llvm-project/llvm/lib/ExecutionEngine/Orc/OrcTargetSupport.cpp

217 lines
7.1 KiB
C++
Raw Normal View History

#include "llvm/ADT/Triple.h"
#include "llvm/ExecutionEngine/Orc/OrcTargetSupport.h"
#include "llvm/Support/Process.h"
#include <array>
using namespace llvm::orc;
namespace {
uint64_t executeCompileCallback(JITCompileCallbackManagerBase *JCBM,
TargetAddress CallbackID) {
return JCBM->executeCompileCallback(CallbackID);
}
}
namespace llvm {
namespace orc {
const char* OrcX86_64::ResolverBlockName = "orc_resolver_block";
void OrcX86_64::insertResolverBlock(
Module &M, JITCompileCallbackManagerBase &JCBM) {
// Trampoline code-sequence length, used to get trampoline address from return
// address.
const unsigned X86_64_TrampolineLength = 6;
// List of x86-64 GPRs to save. Note - RBP saved separately below.
std::array<const char *, 14> GPRs = {{
"rax", "rbx", "rcx", "rdx",
"rsi", "rdi", "r8", "r9",
"r10", "r11", "r12", "r13",
"r14", "r15"
}};
// Address of the executeCompileCallback function.
uint64_t CallbackAddr =
static_cast<uint64_t>(
reinterpret_cast<uintptr_t>(executeCompileCallback));
std::ostringstream AsmStream;
Triple TT(M.getTargetTriple());
// Switch to text section.
if (TT.getOS() == Triple::Darwin)
AsmStream << ".section __TEXT,__text,regular,pure_instructions\n"
<< ".align 4, 0x90\n";
else
AsmStream << ".text\n"
<< ".align 16, 0x90\n";
// Bake in a pointer to the callback manager immediately before the
// start of the resolver function.
AsmStream << "jit_callback_manager_addr:\n"
<< " .quad " << &JCBM << "\n";
// Start the resolver function.
AsmStream << ResolverBlockName << ":\n"
<< " pushq %rbp\n"
<< " movq %rsp, %rbp\n";
// Store the GPRs.
for (const auto &GPR : GPRs)
AsmStream << " pushq %" << GPR << "\n";
// Store floating-point state with FXSAVE.
// Note: We need to keep the stack 16-byte aligned, so if we've emitted an odd
// number of 64-bit pushes so far (GPRs.size() plus 1 for RBP) then add
// an extra 64 bits of padding to the FXSave area.
unsigned Padding = (GPRs.size() + 1) % 2 ? 8 : 0;
unsigned FXSaveSize = 512 + Padding;
AsmStream << " subq $" << FXSaveSize << ", %rsp\n"
<< " fxsave64 (%rsp)\n"
// Load callback manager address, compute trampoline address, call JIT.
<< " lea jit_callback_manager_addr(%rip), %rdi\n"
<< " movq (%rdi), %rdi\n"
<< " movq 0x8(%rbp), %rsi\n"
<< " subq $" << X86_64_TrampolineLength << ", %rsi\n"
<< " movabsq $" << CallbackAddr << ", %rax\n"
<< " callq *%rax\n"
// Replace the return to the trampoline with the return address of the
// compiled function body.
<< " movq %rax, 0x8(%rbp)\n"
// Restore the floating point state.
<< " fxrstor64 (%rsp)\n"
<< " addq $" << FXSaveSize << ", %rsp\n";
for (const auto &GPR : make_range(GPRs.rbegin(), GPRs.rend()))
AsmStream << " popq %" << GPR << "\n";
// Restore original RBP and return to compiled function body.
AsmStream << " popq %rbp\n"
<< " retq\n";
M.appendModuleInlineAsm(AsmStream.str());
}
OrcX86_64::LabelNameFtor
OrcX86_64::insertCompileCallbackTrampolines(Module &M,
TargetAddress ResolverBlockAddr,
unsigned NumCalls,
unsigned StartIndex) {
const char *ResolverBlockPtrName = "Lorc_resolve_block_addr";
std::ostringstream AsmStream;
Triple TT(M.getTargetTriple());
if (TT.getOS() == Triple::Darwin)
AsmStream << ".section __TEXT,__text,regular,pure_instructions\n"
<< ".align 4, 0x90\n";
else
AsmStream << ".text\n"
<< ".align 16, 0x90\n";
AsmStream << ResolverBlockPtrName << ":\n"
<< " .quad " << ResolverBlockAddr << "\n";
auto GetLabelName =
[=](unsigned I) {
std::ostringstream LabelStream;
LabelStream << "orc_jcc_" << (StartIndex + I);
return LabelStream.str();
};
for (unsigned I = 0; I < NumCalls; ++I)
AsmStream << GetLabelName(I) << ":\n"
<< " callq *" << ResolverBlockPtrName << "(%rip)\n";
M.appendModuleInlineAsm(AsmStream.str());
return GetLabelName;
}
OrcX86_64::IndirectStubsInfo::~IndirectStubsInfo() {
sys::Memory::releaseMappedMemory(StubsBlock);
sys::Memory::releaseMappedMemory(PtrsBlock);
}
std::error_code OrcX86_64::emitIndirectStubsBlock(IndirectStubsInfo &StubsInfo,
unsigned MinStubs,
void *InitialPtrVal) {
// Stub format is:
//
// .section __orc_stubs
// stub1:
// jmpq *ptr1(%rip)
// .byte 0xC4 ; <- Invalid opcode padding.
// .byte 0xF1
// stub2:
// jmpq *ptr2(%rip)
//
// ...
//
// .section __orc_ptrs
// ptr1:
// .quad 0x0
// ptr2:
// .quad 0x0
//
// ...
const unsigned StubSize = IndirectStubsInfo::StubSize;
// Emit at least MinStubs, rounded up to fill the pages allocated.
unsigned PageSize = sys::Process::getPageSize();
unsigned NumPages = ((MinStubs * StubSize) + (PageSize - 1)) / PageSize;
unsigned NumStubs = (NumPages * PageSize) / StubSize;
// Allocate memory for stubs and pointers in one call.
std::error_code EC;
auto InitialBlock = sys::Memory::allocateMappedMemory(2 * NumPages * PageSize,
nullptr,
sys::Memory::MF_READ |
sys::Memory::MF_WRITE,
EC);
if (EC)
return EC;
// Create separate MemoryBlocks representing the stubs and pointers.
sys::MemoryBlock StubsBlock(InitialBlock.base(), NumPages * PageSize);
sys::MemoryBlock PtrsBlock(static_cast<char*>(InitialBlock.base()) +
NumPages * PageSize,
NumPages * PageSize);
// Populate the stubs page stubs and mark it executable.
uint64_t *Stub = reinterpret_cast<uint64_t*>(StubsBlock.base());
uint64_t PtrOffsetField =
static_cast<uint64_t>(NumPages * PageSize - 6) << 16;
for (unsigned I = 0; I < NumStubs; ++I)
Stub[I] = 0xF1C40000000025ff | PtrOffsetField;
if (auto EC = sys::Memory::protectMappedMemory(StubsBlock,
sys::Memory::MF_READ |
sys::Memory::MF_EXEC))
return EC;
// Initialize all pointers to point at FailureAddress.
void **Ptr = reinterpret_cast<void**>(PtrsBlock.base());
for (unsigned I = 0; I < NumStubs; ++I)
Ptr[I] = InitialPtrVal;
StubsInfo.NumStubs = NumStubs;
StubsInfo.StubsBlock = std::move(StubsBlock);
StubsInfo.PtrsBlock = std::move(PtrsBlock);
return std::error_code();
}
} // End namespace orc.
} // End namespace llvm.