From: Leo Duran <leo.duran@amd.com>
To: edk2-devel@lists.01.org
Cc: Leo Duran <leo.duran@amd.com>,
Jordan Justen <jordan.l.justen@intel.com>,
Jeff Fan <jeff.fan@intel.com>, Liming Gao <liming.gao@intel.com>,
Brijesh Singh <brijesh.singh@amd.com>
Subject: [PATCH v4 2/2] UefiCpuPkg: Modify GetProcessorLocationByApicId() to support AMD.
Date: Fri, 16 Jun 2017 17:57:00 -0500 [thread overview]
Message-ID: <1497653820-15192-3-git-send-email-leo.duran@amd.com> (raw)
In-Reply-To: <1497653820-15192-1-git-send-email-leo.duran@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Leo Duran <leo.duran@amd.com>
---
UefiCpuPkg/Library/BaseXApicLib/BaseXApicLib.c | 140 ++++++++++++++++-----
.../BaseXApicX2ApicLib/BaseXApicX2ApicLib.c | 140 ++++++++++++++++-----
2 files changed, 216 insertions(+), 64 deletions(-)
diff --git a/UefiCpuPkg/Library/BaseXApicLib/BaseXApicLib.c b/UefiCpuPkg/Library/BaseXApicLib/BaseXApicLib.c
index f81bbb2..898d844 100644
--- a/UefiCpuPkg/Library/BaseXApicLib/BaseXApicLib.c
+++ b/UefiCpuPkg/Library/BaseXApicLib/BaseXApicLib.c
@@ -4,6 +4,8 @@
This local APIC library instance supports xAPIC mode only.
Copyright (c) 2010 - 2016, Intel Corporation. All rights reserved.<BR>
+ Copyright (c) 2017, AMD Inc. All rights reserved.<BR>
+
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
@@ -15,6 +17,7 @@
**/
#include <Register/Cpuid.h>
+#include <Register/Amd/Cpuid.h>
#include <Register/Msr.h>
#include <Register/LocalApic.h>
@@ -966,20 +969,33 @@ GetProcessorLocationByApicId (
OUT UINT32 *Thread OPTIONAL
)
{
- BOOLEAN TopologyLeafSupported;
- UINTN ThreadBits;
- UINTN CoreBits;
- CPUID_VERSION_INFO_EBX VersionInfoEbx;
- CPUID_VERSION_INFO_EDX VersionInfoEdx;
- CPUID_CACHE_PARAMS_EAX CacheParamsEax;
- CPUID_EXTENDED_TOPOLOGY_EAX ExtendedTopologyEax;
- CPUID_EXTENDED_TOPOLOGY_EBX ExtendedTopologyEbx;
- CPUID_EXTENDED_TOPOLOGY_ECX ExtendedTopologyEcx;
- UINT32 MaxCpuIdIndex;
- UINT32 SubIndex;
- UINTN LevelType;
- UINT32 MaxLogicProcessorsPerPackage;
- UINT32 MaxCoresPerPackage;
+ BOOLEAN TopologyLeafSupported;
+ CPUID_VERSION_INFO_EBX VersionInfoEbx;
+ CPUID_VERSION_INFO_EDX VersionInfoEdx;
+ CPUID_CACHE_PARAMS_EAX CacheParamsEax;
+ CPUID_EXTENDED_TOPOLOGY_EAX ExtendedTopologyEax;
+ CPUID_EXTENDED_TOPOLOGY_EBX ExtendedTopologyEbx;
+ CPUID_EXTENDED_TOPOLOGY_ECX ExtendedTopologyEcx;
+ CPUID_AMD_EXTENDED_CPU_SIG_ECX AmdExtendedCpuSigEcx;
+ CPUID_AMD_PROCESSOR_TOPOLOGY_EBX AmdProcessorTopologyEbx;
+ CPUID_AMD_PROCESSOR_TOPOLOGY_ECX AmdProcessorTopologyEcx;
+ CPUID_AMD_VIR_PHY_ADDRESS_SIZE_ECX AmdVirPhyAddressSizeEcx;
+ UINT32 SignatureEbx;
+ UINT32 SignatureEcx;
+ UINT32 SignatureEdx;
+ UINT32 MaxStandardCpuIdIndex;
+ UINT32 MaxExtendedCpuIdIndex;
+ UINT32 SubIndex;
+ UINTN LevelType;
+ UINT32 MaxLogicProcessorsPerPackage;
+ UINT32 MaxCoresPerPackage;
+ UINT32 MaxThreadPerPackageMask;
+ UINT32 ActualThreadPerPackageMask;
+ UINT32 MaxCoresPerNode;
+ UINT32 CorePerNodeMask;
+ UINT32 ApicIdShift;
+ UINTN ThreadBits;
+ UINTN CoreBits;
//
// Check if the processor is capable of supporting more than one logical processor.
@@ -987,10 +1003,10 @@ GetProcessorLocationByApicId (
AsmCpuid(CPUID_VERSION_INFO, NULL, NULL, NULL, &VersionInfoEdx.Uint32);
if (VersionInfoEdx.Bits.HTT == 0) {
if (Thread != NULL) {
- *Thread = 0;
+ *Thread = 0;
}
if (Core != NULL) {
- *Core = 0;
+ *Core = 0;
}
if (Package != NULL) {
*Package = 0;
@@ -998,24 +1014,24 @@ GetProcessorLocationByApicId (
return;
}
+ //
+ // Assume three-level mapping of APIC ID: Package|Core|Thread.
+ //
ThreadBits = 0;
CoreBits = 0;
//
- // Assume three-level mapping of APIC ID: Package:Core:SMT.
+ // Get max index of CPUID and vendor's signature
//
- TopologyLeafSupported = FALSE;
-
- //
- // Get the max index of basic CPUID
- //
- AsmCpuid(CPUID_SIGNATURE, &MaxCpuIdIndex, NULL, NULL, NULL);
+ AsmCpuid(CPUID_SIGNATURE, &MaxStandardCpuIdIndex, &SignatureEbx, &SignatureEcx, &SignatureEdx);
+ AsmCpuid(CPUID_EXTENDED_FUNCTION, &MaxExtendedCpuIdIndex, NULL, NULL, NULL);
//
// If the extended topology enumeration leaf is available, it
// is the preferred mechanism for enumerating topology.
//
- if (MaxCpuIdIndex >= CPUID_EXTENDED_TOPOLOGY) {
+ TopologyLeafSupported = FALSE;
+ if (MaxStandardCpuIdIndex >= CPUID_EXTENDED_TOPOLOGY) {
AsmCpuidEx(
CPUID_EXTENDED_TOPOLOGY,
0,
@@ -1065,27 +1081,87 @@ GetProcessorLocationByApicId (
}
if (!TopologyLeafSupported) {
+ //
+ // Get logical processor count
+ //
AsmCpuid(CPUID_VERSION_INFO, NULL, &VersionInfoEbx.Uint32, NULL, NULL);
MaxLogicProcessorsPerPackage = VersionInfoEbx.Bits.MaximumAddressableIdsForLogicalProcessors;
- if (MaxCpuIdIndex >= CPUID_CACHE_PARAMS) {
- AsmCpuidEx(CPUID_CACHE_PARAMS, 0, &CacheParamsEax.Uint32, NULL, NULL, NULL);
- MaxCoresPerPackage = CacheParamsEax.Bits.MaximumAddressableIdsForLogicalProcessors + 1;
+
+ //
+ // Assume single-core processor
+ //
+ MaxCoresPerPackage = 1;
+
+ //
+ // Check for topology extensions on AMD processor
+ //
+ if (SignatureEbx == CPUID_SIGNATURE_GENUINE_AMD_EBX &&
+ SignatureEcx == CPUID_SIGNATURE_GENUINE_AMD_ECX &&
+ SignatureEdx == CPUID_SIGNATURE_GENUINE_AMD_EDX) {
+ if (MaxExtendedCpuIdIndex >= CPUID_AMD_PROCESSOR_TOPOLOGY) {
+ AsmCpuid(CPUID_EXTENDED_CPU_SIG, NULL, NULL, &AmdExtendedCpuSigEcx.Uint32, NULL);
+ if (AmdExtendedCpuSigEcx.Bits.TopologyExtensions != 0) {
+ AsmCpuid(CPUID_AMD_PROCESSOR_TOPOLOGY, NULL, &AmdProcessorTopologyEbx.Uint32,
+ &AmdProcessorTopologyEcx.Uint32, NULL);
+ //
+ // Get cores per processor package
+ //
+ MaxCoresPerPackage = MaxLogicProcessorsPerPackage / (AmdProcessorTopologyEbx.Bits.ThreadsPerCore + 1);
+
+ //
+ // Account for actual thread count (e.g., SMT disabled)
+ //
+ AsmCpuid(CPUID_VIR_PHY_ADDRESS_SIZE, NULL, NULL, &AmdVirPhyAddressSizeEcx.Uint32, NULL);
+ MaxThreadPerPackageMask = 1 << AmdVirPhyAddressSizeEcx.Bits.ApicIdCoreIdSize;
+ ActualThreadPerPackageMask = 1;
+ while (ActualThreadPerPackageMask < MaxLogicProcessorsPerPackage) {
+ ActualThreadPerPackageMask <<= 1;
+ }
+
+ //
+ // Adjust APIC Id to report concatenation of Package|Core|Thread.
+ //
+ if (ActualThreadPerPackageMask < MaxThreadPerPackageMask) {
+ MaxCoresPerNode = MaxCoresPerPackage / (AmdProcessorTopologyEcx.Bits.NodesPerProcessor + 1);
+
+ CorePerNodeMask = 1;
+ while (CorePerNodeMask < MaxCoresPerNode) {
+ CorePerNodeMask <<= 1;
+ }
+ CorePerNodeMask -= 1;
+
+ ApicIdShift = 0;
+ do {
+ ApicIdShift += 1;
+ ActualThreadPerPackageMask <<= 1;
+ } while (ActualThreadPerPackageMask < MaxThreadPerPackageMask);
+
+ InitialApicId = ((InitialApicId & ~CorePerNodeMask) >> ApicIdShift) | (InitialApicId & CorePerNodeMask);
+ }
+ }
+ }
}
else {
//
- // Must be a single-core processor.
+ // Extract core count based on CACHE information
//
- MaxCoresPerPackage = 1;
+ if (MaxStandardCpuIdIndex >= CPUID_CACHE_PARAMS) {
+ AsmCpuidEx(CPUID_CACHE_PARAMS, 0, &CacheParamsEax.Uint32, NULL, NULL, NULL);
+ if (CacheParamsEax.Uint32 != 0) {
+ MaxCoresPerPackage = CacheParamsEax.Bits.MaximumAddressableIdsForLogicalProcessors + 1;
+ }
+ }
}
ThreadBits = (UINTN)(HighBitSet32(MaxLogicProcessorsPerPackage / MaxCoresPerPackage - 1) + 1);
- CoreBits = (UINTN)(HighBitSet32(MaxCoresPerPackage - 1) + 1); }
+ CoreBits = (UINTN)(HighBitSet32(MaxCoresPerPackage - 1) + 1);
+ }
if (Thread != NULL) {
- *Thread = InitialApicId & ((1 << ThreadBits) - 1);
+ *Thread = InitialApicId & ((1 << ThreadBits) - 1);
}
if (Core != NULL) {
- *Core = (InitialApicId >> ThreadBits) & ((1 << CoreBits) - 1);
+ *Core = (InitialApicId >> ThreadBits) & ((1 << CoreBits) - 1);
}
if (Package != NULL) {
*Package = (InitialApicId >> (ThreadBits + CoreBits));
diff --git a/UefiCpuPkg/Library/BaseXApicX2ApicLib/BaseXApicX2ApicLib.c b/UefiCpuPkg/Library/BaseXApicX2ApicLib/BaseXApicX2ApicLib.c
index e690d2a..9d3b82f 100644
--- a/UefiCpuPkg/Library/BaseXApicX2ApicLib/BaseXApicX2ApicLib.c
+++ b/UefiCpuPkg/Library/BaseXApicX2ApicLib/BaseXApicX2ApicLib.c
@@ -5,6 +5,8 @@
which have xAPIC and x2APIC modes.
Copyright (c) 2010 - 2016, Intel Corporation. All rights reserved.<BR>
+ Copyright (c) 2017, AMD Inc. All rights reserved.<BR>
+
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
@@ -16,6 +18,7 @@
**/
#include <Register/Cpuid.h>
+#include <Register/Amd/Cpuid.h>
#include <Register/Msr.h>
#include <Register/LocalApic.h>
@@ -1061,20 +1064,33 @@ GetProcessorLocationByApicId (
OUT UINT32 *Thread OPTIONAL
)
{
- BOOLEAN TopologyLeafSupported;
- UINTN ThreadBits;
- UINTN CoreBits;
- CPUID_VERSION_INFO_EBX VersionInfoEbx;
- CPUID_VERSION_INFO_EDX VersionInfoEdx;
- CPUID_CACHE_PARAMS_EAX CacheParamsEax;
- CPUID_EXTENDED_TOPOLOGY_EAX ExtendedTopologyEax;
- CPUID_EXTENDED_TOPOLOGY_EBX ExtendedTopologyEbx;
- CPUID_EXTENDED_TOPOLOGY_ECX ExtendedTopologyEcx;
- UINT32 MaxCpuIdIndex;
- UINT32 SubIndex;
- UINTN LevelType;
- UINT32 MaxLogicProcessorsPerPackage;
- UINT32 MaxCoresPerPackage;
+ BOOLEAN TopologyLeafSupported;
+ CPUID_VERSION_INFO_EBX VersionInfoEbx;
+ CPUID_VERSION_INFO_EDX VersionInfoEdx;
+ CPUID_CACHE_PARAMS_EAX CacheParamsEax;
+ CPUID_EXTENDED_TOPOLOGY_EAX ExtendedTopologyEax;
+ CPUID_EXTENDED_TOPOLOGY_EBX ExtendedTopologyEbx;
+ CPUID_EXTENDED_TOPOLOGY_ECX ExtendedTopologyEcx;
+ CPUID_AMD_EXTENDED_CPU_SIG_ECX AmdExtendedCpuSigEcx;
+ CPUID_AMD_PROCESSOR_TOPOLOGY_EBX AmdProcessorTopologyEbx;
+ CPUID_AMD_PROCESSOR_TOPOLOGY_ECX AmdProcessorTopologyEcx;
+ CPUID_AMD_VIR_PHY_ADDRESS_SIZE_ECX AmdVirPhyAddressSizeEcx;
+ UINT32 SignatureEbx;
+ UINT32 SignatureEcx;
+ UINT32 SignatureEdx;
+ UINT32 MaxStandardCpuIdIndex;
+ UINT32 MaxExtendedCpuIdIndex;
+ UINT32 SubIndex;
+ UINTN LevelType;
+ UINT32 MaxLogicProcessorsPerPackage;
+ UINT32 MaxCoresPerPackage;
+ UINT32 MaxThreadPerPackageMask;
+ UINT32 ActualThreadPerPackageMask;
+ UINT32 MaxCoresPerNode;
+ UINT32 CorePerNodeMask;
+ UINT32 ApicIdShift;
+ UINTN ThreadBits;
+ UINTN CoreBits;
//
// Check if the processor is capable of supporting more than one logical processor.
@@ -1082,10 +1098,10 @@ GetProcessorLocationByApicId (
AsmCpuid(CPUID_VERSION_INFO, NULL, NULL, NULL, &VersionInfoEdx.Uint32);
if (VersionInfoEdx.Bits.HTT == 0) {
if (Thread != NULL) {
- *Thread = 0;
+ *Thread = 0;
}
if (Core != NULL) {
- *Core = 0;
+ *Core = 0;
}
if (Package != NULL) {
*Package = 0;
@@ -1093,24 +1109,24 @@ GetProcessorLocationByApicId (
return;
}
+ //
+ // Assume three-level mapping of APIC ID: Package|Core|Thread.
+ //
ThreadBits = 0;
CoreBits = 0;
//
- // Assume three-level mapping of APIC ID: Package:Core:SMT.
+ // Get max index of CPUID and vendor's signature
//
- TopologyLeafSupported = FALSE;
-
- //
- // Get the max index of basic CPUID
- //
- AsmCpuid(CPUID_SIGNATURE, &MaxCpuIdIndex, NULL, NULL, NULL);
+ AsmCpuid(CPUID_SIGNATURE, &MaxStandardCpuIdIndex, &SignatureEbx, &SignatureEcx, &SignatureEdx);
+ AsmCpuid(CPUID_EXTENDED_FUNCTION, &MaxExtendedCpuIdIndex, NULL, NULL, NULL);
//
// If the extended topology enumeration leaf is available, it
// is the preferred mechanism for enumerating topology.
//
- if (MaxCpuIdIndex >= CPUID_EXTENDED_TOPOLOGY) {
+ TopologyLeafSupported = FALSE;
+ if (MaxStandardCpuIdIndex >= CPUID_EXTENDED_TOPOLOGY) {
AsmCpuidEx(
CPUID_EXTENDED_TOPOLOGY,
0,
@@ -1160,27 +1176,87 @@ GetProcessorLocationByApicId (
}
if (!TopologyLeafSupported) {
+ //
+ // Get logical processor count
+ //
AsmCpuid(CPUID_VERSION_INFO, NULL, &VersionInfoEbx.Uint32, NULL, NULL);
MaxLogicProcessorsPerPackage = VersionInfoEbx.Bits.MaximumAddressableIdsForLogicalProcessors;
- if (MaxCpuIdIndex >= CPUID_CACHE_PARAMS) {
- AsmCpuidEx(CPUID_CACHE_PARAMS, 0, &CacheParamsEax.Uint32, NULL, NULL, NULL);
- MaxCoresPerPackage = CacheParamsEax.Bits.MaximumAddressableIdsForLogicalProcessors + 1;
+
+ //
+ // Assume single-core processor
+ //
+ MaxCoresPerPackage = 1;
+
+ //
+ // Check for topology extensions on AMD processor
+ //
+ if (SignatureEbx == CPUID_SIGNATURE_GENUINE_AMD_EBX &&
+ SignatureEcx == CPUID_SIGNATURE_GENUINE_AMD_ECX &&
+ SignatureEdx == CPUID_SIGNATURE_GENUINE_AMD_EDX) {
+ if (MaxExtendedCpuIdIndex >= CPUID_AMD_PROCESSOR_TOPOLOGY) {
+ AsmCpuid(CPUID_EXTENDED_CPU_SIG, NULL, NULL, &AmdExtendedCpuSigEcx.Uint32, NULL);
+ if (AmdExtendedCpuSigEcx.Bits.TopologyExtensions != 0) {
+ AsmCpuid(CPUID_AMD_PROCESSOR_TOPOLOGY, NULL, &AmdProcessorTopologyEbx.Uint32,
+ &AmdProcessorTopologyEcx.Uint32, NULL);
+ //
+ // Get cores per processor package
+ //
+ MaxCoresPerPackage = MaxLogicProcessorsPerPackage / (AmdProcessorTopologyEbx.Bits.ThreadsPerCore + 1);
+
+ //
+ // Account for actual thread count (e.g., SMT disabled)
+ //
+ AsmCpuid(CPUID_VIR_PHY_ADDRESS_SIZE, NULL, NULL, &AmdVirPhyAddressSizeEcx.Uint32, NULL);
+ MaxThreadPerPackageMask = 1 << AmdVirPhyAddressSizeEcx.Bits.ApicIdCoreIdSize;
+ ActualThreadPerPackageMask = 1;
+ while (ActualThreadPerPackageMask < MaxLogicProcessorsPerPackage) {
+ ActualThreadPerPackageMask <<= 1;
+ }
+
+ //
+ // Adjust APIC Id to report concatenation of Package|Core|Thread.
+ //
+ if (ActualThreadPerPackageMask < MaxThreadPerPackageMask) {
+ MaxCoresPerNode = MaxCoresPerPackage / (AmdProcessorTopologyEcx.Bits.NodesPerProcessor + 1);
+
+ CorePerNodeMask = 1;
+ while (CorePerNodeMask < MaxCoresPerNode) {
+ CorePerNodeMask <<= 1;
+ }
+ CorePerNodeMask -= 1;
+
+ ApicIdShift = 0;
+ do {
+ ApicIdShift += 1;
+ ActualThreadPerPackageMask <<= 1;
+ } while (ActualThreadPerPackageMask < MaxThreadPerPackageMask);
+
+ InitialApicId = ((InitialApicId & ~CorePerNodeMask) >> ApicIdShift) | (InitialApicId & CorePerNodeMask);
+ }
+ }
+ }
}
else {
//
- // Must be a single-core processor.
+ // Extract core count based on CACHE information
//
- MaxCoresPerPackage = 1;
+ if (MaxStandardCpuIdIndex >= CPUID_CACHE_PARAMS) {
+ AsmCpuidEx(CPUID_CACHE_PARAMS, 0, &CacheParamsEax.Uint32, NULL, NULL, NULL);
+ if (CacheParamsEax.Uint32 != 0) {
+ MaxCoresPerPackage = CacheParamsEax.Bits.MaximumAddressableIdsForLogicalProcessors + 1;
+ }
+ }
}
ThreadBits = (UINTN)(HighBitSet32(MaxLogicProcessorsPerPackage / MaxCoresPerPackage - 1) + 1);
- CoreBits = (UINTN)(HighBitSet32(MaxCoresPerPackage - 1) + 1); }
+ CoreBits = (UINTN)(HighBitSet32(MaxCoresPerPackage - 1) + 1);
+ }
if (Thread != NULL) {
- *Thread = InitialApicId & ((1 << ThreadBits) - 1);
+ *Thread = InitialApicId & ((1 << ThreadBits) - 1);
}
if (Core != NULL) {
- *Core = (InitialApicId >> ThreadBits) & ((1 << CoreBits) - 1);
+ *Core = (InitialApicId >> ThreadBits) & ((1 << CoreBits) - 1);
}
if (Package != NULL) {
*Package = (InitialApicId >> (ThreadBits + CoreBits));
--
2.7.4
next prev parent reply other threads:[~2017-06-16 22:55 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-16 22:56 [PATCH v4 0/2] UefiCpuPkg: Add CPUID support for AMD Leo Duran
2017-06-16 22:56 ` [PATCH v4 1/2] UefiCpuPkg: Add CPUID definitions " Leo Duran
2017-06-16 22:57 ` Leo Duran [this message]
2017-08-01 19:32 ` [PATCH v4 0/2] UefiCpuPkg: Add CPUID support " Duran, Leo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-list from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1497653820-15192-3-git-send-email-leo.duran@amd.com \
--to=devel@edk2.groups.io \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox