In the original design, once the APs finished executing their assembly
code and switched to executing C code, they would enter a continuous
loop within a function. In this function, they would collect CpuMpData
using the MP_CPU_EXCHANGE_INFO mechanism. However, in the updated
approach, CpuMpData can now be passed directly to the ApWakeUpFunction,
bypassing the need for MP_CPU_EXCHANGE_INFO. This modification is made
in preparation for eliminating the requirement of a second
INIT-SIPI-SIPI sequence in the DXE phase.
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=4353
Due to AMD erratum #1467, an SEV-SNP VMSA should not be 2MB aligned. To
work around this issue, allocate two pages instead of one. Because of the
way that page allocation is implemented, always try to use the second
page. If the second page is not 2MB aligned, free the first page and use
the second page. If the second page is 2MB aligned, free the second page
and use the first page. Freeing in this way reduces holes in the memory
map.
Fixes: 06544455d0 ("UefiCpuPkg/MpInitLib: Use SEV-SNP AP Creation ...")
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Ray Ni <ray.ni@intel.com>
https://bugzilla.tianocore.org/show_bug.cgi?id=4353
When parking the APs on exiting from UEFI, a new page allocation is made.
This allocation, however, does not end up being marked reserved in the
memory map supplied to the OS. To avoid this, re-use the VMSA by clearing
the VMSA RMP flag, updating the page contents and re-setting the VMSA RMP
flag.
Fixes: 06544455d0 ("UefiCpuPkg/MpInitLib: Use SEV-SNP AP Creation ...")
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Ray Ni <ray.ni@intel.com>
Add OUTPUT IsModified parameter in PageTableMap() to indicate
if page table has been modified. With this parameter, caller
can know if need to call FlushTlb when the page table is in CR3.
Signed-off-by: Dun Tan <dun.tan@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
In function CreatePageTable(), add code to initialize MapMask to
MAX_UINT64. When creating new page table or map non-present range
to present, all attributes should be provided.
Signed-off-by: Dun Tan <dun.tan@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
For the case CPU logic index is 0, RSP points to the very top of all AP
stacks. That address is not mapped in page table.
Cc: Guo Dong <guo.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Sean Rhodes <sean@starlabs.systems>
Cc: James Lu <james.lu@intel.com>
Cc: Gua Guo <gua.guo@intel.com>
Signed-off-by: Ted Kuo <ted.kuo@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Rename AsmRelocateApLoopStart to AsmRelocateApLoopStartAmdSev
Cc: Guo Dong <guo.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Sean Rhodes <sean@starlabs.systems>
Cc: James Lu <james.lu@intel.com>
Cc: Gua Guo <gua.guo@intel.com>
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Add the 'AsmRelocateApLoopStartGeneric' for X64 processors except 64-bit
AMD processors with SEV-ES.
Remove the unused arguments of AsmRelocateApLoopStartGeneric, updated
the stack offset.
Create PageTable for the allocated reserved memory.
Only keep 4GB limitation of memory allocation for the case APs still
need to be transferred to 32-bit mode before OS.
Cc: Guo Dong <guo.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Sean Rhodes <sean@starlabs.systems>
Cc: James Lu <james.lu@intel.com>
Cc: Gua Guo <gua.guo@intel.com>
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
This reverts commit 73ccde8f6d since it
results in a hang of the IA32 processor and needs further clean-up.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=4234
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
During the finalization of Mp initialization before booting into the OS,
depending on whether Mwait is supported or not, AsmRelocateApLoop
places Aps in MWAIT-loop or HLT-loop.
Since paging is necessary for long mode, the original implementation of
moving APs to 32-bit was to disable paging to ensure that the booting
does not crash.
The current modification creates a page table in reserved memory,
avoiding switching modes and reclaiming memory by OS. This modification
is only for 64 bit mode.
More specifically, we keep the AMD logic as the original code flow,
extract and update the Intel-related code, where the APs would stay
in 64-bit, and run in a Mwait or Hlt loop until the OS wake them up.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
AsmRelocateApLoop is replicated for future Intel Logic Extraction,
further brings AP into 64-bit, and enables paging.
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=4123
APIs which are defined in CcExitLib.h are added with the CcExit prefix.
This is to make the APIs' name more meaningful.
This change impacts OvmfPkg/UefiCpuPkg.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Signed-off-by: Min Xu <min.m.xu@intel.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=4123
VmgExitLib once was designed to provide interfaces to support #VC handler
and issue VMGEXIT instruction. After TDVF (enable TDX feature in OVMF) is
introduced, this library is updated to support #VE as well. Now the name
of VmgExitLib cannot reflect what the lib does.
This patch renames VmgExitLib to CcExitLib (Cc means Confidential
Computing). This is a simple renaming and there is no logic changes.
After renaming all the VmgExitLib related codes are updated with
CcExitLib. These changes are in OvmfPkg/UefiCpuPkg/UefiPayloadPkg.
Cc: Guo Dong <guo.dong@intel.com>
Cc: Sean Rhodes <sean@starlabs.systems>
Cc: James Lu <james.lu@intel.com>
Cc: Gua Guo <gua.guo@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Reviewed-by: James Lu <james.lu@intel.com>
Reviewed-by: Gua Guo <gua.guo@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Signed-off-by: Min Xu <min.m.xu@intel.com>
To remove the dependency of CPU register, 4/8 byte at the top of the
stack is occupied for CpuMpData. BIST information is also taken care
here. This modification is only for PEI phase, since in DXE phase
CpuMpData is accessed via global variable.
Signed-off-by: Yuanhao Xie <yuanhao.xie@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
When switch bsp, old bsp and new bsp put CR0/CR4 into stack, and put IDT
and GDT register into a structure. After they exchange their stack, they
restore these registers. This logic is now implemented by assembly code.
This patch aims to reuse (Save/Restore)VolatileRegisters function to
replace such assembly code for better code readability.
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Signed-off-by: Zhiguang Liu <zhiguang.liu@intel.com>
Currently, when waking up AP, IDT table of AP will be set in 16 bit code,
and assume the IDT table base is 32 bit. However, the IDT table is created
by BSP. Issue will happen if the BSP allocates memory above 4G for BSP's
IDT table. Moreover, even the IDT table location is below 4G, the handler
function inside the IDT table is 64 bit, and it won't take effect until
CPU transfers to 64 bit long mode. There is no benefit to set IDT table in
such an early phase.
To avoid such issue, this patch moves the LIDT instruction into 64 bit
code.
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Signed-off-by: Zhiguang Liu <zhiguang.liu@intel.com>
The patch does several simplifications:
1. Treat SwitchToRealProc as part of RendezvousFunnelProc.
So the common logic in MpLib.c doesn't need to be aware of
SwitchToRealProc.
As a result, SwitchToRealSize/Offset are removed from
MP_ASSEMBLY_ADDRESS_MAP.
2. Move SwitchToRealProc to AmdSev.nasm.
All other assembly code in AmdSev.nasm is called through
OneTimeCall.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Tested-by: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: Michael Roth <michael.roth@amd.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
global in NASM file is used for symbols that are
referenced in C files.
Remove unneeded global keyword in NASM file.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3790
Replace Opcode with the corresponding instructions.
The code changes have been verified with CompareBuild.py tool, which
can be used to compare the results of two different EDK II builds to
determine if they generate the same binaries.
(tool link: https://github.com/mdkinney/edk2/tree/sandbox/CompareBuild)
Signed-off-by: Jason Lou <yun.lou@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3275
Use the SEV-SNP AP Creation NAE event to create and launch APs under
SEV-SNP. This capability will be advertised in the SEV Hypervisor
Feature Support PCD (PcdSevEsHypervisorFeatures).
Cc: Michael Roth <michael.roth@amd.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Ray Ni <ray.ni@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
During AP bringup, just after switching to long mode, APs will do some
cpuid calls to verify that the extended topology leaf (0xB) is available
so they can fetch their x2 APIC IDs from it. In the case of SEV-ES,
these cpuid instructions must be handled by direct use of the GHCB MSR
protocol to fetch the values from the hypervisor, since a #VC handler
is not yet available due to the AP's stack not being set up yet.
For SEV-SNP, rather than relying on the GHCB MSR protocol, it is
expected that these values would be obtained from the SEV-SNP CPUID
table instead. The actual x2 APIC ID (and 8-bit APIC IDs) would still
be fetched from hypervisor using the GHCB MSR protocol however, so
introducing support for the SEV-SNP CPUID table in that part of the AP
bring-up code would only be to handle the checks/validation of the
extended topology leaf.
Rather than introducing all the added complexity needed to handle these
checks via the CPUID table, instead let the BSP do the check in advance,
since it can make use of the #VC handler to avoid the need to scan the
SNP CPUID table directly, and add a flag in ExchangeInfo to communicate
the result of this check to APs.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Ray Ni <ray.ni@intel.com>
Suggested-by: Brijesh Singh <brijesh.singh@amd.com>
Signed-off-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3275
An SEV-SNP guest requires that the physical address of the GHCB must
be registered with the hypervisor before using it. See the GHCB
specification section 2.3.2 for more details.
Cc: Michael Roth <michael.roth@amd.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Ray Ni <ray.ni@Intel.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3275
Move all the SEV specific function in AmdSev.c.
No functional change intended.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Cc: Michael Roth <michael.roth@amd.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Suggested-by: Jiewen Yao <Jiewen.yao@intel.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
The Lock is no longer needed since "LOCK XADD" was used in
MpFuncs.nasm for ApIndex atomic increment.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
In Windows environment, "dumpbin /disasm" is used to verify the
disassembly before and after using NASM struc doesn't change.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
When AP firstly wakes up, MpFuncs.nasm contains below logic to assign
an unique ApIndex to each AP according to who comes first:
---ASM---
TestLock:
xchg [edi], eax
cmp eax, NotVacantFlag
jz TestLock
mov ecx, esi
add ecx, ApIndexLocation
inc dword [ecx]
mov ebx, [ecx]
Releaselock:
mov eax, VacantFlag
xchg [edi], eax
---ASM END---
"lock inc" cannot be used to increase ApIndex because not only the
global ApIndex should be increased, but also the result should be
stored to a local general purpose register EBX.
This patch learns from the NASM implementation of
InternalSyncIncrement() to use "XADD" instruction which can increase
the global ApIndex and store the original ApIndex to EBX in one
instruction.
With this patch, OVMF when running in a 255 threads QEMU spends about
one second to wakeup all APs. Original implementation needs more than
10 seconds.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
This patch fixed the hang in UEFICpuPkg when it is dispatched above 4GB.
In UEFI BIOS case CpuInfoInHob is provided to DXE under 4GB from PEI.
When using UEFI payload and bootloaders, CpuInfoInHob will be allocated
above 4GB since it is not provided from bootloader. so we need update
the code to make sure this hob could be accessed correctly in this case.
Signed-off-by: Guo Dong <guo.dong@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3008
All fields that are set in the GHCB should have their associated bit in
the GHCB ValidBitmap field set. Add support to set the bits for the
software exit information fields when performing a VMGEXIT (SwExitCode,
SwExitInfo1, SwExitInfo2).
Fixes: 20da7ca42a
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Rahul Kumar <rahul1.kumar@intel.com>
Acked-by: Ray Ni <ray.ni@intel.com>
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Message-Id: <6e11dd7e161bddeacc3fb4817467cef24510c31c.1604685192.git.thomas.lendacky@amd.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=2198
Before UEFI transfers control to the OS, it must park the AP. This is
done using the AsmRelocateApLoop function to transition into 32-bit
non-paging mode. For an SEV-ES guest, a few additional things must be
done:
- AsmRelocateApLoop must be updated to support SEV-ES. This means
performing a VMGEXIT AP Reset Hold instead of an MWAIT or HLT loop.
- Since the AP must transition to real mode, a small routine is copied
to the WakeupBuffer area. Since the WakeupBuffer will be used by
the AP during OS booting, it must be placed in reserved memory.
Additionally, the AP stack must be located where it can be accessed
in real mode.
- Once the AP is in real mode it will transfer control to the
destination specified by the OS in the SEV-ES AP Jump Table. The
SEV-ES AP Jump Table address is saved by the hypervisor for the OS
using the GHCB VMGEXIT AP Jump Table exit code.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=2198
Typically, an AP is booted using the INIT-SIPI-SIPI sequence. This
sequence is intercepted by the hypervisor, which sets the AP's registers
to the values requested by the sequence. At that point, the hypervisor can
start the AP, which will then begin execution at the appropriate location.
Under SEV-ES, AP booting presents some challenges since the hypervisor is
not allowed to alter the AP's register state. In this situation, we have
to distinguish between the AP's first boot and AP's subsequent boots.
First boot:
Once the AP's register state has been defined (which is before the guest
is first booted) it cannot be altered. Should the hypervisor attempt to
alter the register state, the change would be detected by the hardware
and the VMRUN instruction would fail. Given this, the first boot for the
AP is required to begin execution with this initial register state, which
is typically the reset vector. This prevents the BSP from directing the
AP startup location through the INIT-SIPI-SIPI sequence.
To work around this, the firmware will provide a build time reserved area
that can be used as the initial IP value. The hypervisor can extract this
location value by checking for the SEV-ES reset block GUID that must be
located 48-bytes from the end of the firmware. The format of the SEV-ES
reset block area is:
0x00 - 0x01 - SEV-ES Reset IP
0x02 - 0x03 - SEV-ES Reset CS Segment Base[31:16]
0x04 - 0x05 - Size of the SEV-ES reset block
0x06 - 0x15 - SEV-ES Reset Block GUID
(00f771de-1a7e-4fcb-890e-68c77e2fb44e)
The total size is 22 bytes. Any expansion to this block must be done
by adding new values before existing values.
The hypervisor will use the IP and CS values obtained from the SEV-ES
reset block to set as the AP's initial values. The CS Segment Base
represents the upper 16 bits of the CS segment base and must be left
shifted by 16 bits to form the complete CS segment base value.
Before booting the AP for the first time, the BSP must initialize the
SEV-ES reset area. This consists of programming a FAR JMP instruction
to the contents of a memory location that is also located in the SEV-ES
reset area. The BSP must program the IP and CS values for the FAR JMP
based on values drived from the INIT-SIPI-SIPI sequence.
Subsequent boots:
Again, the hypervisor cannot alter the AP register state, so a method is
required to take the AP out of halt state and redirect it to the desired
IP location. If it is determined that the AP is running in an SEV-ES
guest, then instead of calling CpuSleep(), a VMGEXIT is issued with the
AP Reset Hold exit code (0x80000004). The hypervisor will put the AP in
a halt state, waiting for an INIT-SIPI-SIPI sequence. Once the sequence
is recognized, the hypervisor will resume the AP. At this point the AP
must transition from the current 64-bit long mode down to 16-bit real
mode and begin executing at the derived location from the INIT-SIPI-SIPI
sequence.
Another change is around the area of obtaining the (x2)APIC ID during AP
startup. During AP startup, the AP can't take a #VC exception before the
AP has established a stack. However, the AP stack is set by using the
(x2)APIC ID, which is obtained through CPUID instructions. A CPUID
instruction will cause a #VC, so a different method must be used. The
GHCB protocol supports a method to obtain CPUID information from the
hypervisor through the GHCB MSR. This method does not require a stack,
so it is used to obtain the necessary CPUID information to determine the
(x2)APIC ID.
The new 16-bit protected mode GDT entry is used in order to transition
from 64-bit long mode down to 16-bit real mode.
A new assembler routine is created that takes the AP from 64-bit long mode
to 16-bit real mode. This is located under 1MB in memory and transitions
from 64-bit long mode to 32-bit compatibility mode to 16-bit protected
mode and finally 16-bit real mode.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2008
MpInitLib is the library that's responsible to wake up APs to provide
MP PPI and Protocol services.
The patch synchronizes BSP's CR4.LA57 to each AP's CR4.LA57.
Without this change, AP may enter to GP fault when BSP's 5-level page
table is set to AP during AP wakes up.
Signed-off-by: Ray Ni <ray.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Eric Dong <eric.dong@intel.com>
1. Do not use tab characters
2. No trailing white space in one line
3. All files must end with CRLF
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Within function ApWakeupFunction():
When source level debugger is enabled, AP interrupts will be enabled by
EnableDebugAgent(). Then the AP function will be executed by:
Procedure (Parameter);
After the AP function returns, AP interrupts will be disabled when the
APs are placed in loop mode (both HltLoop and MwaiLoop).
However, at ExitBootServices, ApWakeupFunction() is called with
'Procedure' equals to RelocateApLoop().
(ExitBootServices callback registered within InitMpGlobalData())
RelocateApLoop() never returns, so it has to disable the AP interrupts by
itself. However, we find that interrupts are only disabled for the
HltLoop case, but not for the MwaitLoop case (within file MpFuncs.nasm).
This commit adds the missing disabling of AP interrupts for MwaitLoop.
Also, for X64, this commit will disable the interrupts before switching to
32-bit mode.
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Jeff Fan <vanjeff_919@hotmail.com>
If PcdDxeNxMemoryProtectionPolicy is set to enable protection for memory
of EfiBootServicesCode, EfiConventionalMemory, the BIOS will hang at a page
fault exception during MP initialization.
The root cause is that the AP wake up buffer, which is below 1MB and used
to hold both AP init code and data, is type of EfiConventionalMemory (not
really allocated because of potential conflict with legacy code), and is
marked as non-executable. During the transition from real address mode
to long mode, the AP init code has to enable paging which will then cause
itself a page fault exception because it's just running in non-executable
memory.
The solution is splitting AP wake up buffer into two part: lower part is
still below 1MB and shared with legacy system, higher part is really
allocated memory of BootServicesCode type. The init code in the memory
below 1MB will not enable paging but just switch to protected mode and
jump to higher memory, in which the init code will enable paging and
switch to long mode.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Current logic always waiting for a specific value to collect all APs
count. This logic may caused some platforms cost too much time to
wait for time out.
This patch add new logic to collect APs count. It adds new variable
NumApsExecuting to detect whether all APs have finished initialization.
Each AP let NumApsExecuting++ when begin to initialize itself and let
NumApsExecuting-- when it finish the initialization. BSP base on whether
NumApsExecuting == 0 to finished the collect AP process.
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jeff Fan <vanjeff_919@hotmail.com>
Original AP index variable name not well express the meaning
of the variable. Also this name is better used in later patch.
So change the variable name for better understanding.
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jeff Fan <vanjeff_919@hotmail.com>
https://bugzilla.tianocore.org/show_bug.cgi?id=565
Fix NASM compatibility issues with XCODE5 tool chain.
The XCODE5 tool chain for X64 builds using PIE (Position
Independent Executable). For most assembly sources using
PIE mode does not cause any issues.
However, if assembly code is copied to a different address
(such as AP startup code in the MpInitLib), then the
X64 assembly source must be implemented to be compatible
with PIE mode that uses RIP relative addressing.
The specific changes in this patch are:
* Use LEA instruction instead of MOV instruction to lookup
the addresses of functions.
* The assembly function RendezvousFunnelProc() is copied
below 1MB so it can be executed as part of the MpInitLib
AP startup sequence. RendezvousFunnelProc() calls the
external function InitializeFloatingPointUnits(). The
absolute address of InitializeFloatingPointUnits() is
added to the MP_CPU_EXCHANGE_INFO structure that is passed
to RendezvousFunnelProc().
Cc: Andrew Fish <afish@apple.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>
AP will get target C-State from eax[7:4]. We do shift in ebx firstly before set
to eax. It will lead ebx is incorrect in the next time.
The fix is to set ebx to eax firstly and does shift in eax. Thus, ebx could keep
original value.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
Add one semaphore to make sure BSP to wait till all APs run in AP safe loop
code.
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
For long mode DXE, we will disable paging on AP to protected mode to execute AP
safe loop code in reserved memory range under 4GB. But we forget to allocate
stack for AP under 4GB and AP still are using original AP stack. If original AP
stack is larger than 4GB, it cannot be used after AP is transferred to protected
mode. Besides MwaitSupport == TRUE, AP stack is still required during phase of
disabling paging in long mode DXE.
Moreover, even though AP stack is always under 4GB (a) in Ia32 DXE and (b) with
this patch, after transferring to protected mode from X64 DXE, AP stack
(in BootServiceData) maybe crashed by OS after Exit Boot Service event.
This fix is to allocate reserved memory range under 4GB together with AP safe
loop code. APs will switch to new stack in safe loop code.
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
According to the Intel SDM (325462-060US / September 2016),
> INPUT EAX = 0BH: Returns Extended Topology Information
>
> [...] Software must detect the presence of CPUID leaf 0BH by verifying
> (a) the highest leaf index supported by CPUID is >= 0BH, and
> (b) CPUID.0BH:EBX[15:0] reports a non-zero value. [...]
The "GetApicId" sections in the Ia32 and X64 "MpFuncs.nasm" files do not
perform check (b).
This causes an actual bug in the following OVMF setup:
- Intel W3550 host processor <http://ark.intel.com/products/39720/>,
- the QEMU/KVM guest's VCPU model is set to "host", that is, "the CPU
visible to the guest should be exactly the same as the host CPU".
Under "GetApicId", check (a) passes: the CPUID level of the W3550 is
exactly 11 decimal. However, leaf 11 itself is not supported, therefore
EDX is set to zero:
> If a value entered for CPUID.EAX is less than or equal to the maximum
> input value and the leaf is not supported on that processor then 0 is
> returned in all the registers.
Because we don't check (b), the "GetProcessorNumber" section of the code
is reached with an initial APIC ID of 0 in EDX on all of the APs. Given
that "GetProcessorNumber" searches the
"MP_CPU_EXCHANGE_INFO.CpuInfo[*].InitialApicId" fields for a match, all
APs enter ApWakeupFunction() with an identical "NumApsExecuting"
parameter. This results in unpredictable guest behavior (crashes, reboots,
hangs etc).
Reorganize the "GetApicId" section and add the missing check in both
assembly files.
Cc: Jeff Fan <jeff.fan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
The cached "CPU_INFO_IN_HOB.ApTopOfStack" field currently has type UINT32.
This is not ideal because the AP stacks are located within
"CpuMpData->Buffer", which is allocated with a plain AllocatePages() call
in MpInitLibInitialize():
platform CpuMpPei included PEI RAM > 4GB result
-------- ----------------- ------------- ------
Ia32 * n/a good
Ia32X64 no n/a BAD
Ia32X64 yes n/a good
X64 no * BAD
X64 yes no good
X64 yes yes BAD
- If we are on an Ia32X64 or X64 platform that does not include CpuMpPei,
then CpuDxe cannot reuse the CPU_INFO_IN_HOB structures preallocated by
CpuMpPei (through the CpuInitMpLib GUID HOB), and then AllocatePages()
-- invoked first in 64-bit DXE -- could return an address outside of
32-bit address space.
- If we are on an X64 platform where the permanent PEI RAM extends above
the 32-bit address space, then the same issue can surface even if
CpuMpPei is included: even the original allocation of the
CPU_INFO_IN_HOB structures, by CpuMpPei, could be satisfied from above
4GB.
The original "AP init" branch in "X64/MpFuncs.nasm" correctly considers a
64-bit stack start: the "MP_CPU_EXCHANGE_INFO.StackStart" field has type
UINTN, and the code uses QWORD addition and movement to set RSP from it.
Adapt the "GetApicId" branch of "X64/MpFuncs.nasm":
- change the type of "CPU_INFO_IN_HOB.ApTopOfStack" to UINT64,
- remove the explicit truncation to UINT32 in InitializeApData(),
- update the "GetNextProcNumber" iteration size to the new size of
"CPU_INFO_IN_HOB",
- set RSP with a QWORD movement from "CPU_INFO_IN_HOB.ApTopOfStack".
Because the same CPU_INFO_IN_HOB structure is used by "Ia32/MpFuncs.nasm",
we have to update the "GetNextProcNumber" iteration size there as well.
The ESP setting can be preserved as a DWORD movement from the original
offset (decimal 12), since our integers are little endian.
Cc: Jeff Fan <jeff.fan@intel.com>
Fixes: 845c5be1fd
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
The recent patch "UefiCpuPkg/MpInitLib: Program AP stack in fixed address"
inadvertently broke the first startup of APs during X64 PEI, because in
the TestLock section of the code, it replaced the access to the
NumApsExecuting counter with an access to the unrelated InitFlag field.
Cc: Jeff Fan <jeff.fan@intel.com>
Fixes: 845c5be1fd
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
At this point, ESI still has the value from EBX.
Cc: Jeff Fan <jeff.fan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Currently, MpInitLib will program AP stack in dynamic address. Each processor
will calculate its stack address by adding stack size based on the last stack
address. That means AP may have the different stack address everytime it is
wakeup by INIT-SIPI-SIPI.
When all APs have wakeup to execute AP task, each each has been assigned one
stack address. Once the timeout happened on some of APs, BSP will send INIT-
SIPI-SIPI to wake up APs. We need to re-assign stack for APs. Based on the
current implementation, we might assign one stack address used by other APs.
It will cause the unexpected stack overlapped issue.
This fix changed the stack assignment policy. We will record the stack address
assigned to AP at first time AP wakeup. When AP failed on AP task, BSP could
reassigned the same stack for it.
Getting initial APIC ID in assembly code could help AP to get saved its stack
address.
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Xie, Yuanhao
Lendacky, Thomas via groups.io
Dun Tan
Min M Xu
Liu, Zhiguang
Ray Ni
Jason
Tom Lendacky
Michael Roth
Brijesh Singh
Guo Dong
Michael D Kinney
Liming Gao
Hao Wu
Jian J Wang
Eric Dong
Jeff Fan
Laszlo Ersek