diff options
Diffstat (limited to 'lib')
| -rw-r--r-- | lib/ExecutionEngine/JIT/CMakeLists.txt | 1 | ||||
| -rw-r--r-- | lib/ExecutionEngine/JIT/Intercept.cpp | 162 | ||||
| -rw-r--r-- | lib/ExecutionEngine/JIT/JIT.cpp | 28 | ||||
| -rw-r--r-- | lib/ExecutionEngine/JIT/JIT.h | 1 | ||||
| -rw-r--r-- | lib/ExecutionEngine/JIT/JITMemoryManager.cpp | 153 | ||||
| -rw-r--r-- | lib/ExecutionEngine/MCJIT/CMakeLists.txt | 1 | ||||
| -rw-r--r-- | lib/ExecutionEngine/MCJIT/Intercept.cpp | 162 | ||||
| -rw-r--r-- | lib/ExecutionEngine/MCJIT/MCJIT.cpp | 20 | ||||
| -rw-r--r-- | lib/ExecutionEngine/MCJIT/MCJIT.h | 1 | ||||
| -rw-r--r-- | lib/ExecutionEngine/MCJIT/MCJITMemoryManager.h | 39 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp | 295 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp | 408 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.h | 176 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldImpl.h | 168 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.cpp | 622 | ||||
| -rw-r--r-- | lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h | 86 |
16 files changed, 1371 insertions, 952 deletions
diff --git a/lib/ExecutionEngine/JIT/CMakeLists.txt b/lib/ExecutionEngine/JIT/CMakeLists.txt index 52bb389..0972872 100644 --- a/lib/ExecutionEngine/JIT/CMakeLists.txt +++ b/lib/ExecutionEngine/JIT/CMakeLists.txt @@ -2,6 +2,7 @@ add_definitions(-DENABLE_X86_JIT) add_llvm_library(LLVMJIT + Intercept.cpp JIT.cpp JITDwarfEmitter.cpp JITEmitter.cpp diff --git a/lib/ExecutionEngine/JIT/Intercept.cpp b/lib/ExecutionEngine/JIT/Intercept.cpp new file mode 100644 index 0000000..2251a8e --- /dev/null +++ b/lib/ExecutionEngine/JIT/Intercept.cpp @@ -0,0 +1,162 @@ +//===-- Intercept.cpp - System function interception routines -------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// +// If a function call occurs to an external function, the JIT is designed to use +// the dynamic loader interface to find a function to call. This is useful for +// calling system calls and library functions that are not available in LLVM. +// Some system calls, however, need to be handled specially. For this reason, +// we intercept some of them here and use our own stubs to handle them. +// +//===----------------------------------------------------------------------===// + +#include "JIT.h" +#include "llvm/Support/ErrorHandling.h" +#include "llvm/Support/DynamicLibrary.h" +#include "llvm/Config/config.h" +using namespace llvm; + +// AtExitHandlers - List of functions to call when the program exits, +// registered with the atexit() library function. +static std::vector<void (*)()> AtExitHandlers; + +/// runAtExitHandlers - Run any functions registered by the program's +/// calls to atexit(3), which we intercept and store in +/// AtExitHandlers. +/// +static void runAtExitHandlers() { + while (!AtExitHandlers.empty()) { + void (*Fn)() = AtExitHandlers.back(); + AtExitHandlers.pop_back(); + Fn(); + } +} + +//===----------------------------------------------------------------------===// +// Function stubs that are invoked instead of certain library calls +//===----------------------------------------------------------------------===// + +// Force the following functions to be linked in to anything that uses the +// JIT. This is a hack designed to work around the all-too-clever Glibc +// strategy of making these functions work differently when inlined vs. when +// not inlined, and hiding their real definitions in a separate archive file +// that the dynamic linker can't see. For more info, search for +// 'libc_nonshared.a' on Google, or read http://llvm.org/PR274. +#if defined(__linux__) +#if defined(HAVE_SYS_STAT_H) +#include <sys/stat.h> +#endif +#include <fcntl.h> +#include <unistd.h> +/* stat functions are redirecting to __xstat with a version number. On x86-64 + * linking with libc_nonshared.a and -Wl,--export-dynamic doesn't make 'stat' + * available as an exported symbol, so we have to add it explicitly. + */ +namespace { +class StatSymbols { +public: + StatSymbols() { + sys::DynamicLibrary::AddSymbol("stat", (void*)(intptr_t)stat); + sys::DynamicLibrary::AddSymbol("fstat", (void*)(intptr_t)fstat); + sys::DynamicLibrary::AddSymbol("lstat", (void*)(intptr_t)lstat); + sys::DynamicLibrary::AddSymbol("stat64", (void*)(intptr_t)stat64); + sys::DynamicLibrary::AddSymbol("\x1stat64", (void*)(intptr_t)stat64); + sys::DynamicLibrary::AddSymbol("\x1open64", (void*)(intptr_t)open64); + sys::DynamicLibrary::AddSymbol("\x1lseek64", (void*)(intptr_t)lseek64); + sys::DynamicLibrary::AddSymbol("fstat64", (void*)(intptr_t)fstat64); + sys::DynamicLibrary::AddSymbol("lstat64", (void*)(intptr_t)lstat64); + sys::DynamicLibrary::AddSymbol("atexit", (void*)(intptr_t)atexit); + sys::DynamicLibrary::AddSymbol("mknod", (void*)(intptr_t)mknod); + } +}; +} +static StatSymbols initStatSymbols; +#endif // __linux__ + +// jit_exit - Used to intercept the "exit" library call. +static void jit_exit(int Status) { + runAtExitHandlers(); // Run atexit handlers... + exit(Status); +} + +// jit_atexit - Used to intercept the "atexit" library call. +static int jit_atexit(void (*Fn)()) { + AtExitHandlers.push_back(Fn); // Take note of atexit handler... + return 0; // Always successful +} + +static int jit_noop() { + return 0; +} + +//===----------------------------------------------------------------------===// +// +/// getPointerToNamedFunction - This method returns the address of the specified +/// function by using the dynamic loader interface. As such it is only useful +/// for resolving library symbols, not code generated symbols. +/// +void *JIT::getPointerToNamedFunction(const std::string &Name, + bool AbortOnFailure) { + if (!isSymbolSearchingDisabled()) { + // Check to see if this is one of the functions we want to intercept. Note, + // we cast to intptr_t here to silence a -pedantic warning that complains + // about casting a function pointer to a normal pointer. + if (Name == "exit") return (void*)(intptr_t)&jit_exit; + if (Name == "atexit") return (void*)(intptr_t)&jit_atexit; + + // We should not invoke parent's ctors/dtors from generated main()! + // On Mingw and Cygwin, the symbol __main is resolved to + // callee's(eg. tools/lli) one, to invoke wrong duplicated ctors + // (and register wrong callee's dtors with atexit(3)). + // We expect ExecutionEngine::runStaticConstructorsDestructors() + // is called before ExecutionEngine::runFunctionAsMain() is called. + if (Name == "__main") return (void*)(intptr_t)&jit_noop; + + const char *NameStr = Name.c_str(); + // If this is an asm specifier, skip the sentinal. + if (NameStr[0] == 1) ++NameStr; + + // If it's an external function, look it up in the process image... + void *Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr); + if (Ptr) return Ptr; + + // If it wasn't found and if it starts with an underscore ('_') character, + // and has an asm specifier, try again without the underscore. + if (Name[0] == 1 && NameStr[0] == '_') { + Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr+1); + if (Ptr) return Ptr; + } + + // Darwin/PPC adds $LDBLStub suffixes to various symbols like printf. These + // are references to hidden visibility symbols that dlsym cannot resolve. + // If we have one of these, strip off $LDBLStub and try again. +#if defined(__APPLE__) && defined(__ppc__) + if (Name.size() > 9 && Name[Name.size()-9] == '$' && + memcmp(&Name[Name.size()-8], "LDBLStub", 8) == 0) { + // First try turning $LDBLStub into $LDBL128. If that fails, strip it off. + // This mirrors logic in libSystemStubs.a. + std::string Prefix = std::string(Name.begin(), Name.end()-9); + if (void *Ptr = getPointerToNamedFunction(Prefix+"$LDBL128", false)) + return Ptr; + if (void *Ptr = getPointerToNamedFunction(Prefix, false)) + return Ptr; + } +#endif + } + + /// If a LazyFunctionCreator is installed, use it to get/create the function. + if (LazyFunctionCreator) + if (void *RP = LazyFunctionCreator(Name)) + return RP; + + if (AbortOnFailure) { + report_fatal_error("Program used external function '"+Name+ + "' which could not be resolved!"); + } + return 0; +} diff --git a/lib/ExecutionEngine/JIT/JIT.cpp b/lib/ExecutionEngine/JIT/JIT.cpp index 16b8ee2..f715f6f 100644 --- a/lib/ExecutionEngine/JIT/JIT.cpp +++ b/lib/ExecutionEngine/JIT/JIT.cpp @@ -23,7 +23,6 @@ #include "llvm/CodeGen/MachineCodeInfo.h" #include "llvm/ExecutionEngine/GenericValue.h" #include "llvm/ExecutionEngine/JITEventListener.h" -#include "llvm/ExecutionEngine/JITMemoryManager.h" #include "llvm/Target/TargetData.h" #include "llvm/Target/TargetMachine.h" #include "llvm/Target/TargetJITInfo.h" @@ -268,9 +267,9 @@ extern "C" { } JIT::JIT(Module *M, TargetMachine &tm, TargetJITInfo &tji, - JITMemoryManager *jmm, bool GVsWithCode) - : ExecutionEngine(M), TM(tm), TJI(tji), JMM(jmm), - AllocateGVsWithCode(GVsWithCode), isAlreadyCodeGenerating(false) { + JITMemoryManager *JMM, bool GVsWithCode) + : ExecutionEngine(M), TM(tm), TJI(tji), AllocateGVsWithCode(GVsWithCode), + isAlreadyCodeGenerating(false) { setTargetData(TM.getTargetData()); jitstate = new JITState(M); @@ -712,27 +711,6 @@ void *JIT::getPointerToBasicBlock(BasicBlock *BB) { } } -void *JIT::getPointerToNamedFunction(const std::string &Name, - bool AbortOnFailure){ - if (!isSymbolSearchingDisabled()) { - void *ptr = JMM->getPointerToNamedFunction(Name, false); - if (ptr) - return ptr; - } - - /// If a LazyFunctionCreator is installed, use it to get/create the function. - if (LazyFunctionCreator) - if (void *RP = LazyFunctionCreator(Name)) - return RP; - - if (AbortOnFailure) { - report_fatal_error("Program used external function '"+Name+ - "' which could not be resolved!"); - } - return 0; -} - - /// getOrEmitGlobalVariable - Return the address of the specified global /// variable, possibly emitting it to memory if needed. This is used by the /// Emitter. diff --git a/lib/ExecutionEngine/JIT/JIT.h b/lib/ExecutionEngine/JIT/JIT.h index c557981..17d33fe 100644 --- a/lib/ExecutionEngine/JIT/JIT.h +++ b/lib/ExecutionEngine/JIT/JIT.h @@ -58,7 +58,6 @@ class JIT : public ExecutionEngine { TargetMachine &TM; // The current target we are compiling to TargetJITInfo &TJI; // The JITInfo for the target we are compiling to JITCodeEmitter *JCE; // JCE object - JITMemoryManager *JMM; std::vector<JITEventListener*> EventListeners; /// AllocateGVsWithCode - Some applications require that global variables and diff --git a/lib/ExecutionEngine/JIT/JITMemoryManager.cpp b/lib/ExecutionEngine/JIT/JITMemoryManager.cpp index d404d0c..efd570d 100644 --- a/lib/ExecutionEngine/JIT/JITMemoryManager.cpp +++ b/lib/ExecutionEngine/JIT/JITMemoryManager.cpp @@ -314,17 +314,6 @@ namespace { /// should allocate a separate slab. static const size_t DefaultSizeThreshold; - /// getPointerToNamedFunction - This method returns the address of the - /// specified function by using the dlsym function call. As such it is only - /// useful for resolving library symbols, not code generated symbols. - /// - /// If AbortOnFailure is false and no function with the given name is - /// found, this function silently returns a null pointer. Otherwise, - /// it prints a message to stderr and aborts. - /// - virtual void *getPointerToNamedFunction(const std::string &Name, - bool AbortOnFailure = true); - void AllocateGOT(); // Testing methods. @@ -768,148 +757,6 @@ bool DefaultJITMemoryManager::CheckInvariants(std::string &ErrorStr) { return true; } -//===----------------------------------------------------------------------===// -// getPointerToNamedFunction() implementation. -//===----------------------------------------------------------------------===// -#include "llvm/Support/ErrorHandling.h" -#include "llvm/Support/DynamicLibrary.h" -#include "llvm/Config/config.h" - -// AtExitHandlers - List of functions to call when the program exits, -// registered with the atexit() library function. -static std::vector<void (*)()> AtExitHandlers; - -/// runAtExitHandlers - Run any functions registered by the program's -/// calls to atexit(3), which we intercept and store in -/// AtExitHandlers. -/// -static void runAtExitHandlers() { - while (!AtExitHandlers.empty()) { - void (*Fn)() = AtExitHandlers.back(); - AtExitHandlers.pop_back(); - Fn(); - } -} - -//===----------------------------------------------------------------------===// -// Function stubs that are invoked instead of certain library calls -//===----------------------------------------------------------------------===// - -// Force the following functions to be linked in to anything that uses the -// JIT. This is a hack designed to work around the all-too-clever Glibc -// strategy of making these functions work differently when inlined vs. when -// not inlined, and hiding their real definitions in a separate archive file -// that the dynamic linker can't see. For more info, search for -// 'libc_nonshared.a' on Google, or read http://llvm.org/PR274. -#if defined(__linux__) -#if defined(HAVE_SYS_STAT_H) -#include <sys/stat.h> -#endif -#include <fcntl.h> -#include <unistd.h> -/* stat functions are redirecting to __xstat with a version number. On x86-64 - * linking with libc_nonshared.a and -Wl,--export-dynamic doesn't make 'stat' - * available as an exported symbol, so we have to add it explicitly. - */ -namespace { -class StatSymbols { -public: - StatSymbols() { - sys::DynamicLibrary::AddSymbol("stat", (void*)(intptr_t)stat); - sys::DynamicLibrary::AddSymbol("fstat", (void*)(intptr_t)fstat); - sys::DynamicLibrary::AddSymbol("lstat", (void*)(intptr_t)lstat); - sys::DynamicLibrary::AddSymbol("stat64", (void*)(intptr_t)stat64); - sys::DynamicLibrary::AddSymbol("\x1stat64", (void*)(intptr_t)stat64); - sys::DynamicLibrary::AddSymbol("\x1open64", (void*)(intptr_t)open64); - sys::DynamicLibrary::AddSymbol("\x1lseek64", (void*)(intptr_t)lseek64); - sys::DynamicLibrary::AddSymbol("fstat64", (void*)(intptr_t)fstat64); - sys::DynamicLibrary::AddSymbol("lstat64", (void*)(intptr_t)lstat64); - sys::DynamicLibrary::AddSymbol("atexit", (void*)(intptr_t)atexit); - sys::DynamicLibrary::AddSymbol("mknod", (void*)(intptr_t)mknod); - } -}; -} -static StatSymbols initStatSymbols; -#endif // __linux__ - -// jit_exit - Used to intercept the "exit" library call. -static void jit_exit(int Status) { - runAtExitHandlers(); // Run atexit handlers... - exit(Status); -} - -// jit_atexit - Used to intercept the "atexit" library call. -static int jit_atexit(void (*Fn)()) { - AtExitHandlers.push_back(Fn); // Take note of atexit handler... - return 0; // Always successful -} - -static int jit_noop() { - return 0; -} - -//===----------------------------------------------------------------------===// -// -/// getPointerToNamedFunction - This method returns the address of the specified -/// function by using the dynamic loader interface. As such it is only useful -/// for resolving library symbols, not code generated symbols. -/// -void *DefaultJITMemoryManager::getPointerToNamedFunction(const std::string &Name, - bool AbortOnFailure) { - // Check to see if this is one of the functions we want to intercept. Note, - // we cast to intptr_t here to silence a -pedantic warning that complains - // about casting a function pointer to a normal pointer. - if (Name == "exit") return (void*)(intptr_t)&jit_exit; - if (Name == "atexit") return (void*)(intptr_t)&jit_atexit; - - // We should not invoke parent's ctors/dtors from generated main()! - // On Mingw and Cygwin, the symbol __main is resolved to - // callee's(eg. tools/lli) one, to invoke wrong duplicated ctors - // (and register wrong callee's dtors with atexit(3)). - // We expect ExecutionEngine::runStaticConstructorsDestructors() - // is called before ExecutionEngine::runFunctionAsMain() is called. - if (Name == "__main") return (void*)(intptr_t)&jit_noop; - - const char *NameStr = Name.c_str(); - // If this is an asm specifier, skip the sentinal. - if (NameStr[0] == 1) ++NameStr; - - // If it's an external function, look it up in the process image... - void *Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr); - if (Ptr) return Ptr; - - // If it wasn't found and if it starts with an underscore ('_') character, - // try again without the underscore. - if (NameStr[0] == '_') { - Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr+1); - if (Ptr) return Ptr; - } - - // Darwin/PPC adds $LDBLStub suffixes to various symbols like printf. These - // are references to hidden visibility symbols that dlsym cannot resolve. - // If we have one of these, strip off $LDBLStub and try again. -#if defined(__APPLE__) && defined(__ppc__) - if (Name.size() > 9 && Name[Name.size()-9] == '$' && - memcmp(&Name[Name.size()-8], "LDBLStub", 8) == 0) { - // First try turning $LDBLStub into $LDBL128. If that fails, strip it off. - // This mirrors logic in libSystemStubs.a. - std::string Prefix = std::string(Name.begin(), Name.end()-9); - if (void *Ptr = getPointerToNamedFunction(Prefix+"$LDBL128", false)) - return Ptr; - if (void *Ptr = getPointerToNamedFunction(Prefix, false)) - return Ptr; - } -#endif - - if (AbortOnFailure) { - report_fatal_error("Program used external function '"+Name+ - "' which could not be resolved!"); - } - return 0; -} - - - JITMemoryManager *JITMemoryManager::CreateDefaultMemManager() { return new DefaultJITMemoryManager(); } diff --git a/lib/ExecutionEngine/MCJIT/CMakeLists.txt b/lib/ExecutionEngine/MCJIT/CMakeLists.txt index fef7176..2c0f8d6 100644 --- a/lib/ExecutionEngine/MCJIT/CMakeLists.txt +++ b/lib/ExecutionEngine/MCJIT/CMakeLists.txt @@ -1,4 +1,5 @@ add_llvm_library(LLVMMCJIT MCJIT.cpp MCJITMemoryManager.cpp + Intercept.cpp ) diff --git a/lib/ExecutionEngine/MCJIT/Intercept.cpp b/lib/ExecutionEngine/MCJIT/Intercept.cpp new file mode 100644 index 0000000..f83f428 --- /dev/null +++ b/lib/ExecutionEngine/MCJIT/Intercept.cpp @@ -0,0 +1,162 @@ +//===-- Intercept.cpp - System function interception routines -------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// +// If a function call occurs to an external function, the JIT is designed to use +// the dynamic loader interface to find a function to call. This is useful for +// calling system calls and library functions that are not available in LLVM. +// Some system calls, however, need to be handled specially. For this reason, +// we intercept some of them here and use our own stubs to handle them. +// +//===----------------------------------------------------------------------===// + +#include "MCJIT.h" +#include "llvm/Support/ErrorHandling.h" +#include "llvm/Support/DynamicLibrary.h" +#include "llvm/Config/config.h" +using namespace llvm; + +// AtExitHandlers - List of functions to call when the program exits, +// registered with the atexit() library function. +static std::vector<void (*)()> AtExitHandlers; + +/// runAtExitHandlers - Run any functions registered by the program's +/// calls to atexit(3), which we intercept and store in +/// AtExitHandlers. +/// +static void runAtExitHandlers() { + while (!AtExitHandlers.empty()) { + void (*Fn)() = AtExitHandlers.back(); + AtExitHandlers.pop_back(); + Fn(); + } +} + +//===----------------------------------------------------------------------===// +// Function stubs that are invoked instead of certain library calls +//===----------------------------------------------------------------------===// + +// Force the following functions to be linked in to anything that uses the +// JIT. This is a hack designed to work around the all-too-clever Glibc +// strategy of making these functions work differently when inlined vs. when +// not inlined, and hiding their real definitions in a separate archive file +// that the dynamic linker can't see. For more info, search for +// 'libc_nonshared.a' on Google, or read http://llvm.org/PR274. +#if defined(__linux__) +#if defined(HAVE_SYS_STAT_H) +#include <sys/stat.h> +#endif +#include <fcntl.h> +#include <unistd.h> +/* stat functions are redirecting to __xstat with a version number. On x86-64 + * linking with libc_nonshared.a and -Wl,--export-dynamic doesn't make 'stat' + * available as an exported symbol, so we have to add it explicitly. + */ +namespace { +class StatSymbols { +public: + StatSymbols() { + sys::DynamicLibrary::AddSymbol("stat", (void*)(intptr_t)stat); + sys::DynamicLibrary::AddSymbol("fstat", (void*)(intptr_t)fstat); + sys::DynamicLibrary::AddSymbol("lstat", (void*)(intptr_t)lstat); + sys::DynamicLibrary::AddSymbol("stat64", (void*)(intptr_t)stat64); + sys::DynamicLibrary::AddSymbol("\x1stat64", (void*)(intptr_t)stat64); + sys::DynamicLibrary::AddSymbol("\x1open64", (void*)(intptr_t)open64); + sys::DynamicLibrary::AddSymbol("\x1lseek64", (void*)(intptr_t)lseek64); + sys::DynamicLibrary::AddSymbol("fstat64", (void*)(intptr_t)fstat64); + sys::DynamicLibrary::AddSymbol("lstat64", (void*)(intptr_t)lstat64); + sys::DynamicLibrary::AddSymbol("atexit", (void*)(intptr_t)atexit); + sys::DynamicLibrary::AddSymbol("mknod", (void*)(intptr_t)mknod); + } +}; +} +static StatSymbols initStatSymbols; +#endif // __linux__ + +// jit_exit - Used to intercept the "exit" library call. +static void jit_exit(int Status) { + runAtExitHandlers(); // Run atexit handlers... + exit(Status); +} + +// jit_atexit - Used to intercept the "atexit" library call. +static int jit_atexit(void (*Fn)()) { + AtExitHandlers.push_back(Fn); // Take note of atexit handler... + return 0; // Always successful +} + +static int jit_noop() { + return 0; +} + +//===----------------------------------------------------------------------===// +// +/// getPointerToNamedFunction - This method returns the address of the specified +/// function by using the dynamic loader interface. As such it is only useful +/// for resolving library symbols, not code generated symbols. +/// +void *MCJIT::getPointerToNamedFunction(const std::string &Name, + bool AbortOnFailure) { + if (!isSymbolSearchingDisabled()) { + // Check to see if this is one of the functions we want to intercept. Note, + // we cast to intptr_t here to silence a -pedantic warning that complains + // about casting a function pointer to a normal pointer. + if (Name == "exit") return (void*)(intptr_t)&jit_exit; + if (Name == "atexit") return (void*)(intptr_t)&jit_atexit; + + // We should not invoke parent's ctors/dtors from generated main()! + // On Mingw and Cygwin, the symbol __main is resolved to + // callee's(eg. tools/lli) one, to invoke wrong duplicated ctors + // (and register wrong callee's dtors with atexit(3)). + // We expect ExecutionEngine::runStaticConstructorsDestructors() + // is called before ExecutionEngine::runFunctionAsMain() is called. + if (Name == "__main") return (void*)(intptr_t)&jit_noop; + + const char *NameStr = Name.c_str(); + // If this is an asm specifier, skip the sentinal. + if (NameStr[0] == 1) ++NameStr; + + // If it's an external function, look it up in the process image... + void *Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr); + if (Ptr) return Ptr; + + // If it wasn't found and if it starts with an underscore ('_') character, + // and has an asm specifier, try again without the underscore. + if (Name[0] == 1 && NameStr[0] == '_') { + Ptr = sys::DynamicLibrary::SearchForAddressOfSymbol(NameStr+1); + if (Ptr) return Ptr; + } + + // Darwin/PPC adds $LDBLStub suffixes to various symbols like printf. These + // are references to hidden visibility symbols that dlsym cannot resolve. + // If we have one of these, strip off $LDBLStub and try again. +#if defined(__APPLE__) && defined(__ppc__) + if (Name.size() > 9 && Name[Name.size()-9] == '$' && + memcmp(&Name[Name.size()-8], "LDBLStub", 8) == 0) { + // First try turning $LDBLStub into $LDBL128. If that fails, strip it off. + // This mirrors logic in libSystemStubs.a. + std::string Prefix = std::string(Name.begin(), Name.end()-9); + if (void *Ptr = getPointerToNamedFunction(Prefix+"$LDBL128", false)) + return Ptr; + if (void *Ptr = getPointerToNamedFunction(Prefix, false)) + return Ptr; + } +#endif + } + + /// If a LazyFunctionCreator is installed, use it to get/create the function. + if (LazyFunctionCreator) + if (void *RP = LazyFunctionCreator(Name)) + return RP; + + if (AbortOnFailure) { + report_fatal_error("Program used external function '"+Name+ + "' which could not be resolved!"); + } + return 0; +} diff --git a/lib/ExecutionEngine/MCJIT/MCJIT.cpp b/lib/ExecutionEngine/MCJIT/MCJIT.cpp index cbb23d3..5f93a8d 100644 --- a/lib/ExecutionEngine/MCJIT/MCJIT.cpp +++ b/lib/ExecutionEngine/MCJIT/MCJIT.cpp @@ -215,23 +215,3 @@ GenericValue MCJIT::runFunction(Function *F, llvm_unreachable("Full-featured argument passing not supported yet!"); } - -void *MCJIT::getPointerToNamedFunction(const std::string &Name, - bool AbortOnFailure){ - if (!isSymbolSearchingDisabled()) { - void *ptr = MemMgr->getPointerToNamedFunction(Name, false); - if (ptr) - return ptr; - } - - /// If a LazyFunctionCreator is installed, use it to get/create the function. - if (LazyFunctionCreator) - if (void *RP = LazyFunctionCreator(Name)) - return RP; - - if (AbortOnFailure) { - report_fatal_error("Program used external function '"+Name+ - "' which could not be resolved!"); - } - return 0; -} diff --git a/lib/ExecutionEngine/MCJIT/MCJIT.h b/lib/ExecutionEngine/MCJIT/MCJIT.h index 2b3df98..7f4ae77 100644 --- a/lib/ExecutionEngine/MCJIT/MCJIT.h +++ b/lib/ExecutionEngine/MCJIT/MCJIT.h @@ -67,7 +67,6 @@ public: /// virtual void *getPointerToNamedFunction(const std::string &Name, bool AbortOnFailure = true); - /// mapSectionAddress - map a section to its target address space value. /// Map the address of a JIT section as returned from the memory manager /// to the address in the target process as the running code will see it. diff --git a/lib/ExecutionEngine/MCJIT/MCJITMemoryManager.h b/lib/ExecutionEngine/MCJIT/MCJITMemoryManager.h index dac8b26..ac8c155 100644 --- a/lib/ExecutionEngine/MCJIT/MCJITMemoryManager.h +++ b/lib/ExecutionEngine/MCJIT/MCJITMemoryManager.h @@ -33,17 +33,46 @@ public: uint8_t *allocateDataSection(uintptr_t Size, unsigned Alignment, unsigned SectionID) { - return JMM->allocateSpace(Size, Alignment); + return JMM->allocateDataSection(Size, Alignment, SectionID); } uint8_t *allocateCodeSection(uintptr_t Size, unsigned Alignment, unsigned SectionID) { - return JMM->allocateSpace(Size, Alignment); + return JMM->allocateCodeSection(Size, Alignment, SectionID); } - virtual void *getPointerToNamedFunction(const std::string &Name, - bool AbortOnFailure = true) { - return JMM->getPointerToNamedFunction(Name, AbortOnFailure); + // Allocate ActualSize bytes, or more, for the named function. Return + // a pointer to the allocated memory and update Size to reflect how much + // memory was acutally allocated. + uint8_t *startFunctionBody(const char *Name, uintptr_t &Size) { + // FIXME: This should really reference the MCAsmInfo to get the global + // prefix. + if (Name[0] == '_') ++Name; + Function *F = M->getFunction(Name); + // Some ObjC names have a prefixed \01 in the IR. If we failed to find + // the symbol and it's of the ObjC conventions (starts with "-" or + // "+"), try prepending a \01 and see if we can find it that way. + if (!F && (Name[0] == '-' || Name[0] == '+')) + F = M->getFunction((Twine("\1") + Name).str()); + assert(F && "No matching function in JIT IR Module!"); + return JMM->startFunctionBody(F, Size); + } + + // Mark the end of the function, including how much of the allocated + // memory was actually used. + void endFunctionBody(const char *Name, uint8_t *FunctionStart, + uint8_t *FunctionEnd) { + // FIXME: This should really reference the MCAsmInfo to get the global + // prefix. + if (Name[0] == '_') ++Name; + Function *F = M->getFunction(Name); + // Some ObjC names have a prefixed \01 in the IR. If we failed to find + // the symbol and it's of the ObjC conventions (starts with "-" or + // "+"), try prepending a \01 and see if we can find it that way. + if (!F && (Name[0] == '-' || Name[0] == '+')) + F = M->getFunction((Twine("\1") + Name).str()); + assert(F && "No matching function in JIT IR Module!"); + JMM->endFunctionBody(F, FunctionStart, FunctionEnd); } }; diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp index ff4a2c8..2896c2d 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp @@ -26,290 +26,45 @@ RuntimeDyldImpl::~RuntimeDyldImpl() {} namespace llvm { +void RuntimeDyldImpl::extractFunction(StringRef Name, uint8_t *StartAddress, + uint8_t *EndAddress) { + // FIXME: DEPRECATED in favor of by-section allocation. + // Allocate memory for the function via the memory manager. + uintptr_t Size = EndAddress - StartAddress + 1; + uintptr_t AllocSize = Size; + uint8_t *Mem = MemMgr->startFunctionBody(Name.data(), AllocSize); + assert(Size >= (uint64_t)(EndAddress - StartAddress + 1) && + "Memory manager failed to allocate enough memory!"); + // Copy the function payload into the memory block. + memcpy(Mem, StartAddress, Size); + MemMgr->endFunctionBody(Name.data(), Mem, Mem + Size); + // Remember where we put it. + unsigned SectionID = Sections.size(); + Sections.push_back(sys::MemoryBlock(Mem, Size)); + // Default the assigned address for this symbol to wherever this + // allocated it. + SymbolTable[Name] = SymbolLoc(SectionID, 0); + DEBUG(dbgs() << " allocated to [" << Mem << ", " << Mem + Size << "]\n"); +} // Resolve the relocations for all symbols we currently know about. void RuntimeDyldImpl::resolveRelocations() { - // First, resolve relocations assotiated with external symbols. - resolveSymbols(); - // Just iterate over the sections we have and resolve all the relocations // in them. Gross overkill, but it gets the job done. for (int i = 0, e = Sections.size(); i != e; ++i) { - reassignSectionAddress(i, Sections[i].LoadAddress); + reassignSectionAddress(i, SectionLoadAddress[i]); } } void RuntimeDyldImpl::mapSectionAddress(void *LocalAddress, uint64_t TargetAddress) { - for (unsigned i = 0, e = Sections.size(); i != e; ++i) { - if (Sections[i].Address == LocalAddress) { - reassignSectionAddress(i, TargetAddress); - return; - } - } - llvm_unreachable("Attempting to remap address of unknown section!"); -} - -bool RuntimeDyldImpl::loadObject(const MemoryBuffer *InputBuffer) { - // FIXME: ObjectFile don't modify MemoryBuffer. - // It should use const MemoryBuffer as parameter. - ObjectFile *obj = ObjectFile:: - createObjectFile(const_cast<MemoryBuffer*>(InputBuffer)); - - Arch = (Triple::ArchType)obj->getArch(); - - LocalSymbolMap LocalSymbols; // Functions and data symbols from the - // object file. - ObjSectionToIDMap LocalSections; // Used sections from the object file - - error_code err; - - - // Parse symbols - DEBUG(dbgs() << "Parse symbols:\n"); - for (symbol_iterator it = obj->begin_symbols(), itEnd = obj->end_symbols(); - it != itEnd; it.increment(err)) { - if (err) break; - object::SymbolRef::Type SymType; - StringRef Name; - if ((bool)(err = it->getType(SymType))) break; - if ((bool)(err = it->getName(Name))) break; - - if (SymType == object::SymbolRef::ST_Function || - SymType == object::SymbolRef::ST_Data) { - uint64_t FileOffset; - uint32_t flags; - StringRef sData; - section_iterator sIt = obj->end_sections(); - if ((bool)(err = it->getFileOffset(FileOffset))) break; - if ((bool)(err = it->getFlags(flags))) break; - if ((bool)(err = it->getSection(sIt))) break; - if (sIt == obj->end_sections()) continue; - if ((bool)(err = sIt->getContents(sData))) break; - const uint8_t* SymPtr = (const uint8_t*)InputBuffer->getBufferStart() + - (uintptr_t)FileOffset; - uintptr_t SectOffset = (uintptr_t)(SymPtr - (const uint8_t*)sData.begin()); - unsigned SectionID = - findOrEmitSection(*sIt, - SymType == object::SymbolRef::ST_Function, - LocalSections); - bool isGlobal = flags & SymbolRef::SF_Global; - LocalSymbols[Name.data()] = SymbolLoc(SectionID, SectOffset); - DEBUG(dbgs() << "\tFileOffset: " << format("%p", (uintptr_t)FileOffset) - << " flags: " << flags - << " SID: " << SectionID - << " Offset: " << format("%p", SectOffset)); - if (isGlobal) - SymbolTable[Name] = SymbolLoc(SectionID, SectOffset); - } - DEBUG(dbgs() << "\tType: " << SymType << " Name: " << Name << "\n"); - } - if (err) { - report_fatal_error(err.message()); - } - - // Parse and proccess relocations - DEBUG(dbgs() << "Parse relocations:\n"); - for (section_iterator sIt = obj->begin_sections(), - sItEnd = obj->end_sections(); sIt != sItEnd; sIt.increment(err)) { - if (err) break; - bool isFirstRelocation = true; - unsigned SectionID = 0; - StubMap Stubs; - - for (relocation_iterator it = sIt->begin_relocations(), - itEnd = sIt->end_relocations(); it != itEnd; it.increment(err)) { - if (err) break; - - // If it's first relocation in this section, find its SectionID - if (isFirstRelocation) { - SectionID = findOrEmitSection(*sIt, true, LocalSections); - DEBUG(dbgs() << "\tSectionID: " << SectionID << "\n"); - isFirstRelocation = false; - } - - ObjRelocationInfo RI; - RI.SectionID = SectionID; - if ((bool)(err = it->getAdditionalInfo(RI.AdditionalInfo))) break; - if ((bool)(err = it->getOffset(RI.Offset))) break; - if ((bool)(err = it->getSymbol(RI.Symbol))) break; - if ((bool)(err = it->getType(RI.Type))) break; - - DEBUG(dbgs() << "\t\tAddend: " << RI.AdditionalInfo - << " Offset: " << format("%p", (uintptr_t)RI.Offset) - << " Type: " << (uint32_t)(RI.Type & 0xffffffffL) - << "\n"); - processRelocationRef(RI, *obj, LocalSections, LocalSymbols, Stubs); - } - if (err) { - report_fatal_error(err.message()); - } - } - return false; -} - -unsigned RuntimeDyldImpl::emitSection(const SectionRef &Section, - bool IsCode) { - - unsigned StubBufSize = 0, - StubSize = getMaxStubSize(); - error_code err; - if (StubSize > 0) { - for (relocation_iterator it = Section.begin_relocations(), - itEnd = Section.end_relocations(); it != itEnd; it.increment(err)) - StubBufSize += StubSize; - } - StringRef data; - uint64_t Alignment64; - if ((bool)(err = Section.getContents(data))) report_fatal_error(err.message()); - if ((bool)(err = Section.getAlignment(Alignment64))) - report_fatal_error(err.message()); - - unsigned Alignment = (unsigned)Alignment64 & 0xffffffffL; - unsigned DataSize = data.size(); - unsigned Allocate = DataSize + StubBufSize; - unsigned SectionID = Sections.size(); - const char *pData = data.data(); - uint8_t *Addr = IsCode - ? MemMgr->allocateCodeSection(Allocate, Alignment, SectionID) - : MemMgr->allocateDataSection(Allocate, Alignment, SectionID); - - memcpy(Addr, pData, DataSize); - DEBUG(dbgs() << "emitSection SectionID: " << SectionID - << " obj addr: " << format("%p", pData) - << " new addr: " << format("%p", Addr) - << " DataSize: " << DataSize - << " StubBufSize: " << StubBufSize - << " Allocate: " << Allocate - << "\n"); - Sections.push_back(SectionEntry(Addr, Allocate, DataSize,(uintptr_t)pData)); - return SectionID; -} - -unsigned RuntimeDyldImpl:: -findOrEmitSection(const SectionRef &Section, bool IsCode, - ObjSectionToIDMap &LocalSections) { - - unsigned SectionID = 0; - ObjSectionToIDMap::iterator sIDIt = LocalSections.find(Section); - if (sIDIt != LocalSections.end()) - SectionID = sIDIt->second; - else { - SectionID = emitSection(Section, IsCode); - LocalSections[Section] = SectionID; - } - return SectionID; -} - -void RuntimeDyldImpl::AddRelocation(const RelocationValueRef &Value, - unsigned SectionID, uintptr_t Offset, - uint32_t RelType) { - DEBUG(dbgs() << "AddRelocation SymNamePtr: " << format("%p", Value.SymbolName) - << " SID: " << Value.SectionID - << " Addend: " << format("%p", Value.Addend) - << " Offset: " << format("%p", Offset) - << " RelType: " << format("%x", RelType) - << "\n"); - - if (Value.SymbolName == 0) { - Relocations[Value.SectionID].push_back(RelocationEntry( - SectionID, - Offset, - RelType, - Value.Addend)); - } else - SymbolRelocations[Value.SymbolName].push_back(RelocationEntry( - SectionID, - Offset, - RelType, - Value.Addend)); -} - -uint8_t *RuntimeDyldImpl::createStubFunction(uint8_t *Addr) { - // TODO: There is only ARM far stub now. We should add the Thumb stub, - // and stubs for branches Thumb - ARM and ARM - Thumb. - if (Arch == Triple::arm) { - uint32_t *StubAddr = (uint32_t*)Addr; - *StubAddr = 0xe51ff004; // ldr pc,<label> - return (uint8_t*)++StubAddr; - } - else - return Addr; + assert(SectionLocalMemToID.count(LocalAddress) && + "Attempting to remap address of unknown section!"); + unsigned SectionID = SectionLocalMemToID[LocalAddress]; + reassignSectionAddress(SectionID, TargetAddress); } -// Assign an address to a symbol name and resolve all the relocations -// associated with it. -void RuntimeDyldImpl::reassignSectionAddress(unsigned SectionID, - uint64_t Addr) { - // The address to use for relocation resolution is not - // the address of the local section buffer. We must be doing - // a remote execution environment of some sort. Re-apply any - // relocations referencing this section with the given address. - // - // Addr is a uint64_t because we can't assume the pointer width - // of the target is the same as that of the host. Just use a generic - // "big enough" type. - Sections[SectionID].LoadAddress = Addr; - DEBUG(dbgs() << "Resolving relocations Section #" << SectionID - << "\t" << format("%p", (uint8_t *)Addr) - << "\n"); - resolveRelocationList(Relocations[SectionID], Addr); -} - -void RuntimeDyldImpl::resolveRelocationEntry(const RelocationEntry &RE, - uint64_t Value) { - uint8_t *Target = Sections[RE.SectionID].Address + RE.Offset; - DEBUG(dbgs() << "\tSectionID: " << RE.SectionID - << " + " << RE.Offset << " (" << format("%p", Target) << ")" - << " Data: " << RE.Data - << " Addend: " << RE.Addend - << "\n"); - - resolveRelocation(Target, Sections[RE.SectionID].LoadAddress + RE.Offset, - Value, RE.Data, RE.Addend); -} - -void RuntimeDyldImpl::resolveRelocationList(const RelocationList &Relocs, - uint64_t Value) { - for (unsigned i = 0, e = Relocs.size(); i != e; ++i) { - resolveRelocationEntry(Relocs[i], Value); - } -} - -// resolveSymbols - Resolve any relocations to the specified symbols if -// we know where it lives. -void RuntimeDyldImpl::resolveSymbols() { - StringMap<RelocationList>::iterator it = SymbolRelocations.begin(), - itEnd = SymbolRelocations.end(); - for (; it != itEnd; it++) { - StringRef Name = it->first(); - RelocationList &Relocs = it->second; - StringMap<SymbolLoc>::const_iterator Loc = SymbolTable.find(Name); - if (Loc == SymbolTable.end()) { - // This is an external symbol, try to get it address from - // MemoryManager. - uint8_t *Addr = (uint8_t*) MemMgr->getPointerToNamedFunction(Name.data(), - true); - DEBUG(dbgs() << "Resolving relocations Name: " << Name - << "\t" << format("%p", Addr) - << "\n"); - resolveRelocationList(Relocs, (uintptr_t)Addr); - } else { - // Change the relocation to be section relative rather than symbol - // relative and move it to the resolved relocation list. - DEBUG(dbgs() << "Resolving symbol '" << Name << "'\n"); - for (int i = 0, e = Relocs.size(); i != e; ++i) { - RelocationEntry Entry = Relocs[i]; - Entry.Addend += Loc->second.second; - Relocations[Loc->second.first].push_back(Entry); - } - Relocs.clear(); - } - } -} - - //===----------------------------------------------------------------------===// // RuntimeDyld class implementation RuntimeDyld::RuntimeDyld(RTDyldMemoryManager *mm) { diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp index 9351b6c..e15b200 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp @@ -25,58 +25,222 @@ using namespace llvm::object; namespace llvm { +namespace { -void RuntimeDyldELF::resolveX86_64Relocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend) { - switch (Type) { - default: - llvm_unreachable("Relocation type not implemented yet!"); - break; +// FIXME: this function should probably not live here... +// +// Returns the name and address of an unrelocated symbol in an ELF section +void getSymbolInfo(symbol_iterator Sym, uint64_t &Addr, StringRef &Name) { + //FIXME: error checking here required to catch corrupt ELF objects... + error_code Err = Sym->getName(Name); + + uint64_t AddrInSection; + Err = Sym->getAddress(AddrInSection); + + SectionRef empty_section; + section_iterator Section(empty_section); + Err = Sym->getSection(Section); + + StringRef SectionContents; + Section->getContents(SectionContents); + + Addr = reinterpret_cast<uint64_t>(SectionContents.data()) + AddrInSection; +} + +} + +bool RuntimeDyldELF::loadObject(MemoryBuffer *InputBuffer) { + if (!isCompatibleFormat(InputBuffer)) + return true; + + OwningPtr<ObjectFile> Obj(ObjectFile::createELFObjectFile(InputBuffer)); + + Arch = Obj->getArch(); + + // Map address in the Object file image to function names + IntervalMap<uint64_t, StringRef>::Allocator A; + IntervalMap<uint64_t, StringRef> FuncMap(A); + + // This is a bit of a hack. The ObjectFile we've just loaded reports + // section addresses as 0 and doesn't provide access to the section + // offset (from which we could calculate the address. Instead, + // we're storing the address when it comes up in the ST_Debug case + // below. + // + StringMap<uint64_t> DebugSymbolMap; + + symbol_iterator SymEnd = Obj->end_symbols(); + error_code Err; + for (symbol_iterator Sym = Obj->begin_symbols(); + Sym != SymEnd; Sym.increment(Err)) { + SymbolRef::Type Type; + Sym->getType(Type); + if (Type == SymbolRef::ST_Function) { + StringRef Name; + uint64_t Addr; + getSymbolInfo(Sym, Addr, Name); + + uint64_t Size; + Err = Sym->getSize(Size); + + uint8_t *Start; + uint8_t *End; + Start = reinterpret_cast<uint8_t*>(Addr); + End = reinterpret_cast<uint8_t*>(Addr + Size - 1); + + extractFunction(Name, Start, End); + FuncMap.insert(Addr, Addr + Size - 1, Name); + } else if (Type == SymbolRef::ST_Debug) { + // This case helps us find section addresses + StringRef Name; + uint64_t Addr; + getSymbolInfo(Sym, Addr, Name); + DebugSymbolMap[Name] = Addr; + } + } + + // Iterate through the relocations for this object + section_iterator SecEnd = Obj->end_sections(); + for (section_iterator Sec = Obj->begin_sections(); + Sec != SecEnd; Sec.increment(Err)) { + StringRef SecName; + uint64_t SecAddr; + Sec->getName(SecName); + // Ignore sections that aren't in our map + if (DebugSymbolMap.find(SecName) == DebugSymbolMap.end()) { + continue; + } + SecAddr = DebugSymbolMap[SecName]; + relocation_iterator RelEnd = Sec->end_relocations(); + for (relocation_iterator Rel = Sec->begin_relocations(); + Rel != RelEnd; Rel.increment(Err)) { + uint64_t RelOffset; + uint64_t RelType; + int64_t RelAddend; + SymbolRef RelSym; + StringRef SymName; + uint64_t SymAddr; + uint64_t SymOffset; + + Rel->getAddress(RelOffset); + Rel->getType(RelType); + Rel->getAdditionalInfo(RelAddend); + Rel->getSymbol(RelSym); + RelSym.getName(SymName); + RelSym.getAddress(SymAddr); + RelSym.getFileOffset(SymOffset); + + // If this relocation is inside a function, we want to store the + // function name and a function-relative offset + IntervalMap<uint64_t, StringRef>::iterator ContainingFunc + = FuncMap.find(SecAddr + RelOffset); + if (ContainingFunc.valid()) { + // Re-base the relocation to make it relative to the target function + RelOffset = (SecAddr + RelOffset) - ContainingFunc.start(); + Relocations[SymName].push_back(RelocationEntry(ContainingFunc.value(), + RelOffset, + RelType, + RelAddend, + true)); + } else { + Relocations[SymName].push_back(RelocationEntry(SecName, + RelOffset, + RelType, + RelAddend, + false)); + } + } + } + return false; +} + +void RuntimeDyldELF::resolveRelocations() { + // FIXME: deprecated. should be changed to use the by-section + // allocation and relocation scheme. + + // Just iterate over the symbols in our symbol table and assign their + // addresses. + StringMap<SymbolLoc>::iterator i = SymbolTable.begin(); + StringMap<SymbolLoc>::iterator e = SymbolTable.end(); + for (;i != e; ++i) { + assert (i->getValue().second == 0 && "non-zero offset in by-function sym!"); + reassignSymbolAddress(i->getKey(), + (uint8_t*)Sections[i->getValue().first].base()); + } +} + +void RuntimeDyldELF::resolveX86_64Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE) { + uint8_t *TargetAddr; + if (RE.IsFunctionRelative) { + StringMap<SymbolLoc>::const_iterator Loc = SymbolTable.find(RE.Target); + assert(Loc != SymbolTable.end() && "Function for relocation not found"); + TargetAddr = + reinterpret_cast<uint8_t*>(Sections[Loc->second.first].base()) + + Loc->second.second + RE.Offset; + } else { + // FIXME: Get the address of the target section and add that to RE.Offset + llvm_unreachable("Non-function relocation not implemented yet!"); + } + + switch (RE.Type) { + default: llvm_unreachable("Relocation type not implemented yet!"); case ELF::R_X86_64_64: { - uint64_t *Target = (uint64_t*)(LocalAddress); - *Target = Value + Addend; + uint8_t **Target = reinterpret_cast<uint8_t**>(TargetAddr); + *Target = Addr + RE.Addend; break; } case ELF::R_X86_64_32: case ELF::R_X86_64_32S: { - Value += Addend; + uint64_t Value = reinterpret_cast<uint64_t>(Addr) + RE.Addend; // FIXME: Handle the possibility of this assertion failing - assert((Type == ELF::R_X86_64_32 && !(Value & 0xFFFFFFFF00000000ULL)) || - (Type == ELF::R_X86_64_32S && + assert((RE.Type == ELF::R_X86_64_32 && !(Value & 0xFFFFFFFF00000000ULL)) || + (RE.Type == ELF::R_X86_64_32S && (Value & 0xFFFFFFFF00000000ULL) == 0xFFFFFFFF00000000ULL)); uint32_t TruncatedAddr = (Value & 0xFFFFFFFF); - uint32_t *Target = reinterpret_cast<uint32_t*>(LocalAddress); + uint32_t *Target = reinterpret_cast<uint32_t*>(TargetAddr); *Target = TruncatedAddr; break; } case ELF::R_X86_64_PC32: { - uint32_t *Placeholder = reinterpret_cast<uint32_t*>(LocalAddress); - int64_t RealOffset = *Placeholder + Value + Addend - FinalAddress; - assert(RealOffset <= 214783647 && RealOffset >= -214783648); - int32_t TruncOffset = (RealOffset & 0xFFFFFFFF); + uint32_t *Placeholder = reinterpret_cast<uint32_t*>(TargetAddr); + uint64_t RealOffset = *Placeholder + + reinterpret_cast<uint64_t>(Addr) + + RE.Addend - reinterpret_cast<uint64_t>(TargetAddr); + assert((RealOffset & 0xFFFFFFFF) == RealOffset); + uint32_t TruncOffset = (RealOffset & 0xFFFFFFFF); *Placeholder = TruncOffset; break; } } } -void RuntimeDyldELF::resolveX86Relocation(uint8_t *LocalAddress, - uint32_t FinalAddress, - uint32_t Value, - uint32_t Type, - int32_t Addend) { - switch (Type) { +void RuntimeDyldELF::resolveX86Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE) { + uint8_t *TargetAddr; + if (RE.IsFunctionRelative) { + StringMap<SymbolLoc>::const_iterator Loc = SymbolTable.find(RE.Target); + assert(Loc != SymbolTable.end() && "Function for relocation not found"); + TargetAddr = + reinterpret_cast<uint8_t*>(Sections[Loc->second.first].base()) + + Loc->second.second + RE.Offset; + } else { + // FIXME: Get the address of the target section and add that to RE.Offset + llvm_unreachable("Non-function relocation not implemented yet!"); + } + + switch (RE.Type) { case ELF::R_386_32: { - uint32_t *Target = (uint32_t*)(LocalAddress); - *Target = Value + Addend; + uint8_t **Target = reinterpret_cast<uint8_t**>(TargetAddr); + *Target = Addr + RE.Addend; break; } case ELF::R_386_PC32: { - uint32_t *Placeholder = reinterpret_cast<uint32_t*>(LocalAddress); - uint32_t RealOffset = *Placeholder + Value + Addend - FinalAddress; + uint32_t *Placeholder = reinterpret_cast<uint32_t*>(TargetAddr); + uint32_t RealOffset = *Placeholder + reinterpret_cast<uintptr_t>(Addr) + + RE.Addend - reinterpret_cast<uintptr_t>(TargetAddr); *Placeholder = RealOffset; break; } @@ -84,173 +248,57 @@ void RuntimeDyldELF::resolveX86Relocation(uint8_t *LocalAddress, // There are other relocation types, but it appears these are the // only ones currently used by the LLVM ELF object writer llvm_unreachable("Relocation type not implemented yet!"); - break; } } -void RuntimeDyldELF::resolveARMRelocation(uint8_t *LocalAddress, - uint32_t FinalAddress, - uint32_t Value, - uint32_t Type, - int32_t Addend) { - // TODO: Add Thumb relocations. - uint32_t* TargetPtr = (uint32_t*)LocalAddress; - Value += Addend; - - DEBUG(dbgs() << "resolveARMRelocation, LocalAddress: " << LocalAddress - << " FinalAddress: " << format("%p",FinalAddress) - << " Value: " << format("%x",Value) - << " Type: " << format("%x",Type) - << " Addend: " << format("%x",Addend) - << "\n"); - - switch(Type) { - default: - llvm_unreachable("Not implemented relocation type!"); - - // Just write 32bit value to relocation address - case ELF::R_ARM_ABS32 : - *TargetPtr = Value; - break; - - // Write first 16 bit of 32 bit value to the mov instruction. - // Last 4 bit should be shifted. - case ELF::R_ARM_MOVW_ABS_NC : - Value = Value & 0xFFFF; - *TargetPtr |= Value & 0xFFF; - *TargetPtr |= ((Value >> 12) & 0xF) << 16; - break; - - // Write last 16 bit of 32 bit value to the mov instruction. - // Last 4 bit should be shifted. - case ELF::R_ARM_MOVT_ABS : - Value = (Value >> 16) & 0xFFFF; - *TargetPtr |= Value & 0xFFF; - *TargetPtr |= ((Value >> 12) & 0xF) << 16; - break; - - // Write 24 bit relative value to the branch instruction. - case ELF::R_ARM_PC24 : // Fall through. - case ELF::R_ARM_CALL : // Fall through. - case ELF::R_ARM_JUMP24 : - int32_t RelValue = static_cast<int32_t>(Value - FinalAddress - 8); - RelValue = (RelValue & 0x03FFFFFC) >> 2; - *TargetPtr &= 0xFF000000; - *TargetPtr |= RelValue; - break; - } +void RuntimeDyldELF::resolveArmRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE) { } -void RuntimeDyldELF::resolveRelocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend) { +void RuntimeDyldELF::resolveRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE) { switch (Arch) { case Triple::x86_64: - resolveX86_64Relocation(LocalAddress, FinalAddress, Value, Type, Addend); + resolveX86_64Relocation(Name, Addr, RE); break; case Triple::x86: - resolveX86Relocation(LocalAddress, (uint32_t)(FinalAddress & 0xffffffffL), - (uint32_t)(Value & 0xffffffffL), Type, - (uint32_t)(Addend & 0xffffffffL)); + resolveX86Relocation(Name, Addr, RE); break; - case Triple::arm: // Fall through. - case Triple::thumb: - resolveARMRelocation(LocalAddress, (uint32_t)(FinalAddress & 0xffffffffL), - (uint32_t)(Value & 0xffffffffL), Type, - (uint32_t)(Addend & 0xffffffffL)); + case Triple::arm: + resolveArmRelocation(Name, Addr, RE); break; default: llvm_unreachable("Unsupported CPU type!"); } } -void RuntimeDyldELF:: -processRelocationRef(const ObjRelocationInfo &Rel, const ObjectFile &Obj, - ObjSectionToIDMap &ObjSectionToID, - LocalSymbolMap &Symbols, StubMap &Stubs) { - - uint32_t RelType = (uint32_t)(Rel.Type & 0xffffffffL); - intptr_t Addend = (intptr_t)Rel.AdditionalInfo; - RelocationValueRef Value; - StringRef TargetName; - const SymbolRef &Symbol = Rel.Symbol; - Symbol.getName(TargetName); - DEBUG(dbgs() << "\t\tRelType: " << RelType - << " Addend: " << Addend - << " TargetName: " << TargetName - << "\n"); - // First look the symbol in object file symbols. - LocalSymbolMap::iterator it = Symbols.find(TargetName.data()); - if (it != Symbols.end()) { - Value.SectionID = it->second.first; - Value.Addend = it->second.second; - } else { - // Second look the symbol in global symbol table. - StringMap<SymbolLoc>::iterator itS = SymbolTable.find(TargetName.data()); - if (itS != SymbolTable.end()) { - Value.SectionID = itS->second.first; - Value.Addend = itS->second.second; - } else { - SymbolRef::Type SymType; - Symbol.getType(SymType); - switch (SymType) { - case SymbolRef::ST_Debug: { - // TODO: Now ELF SymbolRef::ST_Debug = STT_SECTION, it's not obviously - // and can be changed by another developers. Maybe best way is add - // a new symbol type ST_Section to SymbolRef and use it. - section_iterator sIt = Obj.end_sections(); - Symbol.getSection(sIt); - if (sIt == Obj.end_sections()) - llvm_unreachable("Symbol section not found, bad object file format!"); - DEBUG(dbgs() << "\t\tThis is section symbol\n"); - Value.SectionID = findOrEmitSection((*sIt), true, ObjSectionToID); - Value.Addend = Addend; - break; - } - case SymbolRef::ST_Unknown: { - Value.SymbolName = TargetName.data(); - Value.Addend = Addend; - break; - } - default: - llvm_unreachable("Unresolved symbol type!"); - break; - } - } +void RuntimeDyldELF::reassignSymbolAddress(StringRef Name, uint8_t *Addr) { + // FIXME: deprecated. switch to reassignSectionAddress() instead. + // + // Actually moving the symbol address requires by-section mapping. + assert(Sections[SymbolTable.lookup(Name).first].base() == (void*)Addr && + "Unable to relocate section in by-function JIT allocation model!"); + + RelocationList &Relocs = Relocations[Name]; + for (unsigned i = 0, e = Relocs.size(); i != e; ++i) { + RelocationEntry &RE = Relocs[i]; + resolveRelocation(Name, Addr, RE); } - DEBUG(dbgs() << "\t\tRel.SectionID: " << Rel.SectionID - << " Rel.Offset: " << Rel.Offset - << "\n"); - if (Arch == Triple::arm && - (RelType == ELF::R_ARM_PC24 || - RelType == ELF::R_ARM_CALL || - RelType == ELF::R_ARM_JUMP24)) { - // This is an ARM branch relocation, need to use a stub function. - DEBUG(dbgs() << "\t\tThis is an ARM branch relocation."); - SectionEntry &Section = Sections[Rel.SectionID]; - uint8_t *Target = Section.Address + Rel.Offset; - - // Look up for existing stub. - StubMap::const_iterator stubIt = Stubs.find(Value); - if (stubIt != Stubs.end()) { - resolveRelocation(Target, Section.LoadAddress, (uint64_t)Section.Address + - stubIt->second, RelType, 0); - DEBUG(dbgs() << " Stub function found\n"); - } else { - // Create a new stub function. - DEBUG(dbgs() << " Create a new stub function\n"); - Stubs[Value] = Section.StubOffset; - uint8_t *StubTargetAddr = createStubFunction(Section.Address + - Section.StubOffset); - AddRelocation(Value, Rel.SectionID, - StubTargetAddr - Section.Address, ELF::R_ARM_ABS32); - resolveRelocation(Target, Section.LoadAddress, (uint64_t)Section.Address + - Section.StubOffset, RelType, 0); - Section.StubOffset += getMaxStubSize(); - } - } else - AddRelocation(Value, Rel.SectionID, Rel.Offset, RelType); +} + +// Assign an address to a symbol name and resolve all the relocations +// associated with it. +void RuntimeDyldELF::reassignSectionAddress(unsigned SectionID, uint64_t Addr) { + // The address to use for relocation resolution is not + // the address of the local section buffer. We must be doing + // a remote execution environment of some sort. Re-apply any + // relocations referencing this section with the given address. + // + // Addr is a uint64_t because we can't assume the pointer width + // of the target is the same as that of the host. Just use a generic + // "big enough" type. + assert(0); } bool RuntimeDyldELF::isCompatibleFormat(const MemoryBuffer *InputBuffer) const { diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.h b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.h index 36566da..e0f7d54 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.h +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.h @@ -21,42 +21,158 @@ using namespace llvm; namespace llvm { class RuntimeDyldELF : public RuntimeDyldImpl { -protected: - void resolveX86_64Relocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend); - - void resolveX86Relocation(uint8_t *LocalAddress, - uint32_t FinalAddress, - uint32_t Value, - uint32_t Type, - int32_t Addend); - - void resolveARMRelocation(uint8_t *LocalAddress, - uint32_t FinalAddress, - uint32_t Value, - uint32_t Type, - int32_t Addend); - - virtual void resolveRelocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend); - - virtual void processRelocationRef(const ObjRelocationInfo &Rel, - const ObjectFile &Obj, - ObjSectionToIDMap &ObjSectionToID, - LocalSymbolMap &Symbols, StubMap &Stubs); + // For each symbol, keep a list of relocations based on it. Anytime + // its address is reassigned (the JIT re-compiled the function, e.g.), + // the relocations get re-resolved. + struct RelocationEntry { + // Function or section this relocation is contained in. + std::string Target; + // Offset into the target function or section for the relocation. + uint32_t Offset; + // Relocation type + uint32_t Type; + // Addend encoded in the instruction itself, if any. + int32_t Addend; + // Has the relocation been recalcuated as an offset within a function? + bool IsFunctionRelative; + // Has this relocation been resolved previously? + bool isResolved; + + RelocationEntry(StringRef t, + uint32_t offset, + uint32_t type, + int32_t addend, + bool isFunctionRelative) + : Target(t) + , Offset(offset) + , Type(type) + , Addend(addend) + , IsFunctionRelative(isFunctionRelative) + , isResolved(false) { } + }; + typedef SmallVector<RelocationEntry, 4> RelocationList; + StringMap<RelocationList> Relocations; + unsigned Arch; + + void resolveRelocations(); + + void resolveX86_64Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveX86Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveArmRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + +public: + RuntimeDyldELF(RTDyldMemoryManager *mm) : RuntimeDyldImpl(mm) {} + + bool loadObject(MemoryBuffer *InputBuffer); + + void reassignSymbolAddress(StringRef Name, uint8_t *Addr); + void reassignSectionAddress(unsigned SectionID, uint64_t Addr); + + bool isCompatibleFormat(const MemoryBuffer *InputBuffer) const; +}; + +} // end namespace llvm + +#endif + +//===-- RuntimeDyldELF.h - Run-time dynamic linker for MC-JIT ---*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// +// ELF support for MC-JIT runtime dynamic linker. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_RUNTIME_DYLD_ELF_H +#define LLVM_RUNTIME_DYLD_ELF_H + +#include "RuntimeDyldImpl.h" + +using namespace llvm; + + +namespace llvm { +class RuntimeDyldELF : public RuntimeDyldImpl { + // For each symbol, keep a list of relocations based on it. Anytime + // its address is reassigned (the JIT re-compiled the function, e.g.), + // the relocations get re-resolved. + struct RelocationEntry { + // Function or section this relocation is contained in. + std::string Target; + // Offset into the target function or section for the relocation. + uint32_t Offset; + // Relocation type + uint32_t Type; + // Addend encoded in the instruction itself, if any. + int32_t Addend; + // Has the relocation been recalcuated as an offset within a function? + bool IsFunctionRelative; + // Has this relocation been resolved previously? + bool isResolved; + + RelocationEntry(StringRef t, + uint32_t offset, + uint32_t type, + int32_t addend, + bool isFunctionRelative) + : Target(t) + , Offset(offset) + , Type(type) + , Addend(addend) + , IsFunctionRelative(isFunctionRelative) + , isResolved(false) { } + }; + typedef SmallVector<RelocationEntry, 4> RelocationList; + StringMap<RelocationList> Relocations; + unsigned Arch; + + void resolveRelocations(); + + void resolveX86_64Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveX86Relocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveArmRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); + + void resolveRelocation(StringRef Name, + uint8_t *Addr, + const RelocationEntry &RE); public: RuntimeDyldELF(RTDyldMemoryManager *mm) : RuntimeDyldImpl(mm) {} + bool loadObject(MemoryBuffer *InputBuffer); + + void reassignSymbolAddress(StringRef Name, uint8_t *Addr); + void reassignSectionAddress(unsigned SectionID, uint64_t Addr); + bool isCompatibleFormat(const MemoryBuffer *InputBuffer) const; }; } // end namespace llvm -#endif +#endif + diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldImpl.h b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldImpl.h index d6430a9..28e99be 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldImpl.h +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldImpl.h @@ -15,125 +15,45 @@ #define LLVM_RUNTIME_DYLD_IMPL_H #include "llvm/ExecutionEngine/RuntimeDyld.h" -#include "llvm/Object/ObjectFile.h" #include "llvm/ADT/DenseMap.h" #include "llvm/ADT/StringMap.h" #include "llvm/ADT/Twine.h" #include "llvm/ADT/SmallVector.h" +#include "llvm/ExecutionEngine/ExecutionEngine.h" #include "llvm/Support/Memory.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/system_error.h" #include "llvm/Support/raw_ostream.h" #include "llvm/Support/Debug.h" #include "llvm/Support/ErrorHandling.h" -#include "llvm/ADT/Triple.h" -#include <map> -#include "llvm/Support/Format.h" using namespace llvm; -using namespace llvm::object; namespace llvm { - -class SectionEntry { -public: - uint8_t* Address; - size_t Size; - uint64_t LoadAddress; // For each section, the address it will be - // considered to live at for relocations. The same - // as the pointer to the above memory block for - // hosted JITs. - uintptr_t StubOffset; // It's used for architecturies with stub - // functions for far relocations like ARM. - uintptr_t ObjAddress; // Section address in object file. It's use for - // calculate MachO relocation addend - SectionEntry(uint8_t* address, size_t size, uintptr_t stubOffset, - uintptr_t objAddress) - : Address(address), Size(size), LoadAddress((uintptr_t)address), - StubOffset(stubOffset), ObjAddress(objAddress) {} -}; - -class RelocationEntry { -public: - unsigned SectionID; // Section the relocation is contained in. - uintptr_t Offset; // Offset into the section for the relocation. - uint32_t Data; // Relocatino data. Including type of relocation - // and another flags and parameners from - intptr_t Addend; // Addend encoded in the instruction itself, if any, - // plus the offset into the source section for - // the symbol once the relocation is resolvable. - RelocationEntry(unsigned id, uint64_t offset, uint32_t data, int64_t addend) - : SectionID(id), Offset(offset), Data(data), Addend(addend) {} -}; - -// Raw relocation data from object file -class ObjRelocationInfo { -public: - unsigned SectionID; - uint64_t Offset; - SymbolRef Symbol; - uint64_t Type; - int64_t AdditionalInfo; -}; - -class RelocationValueRef { -public: - unsigned SectionID; - intptr_t Addend; - const char *SymbolName; - RelocationValueRef(): SectionID(0), Addend(0), SymbolName(0) {} - - inline bool operator==(const RelocationValueRef &Other) const { - return std::memcmp(this, &Other, sizeof(RelocationValueRef)) == 0; - } - inline bool operator <(const RelocationValueRef &Other) const { - return std::memcmp(this, &Other, sizeof(RelocationValueRef)) < 0; - } -}; - class RuntimeDyldImpl { protected: + unsigned CPUType; + unsigned CPUSubtype; + // The MemoryManager to load objects into. RTDyldMemoryManager *MemMgr; - // A list of emmitted sections. - typedef SmallVector<SectionEntry, 64> SectionList; - SectionList Sections; + // For each section, we have a MemoryBlock of it's data. + // Indexed by SectionID. + SmallVector<sys::MemoryBlock, 32> Sections; + // For each section, the address it will be considered to live at for + // relocations. The same as the pointer to the above memory block for hosted + // JITs. Indexed by SectionID. + SmallVector<uint64_t, 32> SectionLoadAddress; - // Keep a map of sections from object file to the SectionID which - // references it. - typedef std::map<SectionRef, unsigned> ObjSectionToIDMap; + // Keep a map of starting local address to the SectionID which references it. + // Lookup function for when we assign virtual addresses. + DenseMap<void *, unsigned> SectionLocalMemToID; // Master symbol table. As modules are loaded and external symbols are // resolved, their addresses are stored here as a SectionID/Offset pair. - typedef std::pair<unsigned, uintptr_t> SymbolLoc; + typedef std::pair<unsigned, uint64_t> SymbolLoc; StringMap<SymbolLoc> SymbolTable; - typedef DenseMap<const char*, SymbolLoc> LocalSymbolMap; - - // For each symbol, keep a list of relocations based on it. Anytime - // its address is reassigned (the JIT re-compiled the function, e.g.), - // the relocations get re-resolved. - // The symbol (or section) the relocation is sourced from is the Key - // in the relocation list where it's stored. - typedef SmallVector<RelocationEntry, 64> RelocationList; - // Relocations to sections already loaded. Indexed by SectionID which is the - // source of the address. The target where the address will be writen is - // SectionID/Offset in the relocation itself. - DenseMap<unsigned, RelocationList> Relocations; - // Relocations to external symbols that are not yet resolved. - // Indexed by symbol name. - StringMap<RelocationList> SymbolRelocations; - - typedef std::map<RelocationValueRef, uintptr_t> StubMap; - - Triple::ArchType Arch; - - inline unsigned getMaxStubSize() { - if (Arch == Triple::arm || Arch == Triple::thumb) - return 8; // 32-bit instruction and 32-bit address - else - return 0; - } bool HasError; std::string ErrorStr; @@ -146,62 +66,17 @@ protected: } uint8_t *getSectionAddress(unsigned SectionID) { - return (uint8_t*)Sections[SectionID].Address; + return (uint8_t*)Sections[SectionID].base(); } + void extractFunction(StringRef Name, uint8_t *StartAddress, + uint8_t *EndAddress); - /// \brief Emits section data from the object file to the MemoryManager. - /// \param IsCode if it's true then allocateCodeSection() will be - /// used for emmits, else allocateDataSection() will be used. - /// \return SectionID. - unsigned emitSection(const SectionRef &Section, bool IsCode); - - /// \brief Find Section in LocalSections. If the secton is not found - emit - /// it and store in LocalSections. - /// \param IsCode if it's true then allocateCodeSection() will be - /// used for emmits, else allocateDataSection() will be used. - /// \return SectionID. - unsigned findOrEmitSection(const SectionRef &Section, bool IsCode, - ObjSectionToIDMap &LocalSections); - - /// \brief If Value.SymbolName is NULL then store relocation to the - /// Relocations, else store it in the SymbolRelocations. - void AddRelocation(const RelocationValueRef &Value, unsigned SectionID, - uintptr_t Offset, uint32_t RelType); - - /// \brief Emits long jump instruction to Addr. - /// \return Pointer to the memory area for emitting target address. - uint8_t* createStubFunction(uint8_t *Addr); - - /// \brief Resolves relocations from Relocs list with address from Value. - void resolveRelocationList(const RelocationList &Relocs, uint64_t Value); - void resolveRelocationEntry(const RelocationEntry &RE, uint64_t Value); - - /// \brief A object file specific relocation resolver - /// \param Address Address to apply the relocation action - /// \param Value Target symbol address to apply the relocation action - /// \param Type object file specific relocation type - /// \param Addend A constant addend used to compute the value to be stored - /// into the relocatable field - virtual void resolveRelocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend) = 0; - - /// \brief Parses the object file relocation and store it to Relocations - /// or SymbolRelocations. Its depend from object file type. - virtual void processRelocationRef(const ObjRelocationInfo &Rel, - const ObjectFile &Obj, - ObjSectionToIDMap &ObjSectionToID, - LocalSymbolMap &Symbols, StubMap &Stubs) = 0; - - void resolveSymbols(); public: RuntimeDyldImpl(RTDyldMemoryManager *mm) : MemMgr(mm), HasError(false) {} virtual ~RuntimeDyldImpl(); - bool loadObject(const MemoryBuffer *InputBuffer); + virtual bool loadObject(MemoryBuffer *InputBuffer) = 0; void *getSymbolAddress(StringRef Name) { // FIXME: Just look up as a function for now. Overly simple of course. @@ -212,9 +87,9 @@ public: return getSectionAddress(Loc.first) + Loc.second; } - void resolveRelocations(); + virtual void resolveRelocations(); - void reassignSectionAddress(unsigned SectionID, uint64_t Addr); + virtual void reassignSectionAddress(unsigned SectionID, uint64_t Addr) = 0; void mapSectionAddress(void *LocalAddress, uint64_t TargetAddress); @@ -228,7 +103,6 @@ public: StringRef getErrorString() { return ErrorStr; } virtual bool isCompatibleFormat(const MemoryBuffer *InputBuffer) const = 0; - }; } // end namespace llvm diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.cpp b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.cpp index 24437e0..0b72b56 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.cpp +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.cpp @@ -21,47 +21,33 @@ using namespace llvm::object; namespace llvm { -void RuntimeDyldMachO::resolveRelocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend) { - bool isPCRel = (Type >> 24) & 1; - unsigned MachoType = (Type >> 28) & 0xf; - unsigned Size = 1 << ((Type >> 25) & 3); - - DEBUG(dbgs() << "resolveRelocation LocalAddress: " << format("%p", LocalAddress) - << " FinalAddress: " << format("%p", FinalAddress) - << " Value: " << format("%p", Value) - << " Addend: " << Addend - << " isPCRel: " << isPCRel - << " MachoType: " << MachoType - << " Size: " << Size - << "\n"); - +bool RuntimeDyldMachO:: +resolveRelocation(uint8_t *LocalAddress, + uint64_t FinalAddress, + uint64_t Value, + bool isPCRel, + unsigned Type, + unsigned Size, + int64_t Addend) { // This just dispatches to the proper target specific routine. - switch (Arch) { + switch (CPUType) { default: llvm_unreachable("Unsupported CPU type!"); - case Triple::x86_64: // Fall through. - case Triple::x86: - resolveX86_64Relocation(LocalAddress, - FinalAddress, - (uintptr_t)Value, - isPCRel, - MachoType, - Size, - Addend); - break; - case Triple::arm: // Fall through. - case Triple::thumb: - resolveARMRelocation(LocalAddress, - FinalAddress, - (uintptr_t)Value, - isPCRel, - MachoType, - Size, - Addend); - break; + case mach::CTM_x86_64: + return resolveX86_64Relocation(LocalAddress, + FinalAddress, + (uintptr_t)Value, + isPCRel, + Type, + Size, + Addend); + case mach::CTM_ARM: + return resolveARMRelocation(LocalAddress, + FinalAddress, + (uintptr_t)Value, + isPCRel, + Type, + Size, + Addend); } } @@ -167,83 +153,503 @@ resolveARMRelocation(uint8_t *LocalAddress, return false; } -void RuntimeDyldMachO:: -processRelocationRef(const ObjRelocationInfo &Rel, const ObjectFile &Obj, - ObjSectionToIDMap &ObjSectionToID, - LocalSymbolMap &Symbols, StubMap &Stubs) { - - uint32_t RelType = (uint32_t) (Rel.Type & 0xffffffffL); - RelocationValueRef Value; - SectionEntry &Section = Sections[Rel.SectionID]; - uint8_t *Target = Section.Address + Rel.Offset; - - bool isExtern = (RelType >> 27) & 1; - if (isExtern) { - StringRef TargetName; - const SymbolRef &Symbol = Rel.Symbol; - Symbol.getName(TargetName); - // First look the symbol in object file symbols. - LocalSymbolMap::iterator it = Symbols.find(TargetName.data()); - if (it != Symbols.end()) { - Value.SectionID = it->second.first; - Value.Addend = it->second.second; - } else { - // Second look the symbol in global symbol table. - StringMap<SymbolLoc>::iterator itS = SymbolTable.find(TargetName.data()); - if (itS != SymbolTable.end()) { - Value.SectionID = itS->second.first; - Value.Addend = itS->second.second; - } else - Value.SymbolName = TargetName.data(); - } - } else { - error_code err; - uint8_t sIdx = static_cast<uint8_t>(RelType & 0xFF); - section_iterator sIt = Obj.begin_sections(), - sItEnd = Obj.end_sections(); - for (uint8_t i = 1; i < sIdx; i++) { - error_code err; - sIt.increment(err); - if (sIt == sItEnd) - break; +bool RuntimeDyldMachO:: +loadSegment32(const MachOObject *Obj, + const MachOObject::LoadCommandInfo *SegmentLCI, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC) { + // FIXME: This should really be combined w/ loadSegment64. Templatized + // function on the 32/64 datatypes maybe? + InMemoryStruct<macho::SegmentLoadCommand> SegmentLC; + Obj->ReadSegmentLoadCommand(*SegmentLCI, SegmentLC); + if (!SegmentLC) + return Error("unable to load segment load command"); + + + SmallVector<unsigned, 16> SectionMap; + for (unsigned SectNum = 0; SectNum != SegmentLC->NumSections; ++SectNum) { + InMemoryStruct<macho::Section> Sect; + Obj->ReadSection(*SegmentLCI, SectNum, Sect); + if (!Sect) + return Error("unable to load section: '" + Twine(SectNum) + "'"); + + // Allocate memory via the MM for the section. + uint8_t *Buffer; + uint32_t SectionID = Sections.size(); + if (Sect->Flags == 0x80000400) + Buffer = MemMgr->allocateCodeSection(Sect->Size, Sect->Align, SectionID); + else + Buffer = MemMgr->allocateDataSection(Sect->Size, Sect->Align, SectionID); + + DEBUG(dbgs() << "Loading " + << ((Sect->Flags == 0x80000400) ? "text" : "data") + << " (ID #" << SectionID << ")" + << " '" << Sect->SegmentName << "," + << Sect->Name << "' of size " << Sect->Size + << " to address " << Buffer << ".\n"); + + // Copy the payload from the object file into the allocated buffer. + uint8_t *Base = (uint8_t*)Obj->getData(SegmentLC->FileOffset, + SegmentLC->FileSize).data(); + memcpy(Buffer, Base + Sect->Address, Sect->Size); + + // Remember what got allocated for this SectionID. + Sections.push_back(sys::MemoryBlock(Buffer, Sect->Size)); + SectionLocalMemToID[Buffer] = SectionID; + + // By default, the load address of a section is its memory buffer. + SectionLoadAddress.push_back((uint64_t)Buffer); + + // Keep a map of object file section numbers to corresponding SectionIDs + // while processing the file. + SectionMap.push_back(SectionID); + } + + // Process the symbol table. + SmallVector<StringRef, 64> SymbolNames; + processSymbols32(Obj, SectionMap, SymbolNames, SymtabLC); + + // Process the relocations for each section we're loading. + Relocations.grow(Relocations.size() + SegmentLC->NumSections); + Referrers.grow(Referrers.size() + SegmentLC->NumSections); + for (unsigned SectNum = 0; SectNum != SegmentLC->NumSections; ++SectNum) { + InMemoryStruct<macho::Section> Sect; + Obj->ReadSection(*SegmentLCI, SectNum, Sect); + if (!Sect) + return Error("unable to load section: '" + Twine(SectNum) + "'"); + for (unsigned j = 0; j != Sect->NumRelocationTableEntries; ++j) { + InMemoryStruct<macho::RelocationEntry> RE; + Obj->ReadRelocationEntry(Sect->RelocationTableOffset, j, RE); + if (RE->Word0 & macho::RF_Scattered) + return Error("NOT YET IMPLEMENTED: scattered relocations."); + // Word0 of the relocation is the offset into the section where the + // relocation should be applied. We need to translate that into an + // offset into a function since that's our atom. + uint32_t Offset = RE->Word0; + bool isExtern = (RE->Word1 >> 27) & 1; + + // FIXME: Get the relocation addend from the target address. + // FIXME: VERY imporant for internal relocations. + + // Figure out the source symbol of the relocation. If isExtern is true, + // this relocation references the symbol table, otherwise it references + // a section in the same object, numbered from 1 through NumSections + // (SectionBases is [0, NumSections-1]). + uint32_t SourceNum = RE->Word1 & 0xffffff; // 24-bit value + if (!isExtern) { + assert(SourceNum > 0 && "Invalid relocation section number!"); + unsigned SectionID = SectionMap[SourceNum - 1]; + unsigned TargetID = SectionMap[SectNum]; + DEBUG(dbgs() << "Internal relocation at Section #" + << TargetID << " + " << Offset + << " from Section #" + << SectionID << " (Word1: " + << format("0x%x", RE->Word1) << ")\n"); + + // Store the relocation information. It will get resolved when + // the section addresses are assigned. + uint32_t RelocationIndex = Relocations[SectionID].size(); + Relocations[SectionID].push_back(RelocationEntry(TargetID, + Offset, + RE->Word1, + 0 /*Addend*/)); + Referrers[TargetID].push_back(Referrer(SectionID, RelocationIndex)); + } else { + StringRef SourceName = SymbolNames[SourceNum]; + + // Now store the relocation information. Associate it with the source + // symbol. Just add it to the unresolved list and let the general + // path post-load resolve it if we know where the symbol is. + UnresolvedRelocations[SourceName].push_back(RelocationEntry(SectNum, + Offset, + RE->Word1, + 0 /*Addend*/)); + DEBUG(dbgs() << "Relocation at Section #" << SectNum << " + " << Offset + << " from '" << SourceName << "(Word1: " + << format("0x%x", RE->Word1) << ")\n"); + } } - assert(sIt != sItEnd && "No section containing relocation!"); - Value.SectionID = findOrEmitSection(*sIt, true, ObjSectionToID); - Value.Addend = *(const intptr_t *)Target; - if (Value.Addend) { - // The MachO addend is offset from the current section, we need set it - // as offset from destination section - Value.Addend += Section.ObjAddress - Sections[Value.SectionID].ObjAddress; + } + + // Resolve the addresses of any symbols that were defined in this segment. + for (int i = 0, e = SymbolNames.size(); i != e; ++i) + resolveSymbol(SymbolNames[i]); + + return false; +} + + +bool RuntimeDyldMachO:: +loadSegment64(const MachOObject *Obj, + const MachOObject::LoadCommandInfo *SegmentLCI, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC) { + InMemoryStruct<macho::Segment64LoadCommand> Segment64LC; + Obj->ReadSegment64LoadCommand(*SegmentLCI, Segment64LC); + if (!Segment64LC) + return Error("unable to load segment load command"); + + + SmallVector<unsigned, 16> SectionMap; + for (unsigned SectNum = 0; SectNum != Segment64LC->NumSections; ++SectNum) { + InMemoryStruct<macho::Section64> Sect; + Obj->ReadSection64(*SegmentLCI, SectNum, Sect); + if (!Sect) + return Error("unable to load section: '" + Twine(SectNum) + "'"); + + // Allocate memory via the MM for the section. + uint8_t *Buffer; + uint32_t SectionID = Sections.size(); + unsigned Align = 1 << Sect->Align; // .o file has log2 alignment. + if (Sect->Flags == 0x80000400) + Buffer = MemMgr->allocateCodeSection(Sect->Size, Align, SectionID); + else + Buffer = MemMgr->allocateDataSection(Sect->Size, Align, SectionID); + + DEBUG(dbgs() << "Loading " + << ((Sect->Flags == 0x80000400) ? "text" : "data") + << " (ID #" << SectionID << ")" + << " '" << Sect->SegmentName << "," + << Sect->Name << "' of size " << Sect->Size + << " (align " << Align << ")" + << " to address " << Buffer << ".\n"); + + // Copy the payload from the object file into the allocated buffer. + uint8_t *Base = (uint8_t*)Obj->getData(Segment64LC->FileOffset, + Segment64LC->FileSize).data(); + memcpy(Buffer, Base + Sect->Address, Sect->Size); + + // Remember what got allocated for this SectionID. + Sections.push_back(sys::MemoryBlock(Buffer, Sect->Size)); + SectionLocalMemToID[Buffer] = SectionID; + + // By default, the load address of a section is its memory buffer. + SectionLoadAddress.push_back((uint64_t)Buffer); + + // Keep a map of object file section numbers to corresponding SectionIDs + // while processing the file. + SectionMap.push_back(SectionID); + } + + // Process the symbol table. + SmallVector<StringRef, 64> SymbolNames; + processSymbols64(Obj, SectionMap, SymbolNames, SymtabLC); + + // Process the relocations for each section we're loading. + Relocations.grow(Relocations.size() + Segment64LC->NumSections); + Referrers.grow(Referrers.size() + Segment64LC->NumSections); + for (unsigned SectNum = 0; SectNum != Segment64LC->NumSections; ++SectNum) { + InMemoryStruct<macho::Section64> Sect; + Obj->ReadSection64(*SegmentLCI, SectNum, Sect); + if (!Sect) + return Error("unable to load section: '" + Twine(SectNum) + "'"); + for (unsigned j = 0; j != Sect->NumRelocationTableEntries; ++j) { + InMemoryStruct<macho::RelocationEntry> RE; + Obj->ReadRelocationEntry(Sect->RelocationTableOffset, j, RE); + if (RE->Word0 & macho::RF_Scattered) + return Error("NOT YET IMPLEMENTED: scattered relocations."); + // Word0 of the relocation is the offset into the section where the + // relocation should be applied. We need to translate that into an + // offset into a function since that's our atom. + uint32_t Offset = RE->Word0; + bool isExtern = (RE->Word1 >> 27) & 1; + + // FIXME: Get the relocation addend from the target address. + // FIXME: VERY imporant for internal relocations. + + // Figure out the source symbol of the relocation. If isExtern is true, + // this relocation references the symbol table, otherwise it references + // a section in the same object, numbered from 1 through NumSections + // (SectionBases is [0, NumSections-1]). + uint32_t SourceNum = RE->Word1 & 0xffffff; // 24-bit value + if (!isExtern) { + assert(SourceNum > 0 && "Invalid relocation section number!"); + unsigned SectionID = SectionMap[SourceNum - 1]; + unsigned TargetID = SectionMap[SectNum]; + DEBUG(dbgs() << "Internal relocation at Section #" + << TargetID << " + " << Offset + << " from Section #" + << SectionID << " (Word1: " + << format("0x%x", RE->Word1) << ")\n"); + + // Store the relocation information. It will get resolved when + // the section addresses are assigned. + uint32_t RelocationIndex = Relocations[SectionID].size(); + Relocations[SectionID].push_back(RelocationEntry(TargetID, + Offset, + RE->Word1, + 0 /*Addend*/)); + Referrers[TargetID].push_back(Referrer(SectionID, RelocationIndex)); + } else { + StringRef SourceName = SymbolNames[SourceNum]; + + // Now store the relocation information. Associate it with the source + // symbol. Just add it to the unresolved list and let the general + // path post-load resolve it if we know where the symbol is. + UnresolvedRelocations[SourceName].push_back(RelocationEntry(SectNum, + Offset, + RE->Word1, + 0 /*Addend*/)); + DEBUG(dbgs() << "Relocation at Section #" << SectNum << " + " << Offset + << " from '" << SourceName << "(Word1: " + << format("0x%x", RE->Word1) << ")\n"); + } } } - if (Arch == Triple::arm && RelType == macho::RIT_ARM_Branch24Bit) { - // This is an ARM branch relocation, need to use a stub function. - - // Look up for existing stub. - StubMap::const_iterator stubIt = Stubs.find(Value); - if (stubIt != Stubs.end()) - resolveRelocation(Target, (uint64_t)Target, - (uint64_t)Section.Address + stubIt->second, - RelType, 0); - else { - // Create a new stub function. - Stubs[Value] = Section.StubOffset; - uint8_t *StubTargetAddr = createStubFunction(Section.Address + - Section.StubOffset); - AddRelocation(Value, Rel.SectionID, StubTargetAddr - Section.Address, - macho::RIT_Vanilla); - resolveRelocation(Target, (uint64_t)Target, - (uint64_t)Section.Address + Section.StubOffset, - RelType, 0); - Section.StubOffset += getMaxStubSize(); + // Resolve the addresses of any symbols that were defined in this segment. + for (int i = 0, e = SymbolNames.size(); i != e; ++i) + resolveSymbol(SymbolNames[i]); + + return false; +} + +bool RuntimeDyldMachO:: +processSymbols32(const MachOObject *Obj, + SmallVectorImpl<unsigned> &SectionMap, + SmallVectorImpl<StringRef> &SymbolNames, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC) { + // FIXME: Combine w/ processSymbols64. Factor 64/32 datatype and such. + for (unsigned i = 0; i != SymtabLC->NumSymbolTableEntries; ++i) { + InMemoryStruct<macho::SymbolTableEntry> STE; + Obj->ReadSymbolTableEntry(SymtabLC->SymbolTableOffset, i, STE); + if (!STE) + return Error("unable to read symbol: '" + Twine(i) + "'"); + // Get the symbol name. + StringRef Name = Obj->getStringAtIndex(STE->StringIndex); + SymbolNames.push_back(Name); + + // FIXME: Check the symbol type and flags. + if (STE->Type != 0xF) // external, defined in this segment. + continue; + // Flags in the upper nibble we don't care about. + if ((STE->Flags & 0xf) != 0x0) + continue; + + // Remember the symbol. + uint32_t SectionID = SectionMap[STE->SectionIndex - 1]; + SymbolTable[Name] = SymbolLoc(SectionID, STE->Value); + + DEBUG(dbgs() << "Symbol: '" << Name << "' @ " + << (getSectionAddress(SectionID) + STE->Value) + << "\n"); + } + return false; +} + +bool RuntimeDyldMachO:: +processSymbols64(const MachOObject *Obj, + SmallVectorImpl<unsigned> &SectionMap, + SmallVectorImpl<StringRef> &SymbolNames, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC) { + for (unsigned i = 0; i != SymtabLC->NumSymbolTableEntries; ++i) { + InMemoryStruct<macho::Symbol64TableEntry> STE; + Obj->ReadSymbol64TableEntry(SymtabLC->SymbolTableOffset, i, STE); + if (!STE) + return Error("unable to read symbol: '" + Twine(i) + "'"); + // Get the symbol name. + StringRef Name = Obj->getStringAtIndex(STE->StringIndex); + SymbolNames.push_back(Name); + + // FIXME: Check the symbol type and flags. + if (STE->Type != 0xF) // external, defined in this segment. + continue; + // Flags in the upper nibble we don't care about. + if ((STE->Flags & 0xf) != 0x0) + continue; + + // Remember the symbol. + uint32_t SectionID = SectionMap[STE->SectionIndex - 1]; + SymbolTable[Name] = SymbolLoc(SectionID, STE->Value); + + DEBUG(dbgs() << "Symbol: '" << Name << "' @ " + << (getSectionAddress(SectionID) + STE->Value) + << "\n"); + } + return false; +} + +// resolveSymbol - Resolve any relocations to the specified symbol if +// we know where it lives. +void RuntimeDyldMachO::resolveSymbol(StringRef Name) { + StringMap<SymbolLoc>::const_iterator Loc = SymbolTable.find(Name); + if (Loc == SymbolTable.end()) + return; + + RelocationList &Relocs = UnresolvedRelocations[Name]; + DEBUG(dbgs() << "Resolving symbol '" << Name << "'\n"); + for (int i = 0, e = Relocs.size(); i != e; ++i) { + // Change the relocation to be section relative rather than symbol + // relative and move it to the resolved relocation list. + RelocationEntry Entry = Relocs[i]; + Entry.Addend += Loc->second.second; + uint32_t RelocationIndex = Relocations[Loc->second.first].size(); + Relocations[Loc->second.first].push_back(Entry); + Referrers[Entry.SectionID].push_back(Referrer(Loc->second.first, RelocationIndex)); + } + // FIXME: Keep a worklist of the relocations we've added so that we can + // resolve more selectively later. + Relocs.clear(); +} + +bool RuntimeDyldMachO::loadObject(MemoryBuffer *InputBuffer) { + // If the linker is in an error state, don't do anything. + if (hasError()) + return true; + // Load the Mach-O wrapper object. + std::string ErrorStr; + OwningPtr<MachOObject> Obj( + MachOObject::LoadFromBuffer(InputBuffer, &ErrorStr)); + if (!Obj) + return Error("unable to load object: '" + ErrorStr + "'"); + + // Get the CPU type information from the header. + const macho::Header &Header = Obj->getHeader(); + + // FIXME: Error checking that the loaded object is compatible with + // the system we're running on. + CPUType = Header.CPUType; + CPUSubtype = Header.CPUSubtype; + + // Validate that the load commands match what we expect. + const MachOObject::LoadCommandInfo *SegmentLCI = 0, *SymtabLCI = 0, + *DysymtabLCI = 0; + for (unsigned i = 0; i != Header.NumLoadCommands; ++i) { + const MachOObject::LoadCommandInfo &LCI = Obj->getLoadCommandInfo(i); + switch (LCI.Command.Type) { + case macho::LCT_Segment: + case macho::LCT_Segment64: + if (SegmentLCI) + return Error("unexpected input object (multiple segments)"); + SegmentLCI = &LCI; + break; + case macho::LCT_Symtab: + if (SymtabLCI) + return Error("unexpected input object (multiple symbol tables)"); + SymtabLCI = &LCI; + break; + case macho::LCT_Dysymtab: + if (DysymtabLCI) + return Error("unexpected input object (multiple symbol tables)"); + DysymtabLCI = &LCI; + break; + default: + return Error("unexpected input object (unexpected load command"); } - } else - AddRelocation(Value, Rel.SectionID, Rel.Offset, RelType); + } + + if (!SymtabLCI) + return Error("no symbol table found in object"); + if (!SegmentLCI) + return Error("no segments found in object"); + + // Read and register the symbol table data. + InMemoryStruct<macho::SymtabLoadCommand> SymtabLC; + Obj->ReadSymtabLoadCommand(*SymtabLCI, SymtabLC); + if (!SymtabLC) + return Error("unable to load symbol table load command"); + Obj->RegisterStringTable(*SymtabLC); + + // Read the dynamic link-edit information, if present (not present in static + // objects). + if (DysymtabLCI) { + InMemoryStruct<macho::DysymtabLoadCommand> DysymtabLC; + Obj->ReadDysymtabLoadCommand(*DysymtabLCI, DysymtabLC); + if (!DysymtabLC) + return Error("unable to load dynamic link-exit load command"); + + // FIXME: We don't support anything interesting yet. +// if (DysymtabLC->LocalSymbolsIndex != 0) +// return Error("NOT YET IMPLEMENTED: local symbol entries"); +// if (DysymtabLC->ExternalSymbolsIndex != 0) +// return Error("NOT YET IMPLEMENTED: non-external symbol entries"); +// if (DysymtabLC->UndefinedSymbolsIndex != SymtabLC->NumSymbolTableEntries) +// return Error("NOT YET IMPLEMENTED: undefined symbol entries"); + } + + // Load the segment load command. + if (SegmentLCI->Command.Type == macho::LCT_Segment) { + if (loadSegment32(Obj.get(), SegmentLCI, SymtabLC)) + return true; + } else { + if (loadSegment64(Obj.get(), SegmentLCI, SymtabLC)) + return true; + } + + // Assign the addresses of the sections from the object so that any + // relocations to them get set properly. + // FIXME: This is done directly from the client at the moment. We should + // default the values to the local storage, at least when the target arch + // is the same as the host arch. + + return false; } +// Assign an address to a symbol name and resolve all the relocations +// associated with it. +void RuntimeDyldMachO::reassignSectionAddress(unsigned SectionID, + uint64_t Addr) { + // The address to use for relocation resolution is not + // the address of the local section buffer. We must be doing + // a remote execution environment of some sort. Re-apply any + // relocations referencing this section with the given address. + // + // Addr is a uint64_t because we can't assume the pointer width + // of the target is the same as that of the host. Just use a generic + // "big enough" type. + + SectionLoadAddress[SectionID] = Addr; + + RelocationList &Relocs = Relocations[SectionID]; + for (unsigned i = 0, e = Relocs.size(); i != e; ++i) { + RelocationEntry &RE = Relocs[i]; + uint8_t *Target = (uint8_t*)Sections[RE.SectionID].base() + RE.Offset; + uint64_t FinalTarget = (uint64_t)SectionLoadAddress[RE.SectionID] + RE.Offset; + bool isPCRel = (RE.Data >> 24) & 1; + unsigned Type = (RE.Data >> 28) & 0xf; + unsigned Size = 1 << ((RE.Data >> 25) & 3); + + DEBUG(dbgs() << "Resolving relocation at Section #" << RE.SectionID + << " + " << RE.Offset << " (" << format("%p", Target) << ")" + << " from Section #" << SectionID << " (" << format("%p", Addr) << ")" + << "(" << (isPCRel ? "pcrel" : "absolute") + << ", type: " << Type << ", Size: " << Size << ", Addend: " + << RE.Addend << ").\n"); + + resolveRelocation(Target, + FinalTarget, + Addr, + isPCRel, + Type, + Size, + RE.Addend); + } + ReferrerList &Refers = Referrers[SectionID]; + for (unsigned i = 0, e = Refers.size(); i != e; ++i) { + Referrer &R = Refers[i]; + RelocationEntry &RE = Relocations[R.SectionID][R.Index]; + uint8_t *Target = (uint8_t*)Sections[RE.SectionID].base() + RE.Offset; + uint64_t FinalTarget = (uint64_t)SectionLoadAddress[RE.SectionID] + RE.Offset; + bool isPCRel = (RE.Data >> 24) & 1; + unsigned Type = (RE.Data >> 28) & 0xf; + unsigned Size = 1 << ((RE.Data >> 25) & 3); + + DEBUG(dbgs() << "Resolving relocation at Section #" << RE.SectionID + << " + " << RE.Offset << " (" << format("%p", Target) << ")" + << " from Section #" << SectionID << " (" << format("%p", Addr) << ")" + << "(" << (isPCRel ? "pcrel" : "absolute") + << ", type: " << Type << ", Size: " << Size << ", Addend: " + << RE.Addend << ").\n"); + + resolveRelocation(Target, + FinalTarget, + Addr, + isPCRel, + Type, + Size, + RE.Addend); + } +} -bool RuntimeDyldMachO::isCompatibleFormat(const MemoryBuffer *InputBuffer) const { +bool RuntimeDyldMachO::isKnownFormat(const MemoryBuffer *InputBuffer) { StringRef Magic = InputBuffer->getBuffer().slice(0, 4); if (Magic == "\xFE\xED\xFA\xCE") return true; if (Magic == "\xCE\xFA\xED\xFE") return true; diff --git a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h index 36b39dd..6721703 100644 --- a/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h +++ b/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldMachO.h @@ -25,7 +25,55 @@ using namespace llvm::object; namespace llvm { class RuntimeDyldMachO : public RuntimeDyldImpl { -protected: + + // For each symbol, keep a list of relocations based on it. Anytime + // its address is reassigned (the JIT re-compiled the function, e.g.), + // the relocations get re-resolved. + // The symbol (or section) the relocation is sourced from is the Key + // in the relocation list where it's stored. + struct RelocationEntry { + unsigned SectionID; // Section the relocation is contained in. + uint64_t Offset; // Offset into the section for the relocation. + uint32_t Data; // Second word of the raw macho relocation entry. + int64_t Addend; // Addend encoded in the instruction itself, if any, + // plus the offset into the source section for + // the symbol once the relocation is resolvable. + + RelocationEntry(unsigned id, uint64_t offset, uint32_t data, int64_t addend) + : SectionID(id), Offset(offset), Data(data), Addend(addend) {} + }; + typedef SmallVector<RelocationEntry, 4> RelocationList; + + // For each section, keep a list of referrers in that section that are clients + // of relocations in other sections. Whenever a relocation gets created, + // create a corresponding referrer. Whenever relocations are re-resolved, + // re-resolve the referrers' relocations as well. + struct Referrer { + unsigned SectionID; // Section whose RelocationList contains the relocation. + uint32_t Index; // Index of the RelocatonEntry in that RelocationList. + + Referrer(unsigned id, uint32_t index) + : SectionID(id), Index(index) {} + }; + typedef SmallVector<Referrer, 4> ReferrerList; + + // Relocations to sections already loaded. Indexed by SectionID which is the + // source of the address. The target where the address will be writen is + // SectionID/Offset in the relocation itself. + IndexedMap<RelocationList> Relocations; + // Referrers corresponding to Relocations. + IndexedMap<ReferrerList> Referrers; + // Relocations to symbols that are not yet resolved. Must be external + // relocations by definition. Indexed by symbol name. + StringMap<RelocationList> UnresolvedRelocations; + + bool resolveRelocation(uint8_t *LocalAddress, + uint64_t FinalAddress, + uint64_t Value, + bool isPCRel, + unsigned Type, + unsigned Size, + int64_t Addend); bool resolveX86_64Relocation(uint8_t *LocalAddress, uint64_t FinalAddress, uint64_t Value, @@ -41,21 +89,35 @@ protected: unsigned Size, int64_t Addend); - virtual void processRelocationRef(const ObjRelocationInfo &Rel, - const ObjectFile &Obj, - ObjSectionToIDMap &ObjSectionToID, - LocalSymbolMap &Symbols, StubMap &Stubs); + bool loadSegment32(const MachOObject *Obj, + const MachOObject::LoadCommandInfo *SegmentLCI, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC); + bool loadSegment64(const MachOObject *Obj, + const MachOObject::LoadCommandInfo *SegmentLCI, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC); + bool processSymbols32(const MachOObject *Obj, + SmallVectorImpl<unsigned> &SectionMap, + SmallVectorImpl<StringRef> &SymbolNames, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC); + bool processSymbols64(const MachOObject *Obj, + SmallVectorImpl<unsigned> &SectionMap, + SmallVectorImpl<StringRef> &SymbolNames, + const InMemoryStruct<macho::SymtabLoadCommand> &SymtabLC); + + void resolveSymbol(StringRef Name); public: - virtual void resolveRelocation(uint8_t *LocalAddress, - uint64_t FinalAddress, - uint64_t Value, - uint32_t Type, - int64_t Addend); - RuntimeDyldMachO(RTDyldMemoryManager *mm) : RuntimeDyldImpl(mm) {} - bool isCompatibleFormat(const MemoryBuffer *InputBuffer) const; + bool loadObject(MemoryBuffer *InputBuffer); + + void reassignSectionAddress(unsigned SectionID, uint64_t Addr); + + static bool isKnownFormat(const MemoryBuffer *InputBuffer); + + bool isCompatibleFormat(const MemoryBuffer *InputBuffer) const { + return isKnownFormat(InputBuffer); + } }; } // end namespace llvm |