[edk2-devel] [PATCH v12 27/32] UefiCpuPkg/MpInitLib: use BSP to do extended topology check
Brijesh Singh via groups.io
brijesh.singh=amd.com at groups.io
Wed Nov 10 22:14:52 UTC 2021
From: Michael Roth <michael.roth at 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 at intel.com>
Cc: Ray Ni <ray.ni at intel.com>
Cc: Rahul Kumar <rahul1.kumar at intel.com>
Cc: James Bottomley <jejb at linux.ibm.com>
Cc: Min Xu <min.m.xu at intel.com>
Cc: Jiewen Yao <jiewen.yao at intel.com>
Cc: Tom Lendacky <thomas.lendacky at amd.com>
Cc: Jordan Justen <jordan.l.justen at intel.com>
Cc: Ard Biesheuvel <ardb+tianocore at kernel.org>
Cc: Erdem Aktas <erdemaktas at google.com>
Cc: Gerd Hoffmann <kraxel at redhat.com>
Acked-by: Gerd Hoffmann <kraxel at redhat.com>
Suggested-by: Brijesh Singh <brijesh.singh at amd.com>
Signed-off-by: Michael Roth <michael.roth at amd.com>
Signed-off-by: Brijesh Singh <brijesh.singh at amd.com>
---
UefiCpuPkg/Library/MpInitLib/MpLib.h | 1 +
UefiCpuPkg/Library/MpInitLib/MpLib.c | 11 ++++++++
UefiCpuPkg/Library/MpInitLib/MpEqu.inc | 1 +
UefiCpuPkg/Library/MpInitLib/X64/AmdSev.nasm | 27 ++++++++++++++++++++
4 files changed, 40 insertions(+)
diff --git a/UefiCpuPkg/Library/MpInitLib/MpLib.h b/UefiCpuPkg/Library/MpInitLib/MpLib.h
index 45bc1de23e3c..c52b6157429b 100644
--- a/UefiCpuPkg/Library/MpInitLib/MpLib.h
+++ b/UefiCpuPkg/Library/MpInitLib/MpLib.h
@@ -224,6 +224,7 @@ typedef struct {
BOOLEAN SevEsIsEnabled;
BOOLEAN SevSnpIsEnabled;
UINTN GhcbBase;
+ BOOLEAN ExtTopoAvail;
} MP_CPU_EXCHANGE_INFO;
#pragma pack()
diff --git a/UefiCpuPkg/Library/MpInitLib/MpLib.c b/UefiCpuPkg/Library/MpInitLib/MpLib.c
index ab838cbc0ff6..d8372ffa9d5a 100644
--- a/UefiCpuPkg/Library/MpInitLib/MpLib.c
+++ b/UefiCpuPkg/Library/MpInitLib/MpLib.c
@@ -853,6 +853,7 @@ FillExchangeInfoData (
UINTN Size;
IA32_SEGMENT_DESCRIPTOR *Selector;
IA32_CR4 Cr4;
+ UINT32 StdRangeMax;
ExchangeInfo = CpuMpData->MpCpuExchangeInfo;
ExchangeInfo->StackStart = CpuMpData->Buffer;
@@ -892,6 +893,16 @@ FillExchangeInfoData (
ExchangeInfo->SevSnpIsEnabled = CpuMpData->SevSnpIsEnabled;
ExchangeInfo->GhcbBase = (UINTN) CpuMpData->GhcbBase;
+ if (ExchangeInfo->SevSnpIsEnabled) {
+ AsmCpuid (CPUID_SIGNATURE, &StdRangeMax, NULL, NULL, NULL);
+ if (StdRangeMax >= CPUID_EXTENDED_TOPOLOGY) {
+ CPUID_EXTENDED_TOPOLOGY_EBX ExtTopoEbx;
+
+ AsmCpuid (CPUID_EXTENDED_TOPOLOGY, NULL, &ExtTopoEbx.Uint32, NULL, NULL);
+ ExchangeInfo->ExtTopoAvail = !!ExtTopoEbx.Bits.LogicalProcessors;
+ }
+ }
+
//
// Get the BSP's data of GDT and IDT
//
diff --git a/UefiCpuPkg/Library/MpInitLib/MpEqu.inc b/UefiCpuPkg/Library/MpInitLib/MpEqu.inc
index 01668638f245..aba53f57201c 100644
--- a/UefiCpuPkg/Library/MpInitLib/MpEqu.inc
+++ b/UefiCpuPkg/Library/MpInitLib/MpEqu.inc
@@ -94,6 +94,7 @@ struc MP_CPU_EXCHANGE_INFO
.SevEsIsEnabled: CTYPE_BOOLEAN 1
.SevSnpIsEnabled CTYPE_BOOLEAN 1
.GhcbBase: CTYPE_UINTN 1
+ .ExtTopoAvail: CTYPE_BOOLEAN 1
endstruc
MP_CPU_EXCHANGE_INFO_OFFSET equ (SwitchToRealProcEnd - RendezvousFunnelProcStart)
diff --git a/UefiCpuPkg/Library/MpInitLib/X64/AmdSev.nasm b/UefiCpuPkg/Library/MpInitLib/X64/AmdSev.nasm
index 0034920b2f6b..8bb1161fa0f7 100644
--- a/UefiCpuPkg/Library/MpInitLib/X64/AmdSev.nasm
+++ b/UefiCpuPkg/Library/MpInitLib/X64/AmdSev.nasm
@@ -118,6 +118,32 @@ SevEsGetApicId:
or rax, rdx
mov rdi, rax ; RDI now holds the original GHCB GPA
+ ;
+ ; For SEV-SNP, the recommended handling for getting the x2APIC ID
+ ; would be to use the SNP CPUID table to fetch CPUID.00H:EAX and
+ ; CPUID:0BH:EBX[15:0] instead of the GHCB MSR protocol vmgexits
+ ; below.
+ ;
+ ; To avoid the unecessary ugliness to accomplish that here, the BSP
+ ; has performed these checks in advance (where #VC handler handles
+ ; the CPUID table lookups automatically) and cached them in a flag
+ ; so those checks can be skipped here.
+ ;
+ mov eax, [esi + MP_CPU_EXCHANGE_INFO_FIELD (SevSnpIsEnabled)]
+ cmp al, 1
+ jne CheckExtTopoAvail
+
+ ;
+ ; Even with SEV-SNP, the actual x2APIC ID in CPUID.0BH:EDX
+ ; fetched from the hypervisor the same way SEV-ES does it.
+ ;
+ mov eax, [esi + MP_CPU_EXCHANGE_INFO_FIELD (ExtTopoAvail)]
+ cmp al, 1
+ je GetApicIdSevEs
+ ; The 8-bit APIC ID fallback is also the same as with SEV-ES
+ jmp NoX2ApicSevEs
+
+CheckExtTopoAvail:
mov rdx, 0 ; CPUID function 0
mov rax, 0 ; RAX register requested
or rax, 4
@@ -136,6 +162,7 @@ SevEsGetApicId:
test edx, 0ffffh
jz NoX2ApicSevEs ; CPUID.0BH:EBX[15:0] is zero
+GetApicIdSevEs:
mov rdx, 0bh ; CPUID function 0x0b
mov rax, 0c0000000h ; RDX register requested
or rax, 4
--
2.25.1
-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#83601): https://edk2.groups.io/g/devel/message/83601
Mute This Topic: https://groups.io/mt/86969151/1813853
Group Owner: devel+owner at edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [edk2-devel-archive at redhat.com]
-=-=-=-=-=-=-=-=-=-=-=-
More information about the edk2-devel-archive
mailing list