audk/UefiCpuPkg/PiSmmCpuDxeSmm/SmmProfile.c

1385 lines
40 KiB
C
Raw Normal View History

UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
/** @file
Enable SMM profile.
Copyright (c) 2012 - 2016, Intel Corporation. All rights reserved.<BR>
Copyright (c) 2017, AMD Incorporated. All rights reserved.<BR>
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
This program and the accompanying materials
are licensed and made available under the terms and conditions of the BSD License
which accompanies this distribution. The full text of the license may be found at
http://opensource.org/licenses/bsd-license.php
THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,
WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.
**/
#include "PiSmmCpuDxeSmm.h"
#include "SmmProfileInternal.h"
UINT32 mSmmProfileCr3;
SMM_PROFILE_HEADER *mSmmProfileBase;
MSR_DS_AREA_STRUCT *mMsrDsAreaBase;
//
// The buffer to store SMM profile data.
//
UINTN mSmmProfileSize;
//
// The buffer to enable branch trace store.
//
UINTN mMsrDsAreaSize = SMM_PROFILE_DTS_SIZE;
//
// The flag indicates if execute-disable is enabled on processor.
//
BOOLEAN mXdEnabled = FALSE;
//
// The flag indicates if BTS is supported by processor.
//
BOOLEAN mBtsSupported = TRUE;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
//
// The flag indicates if SMM profile starts to record data.
//
BOOLEAN mSmmProfileStart = FALSE;
//
// Record the page fault exception count for one instruction execution.
//
UINTN *mPFEntryCount;
UINT64 (*mLastPFEntryValue)[MAX_PF_ENTRY_COUNT];
UINT64 *(*mLastPFEntryPointer)[MAX_PF_ENTRY_COUNT];
MSR_DS_AREA_STRUCT **mMsrDsArea;
BRANCH_TRACE_RECORD **mMsrBTSRecord;
UINTN mBTSRecordNumber;
PEBS_RECORD **mMsrPEBSRecord;
//
// These memory ranges are always present, they does not generate the access type of page fault exception,
// but they possibly generate instruction fetch type of page fault exception.
//
MEMORY_PROTECTION_RANGE *mProtectionMemRange = NULL;
UINTN mProtectionMemRangeCount = 0;
//
// Some predefined memory ranges.
//
MEMORY_PROTECTION_RANGE mProtectionMemRangeTemplate[] = {
//
// SMRAM range (to be fixed in runtime).
// It is always present and instruction fetches are allowed.
//
{{0x00000000, 0x00000000},TRUE,FALSE},
//
// SMM profile data range( to be fixed in runtime).
// It is always present and instruction fetches are not allowed.
//
{{0x00000000, 0x00000000},TRUE,TRUE},
//
// Future extended range could be added here.
//
//
// PCI MMIO ranges (to be added in runtime).
// They are always present and instruction fetches are not allowed.
//
};
//
// These memory ranges are mapped by 4KB-page instead of 2MB-page.
//
MEMORY_RANGE *mSplitMemRange = NULL;
UINTN mSplitMemRangeCount = 0;
//
// SMI command port.
//
UINT32 mSmiCommandPort;
/**
Disable branch trace store.
**/
VOID
DisableBTS (
VOID
)
{
AsmMsrAnd64 (MSR_DEBUG_CTL, ~((UINT64)(MSR_DEBUG_CTL_BTS | MSR_DEBUG_CTL_TR)));
}
/**
Enable branch trace store.
**/
VOID
EnableBTS (
VOID
)
{
AsmMsrOr64 (MSR_DEBUG_CTL, (MSR_DEBUG_CTL_BTS | MSR_DEBUG_CTL_TR));
}
/**
Get CPU Index from APIC ID.
**/
UINTN
GetCpuIndex (
VOID
)
{
UINTN Index;
UINT32 ApicId;
ApicId = GetApicId ();
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
for (Index = 0; Index < mMaxNumberOfCpus; Index++) {
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if (gSmmCpuPrivate->ProcessorInfo[Index].ProcessorId == ApicId) {
return Index;
}
}
ASSERT (FALSE);
return 0;
}
/**
Get the source of IP after execute-disable exception is triggered.
@param CpuIndex The index of CPU.
@param DestinationIP The destination address.
**/
UINT64
GetSourceFromDestinationOnBts (
UINTN CpuIndex,
UINT64 DestinationIP
)
{
BRANCH_TRACE_RECORD *CurrentBTSRecord;
UINTN Index;
BOOLEAN FirstMatch;
FirstMatch = FALSE;
CurrentBTSRecord = (BRANCH_TRACE_RECORD *)mMsrDsArea[CpuIndex]->BTSIndex;
for (Index = 0; Index < mBTSRecordNumber; Index++) {
if ((UINTN)CurrentBTSRecord < (UINTN)mMsrBTSRecord[CpuIndex]) {
//
// Underflow
//
CurrentBTSRecord = (BRANCH_TRACE_RECORD *)((UINTN)mMsrDsArea[CpuIndex]->BTSAbsoluteMaximum - 1);
CurrentBTSRecord --;
}
if (CurrentBTSRecord->LastBranchTo == DestinationIP) {
//
// Good! find 1st one, then find 2nd one.
//
if (!FirstMatch) {
//
// The first one is DEBUG exception
//
FirstMatch = TRUE;
} else {
//
// Good find proper one.
//
return CurrentBTSRecord->LastBranchFrom;
}
}
CurrentBTSRecord--;
}
return 0;
}
/**
SMM profile specific INT 1 (single-step) exception handler.
@param InterruptType Defines the type of interrupt or exception that
occurred on the processor.This parameter is processor architecture specific.
@param SystemContext A pointer to the processor context when
the interrupt occurred on the processor.
**/
VOID
EFIAPI
DebugExceptionHandler (
IN EFI_EXCEPTION_TYPE InterruptType,
IN EFI_SYSTEM_CONTEXT SystemContext
)
{
UINTN CpuIndex;
UINTN PFEntry;
if (!mSmmProfileStart) {
return;
}
CpuIndex = GetCpuIndex ();
//
// Clear last PF entries
//
for (PFEntry = 0; PFEntry < mPFEntryCount[CpuIndex]; PFEntry++) {
*mLastPFEntryPointer[CpuIndex][PFEntry] = mLastPFEntryValue[CpuIndex][PFEntry];
}
//
// Reset page fault exception count for next page fault.
//
mPFEntryCount[CpuIndex] = 0;
//
// Flush TLB
//
CpuFlushTlb ();
//
// Clear TF in EFLAGS
//
ClearTrapFlag (SystemContext);
}
/**
Check if the memory address will be mapped by 4KB-page.
@param Address The address of Memory.
@param Nx The flag indicates if the memory is execute-disable.
**/
BOOLEAN
IsAddressValid (
IN EFI_PHYSICAL_ADDRESS Address,
IN BOOLEAN *Nx
)
{
UINTN Index;
*Nx = FALSE;
if (FeaturePcdGet (PcdCpuSmmProfileEnable)) {
//
// Check configuration
//
for (Index = 0; Index < mProtectionMemRangeCount; Index++) {
if ((Address >= mProtectionMemRange[Index].Range.Base) && (Address < mProtectionMemRange[Index].Range.Top)) {
*Nx = mProtectionMemRange[Index].Nx;
return mProtectionMemRange[Index].Present;
}
}
*Nx = TRUE;
return FALSE;
} else {
if ((Address < mCpuHotPlugData.SmrrBase) ||
(Address >= mCpuHotPlugData.SmrrBase + mCpuHotPlugData.SmrrSize)) {
*Nx = TRUE;
}
return TRUE;
}
}
/**
Check if the memory address will be mapped by 4KB-page.
@param Address The address of Memory.
**/
BOOLEAN
IsAddressSplit (
IN EFI_PHYSICAL_ADDRESS Address
)
{
UINTN Index;
if (FeaturePcdGet (PcdCpuSmmProfileEnable)) {
//
// Check configuration
//
for (Index = 0; Index < mSplitMemRangeCount; Index++) {
if ((Address >= mSplitMemRange[Index].Base) && (Address < mSplitMemRange[Index].Top)) {
return TRUE;
}
}
} else {
if (Address < mCpuHotPlugData.SmrrBase) {
if ((mCpuHotPlugData.SmrrBase - Address) < BASE_2MB) {
return TRUE;
}
} else if (Address > (mCpuHotPlugData.SmrrBase + mCpuHotPlugData.SmrrSize - BASE_2MB)) {
if ((Address - (mCpuHotPlugData.SmrrBase + mCpuHotPlugData.SmrrSize - BASE_2MB)) < BASE_2MB) {
return TRUE;
}
}
}
//
// Return default
//
return FALSE;
}
/**
Initialize the protected memory ranges and the 4KB-page mapped memory ranges.
**/
VOID
InitProtectedMemRange (
VOID
)
{
UINTN Index;
UINTN NumberOfDescriptors;
UINTN NumberOfMmioDescriptors;
UINTN NumberOfProtectRange;
UINTN NumberOfSpliteRange;
EFI_GCD_MEMORY_SPACE_DESCRIPTOR *MemorySpaceMap;
UINTN TotalSize;
EFI_PHYSICAL_ADDRESS ProtectBaseAddress;
EFI_PHYSICAL_ADDRESS ProtectEndAddress;
EFI_PHYSICAL_ADDRESS Top2MBAlignedAddress;
EFI_PHYSICAL_ADDRESS Base2MBAlignedAddress;
UINT64 High4KBPageSize;
UINT64 Low4KBPageSize;
NumberOfDescriptors = 0;
NumberOfMmioDescriptors = 0;
NumberOfSpliteRange = 0;
MemorySpaceMap = NULL;
//
// Get MMIO ranges from GCD and add them into protected memory ranges.
//
gDS->GetMemorySpaceMap (
&NumberOfDescriptors,
&MemorySpaceMap
);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
for (Index = 0; Index < NumberOfDescriptors; Index++) {
if (MemorySpaceMap[Index].GcdMemoryType == EfiGcdMemoryTypeMemoryMappedIo) {
NumberOfMmioDescriptors++;
}
}
if (NumberOfMmioDescriptors != 0) {
TotalSize = NumberOfMmioDescriptors * sizeof (MEMORY_PROTECTION_RANGE) + sizeof (mProtectionMemRangeTemplate);
mProtectionMemRange = (MEMORY_PROTECTION_RANGE *) AllocateZeroPool (TotalSize);
ASSERT (mProtectionMemRange != NULL);
mProtectionMemRangeCount = TotalSize / sizeof (MEMORY_PROTECTION_RANGE);
//
// Copy existing ranges.
//
CopyMem (mProtectionMemRange, mProtectionMemRangeTemplate, sizeof (mProtectionMemRangeTemplate));
//
// Create split ranges which come from protected ranges.
//
TotalSize = (TotalSize / sizeof (MEMORY_PROTECTION_RANGE)) * sizeof (MEMORY_RANGE);
mSplitMemRange = (MEMORY_RANGE *) AllocateZeroPool (TotalSize);
ASSERT (mSplitMemRange != NULL);
//
// Create MMIO ranges which are set to present and execution-disable.
//
NumberOfProtectRange = sizeof (mProtectionMemRangeTemplate) / sizeof (MEMORY_PROTECTION_RANGE);
for (Index = 0; Index < NumberOfDescriptors; Index++) {
if (MemorySpaceMap[Index].GcdMemoryType != EfiGcdMemoryTypeMemoryMappedIo) {
continue;
}
mProtectionMemRange[NumberOfProtectRange].Range.Base = MemorySpaceMap[Index].BaseAddress;
mProtectionMemRange[NumberOfProtectRange].Range.Top = MemorySpaceMap[Index].BaseAddress + MemorySpaceMap[Index].Length;
mProtectionMemRange[NumberOfProtectRange].Present = TRUE;
mProtectionMemRange[NumberOfProtectRange].Nx = TRUE;
NumberOfProtectRange++;
}
}
//
// According to protected ranges, create the ranges which will be mapped by 2KB page.
//
NumberOfSpliteRange = 0;
NumberOfProtectRange = mProtectionMemRangeCount;
for (Index = 0; Index < NumberOfProtectRange; Index++) {
//
// If MMIO base address is not 2MB alignment, make 2MB alignment for create 4KB page in page table.
//
ProtectBaseAddress = mProtectionMemRange[Index].Range.Base;
ProtectEndAddress = mProtectionMemRange[Index].Range.Top;
if (((ProtectBaseAddress & (SIZE_2MB - 1)) != 0) || ((ProtectEndAddress & (SIZE_2MB - 1)) != 0)) {
//
// Check if it is possible to create 4KB-page for not 2MB-aligned range and to create 2MB-page for 2MB-aligned range.
// A mix of 4KB and 2MB page could save SMRAM space.
//
Top2MBAlignedAddress = ProtectEndAddress & ~(SIZE_2MB - 1);
Base2MBAlignedAddress = (ProtectBaseAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1);
if ((Top2MBAlignedAddress > Base2MBAlignedAddress) &&
((Top2MBAlignedAddress - Base2MBAlignedAddress) >= SIZE_2MB)) {
//
// There is an range which could be mapped by 2MB-page.
//
High4KBPageSize = ((ProtectEndAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1)) - (ProtectEndAddress & ~(SIZE_2MB - 1));
Low4KBPageSize = ((ProtectBaseAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1)) - (ProtectBaseAddress & ~(SIZE_2MB - 1));
if (High4KBPageSize != 0) {
//
// Add not 2MB-aligned range to be mapped by 4KB-page.
//
mSplitMemRange[NumberOfSpliteRange].Base = ProtectEndAddress & ~(SIZE_2MB - 1);
mSplitMemRange[NumberOfSpliteRange].Top = (ProtectEndAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1);
NumberOfSpliteRange++;
}
if (Low4KBPageSize != 0) {
//
// Add not 2MB-aligned range to be mapped by 4KB-page.
//
mSplitMemRange[NumberOfSpliteRange].Base = ProtectBaseAddress & ~(SIZE_2MB - 1);
mSplitMemRange[NumberOfSpliteRange].Top = (ProtectBaseAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1);
NumberOfSpliteRange++;
}
} else {
//
// The range could only be mapped by 4KB-page.
//
mSplitMemRange[NumberOfSpliteRange].Base = ProtectBaseAddress & ~(SIZE_2MB - 1);
mSplitMemRange[NumberOfSpliteRange].Top = (ProtectEndAddress + SIZE_2MB - 1) & ~(SIZE_2MB - 1);
NumberOfSpliteRange++;
}
}
}
mSplitMemRangeCount = NumberOfSpliteRange;
DEBUG ((EFI_D_INFO, "SMM Profile Memory Ranges:\n"));
for (Index = 0; Index < mProtectionMemRangeCount; Index++) {
DEBUG ((EFI_D_INFO, "mProtectionMemRange[%d].Base = %lx\n", Index, mProtectionMemRange[Index].Range.Base));
DEBUG ((EFI_D_INFO, "mProtectionMemRange[%d].Top = %lx\n", Index, mProtectionMemRange[Index].Range.Top));
}
for (Index = 0; Index < mSplitMemRangeCount; Index++) {
DEBUG ((EFI_D_INFO, "mSplitMemRange[%d].Base = %lx\n", Index, mSplitMemRange[Index].Base));
DEBUG ((EFI_D_INFO, "mSplitMemRange[%d].Top = %lx\n", Index, mSplitMemRange[Index].Top));
}
}
/**
Update page table according to protected memory ranges and the 4KB-page mapped memory ranges.
**/
VOID
InitPaging (
VOID
)
{
UINT64 *Pml4;
UINT64 *Pde;
UINT64 *Pte;
UINT64 *Pt;
UINTN Address;
UINTN Level1;
UINTN Level2;
UINTN Level3;
UINTN Level4;
UINTN NumberOfPdpEntries;
UINTN NumberOfPml4Entries;
UINTN SizeOfMemorySpace;
BOOLEAN Nx;
if (sizeof (UINTN) == sizeof (UINT64)) {
Pml4 = (UINT64*)(UINTN)mSmmProfileCr3;
SizeOfMemorySpace = HighBitSet64 (gPhyMask) + 1;
//
// Calculate the table entries of PML4E and PDPTE.
//
if (SizeOfMemorySpace <= 39 ) {
NumberOfPml4Entries = 1;
NumberOfPdpEntries = (UINT32)LShiftU64 (1, (SizeOfMemorySpace - 30));
} else {
NumberOfPml4Entries = (UINT32)LShiftU64 (1, (SizeOfMemorySpace - 39));
NumberOfPdpEntries = 512;
}
} else {
NumberOfPml4Entries = 1;
NumberOfPdpEntries = 4;
}
//
// Go through page table and change 2MB-page into 4KB-page.
//
for (Level1 = 0; Level1 < NumberOfPml4Entries; Level1++) {
if (sizeof (UINTN) == sizeof (UINT64)) {
if ((Pml4[Level1] & IA32_PG_P) == 0) {
//
// If Pml4 entry does not exist, skip it
//
continue;
}
Pde = (UINT64 *)(UINTN)(Pml4[Level1] & ~mAddressEncMask & PHYSICAL_ADDRESS_MASK);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
} else {
Pde = (UINT64*)(UINTN)mSmmProfileCr3;
}
for (Level2 = 0; Level2 < NumberOfPdpEntries; Level2++, Pde++) {
if ((*Pde & IA32_PG_P) == 0) {
//
// If PDE entry does not exist, skip it
//
continue;
}
if ((*Pde & IA32_PG_PS) != 0) {
//
// This is 1G entry, skip it
//
continue;
}
Pte = (UINT64 *)(UINTN)(*Pde & ~mAddressEncMask & PHYSICAL_ADDRESS_MASK);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if (Pte == 0) {
continue;
}
for (Level3 = 0; Level3 < SIZE_4KB / sizeof (*Pte); Level3++, Pte++) {
if ((*Pte & IA32_PG_P) == 0) {
//
// If PTE entry does not exist, skip it
//
continue;
}
Address = (((Level2 << 9) + Level3) << 21);
//
// If it is 2M page, check IsAddressSplit()
//
if (((*Pte & IA32_PG_PS) != 0) && IsAddressSplit (Address)) {
//
// Based on current page table, create 4KB page table for split area.
//
ASSERT (Address == (*Pte & PHYSICAL_ADDRESS_MASK));
Pt = AllocatePageTableMemory (1);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (Pt != NULL);
// Split it
for (Level4 = 0; Level4 < SIZE_4KB / sizeof(*Pt); Level4++) {
Pt[Level4] = Address + ((Level4 << 12) | mAddressEncMask | PAGE_ATTRIBUTE_BITS);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
} // end for PT
*Pte = (UINT64)(UINTN)Pt | mAddressEncMask | PAGE_ATTRIBUTE_BITS;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
} // end if IsAddressSplit
} // end for PTE
} // end for PDE
}
//
// Go through page table and set several page table entries to absent or execute-disable.
//
DEBUG ((EFI_D_INFO, "Patch page table start ...\n"));
for (Level1 = 0; Level1 < NumberOfPml4Entries; Level1++) {
if (sizeof (UINTN) == sizeof (UINT64)) {
if ((Pml4[Level1] & IA32_PG_P) == 0) {
//
// If Pml4 entry does not exist, skip it
//
continue;
}
Pde = (UINT64 *)(UINTN)(Pml4[Level1] & ~mAddressEncMask & PHYSICAL_ADDRESS_MASK);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
} else {
Pde = (UINT64*)(UINTN)mSmmProfileCr3;
}
for (Level2 = 0; Level2 < NumberOfPdpEntries; Level2++, Pde++) {
if ((*Pde & IA32_PG_P) == 0) {
//
// If PDE entry does not exist, skip it
//
continue;
}
if ((*Pde & IA32_PG_PS) != 0) {
//
// This is 1G entry, set NX bit and skip it
//
if (mXdSupported) {
*Pde = *Pde | IA32_PG_NX;
}
continue;
}
Pte = (UINT64 *)(UINTN)(*Pde & ~mAddressEncMask & PHYSICAL_ADDRESS_MASK);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if (Pte == 0) {
continue;
}
for (Level3 = 0; Level3 < SIZE_4KB / sizeof (*Pte); Level3++, Pte++) {
if ((*Pte & IA32_PG_P) == 0) {
//
// If PTE entry does not exist, skip it
//
continue;
}
Address = (((Level2 << 9) + Level3) << 21);
if ((*Pte & IA32_PG_PS) != 0) {
// 2MB page
if (!IsAddressValid (Address, &Nx)) {
//
// Patch to remove Present flag and RW flag
//
*Pte = *Pte & (INTN)(INT32)(~PAGE_ATTRIBUTE_BITS);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
}
if (Nx && mXdSupported) {
*Pte = *Pte | IA32_PG_NX;
}
} else {
// 4KB page
Pt = (UINT64 *)(UINTN)(*Pte & ~mAddressEncMask & PHYSICAL_ADDRESS_MASK);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if (Pt == 0) {
continue;
}
for (Level4 = 0; Level4 < SIZE_4KB / sizeof(*Pt); Level4++, Pt++) {
if (!IsAddressValid (Address, &Nx)) {
*Pt = *Pt & (INTN)(INT32)(~PAGE_ATTRIBUTE_BITS);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
}
if (Nx && mXdSupported) {
*Pt = *Pt | IA32_PG_NX;
}
Address += SIZE_4KB;
} // end for PT
} // end if PS
} // end for PTE
} // end for PDE
}
//
// Flush TLB
//
CpuFlushTlb ();
DEBUG ((EFI_D_INFO, "Patch page table done!\n"));
//
// Set execute-disable flag
//
mXdEnabled = TRUE;
return ;
}
/**
To find FADT in ACPI tables.
@param AcpiTableGuid The GUID used to find ACPI table in UEFI ConfigurationTable.
@return FADT table pointer.
**/
EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *
FindAcpiFadtTableByAcpiGuid (
IN EFI_GUID *AcpiTableGuid
)
{
EFI_ACPI_2_0_ROOT_SYSTEM_DESCRIPTION_POINTER *Rsdp;
EFI_ACPI_DESCRIPTION_HEADER *Rsdt;
EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *Fadt;
UINTN Index;
UINT32 Data32;
Rsdp = NULL;
Rsdt = NULL;
Fadt = NULL;
//
// found ACPI table RSD_PTR from system table
//
for (Index = 0; Index < gST->NumberOfTableEntries; Index++) {
if (CompareGuid (&(gST->ConfigurationTable[Index].VendorGuid), AcpiTableGuid)) {
//
// A match was found.
//
Rsdp = gST->ConfigurationTable[Index].VendorTable;
break;
}
}
if (Rsdp == NULL) {
return NULL;
}
Rsdt = (EFI_ACPI_DESCRIPTION_HEADER *)(UINTN) Rsdp->RsdtAddress;
if (Rsdt == NULL || Rsdt->Signature != EFI_ACPI_2_0_ROOT_SYSTEM_DESCRIPTION_TABLE_SIGNATURE) {
return NULL;
}
for (Index = sizeof (EFI_ACPI_DESCRIPTION_HEADER); Index < Rsdt->Length; Index = Index + sizeof (UINT32)) {
Data32 = *(UINT32 *) ((UINT8 *) Rsdt + Index);
Fadt = (EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *) (UINT32 *) (UINTN) Data32;
if (Fadt->Header.Signature == EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE_SIGNATURE) {
break;
}
}
if (Fadt == NULL || Fadt->Header.Signature != EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE_SIGNATURE) {
return NULL;
}
return Fadt;
}
/**
To find FADT in ACPI tables.
@return FADT table pointer.
**/
EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *
FindAcpiFadtTable (
VOID
)
{
EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *Fadt;
Fadt = FindAcpiFadtTableByAcpiGuid (&gEfiAcpi20TableGuid);
if (Fadt != NULL) {
return Fadt;
}
return FindAcpiFadtTableByAcpiGuid (&gEfiAcpi10TableGuid);
}
/**
To get system port address of the SMI Command Port in FADT table.
**/
VOID
GetSmiCommandPort (
VOID
)
{
EFI_ACPI_2_0_FIXED_ACPI_DESCRIPTION_TABLE *Fadt;
Fadt = FindAcpiFadtTable ();
ASSERT (Fadt != NULL);
mSmiCommandPort = Fadt->SmiCmd;
DEBUG ((EFI_D_INFO, "mSmiCommandPort = %x\n", mSmiCommandPort));
}
/**
Updates page table to make some memory ranges (like system memory) absent
and make some memory ranges (like MMIO) present and execute disable. It also
update 2MB-page to 4KB-page for some memory ranges.
**/
VOID
SmmProfileStart (
VOID
)
{
//
// The flag indicates SMM profile starts to work.
//
mSmmProfileStart = TRUE;
}
/**
Initialize SMM profile in SmmReadyToLock protocol callback function.
@param Protocol Points to the protocol's unique identifier.
@param Interface Points to the interface instance.
@param Handle The handle on which the interface was installed.
@retval EFI_SUCCESS SmmReadyToLock protocol callback runs successfully.
**/
EFI_STATUS
EFIAPI
InitSmmProfileCallBack (
IN CONST EFI_GUID *Protocol,
IN VOID *Interface,
IN EFI_HANDLE Handle
)
{
//
// Save to variable so that SMM profile data can be found.
//
gRT->SetVariable (
SMM_PROFILE_NAME,
&gEfiCallerIdGuid,
EFI_VARIABLE_BOOTSERVICE_ACCESS | EFI_VARIABLE_RUNTIME_ACCESS,
sizeof(mSmmProfileBase),
&mSmmProfileBase
);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
//
// Get Software SMI from FADT
//
GetSmiCommandPort ();
//
// Initialize protected memory range for patching page table later.
//
InitProtectedMemRange ();
return EFI_SUCCESS;
}
/**
Initialize SMM profile data structures.
**/
VOID
InitSmmProfileInternal (
VOID
)
{
EFI_STATUS Status;
EFI_PHYSICAL_ADDRESS Base;
VOID *Registration;
UINTN Index;
UINTN MsrDsAreaSizePerCpu;
UINTN TotalSize;
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
mPFEntryCount = (UINTN *)AllocateZeroPool (sizeof (UINTN) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mPFEntryCount != NULL);
mLastPFEntryValue = (UINT64 (*)[MAX_PF_ENTRY_COUNT])AllocateZeroPool (
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
sizeof (mLastPFEntryValue[0]) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mLastPFEntryValue != NULL);
mLastPFEntryPointer = (UINT64 *(*)[MAX_PF_ENTRY_COUNT])AllocateZeroPool (
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
sizeof (mLastPFEntryPointer[0]) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mLastPFEntryPointer != NULL);
//
// Allocate memory for SmmProfile below 4GB.
// The base address
//
mSmmProfileSize = PcdGet32 (PcdCpuSmmProfileSize);
ASSERT ((mSmmProfileSize & 0xFFF) == 0);
if (mBtsSupported) {
TotalSize = mSmmProfileSize + mMsrDsAreaSize;
} else {
TotalSize = mSmmProfileSize;
}
Base = 0xFFFFFFFF;
Status = gBS->AllocatePages (
AllocateMaxAddress,
EfiReservedMemoryType,
EFI_SIZE_TO_PAGES (TotalSize),
&Base
);
ASSERT_EFI_ERROR (Status);
ZeroMem ((VOID *)(UINTN)Base, TotalSize);
mSmmProfileBase = (SMM_PROFILE_HEADER *)(UINTN)Base;
//
// Initialize SMM profile data header.
//
mSmmProfileBase->HeaderSize = sizeof (SMM_PROFILE_HEADER);
mSmmProfileBase->MaxDataEntries = (UINT64)((mSmmProfileSize - sizeof(SMM_PROFILE_HEADER)) / sizeof (SMM_PROFILE_ENTRY));
mSmmProfileBase->MaxDataSize = MultU64x64 (mSmmProfileBase->MaxDataEntries, sizeof(SMM_PROFILE_ENTRY));
mSmmProfileBase->CurDataEntries = 0;
mSmmProfileBase->CurDataSize = 0;
mSmmProfileBase->TsegStart = mCpuHotPlugData.SmrrBase;
mSmmProfileBase->TsegSize = mCpuHotPlugData.SmrrSize;
mSmmProfileBase->NumSmis = 0;
mSmmProfileBase->NumCpus = gSmmCpuPrivate->SmmCoreEntryContext.NumberOfCpus;
if (mBtsSupported) {
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
mMsrDsArea = (MSR_DS_AREA_STRUCT **)AllocateZeroPool (sizeof (MSR_DS_AREA_STRUCT *) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mMsrDsArea != NULL);
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
mMsrBTSRecord = (BRANCH_TRACE_RECORD **)AllocateZeroPool (sizeof (BRANCH_TRACE_RECORD *) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mMsrBTSRecord != NULL);
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
mMsrPEBSRecord = (PEBS_RECORD **)AllocateZeroPool (sizeof (PEBS_RECORD *) * mMaxNumberOfCpus);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
ASSERT (mMsrPEBSRecord != NULL);
mMsrDsAreaBase = (MSR_DS_AREA_STRUCT *)((UINTN)Base + mSmmProfileSize);
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
MsrDsAreaSizePerCpu = mMsrDsAreaSize / mMaxNumberOfCpus;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
mBTSRecordNumber = (MsrDsAreaSizePerCpu - sizeof(PEBS_RECORD) * PEBS_RECORD_NUMBER - sizeof(MSR_DS_AREA_STRUCT)) / sizeof(BRANCH_TRACE_RECORD);
UefiCpuPkg/PiSmmCpuDxeSmm: handle dynamic PcdCpuMaxLogicalProcessorNumber "UefiCpuPkg/UefiCpuPkg.dec" already allows platforms to make PcdCpuMaxLogicalProcessorNumber dynamic, however PiSmmCpuDxeSmm does not take this into account everywhere. As soon as a platform turns the PCD into a dynamic one, at least S3 fails. When the PCD is dynamic, all PcdGet() calls translate into PCD DXE protocol calls, which are only permitted at boot time, not at runtime or during S3 resume. We already have a variable called mMaxNumberOfCpus; it is initialized in the entry point function like this: > // > // If support CPU hot plug, we need to allocate resources for possibly > // hot-added processors > // > if (FeaturePcdGet (PcdCpuHotPlugSupport)) { > mMaxNumberOfCpus = PcdGet32 (PcdCpuMaxLogicalProcessorNumber); > } else { > mMaxNumberOfCpus = mNumberOfCpus; > } There's another use of the PCD a bit higher up, also in the entry point function: > // > // Use MP Services Protocol to retrieve the number of processors and > // number of enabled processors > // > Status = MpServices->GetNumberOfProcessors (MpServices, &mNumberOfCpus, > &NumberOfEnabledProcessors); > ASSERT_EFI_ERROR (Status); > ASSERT (mNumberOfCpus <= PcdGet32 (PcdCpuMaxLogicalProcessorNumber)); Preserve these calls in the entry point function, and replace all other uses of PcdCpuMaxLogicalProcessorNumber -- there are only reads -- with mMaxNumberOfCpus. For PcdCpuHotPlugSupport==TRUE, this is an unobservable change. For PcdCpuHotPlugSupport==FALSE, we even save SMRAM, because we no longer allocate resources needlessly for CPUs that can never appear in the system. PcdCpuMaxLogicalProcessorNumber is also retrieved in "UefiCpuPkg/Library/SmmCpuFeaturesLib/SmmCpuFeaturesLib.c", but only in the library instance constructor, which runs even before the entry point function is called. Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jeff Fan <jeff.fan@intel.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Bugzilla: https://bugzilla.tianocore.org/show_bug.cgi?id=116 Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
2016-11-24 20:49:43 +01:00
for (Index = 0; Index < mMaxNumberOfCpus; Index++) {
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
mMsrDsArea[Index] = (MSR_DS_AREA_STRUCT *)((UINTN)mMsrDsAreaBase + MsrDsAreaSizePerCpu * Index);
mMsrBTSRecord[Index] = (BRANCH_TRACE_RECORD *)((UINTN)mMsrDsArea[Index] + sizeof(MSR_DS_AREA_STRUCT));
mMsrPEBSRecord[Index] = (PEBS_RECORD *)((UINTN)mMsrDsArea[Index] + MsrDsAreaSizePerCpu - sizeof(PEBS_RECORD) * PEBS_RECORD_NUMBER);
mMsrDsArea[Index]->BTSBufferBase = (UINTN)mMsrBTSRecord[Index];
mMsrDsArea[Index]->BTSIndex = mMsrDsArea[Index]->BTSBufferBase;
mMsrDsArea[Index]->BTSAbsoluteMaximum = mMsrDsArea[Index]->BTSBufferBase + mBTSRecordNumber * sizeof(BRANCH_TRACE_RECORD) + 1;
mMsrDsArea[Index]->BTSInterruptThreshold = mMsrDsArea[Index]->BTSAbsoluteMaximum + 1;
mMsrDsArea[Index]->PEBSBufferBase = (UINTN)mMsrPEBSRecord[Index];
mMsrDsArea[Index]->PEBSIndex = mMsrDsArea[Index]->PEBSBufferBase;
mMsrDsArea[Index]->PEBSAbsoluteMaximum = mMsrDsArea[Index]->PEBSBufferBase + PEBS_RECORD_NUMBER * sizeof(PEBS_RECORD) + 1;
mMsrDsArea[Index]->PEBSInterruptThreshold = mMsrDsArea[Index]->PEBSAbsoluteMaximum + 1;
}
}
mProtectionMemRange = mProtectionMemRangeTemplate;
mProtectionMemRangeCount = sizeof (mProtectionMemRangeTemplate) / sizeof (MEMORY_PROTECTION_RANGE);
//
// Update TSeg entry.
//
mProtectionMemRange[0].Range.Base = mCpuHotPlugData.SmrrBase;
mProtectionMemRange[0].Range.Top = mCpuHotPlugData.SmrrBase + mCpuHotPlugData.SmrrSize;
//
// Update SMM profile entry.
//
mProtectionMemRange[1].Range.Base = (EFI_PHYSICAL_ADDRESS)(UINTN)mSmmProfileBase;
mProtectionMemRange[1].Range.Top = (EFI_PHYSICAL_ADDRESS)(UINTN)mSmmProfileBase + TotalSize;
//
// Allocate memory reserved for creating 4KB pages.
//
InitPagesForPFHandler ();
//
// Start SMM profile when SmmReadyToLock protocol is installed.
//
Status = gSmst->SmmRegisterProtocolNotify (
&gEfiSmmReadyToLockProtocolGuid,
InitSmmProfileCallBack,
&Registration
);
ASSERT_EFI_ERROR (Status);
return ;
}
/**
Check if XD feature is supported by a processor.
**/
VOID
CheckFeatureSupported (
VOID
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
)
{
UINT32 RegEax;
UINT32 RegEdx;
MSR_IA32_MISC_ENABLE_REGISTER MiscEnableMsr;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if (mXdSupported) {
AsmCpuid (CPUID_EXTENDED_FUNCTION, &RegEax, NULL, NULL, NULL);
if (RegEax <= CPUID_EXTENDED_FUNCTION) {
//
// Extended CPUID functions are not supported on this processor.
//
mXdSupported = FALSE;
}
AsmCpuid (CPUID_EXTENDED_CPU_SIG, NULL, NULL, NULL, &RegEdx);
if ((RegEdx & CPUID1_EDX_XD_SUPPORT) == 0) {
//
// Execute Disable Bit feature is not supported on this processor.
//
mXdSupported = FALSE;
}
}
if (mBtsSupported) {
AsmCpuid (CPUID_VERSION_INFO, NULL, NULL, NULL, &RegEdx);
if ((RegEdx & CPUID1_EDX_BTS_AVAILABLE) != 0) {
//
// Per IA32 manuals:
// When CPUID.1:EDX[21] is set, the following BTS facilities are available:
// 1. The BTS_UNAVAILABLE flag in the IA32_MISC_ENABLE MSR indicates the
// availability of the BTS facilities, including the ability to set the BTS and
// BTINT bits in the MSR_DEBUGCTLA MSR.
// 2. The IA32_DS_AREA MSR can be programmed to point to the DS save area.
//
MiscEnableMsr.Uint64 = AsmReadMsr64 (MSR_IA32_MISC_ENABLE);
if (MiscEnableMsr.Bits.BTS == 1) {
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
//
// BTS facilities is not supported if MSR_IA32_MISC_ENABLE.BTS bit is set.
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
//
mBtsSupported = FALSE;
}
}
}
}
/**
Enable single step.
**/
VOID
ActivateSingleStepDB (
VOID
)
{
UINTN Dr6;
Dr6 = AsmReadDr6 ();
if ((Dr6 & DR6_SINGLE_STEP) != 0) {
return;
}
Dr6 |= DR6_SINGLE_STEP;
AsmWriteDr6 (Dr6);
}
/**
Enable last branch.
**/
VOID
ActivateLBR (
VOID
)
{
UINT64 DebugCtl;
DebugCtl = AsmReadMsr64 (MSR_DEBUG_CTL);
if ((DebugCtl & MSR_DEBUG_CTL_LBR) != 0) {
return ;
}
DebugCtl |= MSR_DEBUG_CTL_LBR;
AsmWriteMsr64 (MSR_DEBUG_CTL, DebugCtl);
}
/**
Enable branch trace store.
@param CpuIndex The index of the processor.
**/
VOID
ActivateBTS (
IN UINTN CpuIndex
)
{
UINT64 DebugCtl;
DebugCtl = AsmReadMsr64 (MSR_DEBUG_CTL);
if ((DebugCtl & MSR_DEBUG_CTL_BTS) != 0) {
return ;
}
AsmWriteMsr64 (MSR_DS_AREA, (UINT64)(UINTN)mMsrDsArea[CpuIndex]);
DebugCtl |= (UINT64)(MSR_DEBUG_CTL_BTS | MSR_DEBUG_CTL_TR);
DebugCtl &= ~((UINT64)MSR_DEBUG_CTL_BTINT);
AsmWriteMsr64 (MSR_DEBUG_CTL, DebugCtl);
}
/**
Increase SMI number in each SMI entry.
**/
VOID
SmmProfileRecordSmiNum (
VOID
)
{
if (mSmmProfileStart) {
mSmmProfileBase->NumSmis++;
}
}
/**
Initialize processor environment for SMM profile.
@param CpuIndex The index of the processor.
**/
VOID
ActivateSmmProfile (
IN UINTN CpuIndex
)
{
//
// Enable Single Step DB#
//
ActivateSingleStepDB ();
if (mBtsSupported) {
//
// We can not get useful information from LER, so we have to use BTS.
//
ActivateLBR ();
//
// Enable BTS
//
ActivateBTS (CpuIndex);
}
}
/**
Initialize SMM profile in SMM CPU entry point.
@param[in] Cr3 The base address of the page tables to use in SMM.
**/
VOID
InitSmmProfile (
UINT32 Cr3
)
{
//
// Save Cr3
//
mSmmProfileCr3 = Cr3;
//
// Skip SMM profile initialization if feature is disabled
//
if (!FeaturePcdGet (PcdCpuSmmProfileEnable)) {
return;
}
//
// Initialize SmmProfile here
//
InitSmmProfileInternal ();
//
// Initialize profile IDT.
//
InitIdtr ();
}
/**
Update page table to map the memory correctly in order to make the instruction
which caused page fault execute successfully. And it also save the original page
table to be restored in single-step exception.
@param PageTable PageTable Address.
@param PFAddress The memory address which caused page fault exception.
@param CpuIndex The index of the processor.
@param ErrorCode The Error code of exception.
**/
VOID
RestorePageTableBelow4G (
UINT64 *PageTable,
UINT64 PFAddress,
UINTN CpuIndex,
UINTN ErrorCode
)
{
UINTN PTIndex;
UINTN PFIndex;
//
// PML4
//
if (sizeof(UINT64) == sizeof(UINTN)) {
PTIndex = (UINTN)BitFieldRead64 (PFAddress, 39, 47);
ASSERT (PageTable[PTIndex] != 0);
PageTable = (UINT64*)(UINTN)(PageTable[PTIndex] & PHYSICAL_ADDRESS_MASK);
}
//
// PDPTE
//
PTIndex = (UINTN)BitFieldRead64 (PFAddress, 30, 38);
ASSERT (PageTable[PTIndex] != 0);
PageTable = (UINT64*)(UINTN)(PageTable[PTIndex] & PHYSICAL_ADDRESS_MASK);
//
// PD
//
PTIndex = (UINTN)BitFieldRead64 (PFAddress, 21, 29);
if ((PageTable[PTIndex] & IA32_PG_PS) != 0) {
//
// Large page
//
//
// Record old entries with non-present status
// Old entries include the memory which instruction is at and the memory which instruction access.
//
//
ASSERT (mPFEntryCount[CpuIndex] < MAX_PF_ENTRY_COUNT);
if (mPFEntryCount[CpuIndex] < MAX_PF_ENTRY_COUNT) {
PFIndex = mPFEntryCount[CpuIndex];
mLastPFEntryValue[CpuIndex][PFIndex] = PageTable[PTIndex];
mLastPFEntryPointer[CpuIndex][PFIndex] = &PageTable[PTIndex];
mPFEntryCount[CpuIndex]++;
}
//
// Set new entry
//
PageTable[PTIndex] = (PFAddress & ~((1ull << 21) - 1));
PageTable[PTIndex] |= (UINT64)IA32_PG_PS;
PageTable[PTIndex] |= (UINT64)PAGE_ATTRIBUTE_BITS;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if ((ErrorCode & IA32_PF_EC_ID) != 0) {
PageTable[PTIndex] &= ~IA32_PG_NX;
}
} else {
//
// Small page
//
ASSERT (PageTable[PTIndex] != 0);
PageTable = (UINT64*)(UINTN)(PageTable[PTIndex] & PHYSICAL_ADDRESS_MASK);
//
// 4K PTE
//
PTIndex = (UINTN)BitFieldRead64 (PFAddress, 12, 20);
//
// Record old entries with non-present status
// Old entries include the memory which instruction is at and the memory which instruction access.
//
//
ASSERT (mPFEntryCount[CpuIndex] < MAX_PF_ENTRY_COUNT);
if (mPFEntryCount[CpuIndex] < MAX_PF_ENTRY_COUNT) {
PFIndex = mPFEntryCount[CpuIndex];
mLastPFEntryValue[CpuIndex][PFIndex] = PageTable[PTIndex];
mLastPFEntryPointer[CpuIndex][PFIndex] = &PageTable[PTIndex];
mPFEntryCount[CpuIndex]++;
}
//
// Set new entry
//
PageTable[PTIndex] = (PFAddress & ~((1ull << 12) - 1));
PageTable[PTIndex] |= (UINT64)PAGE_ATTRIBUTE_BITS;
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
if ((ErrorCode & IA32_PF_EC_ID) != 0) {
PageTable[PTIndex] &= ~IA32_PG_NX;
}
}
}
/**
The Page fault handler to save SMM profile data.
@param Rip The RIP when exception happens.
@param ErrorCode The Error code of exception.
**/
VOID
SmmProfilePFHandler (
UINTN Rip,
UINTN ErrorCode
)
{
UINT64 *PageTable;
UINT64 PFAddress;
UINTN CpuIndex;
UINTN Index;
UINT64 InstructionAddress;
UINTN MaxEntryNumber;
UINTN CurrentEntryNumber;
BOOLEAN IsValidPFAddress;
SMM_PROFILE_ENTRY *SmmProfileEntry;
UINT64 SmiCommand;
EFI_STATUS Status;
UINT8 SoftSmiValue;
EFI_SMM_SAVE_STATE_IO_INFO IoInfo;
if (!mSmmProfileStart) {
//
// If SMM profile does not start, call original page fault handler.
//
SmiDefaultPFHandler ();
return;
}
if (mBtsSupported) {
DisableBTS ();
}
IsValidPFAddress = FALSE;
PageTable = (UINT64 *)AsmReadCr3 ();
PFAddress = AsmReadCr2 ();
CpuIndex = GetCpuIndex ();
if (PFAddress <= 0xFFFFFFFF) {
RestorePageTableBelow4G (PageTable, PFAddress, CpuIndex, ErrorCode);
} else {
RestorePageTableAbove4G (PageTable, PFAddress, CpuIndex, ErrorCode, &IsValidPFAddress);
}
if (!IsValidPFAddress) {
InstructionAddress = Rip;
if ((ErrorCode & IA32_PF_EC_ID) != 0 && (mBtsSupported)) {
//
// If it is instruction fetch failure, get the correct IP from BTS.
//
InstructionAddress = GetSourceFromDestinationOnBts (CpuIndex, Rip);
if (InstructionAddress == 0) {
//
// It indicates the instruction which caused page fault is not a jump instruction,
// set instruction address same as the page fault address.
//
InstructionAddress = PFAddress;
}
}
//
// Indicate it is not software SMI
//
SmiCommand = 0xFFFFFFFFFFFFFFFFULL;
for (Index = 0; Index < gSmst->NumberOfCpus; Index++) {
Status = SmmReadSaveState(&mSmmCpu, sizeof(IoInfo), EFI_SMM_SAVE_STATE_REGISTER_IO, Index, &IoInfo);
if (EFI_ERROR (Status)) {
continue;
}
if (IoInfo.IoPort == mSmiCommandPort) {
//
// A software SMI triggered by SMI command port has been found, get SmiCommand from SMI command port.
//
SoftSmiValue = IoRead8 (mSmiCommandPort);
SmiCommand = (UINT64)SoftSmiValue;
break;
}
}
SmmProfileEntry = (SMM_PROFILE_ENTRY *)(UINTN)(mSmmProfileBase + 1);
//
// Check if there is already a same entry in profile data.
//
for (Index = 0; Index < (UINTN) mSmmProfileBase->CurDataEntries; Index++) {
if ((SmmProfileEntry[Index].ErrorCode == (UINT64)ErrorCode) &&
(SmmProfileEntry[Index].Address == PFAddress) &&
(SmmProfileEntry[Index].CpuNum == (UINT64)CpuIndex) &&
(SmmProfileEntry[Index].Instruction == InstructionAddress) &&
(SmmProfileEntry[Index].SmiCmd == SmiCommand)) {
//
// Same record exist, need not save again.
//
break;
}
}
if (Index == mSmmProfileBase->CurDataEntries) {
CurrentEntryNumber = (UINTN) mSmmProfileBase->CurDataEntries;
MaxEntryNumber = (UINTN) mSmmProfileBase->MaxDataEntries;
if (FeaturePcdGet (PcdCpuSmmProfileRingBuffer)) {
CurrentEntryNumber = CurrentEntryNumber % MaxEntryNumber;
}
if (CurrentEntryNumber < MaxEntryNumber) {
//
// Log the new entry
//
SmmProfileEntry[CurrentEntryNumber].SmiNum = mSmmProfileBase->NumSmis;
SmmProfileEntry[CurrentEntryNumber].ErrorCode = (UINT64)ErrorCode;
SmmProfileEntry[CurrentEntryNumber].ApicId = (UINT64)GetApicId ();
SmmProfileEntry[CurrentEntryNumber].CpuNum = (UINT64)CpuIndex;
SmmProfileEntry[CurrentEntryNumber].Address = PFAddress;
SmmProfileEntry[CurrentEntryNumber].Instruction = InstructionAddress;
SmmProfileEntry[CurrentEntryNumber].SmiCmd = SmiCommand;
//
// Update current entry index and data size in the header.
//
mSmmProfileBase->CurDataEntries++;
mSmmProfileBase->CurDataSize = MultU64x64 (mSmmProfileBase->CurDataEntries, sizeof (SMM_PROFILE_ENTRY));
}
}
}
//
// Flush TLB
//
CpuFlushTlb ();
if (mBtsSupported) {
EnableBTS ();
}
}
/**
Replace INT1 exception handler to restore page table to absent/execute-disable state
in order to trigger page fault again to save SMM profile data..
**/
VOID
InitIdtr (
VOID
)
{
EFI_STATUS Status;
Status = SmmRegisterExceptionHandler (&mSmmCpuService, EXCEPT_IA32_DEBUG, DebugExceptionHandler);
ASSERT_EFI_ERROR (Status);
UefiCpuPkg: Add PiSmmCpuDxeSmm module no IA32/X64 files Add module that initializes a CPU for the SMM environment and installs the first level SMI handler. This module along with the SMM IPL and SMM Core provide the services required for DXE_SMM_DRIVERS to register hardware and software SMI handlers. CPU specific features are abstracted through the SmmCpuFeaturesLib Platform specific features are abstracted through the SmmCpuPlatformHookLib Several PCDs are added to enable/disable features and configure settings for the PiSmmCpuDxeSmm module Changes between [PATCH v1] and [PATCH v2]: 1) Swap PTE init order for QEMU compatibility. Current PTE initialization algorithm works on HW but breaks QEMU emulator. Update the PTE initialization order to be compatible with both. 2) Update comment block that describes 32KB SMBASE alignment requirement to match contents of Intel(R) 64 and IA-32 Architectures Software Developer's Manual 3) Remove BUGBUG comment and call to ClearSmi() that is not required. SMI should be cleared by root SMI handler. [jeff.fan@intel.com: Fix code style issues reported by ECC] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Kinney <michael.d.kinney@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> [pbonzini@redhat.com: InitPaging: prepare PT before filling in PDE] Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com> git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18645 6f19259b-4bc3-4df7-8a09-765794883524
2015-10-19 21:12:53 +02:00
}