首页
社区
课程
招聘
[原创]EFER hook
发表于: 2022-11-4 22:17 29642

[原创]EFER hook

2022-11-4 22:17
29642

最近在看HyperDbg这个项目,看到了一种基于VT的syscall挂钩方法,记录一下。EFER(Extended Feature Enable Register)是MSR寄存器的一种,相关功能可以在intel白皮书卷3A,第2.2.1节中找到。

EFER寄存器的SCE位用来控制对SYSCALL/SYSRET的支持。
在卷2B,第4.3节可以找到SYSCALL指令的操作

可以看出当EFER寄存器的SCE位为0时会产生#UD异常,如果我们手动将SCE清零,那么就会导致#UD异常,再通过设置VT中对#UD异常的响应事件就可以拦截SYSCALL。
用代码模拟SYSCALL:

看下SYSRET的操作,可以看出也是SCE位置0时抛出#UD异常

用代码模拟:

手动清除SCE位并设置拦截:

对#UD事件的响应:

IF (CS.L ≠ 1 ) or (IA32_EFER.LMA ≠ 1) or (IA32_EFER.SCE ≠ 1)
(* Not in 64-Bit Mode or SYSCALL/SYSRET not enabled in IA32_EFER *)
  THEN #UD;
FI;
RCX ← RIP; (* Will contain address of next instruction *)
RIP ← IA32_LSTAR;
R11 ← RFLAGS;
RFLAGS ← RFLAGS AND NOT(IA32_FMASK);
CS.Selector ← IA32_STAR[47:32] AND FFFCH (* Operating system provides CS; RPL forced to 0 *)
(* Set rest of CS to a fixed value *)
CS.Base ← 0; (* Flat segment *)
CS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
CS.Type 11; (* Execute/read code, accessed *)
CS.S ← 1;
CS.DPL ← 0;
CS.P ← 1;
CS.L ← 1; (* Entry is to 64-bit mode *)
CS.D ← 0; (* Required if CS.L = 1 *)
CS.G ← 1; (* 4-KByte granularity *)
CPL ← 0;
SS.Selector ← IA32_STAR[47:32] + 8; (* SS just above CS *)
(* Set rest of SS to a fixed value *)
SS.Base ← 0; (* Flat segment *)
SS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
SS.Type 3; (* Read/write data, accessed *)
SS.S ← 1;
SS.DPL ← 0;
SS.P ← 1;
SS.B ← 1; (* 32-bit stack segment *)
SS.G ← 1; (* 4-KByte granularity *)
IF (CS.L ≠ 1 ) or (IA32_EFER.LMA ≠ 1) or (IA32_EFER.SCE ≠ 1)
(* Not in 64-Bit Mode or SYSCALL/SYSRET not enabled in IA32_EFER *)
  THEN #UD;
FI;
RCX ← RIP; (* Will contain address of next instruction *)
RIP ← IA32_LSTAR;
R11 ← RFLAGS;
RFLAGS ← RFLAGS AND NOT(IA32_FMASK);
CS.Selector ← IA32_STAR[47:32] AND FFFCH (* Operating system provides CS; RPL forced to 0 *)
(* Set rest of CS to a fixed value *)
CS.Base ← 0; (* Flat segment *)
CS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
CS.Type 11; (* Execute/read code, accessed *)
CS.S ← 1;
CS.DPL ← 0;
CS.P ← 1;
CS.L ← 1; (* Entry is to 64-bit mode *)
CS.D ← 0; (* Required if CS.L = 1 *)
CS.G ← 1; (* 4-KByte granularity *)
CPL ← 0;
SS.Selector ← IA32_STAR[47:32] + 8; (* SS just above CS *)
(* Set rest of SS to a fixed value *)
SS.Base ← 0; (* Flat segment *)
SS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
SS.Type 3; (* Read/write data, accessed *)
SS.S ← 1;
SS.DPL ← 0;
SS.P ← 1;
SS.B ← 1; (* 32-bit stack segment *)
SS.G ← 1; (* 4-KByte granularity *)
_Use_decl_annotations_
BOOLEAN
SyscallHookEmulateSYSCALL(PGUEST_REGS Regs)
{
    VMX_SEGMENT_SELECTOR Cs, Ss;
    UINT32               InstructionLength;
    UINT64               MsrValue;
    UINT64               GuestRip;
    UINT64               GuestRflags;
 
    //
    // Reading guest's RIP
    //
    __vmx_vmread(VMCS_GUEST_RIP, &GuestRip);
 
    //
    // Reading instruction length
    //
    __vmx_vmread(VMCS_VMEXIT_INSTRUCTION_LENGTH, &InstructionLength);
 
    //
    // Reading guest's Rflags
    //
    __vmx_vmread(VMCS_GUEST_RFLAGS, &GuestRflags);
 
    //
    // Save the address of the instruction following SYSCALL into RCX and then
    // load RIP from IA32_LSTAR.
    //
    MsrValue  = __readmsr(IA32_LSTAR);
    Regs->rcx = GuestRip + InstructionLength;
    GuestRip  = MsrValue;
    __vmx_vmwrite(VMCS_GUEST_RIP, GuestRip);
 
    //
    // Save RFLAGS into R11 and then mask RFLAGS using IA32_FMASK
    //
    MsrValue  = __readmsr(IA32_FMASK);
    Regs->r11 = GuestRflags;
    GuestRflags &= ~(MsrValue | X86_FLAGS_RF);
    __vmx_vmwrite(VMCS_GUEST_RFLAGS, GuestRflags);
 
    //
    // Load the CS and SS selectors with values derived from bits 47:32 of IA32_STAR
    //
    MsrValue             = __readmsr(IA32_STAR);
    Cs.Selector          = (UINT16)((MsrValue >> 32) & ~3); // STAR[47:32] & ~RPL3
    Cs.Base              = 0;                               // flat segment
    Cs.Limit             = (UINT32)~0;                      // 4GB limit
    Cs.Attributes.AsUInt = 0xA09B;                          // L+DB+P+S+DPL0+Code
    SetGuestCs(&Cs);
 
    Ss.Selector          = (UINT16)(((MsrValue >> 32) & ~3) + 8); // STAR[47:32] + 8
    Ss.Base              = 0;                                     // flat segment
    Ss.Limit             = (UINT32)~0;                            // 4GB limit
    Ss.Attributes.AsUInt = 0xC093;                                // G+DB+P+S+DPL0+Data
    SetGuestSs(&Ss);
 
    return TRUE;
}
_Use_decl_annotations_
BOOLEAN
SyscallHookEmulateSYSCALL(PGUEST_REGS Regs)
{
    VMX_SEGMENT_SELECTOR Cs, Ss;
    UINT32               InstructionLength;
    UINT64               MsrValue;
    UINT64               GuestRip;
    UINT64               GuestRflags;
 
    //
    // Reading guest's RIP
    //
    __vmx_vmread(VMCS_GUEST_RIP, &GuestRip);
 
    //
    // Reading instruction length
    //
    __vmx_vmread(VMCS_VMEXIT_INSTRUCTION_LENGTH, &InstructionLength);
 
    //
    // Reading guest's Rflags
    //
    __vmx_vmread(VMCS_GUEST_RFLAGS, &GuestRflags);
 
    //
    // Save the address of the instruction following SYSCALL into RCX and then
    // load RIP from IA32_LSTAR.
    //
    MsrValue  = __readmsr(IA32_LSTAR);
    Regs->rcx = GuestRip + InstructionLength;
    GuestRip  = MsrValue;
    __vmx_vmwrite(VMCS_GUEST_RIP, GuestRip);
 
    //
    // Save RFLAGS into R11 and then mask RFLAGS using IA32_FMASK
    //
    MsrValue  = __readmsr(IA32_FMASK);
    Regs->r11 = GuestRflags;
    GuestRflags &= ~(MsrValue | X86_FLAGS_RF);
    __vmx_vmwrite(VMCS_GUEST_RFLAGS, GuestRflags);
 
    //
    // Load the CS and SS selectors with values derived from bits 47:32 of IA32_STAR
    //
    MsrValue             = __readmsr(IA32_STAR);
    Cs.Selector          = (UINT16)((MsrValue >> 32) & ~3); // STAR[47:32] & ~RPL3
    Cs.Base              = 0;                               // flat segment
    Cs.Limit             = (UINT32)~0;                      // 4GB limit
    Cs.Attributes.AsUInt = 0xA09B;                          // L+DB+P+S+DPL0+Code
    SetGuestCs(&Cs);
 
    Ss.Selector          = (UINT16)(((MsrValue >> 32) & ~3) + 8); // STAR[47:32] + 8
    Ss.Base              = 0;                                     // flat segment
    Ss.Limit             = (UINT32)~0;                            // 4GB limit
    Ss.Attributes.AsUInt = 0xC093;                                // G+DB+P+S+DPL0+Data
    SetGuestSs(&Ss);
 
    return TRUE;
}
IF (CS.L ≠ 1 ) or (IA32_EFER.LMA ≠ 1) or (IA32_EFER.SCE ≠ 1)
(* Not in 64-Bit Mode or SYSCALL/SYSRET not enabled in IA32_EFER *)
  THEN #UD; FI;
IF (CPL ≠ 0) OR (RCX is not canonical) THEN #GP(0); FI;
IF (operand size is 64-bit)
  THEN (* Return to 64-Bit Mode *)
    RIP ← RCX;
  ELSE (* Return to Compatibility Mode *)
    RIP ← ECX;
FI;
RFLAGS ← (R11 & 3C7FD7H) | 2; (* Clear RF, VM, reserved bits; set bit 2 *)
IF (operand size is 64-bit)
  THEN CS.Selector ← IA32_STAR[63:48]+16;
  ELSE CS.Selector ← IA32_STAR[63:48];
FI;
CS.Selector ← CS.Selector OR 3; (* RPL forced to 3 *)
(* Set rest of CS to a fixed value *)
CS.Base ← 0; (* Flat segment *)
CS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
CS.Type 11; (* Execute/read code, accessed *)
CS.S ← 1;
CS.DPL ← 3;
CS.P ← 1;
IF (operand size is 64-bit)
  THEN (* Return to 64-Bit Mode *)
    CS.L ← 1; (* 64-bit code segment *)
    CS.D ← 0; (* Required if CS.L = 1 *)
  ELSE (* Return to Compatibility Mode *)
    CS.L ← 0; (* Compatibility mode *)
    CS.D ← 1; (* 32-bit code segment *)
FI;
CS.G ← 1; (* 4-KByte granularity *)
CPL ← 3;
SS.Selector ← (IA32_STAR[63:48]+8) OR 3; (* RPL forced to 3 *)
(* Set rest of SS to a fixed value *)
SS.Base ← 0; (* Flat segment *)
SS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
SS.Type 3; (* Read/write data, accessed *)
SS.S ← 1;
SS.DPL ← 3;
SS.P ← 1;
SS.B ← 1; (* 32-bit stack segment*)
SS.G ← 1; (* 4-KByte granularity *)
IF (CS.L ≠ 1 ) or (IA32_EFER.LMA ≠ 1) or (IA32_EFER.SCE ≠ 1)
(* Not in 64-Bit Mode or SYSCALL/SYSRET not enabled in IA32_EFER *)
  THEN #UD; FI;
IF (CPL ≠ 0) OR (RCX is not canonical) THEN #GP(0); FI;
IF (operand size is 64-bit)
  THEN (* Return to 64-Bit Mode *)
    RIP ← RCX;
  ELSE (* Return to Compatibility Mode *)
    RIP ← ECX;
FI;
RFLAGS ← (R11 & 3C7FD7H) | 2; (* Clear RF, VM, reserved bits; set bit 2 *)
IF (operand size is 64-bit)
  THEN CS.Selector ← IA32_STAR[63:48]+16;
  ELSE CS.Selector ← IA32_STAR[63:48];
FI;
CS.Selector ← CS.Selector OR 3; (* RPL forced to 3 *)
(* Set rest of CS to a fixed value *)
CS.Base ← 0; (* Flat segment *)
CS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
CS.Type 11; (* Execute/read code, accessed *)
CS.S ← 1;
CS.DPL ← 3;
CS.P ← 1;
IF (operand size is 64-bit)
  THEN (* Return to 64-Bit Mode *)
    CS.L ← 1; (* 64-bit code segment *)
    CS.D ← 0; (* Required if CS.L = 1 *)
  ELSE (* Return to Compatibility Mode *)
    CS.L ← 0; (* Compatibility mode *)
    CS.D ← 1; (* 32-bit code segment *)
FI;
CS.G ← 1; (* 4-KByte granularity *)
CPL ← 3;
SS.Selector ← (IA32_STAR[63:48]+8) OR 3; (* RPL forced to 3 *)
(* Set rest of SS to a fixed value *)
SS.Base ← 0; (* Flat segment *)
SS.Limit ← FFFFFH; (* With 4-KByte granularity, implies a 4-GByte limit *)
SS.Type 3; (* Read/write data, accessed *)
SS.S ← 1;
SS.DPL ← 3;
SS.P ← 1;
SS.B ← 1; (* 32-bit stack segment*)
SS.G ← 1; (* 4-KByte granularity *)
_Use_decl_annotations_
BOOLEAN
SyscallHookEmulateSYSRET(PGUEST_REGS Regs)
{
    VMX_SEGMENT_SELECTOR Cs, Ss;
    UINT64               MsrValue;
    UINT64               GuestRip;
    UINT64               GuestRflags;
 
    //
    // Load RIP from RCX
    //
    GuestRip = Regs->rcx;
    __vmx_vmwrite(VMCS_GUEST_RIP, GuestRip);
 
    //
    // Load RFLAGS from R11. Clear RF, VM, reserved bits
    //
    GuestRflags = (Regs->r11 & ~(X86_FLAGS_RF | X86_FLAGS_VM | X86_FLAGS_RESERVED_BITS)) | X86_FLAGS_FIXED;
    __vmx_vmwrite(VMCS_GUEST_RFLAGS, GuestRflags);
 
    //
    // SYSRET loads the CS and SS selectors with values derived from bits 63:48 of IA32_STAR
    //
    MsrValue             = __readmsr(IA32_STAR);
    Cs.Selector          = (UINT16)(((MsrValue >> 48) + 16) | 3); // (STAR[63:48]+16) | 3 (* RPL forced to 3 *)
    Cs.Base              = 0;                                     // Flat segment
    Cs.Limit             = (UINT32)~0;                            // 4GB limit
    Cs.Attributes.AsUInt = 0xA0FB;                                // L+DB+P+S+DPL3+Code
    SetGuestCs(&Cs);
 
    Ss.Selector          = (UINT16)(((MsrValue >> 48) + 8) | 3); // (STAR[63:48]+8) | 3 (* RPL forced to 3 *)
    Ss.Base              = 0;                                    // Flat segment
    Ss.Limit             = (UINT32)~0;                           // 4GB limit
    Ss.Attributes.AsUInt = 0xC0F3;                               // G+DB+P+S+DPL3+Data
    SetGuestSs(&Ss);
 
    return TRUE;
}
_Use_decl_annotations_
BOOLEAN
SyscallHookEmulateSYSRET(PGUEST_REGS Regs)
{
    VMX_SEGMENT_SELECTOR Cs, Ss;
    UINT64               MsrValue;
    UINT64               GuestRip;
    UINT64               GuestRflags;
 
    //
    // Load RIP from RCX
    //
    GuestRip = Regs->rcx;
    __vmx_vmwrite(VMCS_GUEST_RIP, GuestRip);
 
    //
    // Load RFLAGS from R11. Clear RF, VM, reserved bits
    //
    GuestRflags = (Regs->r11 & ~(X86_FLAGS_RF | X86_FLAGS_VM | X86_FLAGS_RESERVED_BITS)) | X86_FLAGS_FIXED;
    __vmx_vmwrite(VMCS_GUEST_RFLAGS, GuestRflags);
 
    //
    // SYSRET loads the CS and SS selectors with values derived from bits 63:48 of IA32_STAR
    //
    MsrValue             = __readmsr(IA32_STAR);
    Cs.Selector          = (UINT16)(((MsrValue >> 48) + 16) | 3); // (STAR[63:48]+16) | 3 (* RPL forced to 3 *)
    Cs.Base              = 0;                                     // Flat segment
    Cs.Limit             = (UINT32)~0;                            // 4GB limit
    Cs.Attributes.AsUInt = 0xA0FB;                                // L+DB+P+S+DPL3+Code
    SetGuestCs(&Cs);
 
    Ss.Selector          = (UINT16)(((MsrValue >> 48) + 8) | 3); // (STAR[63:48]+8) | 3 (* RPL forced to 3 *)
    Ss.Base              = 0;                                    // Flat segment
    Ss.Limit             = (UINT32)~0;                           // 4GB limit
    Ss.Attributes.AsUInt = 0xC0F3;                               // G+DB+P+S+DPL3+Data
    SetGuestSs(&Ss);
 
    return TRUE;
}
VOID
SyscallHookConfigureEFER(BOOLEAN EnableEFERSyscallHook)
{
    IA32_EFER_REGISTER      MsrValue;
    IA32_VMX_BASIC_REGISTER VmxBasicMsr     = {0};
    UINT32                  VmEntryControls = 0;
    UINT32                  VmExitControls  = 0;
 
    //
    // Reading IA32_VMX_BASIC_MSR
    //
    VmxBasicMsr.AsUInt = __readmsr(IA32_VMX_BASIC);
 
    //
    // Read previous VM-Entry and VM-Exit controls
    //
    __vmx_vmread(VMCS_CTRL_VMENTRY_CONTROLS, &VmEntryControls);
    __vmx_vmread(VMCS_CTRL_PRIMARY_VMEXIT_CONTROLS, &VmExitControls);
 
    MsrValue.AsUInt = __readmsr(IA32_EFER);
 
    if (EnableEFERSyscallHook)
    {
        MsrValue.SyscallEnable = FALSE;
 
        //
        // Set VM-Entry controls to load EFER
        //
        __vmx_vmwrite(VMCS_CTRL_VMENTRY_CONTROLS, HvAdjustControls(VmEntryControls | VM_ENTRY_LOAD_IA32_EFER, VmxBasicMsr.VmxControls ? IA32_VMX_TRUE_ENTRY_CTLS : IA32_VMX_ENTRY_CTLS));
 
        //
        // Set VM-Exit controls to save EFER
        //
        __vmx_vmwrite(VMCS_CTRL_PRIMARY_VMEXIT_CONTROLS, HvAdjustControls(VmExitControls | VM_EXIT_SAVE_IA32_EFER, VmxBasicMsr.VmxControls ? IA32_VMX_TRUE_EXIT_CTLS : IA32_VMX_EXIT_CTLS));
 
        //
        // Set the GUEST EFER to use this value as the EFER
        //
        __vmx_vmwrite(VMCS_GUEST_EFER, MsrValue.AsUInt);
 
        //
        // also, we have to set exception bitmap to cause vm-exit on #UDs
        //
        HvSetExceptionBitmap(EXCEPTION_VECTOR_UNDEFINED_OPCODE);
    }
    else
    {
        MsrValue.SyscallEnable = TRUE;
 
        //
        // Set VM-Entry controls to load EFER
        //
        __vmx_vmwrite(VMCS_CTRL_VMENTRY_CONTROLS, HvAdjustControls(VmEntryControls & ~VM_ENTRY_LOAD_IA32_EFER, VmxBasicMsr.VmxControls ? IA32_VMX_TRUE_ENTRY_CTLS : IA32_VMX_ENTRY_CTLS));
 
        //
        // Set VM-Exit controls to save EFER
        //
        __vmx_vmwrite(VMCS_CTRL_PRIMARY_VMEXIT_CONTROLS, HvAdjustControls(VmExitControls & ~VM_EXIT_SAVE_IA32_EFER, VmxBasicMsr.VmxControls ? IA32_VMX_TRUE_EXIT_CTLS : IA32_VMX_EXIT_CTLS));
 
        //
        // Set the GUEST EFER to use this value as the EFER
        //
        __vmx_vmwrite(VMCS_GUEST_EFER, MsrValue.AsUInt);
 
        //
        // Because we're not save or load EFER on vm-exits so
        // we have to set it manually
        //
        __writemsr(IA32_EFER, MsrValue.AsUInt);
 
        //
        // unset the exception to not cause vm-exit on #UDs
        //
        ProtectedHvRemoveUndefinedInstructionForDisablingSyscallSysretCommands();
    }
}
VOID
SyscallHookConfigureEFER(BOOLEAN EnableEFERSyscallHook)
{
    IA32_EFER_REGISTER      MsrValue;
    IA32_VMX_BASIC_REGISTER VmxBasicMsr     = {0};
    UINT32                  VmEntryControls = 0;
    UINT32                  VmExitControls  = 0;
 
    //
    // Reading IA32_VMX_BASIC_MSR
    //
    VmxBasicMsr.AsUInt = __readmsr(IA32_VMX_BASIC);
 
    //
    // Read previous VM-Entry and VM-Exit controls
    //
    __vmx_vmread(VMCS_CTRL_VMENTRY_CONTROLS, &VmEntryControls);
    __vmx_vmread(VMCS_CTRL_PRIMARY_VMEXIT_CONTROLS, &VmExitControls);
 
    MsrValue.AsUInt = __readmsr(IA32_EFER);
 
    if (EnableEFERSyscallHook)
    {
        MsrValue.SyscallEnable = FALSE;

[培训]内核驱动高级班,冲击BAT一流互联网大厂工作,每周日13:00-18:00直播授课

最后于 2022-11-4 22:19 被N1ptune编辑 ,原因: 修改格式
收藏
免费 9
支持
分享
最新回复 (4)
雪    币: 2492
活跃值: (1685)
能力值: ( LV2,RANK:10 )
在线值:
发帖
回帖
粉丝
2

执行性能如何?

ps:https://revers.engineering/syscall-hooking-via-extended-feature-enable-register-efer/

最后于 2022-11-6 17:10 被maxwudi编辑 ,原因:
2022-11-6 17:01
0
雪    币: 1829
活跃值: (4045)
能力值: ( LV2,RANK:10 )
在线值:
发帖
回帖
粉丝
3
mark
2022-11-10 09:51
0
雪    币: 137
活跃值: (1380)
能力值: ( LV2,RANK:10 )
在线值:
发帖
回帖
粉丝
4
amd似乎对syscall这块有更多逻辑.
2022-11-10 16:17
0
雪    币: 12
活跃值: (409)
能力值: ( LV2,RANK:10 )
在线值:
发帖
回帖
粉丝
5
在VM试了一下,貌似不稳定,接近500~1000万次syscall的时候会引起异常炸掉
2022-12-1 11:47
0
游客
登录 | 注册 方可回帖
返回
//