UefiCpuPkg/MpInitLib: split wake up buffer into two parts

If PcdDxeNxMemoryProtectionPolicy is set to enable protection for memory
of EfiBootServicesCode, EfiConventionalMemory, the BIOS will hang at a page
fault exception during MP initialization.

The root cause is that the AP wake up buffer, which is below 1MB and used
to hold both AP init code and data, is type of EfiConventionalMemory (not
really allocated because of potential conflict with legacy code), and is
marked as non-executable. During the transition from real address mode
to long mode, the AP init code has to enable paging which will then cause
itself a page fault exception because it's just running in non-executable
memory.

The solution is splitting AP wake up buffer into two part: lower part is
still below 1MB and shared with legacy system, higher part is really
allocated memory of BootServicesCode type. The init code in the memory
below 1MB will not enable paging but just switch to protected mode and
jump to higher memory, in which the init code will enable paging and
switch to long mode.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

UefiCpuPkg/Library/MpInitLib/DxeMpLib.c

@@ -113,6 +113,40 @@ GetWakeupBuffer (
   return (UINTN) StartAddress;
 }
 
+/**
+  Get available EfiBootServicesCode memory below 4GB by specified size.
+
+  This buffer is required to safely transfer AP from real address mode to
+  protected mode or long mode, due to the fact that the buffer returned by
+  GetWakeupBuffer() may be marked as non-executable.
+
+  @param[in] BufferSize   Wakeup transition buffer size.
+
+  @retval other   Return wakeup transition buffer address below 4GB.
+  @retval 0       Cannot find free memory below 4GB.
+**/
+UINTN
+GetModeTransitionBuffer (
+  IN UINTN                BufferSize
+  )
+{
+  EFI_STATUS              Status;
+  EFI_PHYSICAL_ADDRESS    StartAddress;
+
+  StartAddress = BASE_4GB - 1;
+  Status = gBS->AllocatePages (
+                  AllocateMaxAddress,
+                  EfiBootServicesCode,
+                  EFI_SIZE_TO_PAGES (BufferSize),
+                  &StartAddress
+                  );
+  if (EFI_ERROR (Status)) {
+    StartAddress = 0;
+  }
+
+  return (UINTN)StartAddress;
+}
+
 /**
   Checks APs status and updates APs status if needed.
 

UefiCpuPkg/Library/MpInitLib/Ia32/MpEqu.inc

@@ -41,4 +41,9 @@ Cr3Location                   equ        LockLocation + 34h
 InitFlagLocation              equ        LockLocation + 38h
 CpuInfoLocation               equ        LockLocation + 3Ch
 NumApsExecutingLocation       equ        LockLocation + 40h
+InitializeFloatingPointUnitsAddress equ  LockLocation + 48h
+ModeTransitionMemoryLocation        equ  LockLocation + 4Ch
+ModeTransitionSegmentLocation       equ  LockLocation + 50h
+ModeHighMemoryLocation              equ  LockLocation + 52h
+ModeHighSegmentLocation             equ  LockLocation + 56h
 

UefiCpuPkg/Library/MpInitLib/Ia32/MpFuncs.nasm

@@ -48,34 +48,35 @@ BITS 16
     mov        si,  BufferStartLocation
     mov        ebx, [si]
 
-    mov        si,  ModeOffsetLocation
-    mov        eax, [si]
-    mov        si,  CodeSegmentLocation
-    mov        edx, [si]
-    mov        di,  ax
-    sub        di,  02h
-    mov        [di], dx
-    sub        di,  04h
-    add        eax, ebx
-    mov        [di],eax
-
     mov        si,  DataSegmentLocation
     mov        edx, [si]
 
+    ;
+    ; Get start address of 32-bit code in low memory (<1MB)
+    ;
+    mov        edi, ModeTransitionMemoryLocation
+
     mov        si, GdtrLocation
 o32 lgdt       [cs:si]
 
     mov        si, IdtrLocation
 o32 lidt       [cs:si]
 
-    xor        ax,  ax
+    ;
-    mov        ds,  ax
+    ; Switch to protected mode
-
+    ;
     mov        eax, cr0                        ; Get control register 0
     or         eax, 000000003h                 ; Set PE bit (bit #0) & MP
     mov        cr0, eax
 
-    jmp        0:strict dword 0                ; far jump to protected mode
+    ; Switch to 32-bit code in executable memory (>1MB)
+o32 jmp far    [cs:di]
+
+;
+; Following code may be copied to memory with type of EfiBootServicesCode.
+; This is required at DXE phase if NX is enabled for EfiBootServicesCode of
+; memory.
+;
 BITS 32
 Flat32Start:                                   ; protected mode entry point
     mov        ds, dx
@@ -266,6 +267,7 @@ ASM_PFX(AsmGetAddressMap):
     mov        dword [ebx +  8h], RendezvousFunnelProcEnd - RendezvousFunnelProcStart
     mov        dword [ebx + 0Ch], AsmRelocateApLoopStart
     mov        dword [ebx + 10h], AsmRelocateApLoopEnd - AsmRelocateApLoopStart
+    mov        dword [ebx + 14h], Flat32Start - RendezvousFunnelProcStart
 
     popad
     ret

UefiCpuPkg/Library/MpInitLib/MpLib.c

@@ -772,6 +772,8 @@ FillExchangeInfoData (
   )
 {
   volatile MP_CPU_EXCHANGE_INFO    *ExchangeInfo;
+  UINTN                            Size;
+  IA32_SEGMENT_DESCRIPTOR          *Selector;
 
   ExchangeInfo                  = CpuMpData->MpCpuExchangeInfo;
   ExchangeInfo->Lock            = 0;
@@ -801,6 +803,44 @@ FillExchangeInfoData (
   //
   AsmReadGdtr ((IA32_DESCRIPTOR *) &ExchangeInfo->GdtrProfile);
   AsmReadIdtr ((IA32_DESCRIPTOR *) &ExchangeInfo->IdtrProfile);
+
+  //
+  // Find a 32-bit code segment
+  //
+  Selector = (IA32_SEGMENT_DESCRIPTOR *)ExchangeInfo->GdtrProfile.Base;
+  Size = ExchangeInfo->GdtrProfile.Limit + 1;
+  while (Size > 0) {
+    if (Selector->Bits.L == 0 && Selector->Bits.Type >= 8) {
+      ExchangeInfo->ModeTransitionSegment =
+        (UINT16)((UINTN)Selector - ExchangeInfo->GdtrProfile.Base);
+      break;
+    }
+    Selector += 1;
+    Size -= sizeof (IA32_SEGMENT_DESCRIPTOR);
+  }
+
+  //
+  // Copy all 32-bit code and 64-bit code into memory with type of
+  // EfiBootServicesCode to avoid page fault if NX memory protection is enabled.
+  //
+  if (ExchangeInfo->ModeTransitionMemory != 0) {
+    Size = CpuMpData->AddressMap.RendezvousFunnelSize -
+           CpuMpData->AddressMap.ModeTransitionOffset;
+    CopyMem (
+      (VOID *)(UINTN)ExchangeInfo->ModeTransitionMemory,
+      CpuMpData->AddressMap.RendezvousFunnelAddress +
+      CpuMpData->AddressMap.ModeTransitionOffset,
+      Size
+      );
+
+    ExchangeInfo->ModeHighMemory = ExchangeInfo->ModeTransitionMemory;
+    ExchangeInfo->ModeHighMemory += (UINT32)ExchangeInfo->ModeOffset -
+               (UINT32)CpuMpData->AddressMap.ModeTransitionOffset;
+    ExchangeInfo->ModeHighSegment = (UINT16)ExchangeInfo->CodeSegment;
+  } else {
+    ExchangeInfo->ModeTransitionMemory = (UINT32)
+      (ExchangeInfo->BufferStart + CpuMpData->AddressMap.ModeTransitionOffset);
+  }
 }
 
 /**
@@ -876,6 +916,11 @@ AllocateResetVector (
     CpuMpData->WakeupBuffer      = GetWakeupBuffer (ApResetVectorSize);
     CpuMpData->MpCpuExchangeInfo = (MP_CPU_EXCHANGE_INFO *) (UINTN)
                     (CpuMpData->WakeupBuffer + CpuMpData->AddressMap.RendezvousFunnelSize);
+    CpuMpData->MpCpuExchangeInfo->ModeTransitionMemory = (UINT32)
+                    GetModeTransitionBuffer (
+                      CpuMpData->AddressMap.RendezvousFunnelSize -
+                      CpuMpData->AddressMap.ModeTransitionOffset
+                      );
   }
   BackupAndPrepareWakeupBuffer (CpuMpData);
 }

UefiCpuPkg/Library/MpInitLib/MpLib.h

@@ -152,6 +152,7 @@ typedef struct {
   UINTN             RendezvousFunnelSize;
   UINT8             *RelocateApLoopFuncAddress;
   UINTN             RelocateApLoopFuncSize;
+  UINTN             ModeTransitionOffset;
 } MP_ASSEMBLY_ADDRESS_MAP;
 
 typedef struct _CPU_MP_DATA  CPU_MP_DATA;
@@ -182,6 +183,10 @@ typedef struct {
   UINTN                 NumApsExecuting;
   CPU_MP_DATA           *CpuMpData;
   UINTN                 InitializeFloatingPointUnitsAddress;
+  UINT32                ModeTransitionMemory;
+  UINT16                ModeTransitionSegment;
+  UINT32                ModeHighMemory;
+  UINT16                ModeHighSegment;
 } MP_CPU_EXCHANGE_INFO;
 
 #pragma pack()
@@ -329,6 +334,23 @@ GetWakeupBuffer (
   IN UINTN                WakeupBufferSize
   );
 
+/**
+  Get available EfiBootServicesCode memory below 4GB by specified size.
+
+  This buffer is required to safely transfer AP from real address mode to
+  protected mode or long mode, due to the fact that the buffer returned by
+  GetWakeupBuffer() may be marked as non-executable.
+
+  @param[in] BufferSize   Wakeup transition buffer size.
+
+  @retval other   Return wakeup transition buffer address below 4GB.
+  @retval 0       Cannot find free memory below 4GB.
+**/
+UINTN
+GetModeTransitionBuffer (
+  IN UINTN                BufferSize
+  );
+
 /**
   This function will be called by BSP to wakeup AP.
 

UefiCpuPkg/Library/MpInitLib/PeiMpLib.c

@@ -187,6 +187,29 @@ GetWakeupBuffer (
   return (UINTN) -1;
 }
 
+/**
+  Get available EfiBootServicesCode memory below 4GB by specified size.
+
+  This buffer is required to safely transfer AP from real address mode to
+  protected mode or long mode, due to the fact that the buffer returned by
+  GetWakeupBuffer() may be marked as non-executable.
+
+  @param[in] BufferSize   Wakeup transition buffer size.
+
+  @retval other   Return wakeup transition buffer address below 4GB.
+  @retval 0       Cannot find free memory below 4GB.
+**/
+UINTN
+GetModeTransitionBuffer (
+  IN UINTN                BufferSize
+  )
+{
+  //
+  // PEI phase doesn't need to do such transition. So simply return 0.
+  //
+  return 0;
+}
+
 /**
   Checks APs status and updates APs status if needed.
 

UefiCpuPkg/Library/MpInitLib/X64/MpEqu.inc

@@ -42,4 +42,7 @@ InitFlagLocation              equ        LockLocation + 6Ch
 CpuInfoLocation               equ        LockLocation + 74h
 NumApsExecutingLocation       equ        LockLocation + 7Ch
 InitializeFloatingPointUnitsAddress equ  LockLocation + 8Ch
-
+ModeTransitionMemoryLocation        equ  LockLocation + 94h
+ModeTransitionSegmentLocation       equ  LockLocation + 98h
+ModeHighMemoryLocation              equ  LockLocation + 9Ah
+ModeHighSegmentLocation             equ  LockLocation + 9Eh

UefiCpuPkg/Library/MpInitLib/X64/MpFuncs.nasm

@@ -52,16 +52,13 @@ BITS 16
     mov        si,  BufferStartLocation
     mov        ebx, [si]
 
-    mov        di,  ModeOffsetLocation
+    mov        si,  DataSegmentLocation
-    mov        eax, [di]
+    mov        edx, [si]
-    mov        di,  CodeSegmentLocation
+
-    mov        edx, [di]
+    ;
-    mov        di,  ax
+    ; Get start address of 32-bit code in low memory (<1MB)
-    sub        di,  02h
+    ;
-    mov        [di],dx                         ; Patch long mode CS
+    mov        edi, ModeTransitionMemoryLocation
-    sub        di,  04h
-    add        eax, ebx
-    mov        [di],eax                        ; Patch address
 
     mov        si, GdtrLocation
 o32 lgdt       [cs:si]
@@ -69,56 +66,79 @@ o32 lgdt       [cs:si]
     mov        si, IdtrLocation
 o32 lidt       [cs:si]
 
-    mov        si, EnableExecuteDisableLocation
+    ;
-    cmp        byte [si], 0
+    ; Switch to protected mode
-    jz         SkipEnableExecuteDisableBit
+    ;
+    mov        eax, cr0                    ; Get control register 0
+    or         eax, 000000003h             ; Set PE bit (bit #0) & MP
+    mov        cr0, eax
+
+    ; Switch to 32-bit code (>1MB)
+o32 jmp far    [cs:di]
+
+;
+; Following code must be copied to memory with type of EfiBootServicesCode.
+; This is required if NX is enabled for EfiBootServicesCode of memory.
+;
+BITS 32
+Flat32Start:                                   ; protected mode entry point
+    mov        ds, dx
+    mov        es, dx
+    mov        fs, dx
+    mov        gs, dx
+    mov        ss, dx
 
     ;
     ; Enable execute disable bit
     ;
+    mov        esi, EnableExecuteDisableLocation
+    cmp        byte [ebx + esi], 0
+    jz         SkipEnableExecuteDisableBit
+
     mov        ecx, 0c0000080h             ; EFER MSR number
     rdmsr                                  ; Read EFER
     bts        eax, 11                     ; Enable Execute Disable Bit
     wrmsr                                  ; Write EFER
 
 SkipEnableExecuteDisableBit:
-
+    ;
-    mov        di,  DataSegmentLocation
+    ; Enable PAE
-    mov        edi, [di]                   ; Save long mode DS in edi
+    ;
-
-    mov        si, Cr3Location             ; Save CR3 in ecx
-    mov        ecx, [si]
-
-    xor        ax,  ax
-    mov        ds,  ax                     ; Clear data segment
-
-    mov        eax, cr0                    ; Get control register 0
-    or         eax, 000000003h             ; Set PE bit (bit #0) & MP
-    mov        cr0, eax
-
     mov        eax, cr4
     bts        eax, 5
     mov        cr4, eax
 
+    ;
+    ; Load page table
+    ;
+    mov        esi, Cr3Location             ; Save CR3 in ecx
+    mov        ecx, [ebx + esi]
     mov        cr3, ecx                    ; Load CR3
 
+    ;
+    ; Enable long mode
+    ;
     mov        ecx, 0c0000080h             ; EFER MSR number
     rdmsr                                  ; Read EFER
     bts        eax, 8                      ; Set LME=1
     wrmsr                                  ; Write EFER
 
+    ;
+    ; Enable paging
+    ;
     mov        eax, cr0                    ; Read CR0
     bts        eax, 31                     ; Set PG=1
     mov        cr0, eax                    ; Write CR0
 
-    jmp        0:strict dword 0  ; far jump to long mode
+    ;
+    ; Far jump to 64-bit code
+    ;
+    mov        edi, ModeHighMemoryLocation
+    add        edi, ebx
+    jmp far    [edi]
+
 BITS 64
 LongModeStart:
-    mov        eax, edi
-    mov        ds,  ax
-    mov        es,  ax
-    mov        ss,  ax
-
     mov        esi, ebx
     lea        edi, [esi + InitFlagLocation]
     cmp        qword [edi], 1       ; ApInitConfig
@@ -295,6 +315,7 @@ ASM_PFX(AsmGetAddressMap):
     lea        rax, [ASM_PFX(AsmRelocateApLoop)]
     mov        qword [rcx + 18h], rax
     mov        qword [rcx + 20h], AsmRelocateApLoopEnd - AsmRelocateApLoopStart
+    mov        qword [rcx + 28h], Flat32Start - RendezvousFunnelProcStart
     ret
 
 ;-------------------------------------------------------------------------------------