[EVM] Add datasize/dataoffset intrinsics

PavelKopyl · PavelKopyl · commit bd356fe5f3d1 · 2024-10-09T16:47:19.000+02:00
This includes:
 - datasize/dataoffset intrinsics
 - EVM specific ELF relocations
diff --git a/llvm/include/llvm/BinaryFormat/ELF.h b/llvm/include/llvm/BinaryFormat/ELF.h
@@ -323,6 +323,9 @@ enum {
   // EraVM local begin
   EM_ERAVM = 260,         // EraVM
   // EraVM local end
+  // EVM local begin
+  EM_EVM = 261            // EVM
+  // EVM local end
 };
 
 // Object file classes.
@@ -426,6 +429,13 @@ enum {
 #include "ELFRelocs/AArch64.def"
 };
 
+// EVM local begin
+// ELF Relocation types for EVM
+enum {
+#include "ELFRelocs/EVM.def"
+};
+// EVM local end
+
 // Special values for the st_other field in the symbol table entry for AArch64.
 enum {
   // Symbol may follow different calling convention than base PCS.
diff --git a/llvm/include/llvm/BinaryFormat/ELFRelocs/EVM.def b/llvm/include/llvm/BinaryFormat/ELFRelocs/EVM.def
@@ -0,0 +1,7 @@
+
+#ifndef ELF_RELOC
+#error "ELF_RELOC must be defined"
+#endif
+
+ELF_RELOC(R_EVM_NONE,                  0)
+ELF_RELOC(R_EVM_DATA,                  1)
diff --git a/llvm/include/llvm/IR/IntrinsicsEVM.td b/llvm/include/llvm/IR/IntrinsicsEVM.td
@@ -246,4 +246,20 @@ def int_evm_memcpyas1as4
                WriteOnly<ArgIndex<0>>, ReadOnly<ArgIndex<1>>,
                ImmArg<ArgIndex<3>>]>;
 
+// The following intrinsics are used for linking purposes.
+
+// Returns a code size of the Yul object with the name
+// passed in the metadata.
+def int_evm_datasize : DefaultAttrsIntrinsic<
+  [llvm_i256_ty], [llvm_metadata_ty],
+  [IntrNoMem, IntrSpeculatable]
+>;
+
+// Returns a code offset of the Yul object with the name
+// passed in the metadata.
+def int_evm_dataoffset : DefaultAttrsIntrinsic<
+  [llvm_i256_ty], [llvm_metadata_ty],
+  [IntrNoMem, IntrSpeculatable]
+>;
+
 } // TargetPrefix = "evm"
diff --git a/llvm/include/llvm/MC/MCExpr.h b/llvm/include/llvm/MC/MCExpr.h
@@ -357,7 +357,10 @@ class MCSymbolRefExpr : public MCExpr {
     VK_VE_TPOFF_LO32,  // symbol@tpoff_lo
 
     VK_TPREL,
-    VK_DTPREL
+    VK_DTPREL,
+    // EVM local begin
+    VK_EVM_DATA,
+    // EVM local end
   };
 
 private:
diff --git a/llvm/include/module.modulemap b/llvm/include/module.modulemap
@@ -107,6 +107,7 @@ module LLVM_BinaryFormat {
     textual header "llvm/BinaryFormat/ELFRelocs/Xtensa.def"
     textual header "llvm/BinaryFormat/WasmRelocs.def"
     textual header "llvm/BinaryFormat/MsgPack.def"
+    textual header "llvm/BinaryFormat/EVM.def"
 }
 
 module LLVM_Config {
diff --git a/llvm/lib/BinaryFormat/ELF.cpp b/llvm/lib/BinaryFormat/ELF.cpp
@@ -201,6 +201,9 @@ uint16_t ELF::convertArchNameToEMachine(StringRef Arch) {
       // EraVM local begin
       .Case("eravm", EM_ERAVM)
       // EraVM local end
+      // EVM local begin
+      .Case("evm", EM_EVM)
+      // EVM local end
       .Default(EM_NONE);
 }
 
@@ -571,6 +574,10 @@ StringRef ELF::convertEMachineToArchName(uint16_t EMachine) {
   case EM_ERAVM:
     return "eravm";
   // EraVM local end
+  // EVM local begin
+  case EM_EVM:
+    return "evm";
+  // EVM local end
   default:
     return "None";
   }
diff --git a/llvm/lib/MC/ELFObjectWriter.cpp b/llvm/lib/MC/ELFObjectWriter.cpp
@@ -1096,14 +1096,7 @@ uint64_t ELFWriter::writeObject(MCAssembler &Asm, const MCAsmLayout &Layout) {
   std::map<const MCSymbol *, std::vector<const MCSectionELF *>> GroupMembers;
 
   // Write out the ELF header ...
-  // EVM local begin
-  // HACK!!! For EVM target we don't need the whole EFL file,
-  // but just its .text section. The natural way would be to extract
-  // it using objdump utility, but design of our FE doesn't admit
-  // usage of any other tool besides the BE itslef.
-  if (!Ctx.getTargetTriple().isEVM())
-    writeHeader(Asm);
-  // EVM local end
+  writeHeader(Asm);
 
   // ... then the sections ...
   SectionOffsetsTy SectionOffsets;
@@ -1120,14 +1113,6 @@ uint64_t ELFWriter::writeObject(MCAssembler &Asm, const MCAsmLayout &Layout) {
     const uint64_t SecStart = align(Section.getAlign());
 
     const MCSymbolELF *SignatureSymbol = Section.getGroup();
-    // EVM local begin
-    if (Ctx.getTargetTriple().isEVM()) {
-      if (Section.getName() == ".text")
-        writeSectionData(Asm, Section, Layout);
-      continue;
-    }
-    // EVM local end
-
     writeSectionData(Asm, Section, Layout);
 
     uint64_t SecEnd = W.OS.tell();
@@ -1160,11 +1145,6 @@ uint64_t ELFWriter::writeObject(MCAssembler &Asm, const MCAsmLayout &Layout) {
     OWriter.TargetObjectWriter->addTargetSectionFlags(Ctx, Section);
   }
 
-  // EVM local begin
-  if (Ctx.getTargetTriple().isEVM())
-    return W.OS.tell() - StartOffset;
-  // EVM local end
-
   for (MCSectionELF *Group : Groups) {
     // Remember the offset into the file for this section.
     const uint64_t SecStart = align(Group->getAlign());
diff --git a/llvm/lib/MC/MCExpr.cpp b/llvm/lib/MC/MCExpr.cpp
@@ -388,6 +388,10 @@ StringRef MCSymbolRefExpr::getVariantKindName(VariantKind Kind) {
   case VK_VE_TLS_GD_LO32: return "tls_gd_lo";
   case VK_VE_TPOFF_HI32: return "tpoff_hi";
   case VK_VE_TPOFF_LO32: return "tpoff_lo";
+  // EVM local begin
+  case VK_EVM_DATA:
+    return "evm_data";
+    // EVM local end
   }
   llvm_unreachable("Invalid variant kind");
 }
@@ -532,6 +536,9 @@ MCSymbolRefExpr::getVariantKindForName(StringRef Name) {
     .Case("tls_gd_lo", VK_VE_TLS_GD_LO32)
     .Case("tpoff_hi", VK_VE_TPOFF_HI32)
     .Case("tpoff_lo", VK_VE_TPOFF_LO32)
+    // EVM local begin
+    .Case("evm_data", VK_EVM_DATA)
+    // EVM local end
     .Default(VK_Invalid);
 }
 
diff --git a/llvm/lib/MC/MCObjectFileInfo.cpp b/llvm/lib/MC/MCObjectFileInfo.cpp
@@ -386,6 +386,10 @@ void MCObjectFileInfo::initELFMCObjectFileInfo(const Triple &T, bool Large) {
 
   TextSection = Ctx->getELFSection(".text", ELF::SHT_PROGBITS,
                                    ELF::SHF_EXECINSTR | ELF::SHF_ALLOC);
+  // EVM local begin
+  if (T.getArch() == Triple::evm)
+    TextSection->setAlignment(Align(1));
+  // EVM local end
 
   DataSection = Ctx->getELFSection(".data", ELF::SHT_PROGBITS,
                                    ELF::SHF_WRITE | ELF::SHF_ALLOC);
diff --git a/llvm/lib/Object/ELF.cpp b/llvm/lib/Object/ELF.cpp
@@ -185,11 +185,20 @@ StringRef llvm::object::getELFRelocationTypeName(uint32_t Machine,
   case ELF::EM_ERAVM:
     switch (Type) {
 #include "llvm/BinaryFormat/ELFRelocs/EraVM.def"
-    default:
+  default:
       break;
     }
     break;
   // EraVM local end
+  // EVM local begin
+  case ELF::EM_EVM:
+    switch (Type) {
+#include "llvm/BinaryFormat/ELFRelocs/EVM.def"
+    default:
+      break;
+    }
+    break;
+  // EVM local end
   default:
     break;
   }
@@ -242,6 +251,10 @@ uint32_t llvm::object::getELFRelativeRelocationType(uint32_t Machine) {
     break;
   case ELF::EM_LOONGARCH:
     return ELF::R_LARCH_RELATIVE;
+  // EVM local begin
+  case ELF::EM_EVM:
+    break;
+  // EVM local end
   default:
     break;
   }
diff --git a/llvm/lib/Target/EVM/CMakeLists.txt b/llvm/lib/Target/EVM/CMakeLists.txt
@@ -36,6 +36,7 @@ add_llvm_target(EVMCodeGen
   EVMStackify.cpp
   EVMSubtarget.cpp
   EVMTargetMachine.cpp
+  EVMTargetObjectFile.cpp
   EVMTargetTransformInfo.cpp
 
   LINK_COMPONENTS
diff --git a/llvm/lib/Target/EVM/EVMISelLowering.cpp b/llvm/lib/Target/EVM/EVMISelLowering.cpp
@@ -18,6 +18,8 @@
 #include "MCTargetDesc/EVMMCTargetDesc.h"
 #include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/IntrinsicsEVM.h"
+#include "llvm/MC/MCContext.h"
+#include "llvm/MC/MCSymbol.h"
 
 using namespace llvm;
 
@@ -82,6 +84,7 @@ EVMTargetLowering::EVMTargetLowering(const TargetMachine &TM,
                      Custom);
 
   setOperationAction(ISD::INTRINSIC_VOID, MVT::Other, Custom);
+  setOperationAction(ISD::INTRINSIC_WO_CHAIN, MVT::Other, Custom);
 
   setJumpIsExpensive(false);
   setMaximumJumpTableSize(0);
@@ -117,6 +120,8 @@ SDValue EVMTargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const {
     llvm_unreachable("Unimplemented operation lowering");
   case ISD::GlobalAddress:
     return LowerGlobalAddress(Op, DAG);
+  case ISD::INTRINSIC_WO_CHAIN:
+    return lowerINTRINSIC_WO_CHAIN(Op, DAG);
   case ISD::LOAD:
     return LowerLOAD(Op, DAG);
   case ISD::STORE:
@@ -141,6 +146,35 @@ SDValue EVMTargetLowering::LowerGlobalAddress(SDValue Op,
       DAG.getTargetGlobalAddress(GA->getGlobal(), DL, VT, GA->getOffset()));
 }
 
+SDValue EVMTargetLowering::lowerINTRINSIC_WO_CHAIN(SDValue Op,
+                                                   SelectionDAG &DAG) const {
+  unsigned IntrID = cast<ConstantSDNode>(Op.getOperand(0))->getZExtValue();
+  switch (IntrID) {
+  default:
+    return SDValue();
+  case Intrinsic::evm_datasize:
+  case Intrinsic::evm_dataoffset:
+    return lowerIntrinsicDataSize(IntrID, Op, DAG);
+  }
+}
+
+SDValue EVMTargetLowering::lowerIntrinsicDataSize(unsigned IntrID, SDValue Op,
+                                                  SelectionDAG &DAG) const {
+  const SDLoc DL(Op);
+  EVT Ty = Op.getValueType();
+
+  MachineFunction &MF = DAG.getMachineFunction();
+  const MDNode *Metadata = cast<MDNodeSDNode>(Op.getOperand(1))->getMD();
+  StringRef ContractID = cast<MDString>(Metadata->getOperand(0))->getString();
+  bool IsDataSize = IntrID == Intrinsic::evm_datasize;
+  Twine SymbolReloc =
+      Twine(IsDataSize ? "__datasize_" : "__dataoffset_") + ContractID;
+  MCSymbol *Sym = MF.getContext().getOrCreateSymbol(SymbolReloc);
+  return SDValue(
+      DAG.getMachineNode(EVM::DATA, DL, Ty, DAG.getMCSymbol(Sym, MVT::i256)),
+      0);
+}
+
 SDValue EVMTargetLowering::LowerLOAD(SDValue Op, SelectionDAG &DAG) const {
   const SDLoc DL(Op);
   auto *Load = cast<LoadSDNode>(Op);
diff --git a/llvm/lib/Target/EVM/EVMISelLowering.h b/llvm/lib/Target/EVM/EVMISelLowering.h
@@ -131,6 +131,11 @@ class EVMTargetLowering final : public TargetLowering {
 
   SDValue LowerGlobalAddress(SDValue Op, SelectionDAG &DAG) const;
 
+  SDValue lowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG) const;
+
+  SDValue lowerIntrinsicDataSize(unsigned IntrID, SDValue Op,
+                                 SelectionDAG &DAG) const;
+
   SDValue LowerLOAD(SDValue Op, SelectionDAG &DAG) const;
 
   SDValue LowerSTORE(SDValue Op, SelectionDAG &DAG) const;
diff --git a/llvm/lib/Target/EVM/EVMInstrInfo.td b/llvm/lib/Target/EVM/EVMInstrInfo.td
@@ -1120,3 +1120,9 @@ def PUSH32_S : NI<(outs), (ins i256imm:$imm), [], true, "PUSH32 $imm",
   let BaseName = "PUSH32";
   let DecoderMethod = "decodePUSH<32>";
 }
+
+// Pseudo instructions for linkage
+let isCodeGenOnly = 1 in
+defm DATA
+  : I<(outs GPR:$dst), (ins jmptarget:$reloc), [], "",
+       "", 0, 0>;
diff --git a/llvm/lib/Target/EVM/EVMMCInstLower.cpp b/llvm/lib/Target/EVM/EVMMCInstLower.cpp
@@ -51,6 +51,9 @@ static void removeRegisterOperands(const MachineInstr *MI, MCInst &OutMI) {
       OutMI.erase(&MO);
     }
   }
+
+  if (RegOpcode == EVM::DATA)
+    OutMI.setOpcode(EVM::PUSH4_S);
 }
 
 MCSymbol *
@@ -130,10 +133,15 @@ void EVMMCInstLower::Lower(const MachineInstr *MI, MCInst &OutMI) {
         MCOp = MCOperand::createExpr(EVMCImmMCExpr::create(*Str, Ctx));
       }
     } break;
-    case MachineOperand::MO_MCSymbol:
-      MCOp =
-          MCOperand::createExpr(MCSymbolRefExpr::create(MO.getMCSymbol(), Ctx));
-      break;
+    case MachineOperand::MO_MCSymbol: {
+      MCSymbolRefExpr::VariantKind Kind = MCSymbolRefExpr::VariantKind::VK_None;
+      unsigned Opc = MI->getOpcode();
+      if (Opc == EVM::DATA)
+        Kind = MCSymbolRefExpr::VariantKind::VK_EVM_DATA;
+
+      MCOp = MCOperand::createExpr(
+          MCSymbolRefExpr::create(MO.getMCSymbol(), Kind, Ctx));
+    } break;
     case MachineOperand::MO_MachineBasicBlock:
       MCOp = MCOperand::createExpr(
           MCSymbolRefExpr::create(MO.getMBB()->getSymbol(), Ctx));
diff --git a/llvm/lib/Target/EVM/EVMTargetMachine.cpp b/llvm/lib/Target/EVM/EVMTargetMachine.cpp
@@ -14,10 +14,10 @@
 
 #include "EVMMachineFunctionInfo.h"
 #include "EVMTargetMachine.h"
+#include "EVMTargetObjectFile.h"
 #include "EVMTargetTransformInfo.h"
 #include "TargetInfo/EVMTargetInfo.h"
 #include "llvm/CodeGen/Passes.h"
-#include "llvm/CodeGen/TargetLoweringObjectFileImpl.h"
 #include "llvm/CodeGen/TargetPassConfig.h"
 #include "llvm/InitializePasses.h"
 #include "llvm/MC/MCAsmInfo.h"
@@ -70,7 +70,7 @@ EVMTargetMachine::EVMTargetMachine(const Target &T, const Triple &TT,
     : LLVMTargetMachine(T, computeDataLayout(), TT, CPU, FS, Options,
                         getEffectiveRelocModel(RM),
                         getEffectiveCodeModel(CM, CodeModel::Small), OL),
-      TLOF(std::make_unique<TargetLoweringObjectFileELF>()),
+      TLOF(std::make_unique<EVMELFTargetObjectFile>()),
       Subtarget(TT, std::string(CPU), std::string(FS), *this) {
   setRequiresStructuredCFG(true);
   initAsmInfo();
diff --git a/llvm/lib/Target/EVM/EVMTargetObjectFile.cpp b/llvm/lib/Target/EVM/EVMTargetObjectFile.cpp
@@ -0,0 +1,20 @@
+//===------------- EVMTargetObjectFile.cpp - EVM Object Info --------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines the functions of the EVM-specific subclass
+// of TargetLoweringObjectFile.
+//
+//===----------------------------------------------------------------------===//
+
+#include "EVMTargetObjectFile.h"
+
+using namespace llvm;
+
+// Code sections need to be aligned on 1, otherwise linker will add padding
+// between .text sections of the object files being linked.
+unsigned EVMELFTargetObjectFile::getTextSectionAlignment() const { return 1; }
diff --git a/llvm/lib/Target/EVM/EVMTargetObjectFile.h b/llvm/lib/Target/EVM/EVMTargetObjectFile.h
@@ -0,0 +1,30 @@
+//===---------- EVMTargetObjectFile.h - EVM Object Info -*- C++ ---------*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file declares the EVM-specific subclass of
+// TargetLoweringObjectFile.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_EVM_EVMTARGETOBJECTFILE_H
+#define LLVM_LIB_TARGET_EVM_EVMTARGETOBJECTFILE_H
+
+#include "llvm/CodeGen/TargetLoweringObjectFileImpl.h"
+
+namespace llvm {
+
+class EVMELFTargetObjectFile final : public TargetLoweringObjectFileELF {
+public:
+  EVMELFTargetObjectFile() = default;
+
+  unsigned getTextSectionAlignment() const override;
+};
+
+} // end namespace llvm
+
+#endif
diff --git a/llvm/lib/Target/EVM/MCTargetDesc/EVMAsmBackend.cpp b/llvm/lib/Target/EVM/MCTargetDesc/EVMAsmBackend.cpp
diff --git a/llvm/lib/Target/EVM/MCTargetDesc/EVMELFObjectWriter.cpp b/llvm/lib/Target/EVM/MCTargetDesc/EVMELFObjectWriter.cpp
diff --git a/llvm/lib/Target/EVM/MCTargetDesc/EVMFixupKinds.h b/llvm/lib/Target/EVM/MCTargetDesc/EVMFixupKinds.h
diff --git a/llvm/lib/Target/EVM/MCTargetDesc/EVMMCCodeEmitter.cpp b/llvm/lib/Target/EVM/MCTargetDesc/EVMMCCodeEmitter.cpp
diff --git a/llvm/test/MC/EVM/datasize_relocation.ll b/llvm/test/MC/EVM/datasize_relocation.ll
diff --git a/llvm/tools/llvm-readobj/ELFDumper.cpp b/llvm/tools/llvm-readobj/ELFDumper.cpp

Original file line number	Diff line number	Diff line change
`@@ -107,6 +107,7 @@ module LLVM_BinaryFormat {`
`107`	`107`	`textual header "llvm/BinaryFormat/ELFRelocs/Xtensa.def"`
`108`	`108`	`textual header "llvm/BinaryFormat/WasmRelocs.def"`
`109`	`109`	`textual header "llvm/BinaryFormat/MsgPack.def"`
	`110`	`+ textual header "llvm/BinaryFormat/EVM.def"`
`110`	`111`	`}`
`111`	`112`
`112`	`113`	`module LLVM_Config {`
Original file line number	Diff line number	Diff line change
`@@ -388,6 +388,10 @@ StringRef MCSymbolRefExpr::getVariantKindName(VariantKind Kind) {`
`388`	`388`	`case VK_VE_TLS_GD_LO32: return "tls_gd_lo";`
`389`	`389`	`case VK_VE_TPOFF_HI32: return "tpoff_hi";`
`390`	`390`	`case VK_VE_TPOFF_LO32: return "tpoff_lo";`
	`391`	`+ // EVM local begin`
	`392`	`+ case VK_EVM_DATA:`
	`393`	`+ return "evm_data";`
	`394`	`+ // EVM local end`
`391`	`395`	`}`
`392`	`396`	`llvm_unreachable("Invalid variant kind");`
`393`	`397`	`}`
`@@ -532,6 +536,9 @@ MCSymbolRefExpr::getVariantKindForName(StringRef Name) {`
`532`	`536`	`.Case("tls_gd_lo", VK_VE_TLS_GD_LO32)`
`533`	`537`	`.Case("tpoff_hi", VK_VE_TPOFF_HI32)`
`534`	`538`	`.Case("tpoff_lo", VK_VE_TPOFF_LO32)`
	`539`	`+ // EVM local begin`
	`540`	`+ .Case("evm_data", VK_EVM_DATA)`
	`541`	`+ // EVM local end`
`535`	`542`	`.Default(VK_Invalid);`
`536`	`543`	`}`
`537`	`544`