forked from OSchip/llvm-project
1278 lines
61 KiB
C++
1278 lines
61 KiB
C++
//===-- ABISysV_x86_64.cpp --------------------------------------*- C++ -*-===//
|
|
//
|
|
// The LLVM Compiler Infrastructure
|
|
//
|
|
// This file is distributed under the University of Illinois Open Source
|
|
// License. See LICENSE.TXT for details.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#include "ABISysV_x86_64.h"
|
|
|
|
#include "lldb/Core/ConstString.h"
|
|
#include "lldb/Core/DataExtractor.h"
|
|
#include "lldb/Core/Error.h"
|
|
#include "lldb/Core/Log.h"
|
|
#include "lldb/Core/Module.h"
|
|
#include "lldb/Core/PluginManager.h"
|
|
#include "lldb/Core/RegisterValue.h"
|
|
#include "lldb/Core/Value.h"
|
|
#include "lldb/Core/ValueObjectConstResult.h"
|
|
#include "lldb/Core/ValueObjectRegister.h"
|
|
#include "lldb/Core/ValueObjectMemory.h"
|
|
#include "lldb/Symbol/ClangASTContext.h"
|
|
#include "lldb/Symbol/UnwindPlan.h"
|
|
#include "lldb/Target/Target.h"
|
|
#include "lldb/Target/Process.h"
|
|
#include "lldb/Target/RegisterContext.h"
|
|
#include "lldb/Target/StackFrame.h"
|
|
#include "lldb/Target/Thread.h"
|
|
|
|
#include "llvm/ADT/STLExtras.h"
|
|
#include "llvm/ADT/Triple.h"
|
|
|
|
using namespace lldb;
|
|
using namespace lldb_private;
|
|
|
|
enum gcc_dwarf_regnums
|
|
{
|
|
gcc_dwarf_rax = 0,
|
|
gcc_dwarf_rdx,
|
|
gcc_dwarf_rcx,
|
|
gcc_dwarf_rbx,
|
|
gcc_dwarf_rsi,
|
|
gcc_dwarf_rdi,
|
|
gcc_dwarf_rbp,
|
|
gcc_dwarf_rsp,
|
|
gcc_dwarf_r8,
|
|
gcc_dwarf_r9,
|
|
gcc_dwarf_r10,
|
|
gcc_dwarf_r11,
|
|
gcc_dwarf_r12,
|
|
gcc_dwarf_r13,
|
|
gcc_dwarf_r14,
|
|
gcc_dwarf_r15,
|
|
gcc_dwarf_rip,
|
|
gcc_dwarf_xmm0,
|
|
gcc_dwarf_xmm1,
|
|
gcc_dwarf_xmm2,
|
|
gcc_dwarf_xmm3,
|
|
gcc_dwarf_xmm4,
|
|
gcc_dwarf_xmm5,
|
|
gcc_dwarf_xmm6,
|
|
gcc_dwarf_xmm7,
|
|
gcc_dwarf_xmm8,
|
|
gcc_dwarf_xmm9,
|
|
gcc_dwarf_xmm10,
|
|
gcc_dwarf_xmm11,
|
|
gcc_dwarf_xmm12,
|
|
gcc_dwarf_xmm13,
|
|
gcc_dwarf_xmm14,
|
|
gcc_dwarf_xmm15,
|
|
gcc_dwarf_stmm0,
|
|
gcc_dwarf_stmm1,
|
|
gcc_dwarf_stmm2,
|
|
gcc_dwarf_stmm3,
|
|
gcc_dwarf_stmm4,
|
|
gcc_dwarf_stmm5,
|
|
gcc_dwarf_stmm6,
|
|
gcc_dwarf_stmm7,
|
|
gcc_dwarf_ymm0,
|
|
gcc_dwarf_ymm1,
|
|
gcc_dwarf_ymm2,
|
|
gcc_dwarf_ymm3,
|
|
gcc_dwarf_ymm4,
|
|
gcc_dwarf_ymm5,
|
|
gcc_dwarf_ymm6,
|
|
gcc_dwarf_ymm7,
|
|
gcc_dwarf_ymm8,
|
|
gcc_dwarf_ymm9,
|
|
gcc_dwarf_ymm10,
|
|
gcc_dwarf_ymm11,
|
|
gcc_dwarf_ymm12,
|
|
gcc_dwarf_ymm13,
|
|
gcc_dwarf_ymm14,
|
|
gcc_dwarf_ymm15
|
|
};
|
|
|
|
enum gdb_regnums
|
|
{
|
|
gdb_rax = 0,
|
|
gdb_rbx = 1,
|
|
gdb_rcx = 2,
|
|
gdb_rdx = 3,
|
|
gdb_rsi = 4,
|
|
gdb_rdi = 5,
|
|
gdb_rbp = 6,
|
|
gdb_rsp = 7,
|
|
gdb_r8 = 8,
|
|
gdb_r9 = 9,
|
|
gdb_r10 = 10,
|
|
gdb_r11 = 11,
|
|
gdb_r12 = 12,
|
|
gdb_r13 = 13,
|
|
gdb_r14 = 14,
|
|
gdb_r15 = 15,
|
|
gdb_rip = 16,
|
|
gdb_rflags = 17,
|
|
gdb_cs = 18,
|
|
gdb_ss = 19,
|
|
gdb_ds = 20,
|
|
gdb_es = 21,
|
|
gdb_fs = 22,
|
|
gdb_gs = 23,
|
|
gdb_stmm0 = 24,
|
|
gdb_stmm1 = 25,
|
|
gdb_stmm2 = 26,
|
|
gdb_stmm3 = 27,
|
|
gdb_stmm4 = 28,
|
|
gdb_stmm5 = 29,
|
|
gdb_stmm6 = 30,
|
|
gdb_stmm7 = 31,
|
|
gdb_fctrl = 32, gdb_fcw = gdb_fctrl,
|
|
gdb_fstat = 33, gdb_fsw = gdb_fstat,
|
|
gdb_ftag = 34, gdb_ftw = gdb_ftag,
|
|
gdb_fiseg = 35, gdb_fpu_cs = gdb_fiseg,
|
|
gdb_fioff = 36, gdb_ip = gdb_fioff,
|
|
gdb_foseg = 37, gdb_fpu_ds = gdb_foseg,
|
|
gdb_fooff = 38, gdb_dp = gdb_fooff,
|
|
gdb_fop = 39,
|
|
gdb_xmm0 = 40,
|
|
gdb_xmm1 = 41,
|
|
gdb_xmm2 = 42,
|
|
gdb_xmm3 = 43,
|
|
gdb_xmm4 = 44,
|
|
gdb_xmm5 = 45,
|
|
gdb_xmm6 = 46,
|
|
gdb_xmm7 = 47,
|
|
gdb_xmm8 = 48,
|
|
gdb_xmm9 = 49,
|
|
gdb_xmm10 = 50,
|
|
gdb_xmm11 = 51,
|
|
gdb_xmm12 = 52,
|
|
gdb_xmm13 = 53,
|
|
gdb_xmm14 = 54,
|
|
gdb_xmm15 = 55,
|
|
gdb_mxcsr = 56,
|
|
gdb_ymm0 = 57,
|
|
gdb_ymm1 = 58,
|
|
gdb_ymm2 = 59,
|
|
gdb_ymm3 = 60,
|
|
gdb_ymm4 = 61,
|
|
gdb_ymm5 = 62,
|
|
gdb_ymm6 = 63,
|
|
gdb_ymm7 = 64,
|
|
gdb_ymm8 = 65,
|
|
gdb_ymm9 = 66,
|
|
gdb_ymm10 = 67,
|
|
gdb_ymm11 = 68,
|
|
gdb_ymm12 = 69,
|
|
gdb_ymm13 = 70,
|
|
gdb_ymm14 = 71,
|
|
gdb_ymm15 = 72
|
|
};
|
|
|
|
|
|
static RegisterInfo g_register_infos[] =
|
|
{
|
|
// NAME ALT SZ OFF ENCODING FORMAT EH_FRAME DWARF GENERIC STABS LLDB NATIVE VALUE REGS INVALIDATE REGS
|
|
// ======== ======= == === ============= =================== ======================= ===================== =========================== ===================== ====================== ========== ===============
|
|
{ "rax" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rax , gcc_dwarf_rax , LLDB_INVALID_REGNUM , gdb_rax , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rbx" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rbx , gcc_dwarf_rbx , LLDB_INVALID_REGNUM , gdb_rbx , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rcx" , "arg4", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rcx , gcc_dwarf_rcx , LLDB_REGNUM_GENERIC_ARG4 , gdb_rcx , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rdx" , "arg3", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rdx , gcc_dwarf_rdx , LLDB_REGNUM_GENERIC_ARG3 , gdb_rdx , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rsi" , "arg2", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rsi , gcc_dwarf_rsi , LLDB_REGNUM_GENERIC_ARG2 , gdb_rsi , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rdi" , "arg1", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rdi , gcc_dwarf_rdi , LLDB_REGNUM_GENERIC_ARG1 , gdb_rdi , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rbp" , "fp", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rbp , gcc_dwarf_rbp , LLDB_REGNUM_GENERIC_FP , gdb_rbp , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rsp" , "sp", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rsp , gcc_dwarf_rsp , LLDB_REGNUM_GENERIC_SP , gdb_rsp , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r8" , "arg5", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r8 , gcc_dwarf_r8 , LLDB_REGNUM_GENERIC_ARG5 , gdb_r8 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r9" , "arg6", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r9 , gcc_dwarf_r9 , LLDB_REGNUM_GENERIC_ARG6 , gdb_r9 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r10" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r10 , gcc_dwarf_r10 , LLDB_INVALID_REGNUM , gdb_r10 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r11" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r11 , gcc_dwarf_r11 , LLDB_INVALID_REGNUM , gdb_r11 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r12" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r12 , gcc_dwarf_r12 , LLDB_INVALID_REGNUM , gdb_r12 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r13" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r13 , gcc_dwarf_r13 , LLDB_INVALID_REGNUM , gdb_r13 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r14" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r14 , gcc_dwarf_r14 , LLDB_INVALID_REGNUM , gdb_r14 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "r15" , NULL, 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_r15 , gcc_dwarf_r15 , LLDB_INVALID_REGNUM , gdb_r15 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rip" , "pc", 8, 0, eEncodingUint , eFormatHex , { gcc_dwarf_rip , gcc_dwarf_rip , LLDB_REGNUM_GENERIC_PC , gdb_rip , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "rflags", NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_REGNUM_GENERIC_FLAGS , gdb_rflags , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "cs" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_cs , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ss" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_ss , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ds" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_ds , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "es" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_es , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fs" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fs , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "gs" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_gs , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm0" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm0 , gcc_dwarf_stmm0 , LLDB_INVALID_REGNUM , gdb_stmm0 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm1" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm1 , gcc_dwarf_stmm1 , LLDB_INVALID_REGNUM , gdb_stmm1 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm2" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm2 , gcc_dwarf_stmm2 , LLDB_INVALID_REGNUM , gdb_stmm2 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm3" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm3 , gcc_dwarf_stmm3 , LLDB_INVALID_REGNUM , gdb_stmm3 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm4" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm4 , gcc_dwarf_stmm4 , LLDB_INVALID_REGNUM , gdb_stmm4 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm5" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm5 , gcc_dwarf_stmm5 , LLDB_INVALID_REGNUM , gdb_stmm5 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm6" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm6 , gcc_dwarf_stmm6 , LLDB_INVALID_REGNUM , gdb_stmm6 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "stmm7" , NULL, 10, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_stmm7 , gcc_dwarf_stmm7 , LLDB_INVALID_REGNUM , gdb_stmm7 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fctrl" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fctrl , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fstat" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fstat , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ftag" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_ftag , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fiseg" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fiseg , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fioff" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fioff , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "foseg" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_foseg , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fooff" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fooff , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "fop" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_fop , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm0" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm0 , gcc_dwarf_xmm0 , LLDB_INVALID_REGNUM , gdb_xmm0 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm1" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm1 , gcc_dwarf_xmm1 , LLDB_INVALID_REGNUM , gdb_xmm1 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm2" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm2 , gcc_dwarf_xmm2 , LLDB_INVALID_REGNUM , gdb_xmm2 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm3" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm3 , gcc_dwarf_xmm3 , LLDB_INVALID_REGNUM , gdb_xmm3 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm4" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm4 , gcc_dwarf_xmm4 , LLDB_INVALID_REGNUM , gdb_xmm4 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm5" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm5 , gcc_dwarf_xmm5 , LLDB_INVALID_REGNUM , gdb_xmm5 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm6" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm6 , gcc_dwarf_xmm6 , LLDB_INVALID_REGNUM , gdb_xmm6 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm7" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm7 , gcc_dwarf_xmm7 , LLDB_INVALID_REGNUM , gdb_xmm7 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm8" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm8 , gcc_dwarf_xmm8 , LLDB_INVALID_REGNUM , gdb_xmm8 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm9" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm9 , gcc_dwarf_xmm9 , LLDB_INVALID_REGNUM , gdb_xmm9 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm10" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm10 , gcc_dwarf_xmm10 , LLDB_INVALID_REGNUM , gdb_xmm10 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm11" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm11 , gcc_dwarf_xmm11 , LLDB_INVALID_REGNUM , gdb_xmm11 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm12" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm12 , gcc_dwarf_xmm12 , LLDB_INVALID_REGNUM , gdb_xmm12 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm13" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm13 , gcc_dwarf_xmm13 , LLDB_INVALID_REGNUM , gdb_xmm13 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm14" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm14 , gcc_dwarf_xmm14 , LLDB_INVALID_REGNUM , gdb_xmm14 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "xmm15" , NULL, 16, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_xmm15 , gcc_dwarf_xmm15 , LLDB_INVALID_REGNUM , gdb_xmm15 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "mxcsr" , NULL, 4, 0, eEncodingUint , eFormatHex , { LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , LLDB_INVALID_REGNUM , gdb_mxcsr , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm0" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm0 , gcc_dwarf_ymm0 , LLDB_INVALID_REGNUM , gdb_ymm0 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm1" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm1 , gcc_dwarf_ymm1 , LLDB_INVALID_REGNUM , gdb_ymm1 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm2" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm2 , gcc_dwarf_ymm2 , LLDB_INVALID_REGNUM , gdb_ymm2 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm3" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm3 , gcc_dwarf_ymm3 , LLDB_INVALID_REGNUM , gdb_ymm3 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm4" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm4 , gcc_dwarf_ymm4 , LLDB_INVALID_REGNUM , gdb_ymm4 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm5" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm5 , gcc_dwarf_ymm5 , LLDB_INVALID_REGNUM , gdb_ymm5 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm6" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm6 , gcc_dwarf_ymm6 , LLDB_INVALID_REGNUM , gdb_ymm6 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm7" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm7 , gcc_dwarf_ymm7 , LLDB_INVALID_REGNUM , gdb_ymm7 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm8" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm8 , gcc_dwarf_ymm8 , LLDB_INVALID_REGNUM , gdb_ymm8 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm9" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm9 , gcc_dwarf_ymm9 , LLDB_INVALID_REGNUM , gdb_ymm9 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm10" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm10 , gcc_dwarf_ymm10 , LLDB_INVALID_REGNUM , gdb_ymm10 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm11" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm11 , gcc_dwarf_ymm11 , LLDB_INVALID_REGNUM , gdb_ymm11 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm12" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm12 , gcc_dwarf_ymm12 , LLDB_INVALID_REGNUM , gdb_ymm12 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm13" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm13 , gcc_dwarf_ymm13 , LLDB_INVALID_REGNUM , gdb_ymm13 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm14" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm14 , gcc_dwarf_ymm14 , LLDB_INVALID_REGNUM , gdb_ymm14 , LLDB_INVALID_REGNUM }, NULL, NULL},
|
|
{ "ymm15" , NULL, 32, 0, eEncodingVector, eFormatVectorOfUInt8, { gcc_dwarf_ymm15 , gcc_dwarf_ymm15 , LLDB_INVALID_REGNUM , gdb_ymm15 , LLDB_INVALID_REGNUM }, NULL, NULL}
|
|
};
|
|
|
|
static const uint32_t k_num_register_infos = llvm::array_lengthof(g_register_infos);
|
|
static bool g_register_info_names_constified = false;
|
|
|
|
const lldb_private::RegisterInfo *
|
|
ABISysV_x86_64::GetRegisterInfoArray (uint32_t &count)
|
|
{
|
|
// Make the C-string names and alt_names for the register infos into const
|
|
// C-string values by having the ConstString unique the names in the global
|
|
// constant C-string pool.
|
|
if (!g_register_info_names_constified)
|
|
{
|
|
g_register_info_names_constified = true;
|
|
for (uint32_t i=0; i<k_num_register_infos; ++i)
|
|
{
|
|
if (g_register_infos[i].name)
|
|
g_register_infos[i].name = ConstString(g_register_infos[i].name).GetCString();
|
|
if (g_register_infos[i].alt_name)
|
|
g_register_infos[i].alt_name = ConstString(g_register_infos[i].alt_name).GetCString();
|
|
}
|
|
}
|
|
count = k_num_register_infos;
|
|
return g_register_infos;
|
|
}
|
|
|
|
|
|
size_t
|
|
ABISysV_x86_64::GetRedZoneSize () const
|
|
{
|
|
return 128;
|
|
}
|
|
|
|
//------------------------------------------------------------------
|
|
// Static Functions
|
|
//------------------------------------------------------------------
|
|
ABISP
|
|
ABISysV_x86_64::CreateInstance (const ArchSpec &arch)
|
|
{
|
|
static ABISP g_abi_sp;
|
|
if (arch.GetTriple().getArch() == llvm::Triple::x86_64)
|
|
{
|
|
if (!g_abi_sp)
|
|
g_abi_sp.reset (new ABISysV_x86_64);
|
|
return g_abi_sp;
|
|
}
|
|
return ABISP();
|
|
}
|
|
|
|
bool
|
|
ABISysV_x86_64::PrepareTrivialCall (Thread &thread,
|
|
addr_t sp,
|
|
addr_t func_addr,
|
|
addr_t return_addr,
|
|
llvm::ArrayRef<addr_t> args) const
|
|
{
|
|
Log *log(lldb_private::GetLogIfAllCategoriesSet (LIBLLDB_LOG_EXPRESSIONS));
|
|
|
|
if (log)
|
|
{
|
|
StreamString s;
|
|
s.Printf("ABISysV_x86_64::PrepareTrivialCall (tid = 0x%" PRIx64 ", sp = 0x%" PRIx64 ", func_addr = 0x%" PRIx64 ", return_addr = 0x%" PRIx64,
|
|
thread.GetID(),
|
|
(uint64_t)sp,
|
|
(uint64_t)func_addr,
|
|
(uint64_t)return_addr);
|
|
|
|
for (size_t i = 0; i < args.size(); ++i)
|
|
s.Printf (", arg%" PRIu64 " = 0x%" PRIx64, static_cast<uint64_t>(i + 1), args[i]);
|
|
s.PutCString (")");
|
|
log->PutCString(s.GetString().c_str());
|
|
}
|
|
|
|
RegisterContext *reg_ctx = thread.GetRegisterContext().get();
|
|
if (!reg_ctx)
|
|
return false;
|
|
|
|
const RegisterInfo *reg_info = NULL;
|
|
|
|
if (args.size() > 6) // TODO handle more than 6 arguments
|
|
return false;
|
|
|
|
for (size_t i = 0; i < args.size(); ++i)
|
|
{
|
|
reg_info = reg_ctx->GetRegisterInfo(eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG1 + i);
|
|
if (log)
|
|
log->Printf("About to write arg%" PRIu64 " (0x%" PRIx64 ") into %s", static_cast<uint64_t>(i + 1), args[i], reg_info->name);
|
|
if (!reg_ctx->WriteRegisterFromUnsigned(reg_info, args[i]))
|
|
return false;
|
|
}
|
|
|
|
// First, align the SP
|
|
|
|
if (log)
|
|
log->Printf("16-byte aligning SP: 0x%" PRIx64 " to 0x%" PRIx64, (uint64_t)sp, (uint64_t)(sp & ~0xfull));
|
|
|
|
sp &= ~(0xfull); // 16-byte alignment
|
|
|
|
sp -= 8;
|
|
|
|
Error error;
|
|
const RegisterInfo *pc_reg_info = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_PC);
|
|
const RegisterInfo *sp_reg_info = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_SP);
|
|
ProcessSP process_sp (thread.GetProcess());
|
|
|
|
RegisterValue reg_value;
|
|
|
|
#if 0
|
|
// This code adds an extra frame so that we don't lose the function that we came from
|
|
// by pushing the PC and the FP and then writing the current FP to point to the FP value
|
|
// we just pushed. It is disabled for now until the stack backtracing code can be debugged.
|
|
|
|
// Save current PC
|
|
const RegisterInfo *fp_reg_info = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_FP);
|
|
if (reg_ctx->ReadRegister(pc_reg_info, reg_value))
|
|
{
|
|
if (log)
|
|
log->Printf("Pushing the current PC onto the stack: 0x%" PRIx64 ": 0x%" PRIx64, (uint64_t)sp, reg_value.GetAsUInt64());
|
|
|
|
if (!process_sp->WritePointerToMemory(sp, reg_value.GetAsUInt64(), error))
|
|
return false;
|
|
|
|
sp -= 8;
|
|
|
|
// Save current FP
|
|
if (reg_ctx->ReadRegister(fp_reg_info, reg_value))
|
|
{
|
|
if (log)
|
|
log->Printf("Pushing the current FP onto the stack: 0x%" PRIx64 ": 0x%" PRIx64, (uint64_t)sp, reg_value.GetAsUInt64());
|
|
|
|
if (!process_sp->WritePointerToMemory(sp, reg_value.GetAsUInt64(), error))
|
|
return false;
|
|
}
|
|
// Setup FP backchain
|
|
reg_value.SetUInt64 (sp);
|
|
|
|
if (log)
|
|
log->Printf("Writing FP: 0x%" PRIx64 " (for FP backchain)", reg_value.GetAsUInt64());
|
|
|
|
if (!reg_ctx->WriteRegister(fp_reg_info, reg_value))
|
|
{
|
|
return false;
|
|
}
|
|
|
|
sp -= 8;
|
|
}
|
|
#endif
|
|
|
|
if (log)
|
|
log->Printf("Pushing the return address onto the stack: 0x%" PRIx64 ": 0x%" PRIx64, (uint64_t)sp, (uint64_t)return_addr);
|
|
|
|
// Save return address onto the stack
|
|
if (!process_sp->WritePointerToMemory(sp, return_addr, error))
|
|
return false;
|
|
|
|
// %rsp is set to the actual stack value.
|
|
|
|
if (log)
|
|
log->Printf("Writing SP: 0x%" PRIx64, (uint64_t)sp);
|
|
|
|
if (!reg_ctx->WriteRegisterFromUnsigned (sp_reg_info, sp))
|
|
return false;
|
|
|
|
// %rip is set to the address of the called function.
|
|
|
|
if (log)
|
|
log->Printf("Writing IP: 0x%" PRIx64, (uint64_t)func_addr);
|
|
|
|
if (!reg_ctx->WriteRegisterFromUnsigned (pc_reg_info, func_addr))
|
|
return false;
|
|
|
|
return true;
|
|
}
|
|
|
|
static bool ReadIntegerArgument(Scalar &scalar,
|
|
unsigned int bit_width,
|
|
bool is_signed,
|
|
Thread &thread,
|
|
uint32_t *argument_register_ids,
|
|
unsigned int ¤t_argument_register,
|
|
addr_t ¤t_stack_argument)
|
|
{
|
|
if (bit_width > 64)
|
|
return false; // Scalar can't hold large integer arguments
|
|
|
|
if (current_argument_register < 6)
|
|
{
|
|
scalar = thread.GetRegisterContext()->ReadRegisterAsUnsigned(argument_register_ids[current_argument_register], 0);
|
|
current_argument_register++;
|
|
if (is_signed)
|
|
scalar.SignExtend (bit_width);
|
|
}
|
|
else
|
|
{
|
|
uint32_t byte_size = (bit_width + (8-1))/8;
|
|
Error error;
|
|
if (thread.GetProcess()->ReadScalarIntegerFromMemory(current_stack_argument, byte_size, is_signed, scalar, error))
|
|
{
|
|
current_stack_argument += byte_size;
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
bool
|
|
ABISysV_x86_64::GetArgumentValues (Thread &thread,
|
|
ValueList &values) const
|
|
{
|
|
unsigned int num_values = values.GetSize();
|
|
unsigned int value_index;
|
|
|
|
// Extract the register context so we can read arguments from registers
|
|
|
|
RegisterContext *reg_ctx = thread.GetRegisterContext().get();
|
|
|
|
if (!reg_ctx)
|
|
return false;
|
|
|
|
// Get the pointer to the first stack argument so we have a place to start
|
|
// when reading data
|
|
|
|
addr_t sp = reg_ctx->GetSP(0);
|
|
|
|
if (!sp)
|
|
return false;
|
|
|
|
addr_t current_stack_argument = sp + 8; // jump over return address
|
|
|
|
uint32_t argument_register_ids[6];
|
|
|
|
argument_register_ids[0] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG1)->kinds[eRegisterKindLLDB];
|
|
argument_register_ids[1] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG2)->kinds[eRegisterKindLLDB];
|
|
argument_register_ids[2] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG3)->kinds[eRegisterKindLLDB];
|
|
argument_register_ids[3] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG4)->kinds[eRegisterKindLLDB];
|
|
argument_register_ids[4] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG5)->kinds[eRegisterKindLLDB];
|
|
argument_register_ids[5] = reg_ctx->GetRegisterInfo (eRegisterKindGeneric, LLDB_REGNUM_GENERIC_ARG6)->kinds[eRegisterKindLLDB];
|
|
|
|
unsigned int current_argument_register = 0;
|
|
|
|
for (value_index = 0;
|
|
value_index < num_values;
|
|
++value_index)
|
|
{
|
|
Value *value = values.GetValueAtIndex(value_index);
|
|
|
|
if (!value)
|
|
return false;
|
|
|
|
// We currently only support extracting values with Clang QualTypes.
|
|
// Do we care about others?
|
|
CompilerType clang_type = value->GetCompilerType();
|
|
if (!clang_type)
|
|
return false;
|
|
bool is_signed;
|
|
|
|
if (clang_type.IsIntegerType (is_signed))
|
|
{
|
|
ReadIntegerArgument(value->GetScalar(),
|
|
clang_type.GetBitSize(&thread),
|
|
is_signed,
|
|
thread,
|
|
argument_register_ids,
|
|
current_argument_register,
|
|
current_stack_argument);
|
|
}
|
|
else if (clang_type.IsPointerType ())
|
|
{
|
|
ReadIntegerArgument(value->GetScalar(),
|
|
clang_type.GetBitSize(&thread),
|
|
false,
|
|
thread,
|
|
argument_register_ids,
|
|
current_argument_register,
|
|
current_stack_argument);
|
|
}
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
Error
|
|
ABISysV_x86_64::SetReturnValueObject(lldb::StackFrameSP &frame_sp, lldb::ValueObjectSP &new_value_sp)
|
|
{
|
|
Error error;
|
|
if (!new_value_sp)
|
|
{
|
|
error.SetErrorString("Empty value object for return value.");
|
|
return error;
|
|
}
|
|
|
|
CompilerType clang_type = new_value_sp->GetCompilerType();
|
|
if (!clang_type)
|
|
{
|
|
error.SetErrorString ("Null clang type for return value.");
|
|
return error;
|
|
}
|
|
|
|
Thread *thread = frame_sp->GetThread().get();
|
|
|
|
bool is_signed;
|
|
uint32_t count;
|
|
bool is_complex;
|
|
|
|
RegisterContext *reg_ctx = thread->GetRegisterContext().get();
|
|
|
|
bool set_it_simple = false;
|
|
if (clang_type.IsIntegerType (is_signed) || clang_type.IsPointerType())
|
|
{
|
|
const RegisterInfo *reg_info = reg_ctx->GetRegisterInfoByName("rax", 0);
|
|
|
|
DataExtractor data;
|
|
Error data_error;
|
|
size_t num_bytes = new_value_sp->GetData(data, data_error);
|
|
if (data_error.Fail())
|
|
{
|
|
error.SetErrorStringWithFormat("Couldn't convert return value to raw data: %s", data_error.AsCString());
|
|
return error;
|
|
}
|
|
lldb::offset_t offset = 0;
|
|
if (num_bytes <= 8)
|
|
{
|
|
uint64_t raw_value = data.GetMaxU64(&offset, num_bytes);
|
|
|
|
if (reg_ctx->WriteRegisterFromUnsigned (reg_info, raw_value))
|
|
set_it_simple = true;
|
|
}
|
|
else
|
|
{
|
|
error.SetErrorString("We don't support returning longer than 64 bit integer values at present.");
|
|
}
|
|
|
|
}
|
|
else if (clang_type.IsFloatingPointType (count, is_complex))
|
|
{
|
|
if (is_complex)
|
|
error.SetErrorString ("We don't support returning complex values at present");
|
|
else
|
|
{
|
|
size_t bit_width = clang_type.GetBitSize(frame_sp.get());
|
|
if (bit_width <= 64)
|
|
{
|
|
const RegisterInfo *xmm0_info = reg_ctx->GetRegisterInfoByName("xmm0", 0);
|
|
RegisterValue xmm0_value;
|
|
DataExtractor data;
|
|
Error data_error;
|
|
size_t num_bytes = new_value_sp->GetData(data, data_error);
|
|
if (data_error.Fail())
|
|
{
|
|
error.SetErrorStringWithFormat("Couldn't convert return value to raw data: %s", data_error.AsCString());
|
|
return error;
|
|
}
|
|
|
|
unsigned char buffer[16];
|
|
ByteOrder byte_order = data.GetByteOrder();
|
|
|
|
data.CopyByteOrderedData (0, num_bytes, buffer, 16, byte_order);
|
|
xmm0_value.SetBytes(buffer, 16, byte_order);
|
|
reg_ctx->WriteRegister(xmm0_info, xmm0_value);
|
|
set_it_simple = true;
|
|
}
|
|
else
|
|
{
|
|
// FIXME - don't know how to do 80 bit long doubles yet.
|
|
error.SetErrorString ("We don't support returning float values > 64 bits at present");
|
|
}
|
|
}
|
|
}
|
|
|
|
if (!set_it_simple)
|
|
{
|
|
// Okay we've got a structure or something that doesn't fit in a simple register.
|
|
// We should figure out where it really goes, but we don't support this yet.
|
|
error.SetErrorString ("We only support setting simple integer and float return types at present.");
|
|
}
|
|
|
|
return error;
|
|
}
|
|
|
|
|
|
ValueObjectSP
|
|
ABISysV_x86_64::GetReturnValueObjectSimple (Thread &thread,
|
|
CompilerType &return_clang_type) const
|
|
{
|
|
ValueObjectSP return_valobj_sp;
|
|
Value value;
|
|
|
|
if (!return_clang_type)
|
|
return return_valobj_sp;
|
|
|
|
//value.SetContext (Value::eContextTypeClangType, return_value_type);
|
|
value.SetCompilerType (return_clang_type);
|
|
|
|
RegisterContext *reg_ctx = thread.GetRegisterContext().get();
|
|
if (!reg_ctx)
|
|
return return_valobj_sp;
|
|
|
|
const uint32_t type_flags = return_clang_type.GetTypeInfo ();
|
|
if (type_flags & eTypeIsScalar)
|
|
{
|
|
value.SetValueType(Value::eValueTypeScalar);
|
|
|
|
bool success = false;
|
|
if (type_flags & eTypeIsInteger)
|
|
{
|
|
// Extract the register context so we can read arguments from registers
|
|
|
|
const size_t byte_size = return_clang_type.GetByteSize(nullptr);
|
|
uint64_t raw_value = thread.GetRegisterContext()->ReadRegisterAsUnsigned(reg_ctx->GetRegisterInfoByName("rax", 0), 0);
|
|
const bool is_signed = (type_flags & eTypeIsSigned) != 0;
|
|
switch (byte_size)
|
|
{
|
|
default:
|
|
break;
|
|
|
|
case sizeof(uint64_t):
|
|
if (is_signed)
|
|
value.GetScalar() = (int64_t)(raw_value);
|
|
else
|
|
value.GetScalar() = (uint64_t)(raw_value);
|
|
success = true;
|
|
break;
|
|
|
|
case sizeof(uint32_t):
|
|
if (is_signed)
|
|
value.GetScalar() = (int32_t)(raw_value & UINT32_MAX);
|
|
else
|
|
value.GetScalar() = (uint32_t)(raw_value & UINT32_MAX);
|
|
success = true;
|
|
break;
|
|
|
|
case sizeof(uint16_t):
|
|
if (is_signed)
|
|
value.GetScalar() = (int16_t)(raw_value & UINT16_MAX);
|
|
else
|
|
value.GetScalar() = (uint16_t)(raw_value & UINT16_MAX);
|
|
success = true;
|
|
break;
|
|
|
|
case sizeof(uint8_t):
|
|
if (is_signed)
|
|
value.GetScalar() = (int8_t)(raw_value & UINT8_MAX);
|
|
else
|
|
value.GetScalar() = (uint8_t)(raw_value & UINT8_MAX);
|
|
success = true;
|
|
break;
|
|
}
|
|
}
|
|
else if (type_flags & eTypeIsFloat)
|
|
{
|
|
if (type_flags & eTypeIsComplex)
|
|
{
|
|
// Don't handle complex yet.
|
|
}
|
|
else
|
|
{
|
|
const size_t byte_size = return_clang_type.GetByteSize(nullptr);
|
|
if (byte_size <= sizeof(long double))
|
|
{
|
|
const RegisterInfo *xmm0_info = reg_ctx->GetRegisterInfoByName("xmm0", 0);
|
|
RegisterValue xmm0_value;
|
|
if (reg_ctx->ReadRegister (xmm0_info, xmm0_value))
|
|
{
|
|
DataExtractor data;
|
|
if (xmm0_value.GetData(data))
|
|
{
|
|
lldb::offset_t offset = 0;
|
|
if (byte_size == sizeof(float))
|
|
{
|
|
value.GetScalar() = (float) data.GetFloat(&offset);
|
|
success = true;
|
|
}
|
|
else if (byte_size == sizeof(double))
|
|
{
|
|
value.GetScalar() = (double) data.GetDouble(&offset);
|
|
success = true;
|
|
}
|
|
else if (byte_size == sizeof(long double))
|
|
{
|
|
// Don't handle long double since that can be encoded as 80 bit floats...
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
if (success)
|
|
return_valobj_sp = ValueObjectConstResult::Create (thread.GetStackFrameAtIndex(0).get(),
|
|
value,
|
|
ConstString(""));
|
|
|
|
}
|
|
else if (type_flags & eTypeIsPointer)
|
|
{
|
|
unsigned rax_id = reg_ctx->GetRegisterInfoByName("rax", 0)->kinds[eRegisterKindLLDB];
|
|
value.GetScalar() = (uint64_t)thread.GetRegisterContext()->ReadRegisterAsUnsigned(rax_id, 0);
|
|
value.SetValueType(Value::eValueTypeScalar);
|
|
return_valobj_sp = ValueObjectConstResult::Create (thread.GetStackFrameAtIndex(0).get(),
|
|
value,
|
|
ConstString(""));
|
|
}
|
|
else if (type_flags & eTypeIsVector)
|
|
{
|
|
const size_t byte_size = return_clang_type.GetByteSize(nullptr);
|
|
if (byte_size > 0)
|
|
{
|
|
const RegisterInfo *altivec_reg = reg_ctx->GetRegisterInfoByName("xmm0", 0);
|
|
if (altivec_reg == nullptr)
|
|
altivec_reg = reg_ctx->GetRegisterInfoByName("mm0", 0);
|
|
|
|
if (altivec_reg)
|
|
{
|
|
if (byte_size <= altivec_reg->byte_size)
|
|
{
|
|
ProcessSP process_sp (thread.GetProcess());
|
|
if (process_sp)
|
|
{
|
|
std::unique_ptr<DataBufferHeap> heap_data_ap (new DataBufferHeap(byte_size, 0));
|
|
const ByteOrder byte_order = process_sp->GetByteOrder();
|
|
RegisterValue reg_value;
|
|
if (reg_ctx->ReadRegister(altivec_reg, reg_value))
|
|
{
|
|
Error error;
|
|
if (reg_value.GetAsMemoryData (altivec_reg,
|
|
heap_data_ap->GetBytes(),
|
|
heap_data_ap->GetByteSize(),
|
|
byte_order,
|
|
error))
|
|
{
|
|
DataExtractor data (DataBufferSP (heap_data_ap.release()),
|
|
byte_order,
|
|
process_sp->GetTarget().GetArchitecture().GetAddressByteSize());
|
|
return_valobj_sp = ValueObjectConstResult::Create (&thread,
|
|
return_clang_type,
|
|
ConstString(""),
|
|
data);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
else if (byte_size <= altivec_reg->byte_size*2)
|
|
{
|
|
const RegisterInfo *altivec_reg2 = reg_ctx->GetRegisterInfoByName("xmm1", 0);
|
|
if (altivec_reg2)
|
|
{
|
|
ProcessSP process_sp (thread.GetProcess());
|
|
if (process_sp)
|
|
{
|
|
std::unique_ptr<DataBufferHeap> heap_data_ap (new DataBufferHeap(byte_size, 0));
|
|
const ByteOrder byte_order = process_sp->GetByteOrder();
|
|
RegisterValue reg_value;
|
|
RegisterValue reg_value2;
|
|
if (reg_ctx->ReadRegister(altivec_reg, reg_value) && reg_ctx->ReadRegister(altivec_reg2, reg_value2))
|
|
{
|
|
|
|
Error error;
|
|
if (reg_value.GetAsMemoryData (altivec_reg,
|
|
heap_data_ap->GetBytes(),
|
|
altivec_reg->byte_size,
|
|
byte_order,
|
|
error) &&
|
|
reg_value2.GetAsMemoryData (altivec_reg2,
|
|
heap_data_ap->GetBytes() + altivec_reg->byte_size,
|
|
heap_data_ap->GetByteSize() - altivec_reg->byte_size,
|
|
byte_order,
|
|
error))
|
|
{
|
|
DataExtractor data (DataBufferSP (heap_data_ap.release()),
|
|
byte_order,
|
|
process_sp->GetTarget().GetArchitecture().GetAddressByteSize());
|
|
return_valobj_sp = ValueObjectConstResult::Create (&thread,
|
|
return_clang_type,
|
|
ConstString(""),
|
|
data);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
return return_valobj_sp;
|
|
}
|
|
|
|
ValueObjectSP
|
|
ABISysV_x86_64::GetReturnValueObjectImpl (Thread &thread, CompilerType &return_clang_type) const
|
|
{
|
|
ValueObjectSP return_valobj_sp;
|
|
|
|
if (!return_clang_type)
|
|
return return_valobj_sp;
|
|
|
|
ExecutionContext exe_ctx (thread.shared_from_this());
|
|
return_valobj_sp = GetReturnValueObjectSimple(thread, return_clang_type);
|
|
if (return_valobj_sp)
|
|
return return_valobj_sp;
|
|
|
|
RegisterContextSP reg_ctx_sp = thread.GetRegisterContext();
|
|
if (!reg_ctx_sp)
|
|
return return_valobj_sp;
|
|
|
|
const size_t bit_width = return_clang_type.GetBitSize(&thread);
|
|
if (return_clang_type.IsAggregateType())
|
|
{
|
|
Target *target = exe_ctx.GetTargetPtr();
|
|
bool is_memory = true;
|
|
if (bit_width <= 128)
|
|
{
|
|
ByteOrder target_byte_order = target->GetArchitecture().GetByteOrder();
|
|
DataBufferSP data_sp (new DataBufferHeap(16, 0));
|
|
DataExtractor return_ext (data_sp,
|
|
target_byte_order,
|
|
target->GetArchitecture().GetAddressByteSize());
|
|
|
|
const RegisterInfo *rax_info = reg_ctx_sp->GetRegisterInfoByName("rax", 0);
|
|
const RegisterInfo *rdx_info = reg_ctx_sp->GetRegisterInfoByName("rdx", 0);
|
|
const RegisterInfo *xmm0_info = reg_ctx_sp->GetRegisterInfoByName("xmm0", 0);
|
|
const RegisterInfo *xmm1_info = reg_ctx_sp->GetRegisterInfoByName("xmm1", 0);
|
|
|
|
RegisterValue rax_value, rdx_value, xmm0_value, xmm1_value;
|
|
reg_ctx_sp->ReadRegister (rax_info, rax_value);
|
|
reg_ctx_sp->ReadRegister (rdx_info, rdx_value);
|
|
reg_ctx_sp->ReadRegister (xmm0_info, xmm0_value);
|
|
reg_ctx_sp->ReadRegister (xmm1_info, xmm1_value);
|
|
|
|
DataExtractor rax_data, rdx_data, xmm0_data, xmm1_data;
|
|
|
|
rax_value.GetData(rax_data);
|
|
rdx_value.GetData(rdx_data);
|
|
xmm0_value.GetData(xmm0_data);
|
|
xmm1_value.GetData(xmm1_data);
|
|
|
|
uint32_t fp_bytes = 0; // Tracks how much of the xmm registers we've consumed so far
|
|
uint32_t integer_bytes = 0; // Tracks how much of the rax/rds registers we've consumed so far
|
|
|
|
const uint32_t num_children = return_clang_type.GetNumFields ();
|
|
|
|
// Since we are in the small struct regime, assume we are not in memory.
|
|
is_memory = false;
|
|
|
|
for (uint32_t idx = 0; idx < num_children; idx++)
|
|
{
|
|
std::string name;
|
|
uint64_t field_bit_offset = 0;
|
|
bool is_signed;
|
|
bool is_complex;
|
|
uint32_t count;
|
|
|
|
CompilerType field_clang_type = return_clang_type.GetFieldAtIndex (idx, name, &field_bit_offset, NULL, NULL);
|
|
const size_t field_bit_width = field_clang_type.GetBitSize(&thread);
|
|
|
|
// if we don't know the size of the field (e.g. invalid type), just bail out
|
|
if (field_bit_width == 0)
|
|
break;
|
|
|
|
// If there are any unaligned fields, this is stored in memory.
|
|
if (field_bit_offset % field_bit_width != 0)
|
|
{
|
|
is_memory = true;
|
|
break;
|
|
}
|
|
|
|
uint32_t field_byte_width = field_bit_width/8;
|
|
uint32_t field_byte_offset = field_bit_offset/8;
|
|
|
|
|
|
DataExtractor *copy_from_extractor = NULL;
|
|
uint32_t copy_from_offset = 0;
|
|
|
|
if (field_clang_type.IsIntegerType (is_signed) || field_clang_type.IsPointerType ())
|
|
{
|
|
if (integer_bytes < 8)
|
|
{
|
|
if (integer_bytes + field_byte_width <= 8)
|
|
{
|
|
// This is in RAX, copy from register to our result structure:
|
|
copy_from_extractor = &rax_data;
|
|
copy_from_offset = integer_bytes;
|
|
integer_bytes += field_byte_width;
|
|
}
|
|
else
|
|
{
|
|
// The next field wouldn't fit in the remaining space, so we pushed it to rdx.
|
|
copy_from_extractor = &rdx_data;
|
|
copy_from_offset = 0;
|
|
integer_bytes = 8 + field_byte_width;
|
|
|
|
}
|
|
}
|
|
else if (integer_bytes + field_byte_width <= 16)
|
|
{
|
|
copy_from_extractor = &rdx_data;
|
|
copy_from_offset = integer_bytes - 8;
|
|
integer_bytes += field_byte_width;
|
|
}
|
|
else
|
|
{
|
|
// The last field didn't fit. I can't see how that would happen w/o the overall size being
|
|
// greater than 16 bytes. For now, return a NULL return value object.
|
|
return return_valobj_sp;
|
|
}
|
|
}
|
|
else if (field_clang_type.IsFloatingPointType (count, is_complex))
|
|
{
|
|
// Structs with long doubles are always passed in memory.
|
|
if (field_bit_width == 128)
|
|
{
|
|
is_memory = true;
|
|
break;
|
|
}
|
|
else if (field_bit_width == 64)
|
|
{
|
|
// These have to be in a single xmm register.
|
|
if (fp_bytes == 0)
|
|
copy_from_extractor = &xmm0_data;
|
|
else
|
|
copy_from_extractor = &xmm1_data;
|
|
|
|
copy_from_offset = 0;
|
|
fp_bytes += field_byte_width;
|
|
}
|
|
else if (field_bit_width == 32)
|
|
{
|
|
// This one is kind of complicated. If we are in an "eightbyte" with another float, we'll
|
|
// be stuffed into an xmm register with it. If we are in an "eightbyte" with one or more ints,
|
|
// then we will be stuffed into the appropriate GPR with them.
|
|
bool in_gpr;
|
|
if (field_byte_offset % 8 == 0)
|
|
{
|
|
// We are at the beginning of one of the eightbytes, so check the next element (if any)
|
|
if (idx == num_children - 1)
|
|
in_gpr = false;
|
|
else
|
|
{
|
|
uint64_t next_field_bit_offset = 0;
|
|
CompilerType next_field_clang_type = return_clang_type.GetFieldAtIndex (idx + 1,
|
|
name,
|
|
&next_field_bit_offset,
|
|
NULL,
|
|
NULL);
|
|
if (next_field_clang_type.IsIntegerType (is_signed))
|
|
in_gpr = true;
|
|
else
|
|
{
|
|
copy_from_offset = 0;
|
|
in_gpr = false;
|
|
}
|
|
}
|
|
|
|
}
|
|
else if (field_byte_offset % 4 == 0)
|
|
{
|
|
// We are inside of an eightbyte, so see if the field before us is floating point:
|
|
// This could happen if somebody put padding in the structure.
|
|
if (idx == 0)
|
|
in_gpr = false;
|
|
else
|
|
{
|
|
uint64_t prev_field_bit_offset = 0;
|
|
CompilerType prev_field_clang_type = return_clang_type.GetFieldAtIndex (idx - 1,
|
|
name,
|
|
&prev_field_bit_offset,
|
|
NULL,
|
|
NULL);
|
|
if (prev_field_clang_type.IsIntegerType (is_signed))
|
|
in_gpr = true;
|
|
else
|
|
{
|
|
copy_from_offset = 4;
|
|
in_gpr = false;
|
|
}
|
|
}
|
|
|
|
}
|
|
else
|
|
{
|
|
is_memory = true;
|
|
continue;
|
|
}
|
|
|
|
// Okay, we've figured out whether we are in GPR or XMM, now figure out which one.
|
|
if (in_gpr)
|
|
{
|
|
if (integer_bytes < 8)
|
|
{
|
|
// This is in RAX, copy from register to our result structure:
|
|
copy_from_extractor = &rax_data;
|
|
copy_from_offset = integer_bytes;
|
|
integer_bytes += field_byte_width;
|
|
}
|
|
else
|
|
{
|
|
copy_from_extractor = &rdx_data;
|
|
copy_from_offset = integer_bytes - 8;
|
|
integer_bytes += field_byte_width;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if (fp_bytes < 8)
|
|
copy_from_extractor = &xmm0_data;
|
|
else
|
|
copy_from_extractor = &xmm1_data;
|
|
|
|
fp_bytes += field_byte_width;
|
|
}
|
|
}
|
|
}
|
|
|
|
// These two tests are just sanity checks. If I somehow get the
|
|
// type calculation wrong above it is better to just return nothing
|
|
// than to assert or crash.
|
|
if (!copy_from_extractor)
|
|
return return_valobj_sp;
|
|
if (copy_from_offset + field_byte_width > copy_from_extractor->GetByteSize())
|
|
return return_valobj_sp;
|
|
|
|
copy_from_extractor->CopyByteOrderedData (copy_from_offset,
|
|
field_byte_width,
|
|
data_sp->GetBytes() + field_byte_offset,
|
|
field_byte_width,
|
|
target_byte_order);
|
|
}
|
|
|
|
if (!is_memory)
|
|
{
|
|
// The result is in our data buffer. Let's make a variable object out of it:
|
|
return_valobj_sp = ValueObjectConstResult::Create (&thread,
|
|
return_clang_type,
|
|
ConstString(""),
|
|
return_ext);
|
|
}
|
|
}
|
|
|
|
|
|
// FIXME: This is just taking a guess, rax may very well no longer hold the return storage location.
|
|
// If we are going to do this right, when we make a new frame we should check to see if it uses a memory
|
|
// return, and if we are at the first instruction and if so stash away the return location. Then we would
|
|
// only return the memory return value if we know it is valid.
|
|
|
|
if (is_memory)
|
|
{
|
|
unsigned rax_id = reg_ctx_sp->GetRegisterInfoByName("rax", 0)->kinds[eRegisterKindLLDB];
|
|
lldb::addr_t storage_addr = (uint64_t)thread.GetRegisterContext()->ReadRegisterAsUnsigned(rax_id, 0);
|
|
return_valobj_sp = ValueObjectMemory::Create (&thread,
|
|
"",
|
|
Address (storage_addr, NULL),
|
|
return_clang_type);
|
|
}
|
|
}
|
|
|
|
return return_valobj_sp;
|
|
}
|
|
|
|
// This defines the CFA as rsp+8
|
|
// the saved pc is at CFA-8 (i.e. rsp+0)
|
|
// The saved rsp is CFA+0
|
|
|
|
bool
|
|
ABISysV_x86_64::CreateFunctionEntryUnwindPlan (UnwindPlan &unwind_plan)
|
|
{
|
|
unwind_plan.Clear();
|
|
unwind_plan.SetRegisterKind (eRegisterKindDWARF);
|
|
|
|
uint32_t sp_reg_num = gcc_dwarf_rsp;
|
|
uint32_t pc_reg_num = gcc_dwarf_rip;
|
|
|
|
UnwindPlan::RowSP row(new UnwindPlan::Row);
|
|
row->GetCFAValue().SetIsRegisterPlusOffset(sp_reg_num, 8);
|
|
row->SetRegisterLocationToAtCFAPlusOffset(pc_reg_num, -8, false);
|
|
row->SetRegisterLocationToIsCFAPlusOffset(sp_reg_num, 0, true);
|
|
unwind_plan.AppendRow (row);
|
|
unwind_plan.SetSourceName ("x86_64 at-func-entry default");
|
|
unwind_plan.SetSourcedFromCompiler (eLazyBoolNo);
|
|
return true;
|
|
}
|
|
|
|
// This defines the CFA as rbp+16
|
|
// The saved pc is at CFA-8 (i.e. rbp+8)
|
|
// The saved rbp is at CFA-16 (i.e. rbp+0)
|
|
// The saved rsp is CFA+0
|
|
|
|
bool
|
|
ABISysV_x86_64::CreateDefaultUnwindPlan (UnwindPlan &unwind_plan)
|
|
{
|
|
unwind_plan.Clear();
|
|
unwind_plan.SetRegisterKind (eRegisterKindDWARF);
|
|
|
|
uint32_t fp_reg_num = gcc_dwarf_rbp;
|
|
uint32_t sp_reg_num = gcc_dwarf_rsp;
|
|
uint32_t pc_reg_num = gcc_dwarf_rip;
|
|
|
|
UnwindPlan::RowSP row(new UnwindPlan::Row);
|
|
|
|
const int32_t ptr_size = 8;
|
|
row->GetCFAValue().SetIsRegisterPlusOffset(gcc_dwarf_rbp, 2 * ptr_size);
|
|
row->SetOffset (0);
|
|
|
|
row->SetRegisterLocationToAtCFAPlusOffset(fp_reg_num, ptr_size * -2, true);
|
|
row->SetRegisterLocationToAtCFAPlusOffset(pc_reg_num, ptr_size * -1, true);
|
|
row->SetRegisterLocationToIsCFAPlusOffset(sp_reg_num, 0, true);
|
|
|
|
unwind_plan.AppendRow (row);
|
|
unwind_plan.SetSourceName ("x86_64 default unwind plan");
|
|
unwind_plan.SetSourcedFromCompiler (eLazyBoolNo);
|
|
unwind_plan.SetUnwindPlanValidAtAllInstructions (eLazyBoolNo);
|
|
return true;
|
|
}
|
|
|
|
bool
|
|
ABISysV_x86_64::RegisterIsVolatile (const RegisterInfo *reg_info)
|
|
{
|
|
return !RegisterIsCalleeSaved (reg_info);
|
|
}
|
|
|
|
|
|
|
|
// See "Register Usage" in the
|
|
// "System V Application Binary Interface"
|
|
// "AMD64 Architecture Processor Supplement"
|
|
// (or "x86-64(tm) Architecture Processor Supplement" in earlier revisions)
|
|
// (this doc is also commonly referred to as the x86-64/AMD64 psABI)
|
|
// Edited by Michael Matz, Jan Hubicka, Andreas Jaeger, and Mark Mitchell
|
|
// current version is 0.99.6 released 2012-07-02 at http://refspecs.linuxfoundation.org/elf/x86-64-abi-0.99.pdf
|
|
// It's being revised & updated at https://github.com/hjl-tools/x86-psABI/
|
|
|
|
bool
|
|
ABISysV_x86_64::RegisterIsCalleeSaved (const RegisterInfo *reg_info)
|
|
{
|
|
if (reg_info)
|
|
{
|
|
// Preserved registers are :
|
|
// rbx, rsp, rbp, r12, r13, r14, r15
|
|
// mxcsr (partially preserved)
|
|
// x87 control word
|
|
|
|
const char *name = reg_info->name;
|
|
if (name[0] == 'r')
|
|
{
|
|
switch (name[1])
|
|
{
|
|
case '1': // r12, r13, r14, r15
|
|
if (name[2] >= '2' && name[2] <= '5')
|
|
return name[3] == '\0';
|
|
break;
|
|
|
|
default:
|
|
break;
|
|
}
|
|
}
|
|
|
|
// Accept shorter-variant versions, rbx/ebx, rip/ eip, etc.
|
|
if (name[0] == 'r' || name[0] == 'e')
|
|
{
|
|
switch (name[1])
|
|
{
|
|
case 'b': // rbp, rbx
|
|
if (name[2] == 'p' || name[2] == 'x')
|
|
return name[3] == '\0';
|
|
break;
|
|
|
|
case 'i': // rip
|
|
if (name[2] == 'p')
|
|
return name[3] == '\0';
|
|
break;
|
|
|
|
case 's': // rsp
|
|
if (name[2] == 'p')
|
|
return name[3] == '\0';
|
|
break;
|
|
|
|
}
|
|
}
|
|
if (name[0] == 's' && name[1] == 'p' && name[2] == '\0') // sp
|
|
return true;
|
|
if (name[0] == 'f' && name[1] == 'p' && name[2] == '\0') // fp
|
|
return true;
|
|
if (name[0] == 'p' && name[1] == 'c' && name[2] == '\0') // pc
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
|
|
|
|
void
|
|
ABISysV_x86_64::Initialize()
|
|
{
|
|
PluginManager::RegisterPlugin (GetPluginNameStatic(),
|
|
"System V ABI for x86_64 targets",
|
|
CreateInstance);
|
|
}
|
|
|
|
void
|
|
ABISysV_x86_64::Terminate()
|
|
{
|
|
PluginManager::UnregisterPlugin (CreateInstance);
|
|
}
|
|
|
|
lldb_private::ConstString
|
|
ABISysV_x86_64::GetPluginNameStatic()
|
|
{
|
|
static ConstString g_name("sysv-x86_64");
|
|
return g_name;
|
|
}
|
|
|
|
//------------------------------------------------------------------
|
|
// PluginInterface protocol
|
|
//------------------------------------------------------------------
|
|
lldb_private::ConstString
|
|
ABISysV_x86_64::GetPluginName()
|
|
{
|
|
return GetPluginNameStatic();
|
|
}
|
|
|
|
uint32_t
|
|
ABISysV_x86_64::GetPluginVersion()
|
|
{
|
|
return 1;
|
|
}
|
|
|