() == BD.getFragment()) {
Addend += (AD.getOffset() - BD.getOffset());
// Pointers to Thumb symbols need to have their low-bit set to allow
// for interworking.
if (Asm->isThumbFunc(&SA))
Addend |= 1;
// Clear the symbol expr pointers to indicate we have folded these
// operands.
A = B = 0;
return;
}
if (!Layout)
return;
const MCSectionData &SecA = *AD.getFragment()->getParent();
const MCSectionData &SecB = *BD.getFragment()->getParent();
if ((&SecA != &SecB) && !Addrs)
return;
// Eagerly evaluate.
Addend += (Layout->getSymbolOffset(&Asm->getSymbolData(A->getSymbol())) -
Layout->getSymbolOffset(&Asm->getSymbolData(B->getSymbol())));
if (Addrs && (&SecA != &SecB))
Addend += (Addrs->lookup(&SecA) - Addrs->lookup(&SecB));
// Pointers to Thumb symbols need to have their low-bit set to allow
// for interworking.
if (Asm->isThumbFunc(&SA))
Addend |= 1;
// Clear the symbol expr pointers to indicate we have folded these
// operands.
A = B = 0;
}
/// \brief Evaluate the result of an add between (conceptually) two MCValues.
///
/// This routine conceptually attempts to construct an MCValue:
/// Result = (Result_A - Result_B + Result_Cst)
/// from two MCValue's LHS and RHS where
/// Result = LHS + RHS
/// and
/// Result = (LHS_A - LHS_B + LHS_Cst) + (RHS_A - RHS_B + RHS_Cst).
///
/// This routine attempts to aggresively fold the operands such that the result
/// is representable in an MCValue, but may not always succeed.
///
/// \returns True on success, false if the result is not representable in an
/// MCValue.
/// NOTE: It is really important to have both the Asm and Layout arguments.
/// They might look redundant, but this function can be used before layout
/// is done (see the object streamer for example) and having the Asm argument
/// lets us avoid relaxations early.
static bool EvaluateSymbolicAdd(const MCAssembler *Asm,
const MCAsmLayout *Layout,
const SectionAddrMap *Addrs,
bool InSet,
const MCValue &LHS,const MCSymbolRefExpr *RHS_A,
const MCSymbolRefExpr *RHS_B, int64_t RHS_Cst,
MCValue &Res) {
// FIXME: This routine (and other evaluation parts) are *incredibly* sloppy
// about dealing with modifiers. This will ultimately bite us, one day.
const MCSymbolRefExpr *LHS_A = LHS.getSymA();
const MCSymbolRefExpr *LHS_B = LHS.getSymB();
int64_t LHS_Cst = LHS.getConstant();
// Fold the result constant immediately.
int64_t Result_Cst = LHS_Cst + RHS_Cst;
assert((!Layout || Asm) &&
"Must have an assembler object if layout is given!");
// If we have a layout, we can fold resolved differences.
if (Asm) {
// First, fold out any differences which are fully resolved. By
// reassociating terms in
// Result = (LHS_A - LHS_B + LHS_Cst) + (RHS_A - RHS_B + RHS_Cst).
// we have the four possible differences:
// (LHS_A - LHS_B),
// (LHS_A - RHS_B),
// (RHS_A - LHS_B),
// (RHS_A - RHS_B).
// Since we are attempting to be as aggressive as possible about folding, we
// attempt to evaluate each possible alternative.
AttemptToFoldSymbolOffsetDifference(Asm, Layout, Addrs, InSet, LHS_A, LHS_B,
Result_Cst);
AttemptToFoldSymbolOffsetDifference(Asm, Layout, Addrs, InSet, LHS_A, RHS_B,
Result_Cst);
AttemptToFoldSymbolOffsetDifference(Asm, Layout, Addrs, InSet, RHS_A, LHS_B,
Result_Cst);
AttemptToFoldSymbolOffsetDifference(Asm, Layout, Addrs, InSet, RHS_A, RHS_B,
Result_Cst);
}
// We can't represent the addition or subtraction of two symbols.
if ((LHS_A && RHS_A) || (LHS_B && RHS_B))
return false;
// At this point, we have at most one additive symbol and one subtractive
// symbol -- find them.
const MCSymbolRefExpr *A = LHS_A ? LHS_A : RHS_A;
const MCSymbolRefExpr *B = LHS_B ? LHS_B : RHS_B;
// If we have a negated symbol, then we must have also have a non-negated
// symbol in order to encode the expression.
if (B && !A)
return false;
Res = MCValue::get(A, B, Result_Cst);
return true;
}
bool MCExpr::EvaluateAsRelocatable(MCValue &Res,
const MCAsmLayout &Layout) const {
return EvaluateAsRelocatableImpl(Res, &Layout.getAssembler(), &Layout,
0, false);
}
bool MCExpr::EvaluateAsRelocatableImpl(MCValue &Res,
const MCAssembler *Asm,
const MCAsmLayout *Layout,
const SectionAddrMap *Addrs,
bool InSet) const {
++stats::MCExprEvaluate;
switch (getKind()) {
case Target:
return cast<MCTargetExpr>(this)->EvaluateAsRelocatableImpl(Res, Layout);
case Constant:
Res = MCValue::get(cast<MCConstantExpr>(this)->getValue());
return true;
case SymbolRef: {
const MCSymbolRefExpr *SRE = cast<MCSymbolRefExpr>(this);
const MCSymbol &Sym = SRE->getSymbol();
// Evaluate recursively if this is a variable.
if (Sym.isVariable() && SRE->getKind() == MCSymbolRefExpr::VK_None) {
bool Ret = Sym.getVariableValue()->EvaluateAsRelocatableImpl(Res, Asm,
Layout,
Addrs,
true);
// If we failed to simplify this to a constant, let the target
// handle it.
if (Ret && !Res.getSymA() && !Res.getSymB())
return true;
}
Res = MCValue::get(SRE, 0, 0);
return true;
}
case Unary: {
const MCUnaryExpr *AUE = cast<MCUnaryExpr>(this);
MCValue Value;
if (!AUE->getSubExpr()->EvaluateAsRelocatableImpl(Value, Asm, Layout,
Addrs, InSet))
return false;
switch (AUE->getOpcode()) {
case MCUnaryExpr::LNot:
if (!Value.isAbsolute())
return false;
Res = MCValue::get(!Value.getConstant());
break;
case MCUnaryExpr::Minus:
/// -(a - b + const) ==> (b - a - const)
if (Value.getSymA() && !Value.getSymB())
return false;
Res = MCValue::get(Value.getSymB(), Value.getSymA(),
-Value.getConstant());
break;
case MCUnaryExpr::Not:
if (!Value.isAbsolute())
return false;
Res = MCValue::get(~Value.getConstant());
break;
case MCUnaryExpr::Plus:
Res = Value;
break;
}
return true;
}
case Binary: {
const MCBinaryExpr *ABE = cast<MCBinaryExpr>(this);
MCValue LHSValue, RHSValue;
if (!ABE->getLHS()->EvaluateAsRelocatableImpl(LHSValue, Asm, Layout,
Addrs, InSet) ||
!ABE->getRHS()->EvaluateAsRelocatableImpl(RHSValue, Asm, Layout,
Addrs, InSet))
return false;
// We only support a few operations on non-constant expressions, handle
// those first.
if (!LHSValue.isAbsolute() || !RHSValue.isAbsolute()) {
switch (ABE->getOpcode()) {
default:
return false;
case MCBinaryExpr::Sub:
// Negate RHS and add.
return EvaluateSymbolicAdd(Asm, Layout, Addrs, InSet, LHSValue,
RHSValue.getSymB(), RHSValue.getSymA(),
-RHSValue.getConstant(),
Res);
case MCBinaryExpr::Add:
return EvaluateSymbolicAdd(Asm, Layout, Addrs, InSet, LHSValue,
RHSValue.getSymA(), RHSValue.getSymB(),
RHSValue.getConstant(),
Res);
}
}
// FIXME: We need target hooks for the evaluation. It may be limited in
// width, and gas defines the result of comparisons and right shifts
// differently from Apple as.
int64_t LHS = LHSValue.getConstant(), RHS = RHSValue.getConstant();
int64_t Result = 0;
switch (ABE->getOpcode()) {
case MCBinaryExpr::Add: Result = LHS + RHS; break;
case MCBinaryExpr::And: Result = LHS & RHS; break;
case MCBinaryExpr::Div: Result = LHS / RHS; break;
case MCBinaryExpr::EQ: Result = LHS == RHS; break;
case MCBinaryExpr::GT: Result = LHS > RHS; break;
case MCBinaryExpr::GTE: Result = LHS >= RHS; break;
case MCBinaryExpr::LAnd: Result = LHS && RHS; break;
case MCBinaryExpr::LOr: Result = LHS || RHS; break;
case MCBinaryExpr::LT: Result = LHS < RHS; break;
case MCBinaryExpr::LTE: Result = LHS <= RHS; break;
case MCBinaryExpr::Mod: Result = LHS % RHS; break;
case MCBinaryExpr::Mul: Result = LHS * RHS; break;
case MCBinaryExpr::NE: Result = LHS != RHS; break;
case MCBinaryExpr::Or: Result = LHS | RHS; break;
case MCBinaryExpr::Shl: Result = LHS << RHS; break;
case MCBinaryExpr::Shr: Result = LHS >> RHS; break;
case MCBinaryExpr::Sub: Result = LHS - RHS; break;
case MCBinaryExpr::Xor: Result = LHS ^ RHS; break;
}
Res = MCValue::get(Result);
return true;
}
}
llvm_unreachable("Invalid assembly expression kind!");
}
const MCSection *MCExpr::FindAssociatedSection() const {
switch (getKind()) {
case Target:
// We never look through target specific expressions.
return cast<MCTargetExpr>(this)->FindAssociatedSection();
case Constant:
return MCSymbol::AbsolutePseudoSection;
case SymbolRef: {
const MCSymbolRefExpr *SRE = cast<MCSymbolRefExpr>(this);
const MCSymbol &Sym = SRE->getSymbol();
if (Sym.isDefined())
return &Sym.getSection();
return 0;
}
case Unary:
return cast<MCUnaryExpr>(this)->getSubExpr()->FindAssociatedSection();
case Binary: {
const MCBinaryExpr *BE = cast<MCBinaryExpr>(this);
const MCSection *LHS_S = BE->getLHS()->FindAssociatedSection();
const MCSection *RHS_S = BE->getRHS()->FindAssociatedSection();
// If either section is absolute, return the other.
if (LHS_S == MCSymbol::AbsolutePseudoSection)
return RHS_S;
if (RHS_S == MCSymbol::AbsolutePseudoSection)
return LHS_S;
// Otherwise, return the first non-null section.
return LHS_S ? LHS_S : RHS_S;
}
}
llvm_unreachable("Invalid assembly expression kind!");
}