ETC 26049

BIOS and Kernel Developer's
Guide for
AMD Athlon 64
and
AMD Opteron Processors
TM
TM
Publication # 26094
Revision: 3.06
Issue Date: September 2003
© 2002, 2003 Advanced Micro Devices, Inc. All rights reserved.
The contents of this document are provided in connection with Advanced Micro Devices,
Inc. (“AMD”) products. AMD makes no representations or warranties with respect to the
accuracy or completeness of the contents of this publication and reserves the right to make
changes to speciﬁcations and product descriptions at any time without notice. No license,
whether express, implied, arising by estoppel or otherwise, to any intellectual property
rights is granted by this publication. Except as set forth in AMD’s Standard Terms and
Conditions of Sale, AMD assumes no liability whatsoever, and disclaims any express or
implied warranty, relating to its products including, but not limited to, the implied warranty of merchantability, ﬁtness for a particular purpose, or infringement of any intellectual property right.
AMD’s products are not designed, intended, authorized or warranted for use as components in systems intended for surgical implant into the body, or in other applications
intended to support or sustain life, or in any other application in which the failure of
AMD’s product could create a situation where personal injury, death, or severe property or
environmental damage may occur. AMD reserves the right to discontinue or make
changes to its products at any time without notice.
Trademarks
AMD, the AMD Arrow logo, AMD Athlon, AMD Opteron, and combinations thereof, 3DNow!, and AMD PowerNow! are trademarks
of Advanced Micro Devices, Inc.
HyperTransport is a licensed trademark of the HyperTransport Technology Consortium.
MMX is a trademark and Pentium is a registered trademark of Intel Corporation.
Microsoft and Windows are registered trademarks of Microsoft Corporation.
Other product names used in this publication are for identiﬁcation purposes only and may be trademarks of their respective companies.
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Contents
Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
1
Introduction and Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
1.1
About This Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
1.2
Related Documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
1.3
Conventions and Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
1.3.1
1.4
2
2.1
Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
Register Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™
processors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
Processor Initialization and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
Bootstrap Processor Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
2.1.1
Detecting AP and Initializing Routing Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
2.1.2
Initializing Noncoherent HyperTransport™ Technology Devices . . . . . . . . . . . . . .22
2.1.3
Initializing Link Width and Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
2.1.4
Initializing the Memory Controller on All Processor Nodes . . . . . . . . . . . . . . . . . .23
2.1.5
Initializing the Address Map Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
2.2
Application Processor Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
2.3
BIOS Requirement for 64-Bit Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
2.3.1
Sizing and Testing Memory above 4 Gbytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
2.3.2
Using BIOS Callbacks in 64-Bit Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
3
3.1
Memory System Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
Configuration Space Accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
3.1.1
Configuration Address Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
3.1.2
Configuration Data Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
3.2
Memory System Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
3.3
Function 0—HyperTransport™ Technology Configuration . . . . . . . . . . . . . . . . . . . . . .27
3.3.1
Device/Vendor ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
3.3.2
Status/Command Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Contents
3
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
3.3.3
Class Code/Revision ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
3.3.4
Header Type Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
3.3.5
Capabilities Pointer Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
3.3.6
Routing Table Node i Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32
3.3.7
Node ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
3.3.8
Unit ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
3.3.9
HyperTransport™ Transaction Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . .36
3.3.10
HyperTransport™ Initialization Control Register . . . . . . . . . . . . . . . . . . . . . . . . . .40
3.3.11
LDTi Capabilities Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
3.3.12
LDTi Link Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43
3.3.13
LDTi Frequency/Revision Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47
3.3.14
LDTi Feature Capability Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48
3.3.15
LDTi Buffer Count Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49
3.3.16
LDTi Bus Number Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
3.3.17
LDTi Type Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
3.4
Function 1—Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53
3.4.1
Device/Vendor ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
3.4.2
Class Code/Revision ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
3.4.3
Header Type Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56
3.4.4
DRAM Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57
3.4.5
Memory-Mapped I/O Address Map Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
3.4.6
PCI I/O Address Map Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62
3.4.7
Configuration Map Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
3.5
Function 2—DRAM Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
3.5.1
Device/Vendor ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67
3.5.2
Class Code/Revision ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68
3.5.3
Header Type Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
3.5.4
DRAM CS Base Address Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
3.5.5
DRAM CS Mask Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
3.5.6
DRAM Bank Address Mapping Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73
4
Contents
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.5.7
DRAM Timing Low Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
3.5.8
DRAM Timing High Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
3.5.9
DRAM Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82
3.5.10
DRAM Delay Line Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88
3.6
Function 3—Miscellaneous Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88
3.6.1
Device/Vendor ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90
3.6.2
Class Code/Revision ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91
3.6.3
Header Type Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91
3.6.4
Machine Check Architecture (MCA) Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
3.6.5
ECC and Chip Kill Error Checking and Correction . . . . . . . . . . . . . . . . . . . . . . . .108
3.6.6
Scrub Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110
3.6.7
DRAM Scrub Address Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
3.6.8
XBAR Flow Control Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113
3.6.9
XBAR-to-SRI Buffer Count Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118
3.6.10
Display Refresh Flow Control Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119
3.6.11
Power Management Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120
3.6.12
GART Aperture Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122
3.6.13
GART Aperture Base Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123
3.6.14
GART Table Base Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124
3.6.15
GART Cache Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
3.6.16
Clock Power/Timing Low Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
3.6.17
Clock Power/Timing High Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127
3.6.18
HyperTransport™ FIFO Read Pointer Optimization Register . . . . . . . . . . . . . . . .128
3.6.19
Thermtrip Status Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129
3.6.20
Northbridge Capabilities Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .131
4
4.1
DRAM Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
Programming Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
4.1.1
SPD ROM-Based Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
4.1.2
Non-SPD ROM-Based Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137
4.1.3
DRAM Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139
Contents
5
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
4.2
5
26094
Rev. 3.06
September 2003
DRAM Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139
Machine Check Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143
5.1
Determining Machine Check Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143
5.2
Machine Check Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143
5.2.1
5.3
Sources of Machine Check Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144
Machine Check Architecture Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146
5.3.1
Global Machine Check Model-Specific Registers (MSRs) . . . . . . . . . . . . . . . . . .146
5.3.2
Error Reporting Bank Machine Check MSRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . .149
5.4
Error Reporting Banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152
5.4.1
Data Cache (DC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152
5.4.2
Instruction Cache (IC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156
5.4.3
Bus Unit (BU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .158
5.4.4
Load Store Unit (LS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161
5.4.5
Northbridge (NB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162
5.5
Initializing the Machine Check Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163
5.6
Using Machine Check Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163
5.6.1
6
Handling Machine Check Exceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
System Management Mode (SMM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167
6.1
SMM Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167
6.2
Operating Mode and Default Register Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167
6.3
SMM State Save Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
6.4
SMM Initial State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171
6.5
SMM-Revision Identifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172
6.6
SMM Base Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173
6.7
Auto Halt Restart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173
6.8
SMM I/O Trap and I/O Restart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .174
6.8.1
SMM I/O Trap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .175
6.8.2
SMM I/O Restart Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .175
6.9
Exceptions and Interrupts in SMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176
6.10
Protected SMM and ASeg/TSeg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176
6
Contents
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
6.10.1
SMM_MASK Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .177
6.10.2
SMM_ADDR Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178
6.10.3
SMM ASeg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178
6.10.4
SMM TSeg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179
6.10.5
Closing SMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
6.10.6
Locking SMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
6.11
7
SMM Special Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
Advanced Programmable Interrupt Controller (APIC) . . . . . . . . . . . . . . . . . . . . . . . . .181
7.1
Interrupt Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
7.2
Vectored Interrupt Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
7.3
Spurious Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .183
7.3.1
Spurious Interrupts Caused by Timer Tick Interrupt . . . . . . . . . . . . . . . . . . . . . . .183
7.4
Lowest-Priority Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .184
7.5
Inter-Processor Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
7.6
APIC Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
7.7
State at Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
7.8
Register Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
7.8.1
APIC ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .187
7.8.2
APIC Version Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .187
7.8.3
Task Priority Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188
7.8.4
Arbitration Priority Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188
7.8.5
Processor Priority Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188
7.8.6
End of Interrupt Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
7.8.7
Logical Destination Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
7.8.8
Destination Format Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190
7.8.9
Spurious Interrupt Vector Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190
7.8.10
In-Service Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
7.8.11
Trigger Mode Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
7.8.12
Interrupt Request Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193
7.8.13
Error Status Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .194
Contents
7
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
7.8.14
Interrupt Command Register Low . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195
7.8.15
Interrupt Command Register High . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .196
7.8.16
Timer Local Vector Table Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197
7.8.17
Performance Counter Local Vector Table Entry . . . . . . . . . . . . . . . . . . . . . . . . . .197
7.8.18
Local Interrupt 0 (Legacy INTR) Local Vector Table Entry Register . . . . . . . . . .198
7.8.19
Local Interrupt 1 (Legacy NMI) Local Vector Table Entry . . . . . . . . . . . . . . . . . .199
7.8.20
Error Local Vector Table Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199
7.8.21
Timer Initial Count Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
7.8.22
Timer Current Count Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
7.8.23
Timer Divide Configuration Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
8
HyperTransport™ Technology Configuration and Enumeration . . . . . . . . . . . . . . . . .203
8.1
Initial Configuration Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203
8.2
One-Node Coherent HyperTransport™ Technology Initialization . . . . . . . . . . . . . . . .204
8.3
Two-Node Coherent HyperTransport™ Technology Initialization . . . . . . . . . . . . . . . .204
8.4
Generic HyperTransport™ Technology Configuration . . . . . . . . . . . . . . . . . . . . . . . . .205
8.4.1
9
Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
Power and Thermal Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .215
9.1
Stop Grant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .216
9.2
C-States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .218
9.2.1
C1 Halt State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .218
9.2.2
C2 and C3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .218
9.3
Throttling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .218
9.4
Processor ACPI Thermal Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .219
9.5
Processor Performance States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .219
9.5.1
BIOS Requirements for P-State Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .219
9.5.2
BIOS-Initiated P-State Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221
9.5.3
BIOS Support for Operating System/CPU Driver-Initiated P-State Transitions . .221
9.5.4
Processor Driver Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
9.5.5
P-State Transition Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
9.6
8
ACPI 2.0 Processor P-State Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230
Contents
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
9.6.1
_PCT (Performance Control) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230
9.6.2
_PSS (Performance-Supported States) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231
9.6.3
_PPC (Performance Present Capabilities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .237
9.6.4
PSTATE_CNT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .239
9.6.5
CST_CNT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .239
9.7
BIOS Support for AMD PowerNow!™ Software with Legacy Operating Systems . . .239
9.8
System Configuration for Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .241
9.8.1
Chipset Configuration for Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . .241
9.8.2
Processor Configuration for Power Management . . . . . . . . . . . . . . . . . . . . . . . . . .241
10
Performance Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245
10.1
Performance Counters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245
10.2
Performance Event-Select Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246
11
BIOS Checklist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
11.1
CPUID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
11.2
CPU Speed Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
11.3
HyperTransport™ Link Frequency Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
11.4
Multiprocessing Capability Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .256
11.5
Model-Specific Registers (MSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .256
11.6
Machine Check Architecture (MCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257
11.7
Memory Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257
11.7.1
I/O and Memory Type and Range Registers (IORRs, MTRRs) . . . . . . . . . . . . . . .257
11.7.2
Memory Map Registers (MMRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .258
11.8
Cache Testing and Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .258
11.9
Memory System Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .258
11.10
XSDT Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259
11.11
Detect Target Operating Mode Callback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259
11.12
SMM Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .260
12
12.1
12.1.1
Processor Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261
General Model-Specific Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261
System Software Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263
Contents
9
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
12.1.2
Memory Typing Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265
12.1.3
APIC Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282
12.1.4
Software Debug Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .283
12.1.5
Performance Monitoring Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284
12.2
AMD Athlon™ 64 processor and AMD Opteron™ Processor Model-Specific
Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284
12.2.1
Feature Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .286
12.2.2
Identification Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .292
12.2.3
Memory Typing Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .292
12.2.4
I/O Range Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .293
12.2.5
System Call Extension Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295
12.2.6
Segmentation Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .297
12.2.7
Power Management Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299
12.2.8
IO and Configuration Space Trapping to SMI . . . . . . . . . . . . . . . . . . . . . . . . . . . .302
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .307
Index of Register Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .313
10
Contents
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
List of Figures
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Interleave Example (IntlvEn Relation to IntlvSel) ...................................................59
Default SMM Memory Map ...................................................................................168
An Eight-Node Configuration.................................................................................205
High-Level P-state Transition Flow........................................................................225
Example P-State Transition Timing Diagram.........................................................226
Phase 1: Core Voltage Transition Flow ..................................................................227
Phase 2: Core Frequency Transition Flow..............................................................228
Phase 3: Core Voltage Transition Flow ..................................................................229
List of Figures
11
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
12
List of Figures
26094
Rev. 3.06
September 2003
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
List of Tables
Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Table 16.
Table 17.
Table 18.
Table 19.
Table 20.
Table 21.
Table 22.
Table 23.
Table 24.
Table 25.
Table 26.
Table 27.
Table 28.
Table 29.
Table 30.
Table 31.
Table 32.
Table 33.
Table 34.
Table 35.
Table 36.
Table 37.
Table 38.
Table 39.
Table 40.
Related Documents .........................................................................................................18
Function 0 Configuration Registers ................................................................................27
Function 1 Configuration Registers ................................................................................53
Function 2 Configuration Registers ................................................................................66
DRAM CS Base Address and DRAM CS Mask Registers.............................................70
Mapping Host Address Lines to Memory Address Lines (64-Bit Interface)..................74
Mapping Host Address Lines to Memory Address Lines (128-Bit Interface)................75
Swapped physical address lines for interleaving with 64-bit interface...........................77
Swapped physical address lines for interleaving with 128-bit interface.........................77
Function 3 Configuration Registers ................................................................................89
Error Code Field Formats................................................................................................99
Transaction Type Bits (TT).............................................................................................99
Cache Level Bits (LL).....................................................................................................99
Memory Transaction Type Bits (RRRR) ........................................................................99
Participation Processor Bits (PP) ..................................................................................100
Time-Out Bit (T) ...........................................................................................................100
Memory or I/O Bits (II).................................................................................................100
Northbridge Error Codes...............................................................................................100
Northbridge Error Status Bit Settings ...........................................................................103
ECC Syndromes ............................................................................................................108
Chip Kill ECC Syndromes ............................................................................................109
Scrub Rate Control Values............................................................................................111
XBAR Input Buffers .....................................................................................................113
Default XBAR Command Buffer Allocation................................................................114
Default Virtual Channel Command Buffer Allocation .................................................114
An Example of a Non Default Virtual Channel Command Buffer Allocation .............115
GART PTE Organization..............................................................................................124
Trwt Values...................................................................................................................137
RdPreamble Values.......................................................................................................138
Sources of Machine Check Errors.................................................................................145
Valid MC0_ADDR Bits................................................................................................156
Valid MC1_ADDR Bits................................................................................................158
Valid MC2_ADDR Bits................................................................................................161
SMM Save State (Offset FE00–FFFFh) .......................................................................169
SMM Entry State...........................................................................................................172
SMM ASeg-Enabled Memory Types............................................................................179
SMM TSeg-Enabled Memory Types ............................................................................179
APIC Register Summary...............................................................................................186
Valid Combinations of ICR Fields................................................................................196
Power Management Categories.....................................................................................215
List of Tables
13
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 41.
Table 42.
Table 43.
Table 44.
Table 45.
Table 46.
Table 47.
Table 48.
Table 49.
Table 50.
Table 51.
Table 52.
Table 53.
Table 54.
Table 55.
Table 56.
Table 57.
Table 58.
14
26094
Rev. 3.06
September 2003
ACPI-State Support by System Class ...........................................................................216
Required SMAF Code to Stop Grant Mapping.............................................................217
Low FID Frequency Table (< 1600 MHz)....................................................................223
High FID Frequency Table (>= 1600 MHz) .................................................................224
Sample VST Values ......................................................................................................233
MVS Values ..................................................................................................................234
RVO Values ..................................................................................................................235
VID Code Voltages .......................................................................................................235
IRT Values ....................................................................................................................236
_PSS Status Field ..........................................................................................................237
Performance State Block Structure ...............................................................................240
Configuration Register Settings for Power Management .............................................242
FADT Table Entries......................................................................................................243
Processor Functional Unit Encoding.............................................................................248
Performance Monitor Events ........................................................................................248
General MSRs ...............................................................................................................261
AMD Athlon™ 64 Processor and AMD Opteron™ Processor MSRs .........................285
FID Code Translations ..................................................................................................300
List of Tables
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Revision History
Date
Rev.
Description
September
2003
3.06
Clarified 32ByteEn (Function 2, Offset 90h) programming in “DRAM Configuration”
section.
Removed SleepVIDCnt (Function 3, Offset D4h) and added a programming
requirement.
Added reference to the AMD Athlon™ 64 Data Sheet, order# 24659 for 3 DIMM configuration to “Programming Interface” section.
Corrected requirement for supported DIMMs in “Registered or Unbuffered DIMMs”
section.
Corrected RdPreamble values for 3 DIMM configuration in “Read Preamble Time”
section.
Added RdPreamble and Trwt values for registered DDR400.
Added EnRefUseFreeBuf and DisCohLdtCfg to NB_CFG register (MSR
C001_001Fh).
Cleanup related to AMD Athlon™ 64 and AMD Opteron™ use.
July 2003
3.04
Added examples to “DRAM Address Mapping in Interleaving Mode” and “DRAM
Address Mapping in Non-Interleaving Mode” sections.
Added recommendation for setting “HyperTransport™ FIFO Read Pointer Optimization Register”.
Added “Spurious Interrupts Caused by Timer Tick Interrupt” section.
Added “XSDT Table” section.
Added “Detect Target Operating Mode Callback” section.
Modified “Multiprocessing Capability Detection” section.
Added requirements for AMD Cool’n’Quiet™ technology and BIOS P-state transitions to “Power and Thermal Management” chapter.
Added ClkRampHyst (Function 3, Offset D4h) setting to “Power and Thermal Management” chapter.
General cleanup in “Power and Thermal Management” chapter.
Added “Twtr (Write to Read Delay)” section.
Added a requirement to “GART Aperture Base Register”.
Added a requirement for GartEn (Function 3, Offset 90h) programming.
Removed recommandation for extended range temperature sensors in DiodeOffset (Function 3, Offset E4h) definition.
Added a requirement for extended configuration space access to “MemoryMapped I/O Address Map Registers” section.
Revision History
15
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
Date
Rev.
Description
May 2003
3.02
Added “ECC and Chip Kill Error Checking and Correction” section.
September 2003
Clarified “Scrub Control Register” and “DRAM Scrub Address Registers” sections.
Clarified LinkFail (Function 0, Offsets 84h, A4h, C4h).
Clarified node interleaving in “DRAM Address Map” section.
Clarified and corrected DRAM address interleaving in “DRAM CS Base Address
Registers” and “DRAM Bank Address Mapping Register” sections.
Clarified “tCL (CAS Latency)” section.
Added “piggyback scrubbing” description to “Sources of Machine Check Errors”
section.
Clarified Table 30, “Sources of Machine Check Errors,” on page 145.
Clarified MCA NB PCI configuration register to MCA MSR mapping in “Northbridge (NB)” section.
Corrected MaxLVTEntry (APIC_VER Register, Offset 30h) default value.
General cleanup in “Power and Thermal Management” chapter.
Added “HyperTransport™ Link Frequency Selection” section.
April 2003
16
3.00
Initial Public release.
Revision History
26094
Rev. 3.06
1
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Introduction and Overview
The BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and AMD Opteron™ Processors
is intended for programmers involved in the development of low-level BIOS (basic input/output
system) functions, drivers, and operating system kernel modules. It assumes previous experience in
microprocessor programming, as well as fundamental knowledge of legacy x86 and AMD64
microprocessor architecture. The reader should also have previous experience in BIOS or OS kernel
design issues, as related to microprocessor systems, and a familiarity with various platform
technologies, such as DDR, HyperTransport™ technology, and JTAG.
1.1
About This Guide
This guide covers the implementation-speciﬁc features of AMD Athlon™ 64 and AMD Opteron™
processors, as opposed to architectural features. AMD64 architectural features include the AMD64
technology, general-purpose, multimedia, and x87 ﬂoating-point registers, as well as other
programmer-visible features deﬁned to be constant across all processors. A subset of implementationspeciﬁc features are not deﬁned by the processor architectural speciﬁcations. These implementationspeciﬁc features may differ in various details from one implementation to the next.
Note: The term processor in this document refers to AMD AthlonTM 64 processor architecture and
AMD OpteronTM processor architecture. This document covers both classes of devices. For
details about differences between them, see the AMD AthlonTM 64 Processor Data Sheet,
order# 24659 and the AMD OpteronTM Processor Data Sheet, order# 23932.
The implementation-speciﬁc features covered in the following chapters include:
•
Model-speciﬁc registers
•
Processor initialization
•
Integrated memory system conﬁguration
•
HyperTransport technology fabric initialization
•
Performance monitoring and special debug features
•
DRAM conﬁguration
•
Machine check error codes
•
Thermal and power management
Chapter 1
Introduction and Overview
17
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
1.2
26094
Rev. 3.06
September 2003
Related Documents
The references listed in Table 1 may prove invaluable towards a complete understanding of the
subject matter in this volume.
Table 1.
Related Documents
Title
Order#
AMD64 Architecture Programmer’s Manual, Volume 1,
Application Programming
24592
AMD64 Architecture Programmer’s Manual, Volume 2,
System Programming
24593
AMD64 Architecture Programmer’s Manual, Volume 3,
General-Purpose and System Instructions
24594
(three-volume kit)
AMD64 Architecture Programmer’s Manual, Volume 4,
128-Bit Media Instructions
AMD64 Architecture Programmer’s Manual, Volume 5,
64-Bit Media and x87 Floating-Point Instructions
AMD OpteronTM Processor Data Sheet
23932
AMD AthlonTM 64 Processor Data Sheet
24659
AMD Processor Recognition Application Note
20734
CPUID Guide for AMD AthlonTM 64 and AMD OpteronTM Processors
25481
Revision Guide for the AMD AthlonTM 64 and AMD OpteronTM Processors
25759
HyperTransport™ I/O Link Specification, Rev. 1.03
http://www.hypertransport.org
1.3
Conventions and Definitions
Some of the following deﬁnitions assume a knowledge of the legacy x86 architecture. See Table 1 for
documents that include information on the legacy x86 architecture.
Additional deﬁnitions are provided in the glossary at the end of this book, beginning on page 307.
That glossary includes the most important terminology of the AMD64 architecture.
1.3.1
Notation
1011b. A binary value. In this example, a 4-bit value is shown.
F0EAh. A hexadecimal value. In this example, a 2-byte value is shown.
[1,2]. A range that includes the left-most value (in this case, 1) but excludes the right-most value (in
this case, 2).
7–4. A bit range, from bit 7 to 4, inclusive. The high-order bit is shown ﬁrst.
#GP(0). Notation indicating a general-protection exception (#GP) with error code of 0.
18
Introduction and Overview
Chapter 1
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
CR0–CR4. A register range, from register CR0 through CR4, inclusive, with the low-order register
ﬁrst.
CR0.PE = 1. Notation indicating that the PE bit of the CR0 register has a value of 1.
DS:rSI. The contents of a memory location whose segment is DS and whose byte address is located in
the rSI register.
EFER.LME = 0. Notation indicating that the LME bit of the EFER register has a value of 0.
FF /0. Notation indicating that FF is the ﬁrst byte of an opcode and a sub-opcode ﬁeld in the MODRM
byte has a value of 0.
1.4
Register Differences in Revisions of the
AMD Athlon™ 64 and AMD Opteron™ processors
Some changes to the register set are introduced with different silicon revisions. Refer to the Revision
Guide for AMD Athlon™ 64 and AMD Opteron™ Processors, order# 25759 for information about
how to identify different processor revisions. The following summarizes register changes after the
initial revision.
Revision C.
•
MSR C001_0043h (ThermTrip_STATUS Register) deleted.
•
Function 2, offset 90h, DramEn (bit 10), MemClrStatus (bit 11) added.
•
Function 2, offset CCh, DisTscCapture (bit 30) added.
•
Function 3, offset 70h, DispRefReq (bits 21-20) added.
•
Function 3, offset 74h, DispRefReq (bits 22-20) added.
•
MSR C001_001Fh (NB_CFG register), DisDatMsk (bit 36), EnRefUseFreeBuf (bit 9) added.
•
Function 3, offset 90h, DisGartTblWlkPrb (bit 6) added.
•
Function 3, offset D4h, ClkRampHyst (bits 10-8) added.
•
MSR C001_0015 (HWCR register), HLTXSYCEN (bit 12) added.
•
MSR C001_0010 (SYSCFG register), ClVicBlkEn (bit 11) deleted.
•
MSRs C001_0050h, C001_0051h, C001_0052h, C001_0053h (IOTRAP_ADDRi registers), and
MSR C001_0054h (IOTRAP_CTL register) added.
Chapter 1
Introduction and Overview
19
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
20
Introduction and Overview
26094
Rev. 3.06
September 2003
Chapter 1
26094
Rev. 3.06
2
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Processor Initialization and
Configuration
Each AMD Athlon™ 64 and AMD Opteron™ processor based system has one processor with its
HyperTransport™ link connected to a HyperTransport™ I/O hub. When a reset signal is applied, this
processor is initialized as the bootstrap processor (BSP). In a multiple processor system, any other
processors are initialized as application processors (APs). The BSP begins executing code from the
reset vector (0xFFFFFFF0), while each of the APs waits for its Request Disable bit to clear to 0. An
AP does not fetch code until its Request Disable bit is cleared. Both BSP and AP operate in 16-bit
Real mode after reset. The BSP has the Boot Strap Processor bit set in its APIC_BASE (001Bh)
MSR, and each AP has this bit cleared.
The processor node is addressed by its Node ID on the HyperTransport link and can be accessed with
a device number in the PCI conﬁguration space on Bus 0. The Node ID 0 is mapped to Device 24, the
Node ID 1 is mapped to Device 25, and so on. The BSP is initialized with Node ID 0 after reset, and
all APs are initialized with Node ID 7.
The BSP is ready to access its Northbridge and memory controller after its routing table is enabled.
The APs are not accessible from the BSP until the links and the routing table are conﬁgured.
The initial processor states are described in the “Processor Initialization State” section of the AMD64
Architecture Programmer's Manual: Volume 2, System Programming.
2.1
Bootstrap Processor Initialization
The BSP is responsible for the execution of the BIOS Power-On Self Test (POST) and initialization
of APs. The BSP must perform the following tasks:
•
AP detection and routing table initialization
•
Noncoherent HyperTransport device initialization
•
Link width and frequency initialization
•
Memory controller initialization on all processor nodes
•
Address map table initialization
2.1.1
Detecting AP and Initializing Routing Table
The AMD Opteron™ processor has three HyperTransport links; therefore, an AMD Opteron™
system can have from one to eight processors. The BSP must detect all APs in the system, starting
from its adjacent processor. Since the APs are initialized at Node 7, the BSP must set entry 7 of its
Routing Table before the AP can be accessed. The adjacent AP will respond to a PCI conﬁguration
Chapter 2
Processor Initialization and Conﬁguration
21
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
read if it is present. A new Node ID can be assigned to the AP after the routing information is updated
in the routing table. Chapter 8, “HyperTransport™ Technology Conﬁguration and Enumeration”
explains the steps for enumerating and initializing routing tables in one-processor (UP), twoprocessor (DP), and n processor systems.
The AMD Athlon™ 64 processor has one HyperTransport link. Because one link must be connected
to the HyperTransport I/O hub, an AMD Athlon™ 64 system can only be used in a uniprocessor
system.
2.1.2
Initializing Noncoherent HyperTransport™ Technology Devices
Once all APs have been detected and the routing tables and link control registers have been
initialized, the BSP must initialize noncoherent HyperTransport technology devices.
A noncoherent HyperTransport technology device has either one or two links. A tunnel device has
two links and a terminal device has one link. The HyperTransport I/O hub is a typical example of a
terminal noncoherent HyperTransport device. The noncoherent HyperTransport device is identiﬁed
by its Unit ID in a noncoherent HyperTransport link and can be accessed with a device number in the
PCI conﬁguration space. The device number in PCI space is the same as the Unit ID assigned to the
noncoherent HyperTransport device.
After reset, the Unit ID of each noncoherent HyperTransport technology device is initialized with a
value of 0. When a PCI conﬁguration read is performed from the BSP through a noncoherent
HyperTransport link, the noncoherent HyperTransport device connected to the port responds. A new
non-zero Unit ID value must be assigned to this device. The BIOS continues this process until no
device responds at Device 0. The bus number range must be set to the noncoherent HyperTransport
link on the BSP and in the address map table prior to the detection. The link control and status
registers of a noncoherent device are implemented in the capability register block.
The noncoherent HyperTransport technology devices connected to a port on the AP node can be
detected in the same way. Again, the bus number range must be set to the noncoherent
HyperTransport link on this AP and in the address map table. The noncoherent device can be detected
with the starting bus number, Device 0 on the PCI conﬁguration space.
A noncoherent HyperTransport technology device may use more than one Unit ID. The new ID
assigned to a device is its starting ID, the Base Unit ID (BaseUID). The next logic device in this
noncoherent HyperTransport device can be identiﬁed with BaseUID + 1, and so on. The Unit ID
Count ﬁeld in the Capabilities register indicates how many Unit IDs this device uses.
2.1.3
Initializing Link Width and Frequency
The HyperTransport link width and frequency are initialized between the adjacent coherent and/or
noncoherent HyperTransport technology devices during the reset sequence. After AP and
noncoherent HyperTransport device detection, the link width and frequency can be changed based on
the capability of the adjacent devices or the implementation of the system. To make the new link
width and frequency take effect, an LDTSTOP_L needs to be asserted or a warm reset must be
22
Processor Initialization and Conﬁguration
Chapter 2
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
performed. The BIOS must guarantee that LDTSTOP_L assertion for link width and frequency
changes does not occur within 200µs of reset.
2.1.4
Initializing the Memory Controller on All Processor Nodes
The BSP is responsible for conﬁguring the memory controller on all processor nodes and for
initializing the DRAM map table. Chapter 4, “DRAM Conﬁguration,” describes the functionality of
the memory controller and the steps required to initialize memory.
After memory conﬁguration, the memory address MSRs must be set accordingly. The Top of Memory
MSR is used for the top of DRAM below 4 Gbytes, and TOP_MEM2 is set to the top of DRAM
above 4 Gbytes.
2.1.5
Initializing the Address Map Table
Each processor node has an address map table. This table contains the address map for the DRAM
area, the address map for PCI memory spaces (MMIO), the address map for PCI I/O spaces, and the
bus number range for each noncoherent HyperTransport link. The address map table must be
duplicated on all processors in a system.
2.2
Application Processor Initialization
The application processor (AP) waits until its request disable bit is cleared to 0. The BIOS may clear
this bit for an AP a soon as the AP node detection is completed and the routing tables are initialized,
or just before the AP register initialization.
When the request disable bit is cleared, the corresponding AP starts to fetch code from the reset
vector (0xFFFFFFF0). The BSP bit in AP APIC_BASE register is cleared at reset, so it can be used to
terminate the AP execution after the initial APIC initialization.
The AP register initialization is the same as that of other AMD x86 processor families, such as the
AMD Athlon™ processors.
Chapter 2
Processor Initialization and Conﬁguration
23
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
2.3
26094
Rev. 3.06
September 2003
BIOS Requirement for 64-Bit Operation
In general, the BIOS need not be aware of 64-bit mode in POST. There are no additional requirements
to support 64-bit mode. The operating system determines whether to enable 64-bit mode after the
BIOS invokes the operating system loader in legacy mode. There are two areas that need additional
explanation:
•
Sizing and testing memory above 4 Gbytes
•
Using BIOS callback when in 64-bit mode
2.3.1
Sizing and Testing Memory above 4 Gbytes
AMD Athlon™ 64 and AMD Opteron™ processors have extended physical address extension (PAE)
to support a 40-bit address space. Thus, the BIOS can set up a 32-bit page table that allows it to size
and test all physical memory in the system.
2.3.2
Using BIOS Callbacks in 64-Bit Mode
A BIOS callback is a mechanism that allows an operating system to call a service routine in the BIOS.
BIOS callbacks on an AMD Athlon™ 64 processor or an AMD Opteron™ processor platform in 64bit mode are not supported by AMD64 operating systems. An operating system loader must be
invoked in legacy mode with paging disabled, so that the loader can use 16/32-bit BIOS callbacks.
ACPI code and operating system drivers are not affected by AMD64 operating systems that do not
use a BIOS callback in 64-bit mode. ACPI is coded in ASL, which is not affected by the operating
system mode. AMD64 operating system drivers are not allowed to use ROM callbacks.
24
Processor Initialization and Conﬁguration
Chapter 2
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
3
Memory System Configuration
3.1
Configuration Space Accesses
The AMD Athlon™ 64 and AMD Opteron™ processors implement configuration space as defined in
the PCI Local Bus Specification, Rev. 2.2, and the HyperTransport™ I/O Link Specification, Rev.
1.03.
Both coherent HyperTransport links and the compatibility bus (the bus to the HyperTransport I/O
hub) are accessed through Bus 0. Configuration accesses to Bus 0, devices 0 to 23, are transmitted to
the compatibility noncoherent HyperTransport bus. Coherent HyperTransport device configuration
space is accessed by using Bus 0, devices 24 to 31, where Device 24 corresponds to Node 0 and
Device 31 corresponds to Node 7.
All configuration cycles are type 1 on coherent HyperTransport links. When transmitted down a
noncoherent HyperTransport chain by the host bridge, configuration cycles are translated to type 0 if
they target the noncoherent HyperTransport bus immediately behind the host bridge. If they target a
subordinate bus, they remain as type 1.
Accesses to memory system configuration registers are controlled through the Configuration Address
register (see “Configuration Address Register” on page 25) and the Configuration Data register (see
“Configuration Data Register” on page 26), which can be accessed through I/O reads/writes to
addresses 0CF8h and 0CFCh, respectively.
3.1.1
Configuration Address Register
Writes to the Configuration Address register specify the target of a configuration access in terms of
the bus, device, function, and register. Access to the Configuration Address register must be full
doubleword reads or writes.
To access one of the memory system configuration registers defined in this chapter, the bus number
should be 0, the device number should be the target processor node number plus 24, and the function
and register number should be set based on the register function and offset as specified in this chapter.
When the Enable bit of the Configuration Address register is set, reads and writes to the Configuration
Data register access the register specified in the Configuration Address register.
EnReg
31
Chapter 3
24 23
reserved
0CF8h (doubleword)
16 15
BusNum
11 10
DevNum
Memory System Configuration
8
FuncNum
7
2
RegNum
1
0
reserved
Configuration Address Register
25
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Bit
Mnemonic
Function
R/W
31
EnReg
Enable Register
R/W
30–24
reserved
23–16
BusNum
Bus Number
R/W
15–11
DevNum
Device Number
R/W
Rev. 3.06
September 2003
R
10–8
FuncNum
Function Number
R/W
7–2
RegNum
Register Number
R/W
1–0
reserved
R
Field Descriptions
Register Number (RegNum)—Bits 7–2. Specifies the doubleword offset of the configuration
address.
Function Number (FuncNum)—Bits 10–8. Specifies the function number of the configuration
address.
Device Number (DevNum)—Bits 15–11. Specifies the device number of the configuration address.
Bus Number (BusNum)—Bits 23–16. Specifies the bus number of the configuration address.
Enable (EnReg)—Bit 31. When this bit is set, aligned doubleword accesses to the Configuration
Data Register result in configuration-space transactions.
0 = Configuration transactions disabled.
1 = Configuration transactions enabled.
3.1.2
Configuration Data Register
Accesses to the Configuration Data Register are translated to configuration-space transactions at the
address specified by the Configuration Address Register.
Configuration Data Register
0CFCh (Doubleword)
31
0
CfgData
Bits
Mnemonic
Function
R/W
31–0
CfgData
Configuration Data
R/W
Field Descriptions
Configuration Data (CfgData)—Bits 31–0.
26
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.2
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Memory System Configuration Registers
The AMD Athlon™ 64 and AMD Opteron™ processors implement memory system configuration
registers in the “Northbridge” block of the processor. These registers are mapped into the coherent
HyperTransport technology configuration space (Bus 0, device numbers 24–31). The number of
devices implemented is equal to the number of processor nodes in the system. Configuration space
Device 24 corresponds to Node 0 and is normally the bootstrap processor (BSP). Configuration space
device numbers 25 through 31 correspond to nodes 1 through 7, which may or may not be
implemented.
The processor implements configuration registers in PCI configuration space using the following four
headers:
•
Function 0: HyperTransport technology configuration (see page 27)
•
Function 1: Address map configuration (see page 53)
•
Function 2: DRAM configuration (see page 66)
•
Function 3: Miscellaneous configuration (see page 88)
Each of these functions are detailed in the following sections.
Note: Contents of some fields in the headers and HyperTransport technology capability blocks are
maintained through a warm reset. If not specified otherwise, a field is initialized by a warm
reset.
3.3
Function 0—HyperTransport™ Technology
Configuration
Table 2. Function 0 Configuration Registers
Offset
Register Name
Reset
Access
Description
00h
Device ID
Vendor ID (AMD)
1100_1022h
RO
page 29
04h
Status
Command1
0010_0000h
RO
page 30
08h
Base
Class
Code
Subclass
Code
Program
Interface
Revision
ID
0600_0000h
RO
page 30
0Ch
BIST
Header
Type
Latency
Timer
Cache
Line Size
0080_0000h
RO
page 31
10h
Base Address 0
0000_0000h
RO
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
Chapter 3
Memory System Configuration
27
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 2. Function 0 Configuration Registers (Continued)
Offset
Register Name
Reset
Access
14h
Base Address 1
0000_0000h
RO
18h
Base Address 2
0000_0000h
RO
1Ch
Base Address 3
0000_0000h
RO
20h
Base Address 4
0000_0000h
RO
24h
Base Address 5
0000_0000h
RO
28h
Card Bus CIS Pointer
0000_0000h
RO
2Ch
Sub-System ID
Sub-System Vendor ID 0000_0000h
RO
30h
ROM Base Address
0000_0000h
RO
34h
Capabilities Pointer
page 31
RO
38h
reserved
0000_0000h
RO
3Ch
Max
Latency
0000_0000h
RO
40h
Routing Table Node 0
0001_0101h
RW
page 32
44h
Routing Table Node 1
0001_0101h
RW
page 32
48h
Routing Table Node 2
0001_0101h
RW
page 32
4Ch
Routing Table Node 3
0001_0101h
RW
page 32
50h
Routing Table Node 4
0001_0101h
RW
page 32
54h
Routing Table Node 5
0001_0101h
RW
page 32
58h
Routing Table Node 6
0001_0101h
RW
page 32
5Ch
Routing Table Node 7
0001_0101h
RW
page 32
60h
Node ID
0000_0000h
RW
page 34
64h
Unit ID
0000_00E4h
RW
page 35
68h
HyperTransport™ Transaction Control
0F00_0000h
RW
page 36
6Ch
HyperTransport™ Initialization Control
page 40
RW
page 40
80h
LDT0 Capabilities
page 42
RO
page 42
84h
LDT0 Link Control
0011_0000h
RW
page 43
88h
LDT0 Frequency/Revision
page 47
RW
page 47
8Ch
LDT0 Feature Capability
0000_0002h
RO
page 48
90h
LDT0 Buffer Count
page 49
RW
page 49
94h
LDT0 Bus Number
0000_0000h
RW
page 51
98h
LDT0 Type
page 51
RO
page 51
A0h
LDT1 Capabilities
page 42
RO
page 42
A4h
LDT1 Link Control
0011_0000h
RW
page 43
Min GNT
Int Pin
Int Line
Description
page 31
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
28
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 2. Function 0 Configuration Registers (Continued)
Offset
Register Name
Reset
Access
Description
A8h
LDT1 Frequency/Revision
page 47
RW
page 47
ACh
LDT1 Feature Capability
0000_0002h
RO
page 48
B0h
LDT1 Buffer Count
page 49
RW
page 49
B4h
LDT1 Bus Number
0000_0000h
RW
page 51
B8h
LDT1 Type
page 51
RO
page 51
C0h
LDT2 Capabilities
page 42
RO
page 42
C4h
LDT2 Link Control
0011_0000h
RW
page 43
C8h
LDT2 Frequency/Revision
page 47
RW
page 47
CCh
LDT2 Feature Capability
0000_0002h
RO
page 48
D0h
LDT2 Buffer Count
page 49
RW
page 49
D4h
LDT2 Bus Number
0000_0000h
RW
page 51
D8h
LDT2 Type
page 51
RO
page 51
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
3.3.1
Device/Vendor ID Register
This register specifies the device and vendor IDs for the Function 0 registers and is part of the
standard PCI configuration header.
Device/Vendor ID Register
Function 0: Offset 00h
31
16 15
DevID
0
VenID
Bits
Mnemonic
Function
R/W
Reset
31–16
DevID
Device ID
R
1100h
15–0
VenID
Vendor ID
R
1022h
Field Descriptions
Vendor ID (VenID)—Bits 15–0. This read-only value is defined as 1022h for AMD.
Device ID (DevID)—Bits 31–16. This read-only value is defined as 1100h for the HyperTransport
technology configuration function.
Chapter 3
Memory System Configuration
29
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.3.2
26094
Rev. 3.06
September 2003
Status/Command Register
This register contains status and command information for the Function 0 registers and is part of the
standard PCI configuration header.
Status/Command Register
Function 0: Offset 04h
31
16 15
0
Status
Bits
Mnemonic
Command
Function
R/W
Reset
31–16
Status
R
0010h
15–0
Command
R
0000h
Field Descriptions
Command—Bits 15–0. This read-only value is defined as 0000h.
Status—Bits 31–16. This read-only value is defined as 0010h to indicate that the processor has a
capabilities list containing configuration information specific to HyperTransport technology.
3.3.3
Class Code/Revision ID Register
This register specifies the class code and revision for the Function 0 registers and is part of the
standard PCI configuration header.
Class Code/Revision ID Register
31
24 23
Function 0: Offset 08h
16 15
BCC
SCC
Bits
Mnemonic
Function
31–24
BCC
8
7
PI
0
RevID
R/W
Reset
Base Class Code
R
06h
23–16
SCC
Subclass Code
R
00h
15–8
PI
Programming Interface
R
00h
7–0
RevID
Revision ID
R
00h
Field Descriptions
Revision ID (RevID)—Bits 7–0.
Programming Interface (PI)—Bits 15–8. This read-only value is defined as 00h.
Sub Class Code (SCC)—Bits 23–16. This read-only value is defined as 00h.
30
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Base Class Code (BCC)—Bits 31–24. This read-only value is defined as 06h for a host bridge
device.
3.3.4
Header Type Register
This register specifies the header type for the Function 0 registers and is part of the standard PCI
configuration header.
Header Type Register
31
Function 0: Offset 0Ch
24 23
16 15
BIST
HType
Bits
Mnemonic
Function
31–24
BIST
23–16
HType
8
7
LatTimer
0
CLS
R/W
Reset
BIST
R
00h
Header Type
R
80h
15–8
LatTimer
Latency Timer
R
00h
7–0
CLS
Cache Line Size
R
00h
Field Descriptions
CacheLineSize (CLS)—Bits 7–0. This read-only value is defined as 00h.
LatencyTimer (LatTimer)—Bits 15–8. This read-only value is defined as 00h.
HeaderType (HType)—Bits 23–16. This read-only value is defined as 80h to indicate that multiple
functions are present in the configuration header and that the header layout corresponds to a
device header as opposed to a bridge header. See “Register Differences in Revisions of the
AMD Athlon™ 64 and AMD Opteron™ processors” on page 19 for revision information
about this field.
BIST—Bits 31–24. This read-only value is defined as 00h.
3.3.5
Capabilities Pointer Register
This register contains a capabilities pointer to a linked capabilities list of registers specific to
HyperTransport technology, with one set for each HyperTransport link. It is part of the standard PCI
configuration header.
Capabilities Pointer Register
Function 0: Offset 34h
31
8
reserved
Chapter 3
Memory System Configuration
7
0
CapPtr
31
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
Mnemonic
31–8
reserved
7–0
CapPtr
Function
Capability Pointer
26094
Rev. 3.06
R/W
Reset
R
0
September 2003
R
Field Descriptions
Capability Pointer (CapPtr)—Bits 7–0. Points to the first available HyperTransport technology
capabilities block. Depending on the specific HyperTransport link configuration physically
present on the processor, this will be either 80h (LDT0), A0h (LDT1) or C0h (LDT2).
3.3.6
Routing Table Node i Registers
The routing table contains three entries—one for requests RQRoute[7:0], one for responses
RPRoute[7:0], and one for broadcasts BCRoute[7:0].
A maximum of eight nodes with three links per node is supported in an AMD Opteron™ system.
These table entries can be read and written, and its contents are not maintained through a warm reset.
At reset, all table entries are initialized to the value 01h, indicating that packets should be accepted by
this node. Care must be exercised when changing table entries after reset to ensure that connectivity is
not lost during the process.
3.3.6.1
Request Routing Table
Each node contains a routing table for use with directed requests, with one entry for each of the eight
possible destination Node IDs. The value in each entry indicates which outgoing link is used for
request packets directed to that particular destination node (DestNode). A 1 in a given bit position
indicates that the request is routed through the corresponding output link. Bit 0, when set to 1,
indicates that the request must be accepted by this node.
3.3.6.2
Response Routing Table
Each node contains a routing table for use with responses, with one entry for each of the eight
possible destination node IDs. The value in each entry indicates which outgoing link is used for
response packets directed to that particular destination node (DestNode). A 1 in a given bit position
indicates that the response is routed through the corresponding output link. Bit 0, when set to 1,
indicates that the response must be accepted by this node.
3.3.6.3
Broadcast Routing Table
Each node contains a routing table for use with broadcast and probe requests with one entry for each
of the eight possible source node IDs. Each entry contains a single bit for each of the outbound links
from the node. The packet is forwarded on all links with their corresponding links set to a 1. Bit 0
when set to 1 indicates that the broadcast must be accepted by this node.
32
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Routing Table Node 0–7 Registers
31
Function 0: Offset 40h, 44h, 48h, 4Ch,
50h, 54h, 58h, 5Ch
20 19
reserved
Bits
Mnemonic
31–20
reserved
19–16
BCRte
15–12
reserved
11–8
RPRte
7–4
reserved
3–0
RQRte
16 15
BCRte
12 11
reserved
8
RPRte
7
4
reserved
Function
R/W
Reset
R
000h
Broadcast Route
R/W
1h
R
0h
Response Route
R/W
1h
R
0h
Request Route
R/W
1h
3
0
RQRte
Field Descriptions
Request Route (RQRte)—Bits 3–0. The Request Route field defines the node or link to which a
request packet is forwarded. Request packets are routed to only one destination and the table
index is based on the destination Node ID.
Bit [0] = Route to this node
Bit [1] = Route to Link 0
Bit [2] = Route to Link 1
Bit [3] = Route to Link 2
Response Route (RPRte)—Bits 11–8. The Response Route field defines the node or link to which a
response packet is forwarded. Response packets are routed to only one destination and the
table index is based on the destination Node ID.
Bit [0] = Route to this node
Bit [1] = Route to Link 0
Bit [2] = Route to Link 1
Bit [3] = Route to Link 2
Broadcast Route (BCRte)—Bits 19–16. The Broadcast Route field defines the node or link(s) to
which a broadcast packet is forwarded. Broadcasts may be routed to more than one
destination. The Node ID that indexes into the table is the source Node ID.
Bit [0] = Route to this node
Bit [1] = Route to Link 0
Bit [2] = Route to Link 1
Bit [3] = Route to Link 2
Chapter 3
Memory System Configuration
33
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.3.7
26094
Rev. 3.06
September 2003
Node ID Register
The Node ID register contains node ID, node count and CPU count information. In general the
number of CPUs will be the same as the number of nodes but separate fields are provided in order to
allow for nodes which may contain more than one CPU per node.
It is expected that system configuration software will program the Node ID as part of coherent
HyperTransport link configuration. Correct system operation depends on an assignment of distinct
node ID values not exceeding NodeCnt[2:0]. For example, the Node IDs in a 4-node system must be
0, 1, 2, and 3; an example of an incorrect Node ID assignment in this system is 0, 1, 3, and 4. The
Node ID will also be used as the initial APIC ID for the local APIC.
Bits
Mnemonic
31–20
reserved
19–16
CPUCnt
15
reserved
14–12
LkNode
11
reserved
10–8
SbNode
7
reserved
6–4
NodeCnt
3
reserved
2–0
NodeID
16 15 14
12 11 10
CpuCnt
LkNode
7
6
4
3
SbNode
NodeCnt
reserved
reserved
8
reserved
20 19
reserved
31
Function 0: Offset 60h
reserved
Node ID Register
Function
R/W
Reset
R
0
CPU Count
R/W
0
R
0
Lock Controller Node ID
R/W
0
R
0
HyperTransport I/O Hub Node ID
R/W
0
R
0
Node Count
R/W
0
R
0
This Node ID
R/W
2
0
NodeId
Field Descriptions
Node ID (NodeID)—Bits 2–0. Defines the Node ID of this node. It resets to 0 for the boot strap
processor (BSP) and to 7h for all other nodes.
Node Count (NodeCnt)—Bits 6–4. Specifies the number of coherent nodes in the system. Note that
the hardware allows only values to be programmed into this field that are consistent with the
multiprocessor capabilities of the device, as specified in “Northbridge Capabilities Register”
on page 131. This field will not be updated if there is an attempt to write a values that is
inconsistent with the specified multiprocessor capabilities.
000b = 1 node
001b = 2 nodes
010b = 3 nodes
011b = 4 nodes
111b = 8 nodes
34
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
HyperTransport I/O Hub Node ID (SbNode)—Bits 10–8. Defines the Node ID of the node that
owns the HyperTransport link to the HyperTransport I/O hub.
Lock Controller Node ID (LkNode)—Bits 14–12. Defines the Node ID of the node that contains the
Lock Controller. The lock controller node (LkNode) must be the same as the HyperTransport
I/O hub node (SbNode).
CPU Count (CPUCnt)—Bits 19–16. Defines the number of CPUs in the system.
0000b = 1 CPU
1111b = 16 CPUs
3.3.8
Unit ID Register
The Unit ID register specifies the Unit ID of each functional unit on the processor. Under normal
operation, there is no need for software to change the default Unit ID assignments. This register also
contains a pointer to the HyperTransport I/O hub link.
Function 0: Offset 64h
Function
6
5
4
Bits
Mnemonic
R/W
Reset
31–10
reserved
R
0
9–8
SbLink
HyperTransport I/O Hub Link ID
R/W
00b
7–6
HbUnit
Host Bridge Unit ID
R/W
11b
5–4
McUnit
Memory Controller Unit ID
R/W
10b
3–2
C1Unit
CPU1 Unit ID
R/W
01b
1–0
C0Unit
CPU0 Unit ID
R/W
00b
3
2
1
0
C0Unit
7
C1Unit
reserved
8
McUnit
10 9
SbLink
31
HbUnit
Unit ID Register
Field Descriptions
CPU0 Unit ID (C0Unit)—Bits 1–0. Defines the Unit ID of CPU0.
CPU1 Unit ID (C1Unit)—Bits 3–2. Defines the Unit ID of CPU1 (if present).
Memory Controller Unit ID (McUnit)—Bits 5–4. Defines the Unit ID of the memory controller.
Host Bridge Unit ID (HbUnit)—Bits 7–6. Defines the Unit ID of the host bridge. The host bridge is
used to bridge between coherent and noncoherent HyperTransport technology domains.
HyperTransport I/O Hub Link ID (SbLink)—Bits 9–8. Defines the link to which the
HyperTransport I/O hub is connected. It is only used by the node which owns the
HyperTransport I/O hub, as indicated in the Node ID register.
00b = LDT0 (default)
01b = LDT1
Chapter 3
Memory System Configuration
35
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
10b
11b
3.3.9
26094
Rev. 3.06
September 2003
= LDT2
= Undefined
HyperTransport™ Transaction Control Register
The HyperTransport Transaction Control register configures the high-level HyperTransport protocols.
It can be used to optimize system performance or change HyperTransport technology transaction
protocols.
Bits
Mnemonic
Function
31
DisIsocWrMedPri
Disable medium priority isochronous writes
R/W
0
30
DisIsocWrHiPri
Disable high priority isochronous writes
R/W
0
29
DisWrLoPri
Disable low priority writes
R/W
0
28
EnCpuRdHiPri
Enable high priority CPU reads
R/W
0
27–26
HiPriBypCnt
High-priority bypass count
R/W
11b
25–24
MedPriBypCnt
Medium-priority bypass count
R/W
11b
23
reserved
R
0
22–21
DsNpReqLmt
Downstream non-posted request limit
R/W
00b
20
SeqIdSrcNodeEn
Sequence ID source node enable
R/W
0
19
ApicExtSpur
APIC extended spurious vector enable
R/W
0
18
ApicExtId
APIC extended ID enable
R/W
0
17
ApicExtBrdCst
APIC extended broadcast enable
R/W
0
16
LintEn
Local interrupt conversion enable
R/W
0
15
LimitCldtCfg
Limit coherent HyperTransport configuration space
range
R/W
0
BufRelPri
Buffer release priority select
R/W
00b
12
ChgIsocToOrd
Change ISOC to Ordered
R/W
0
11
RspPassPW
Response PassPW
R/W
0
10
DisFillP
Disable fill probe
R/W
0
9
DisRmtPMemC
Disable remote probe memory cancel
R/W
0
8
DisPMemC
Disable probe memory cancel
R/W
0
7
CPURdRspPassPW
CPU Read response PassPW
R/W
0
6
CPUReqPassPW
CPU request PassPW
R/W
0
5
Cpu1En
CPU1 enable
R/W
0
4
DisMTS
Disable memory controller target start
R/W
0
Memory System Configuration
1
0
DisRdBP
Reset
14–13
36
2
DisRdDwP
3
DisWrBP
4
DisWrDwP
CPUReqPassPW
R/W
5
DisMTS
6
Cpu1En
7
CPURdRspPassPW
DisFillP
DisRmtPMemC
RspPassPW
ChgIsocToOrd
BufRelPri
LimitCldtCfg
LintEn
ApicExtBrdCst
ApicExtId
ApicExtSpur
SeqIdSrcNodeEn
DsNpReqLmt
reserved
MedPriBypCnt
HiPriBypCnt
EnCpuRdHiPri
DisWrLoPri
8
DisIsocWrHiPri
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
DisPMemC
Function 0: Offset 68h
DisIsocWrMedPri
HyperTransport™ Transaction Control Register
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
Reset
3
DisWrDwP
Disable write doubleword probes
R/W
0
2
DisWrBP
Disable write byte probes
R/W
0
1
DisRdDwP
Disable read doubleword probe
R/W
0
0
DisRdBP
Disable read byte probe
R/W
0
Field Descriptions
Disable Read Byte Probe (DisRdBP)—Bit 0. This bit determines if probes are issued for CPUgenerated RdSized byte (must be 0 for multiprocessor system; recommended to be 1 for
uniprocessor system)
0 = Probes issued
1 = Probes not issued
Disable Read Doubleword Probe (DisRdDwP)—Bit 1. This bit determines if probes are issued for
CPU-generated RdSized Doubleword (must be 0 for multiprocessor system; recommended to
be 1 for uniprocessor system)
0 = Probes issued
1 = Probes not issued
Disable Write Byte Probes (DisWrBP)—Bit 2. This bit determines if probes are issued for CPUgenerated WrSized byte (must be 0 for multiprocessor system; recommended to be 1 for
uniprocessor system)
0 = Probes issued
1 = Probes not issued
Disable Write Doubleword Probes (DisWrDwP)—Bit 3. This bit determines if probes are issued
for CPU-generated WrSized Doubleword (must be 0 for multiprocessor system;
recommended to be 1 for uniprocessor system)
0 = Probes issued
1 = Probes not issued
Disable Memory Controller Target Start (DisMTS)—Bit 4. Disables use of TgtStart. TgtStart is
used to improve scheduling of back-to-back ordered transactions by indicating when the first
transaction is received and ordered at the memory controller.
0 = TgtStart packets are generated
1 = TgtStart packets are not generated
CPU1 Enable (Cpu1En)—Bit 5. Enables a second CPU (if present on the node). This bit should
only be set for nodes which include two CPUs.
0 = Second CPU disabled or not present
1 = Second CPU enabled
CPU Request PassPW (CPUReqPassPW)—Bit 6. Allows CPU-generated requests to pass posted
writes.
Chapter 3
Memory System Configuration
37
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
0
1
26094
Rev. 3.06
September 2003
= CPU requests do not pass posted writes
= CPU requests pass posted writes
CPU Read Response PassPW (CPURdRspPassPW)—Bit 7. Allows RdResponses to CPUgenerated reads to pass posted writes.
0 = CPU responses do not pass posted writes
1 = CPU responses pass posted writes
Disable Probe Memory Cancel (DisPMemC)—Bit 8. Disables generation of the MemCancel
command when a probe hits a dirty cache block. MemCancels are used to attempt to save
DRAM and/or HyperTransport technology bandwidth associated with the transfer of stale
DRAM data.
0 = Probes may generate MemCancels
1 = Probes may not generate MemCancels
Disable Remote Probe Memory Cancel (DisRmtPMemC)—Bit 9. Disables generation of the
MemCancel command when a probe hits a dirty cache block unless the probed cache is on the
same node as the memory controller. MemCancels are used to attempt to save DRAM and/or
HyperTransport technology bandwidth associated with the transfer of stale DRAM data.
0 = Probes hitting dirty blocks generate memory cancel packets, regardless of the location of
the probed cache
1 = Only probed caches on the same node as the target memory controller generate memory
cancel packets
Disable Fill Probe (DisFillP)—Bit 10. Disables probes for CPU-generated fills (must be 0 for
multiprocessor system; recommended to be 1 for uniprocessor system).
0 = Probes issued for cache fills
1 = Probes not issued for cache fills
Response PassPW (RspPassPW)—Bit 11. Causes the host bridge to set the PassPW bit in all
downstream responses.
This technically breaks the PCI ordering rules but it is not expected to be an issue in the
downstream direction. Setting this bit improves the latency of upstream requests by allowing
the downstream responses to pass posted writes.
0 = Downstream response PassPW based on original request
1 = Downstream response PassPW set to 1
Change ISOC to Ordered (ChgIsocToOrd)—Bit 12. Causes bit 1 of RdSz/WrSz commands to be
treated as an ordered bit in coherent HyperTransport technology, instead of as an isochronous
bit. Setting this bit will disable prioritization of isochronous requests, but will enable tracking
of ordered commands across coherent HyperTransport links. This bit should only be set if the
performance of non-ordered peer-to-peer traffic across coherent HyperTransport links is
optimized, or to disable isochronous prioritization.
0 = Bit 1 of coherent HyperTransport technology sized command used for isochronous
prioritization
38
Memory System Configuration
Chapter 3
26094
Rev. 3.06
1
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
= Bit 1 of coherent HyperTransport technology sized command used for ordering
Buffer Release Priority Select (BufRelPri)—Bits 14–13. Selects the number of HyperTransport
technology doublewords sent with a buffer release pending before the buffer release is
inserted into the command/data stream of a busy link.
00b = 64 (default)
01b = 16
10b = 8
11b = 2
This should be set to a value of 10 (8 HyperTransport packets) to maximize HyperTransport
technology bandwidth.
Limit Coherent HyperTransport Configuration Space Range (LimitCldtCfg)—Bit 15. Limits the
extent of the coherent HyperTransport technology configuration space based on the number of
nodes in the system. When this bit is set, configuration accesses that normally map to coherent
HyperTransport space will be sent to noncoherent HyperTransport links instead, if these
accesses attempt to access a non-existent node, as specified through the node count in the
Node ID register. This bit should be set by BIOS once coherent HyperTransport fabric
initialization is complete. Failure to do so will result in PCI configuration accesses to nonexistent nodes being sent into the coherent HyperTransport routing fabric, causing the system
to hang.
0 = No coherent HyperTransport configuration space restrictions
1 = Limit coherent HyperTransport configuration space based on number of nodes
Local Interrupt Conversion Enable (LintEn)—Bit 16. Enables the conversion of broadcast ExtInt/
NMI HyperTransport technology interrupts to LINT0/1 before delivering to the local APIC.
This conversion only takes place if the local APIC is hardware enabled.
0 = ExtInt/NMI interrupts unaffected
1 = ExtInt/NMI broadcast interrupts converted to LINT0/1
APIC Extended Broadcast Enable (ApicExtBrdCst)—Bit 17. Enables extended APIC broadcast
functionality.
0 = APIC broadcast is 0Fh
1 = APIC broadcast is FFh
APIC Extended ID Enable (ApicExtId)—Bit 18. Enables extended APIC ID functionality.
0 = APIC ID is 4 bits
1 = APIC ID is 8 bits
APIC Extended Spurious Vector Enable (ApicExtSpur)—Bit 19. Enables extended APIC
spurious vector functionality.
0 = Lower 4 bits of spurious vector are read-only 1111b
1 = Lower 4 bits of spurious vector are writable
Chapter 3
Memory System Configuration
39
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Sequence ID Source Node Enable (SeqIdSrcNodeEn)—Bit 20. Causes the source node of requests
to be used in the SeqID field of downstream noncoherent HyperTransport technology request
packets. This may be desirable for some debug HyperTransport packet tracing applications, to
match downstream packets with their originating CPU. For normal operation, this bit should
be cleared, since this use of SeqID for source node determination may cause problems with
packet ordering.
Downstream non-posted request limit (DsNpReqLmt)—Bits 22–21. Sets the number of
downstream non-posted requests issued by CPU(s) which may be outstanding on the
noncoherent HyperTransport links attached to this node. Non-posted requests from CPU(s)
will be throttled such that they do not exceed the programmed value in this field.
00b = No limit
01b = Limited to 1
10b = Limited to 4
11b = Limited to 8
Medium-Priority Bypass Count (MedPriBypCnt)—Bits 25–24. The maximum number of times a
medium priority access can pass a low priority access before medium priority mode is
disabled for one access.
High-Priority Bypass Count (HiPriBypCnt)—Bits 27–26. The maximum number of times a high
priority access can pass a medium or low priority access before high priority mode is disabled
for one access.
Enable High Priority CPU Reads (EnCpuRdHiPri)—Bit 28. Enables CPU reads to be treated as
high priority. CPU reads are treated as medium priority by default. Setting EnCpuRdHiPri
changes their priority to high.
Disable Low Priority Writes (DisWrLoPri)—Bit 29. Disables non-isochronous writes from being
treated as low priority. Non-isochronous writes are treated as low priority by default. Setting
DisWrLoPri changes their priority to medium.
Disable High Priority Isochronous Writes (DisIsocWrHiPri)—Bit 30. Disables isochronous
writes from being treated as high priority. Isochronous writes are treated as high priority by
default. Setting DisIsocWrHiPri changes their priority to medium (if DisIsocWrMedPri is
clear) or low (if DisIsocWrMedPri is set).
Disable Medium Priority Isochronous Writes (DisIsocWrMedPri)—Bit 31. Disables isochronous
writes from being treated as medium priority. Isochronous writes are treated as high priority
by default. Setting DisIsocWrMedPri along with DisIsocWrHiPri changes their priority to
low.
3.3.10
HyperTransport™ Initialization Control Register
The HyperTransport Initialization Control register controls the routing and request generation that
follows device initialization. This register also contains a set of scratchpad read/write bits that are
cleared by different initialization conditions and that may be used to distinguish between various
40
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
types of initialization events. Note that the Request Disable and Routing Table Disable bits must not
be cleared until the routing tables have been initialized.
Mnemonic
31–7
reserved
6
InitDet
5
BiosRstDet
4
3–2
Function
R/W
Reset
R
0
INIT Detect
R/W
0
BIOS Reset Detect
R/W
0
ColdRstDet
Cold Reset Detect
R/W
0
DefLnk
Default Link
1
ReqDis
Request Disable
R/W
0
RouteTblDis
Routing Table Disable
R/W
3
2
1
ReqDis
4
0
RouteTblDis
5
DefLnk
reserved
Bits
6
ColdRstDet
7
InitDet
31
Function 0: Offset 6Ch
BiosRstDet
HyperTransport™ Initialization Control Register
R
1
Field Descriptions
Routing Table Disable (RouteTblDis)—Bit 0. This bit determines whether the routing tables are
used or whether default configuration access routing is used. It resets to 1, thereby routing
requests to the Configuration Special Register (CSR) block and routing responses based on
DefLnk. Once the routing tables have been set up this bit should be cleared.
0 = Packets are routed according to the routing tables
1 = Packets are routed according to the default link field (DefLnk)
Request Disable (ReqDis)—Bit 1. This bit determines if the node is allowed to generate request
packets. It resets to 0 for the BSP and to 1 for all other processors. This bit should be cleared
by system initialization firmware once the system has been initialized from the BSP. This bit
is set by hardware and cleared by software.
0 = Request packets may be generated
1 = Request packets may not be generated
Default Link (DefLnk)—Bits 3–2. This field is used after a reset and before the routing tables are
initialized. It is used by hardware to determine which link to route responses to and may be
used by software as part of system connectivity discovery.
The default link field is loaded each time an incoming request is received, with the link ID of
the link on which the packet arrived. It is read-only from software. It is only used to route
packets during initialization, when the Routing Table Disable bit is set to 1, and only one
outstanding request is active in the system at a time. During this interval, the value in the
Default Link field is used to route responses, so that responses are always sent out on the link
from which the last request was received. The register is updated as soon as a request is
Chapter 3
Memory System Configuration
41
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
received, so reads of this register from the fabric return the link ID of the link on which the
read command arrived. Reads from the CPU on the local node cause this field to get loaded
with 11b.
00b = LDT0
01b = LDT1
10b = LDT2
11b = CPU on same node
Cold Reset Detect (ColdRstDet)—Bit 4. This bit may be used to distinguish between a cold versus a
warm reset event by setting the bit to 1 before an initialization event is generated. This bit is
cleared by a cold reset but not by a warm reset.
BIOS Reset Detect (BiosRstDet)—Bit 5. This bit may be used to distinguish between a reset event
generated by the BIOS versus a reset event generated for any other reason by setting the bit to
1 before initiating a BIOS-generated reset event. This bit is cleared by a cold reset but not by a
warm reset.
INIT Detect (InitDet)—Bit 6. This bit may be used to distinguish between an INIT and a warm/cold
reset by setting the bit to 1 before an initialization event is generated. This bit is cleared by a
warm or cold reset but not by an INIT.
3.3.11
LDTi Capabilities Register
These registers define the capabilities of the LDT links. They also contain a capabilities pointer that
points to the next item in the linked capabilities list.
LDT0, LDT1, LDT2 Capabilities Registers
DevNum
Bits
Mnemonic
Function
31–29
CapType
28
27
WarmReset
18 17 16 15
DblEnded
ChainSide
reserved
HostHide
ActAsSlave
CapType
InbndEocErr
29 28 27 26 25 24 23 22
DropOnUnInit
31
Function 0: Offset 80h, A0h, C0h
8
7
CapPtr
0
CapID
R/W
Reset
Capability Type
R
001b
DropOnUnInit
Drop on Uninitialized Link
R
0
InbndEocErr
Inbound End-of-Chain Error
R
0
Act As Slave
26
ActAsSlave
25
reserved
24
HostHide
23
ChainSide
R
0
R
0
Host Hide
R
1
Chain Side
R
0
22–18
DevNum
Device Number
R
0
17
DblEnded
Double Ended
R
0
16
WarmReset
Warm Reset
R
1
42
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
15–8
CapPtr
Capability Pointer
R
7–0
CapID
Capability ID (Always Reads 08h)
R
Reset
08h
Field Descriptions
Capability ID (CapID)—Bits 7–0. A capability ID of 08h indicates HyperTransport technology
capability.
Capability Pointer (CapPtr)—Bits 15–8. Contains a pointer to the next capability in the list.
Depending on the specific HyperTransport link configuration physically present on the
processor, this will be either A0h (LDT1), C0h (LDT2), or 00h, if it is the last one.
Warm Reset (WarmReset)—Bit 16. Allows a reset sequence initiated by the HyperTransport
technology control register to be either warm or cold. Since resets initiated through the
HyperTransport control register are not supported, this field is read-only 1.
Double Ended (DblEnded)—Bit 17. Indicates that there is another bridge at the far end of the
noncoherent HyperTransport chain. Since double-hosted chains are not supported, this field is
read-only 0.
Device Number (DevNum)—Bits 22–18. The device number of configuration accesses which
Northbridge responds to when accessed from the noncoherent HyperTransport technology
chain. Since double-hosted chains are not supported, this field is read-only 0.
Chain Side (ChainSide)—Bit 23. This bit indicates which side of the host bridge is being accessed.
Since double-hosted chains are not supported, this bit is read-only 0.
Host Hide (HostHide)—Bit 24. This bit causes the memory system configuration space to be
inaccessible from the noncoherent HyperTransport technology chain. Since double-hosted
chains are not supported, this bit is read-only 1.
Act As Slave (ActAsSlave)—Bit 26. Since this function is not currently implemented, this bit is readonly 0.
Inbound End-of-Chain Error (InbndEocErr)—Bit 27. Since this function is not currently
implemented, this bit is read-only 0.
Drop on Uninitialized Link (DropOnUnInit)—Bit 28. Since this function is not currently
implemented, this bit is read-only 0.
Capability Type (CapType)—Bits 31–29. The code 001b designates a host interface
HyperTransport technology capability.
3.3.12
LDTi Link Control Registers
These registers control LDT link operation and log link errors.
Chapter 3
Memory System Configuration
43
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
Mnemonic
Function
31
DwFcOutEn
Doubleword Flow Control Out Enable
30–28
WidthOut
Link Width Out
27
DwFcInEn
Doubleword Flow Control In Enable
26–24
WidthIn
Link Width In
23
DwFcOut
Doubleword Flow Control Out
R/W
Reset
R
0
R/W
0
R
0
R/W
0
R
0
22–20
MaxWidthOut
Max Link Width Out
R
001b
19
DwFcIn
Doubleword Flow Control In
R
0
18–16
MaxWidthIn
Max Link Width In
15
reserved
14
ExtCTL
13
LdtStopTriEn
12
IsocEn
Isochronous Enable
11–10
reserved
R
001b
R
0
Extended CTL Time
R/W
0
HyperTransport Stop Tristate Enable
R/W
0
R
0
R
0
9–8
CrcErr
CRC_Error
R/W
0
7
TransOff
Transmitter Off
R/W
0
6
RcvOff
Receiver Off
R/W
0
5
InitComplete
Initialization Complete
R
0
4
LinkFail
Link Failure
R/W
0
3
CrcForceErr
CRC Force Error
R/W
0
2
CrcStartTest
CRC Start Test
1
CrcFloodEn
CRC Flood Enable
0
reserved
R
0
R/W
0
R
0
2
1
0
reserved
3
CrcFloodEn
4
CrcStartTest
RcvOff
5
CrcForceErr
6
LinkFail
7
InitComplete
8
CrcErr
reserved
IsocEn
LdtStopTriEn
ExtCTL
reserved
DwFcIn
September 2003
Function 0: Offset 84h, A4h, C4h
16 15 14 13 12 11 10 9
MaxWidthIn
20 19 18
MaxWidthOut
DwFcOut
24 23 22
WidthIn
DwFcInEn
28 27 26
WidthOut
DwFcOutEn
31 30
Rev. 3.06
TransOff
LDT0, LDT1, LDT2 Link Control Registers
26094
Field Descriptions
CRC Flood Enable (CrcFloodEn)—Bit 1. If cleared to 0, this bit prevents CRC errors both from
generating sync packets and causing the system to come down, and from setting the LinkFail
bit. However, CRC checking logic still runs on all lanes enabled by LinkWidthIn, and detected
errors still set the CRC Error bits to 1. Its reset value is 0.
0 = Do not generate sync packets on CRC error (default)
1 = Generate sync packets on CRC error
CRC Start Test (CrcStartTest)—Bit 2. Since this function is not currently implemented, this bit is
read-only 0.
44
Memory System Configuration
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
CRC Force Error (CrcForceErr)—Bit 3. When this bit is set, a bad CRC is generated on all
transmitting lanes as enabled by LinkWidthOut. The covered data is not affected. This bit is
readable and writable by software and resets to 0.
0 = Do not generate a bad CRC (default)
1 = Generate a bad CRC
Link Failure (LinkFail)—Bit 4. The LinkFail bit is set to 1 when a CRC error or a sync packet is
detected after link initialization or if the link is not connected. See “Machine Check
Architecture (MCA) Registers” on page 92 for more details. This bit is maintained through a
warm reset and is cleared to 0 on a cold reset. LinkFail is set to 1 by hardware in the event of a
link error that results in a sync flood. It can be cleared by software by writing a 1.
0 = No link failure detected
1 = Link failure detected
Initialization Complete (InitComplete)—Bit 5. This read-only bit is reset to 0, and set to 1 by
hardware when low-level link initialization is successfully complete. If there is no device on
the other end of the link, or if that device is unable to properly perform the low-level link
initialization protocol, the bit is not set to 1. Software must not attempt to generate any
packets across the link until this bit is set to 1. CRC checking in the hardware does not begin
until initialization is complete.
0 = Initialization not complete
1 = Initialization is complete
Receiver Off (RcvOff)—Bit 6. This bit disables the receiver from accepting any further packets. This
bit resets to 0, and is set by software writing a 1 to the bit. This bit cannot be cleared by
software—a write of 0 to this bit position has no effect. If the HyperTransport link is not
connected, then this bit is set by hardware.
0 = Receiver on
1 = Receiver off
Transmitter Off (TransOff)—Bit 7. This bit provides a mechanism to shut off a link transmitter for
power savings or EMI reduction. When set to 1, no output signals toggle on the link. This bit
resets to 0, and is set by software writing a 1 to the bit. This bit cannot be cleared by
software—a write of 0 to this bit position has no effect. If the HyperTransport link is not
connected, then this bit is set by hardware.
0 = Transmitter on
1 = Transmitter off
CRC_Error (CrcErr)—Bits 9–8. These bits are set to 1 by hardware when a CRC error is detected
on an incoming link. Errors are detected and reported on a per byte lane basis where bit 8
corresponds to the least significant byte lane. Two bits are required to cover the maximum
HyperTransport link width of 16 bits. Error bits for unimplemented (as specified by
MaxWidthIn) or unused byte lanes return 0 when read.
These bits are maintained through a warm reset and is cleared to 0 on a cold reset. They are
readable from software and are individually cleared to 0 by software by writing a 1.
Chapter 3
Memory System Configuration
45
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
00b = No error
Bit [0] =1 Error on Byte Lane 0
Bit [1] =1 Error on Byte Lane 1
Isochronous Enable (IsocEn)—Bit 12. Since this function is not currently implemented, this bit is
read-only 0.
HyperTransport Stop Tristate Enable (LdtStopTriEn)—Bit 13. This bit controls whether the
transmitter tristates the link during the disconnected state of an LDTSTOP_L sequence. When
the bit is set the transmitter tristates the link. When the bit is clear, it continues to drive the
link. The bit is cleared by a cold reset and its value is maintained through a warm reset.
0 = Driven during disconnect of an LDTSTOP_L
1 = Tristated during disconnect of an LDTSTOP_L
Extended CTL Time (ExtCTL)—Bit 14. When this bit is set, CTL will be asserted for 50µs instead
of 16 bit times during the link initialization sequence. This bit should be set if extended CTL
is required by the device at the other end of the HyperTransport link, as indicated by the
Extended CTL Required bit in the Feature Capability register. The bit is cleared by cold reset
and its value is maintained through warm reset.
Max Link Width In (MaxWidthIn)—Bits 18–16. This field contains three bits that indicate the
physical width of the incoming side of the HyperTransport link implemented by this device.
For the AMD Athlon™ 64 and AMD Opteron™ processors, this field is set to 000 or 001 to
indicate an 8-bit or 16-bit link.
Doubleword Flow Control In (DwFcIn)—Bit 19. This bit is read-only 0 to indicate that this link
does not support doubleword flow control.
Max Link Width Out (MaxWidthOut)—Bits 22–20. This field contains three bits that indicate the
physical width of the outgoing side of the HyperTransport link implemented by this device. It
is read-only. For the AMD Athlon™ 64 and AMD Opteron™ processors, this field is set to
000 or 001 to indicate an 8-bit or 16-bit link.
Doubleword Flow Control Out (DwFcOut)—Bit 23. This bit is read-only 0 to indicate that this link
does not support doubleword flow control.
Link Width In (WidthIn)—Bits 26–24. This field is similar to the WidthOut field, except that it
controls the utilized width of the incoming side of the links implemented by this device.
The hardware will only allow values to be programmed into this field which are consistent
with the width capabilities of the link as specified by MaxWidthIn. Attempts to write values
inconsistent with the capabilities will result in this field not being updated.
Doubleword Flow Control In Enable (DwFcInEn)—Bit 27. Since this function is not currently
implemented, this bit is read-only 0.
Link Width Out (WidthOut)—Bits 30–28. This field controls the utilized width (which may not
exceed the physical width) of the outgoing side of the HyperTransport link. Software can read/
write this field, and its value is maintained through a warm reset. After cold reset, this field is
46
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
initialized by hardware based on the results of the link width negotiation sequence described
in the HyperTransport technology specification. Based on sizing the devices at both ends of
the link, software may then write a different value into the register. The chain must pass
through warm reset or a LDTSTOP_L disconnect sequence for the new width values to be
reflected on the link.
The WidthOut CSR in the link transmitter must match the WidthIn CSR in the link receiver of
the device on the other side of the link. The LinkWidthIn and LinkWidthOut registers within
the same device are not required to have matching values.
The hardware will only allow values to be programmed into this field which are consistent
with the width capabilities of the link as specified by MaxWidthOut. Attempts to write values
inconsistent with the capabilities will result in this field not being updated
000b = 8-bit
001b = 16-bit
010b = reserved
011b = 32-bit
100b = 2-bit
101b = 4-bit
110b = reserved
111b = Link physically not connected
Doubleword Flow Control Out Enable (DwFcOutEn)—Bit 31. Since this function is not currently
implemented, this bit is read-only 0.
3.3.13
LDTi Frequency/Revision Registers
These registers control the link frequency, specify the link frequency capabilities, and describe the
HyperTransport technology specification level to which the specific link conforms.
LDT0, LDT1, LDT2 Frequency/Revision Registers
31
16 15
LnkFreqCap
Function 0: Offset 88h, A8h, C8h
12 11
Error
Bits
Mnemonic
Function
31–16
LnkFreqCap
Link Frequency Capability
8
Freq
7
5
4
MajRev
R/W
MinRev
Reset
R
15–12
Error
Error
R
0
11–8
Freq
Link Frequency
R/W
0
7–5
MajRev
Major Revision
R
001b
4–0
MinRev
Minor Revision
R
00010b
Chapter 3
Memory System Configuration
0
47
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Minor Revision (MinRev)—Bits 4–0. Contains the minor revision of the HyperTransport
technology specification to which the device conforms.
Major Revision (MajRev)—Bits 7–5. Contains the major revision of the HyperTransport technology
specification level to which the device conforms.
Link Frequency (Freq)—Bits 11–8. Specifies the maximum operating frequency of the link's
transmitter clock. The encoding of this field is shown below.
The LinkFreq is cleared by cold reset. Software can write a nonzero value to this register, and
that value takes effect on the next warm reset or LDTSTOP_L disconnect sequence.
The hardware will only allow values to be programmed into this field which are consistent
with the frequency capabilities of the link as specified by LinkFreqCap. Attempts to write
values inconsistent with the capabilities will result in this field not being updated.
It is possible to program this field for a higher frequency than the maximum allowed by the
processor. Refer to the processor data sheet for the maximum operating frequency allowed for
a given processor implementation.
0000b = 200 MHz
0001b = reserved
0010b = 400 MHz
0011b = reserved
0100b = 600 MHz
0101b = 800 MHz
0110b = 1000 MHz
0111b = reserved
1000–1110b = reserved
1111b = 100 MHz
If a link is not connected, then this field should not be written and should be left in its default state.
Error (Error)—Bits 15–12. This function is not currently implemented.
Link Frequency Capability (LnkFreqCap)—Bits 31–16. Indicates the clock frequency capabilities
of the link. Each bit corresponds to one of the 16 possible link frequency encodings.
While the frequency capabilities of different AMD Athlon™ 64 and AMD Opteron™
processors may vary, the 200-MHz and 100-MHz frequencies are always supported.
3.3.14
LDTi Feature Capability Registers
These registers identify features that are supported by the specific link.
48
Memory System Configuration
Chapter 3
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Function 0: Offset 8Ch, ACh, CCh
9
reserved
Bits
Mnemonic
31–9
reserved
8
ExtRegSet
7–4
reserved
3
ExtCTLRqd
2
1
0
Function
Extended Register Set
8
7
4
reserved
R/W
Reset
R
0
R
0
R
0
Extended CTL Required
R
0
CrcTstMode
CRC Test Mode
R
0
LdtStopMode
HyperTransport Stop Mode
R
1
IsocMode
Isochronous Flow Control Mode
R
0
3
2
1
0
IsocMode
31
LdtStopMode
LDT0, LDT1, LDT2 Feature Capability Registers
CrcTstMode
September 2003
ExtCTLRqd
Rev. 3.06
ExtRegSet
26094
Field Descriptions
Isochronous Flow Control Mode (IsocMode)—Bit 0. This is a read-only bit that reflects whether
isochronous flow control operation is supported. This bit is set to 0, indicating no support for
this feature.
HyperTransport Stop Mode (LdtStopMode)—Bit 1. This is a read-only bit that indicates whether
the LDTSTOP_L protocol is supported. This bit is set to 1, indicating support for this feature.
CRC Test Mode (CrcTstMode)—Bit 2. This is a read-only bit that indicates whether CRC test mode
is supported. This bit is set to 0, indicating no support for this feature.
Extended CTL Required (ExtCTLRqd)—Bit 3. This is a read-only bit that indicates whether
extended CTL is required. This bit is set to 0, indicating that extended CTL is not required.
Extended Register Set (ExtRegSet)—Bit 8. This is a read-only bit that indicates whether the
Enumeration Scratchpad, Error Handling and Memory Base/Limit Upper registers are
supported. This bit is set to 0, indicating no support for this feature.
3.3.15
LDTi Buffer Count Registers
These registers specify the number of command and data buffers for each virtual channel available for
use by the transmitter at the other end of the specific link. See “XBAR Flow Control Buffers” on
page 113 for more information on command and data buffers.
Note: The reset values for each of the LDTn Buffer Count registers depend on the link connection
type (coherent HyperTransport or noncoherent HyperTransport technology). Because
hardware attempts to choose optimal settings, this register should not, in general, need to be
changed.
Chapter 3
Memory System Configuration
49
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
27 26
reserved
Bits
Mnemonic
31–27
reserved
26–24
RspD
23
reserved
22–20
PReqD
19
reserved
18–16
ReqD
24 23 22
20 19 18
reserved
31
reserved
LDT0, LDT1, LDT2 Buffer Count Registers
RspD
PReqD
Rev. 3.06
September 2003
Function 0: Offset 90h, B0h, D0h
16 15
ReqD
26094
12 11
Probe
8
Rsp
7
4
3
PReq
0
Req
Coherent
Noncoherent
HyperTransport HyperTransport
Reset
Reset
Function
R/W
R
0
0
Response Data Buffer Count
R/W
100b
010b
R
0
0
Posted Request Data Buffer Count
R/W
001b
101b
R
0
0
Request Data Buffer Count
R/W
011b
001b
15–12
Probe
Probe Buffer Count
R/W
5h
0h
11–8
Rsp
Response Buffer Count
R/W
6h
4h
7–4
PReq
Posted Request Buffer Count
R/W
1h
5h
3–0
Req
Request Buffer Count
R/W
3h
6h
Field Descriptions
Request Buffer Count (Req)—Bits 3–0. Defines the number of request command buffers available
for use by the transmitter at the other end of the link. See the register layout for the default
reset settings for coherent and noncoherent HyperTransport technology links.
Posted Request Buffer Count (PReq)—Bits 7–4. Defines the number of Posted request command
buffers available for use by the transmitter at the other end of the link. See the register layout
for the default reset settings for coherent and noncoherent HyperTransport technology links.
Response Buffer Count (Rsp)—Bits 11–8. Defines the number of response buffers available for use
by the transmitter at the other end of the link. See the register layout for the default reset
settings for coherent and noncoherent HyperTransport technology links.
Probe Buffer Count (Probe)—Bits 15–12. Defines the number of probe buffers available for use by
the transmitter at the other end of the link. This field must be 0 for a noncoherent
HyperTransport link. See the register layout for the default reset settings for coherent and
noncoherent HyperTransport technology links.
Request Data Buffer Count (ReqD)—Bits 18–16. Defines the number of request data buffers
available for use by the transmitter at the other end of the link. See the register layout for the
default reset settings for coherent and noncoherent HyperTransport technology links.
Posted Request Data Buffer Count (PReqD)—Bits 22–20. Defines the number of posted request
data buffers available for use by the transmitter at the other end of the link. See the register
50
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
layout for the default reset settings for coherent and noncoherent HyperTransport technology
links.
Response Data Buffer Count (RspD)—Bits 26–24. Defines the number of response data buffers
available for use by the transmitter at the other end of the link. See the register layout for the
default reset settings for coherent and noncoherent HyperTransport technology links.
3.3.16
LDTi Bus Number Registers
If the specific link (LDT0, LDT1, or LDT2) is noncoherent HyperTransport technology, this register
specifies the bus numbers downstream (behind) the host bridge. If the link is coherent HyperTransport
technology, this register has no meaning.
LDT0, LDT1, LDT2 Bus Number Registers
31
24 23
reserved
Bits
Mnemonic
31–24
reserved
23–16
SubBusNum
Function 0: Offset 94h, B4h, D4h
16 15
SubBusNum
8
7
SecBusNum
0
PriBusNum
Function
R/W
Reset
R
0
Subordinate Bus Number
R/W
0
R/W
0
R
0
15–8
SecBusNum
Secondary Bus Number
7–0
PriBusNum
Primary Bus Number
Field Descriptions
Primary Bus Number (PriBusNum)—Bits 7–0. Defines the primary bus number. Because the
primary bus is the coherent HyperTransport technology fabric, this field always reads 0.
Secondary Bus Number (SecBusNum)—Bits 15–8. Defines the secondary bus number.
The Secondary Bus Number register is used to record the bus number of the bus segment to
which the secondary interface of the host bridge is connected. Configuration software
programs the value in this register. If this link contains the HyperTransport I/O hub, the
secondary bus number must be programmed to 0.
Subordinate Bus Number (SubBusNum)—Bits 23–16. Defines the subordinate bus number.
The Subordinate Bus Number register is used to record the bus number of the highest
numbered bus segment that is behind (or subordinate to) the host bridge. Configuration
software programs the value in this register.
3.3.17
LDTi Type Registers
These registers designate the type of HyperTransport link attached to the specific link. The bits are set
by hardware after link initialization.
Chapter 3
Memory System Configuration
51
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Rev. 3.06
September 2003
Bits
Mnemonic
Function
31–5
reserved
4
LinkConPend
3
UniP-cLDT
UniP-cLDT
R
2
NC
Non Coherent
R
1
InitComplete
Initialization Complete
R
0
LinkCon
Link Connected
R
Link Connect Pending
3
R/W
Reset
R
0
2
1
0
LinkCon
reserved
4
InitComplete
5
NC
31
UniP-cLDT
Function 0: Offset 98h, B8h, D8h
LinkConPend
LDT0, LDT1, LDT2 Type Registers
26094
R
Field Descriptions
Link Connected (LinkCon)—Bit 0. Indicates that the link is connected. It is valid once the
LinkConPend bit is clear.
0 = Not connected
1 = Connected
Initialization Complete (InitComplete)—Bit 1. Set to 1 to indicate that the initialization of the link
has completed. (It is a duplicate of Bit 5 in HyperTransport Link Control). The NC and UniPcLDT bits are invalid until link initialization is complete.
0 = Initialization not complete
1 = Initialization is complete.
Non Coherent (NC)—Bit 2. Defines the link type (coherent versus noncoherent).
0 = Coherent HyperTransport technology
1 = Noncoherent HyperTransport technology
UniP-cLDT (UniP-cLDT)—Bit 3. Further qualifies the NC link type bit. A 1 indicates that this link
is a uniprocessor coherent HyperTransport link connected to an external Northbridge.
0 = Normal coherent HyperTransport or noncoherent HyperTransport link
1 = Uniprocessor coherent HyperTransport link to external Northbridge
Link Connect Pending (LinkConPend)—Bit 4. Qualifies the LinkCon bit and is set to 1 when
hardware is attempting to determine whether a link is connected or not.
0 = Link connection determination complete
1 = Link connection still being determined
52
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.4
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Function 1—Address Map
The address map defines the address spaces assigned to DRAM, memory-mapped I/O, PCI I/O, and
configuration accesses and also specifies destination information for each to facilitate routing of each
access to the appropriate target. Table 3 lists each Function 1 configuration register.
Table 3. Function 1 Configuration Registers
Offset
Register Name
00h
Device ID
1
Reset
Access
Description
Vendor ID (AMD)
1101_1022h
RO
page 29
Command
0000_0000h
RO
04h
Status
08h
Base
Class
Code
Subclass
Code
Program
Interface
Revision
ID
0600_0000h
RO
page 55
0Ch
BIST
Header
Type
Latency
Timer
Cache
Line Size
0080_0000
RO
page 56
10h
Base Address 0
0000_0000h
RO
14h
Base Address 1
0000_0000h
RO
18h
Base Address 2
0000_0000h
RO
1Ch
Base Address 3
0000_0000h
RO
20h
Base Address 4
0000_0000h
RO
24h
Base Address 5
0000_0000h
RO
28h
Card Bus CIS Pointer
0000_0000h
RO
2Ch
Sub-System ID
0000_0000h
RO
30h
ROM Base Address
0000_0000h
RO
34h
Capabilities
0000_0000h
RO
38h
reserved
0000_0000h
RO
3Ch
Max
Latency
0000_0000h
RO
40h
DRAM Base 0
RW
page 57
44h
DRAM Limit 0
RW
page 58
48h
DRAM Base 1
RW
page 57
4Ch
DRAM Limit 1
RW
page 58
50h
DRAM Base 2
RW
page 57
54h
DRAM Limit 2
RW
page 58
58h
DRAM Base 3
RW
page 57
5Ch
DRAM Limit 3
RW
page 58
Min GNT
Sub-System Vendor ID
Int Pin
Int Line
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
Chapter 3
Memory System Configuration
53
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 3. Function 1 Configuration Registers (Continued)
Offset
Register Name
60h
Reset
Access
Description
DRAM Base 4
RW
page 57
64h
DRAM Limit 4
RW
page 58
68h
DRAM Base 5
RW
page 57
6Ch
DRAM Limit 5
RW
page 58
70h
DRAM Base 6
RW
page 57
74h
DRAM Limit 6
RW
page 58
78h
DRAM Base 7
RW
page 57
7Ch
DRAM Limit 7
RW
page 58
80h
Memory-Mapped I/O Base 0
RW
page 60
84h
Memory-Mapped I/O Limit 0
RW
page 61
88h
Memory-Mapped I/O Base 1
RW
page 60
8Ch
Memory-Mapped I/O Limit 1
RW
page 61
90h
Memory-Mapped I/O Base 2
RW
page 60
94h
Memory-Mapped I/O Limit 2
RW
page 61
98h
Memory-Mapped I/O Base 3
RW
page 60
9Ch
Memory-Mapped I/O Limit 3
RW
page 61
A0h
Memory-Mapped I/O Base 4
RW
page 60
A4h
Memory-Mapped I/O Limit 4
RW
page 61
A8h
Memory-Mapped I/O Base 5
RW
page 60
ACh
Memory-Mapped I/O Limit 5
RW
page 61
B0h
Memory-Mapped I/O Base 6
RW
page 60
B4h
Memory-Mapped I/O Limit 6
RW
page 61
B8h
Memory-Mapped I/O Base 7
RW
page 60
BCh
Memory-Mapped I/O Limit 7
RW
page 61
C0h
PCI I/O Base 0
RW
page 62
C4h
PCI I/O Limit 0
RW
page 63
C8h
PCI I/O Base 1
RW
page 62
CCh
PCI I/O Limit 1
RW
page 63
D0h
PCI I/O Base 2
RW
page 62
D4h
PCI I/O Limit 2
RW
page 63
D8h
PCI I/O Base 3
RW
page 62
DCh
PCI I/O Limit 3
RW
page 63
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
54
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 3. Function 1 Configuration Registers (Continued)
Offset
Register Name
E0h
Reset
Access
Description
Configuration Base and Limit 0
RW
page 64
E4h
Configuration Base and Limit 1
RW
page 64
E8h
Configuration Base and Limit 2
RW
page 64
ECh
Configuration Base and Limit 3
RW
page 64
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
3.4.1
Device/Vendor ID Register
This register specifies the device and vendor IDs for the Function 1 registers and is part of the
standard PCI configuration header.
Device/Vendor ID Register
Function 1: Offset 00h
31
16 15
0
Device ID
Vendor ID
Bits
Mnemonic
Function
R/W
Reset
31–16
DevID
Device ID
R
1101h
15–0
VenID
Vendor ID
R
1022h
Field Descriptions
Vendor ID (VenID)—Bits 15–0. This read-only value is defined as 1022h for AMD.
Device ID (DevID)—Bits 31–16. This read-only value is defined as 1101h for the HyperTransport
technology configuration function.
3.4.2
Class Code/Revision ID Register
This register specifies the class code and revision for the Function 1 registers and is part of the
standard PCI configuration header.
Class Code/Revision ID Register
31
24 23
Base Class Code
Chapter 3
Function 1: Offset 08h
16 15
Subclass Code
8
Programming Interface
Memory System Configuration
7
0
Revision ID
55
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
Bits
Mnemonic
Function
R/W
31–24
BCC
Base Class Code
R
06h
23–16
SCC
Subclass Code
R
00h
15–8
PI
Programming Interface
R
00h
7–0
RevID
Revision ID
R
00h
September 2003
Reset
Field Descriptions
Revision ID (RevID)—Bits 7–0.
Programming Interface (PI)—Bits 15–8. This read-only value is defined as 00h.
Sub Class Code (SCC)—Bits 23–16. This read-only value is defined as 00h.
Base Class Code (BCC)—Bits 31–24. This read-only value is defined as 06h for a host bridge
device.
3.4.3
Header Type Register
This register specifies the header type for the Function 1 registers and is part of the standard PCI
configuration header.
Header Type Register
31
Function 1: Offset 0Ch
24 23
16 15
BIST
HType
Bits
Mnemonic
Function
31–24
BIST
8
7
LatTimer
0
CLS
R/W
Reset
BIST
R
00h
23–16
HType
Header Type
R
80h
15–8
LatTimer
Latency Timer
R
00h
7–0
CLS
Cache Line Size
R
00h
Field Descriptions
CacheLineSize (CLS)—Bits 7–0. This read-only value is defined as 00h.
LatencyTimer (LatTimer)—Bits 15–8. This read-only value is defined as 00h.
HeaderType (HType)—Bits 23–16. This read-only value is defined as 80h to indicate that multiple
functions are present in the configuration header and that the header layout corresponds to a
device header as opposed to a bridge header. See “Register Differences in Revisions of the
AMD Athlon™ 64 and AMD Opteron™ processors” on page 19 for revision information
about this field.
56
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
BIST—Bits 31–24. This read-only value is defined as 00h.
3.4.4
DRAM Address Map
These registers define sections of the memory address map for which accesses should be routed to
DRAM. DRAM regions must not overlap each other. For addresses within the specified range of a
base/limit pair, requests are routed to the memory controller on the node specified by the destination
Node ID.
System addresses are considered to be within the defined range if they are greater than or equal to the
base and less than or equal to the limit. For the purposes of this comparison, the lower unspecified bits
of the base are assumed to be 0s and the lower unspecified bits of the limit are assumed to be 1s.
An address that maps to both DRAM and memory-mapped I/O will be routed to MMIO.
Programming of the DRAM address maps must be consistent with the Top Of Memory and Memory
Type Range registers (see Chapter 12, “Processor Configuration Registers”). Accesses from the CPU
can only access the DRAM address maps if the corresponding CPU memory type is DRAM. For
accesses from I/O devices, the lookup is based on address only.
Each base/limit set of DRAM address maps is associated with a particular Node ID (see “Node ID
Register” on page 34). The DRAM Base/Limit 0 registers specify the DRAM attached to node 0.
Similarly, DRAM on nodes 1–7 is described by base/limit registers 1–7. Note that the destination
NodeId field must still be written to ensure correct operation.
When node interleaving is enabled, each node’s DRAM limit must be set to the Top Of Memory and
each node’s DRAM base must be set to 0. The node to which an address is routed to when nodes are
interleaved is defined by IntlvEn (Function 1, Offset 40h, 48h, etc.) and IntlvSel (Function 1, Offset
44h, 4Ch, etc.).
Address routed to the DRAM controller (InputAddr) is calculated from the system address (SysAddr)
in the following way:
DramAddr[39:0] = {SysAddr[39:24] - DRAMBase[39:24], SysAddr[23:0]},
InputAddr[35:0] = {DramAddr[35:12], DramAddr[11:0]} when node memory is not interleaved,
InputAddr[35:0] = {DramAddr[36:13], DramAddr[11:0]} when 2 nodes are interleaved,
InputAddr[35:0] = {DramAddr[37:14], DramAddr[11:0]} when 4 nodes are interleaved,
InputAddr[35:0] = {DramAddr[38:15], DramAddr[11:0]} when 8 nodes are interleaved.
DRAM Base i Registers
31
16 15
DRAMBasei
Chapter 3
Function 1: Offset 40h, 48h, 50h, 58h, 60h, 68h, 70h, 78h
11 10
reserved
Memory System Configuration
8
IntlvEn
7
2
reserved
1
0
RE
DRAM Base 0–7 Registers
WE
3.4.4.1
57
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
26094
Rev. 3.06
Mnemonic
Function
R/W
Reset
31–16
DRAMBasei
DRAM Base Address i Bits 39–24
R/W
X
15–11
reserved
R
0
R/W
X
R
0
10–8
IntlvEn
7–2
reserved
Interleave Enable
1
WE
Write Enable
R/W
0
0
RE
Read Enable
R/W
0
September 2003
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Read Enable (RE)—Bit 0. This bit enables reads to the defined address space.
0 = Disabled
1 = Enabled
Write Enable (WE)—Bit 1. This bit enables writes to the defined address space.
0 = Disabled
1 = Enabled
Interleave Enable (IntlvEn)—Bits 10–8. This field enables interleaving on a 4-Kbyte boundary
between memory on different nodes. The bits are encoded as follows:
000b = No interleave
001b = Interleave on A[12] (2 nodes)
010b = reserved
011b = Interleave on A[12] and A[13] (4 nodes)
100b = reserved
101b = reserved
110b = reserved
111b = Interleave on A[12] and A[13] and A[14] (8 nodes)
DRAM Base Address i Bits 39–24 (DRAMBasei)—Bits 31–16. This field defines the upper address
bits of a 40-bit address that defines the start of DRAM region i (where i = 0, 1, . . . 7).
3.4.4.2
DRAM Limit i Registers
DRAM Limit 0–7 Registers Function 1: Offset 44h, 4Ch, 54h, 5Ch, 64h, 6Ch, 74h, 7Ch
31
16 15
DRAMLimiti
11 10
reserved
8
7
IntlvSel
3
reserved
Bits
Mnemonic
Function
R/W
Reset
31–16
DRAMLimiti
DRAM Limit Address i (39–24)
R/W
X
15–11
reserved
R
0
58
Memory System Configuration
2
0
DstNode
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
Reset
10–8
IntlvSel
Interleave Select
R/W
X
7–3
reserved
R
0
2–0
DstNode
Destination Node ID
R/W
X
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Destination Node ID (DstNode)—Bits 2–0. This field specifies the node that a packet is sent to if it
is within the address range. This field must correspond with the number of the register pair.
The bits are encoded as follows:
000b = Node 0 (register set 0)
001b = Node 1 (register set 1)
...
111b = Node 7 (register set 7)
Interleave Select (IntlvSel)—Bits 10–8. This field specifies the values of address bits A[14:12] to
use with the Interleave Enable field (IntlvEn[2:0]) to determine which 4-Kbyte blocks are
routed to this region (See Figure 1).
IntlvSel[0] corresponds to A[12]
IntlvSel[1] corresponds to A[13]
IntlvSel[2] corresponds to A[14]
DRAM Limit Address i (39–24) (DRAMLimiti)—Bits 31–16. This field defines the upper address
bits of a 40-bit address that defines the end of DRAM region n (where i = 0,1, . . . 7).
Node 0
IntlvEn[2:0] =
IntlvSel[2:0] =
A[13:12] =
011
000
00
Node 1
IntlvEn[2:0] =
IntlvSel[2:0] =
A[13:12] =
011
001
01
Node 2
IntlvEn[2:0] =
IntlvSel[2:0] =
A[13:12] =
011
010
10
Node 3
IntlvEn[2:0] =
IntlvSel[2:0] =
A[13:12] =
011
011
11
Figure 1. Interleave Example (IntlvEn Relation to IntlvSel)
3.4.5
Memory-Mapped I/O Address Map Registers
These registers define sections of the memory address map for which accesses should be routed to
memory-mapped I/O. MMIO regions must not overlap each other. For addresses within the specified
Chapter 3
Memory System Configuration
59
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
range of a base/limit pair, requests are routed to the noncoherent HyperTransport link specified by the
destination Node ID and destination Link ID.
Addresses are considered to be within the defined range if they are greater than or equal to the base
and less than or equal to the limit. For the purposes of this comparison, the lower unspecified bits of
the base are assumed to be 0s and the lower unspecified bits of the limit are assumed to be 1s.
An address that maps to both DRAM and memory-mapped I/O is routed to MMIO.
Programming of the MMIO address maps must be consistent with the Top Of Memory and Memory
Type Range registers (see Chapter 12, “Processor Configuration Registers”). In particular, accesses
from the CPU can only hit in the MMIO address maps if the corresponding CPU memory type is of
type IO. For accesses from I/O devices, the lookup is based on address only.
3.4.5.1
Extended Configuration Space Access
Chipset devices may support PCI-defined extended configuration space through an MMIO range.
Typically, requests to the MMIO range for extended configuration space are required to use the nonposted channel. Therefore, the Non-Posted bit should be set for this MMIO range. Instructions used
to read extended configuration space must be of the following form:
mov EAX/AX/AL, <any_address_mode>;
Instructions used to write extended configuration space must be of the following form:
mov <any_address_mode>, EAX/AX/AL;
In addition, all such accesses are required not to cross any naturally aligned doubleword boundary.
Access to extended configuration registers that do not meet these requirements result in undefined
behavior.
Memory-Mapped I/O Base i Registers
8
MMIOBasei
Bits
7
4
reserved
Mnemonic
Function
R/W
Reset
31–8
MMIOBasei
Memory-Mapped I/O Base Address i (39–16)
R/W
X
7–4
reserved
R
0
3
Lock
Lock
R/W
X
2
CpuDis
CPU Disable
R/W
X
1
WE
Write Enable
R/W
0
0
RE
Read Enable
R/W
0
3
2
1
0
RE
31
WE
Function 1: Offset 80h, 88h, 90h, 98h,
A0h, A8h, B0h, B8h
CpuDis
Memory-Mapped I/O Base 0–7 Registers
Lock
3.4.5.2
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
60
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Field Descriptions
Read Enable (RE)—Bit 0. This bit enables reads to the defined address space.
0 = Disabled
1 = Enabled
Write Enable (WE)—Bit 1. This bit enables writes to the defined address space.
0 = Disabled
1 = Enabled
CPU Disable (CpuDis)—Bit 2. When set, this bit causes this MMIO range to apply to I/O requests
only, not to CPU requests.
Lock (Lock)—Bit 3. Setting this bit, along with either the WE or RE bits, makes the base/limit
registers read-only.
Memory-Mapped I/O Base Address i (39–16) (MMIOBasei)—Bit 31–8. This field defines the
upper address bits of a 40-bit address that defines the start of memory-mapped I/O region n
(where n = 0,1, . . . 7).
Memory-Mapped I/O Limit i Registers
8
MMIOLimiti
Bits
7
6
5
4
DstLink
31
Mnemonic
Function
R/W
Reset
31–8
MMIOLimiti
Memory-Mapped I/O Limit Address i
R/W
X
7
NP
Non-Posted
R/W
X
6
reserved
R
0
5–4
DstLink
Destination Link ID
R/W
X
3
reserved
R
0
2–0
DstNode
Destination Node ID
R/W
X
3
2
reserved
Function 1: Offset 84h, 8Ch, 94h, 9Ch,
A4h, ACh, B4h, BCh
reserved
Memory-Mapped I/O Limit i Registers
NP
3.4.5.3
0
DstNode
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Destination Node ID (DstNode)—Bits 2–0. This field specifies the node that a packet is sent to if it
is within the address range. The bits are encoded as follows:
000b = Node 0
001b = Node 1
...
Chapter 3
Memory System Configuration
61
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
111b = Node 7
Destination Link ID (DstLink)—Bits 5–4. This field specifies the HyperTransport link number to
send the packet to once the packet reaches the destination node.
00b = Link 0
01b = Link 1
10b = Link 2
11b = reserved
Non-Posted (NP)—Bit 7. When set to 1, this bit forces CPU writes to this memory-mapped I/O
region to be non-posted. This may be used to force writes to be non-posted for MMIO regions
which map to the legacy ISA/LPC bus, or in conjunction with the DsNpReqLmt field in the
HyperTransport Transaction Control register (page 36) in order to allow downstream CPU
requests to be counted and thereby limited to a specified number. This latter use of the NP bit
may be used to avoid loop deadlock in systems that implement a reflection region in an I/O
device that reflects downstream accesses back upstream. See the HyperTransport™ I/O Link
Specification summary of deadlock scenarios for more information.
0 = CPU writes may be posted
1 = CPU writes must be non-posted
Memory-Mapped I/O Limit Address i (39–16) (MMIOLimiti)—Bits 31–8. This field defines the
upper address bits of a 40-bit address that defines the end of memory-mapped I/O region n
(where n = 0,1, . . . 7).
3.4.6
PCI I/O Address Map Registers
These registers define sections of the PCI I/O address map for which accesses should be routed to
each noncoherent HyperTransport technology chain. PCI I/O regions must not overlap each other. For
PCI I/O addresses within the specified range of a base/limit pair, requests are routed to the
noncoherent HyperTransport link specified by the destination Node ID and destination Link ID.
Addresses are considered to be within the defined range if they are greater than or equal to the base
and less than or equal to the limit. For the purposes of this comparison, the lower unspecified bits of
the base are assumed to be 0s and the lower unspecified bits of the limit are assumed to be 1s.
PCI I/O accesses are generated from x86 IN/OUT instructions.
PCI I/O Base i Registers
reserved
62
12 11
PCIIOBasei
Memory System Configuration
6
reserved
5
4
IE
3
2
1
0
RE
25 24
WE
31
Function 1: Offset C0h, C8h, D0h, D8h
reserved
PCI I/O Base 0–3 Registers
VE
3.4.6.1
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
31–25
reserved
24–12
PCIIOBasei
11–6
reserved
5
Reset
R
0
PCI I/O Base Address i
R/W
X
R
0
IE
ISA Enable
R/W
X
VGA Enable
R/W
X
R
0
4
VE
3–2
reserved
1
WE
Write Enable
R/W
0
0
RE
Read Enable
R/W
0
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Read Enable (RE)—Bit 0. This bit enables reads to the defined address space.
0 = Disabled
1 = Enabled
Write Enable (WE)—Bit 1. This bit enables writes to the defined address space.
0 = Disabled
1 = Enabled
VGA Enable (VE)—Bit 4. Forces addresses in the first 64 Kbytes of PCI I/O space and where A[9:0]
is in the range 3B0–3BBh or 3C0–3DFh to match against this base/limit pair independent of
the base/limit addresses (see the PCI-to-PCI Bridge Architecture specification for a
description of this function). A WE or RE bit must also be set to enable this feature.
The MMIO 000A_0000h to 000B_FFFFh matching function normally associated with the
VGA enable bit in PCI is NOT included in the VE bit. To enable this behavior, an MMIO
register pair must be used.
ISA Enable (IE)—Bit 5. Blocks addresses in the first 64 Kbytes of PCI I/O space and in the last 768
bytes in each 1-Kbyte block from matching against this base/limit pair (see the PCI-to-PCI
Bridge Architecture specification for a description of this function). A WE or RE bit must also
be set to enable this feature.
PCI I/O Base Address i (PCIIOBasei)—Bits 24–12. This field defines the start of PCI I/O region n
(where n = 0, 1, 2,3).
PCI I/O Limit i Registers
31
Function 1: Offset C4h, CCh, D4h, DCh
25 24
reserved
Chapter 3
12 11
PCIIOLimiti
Memory System Configuration
6
reserved
5
4
3
2
reserved
PCI I/O Limit 0–3 Registers
DstLink
3.4.6.2
0
DstNode
63
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
Mnemonic
31–25
reserved
24–12
PCIIOLimiti
11–6
reserved
5–4
DstLink
3
reserved
2–0
DstNode
26094
Rev. 3.06
Function
R/W
R
0
PCI I/O Limit Address i
R/W
X
R
0
Destination Link ID
R/W
X
R
0
Destination Node ID
R/W
X
September 2003
Reset
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Destination Node ID (DstNode)—Bits 2–0. This field specifies the node to which a packet is sent if
it is within the address range. The bits are encoded as follows:
000b = Node 0
001b = Node 1
...
111b = Node 7
Destination Link ID (DstLink)—Bits 5–4. This field specifies the link to send the packet to once the
packet has reached the destination node.
00b = Link 0
01b = Link 1
10b = Link 2
11b = reserved
PCI I/O Limit Address i (PCI I/OLimiti)—Bits 24–12. This field defines the end of PCI I/O region
3.4.7
Configuration Map Registers
These registers define sections of the configuration bus number map for which configuration accesses
should be routed to each noncoherent HyperTransport technology chain. Configuration regions must
not overlap each other. For configuration accesses with bus numbers within the specified range of a
base/limit pair, requests are routed to the noncoherent HyperTransport link specified by the
destination Node ID and destination Link ID.
Bus numbers are considered to be within the defined range if they are greater than or equal to the base
and less than or equal to the limit.
In device number compare mode (see DevCmpEn bit), the base/limit fields are interpreted as device
number base/limit values for Bus 0 configuration accesses. This may be used to support
configurations where Bus 0 is split between two independent noncoherent HyperTransport
technology chains.
Configuration accesses to Bus 0, device 24–31, are assumed to be targeting coherent HyperTransport
technology devices and are routed to the corresponding coherent HyperTransport node irrespective of
64
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
the values in these registers (see Chapter 3, “Memory System Configuration”). Note that the range of
device numbers affected may be modified based on the number of nodes present (see
“HyperTransport™ Transaction Control Register” on page 36).
Bits
Mnemonic
7
6
4
3
2
1
DstNode
Function
R/W
Reset
31–24
BusNumLimiti
Bus Number Limit i
R/W
X
23–16
BusNumBasei
Bus Number Base i
R/W
X
15–10
reserved
R
0
9–8
DstLink
Destination Link ID
R/W
X
7
reserved
R
0
6–4
DstNode
Destination Node ID
R/W
X
3
reserved
R
0
2
DevCmpEn
Device Number Compare Enable
R/W
X
1
WE
Write Enable
R/W
0
0
RE
Read Enable
R/W
0
0
RE
reserved
8
WE
BusNumBasei
10 9
DevCmpEn
BusNumLimiti
16 15
reserved
24 23
reserved
31
Function 1: Offset E0h, E4h, E8h, ECh
DstLink
Configuration Base and Limit 0–3 Registers
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Read Enable (RE)—Bit 0. This bit enables reads to the defined address space.
0 = Disabled
1 = Enabled
Write Enable (WE)—Bit 1. This bit enables writes to the defined address space.
0 = Disabled
1 = Enabled
Device Number Compare Enable (DevCmpEn)—Bit 2. When this bit is set, this register defines a
device number range rather than a bus number range. To match, configuration cycles must be
to bus number 0 and have device numbers between BusNumBase and BusNumLimit. This is
used to enable multiple noncoherent HyperTransport chains to be configured as Bus 0.
Destination Node ID (DstNode)—Bits 6–4. This field specifies the node that a packet is sent to if it
is within the address range. The bits are encoded as follows:
000b = Node 0
001b = Node 1
...
111b = Node 7
Chapter 3
Memory System Configuration
65
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Destination Link ID (DstLink)—Bits 9–8. This field specifies the link to send the packet to once it
has reached the destination node and unit.
00b = Link 0
01b = Link 1
10b = Link 2
11b = reserved
Bus Number Base i (BusNumBasei)—Bits 23–16. This field defines the lowest bus number in
configuration region i (where i = 0, 1, 2, 3).
Bus Number Limit i (BusNumLimiti)—Bits 31–24. This bit field defines the highest bus number in
configuration region i (where i = 0, 1, 2, 3).
3.5
Function 2—DRAM Controller
Function 2 configuration registers are listed in Table 4.
Table 4. Function 2 Configuration Registers
Offset
Register Name
Reset
Access
Description
00h
Device ID
Vendor ID (AMD)
1102_1022h
RO
page 67
04h
Status1
Command
0000_0000h
RO
08h
Base
Class
Code
Subclass
Code
Program
Interface
Revision
ID
0600_0000h
RO
page 68
0Ch
BIST
Header
Type
Latency
Timer
Cache
Line Size
0080_0000h
RO
page 69
10h
Base Address 0
0000_0000h
RO
14h
Base Address 1
0000_0000h
RO
18h
Base Address 2
0000_0000h
RO
1Ch
Base Address 3
0000_0000h
RO
20h
Base Address 4
0000_0000h
RO
24h
Base Address 5
0000_0000h
RO
28h
Card Bus CIS Pointer
0000_0000h
RO
2Ch
Sub-System ID
0000_0000h
RO
30h
ROM Base Address
0000_0000h
RO
34h
Capabilities
0000_0000h
RO
38h
reserved
0000_0000h
RO
Sub-System Vendor ID
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
66
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 4. Function 2 Configuration Registers (Continued)
Offset
Register Name
Reset
Access
3Ch
Max
Latency
0000_0000h
RO
40h
DRAM CS Base 0
0000_0000h
RW
page 69
44h
DRAM CS Base 1
0000_0000h
RW
page 69
48h
DRAM CS Base 2
0000_0000h
RW
page 69
4Ch
DRAM CS Base 3
0000_0000h
RW
page 69
50h
DRAM CS Base 4
0000_0000h
RW
page 69
54h
DRAM CS Base 5
0000_0000h
RW
page 69
58h
DRAM CS Base 6
0000_0000h
RW
page 69
5Ch
DRAM CS Base 7
0000_0000h
RW
page 69
60h
DRAM CS Mask 0
0000_0000h
RW
page 72
64h
DRAM CS Mask 1
0000_0000h
RW
page 72
68h
DRAM CS Mask 2
0000_0000h
RW
page 72
6Ch
DRAM CS Mask 3
0000_0000h
RW
page 72
70h
DRAM CS Mask 4
0000_0000h
RW
page 72
74h
DRAM CS Mask 5
0000_0000h
RW
page 72
78h
DRAM CS Mask 6
0000_0000h
RW
page 72
7Ch
DRAM CS Mask 7
0000_0000h
RW
page 72
80h
DRAM Bank Address Mapping
0000_0000h
RW
page 73
88h
DRAM Timing Low
0000_0000h
RW
page 78
8Ch
DRAM Timing High
0000_0000h
RW
page 80
90h
DRAM Configuration Low
0000_0000h
RW
page 82
94h
DRAM Configuration High
0000_0000h
RW
page 85
98h
DRAM Delay Line
0000_0000h
RW
page 88
Min GNT
Int Pin
Int Line
Description
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
3.5.1
Device/Vendor ID Register
This register specifies the device and vendor IDs for the Function 2 registers and is part of the
standard PCI configuration header.
Chapter 3
Memory System Configuration
67
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Device/Vendor ID Register
Rev. 3.06
September 2003
Function 2: Offset 00h
31
16 15
0
DevID
VenID
Bits
Mnemonic
Function
R/W
Reset
31–16
DevID
Device ID
R
1102h
15–0
VenID
Vendor ID
R
1022h
Field Descriptions
Vendor ID (VenID)—Bits 15–0. This read-only value is defined as 1022h for AMD.
Device ID (DevID)—Bits 31–16. This read-only value is defined as 1102h for the HyperTransport
technology configuration function.
3.5.2
Class Code/Revision ID Register
This register specifies the class code and revision for the Function 2 registers and is part of the
standard PCI configuration header.
Class Code/Revision ID Register
31
24 23
Function 2: Offset 08h
16 15
BCC
SCC
8
7
PI
0
RevID
Bits
Mnemonic
Function
R/W
Reset
31–24
BCC
Base Class Code
R
06h
23–16
SCC
Subclass Code
R
00h
15–8
PI
Programming Interface
R
00h
7–0
RevID
Revision ID
R
00h
Field Descriptions
Revision ID (RevID)—Bits 7–0.
Programming Interface (PI)—Bits 15–8. This read-only value is defined as 00h.
Sub Class Code (SCC)—Bits 23–16. This read-only value is defined as 00h.
Base Class Code (BCC)—Bits 31–24. This read-only value is defined as 06h for a host bridge
device.
68
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.5.3
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Header Type Register
This register specifies the header type for the Function 2 registers and is part of the standard PCI
configuration header.
Header Type Register
31
Function 2: Offset 0Ch
24 23
16 15
BIST
HType
Bits
Mnemonic
Function
31–24
BIST
23–16
HType
8
7
LatTimer
0
CLS
R/W
Reset
BIST
R
00h
Header Type
R
80h
15–8
LatTimer
Latency Timer
R
00h
7–0
CLS
Cache Line Size
R
00h
Field Descriptions
CacheLineSize (CLS)—Bits 7–0. This read-only value is defined as 00h.
LatencyTimer (LatTimer)—Bits 15–8. This read-only value is defined as 00h.
HeaderType (Htype)—Bits 23–16. This read-only value is defined as 80h to indicate that multiple
functions are present in the configuration header and that the header layout corresponds to a
device header as opposed to a bridge header. See “Register Differences in Revisions of the
AMD Athlon™ 64 and AMD Opteron™ processors” on page 19 for revision information
about this field.
BIST—Bits 31–24. This read-only value is defined as 00h.
3.5.4
DRAM CS Base Address Registers
These registers define DRAM chip select (CS) address mapping. They are programmed based on data
in the Serial Presence Detect (SPD) ROM on each DIMM.
Function 1 address map registers define to which node to route a DRAM request. Function 2 address
map registers define what DRAM chip select to access on that particular node. See “DRAM Address
Map” on page 57. for more information on addresses routed to the DRAM controller (InputAddr).
Memory size of each chip select bank is defined by a DRAM CS Base Address Register and a DRAM
CS Mask register. The chip selects are formed as follows, using the field names from the DRAM CS
Base Address Registers and DRAM CS Mask Registers, where
“,” means “concatenate”
“/” means “not”
Chapter 3
Memory System Configuration
69
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
“==” means “equals”
“&” means “and”
ChipSelect[i] is asserted if the following is true:
CSBE[i] &
( { InputAddr[35:34], (InputAddr[33:25] & /AddrMaskHi[i][33:25]),
( InputAddr[19:13] & /AddrMaskLo[i][19:13]) } ==
{ BaseAddrHi[i][35:34], (BaseAddrHi[i][33:25] & /AddrMaskHi[i][33:25]),
( BaseAddrLo[i][19:13] & /AddrMaskLo[i][19:13]) } ) );
There are eight DRAM CS Base Address and DRAM CS Mask register pairs. DRAM CS Base
Address 0 and DRAM CS Mask 0 define the base address for chip select 0, and so on. DIMM0 is
associated with CS0 and CS1, DIMM1 is associated with CS2 and CS3, and so on. If only DIMM2
and DIMM3 are occupied and DIMM2 is connected to CS[5:4] and DIMM3 is connected to CS[7:6],
then bit 0 of DRAM CS Base Address Register[7-4] is set to 1 to enable these chip-select banks.
Table 5. DRAM CS Base Address and DRAM CS Mask Registers
Memory Module D[127:64]1
Memory Module D[63:0]
DRAM CS Base and Mask 0 CS0
Extended DIMM0
DIMM0
DRAM CS Base and Mask 1 CS1
Extended DIMM0
DIMM0
DRAM CS Base and Mask 2 CS2
Extended DIMM1
DIMM1
DRAM CS Base and Mask 3 CS3
Extended DIMM1
DIMM1
DRAM CS Base and Mask 4 CS4
Extended DIMM2
DIMM2
DRAM CS Base and Mask 5 CS5
Extended DIMM2
DIMM2
DRAM CS Base and Mask 6 CS6
Extended DIMM3
DIMM3
DRAM CS Base and Mask 7 CS7
Extended DIMM3
DIMM3
Registers
Chip Selects
Notes:
1. This column only applies to processors configured for a 128-bit DRAM interface.
DRAM memory can be assigned to chip select banks in two ways: non-interleaving, when contiguous
addresses are assigned to each chip select bank, and interleaving, when non-contiguous addresses are
assigned to each chip select bank.
Non-interleaving mode can always be used. The BIOS must assign the largest DIMM chip-select
range to the lowest address. As addresses increase, the chip select size must remain constant or
decrease. This is necessary to keep DIMM chip select banks on aligned address boundaries as chipselect banks with different depths are added. The masking does not work otherwise.
Memory interleaving mode can be used if the memory system is composed of a single type and size
DRAM, and if the number of chip selects is a power of two. DRAM controller swaps low order
address bits with high order address bits during decoding. As a result, some low order address bits are
used to determine chip select, and some high order address bits are used to determine DIMM row. It
allows that a switch between DIMMs occurs before a page conflict within a DIMM. The one cycle
turnaround saves time needed to close/open new pages. Algorithm for programming DRAM CS Base
70
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Address and DRAM CS Mask registers in interleaving mode are described in section “DRAM
Address Mapping in Interleaving Mode” on page 76.
31
21 20
BaseAddrHi
Bits
Function 2: Offset 40h, 44h, 48h, 4Ch,
50h, 54h, 58h, 5Ch
16 15
reserved
9
8
1
BaseAddrLo
reserved
Mnemonic
Function
R/W
31–21
BaseAddrHi
Base Address (35–25)
R/W
0
20–16
reserved
R
0
15–9
BaseAddrLo
8–1
reserved
0
CSBE
Base Address (19–13)
Chip-Select Bank Enable
0
CSBE
DRAM CS Base Address Registers
Reset
R/W
0
R
0
R/W
0
Field Descriptions
Chip-Select Bank Enable (CSBE)—Bit 0. This bit enables access to the defined address space.
Base Address (19–13) (BaseAddrLo)—Bits 15–9. This field is only used for memory interleaving to
avoid page conflicts. A new chip select bank is accessed before accessing a new row in the old
chip select bank, delaying or avoiding a page conflict.
If contiguous addresses from address 0 are accessed, the controller would first access a row
(e.g., row 0) in bank 0, chip select 0. As the address increases, the controller would access
different columns in the following order:
row 0/bank 0/chip select 0 and then
row 0/bank 1/chip select 0 and then
row 0/bank 2/chip select 0 and then
row 0/bank 3/chip select 0 and then
row 0/bank 0/chip select 1.
A chip select bank does not have a contiguous region of memory assigned to it. Memory is
interleaved between two DIMMs, four DIMMs, or eight DIMMs.
Base Address (35–25) (BaseAddrHi)—Bits 31–21. These bits decode 32-Mbyte blocks of memory.
In the non-interleaving mode (BaseAddrLo has a value of 0), physical addresses map to
DRAM addresses in the following way:
Col—Lowest order physical address bits
Bank—Second lowest order physical address bits (page miss is better than page conflict)
Row—Second highest order physical address bits (page conflict before accessing new chip)
Chip Select—Highest order physical address bits
Chapter 3
Memory System Configuration
71
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
In the non-interleaving mode, contiguous addresses from 0 would first access a row (e.g., row
0) in bank 0, chip select 0. As the address increases, the next access is to a different column
and eventually a different chip select bank, e.g.:
row 0/bank 0/chip select 0 and then
row 0/bank 1/chip select 0 and then
row 0/bank 2/chip select 0 and then
row 0/bank 3/chip select 0 and then
row 1/bank 0/chip select 0
Memory controller accesses all rows and banks in a single chip select bank before accessing a
new chip select bank in the non-interleaving mode.
3.5.5
DRAM CS Mask Registers
The purpose of this register is to exclude the corresponding address bits from the comparison with the
DRAM Base address register.
DRAM CS Mask Registers
reserved
31 30 29
Function 2: Offset 60h, 64h, 68h, 6Ch, 70h, 74h, 78h, 7Ch
21 20
AddrMaskHi
Bits
Mnemonic
31–30
reserved
29–21
AddrMaskHi
20–16
reserved
15–9
AddrMaskLo
8–0
reserved
16 15
reserved
9
8
0
AddrMaskLo
Function
Address Mask (33–25)
Address Mask (19–13)
reserved
R/W
Reset
R
0
R/W
0
R
0
R/W
0
R
0
Field Descriptions
Address Mask (19–13) (AddrMaskLo)—Bits 15–9. This field specifies the addresses to be excluded
for the memory interleaving mode described in the Base Address bit definitions.
Address Mask (33–25) (AddrMaskHi)—Bits 29–21. This field defines the top Address Mask bits.
The bits with an address mask of 1 are excluded from the address comparison. This allows the
memory block size to be larger than 32 Mbytes. If Address Mask bit 25 is set to 1, the memory
block size is 64 Mbytes.
72
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.5.6
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
DRAM Bank Address Mapping Register
The DIMM may have one or two chip-select banks, but it always has support for up to two chip-select
banks. The address mode covers the whole DIMM, regardless of whether there is one or two chipselect banks.
reserved
Bits
Mnemonic
31–15
reserved
14–12
CS7/6
11
reserved
10–8
CS5/4
7
reserved
6–4
CS3/2
3
reserved
2–0
CS1/0
12 11 10
CS7/6
CS5/4
Function
CS7/6
CS5/4
CS3/2
CS1/0
8
7
6
4
CS3/2
R/W
Reset
R
0
R/W
0
R
0
R/W
0
R
0
R/W
0
R
0
R/W
0
3
2
reserved
15 14
reserved
31
Function 2: Offset 80h
reserved
DRAM CS Address Mapping Register
0
CS1/0
Field Descriptions
Chip Select 1/0 (CS1/0)—Bits 2–0. This field specifies the memory module size. This field is
programmed the same regardless of whether the DRAM interface is 64-bits or 128-bits wide.
This field describes the type of DIMMs that compose the chip select.
Chip Select Bank Address Mode Encoding:
000b = 32 Mbyte (Rows =12 & Col=8)
001b = 64 Mbyte (Rows =12 & Col=9)
010b = 128 Mbyte (Rows =13 & Col=9) | (Rows=12 & Col=10)
011b = 256 Mbyte (Rows =13 & Col=10) | (Rows=12 & Col=11)
100b = 512 Mbyte (Rows =13 & Col=11) | (Rows=14 & Col=10)
101b = 1 Gbyte (Rows =14 & Col=11) | (Rows=13 & Col=12)
110b = 2 Gbyte (Rows =14 & Col=12)
111b = reserved
Chip Select 3/2 (CS3/2)—Bits 6–4. This field specifies the memory module size. The bits are
encoded as shown in CS1/0.
Chip Select 5/4 (CS5/4)—Bits 10–8. This field specifies the memory module size. The bits are
encoded as shown in CS1/0.
Chip Select 7/6 (CS7/6)—Bits 14–12. This field specifies the memory module size. The bits are
encoded as shown in CS1/0.
Chapter 3
Memory System Configuration
73
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.5.6.1
26094
Rev. 3.06
September 2003
DRAM Address Mapping in Non-Interleaving Mode
The address line mapping for non-interleaving mode with a 64-bit DRAM interface is shown in Table
6 and with a 128-bit DRAM interface in Table 7. Column A10 in both tables presents the precharge
all signal (PC).
Table 6. Mapping Host Address Lines to Memory Address Lines (64-Bit Interface)
64-Bit Interface to Memory
Reg 80h
CSn[2:0]
B1
B0
A1
3
A1
2
A1
1
A1
0
A9
A8
A7
A6
A5
A4
A3
A2
A1
A0
000b
64Mb x16,
32MB
Ro
12
w
12
Col
11
11
X
X
X
X
18
X
17
PC
16
X
15
X
14
10
13
9
24
8
23
7
22
6
21
5
20
4
19
3
001b
128Mb x16/
64Mb x8,
64MB
Ro
12
w
12
Col
13
13
X
X
X
X
18
X
17
PC
16
X
15
11
14
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
010b
64Mb x4/
256Mb x16/
128Mb x8,
128MB
Ro
12
w
12
Col
13
13
X
X
26
X
18
X
17
PC
16
26
15
11
14
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
011b
128Mb x4/
512Mb x16/
256Mb x8,
256MB
Ro
14
w
14
Col
13
13
X
X
27
X
18
27
17
PC
16
12
15
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
100b
256Mb x4/
512Mb x8/
1Gb x16,
512MB
Ro
14
w
14
Col
13
13
28
X
27
X
18
28
17
PC
16
12
15
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
101b
512Mb x4/
1Gb x8,
1024MB
Ro
14
w
14
Col
15
15
28
X
27
28
18
13
17
PC
16
12
29
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
110b
1Gb x4
2048MB
Ro
14
w
14
Col
15
15
28
X
27
30
18
13
17
PC
16
12
29
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
19
3
74
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 7. Mapping Host Address Lines to Memory Address Lines (128-Bit Interface)
128-Bit Interface to Memory
Reg 80h
CSn[2:0]
B1
B0
A1
3
A1
2
A1
1
A1
0
A9
A8
A7
A6
A5
A4
A3
A2
A1
A0
000b
64Mb x16,
32MB
Ro
w
Col
13
13
12
12
X
X
X
X
19
X
18
PC
17
X
16
X
15
11
14
10
25
9
24
8
23
7
22
6
21
5
20
4
001b
128Mb x16/
64Mb x8,
64MB
Ro
w
Col
13
13
14
14
X
X
X
X
19
X
18
PC
17
X
16
12
15
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
010b
64Mb x4/
256Mb x16/
128Mb x8,
128MB
Ro
w
Col
13
13
14
14
X
X
27
X
19
X
18
PC
17
27
16
12
15
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
011b
128Mb x4/
512Mb x16/
256Mb x8,
256MB
Ro
w
Col
15
15
14
14
X
X
28
X
19
28
18
PC
17
13
16
12
27
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
100b
256Mb x4/
512Mb x8/
1Gb x16,
512MB
Ro
w
Col
15
15
14
14
29
X
28
X
19
29
18
PC
17
13
16
12
27
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
101b
512Mb x4/
1Gb x8,
1024MB
Ro
w
Col
15
15
16
16
29
X
28
29
19
14
18
PC
17
13
30
12
27
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
110b
1Gb x4
2048MB
Ro
w
Col
15
15
16
16
29
X
28
31
19
14
18
PC
17
13
30
12
27
11
26
10
25
9
24
8
23
7
22
6
21
5
20
4
The BIOS sets up the base address and base mask registers after determining the number of populated
DIMMs, how many chip-select banks they have, and their sizes. A single base address and mask are
initialized for each chip select. Hardware decodes addresses and determines to which DIMM chipselect range the address maps, as a function of the base address and mask.
Chapter 3
Memory System Configuration
75
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Example. DRAM memory consists of:
• DIMM0, 2-sided with 128 MBytes on each side (total DIMM0 memory is 256 Mbytes)
• DIMM1, 2-sided with 256 MBytes on each side (total DIMM1 memory: 512 Mbyte)
• DIMM2, 2-sided with 64 MBytes on each side (total DIMM2 memory: 128 Mbyte)
• DIMM3, 2-sided with 128 MBytes on each side (total DIMM3 memory: 256 Mbyte)
A 64-bit interface to DRAM is used.
Because two of the DIMMs are the same size, there are two different ways to program the base
address and base mask registers. One way to program them is as follows:
Function 2, Offset 80h = 0000_2132h // CS0/1 = 128 MB; CS2/3 = 256 MB; CS4/5 = 64 MB;
CS6/7 = 128 MB
Function 2, Offset 5Ch = 0380_0001h // 896 MB base = 768 MB + 128 MB
Function 2, Offset 58h = 0300_0001h // 768 MB base = 640 MB + 128 MB
Function 2, Offset 54h = 0440_0001h // 1088 MB base = 1024 MB + 64 MB
Function 2, Offset 50h = 0400_0001h // 1024 MB base = 896 MB + 128 MB
Function 2, Offset 4Ch = 0100_0001h // 256 MB base = 0 MB + 256 MB
Function 2, Offset 48h = 0000_0001h // 0 MB base
Function 2, Offset 44h = 0280_0001h // 640 MB base = 512 MB + 128 MB
Function 2, Offset 40h = 0200_0001h // 512 MB base = 256 MB + 256 MB
Function 2, Offset 7Ch = 0060_FE00h // CS7 = 128 MB
Function 2, Offset 78h = 0060_FE00h // CS6 = 128 MB
Function 2, Offset 74h = 0020_FE00h // CS5 = 64 MB
Function 2, Offset 70h = 0020_FE00h // CS4 = 64 MB
Function 2, Offset 6Ch = 00E0_FE00h // CS3 = 256 MB
Function 2, Offset 68h = 00E0_FE00h // CS2 = 256 MB
Function 2, Offset 64h = 0060_FE00h // CS1 = 128 MB
Function 2, Offset 60h = 0060_FE00h // CS0 = 128 MB
3.5.6.2
DRAM Address Mapping in Interleaving Mode
In memory interleaving mode all DIMM chip-select ranges are the same size and type, and the
number of chip selects is a power of two. A BIOS algorithm for programming DRAM CS Base
Address and DRAM CS Mask registers in memory interleaving mode is:
1. Program all DRAM CS Base Address and DRAM CS Mask registers using contiguous mapping.
2. For each enabled chip select, swap BaseAddrHi bits with BaseAddrLo bits, defined in Table 8 and
Table 9.
76
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
3. For each enabled chip select, swap AddrMaskHi bits with AddrMaskLo bits, defined in Table 8
and Table 9.
Table 8. Swapped physical address lines for interleaving with 64-bit interface
Swapped Base Address and Address Mask bits
Chip Select Size
8 way interleaving
4 way interleaving
2 way interleaving
32-MByte
[27:25] and [15:13]
[26:25] and [14:13]
[25] and [13]
64-MByte
[28:26] and [16:14]
[27:26] and [15:14]
[26] and [14]
128-MByte
[29:27] and [16:14]
[28:27] and [15:14]
[27] and [14]
256-MByte
[30:28] and [17:15]
[29:28] and [16:15]
[28] and [15]
512-MByte
[31:29] and [17:15]
[30:29] and [16:15]
[29] and [15]
1-GByte
[32:30] and [18:16]
[31:30] and [17:16]
[30] and [16]
2-GByte
[33:31] and [18:16]
[32:31] and [17:16]
[31] and [16]
Table 9. Swapped physical address lines for interleaving with 128-bit interface
Swapped Base Address and Address Mask bits
Chip Select Size
8 way interleaving
4 way interleaving
2 way interleaving
64-MByte
[28:26] and [16:14]
[27:26] and [15:14]
[26] and [14]
128-MByte
[29:27] and [17:15]
[28:27] and [16:15]
[27] and [15]
256-MByte
[30:28] and [17:15]
[29:28] and [16:15]
[28] and [15]
512-MByte
[31:29] and [18:16]
[30:29] and [17:16]
[29] and [16]
1-GByte
[32:30] and [18:16]
[31:30] and [17:16]
[30] and [16]
2-GByte
[33:31] and [19:17]
[32:31] and [18:17]
[31] and [17]
4-GByte
Not implemented.
[33:32] and [18:17]
[32] and [17]
Example. DRAM memory consists of two 2-sided DIMMs with 256 MBytes on each side. A 64-bit
interface to DRAM is used.
1. Register settings for contiguous memory mapping are:
Function 2, Offset 80h = 0000_0033h // CS0/1 = 256 MB; CS2/3 = 256 MB
Function 2, Offset 40h = 0000_0001h // 0 MB base
Function 2, Offset 44h = 0100_0001h // 256 MB base = 0 MB + 256 MB
Function 2, Offset 48h = 0200_0001h // 512 MB base = 256 MB + 256 MB
Function 2, Offset 4Ch = 0300_0001h // 768 MB base = 512 MB + 256 MB
Function 2, Offset 60h = 00e0_fe00h // CS0 = 256 MB
Function 2, Offset 64h = 00e0_fe00h // CS1 = 256 MB
Chapter 3
Memory System Configuration
77
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Function 2, Offset 68h = 00e0_fe00h // CS2 = 256 MB
Function 2, Offset 6Ch = 00e0_fe00h // CS3 = 256 MB
2. Base Address bits to be swapped are defined inTable 8 (64-bit interface), 256MByte row (chip
select size), 4 way interleaving column (4 chip selects are used). Base Address bits [29:28] are
defined with BaseAddrHi[25:24]. Base Address bits [16:15] are defined with
BaseAddrLo[12:11].
Function 2, Offset 40h=0000_0001h
Function 2, Offset 44h=0000_0801h
Function 2, Offset 48h=0000_1001h
Function 2, Offset 4Ch=0000_1801h
3. Address Mask bits to be swapped are the same as the Base Address bits defined in the previous
step. Address Mask bits [29:28] are defined with AddrMaskHi[25:24]. Address Mask bits [16:15]
are defined with AddrMaskLo[12:11].
Function 2, Offset 60h=03e0_e600h
Function 2, Offset 64h=03e0_e600h
Function 2, Offset 68h=03e0_e600h
Function 2, Offset 6Ch=03e0_e600h
3.5.7
DRAM Timing Low Register
This register contains timing parameters specified in a DRAM data sheet.
Bits
Mnemonic
31–29
reserved
Trp
20 19 18
Tras
16 15 14
Trrd
reserved
24 23
reserved
reserved
reserved
29 28 27 26
Twr
31
Function 2: Offset 88h
12 11
Trcd
Function
7
Trfc
4
Trc
R/W
Reset
R
0
28
Twr
27
reserved
26–24
Trp
23–20
Tras
19
reserved
18–16
Trrd
Active-to-active (RAS#-to-RAS#) Delay
15
reserved
R
0
14–12
Trcd
RAS#-active to CAS#-read/write Delay
R/W
0
11–8
Trfc
Row Refresh Cycle Time
R/W
0
78
Write Recovery Time
8
R/W
0
R
0
Row Precharge Time
R/W
0
Minimum RAS# Active Time
R/W
0
R
0
R/W
0
Memory System Configuration
3
reserved
DRAM Timing Low Register
2
0
Tcl
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
Reset
7–4
Trc
Row Cycle Time
R/W
0
3
reserved
R
0
2–0
Tcl
CAS# Latency
R/W
0
Field Descriptions
CAS# Latency (Tcl)—Bits 2–0. Specifies the CAS#-to-read-data-valid.
000b = reserved
001b = CL=2
010b = CL=3
011b = reserved
100b = reserved
101b = CL=2.5
110b = reserved
111b = reserved
Row Cycle Time (Trc)—Bits 7–4. RAS#-active to RAS#-active or auto refresh of the same bank.
Typically ~70 ns. These bits are encoded as follows (only 7–13 are valid; all others are
reserved):
0000b = 7 bus clocks
0001b = 8 bus clocks
...
1110b = 21 bus clocks
1111b = 22 bus clocks
Row Refresh Cycle Time (Trfc)—Bits 11–8. Auto-refresh-active to RAS#-active or RAS# autorefresh.
0000b = 9 bus clocks
0001b = 10 bus clocks
...
1110b = 23 bus clocks
1111b = 24 bus clocks
RAS#-active to CAS#-read/write Delay (Trcd)—Bits 14–12. Specifies the RAS#-active to CAS#read/write delay to the same bank.
000b = reserved
001b = reserved
010b = 2 bus clocks
011b = 3 bus clocks
100b = 4 bus clocks
101b = 5 bus clocks
110b = 6 bus clocks
111b = reserved
Chapter 3
Memory System Configuration
79
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Active-to-active (RAS#-to-RAS#) Delay (Trrd)—Bits 18–16. Specifies the active-to-active delay
(RAS#-to-RAS#) of different banks.
000b = reserved
001b = reserved
010b = 2 bus clocks
011b = 3 bus clocks
100b = 4 bus clocks
101b = reserved
110b = reserved
111b = reserved
Minimum RAS# Active Time (Tras)—Bits 23–20. Specifies the minimum RAS# active time.
0000b to 0100b = reserved
0101b = 5 bus clocks
...
1111b = 15 bus clocks
Row Precharge Time (Trp)—Bits 26–24. Specifies the Row precharge Time. (Precharge-to-Active
or Auto-Refresh of the same bank.
000b = reserved
001b = reserved
010b = 2 bus clocks
011b = 3 bus clocks
100b = 4 bus clocks
101b = 5 bus clocks
110b = 6 bus clocks
111b = reserved
Write Recovery Time (Twr)—Bit 28. Measures when the last write datum is safely registered by the
DRAM. It measures from the last data to precharge (writes can go back-to-back).
0 = 2 bus clocks
1 = 3 bus clocks
3.5.8
DRAM Timing High Register
This register contains the normal timing parameters specified in a DRAM data sheet.
23 22
reserved
80
20 19
Twcl
13 12
reserved
Memory System Configuration
8
Tref
7
6
4
Trwt
1
0
reserved
Twtr
31
Function 2: Offset 8Ch
reserved
DRAM Timing High Register
3
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
31–23
reserved
22–20
Twcl
19–13
reserved
12–8
Tref
7
reserved
6–4
Trwt
3–1
reserved
0
Twtr
Function
R/W
Reset
R
0
Write CAS Latency
R/W
0
R
0
Refresh Rate
R/W
0
R
0
Read-to-Write Delay
R/W
0
R
0
Write-to-Read Delay
R/W
0
Field Descriptions
Write-to-Read Delay (Twtr)—Bit 0. This bit specifies the write-to-read delay. It is measured from
the rising edge following the last non-masked data strobe to the rising edge of the next Read
command.
0 = 1 bus Clocks
1 = 2 bus Clocks
Read-to-Write Delay (Trwt)—Bits 6–4. Specifies the read-to-write delay.
This is not a DRAM-specified timing parameter, but must be considered due to routing
latencies on the clock forwarded bus. It is counted from the first address bus slot that was not
associated with part of the read burst. These bits are encoded as follows:
000b = 1 bus clocks
001b = 2 bus clocks
010b = 3 bus clocks
011b = 4 bus clocks
100b = 5 bus clocks
101b = 6 bus clocks
110b = reserved
111b = reserved
Refresh Rate (Tref)—Bits 12–8. Specifies the refresh rate of the DIMM requiring the most frequent
refresh. These bits are encoded as follows:
00000b = 100 MHz 4K rows
00001b = 133 MHz 4K rows
00010b = 166 MHz 4K rows
00011b = 200 MHz 4K rows
01000b = 100 MHz 8K/16K rows
01001b = 133 MHz 8K/16K rows
01010b = 166 MHz 8K/16K rows
01011b = 200 MHz 8K/16K rows
Chapter 3
Memory System Configuration
81
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Write CAS Latency (Twcl)—Bits 22–20. Specifies the write CAS latency. Unbuffered DIMMs
should be programmed as 0 and registered DIMMs as 1. These bits are encoded as follows:
000b = 1 Mem clock after CAS#
001b = 2 Mem clocks after CAS#
3.5.9
DRAM Configuration Registers
The following registers contain all the bits to configure the DRAMs for proper operation.
3.5.9.1
DRAM Configuration Low Register
This register controls drive strength, refresh, and initialization.
Bits
Mnemonic
31–28
reserved
27–25
24
reserved
Function
R/W
Reset
R
0
BypMax
Bypass Max
R/W
0
DisInRcvrs
Disable DRAM Receivers
R/W
0
23–20
x4DIMMS
x4DIMMS
R/W
0
19
32ByteEn
Enable 32-Byte Granularity
R/W
0
18
UnBuffDimm
Unbuffered DIMMs
R/W
0
17
DimmEcEn
DIMM ECC Enable
R/W
0
16
128/64
128-Bit/64-Bit
R/W
0
15–14
RdWrQByp
Read/Write Queue Bypass Count
R/W
0
13
SR_S
Self-Refresh Status
R/W
0
12
ESR
Exit Self-Refresh
R/W
0
11
MemClrStatus
Memory Clear Status
R
0
10
DramEnable
DRAM Enable
R
0
9
reserved
SBZ
R/W
0
DramInit
R/W
0
R
0
8
DramInit
7–4
reserved
3
DisDqsHys
Disable DQS Hysteresis
R/W
0
2
QFC_EN
QFC_EN Enable
R/W
0
1
D_DRV
D_DRV
R/W
0
0
DLL_Dis
DLL Disable
R/W
0
82
Memory System Configuration
3
2
1
0
D_DRV
4
DLL_Dis
7
QFC_EN
8
DisDqsHys
reserved
DramEnable
MemClrStatus
ESR
SR_S
RdWrQByp
128/64
x4DIMMS
DimmEcEn
BypMax
20 19 18 17 16 15 14 13 12 11 10 9
UnBuffDimm
reserved
25 24 23
32ByteEn
28 27
DisInRcvrs
31
Function 2: Offset 90h
DramInit
DRAM Configuration Low Register
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Field Descriptions
DLL Disable (DLL_Dis)—Bit 0. Disables the Digital Locked Loop that controls relationship
between the memory Data and the memory strobes (DQS).
0 = Enabled (default)
1 = Disabled
D_DRV (D_DRV)—Bit 1. Controls the drive strength of the DRAM control lines. This bit is required
by the DRAM in the Extended Memory Register Set (EMRS) command. It applies to all
outputs.
0 = Normal Drive (default)
1 = Weak Drive (beware)
QFC_EN Enable (QFC_EN)—Bit 2. Causes the corresponding enable bit to be set in the (external)
DRAM extended mode register. The QFC signal is an optional DRAM output control used to
isolate module loads (DIMMs) from the system memory bus by means of FET switches when
the given module is not being accessed.
0 = Disabled (default)
1 = Enabled
Disable DQS Hysteresis (DisDqsHys)—Bit 3. Disables DQS input filtering when set to 1. It is
required that BIOS (1) set this bit high prior to initializing DRAM (through bit 8, DramInit)
and (2), in a subsequent access to this register, clear this bit; the write that clears this bit may
be concurrent with the write that sets DramInit or it may be prior to the write that sets
DramInit.
DramInit (DramInit)—Bit 8. Set by the BIOS to initiate the DRAM initialization sequence. It is
cleared by the DRAM controller when the initialization completes. Bits 3–0 and Tcl must be
properly programmed before (or at same time as) the bit is set to 1. See bit 3 of this register,
DisDqsHys, for BIOS programming requirements related to DramInit.
0 = Initialization done or not started (default)
DRAM Enable (DramEn)—Bit 10. When set, this bit indicates that DRAM is enabled. This bit is
set by hardware at the completion of DRAM initialization or on an exit from self refresh. See
“Register Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™
processors” on page 19 for revision information about this field.
Memory Clear Status (MemClrStatus)—Bit 11. When set, this bit indicates that the memory clear
function is complete. It is only cleared by reset. BIOS should not write or read the DRAM
until this bit is set by hardware. See “Register Differences in Revisions of the AMD Athlon™
64 and AMD Opteron™ processors” on page 19 for revision information about this field.
Exit Self-Refresh (ESR)—Bit 12. BIOS sets either the DramInit bit (bit 8) or the Exit Self-Refresh
bit (bit 12) after reset, depending on whether the part is coming out of Suspend-to-RAM mode
or not (bit 12 is set when coming out of suspend-to-RAM mode). After this bit is set, it reads
back as a 1 until the process of exiting self refresh is complete, at which point it reads back as
a 0.
Chapter 3
Memory System Configuration
83
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Self-Refresh Status (SR_S)—Bit 13. When asserted, indicates that the DRAMs are in self-refresh
mode. When it deasserts, indicates that the memory controller is ready to process requests
again. This bit may be polled until the DRAMs exit the self-refresh state. This bit cannot
accurately reflect that the DRAM is in self-refresh when resuming from suspend-to-RAM
(STR) because the processor loses power in the STR state. Therefore, this bit must be set,
along with bit 12, to properly exit suspend-to-RAM mode.
0 = Normal operation (default)
1 = Self-refresh mode active
Read/Write Queue Bypass Count (RdWrQByp)—Bits 15–14. Specifies the number of times the
oldest operation in the DCI read/write queue can be bypassed before the arbiter is overridden
and the oldest operation is chosen.
00b = 2
01b = 4
10b = 8
11b = 16
128-Bit/64-Bit (128/64)—Bit 16. Indicates a 128-bit interface to DRAM.
0 = 64-bit interface to DRAM
1 = 128-bit interface to DRAM
DIMM ECC Enable (DimmEcEn)—Bit 17. When DimmEcEn is set to 1, it indicates that all
enabled DIMMs have support for ECC check bit storage.
0 = Some DIMMs do not have ECC bits
1 = All DIMMs have ECC bits
Unbuffered DIMMs (UnBuffDimm)—Bit 18. When asserted, the DRAM controller is only
connected to unbuffered DIMMs. In this case the controller drives address/control 3/4
MEMCLK earlier than the clock to account for the address flight time. If deasserted, the
controller drives address/control 1/2 MEMCLK earlier than clock.
0 = Buffered DIMMs
1 = Unbuffered DIMMs
Enable 32-Byte Granularity (32ByteEn)—Bit 19. When asserted, indicates that the burst counter
should be chosen to optimize data bus bandwidth for 32-byte accesses (except when the bus
width is 128 bits). This bit, along with bit 16, produces the following truth table.
Bit 19
Bit 16
0
0
8-beat bursts (64 bytes)
0
1
4-beat bursts (64 bytes)
1
0
4-beat bursts (32 bytes)
1
1
4-beat bursts (64 bytes)
x4DIMMS[3:0]—Bits 23–20. Indicates whether each of the four DIMMs are x4 or not.
84
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
0 = DIMM is not x4
1 = x4 DIMM present
Disable DRAM Receivers (DisInRcvrs)—Bit 24. When the memory controller is disabled, this bit
is asserted and all DRAM receivers are disabled. The inputs/bidirectional DRAM pins can be
left unconnected.
0 = Receivers enabled
1 = Receivers disabled
Bypass Max (BypMax)—Bits 27–25. Specifies the number of times the oldest entry in DCQ can be
bypassed in arbitration before the arbiter's choice is vetoed. When programmed to all 0s, this
feature is disabled, and the choice of the arbiter is always respected.
3.5.9.2
DRAM Configuration High Register
DRAM Configuration High Register
Bits
Mnemonic
31–30
reserved
29
16 15
ILD_lmt
DCC_EN
reser
ved
20 19 18
MemClk
MCR
MC0_EN
MC1_EN
MC2_EN
reser
ved
MC3_EN
31 30 29 28 27 26 25 24 23 22
Function 2: Offset 94h
12 11
reserved
8
RdPreamble
7
4
reserved
Function
R/W
Reset
R
0
MC3_EN
Memory Clock 3 Enabled
R/W
0
28
MC2_EN
Memory Clock 2 Enabled
R/W
0
27
MC1_EN
Memory Clock 1 Enabled
R/W
0
26
MC0_EN
Memory Clock 0 Enabled
R/W
0
25
MCR
Memory Clock Ratio Valid
R/W
0
24–23
reserved
R
0
22–20
MemClk
DRAM MEMCLK Frequency
R/W
0
19
DCC_EN
Dynamic Idle Cycle Center Enable
R/W
0
18–16
ILD_lmt
Idle Cycle Limit
R/W
0
15–12
reserved
R
0
11–8
RdPreamble
Read Preamble
R/W
0
7–4
reserved
R
0
3–0
AsyncLat
Maximum Asynchronous Latency
R/W
0
3
0
AsyncLat
Field Descriptions
Maximum Asynchronous Latency (AsyncLat)—Bits 3–0. This field should be loaded with a 4-bit
value equal to the maximum asynchronous latency in the DRAM read round-trip loop.
0000b = 0 ns
...
1111b = 15 ns
Chapter 3
Memory System Configuration
85
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Read Preamble (RdPreamble)—Bits 11–8. The time prior to the max-read DQS-return when the
DQS receiver should be turned on. This is specified in units of 0.5 ns. The controller needs to
know when to enable its DQS receiver in anticipation of the DRAM DQS driver turning on for
a read. The controller will disable its DQS receiver until the read preamble time and then
enable its DQS receiver while the DRAM asserts DQS.
0000b = 2.0 ns
0001b = 2.5 ns
0010b = 3.0 ns
0011b = 3.5 ns
0100b = 4.0 ns
0101b = 4.5 ns
0110b = 5.0 ns
0111b = 5.5 ns
1000b = 6.0 ns
1001b = 6.5 ns
1010b = 7.0 ns
1011b = 7.5 ns
1100b = 8.0 ns
1101b = 8.5 ns
1110b = 9.0 ns
1111b = 9.5 ns
Idle Cycle Limit (ILD_lmt)—Bits 18–16. Specifies the number of MemCLKs before forcibly
closing (precharging) an open page. If DCC_EN (Function 2, Offset 94h) has a value of 0, the
static counters are loaded with the ILD_lmt and decremented each clock. If DCC_EN
(Function 2, Offset 94h) has a value of 1, the dynamic counters are loaded with the ILD_lmt
and modified as follows:
Increment—When a Page Miss (PM) page hits on an invalid entry in the Page Table. The
presumption is that in the past, that page table entry was occupied by the very same page that
has a Page Miss. Had the old page been kept open longer, it would have been a Page Hit.
Increment the Idle Cycle Limit count to increase the probability of getting a future Page Hit.
Decrement—When a Page Conflict (PC) arrives and hits on an idle entry (obviously an open
page). This Page Conflict can be avoided if the open page is closed earlier. Decrement the Idle
Cycle Limit count to increase the probability of avoiding a future Page Conflict.
000b = 0 cycles
001b = 4 cycles
010b = 8 cycles
011b = 16 cycles
100b = 32 cycles
101b = 64 cycles
110b = 128 cycles
111b = 256 cycles
86
Memory System Configuration
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Dynamic Idle Cycle Center Enable (DCC_EN)—Bit 19. When set to 1, indicates that each entry in
the page table dynamically adjusts the idle cycle limit based on Page Conflict/Page Miss (PC/
PM) traffic.
DRAM MEMCLK Frequency (MemClk)—Bits 22–20. Specifies the DRAM MEMCLK frequency
in MHz. It is possible to program this field for a higher frequency than the maximum allowed
by the processor. Refer to the processor data sheet for the maximum operating frequency
allowed for a given processor implementation.
000b = 100 MHz
001b = reserved
010b = 133 MHz
011b = reserved
100b = reserved
101b = 166 MHz
110b = reserved
111b = 200 MHz
Memory Clock Ratio Valid (MCR)—Bit 25. Indicates to the controller to drive MEMCLK. The
following fields need to be programmed before this bit is set: MemClk (Function 2, Offset
94h), MCi_EN (Function 2, Offset 94h), UnBuffDimm (Function 2, Offset 90h), and 128/64
(Function 2, Offset 90h). This bit is also set after the Base Address registers have been
configured. This allows the system to know if a CS is occupied before clocks start toggling.
(The intent is to not drive clocks to unoccupied slots to reduce the potential for EMI.)
0 = Disable MemCLKs (default)
1 = Enable MemCLKs
Memory Clock 0 Enabled (MC0_EN)—Bit 26. The MC0_EN bit is set to 1 to enable operation of
the MEMCLK signals to DIMM0.
0 = Disabled (default)
1 = Enabled
Memory Clock 1 Enabled (MC1_EN)—Bit 27. The MC1_EN bit is set to 1 to enable operation of
the MEMCLK signals to DIMM1.
0 = Disabled (default)
1 = Enabled
Memory Clock 2 Enabled (MC2_EN)—Bit 28. The MC2_EN bit is set to 1 to enable operation of
the MEMCLK signals to DIMM2.
0 = Disabled (default)
1 = Enabled
Memory Clock 3 Enabled (MC3_EN)—Bit 29. The MC3_EN bit is set to 1 to enable operation of
the MEMCLK signals to DIMM3.
0 = Disabled (default)
1 = Enabled
Chapter 3
Memory System Configuration
87
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.5.10
26094
Rev. 3.06
September 2003
DRAM Delay Line Register
The DRAM Delay Line register is used to adjust the skew of the input DQS strobe relative to the data.
DRAM Delay Line Register
reserved
Bits
Mnemonic
31–26
reserved
25
AdjFaster
26 25 24 23
AdjSlower
31
Function 2: Offset 98h
16 15
Adj
0
reserved
Function
R/W
Reset
R
0
AdjFaster
Adjust Faster
R/W
0
24
AdjSlower
Adjust Slower
R/W
0
23–16
Adj
Delay Line Adjust
R/W
0
15–0
reserved
R
0
Field Descriptions
Delay Line Adjust (Adj)—Bits 23–16. Adjusts the DLL derived PDL delay by one or more delay
stages in either the faster or slower direction.
Adjust Slower (AdjSlower)—Bit 24. Indicates that the value in Adj is used to increase the PDL
delay.
Adjust Faster (AdjFaster)—Bit 25. Indicates that the value in Adj is used to decrease the PDL
delay.
3.6
Function 3—Miscellaneous Control
This function is a collection of the remaining memory system configuration registers. As listed in
Table 10, Function 3 includes the following registers:
•
Machine check architecture for Northbridge errors
•
DRAM scrubber
•
Buffer counts for Northbridge flow control
•
Power management
•
GART
•
Clocking/FIFO control
88
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
•
Thermtrip status
•
Northbridge capabilities
Table 10. Function 3 Configuration Registers
Offset
Register Name
00h
Device ID
1
Reset
Access
Description
Vendor ID (AMD)
1103_1022h
RO
page 90
Command
0000_0000h
RO
04h
Status
08h
Base
Class
Code
Subclass
Code
Program
Interface
Revision
ID
0600_0000h
RO
page 91
0Ch
BIST
Header
Type
Latency
Timer
Cache
Line Size
0080_0000h
RO
page 91
10h
Base Address 0
0000_0000h
RO
14h
Base Address 1
0000_0000h
RO
18h
Base Address 2
0000_0000h
RO
1Ch
Base Address 3
0000_0000h
RO
20h
Base Address 4
0000_0000h
RO
24h
Base Address 5
0000_0000h
RO
28h
Card Bus CIS Pointer
0000_0000h
RO
2Ch
Sub System ID
0000_0000h
RO
30h
ROM Base Address
0000_0000h
RO
34h
Capabilities
0000_0000h
RO
38h
reserved
0000_0000h
RO
3Ch
Max
Latency
0000_0000h
RO
40h
MCA Northbridge Control
0000_0000h
RW
page 92
44h
MCA Northbridge Configuration
0000_0000h
RW
page 92
48h
MCA Northbridge Status Low
page 98
RW
page 98
4Ch
MCA Northbridge Status High
page 101
RW
page 101
50h
MCA Northbridge Address Low
page 103
RW
page 103
54h
MCA Northbridge Address High
page 104
RW
page 104
58h
Scrub Control
0000_0000h
RW
page 110
5Ch
DRAM Scrub Address Low
page 112
RW
page 112
60h
DRAM Scrub Address High
page 113
RW
page 113
70h
SRI-to-XBAR Buffer Counts
5102_0111h
RW
page 115
Sub system Vendor ID
Min GNT
Int Pin
Int Line
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
Chapter 3
Memory System Configuration
89
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 10. Function 3 Configuration Registers (Continued)
Offset
Register Name
Reset
Access
Description
74h
XBAR-to-SRI Buffer Counts
5000_8011h
RW
page 118
78h
MCT-to-XBAR Buffer Counts
0800_3800h
RW
page 116
7Ch
Free List Buffer Counts
0000_221Bh
RW
page 117
80h
Power Management Control Low
0000_0000h
RW
page 121
84h
Power Management Control High
0000_0000h
RW
page 121
90h
GART Aperture Control
0000_0000h
RW
page 122
94h
GART Aperture Base
page 123
RW
page 123
98h
GART Table Base
page 124
RW
page 124
9Ch
GART Cache Control
0000_0000h
RW
page 125
D4h
Clock Power/Timing Low
page 125
RW
page 125
D8h
Clock Power/Timing High
page 127
RW
page 127
DCh
HyperTransport™ Read Pointer Optimization
page 128
RW
page 128
E4h
Thermtrip Status
0000_0000h
RO
page 129
E8h
Northbridge Capabilities
0000_0000h
RO
page 131
Notes:
1. The unimplemented registers in the standard PCI configuration space are implemented as read-only and return
0 if read.
2. Reads and writes to unimplemented registers in the extended PCI configuration space will result in
unpredictable behavior.
3.6.1
Device/Vendor ID Register
This register specifies the device and vendor IDs for the Function 3 registers and is part of the
standard PCI configuration header.
Device/Vendor ID Register
Function 3: Offset 00h
31
16 15
DevID
0
VenID
Bits
Mnemonic
Function
R/W
Reset
31–16
DevID
Device ID
R
1103h
15–0
VenID
Vendor ID
R
1022h
Field Descriptions
Vendor ID (VenID)—Bits 15–0. This read-only value is defined as 1022h for AMD.
90
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Device ID (DevID)—Bits 31–16. This read-only value is defined as 1103h for the HyperTransport
technology configuration function.
3.6.2
Class Code/Revision ID Register
This register specifies the class code and revision for the Function 3 registers and is part of the
standard PCI configuration header.
Class Code/Revision ID Register
31
24 23
Function 3: Offset 08h
16 15
BCC
SCC
8
7
PI
0
RevID
Bits
Mnemonic
Function
R/W
Reset
31–24
BCC
Base Class Code
R
06h
23–16
SCC
Subclass Code
R
00h
15–8
PI
Programming Interface
R
00h
7–0
RevID
Revision ID
R
00h
Field Descriptions
Revision ID (RevID)—Bits 7–0.
Programming Interface (PI)—Bits 15–8. This read-only value is defined as 00h.
Sub Class Code (SCC)—Bits 23–16. This read-only value is defined as 00h.
Base Class Code (BCC)—Bits 31–24. This read-only value is defined as 06h for a host bridge
device.
3.6.3
Header Type Register
This register specifies the header type for the Function 3 registers and is part of the standard PCI
configuration header.
Header Type Register
31
Function 3: Offset 0Ch
24 23
16 15
BIST
HType
Bits
Mnemonic
Function
31–24
BIST
23–16
HType
Chapter 3
8
7
LatTimer
0
CLS
R/W
Reset
BIST
R
00h
Header Type
R
80h
Memory System Configuration
91
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
Mnemonic
Function
15–8
LatTimer
7–0
CLS
26094
Rev. 3.06
R/W
Reset
Latency Timer
R
00h
Cache Line Size
R
00h
September 2003
Field Descriptions
CacheLineSize (CLS)—Bits 7–0. This read-only value is defined as 00h.
LatencyTimer (LatTimer)—Bits 15–8. This read-only value is defined as 00h.
HeaderType (Htype)—Bits 23–16. This read-only value is defined as 80h to indicate that multiple
functions are present in the configuration header and that the header layout corresponds to a
device header as opposed to a bridge header. See “Register Differences in Revisions of the
AMD Athlon™ 64 and AMD Opteron™ processors” on page 19 for revision information
about this field.
BIST—Bits 31–24. This read-only value is defined as 00h.
3.6.4
Machine Check Architecture (MCA) Registers
These registers are used to configure the Machine Check Architecture (MCA) functions of the onchip Northbridge (NB) hardware and to provide a method for the Northbridge hardware to report
errors in a way compatible with MCA. All of the Northbridge MCA registers, except for the MCA
NB Configuration register, are accessible through the MCA-defined MSR method, as well as through
PCI configuration space. The MCA NB Configuration register is accessible only through PCI
configuration space.
The MCA NB Control register enables MCA reporting of each error checked by the Northbridge. The
global MCA error enables must also be set through the global MCA MSRs. The error enables in this
register only affect error reporting through MCA. Actions which the Northbridge may take in
addition to MCA reporting are enabled through the MCA NB Configuration register.
The MCA NB Status registers and MCA NB Address registers provide status and address information
regarding an error which the Northbridge has logged. The values of error-logging registers are
maintained through a warm reset allowing software to identify an error source that resulted in systemwide initialization.
MCA NB Control Register
92
Memory System Configuration
5
4
3
SyncPkt0En
CrcErr2En
CrcErr1En
2
1
0
CorrEccEn
6
UnCorrEccEn
7
CrcErr0En
8
SyncPkt1En
TgtAbrtEn
GartTblWkEn
reserved
AtomicRMWEn
13 12 11 10 9
WchDogTmrEn
31
Function 3: Offset 40h
SyncPkt2En
MCA NB Control Register
MstrAbrtEn
3.6.4.1
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
Reset
31–13
reserved
12
WchDogTmrEn
R
0
Watchdog Timer Error Reporting Enable
R/W
0
11
AtomicRMWEn
Atomic Read-Modify-Write Error Reporting Enable
R/W
0
10
GartTblWkEn
GART Table Walk Error Reporting Enable
R/W
0
9
TgtAbrtEn
Target Abort Error Reporting Enable
R/W
0
8
MstrAbrtEn
Master Abort Error Reporting Enable
R/W
0
7
SyncPkt2En
HyperTransport Link 2 Sync Packet Error Reporting
Enable
R/W
0
6
SyncPkt1En
HyperTransport Link 1 Sync Packet Error Reporting
Enable
R/W
0
5
SyncPkt0En
HyperTransport Link 0 Sync Packet Error Reporting
Enable
R/W
0
4
CrcErr2En
HyperTransport Link 2 CRC Error Reporting Enable
R/W
0
3
CrcErr1En
HyperTransport Link 1 CRC Error Reporting Enable
R/W
0
2
CrcErr0En
HyperTransport Link 0 CRC Error Reporting Enable
R/W
0
1
UnCorrEccEn
Uncorrectable ECC Error Reporting Enable
R/W
0
0
CorrEccEn
Correctable ECC Error Reporting Enable
R/W
0
Field Descriptions
Correctable ECC Error Reporting Enable (CorrEccEn)—Bit 0. Enables MCA reporting of
correctable ECC errors which are detected in the Northbridge. Since correctable errors do not
result in an MCA exception the effect of this bit is limited to the setting of the error enable bit
in the MCA NB Status High register.
Uncorrectable ECC Error Reporting Enable (UnCorrEccEn)—Bit 1. Enables MCA reporting of
uncorrectable ECC errors which are detected in the Northbridge. Note that in some cases data
may be forwarded to the CPU core prior to checking ECC in which case the check takes place
in one of the other error reporting banks.
HyperTransport Link 0 CRC Error Reporting Enable (CrcErr0En)—Bit 2. Enables MCA
reporting of CRC errors detected on HyperTransport link 0. The Northbridge will flood its
outgoing HyperTransport links with sync packets after detecting a CRC error on an incoming
link independent of the state of this bit.
HyperTransport Link 1 CRC Error Reporting Enable (CrcErr1En)—Bit 3. Enables MCA
reporting of CRC errors detected on HyperTransport link 1. The Northbridge will flood its
outgoing HyperTransport links with sync packets after detecting a CRC error on an incoming
link independent of the state of this bit.
HyperTransport Link 2 CRC Error Reporting Enable (CrcErr2En)—Bit 4. Enables MCA
reporting of CRC errors detected on HyperTransport link 2. The Northbridge will flood its
outgoing HyperTransport links with sync packets after detecting a CRC error on an incoming
link independent of the state of this bit.
Chapter 3
Memory System Configuration
93
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
HyperTransport Link 0 Sync Packet Error Reporting Enable (SyncPkt0En)—Bit 5. Enables
MCA reporting of HyperTransport sync error packets detected on HyperTransport link 0. The
Northbridge will flood its outgoing HyperTransport links with sync packets after detecting a
sync packet on an incoming link independent of the state of this bit.
HyperTransport Link 1 Sync Packet Error Reporting Enable (SyncPkt1En)—Bit 6. Enables
MCA reporting of HyperTransport sync error packets detected on HyperTransport link 1. The
Northbridge will flood its outgoing HyperTransport links with sync packets after detecting a
sync packet on an incoming link independent of the state of this bit.
HyperTransport Link 2 Sync Packet Error Reporting Enable (SyncPkt2En)—Bit 7. Enables
MCA reporting of HyperTransport sync error packets detected on HyperTransport link 2. The
Northbridge will flood its outgoing HyperTransport links with sync packets after detecting a
sync packet on an incoming link independent of the state of this bit.
Master Abort Error Reporting Enable (MstrAbrtEn)—Bit 8. Enables MCA reporting of master
aborts (HyperTransport packets that return an error status with the Non Existent Address bit
set). The Northbridge will return an error response back to the requestor with any associated
data all 1s independent of the state of this bit.
Target Abort Error Reporting Enable (TgtAbrtEn)—Bit 9. Enables MCA reporting of target
aborts (HyperTransport technology packets that return an error status with the Non Existent
Address bit clear). The Northbridge will return an error response back to the requestor with
any associated data all 1s independent of the state of this bit.
GART Table Walk Error Reporting Enable (GartTblWkEn)—Bit 10. Enables MCA reporting of
GART cache table walks which encounter a GART PTE entry which is invalid.
Atomic Read-Modify-Write Error Reporting Enable (AtomicRMWEn)—Bit 11. Enables MCA
reporting of atomic read-modify-write (RMW) HyperTransport technology commands
received from an noncoherent HyperTransport link which do not target DRAM. Atomic
RMW commands to memory-mapped I/O are not supported in HyperTransport technology.
An atomic RMW command that targets MMIO results in a HyperTransport error response
being generated back to the requesting I/O device. The generation of the HyperTransport error
response is not affected by this bit.
Watchdog Timer Error Reporting Enable (WchDogTmrEn)—Bit 12. Enables MCA reporting of
watchdog timer errors. The watchdog timer checks for Northbridge system accesses for which
a response is expected and where no response is received. See the MCA NB Configuration
register for information regarding configuration of the watchdog timer duration. Note that this
bit does not affect operation of the watchdog timer in terms of its ability to complete an access
that would otherwise cause a system hang. This bit only affects whether such errors are
reported through MCA.
94
Memory System Configuration
Chapter 3
Rev. 3.06
MCA NB Configuration Register
Function
Bits
Mnemonic
31–26
reserved
25
DisPciCfgCpuErrRsp
PCI configuration CPU Error Response Disable
R/W
0
24
IoRdDatErrEn
I/O Read Data Error Log Enable
R/W
0
23
ChipKillEccEn
Chip-Kill ECC Mode Enable
R/W
0
22
EccEn
ECC Enable
R/W
0
21
SyncOnAnyErrEn
Sync Flood On Any Error Enable
R/W
0
20
SyncOnWdogEn
Sync Flood on Watchdog Timer Error Enable
R/W
0
19–18
reserved
R
0
17
GenCrcErrByte1
Generate CRC Error on Byte Lane 1
R/W
0
16
GenCrcErrByte0
Generate CRC Error on Byte Lane 0
R/W
0
15–14
LdtLinkSel
HyperTransport Link Select for CRC Error Generation
R/W
0
13–12
WdogTmrBaseSel
Watchdog Timer Time Base Select
R/W
0
11–9
WdogTmrCntSel
Watchdog Timer Count Select
R/W
0
8
WdogTmrDis
Watchdog Timer Disable
R/W
0
7
IoErrDis
I/O Error Response Disable
R/W
0
6
CpuErrDis
CPU Error Response Disable
R/W
0
5
IoMstAbortDis
I/O Master Abort Error Response Disable
R/W
0
4
SyncPktPropDis
Sync Packet Propagation Disable
R/W
0
3
SyncPktGenDis
Sync Packet Generation Disable
R/W
0
2
SyncOnUcEccEn
Sync Flood on Uncorrectable ECC Error Enable
R/W
0
1
CpuRdDatErrEn
CPU Read Data Error Log Enable
R/W
0
0
CpuEccErrEn
CPU ECC Error Log Enable
R/W
0
3
2
1
R/W
Reset
R
0
0
CpuEccErrEn
4
CpuRdDatErrEn
5
SyncOnUcEccEn
6
SyncPktGenDis
7
SyncPktPropDis
8
WdogTmrDis
9
WdogTmrCntSel
WdogTmrBaseSel
LdtLinkSel
GenCrcErrByte0
GenCrcErrByte1
reserved
SyncOnWdogEn
SyncOnAnyErrEn
EccEn
ChipKillEccEn
reserved
IoRdDatErrEn
26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11
DisPciCfgCpuErrRsp
31
Function 3: Offset 44h
CpuErrDis
MCA NB Configuration Register
IoMstAbortDis
3.6.4.2
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
IoErrDis
26094
Field Descriptions
CPU ECC Error Log Enable (CpuEccErrEn)—Bit 0. Enables reporting of ECC errors for data
destined for the CPU on this node. This bit should be clear if ECC error logging is enabled for
the remaining error reporting blocks in the CPU. Logging the same error in more than one
block may cause a single error event to be treated as a multiple error event and cause the CPU
to enter shutdown.
Chapter 3
Memory System Configuration
95
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
CPU Read Data Error Log Enable (CpuRdDatErrEn)—Bit 1. Enables reporting of read data
errors (master aborts and target aborts) for data destined for the CPU on this node. This bit
should be clear if read data error logging is enabled for the remaining error reporting blocks in
the CPU. Logging the same error in more than one block may cause a single error event to be
treated as a multiple error event and cause the CPU to enter shutdown.
Sync Flood on Uncorrectable ECC Error Enable (SyncOnUcEccEn)—Bit 2. Enables flooding of
all HyperTransport links with sync packets on detection of an uncorrectable ECC error.
Sync Packet Generation Disable (SyncPktGenDis)—Bit 3. Disables flooding of all outgoing
HyperTransport links with sync packets when a CRC error is detected on an incoming link.
By default, sync packet generation for CRC errors is controlled through the HyperTransport
technology LDTn Link Control registers (see page 43).
Sync Packet Propagation Disable (SyncPktPropDis)—Bit 4. Disables flooding of all outgoing
HyperTransport links with sync packets when a sync packet is detected on an incoming link.
Sync packets are propagated by default.
I/O Master Abort Error Response Disable (IoMstAbortDis)—Bit 5. Disables setting the NXA bit
in HyperTransport response packets to I/O devices on detection of a master abort
HyperTransport technology error condition.
CPU Error Response Disable (CpuErrDis)—Bit 6. Disables generation of a read data error
response to the CPU core on detection of a target or master abort HyperTransport technology
error condition.
I/O Error Response Disable (IoErrDis)—Bit 7. Disables setting the Error bit in HyperTransport
response packets to I/O devices on detection of a target or master abort HyperTransport
technology error condition.
Watchdog Timer Disable (WdogTmrDis)—Bit 8. Disables the watchdog timer. The watchdog
timer is enabled by default and checks for Northbridge system accesses for which a response
is expected and where no response is received. If such a condition is detected the outstanding
access is completed by generating an error response back to the requestor. An MCA error may
also be generated if enabled in the MCA NB Control register.
Watchdog Timer Count Select (WdogTmrCntSel)—Bit 11–9. Selects the count used by the
watchdog timer. The counter selected by WdogTmrCntSel determines the maximum count
value in the time base selected by WdogTmrBaseSel.
000b = 4095 (default)
001b = 2047
010b = 1023
011b = 511
100b = 255
101b = 127
110b = 63
111b = 31
96
Memory System Configuration
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Watchdog Timer Time Base Select (WdogTmrBaseSel)—Bit 13–12. Selects the time base used by
the watchdog timer. The counter selected by WdogTmrCntSel determines the maximum count
value in the time base selected by WdogTmrBaseSel.
00b = 1 ms (default)
01b = 1 µs
10b = 5 ns
11b = reserved
HyperTransport Link Select for CRC Error Generation (LdtLinkSel)—Bit 15–14. Selects the
HyperTransport link to be used for CRC error injection through GenCrcErrByte1/
GenCrcErrByte0.
00b = HyperTransport link 0
01b = HyperTransport link 1
10b = HyperTransport link 2
11b = undefined
Generate CRC Error on Byte Lane 0 (GenCrcErrByte0)—Bit 16. Causes a CRC error to be
injected on Byte Lane 0 of the HyperTransport link specified by LdtLinkSel. The data carried
by the link is unaffected. This bit is cleared after the error has been generated.
Generate CRC Error on Byte Lane 1 (GenCrcErrByte1)—Bit 17. Causes a CRC error to be
injected on Byte Lane 1 of the HyperTransport link specified by LdtLinkSel. The data carried
by the link is unaffected. This bit is cleared after the error has been generated.
Sync Flood on Watchdog Timer Error Enable (SyncOnWdogEn)—Bit 20. Enables flooding of all
HyperTransport links with sync packets on detection of a watchdog timer error.
Sync Flood On Any Error Enable (SyncOnAnyErrEn)—Bit 21. Enables flooding of all
HyperTransport links with sync packets on detection of any MCA error that is uncorrectable.
ECC Enable (EccEn)—Bit 22. Enables ECC check/correct mode. This bit must be set in order for
any ECC checking/correcting to be enabled for any processor block. If set, ECC will be
checked and correctable errors will be corrected irrespective of whether machine check ECC
reporting is enabled. See “ECC and Chip Kill Error Checking and Correction” on page 108
for more details.
The hardware will only allow values to be programmed into this field which are consistent
with the ECC capabilities of the device as specified in the Northbridge Capabilities register
(page 131). Attempts to write values inconsistent with the capabilities will result in this field
not being updated.
Chip-Kill ECC Mode Enable (ChipKillEccEn)—Bit 23. Enables chip-kill ECC mode. Setting this
bit causes ECC checking to be based on a 128/16 data/ECC rather than on a 64/8 data/ECC.
Chip-kill ECC can only be enabled in 128-bit DRAM data width mode. It allows correction of
an error affecting a single symbol rather than the single bit correction available in 64/8 ECC
checking. When this bit is set, EccEn must be set. See “ECC and Chip Kill Error Checking
and Correction” on page 108 for more details.
Chapter 3
Memory System Configuration
97
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
The hardware will only allow values to be programmed into this field which are consistent
with the ECC and Chip Kill ECC capabilities of the device as specified in the Northbridge
Capabilities register (page 131). Attempts to write values inconsistent with the capabilities
will result in this field not being updated.
I/O Read Data Error Log Enable (IoRdDatErrEn)—Bit 24. Enables reporting of read data errors
(master aborts and target aborts) for data destined for I/O devices. If this bit is clear (default
state), master and target aborts for transactions from I/O devices are not logged by MCA,
although error responses may still be generated to the requesting I/O device.
PCI Configuration CPU Error Response Disable (DisPciCfgCpuErrRsp)—Bit 25. Disables
master abort and target abort reporting through the CPU error-reporting banks for PCI
configuration accesses. It is recommended that this bit be set in order to avoid MCA
exceptions being generated from master aborts for PCI configuration accesses (which can be
common during device enumeration).
3.6.4.3
MCA NB Status Low Register
MCA NB Status Low Register
31
24 23
Syndrome[15:8]
Function 3: Offset 48h
20 19
reserved
16 15
ErrorCodeExt
0
ErrorCode
Bits
Mnemonic
Function
R/W
Reset
31–24
Syndrome[15:8]
Syndrome Bits 15–8 for Chip Kill ECC Mode
R/W
X
23–20
reserved
R
0
19–16
ErrorCodeExt
Extended Error Code
R/W
X
15–0
ErrorCode
Error Code
R/W
X
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Error Code (ErrorCode)—Bits 15–0. Logs an error code when an error is detected. See Table 18 on
page 100 for the error codes. Table 11 on page 99 describes the possible error code formats.
Extended Error Code (ErrorCodeExt)—Bits 19–16. Logs an extended error code when an error is
detected. See Table 18 on page 100 for the extended error codes.
Syndrome Bits 15–8 for Chip Kill ECC Mode (Syndrome[15–8])—Bits 31–24. Logs the upper
eight syndrome bits when an ECC error is detected. Only valid for ECC errors in Chip-Kill
ECC mode.
98
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 11. Error Code Field Formats
Error Value
Error Type
0000 0000 0001 TTLL
TLB errors
Description
Errors in the GART TLB cache.
TT = Transaction Type (Table 12 on page 99)
LL = Cache Level (Table 13 on page 99)
Errors in thc cache hierarchy (not applicable for Northbridge
errors)
0000 0001 RRRR TTLL
Memory errors
RRRR = Memory Transaction Type (Table 14 on page 99)
TT = Transaction Type (Table 12 on page 99)
LL = Cache Level (Table 13 on page 99)
General bus errors including errors in the HyperTransport™
link or DRAM.
0000 1PPT RRRR IILL
Bus errors
PP = Participation Processor (Table 15 on page 100)
T = Time-out (Table 16 on page 100)
RRRR = Memory Transaction Type (Table 14 on page 99)
II = Memory or I/O (Table 17 on page 100)
LL = Cache Level (Table 13 on page 99)
Table 12. Transaction Type Bits (TT)
00b
Instruction
01b
Data
10b
Generic
11b
reserved
Table 13. Cache Level Bits (LL)
00b
Level 0 (L0)
01b
Level 1 (L1)
10b
Level 2 (L2)
11b
Generic (LG)
Table 14. Memory Transaction Type Bits (RRRR)
0000b
Generic error (GEN)
0001b
Generic read (RD)
0010b
Generic write (WR)
0011b
Data read
0100b
Data write
0101b
Instruction fetch
Chapter 3
Memory System Configuration
99
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 14. Memory Transaction Type Bits (RRRR) (Continued)
0110b
Prefetch
0111b
Evict
1000b
Snoop (probe)
Table 15. Participation Processor Bits (PP)
00b
Local node originated the request (SRC)
01b
Local node responded to the request (RES)
10b
Local node observed error as 3rd party (OBS)
11b
Generic
Table 16. Time-Out Bit (T)
0
Request did not time out
1
Request timed out (TIMOUT)
Table 17. Memory or I/O Bits (II)
00b
Memory access (MEM)
01b
reserved
10b
I/O access
11b
Generic (GEN)
Table 18 on page 100 lists the errors reported by the Northbridge, along with the error codes for each
error. The error code fields are defined in the table above. See the NB Control register for more
information on errors detected by the Northbridge.
Table 18. Northbridge Error Codes
Extended
Error
Code
Type
PP
T
RRRR
II
LL
TT
ECC error
0000
BUS
SRC/RSP
0
RD/WR
MEM
LG
-
CRC error
0001
BUS
OBS
0
GEN
GEN
GEN
-
Sync error
0010
BUS
OBS
0
GEN
GEN
GEN
-
Mst Abort
0011
BUS
SRC/OBS
0
RD/WR
MEM/IO
GEN
-
Tgt Abort
0100
BUS
SRC/OBS
0
RD/WR
MEM/IO
GEN
-
GART error
0101
TLB
-
-
-
-
GEN
GEN
RMW error
0110
BUS
OBS
0
GEN
IO
GEN
-
Wdog error
0111
BUS
GEN
1
GEN
GEN
GEN
-
Error Type
100
Error Code
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 18. Northbridge Error Codes
Error Type
Extended
Error
Code
Type
PP
T
RRRR
II
LL
TT
1000
BUS
SRC/RSP
0
RD/WR
MEM
LG
-
ChipKill
ECC error
MCA NB Status High Register
reserved
8
7
6
4
LDTLink
Bits
Mnemonic
Function
R/W
Reset
31
ErrValid
Error Valid
R/W
X
30
ErrOver
Error Overflow
R/W
X
29
ErrUnCorr
Error Uncorrected
R/W
X
28
ErrEn
Error Enable
R/W
X
27
ErrMiscVal
Miscellaneous Error Register Valid
R
X
26
ErrAddrVal
Error Address Valid
R/W
X
25
PCC
Processor Context Corrupt
R/W
X
24–23
reserved
R
0
22–15
ECC_Synd (7–0)
Syndrome Bits (7–0) for ECC Errors
R/W
X
14
CorrECC
Correctable ECC Error
R/W
X
13
UnCorrECC
Uncorrectable ECC Error
R/W
X
12–9
reserved
R
0
8
ErrScrub
Error Found by DRAM Scrubber
R/W
X
7
reserved
R
0
6–4
LDTLink
HyperTransport Link Number
R/W
X
3–2
reserved
R
0
1
ErrCPU1
Error Associated with CPU 1
R/W
X
0
ErrCPU0
Error Associated with CPU 0
R/W
X
3
2
1
0
ErrCPU0
9
ErrCPU1
ECC_Synd (7–0)
UnCorrECC
15 14 13 12
CorrECC
reserved
PCC
ErrAddrVal
ErrMiscVal
ErrEn
ErrOver
ErrUnCorr
ErrValid
31 30 29 28 27 26 25 24 23 22
reserved
Function 3: Offset 4Ch
ErrScrub
MCA NB Status High Register
reserved
3.6.4.4
Error Code
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Error Associated with CPU 0 (ErrCPU0)—Bit 0. If set to 1, this bit indicates that the error was
associated with CPU 0.
Error Associated with CPU 1 (ErrCPU1)—Bit 1. If set to 1, this bit indicates that the error was
associated with CPU 1.
Chapter 3
Memory System Configuration
101
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
HyperTransport Link Number (LDTLink[2:0])—Bits 6–4. For errors associated with a
HyperTransport link (e.g., CRC errors), this field indicates which link was associated with the
error.
LDTLink[0] = Error associated with HyperTransport link 0
LDTLink[1] = Error associated with HyperTransport link 1
LDTLink[2] = Error associated with HyperTransport link 2
Error Found by DRAM Scrubber (ErrScrub)—Bit 8. If set to 1, this bit indicates that the error
was found by the DRAM scrubber.
Uncorrectable ECC Error (UnCorrECC)—Bit 13. If set to 1, this bit indicates that the error was an
uncorrectable ECC error.
Correctable ECC Error (CorrECC)—Bit 14. If set to 1, this bit indicates that the error was a
correctable ECC error.
Syndrome Bits 7–0 for ECC Errors (ECC_Synd[7:0])—Bits 22–15. Logs the lower eight
syndrome bits when an ECC error is detected.
Processor Context Corrupt (PCC)—Bit 25. If set to 1, this bit indicates that the state of the
processor may be corrupted by the error condition. Reliable restarting might not be possible.
0 = Processor not corrupted
1 = Processor may be corrupted
Error Address Valid (ErrAddrVal)—Bit 26. If set to 1, this bit indicates that the address saved in
the address register is the address where the error occurred (i.e. it validates the address in the
address register).
0 = Address register not valid
1 = Address register valid
Miscellaneous Error Register Valid (ErrMiscVal)—Bit 27. If set to 1, this bit indicates whether the
Miscellaneous Error register contains valid information for this error. This bit is read-only 0,
since the Miscellaneous Error register is not implemented.
Error Enable (ErrEn)—Bit 28. If set to 1, this bit indicates that MCA error reporting is enabled in
the MCA Control register.
0 = MCA error reporting not enabled
1 = MCA error reporting enabled
Error Uncorrected (ErrUnCorr)—Bit 29. If set to 1, this bit indicates that the error was not
corrected by hardware.
0 = Error corrected
1 = Error not corrected
Error Overflow (ErrOver)—Bit 30. Set to 1 if the Northbridge detects an error with the valid bit of
this register already set. Enabled errors are written over disabled errors, uncorrectable errors
are written over correctable errors. Uncorrectable errors are not overwritten.
102
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
0 = No error overflow
1 = Error overflow
Error Valid (ErrValid)—Bit 31. If set to 1, this bit indicates that a valid error has been detected.
This bit should be cleared to 0 by software after the register is read.
0 = No valid error detected
1 = Valid error detected
Table 19 lists the errors reported by Northbridge along with the status bits logged for each error. The
status bits are defined in the table above. See the Northbridge Control register for more information
on errors detected by Northbridge.
Table 19. Northbridge Error Status Bit Settings
Error
Type
ErrUnCorr
ErrAddr
Val
PCC
ECC_Synd
Valid
ECC
error
If multi-bit
1
If multi-bit
and src
CPU
1
CRC
error
1
0
1
0
0
Sync
error
1
0
1
0
Mst
Abort
1
1
If src CPU
Tgt
Abort
1
1
GART
error
1
RMW
error
Wdog
error
ChipKill
ECC
error
3.6.4.5
LDTLink
Valid
Err
CPU
1/0
0
1/0
0
0
1
0
0
0
0
1
0
0
0
0
0
1
1/0
If src CPU
0
0
0
0
1
1/0
1
If src CPU
0
0
0
0
0
1/0
1
1
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
X
1
If multisymbol
and src
CPU
1
1/0
0
1/0
If multisymbol
Corr
ECC
UnCorr Err
ECC Scrub
If single- If multibit
bit
If single- If multisymbol symbol
MCA NB Address Low Register
MCA NB Address Low Register
Function 3: Offset 50h
31
3
ErrAddrLo
Chapter 3
Memory System Configuration
2
0
reserved
103
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
26094
Rev. 3.06
Mnemonic
Function
R/W
Reset
31–3
ErrAddrLo
Error Address Bits 31–3
R/W
X
2–0
reserved
R
0
September 2003
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Error Address Bits 31–3 (ErrAddrLo)—Bits 31–3. This field specifies bits 31–3 of the address
associated with a machine check error.
3.6.4.6
MCA NB Address High Register
MCA NB Address High Register
Function 3: Offset 54h
31
8
7
0
reserved
Bits
Mnemonic
31–8
reserved
7–0
ErrAddrHi
ErrAddrHi
Function
Error Address Bits 39–32
R/W
Reset
R
0
R/W
X
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
Error Address Bits 39–32 (ErrAddrHi)—Bits 7–0. This field specifies bits 39–32 of the address
associated with a machine check error.
3.6.4.7
Watchdog Timer Errors
For watchdog timer errors the ErrAddrHi and ErrAddrLo fields in the MCA NB Address High/Low
registers log the hardware state information of the request for which the response timed out.
MCA NB Address Low Register for Watchdog Timer Error
Function 3: Offset 50h
20 19 18 17
DstNode
SrcNode
15 14
SrcPtr
12 11
NextAction
25 24 23 22
SrcUnit
WaitDatMove
WaitPW
WaitCpuDat
WaitLock
31 30 29 28 27
DstUnit
Revision B and earlier revisions.
9
OpType
8
3
LdtCmd
2
0
reserved
Revision C.
104
Memory System Configuration
Chapter 3
25 24 23 22
20 19 18 17
DstNode
Mnemonic
SrcNode
15 14
SrcPtr
12 11
NextAction
reserved
WaitPW
WaitCode
31 30 29 28 27
Bit
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SrcUnit
Rev. 3.06
DstUnit
26094
9
OpType
Function
8
3
2
LdtCmd
R/W
0
reserved
Reset
Bits [31:28] in revision B and earlier revisions.
31
WaitLock
Wait For Bus Lock
30
WaitPW
Wait For Posted Write
29
WaitCpuDat
Wait For CPU Write Data
28
WaitDatMove
Wait For Data Movement
Bits [31:28] in revision C.
31-30
WaitCode
Wait Code (Bits 1-0 of WaitCode)
29
WaitPW
Wait For Posted Write
28
reserved
27–25
DstNode
Destination Node ID
24–23
DstUnit
Destination Unit ID
22–20
SrcNode
Source Node ID
19–18
SrcUnit
Source Unit ID
17–15
SrcPtr
Source Pointer
14–12
NextAction
Next Action
11–9
OpType
Operation Type
8–3
LdtCmd
HyperTransport Command
2–0
reserved
Field Descriptions
HyperTransport Command (LdtCmd)—Bit 8–3. The initial command (in HyperTransport
technology format) for the stalled operation.
Operation Type (OpType)—Bit 11–9. Indicates what type of operation is stalled.
000b = Normal operation (i.e., none of the other types listed)
001b = Bus lock
010b = Local APIC access
011b = Interrupt request
100b = System management request
101b = Interrupt broadcast
110b = Stop grant
111b = SMI ack
Next Action (NextAction)—Bit 14–12. Indicates what the stalled operation would have done next
had it not become stalled.
Chapter 3
Memory System Configuration
105
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
000b
001b
010b
011b
100b
101b
110b
111b
26094
Rev. 3.06
September 2003
= Complete
= reserved
= Send request
= Send second request (if any)
= Send final response back to requestor
= Send initial response (if any) back to requestor
= Send final response to the memory controller
= Send initial response (if any) to mem controller
Source Pointer (SrcPtr)—Bit 17–15. Source pointer of the stalled operation.
000b = Host bridge on local node
001b = CPU on local node
010b = Mem controller on local node
101b = HyperTransport link 0
110b = HyperTransport link 1
111b = HyperTransport link 2
Source Unit ID (SrcUnit)—Bit 19–18. Source Unit ID of the stalled operation.
Source Node ID (SrcNode)—Bit 22–20. Source Node ID of the stalled operation.
Destination Unit ID (DstUnit)—Bit 24–23. Destination Unit ID of the stalled operation.
Destination Node ID (DstNode)—Bit 27–25. Destination Node ID of the stalled operation.
Bits [31:28] in revision B and earlier revisions:
Wait For Data Movement (WaitDatMove)—Bit 28. When set, indicates that the stalled operation is
waiting for read or write data to be moved.
Wait For CPU Write Data (WaitCpuDat)—Bit 29. When set, indicates that the stalled operation is
waiting for write data from the CPU.
Wait For Posted Write (WaitPW)—Bit 30. When set, indicates that a response for the stalled
operation is waiting for one or more prior posted writes to complete.
Wait For Bus Lock (WaitLock)—Bit 31. When set, indicates that the stalled operation is waiting for
a bus lock to complete.
Bits [31:28] in revision C:
Wait For Posted Write (WaitPW)—Bit 29. When set, indicates that a response for the stalled
operation is waiting for one or more prior posted writes to complete.
Wait Code (WaitCode)—Bit 31-30. WaitCode is a 5 bit field; WaitCode[1:0] is in Function 3, Offset
50h; WaitCode[4:2] is in Function 3, Offset 54h. WaitCode encodings are implementation
specific. They describe the reason the hardware is waiting. All zeros mean that there is not a
waiting condition. Revision B and earlier revision fields: WaitDatMove, WaitCpuDat,
WaitLock (Function 3, Offset 50h), WaitPrbRsp0, WaitPrbRsp1, and WaitPriorOp (Function
106
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
3, Offset 50h) are replaced with a code in the revision C field WaitCode (Function 3, Offset
50h, 54h).
MCA NB Address High Register for Watchdog Timer Error
Function 3: Offset 54h
6
reserved
3
RspCnt
2
1
0
WaitPrbRsp1
7
WaitPrbRsp0
8
GartTblWkInProg
31
WaitPriorOp
Revision B and earlier revisions.
Revision C.
8
reserved
Bits
Mnemonic
Function
31–8
reserved
7
GartTblWkInProg
GART Table Walk In Progress (Bit 39 of ErrAddr)
6–3
RspCnt
System Response Count (Bits 38–35 of ErrAddr)
7
6
3
GartTblWkInProg
31
RspCnt
R/W
Reset
2
0
WaitCode
Bits [2:0] in revision B and earlier revisions.
2
WaitPriorOp
Wait For Prior Operation (Bit 34 of ErrAddr)
1
WaitPrbRsp0
Wait For CPU0 Probe Response (Bit 33 of ErrAddr)
0
WaitPrbRsp1
Wait For CPU1 Probe Response (Bit 32 of ErrAddr)
Bits [2:0] in revision C.
2-0
WaitCode
Wait Code (Bits 4-2 of WaitCode)
Field Descriptions
Bits [2:0] in revision B and earlier revisions:
Wait For CPU1 Probe Response (WaitPrbRsp1)—Bit 0 (Bit 32 of ErrAddr). When set, indicates
that the stalled operation is waiting on a probe response from CPU1.
Wait For CPU0 Probe Response (WaitPrbRsp0)—Bit 1 (Bit 33 of ErrAddr). When set, indicates
that the stalled operation is waiting on a probe response from CPU0.
Wait For Prior Operation (WaitPriorOp)—Bit 2 (Bit 34 of ErrAddr). When set, indicates that the
stalled operation is waiting for a prior operation to release system resources.
Chapter 3
Memory System Configuration
107
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Bits [2:0] in revision C:
Wait Code (WaitCode)—Bit 2-0. See WaitCode definition in MCA NB Address Low Register for
Watchdog Timer Error (Function 3, Offset 50h).
System Response Count (RspCnt)—Bits 6–3 (Bits 38–35 of ErrAddr). Number of outstanding
system responses for which the stalled operation is waiting.
GART Table Walk In Progress (GartTblWkInProg)—Bit 7 (Bit 39 of ErrAddr). Indicates that a
GART table walk was in progress at the time the error was logged.
3.6.5
ECC and Chip Kill Error Checking and Correction
The memory controller implements two ECC modes: normal ECC and Chip Kill ECC. These error
checking and correction modes can only be used if all DIMMs are ECC capable.
3.6.5.1
Normal ECC
Normal ECC mode is a 64/8 SEC/DED (Single Error Correction/Double Error Detection) Hamming
code. It can detect one bit error and correct it, it can detect two bit errors but it can not correct them,
and it may detect more than two bit errors depending on the position of corrupted bits. Normal ECC
mode is enabled when EccEn, Function 3, Offset 44h is set and ChipKillEccEn, Function 3, Offset
44h is clear.
Error address (Function 3, Offsets 50h and 54h) is valid and uniquely identifies the DIMM with a
correctable or uncorrectable error for 64-bit and 128-bit data width configurations. In a 128-bit data
width configuration, ECC is independently applied to the lower 64 and upper 64 data bits. In this
configuration ECC may detect and correct two bit errors if one error happens on data bits 63-0 and
one error happens on data bits 127-64. Status registers (Function 3, Offsets 48h and 4Ch) and error
address registers will contain information about the second detected error.
Bit position of a correctable error can be determined by matching Ecc_Synd, Function 3, Offset 4Ch
with a value from Table 20.
Table 20.
ECC Syndromes
n=0
n=1
n=2
n=3
n=4
n=5
n=6
n=7
Bit (0+n)
ce
cb
d3
d5
d6
d9
da
dc
Bit (8+n)
23
25
26
29
2a
2c
31
34
Bit (16+n)
0e
0b
13
15
16
19
1a
1c
Bit (24+n)
e3
e5
e6
e9
ea
ec
f1
f4
Bit (32+n)
4f
4a
52
54
57
58
5b
5d
Bit (40+n)
a2
a4
a7
a8
ab
ad
b0
b5
Bit (48+n)
8f
8a
92
94
97
98
9b
9d
Bit (56+n)
62
64
67
68
6b
6d
70
75
108
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.6.5.2
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Chip Kill
Chip Kill ECC mode is a 128/16 SSC/DSD (Single Symbol Correction/Double Symbol Detection)
BCH code. A symbol is a group of 4 bits which are 4 bit aligned (bits 0–3 make symbol 0, bits 4–7
make symbol 1, etc.). A single symbol error is any bit error combination within one symbol (1 to 4
bits can be corrupted). Chip Kill ECC mode can detect one symbol error and correct it, it can detect
two symbol errors but it can not correct them, and it may detect more than two symbol errors
depending on the position of corrupted symbols. Chip Kill ECC mode is enabled when EccEn and
ChipKillEccEn, Function 3, Offset 44h are set.
Chip Kill mode can only be used in a 128-bit data width configuration.
A DIMM with a correctable error is uniquely identified with error address (Function 3, Offsets 50h
and 54h) and Chip Kill syndrome (Syndrome, Function 3, Offset 48h and Ecc_Synd, Function 3,
Offset 4Ch). Error address identifies two DIMMs and Chip Kill syndrome identifies one of them.
Chip Kill syndromes for symbols 00h-0fh, 20h, and 21h map to data bits 63-0 and Chip Kill
syndromes for symbols 10h-1fh, 22h, and 23h map to data bits 127-64 (see Table 21). Corrupted bit
positions within a symbol are determined from the Chip Kill syndrome column number in Table 21.
Bits set in the column number identify corrupted bits within a symbol. For example, if 6913h is a
Chip Kill syndrome, symbol 05h has an error, and bits 0 and 1 within that symbol are corrupted, since
the syndrome is in column 3h. Symbol 05h maps to bits 23-20, so the corrupted bits are 20 and 21.
A DIMM with an uncorrectable error can not be uniquely identified. The error address is valid and
maps to two DIMMs.
Chip Kill mode can be used with any device width. Chip Kill mode can correct all errors in an x4
device, since the device width is one symbol. Chip Kill mode can correct a subset of errors in an x8 or
x16 device.
Table 21.
Chip Kill ECC Syndromes
Symbol 1h
2h
3h
4h
5h
6h
7h
8h
9h
ah
bh
ch
dh
eh
00h
e821 7c32 9413 bb44 5365 c776 2f57
01h
5d31 a612 fb23
02h
0001 0002 0003 0004 0005 0006 0007 0008 0009 000a 000b 000c 000d 000e 000f
03h
2021 3032 1013 4044 6065 7076 5057 8088 a0a9 b0ba 909b c0cc e0ed f0fe
04h
5041 a082 f0c3
05h
be21 d732 6913 2144 9f65
06h
4951 8ea2 c7f3
07h
74e1 9872 ec93 d6b4 a255 4ec6 3a27 6bd8 1f39
08h
15c1 2a42 3f83
09h
3d01 1602 2b03 8504 b805 9306 ae07 ca08 f709
0ah
9801 ec02 7403 6b04 f305
Chapter 3
dd88 35a9 a1ba 499b 66cc 8eed 1afe
fh
9584 c8b5 3396 6ea7 eac8 b7f9
4cda 11eb 7f4c
9054 c015 30d6 6097 e0a8 b0e9 402a 106b 70fc
f676
cef4
227d d95e 846f
db35 e4b6 f177
8706 1f07
f3aa
661b f27c
d0df
20bd d07e 803f
4857 3288 8ca9 e5ba 5b9b 13cc aded c4fe
5394 1ac5 dd36 9467 a1e8 e8b9 2f4a
f2df
7adf
bb2d 7cde 358f
874b bd6c c98d 251e 51ff
4758 5299 6d1a 78db 89ac 9c6d a3ee b62f
dc0a e10b 4f0c
720d 590e 640f
bd08 2509 510a c90b d60c 4e0d 3a0e a20f
Memory System Configuration
109
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 21.
26094
Rev. 3.06
bh
ch
September 2003
Chip Kill ECC Syndromes
Symbol 1h
2h
3h
4h
5h
6h
7h
8h
9h
ah
dh
eh
fh
0bh
d131 6212 b323 3884 e9b5 5a96 8ba7 1cc8 cdf9
7eda afeb
244c f57d
465e 976f
0ch
e1d1 7262 93b3 b834 59e5 ca56 2b87 dc18 3dc9 ae7a 4fab
642c 85fd
164e f79f
0dh
6051 b0a2 d0f3
307c 502d 80de e08f
0eh
a4c1 f842
0fh
11c1 2242 3383 c8f4
10h
45d1 8a62 cfb3
11h
63e1 b172 d293 14b4 7755 a5c6 c627 28d8 4b39 99aa fa4b
12h
b741 d982 6ec3 2254 9515 fbd6
13h
dd41 6682 bbc3 3554 e815 53d6 8e97 1aa8 c7e9 7c2a a16b 2ffc
14h
2bd1 3d62 16b3 4f34
64e5 7256 5987 8518 aec9 b87a 93ab ca2c e1fd
f74e
15h
83c1 c142 4283 a4f4
2735 65b6 e677 f858
7b99 391a badb 5cac df6d
9dee 1e2f
16h
8fd1
c562 4ab3 a934 26e5 6c56 e387 fe18
71c9 3b7a b4ab 572c d8fd
924e 1d9f
17h
4791 89e2 ce73 5264 15f5
18h
5781 a9c2 fe43
19h
bf41
1ah
9391 e1e2 7273 6464 f7f5
1bh
cce1 4472 8893 fdb4
1ch
a761 f9b2
1dh
ff61
1eh
5451 a8a2 fcf3
1fh
6fc1
20h
be01 d702 6903 2104 9f05
21h
4101 8202 c303 5804 1905 da06 9b07 ac08 ed09 2e0a 6f0b
22h
c441 4882 8cc3 f654
23h
7621 9b32 ed13 da44 ac65 4176 3757 6f88
3.6.6
1094 70c5 a036 c067 20e8 40b9 904a f01b
5c83 e6f4
4235 1eb6 ba77 7b58 df99
d935 eab6 fb77
831a 27db 9dac 396d 65ee c12f
4c58 5d99 6e1a 7fdb
84ac 956d a6ee b72f
5e34 1be5 d456 9187 a718 e2c9 2d7a 68ab f92c
734e 369f
3c6c 5f8d
8d1e eeff
4c97 33a8 84e9 ea2a 5d6b 11fc
db86 9c17 a3b8 e429 2a5a 6dcb f1dc
92a4 c525 3b66 6ce7 e3f8
d582 6ac3 2954 9615 fcd6
a6bd c87e 7f3f
f2bd
497e 943f
b64d 783e 3faf
4397 3ea8 81e9 eb2a 546b 17fc
a8bd c27e 7d3f
8586 1617 b8b8 2b29 595a cacb dcdc 4f4d
3d3e aeaf
3155 b9c6 7527 56d8 9a39 12aa de4b ab6c 678d ef1e
55b2 aad3 7914 8675 2ca6 d3c7 9e28 6149 cb9a 34fb
9694 c2c5 3e36 6a67 ebe8 bfb9
23ff
913c 365d 688e cfef
e73c 185d b28e 4def
434a 171b 7d7c 292d d5de 818f
7635 acb6 c377 2e58 4199 9b1a f4db
f606
dc9f
b479 4a3a 1dbb 715c 26dd d89e 8f1f
5ed3 e214 4575 1ba6 bcc7 7328 d449 8a9a 2dfb
b542 da83 19f4
bcfd
37ac 586d 82ee ed2f
4807 3208 8c09 e50a 5b0b 130c ad0d c40e 7a0f
3215 bed6 7a97 5ba8 9fe9
f40c
b50d 760e 370f
132a d76b adfc
69bd e57e 213f
19a9 f4ba
829b b5cc c3ed 2efe
58df
Scrub Control Register
This register specifies the scrub rate for sequential ECC scrubbing of DRAM, the L2 cache and the
DCACHE. The scrub rate specifies the duration between successive scrub events. A value of 0
disables scrubbing.
Each scrub event corresponds to:
•
64 bytes for the DRAM scrubber.
•
64 bits for the data cache scrubber.
110
Memory System Configuration
Chapter 3
26094
•
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
One single L2 cache line tag address for the L2 scrubber.
Scrub Control Register
31
Function 3: Offset 58h
21 20
reserved
Bits
Mnemonic
31–21
reserved
20–16
DcacheScrub
15–13
reserved
12–8
L2Scrub
7–5
reserved
4–0
DramScrub
16 15
DcacheScrub
13 12
reserved
8
L2Scrub
7
5
4
reserved
0
DramScrub
Function
R/W
Reset
R
0
Data Cache Scrub Rate
R/W
0
R
0
L2 Cache Scrub Rate
R/W
0
R
0
DRAM Scrub Rate
R/W
0
Field Descriptions
DRAM Scrub Rate (DramScrub)—Bits 4–0. Specifies the scrub rate for the DRAM. See Table 22.
For example, if 256MByte memory is scrubbed every 12 hours, a 64-byte memory block
should be scrubbed every 10ms, and scrubbing rate of 10.49ms should be selected.
L2 Cache Scrub Rate (L2Scrub)—Bits 12–8. Specifies the scrub rate for the L2 cache. See
Table 22.
Data Cache Scrub Rate (DcacheScrub)—Bits 20–16. Specifies the scrub rate for the data cache.
See Table 22.
Table 22. Scrub Rate Control Values
Scrub Rate
Code
Scrub Rate
00000b
Do not scrub 01100b
81.9 µs
00001b
40.0 ns
01101b
163.8 µs
00010b
80.0 ns
01110b
327.7 µs
00011b
160.0 ns
01111b
655.4 µs
00100b
320.0 ns
10000b
1.31 ms
00101b
640.0 ns
10001b
2.62 ms
00110b
1.28 µs
10010b
5.24 ms
00111b
2.56 µs
10011b
10.49 ms
01000b
5.12 µs
10100b
20.97 ms
01001b
10.2 µs
10101b
42.00 ms
01010b
20.5 µs
10110b
84.00 ms
01011b
41.0 µs
All others
reserved
Chapter 3
Scrub Rate
Code
Scrub Rate
Memory System Configuration
111
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.6.7
26094
Rev. 3.06
September 2003
DRAM Scrub Address Registers
These registers specify the next address to be scrubbed by the DRAM scrubber. They should be
initialized by the BIOS before the scrubber is enabled (the DRAM scrubber is enabled by writing a
valid scrub rate to DramScrub field, Function 3, Offset 58h) and after MemClrStatus, Function 2,
Offset 90h is set. They are then updated by the scrubber hardware as it scrubs successive 64-byte
blocks of DRAM. Once the scrubber reaches the DRAM limit address for the node (see “DRAM
Address Map” on page 57), it wraps to the DRAM base address.
The initial scrubbing address specified by these registers must be between the base and limit address
for the node, defined through the DRAM address maps (see “DRAM Address Map” on page 57). The
recommended initial scrubbing address is the DRAM base address.
In addition to sequential DRAM scrubbing, the DRAM scrubber has a redirect mode for scrubbing
DRAM locations accessed during normal operation. This is enabled by setting ScrubReDirEn
(Function 3, Offset 5Ch). When a DRAM read is generated by any agent other than the DRAM
scrubber, correctable ECC errors are corrected as the data is passed to the requestor, but the data in
DRAM is not corrected if redirect scrubbing mode is disabled. In scrubber redirect mode, correctable
errors detected during normal DRAM read accesses redirect the scrubber to the location of the error.
After the scrubber corrects the location in DRAM, it resumes scrubbing from where it left off. DRAM
scrub address registers are not modified by the redirect scrubbing mode. Sequential scrubbing and
scrubber redirection can be enabled independently or together.
ECC errors detected by the scrubber are logged in the MCA registers (see “Machine Check
Architecture Registers” on page 146).
DRAM Scrub Address Low Register
DRAM Scrub Address Low Register
Function 3: Offset 5Ch
31
6
5
1
reserved
ScrubAddrLo
Bits
Mnemonic
Function
R/W
Reset
31–6
ScrubAddrLo
DRAM Scrub Address Bits 31–6
R/W
X
5–1
reserved
R
0
0
ScrubReDirEn
R/W
0
DRAM Scrubber Redirect Enable
0
ScrubReDirEn
3.6.7.1
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
DRAM Scrubber Redirect Enable (ScrubReDirEn)—Bit 0. If this bit is set, the scrubber is
redirected to correct errors found during normal operation.
112
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
DRAM Scrub Address Bits 31–6 (ScrubAddrLo)—Bits 31–6. This field specifies the low address
bits 31–6 of the next 64-byte block to be scrubbed.
3.6.7.2
DRAM Scrub Address High Register
DRAM Scrub Address High Register
Function 3: Offset 60h
31
8
7
reserved
Bits
Mnemonic
31–8
reserved
7–0
ScrubAddrHi
0
ScrubAddrHi
Function
R/W
Reset
R
0
DRAM Scrub Address Bits 39–32
R/W
X
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
DRAM Scrub Address Bits 39–32 (ScrubAddrHi)—Bits 7–0. This field specifies the high address
bits 39–32 of the next 64-byte block to be scrubbed.
3.6.8
XBAR Flow Control Buffers
The Northbridge interfaces with the CPU core, DRAM controller, and, through three HyperTransport
links, to external chips.
The major Northbridge blocks are: System Request Interface (SRI), Memory Controller (MCT), and
Cross Bar (XBAR). SRI interfaces with the CPU core and connects coherent HyperTransport links
and noncoherent HyperTransport links. MCT maintains cache coherency and interfaces with the
DRAM. XBAR is a five port switch which routes the command packets between SRI, MCT, and the
three HyperTransport links. Not all HyperTransport links have to be active.
The number of buffers available for each link at the XBAR input is shown in Table 23.
Table 23. XBAR Input Buffers
Link
Number of
Command Buffers
Number of
Data Buffers
HyperTransport™ Link 0-2 16 x 3
8x3
SRI
10
5
MCT
12
8
Total
70
37
XBAR command and data buffers are independent. There are 70 command buffers available, but a
maximum of 64 can be used at any given time. Number of used data buffers is not restricted. The
Chapter 3
Memory System Configuration
113
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
default allocation of command buffers when all HyperTransport links are present is shown in
Table 24.
Table 24. Default XBAR Command Buffer Allocation
Link
Number of
Command Buffers
HyperTransport™ Link 0-2 15 x 3
SRI
8
MCT
11
Total
64
The HyperTransport™ I/O Link Specification defines four virtual channels: requests, posted requests,
responses, and probes. The default virtual channel command buffer allocation is shown in Table 25.
At least one command buffer should be allocated to each used virtual channel to avoid deadlock.
Command buffers do not need to be allocated for a virtual channel that is not used by a link. For
example, MCT does not initiate any requests, so there is no need to allocate request or posted request
command buffers for an MCT link. A link should not send transactions through a virtual channel that
does not have at least one command buffer allocated.
Default allocation of SRI buffers can be found in “SRI-to-XBAR Buffer Count Register” on page 115
and “Free List Buffer Count Register” on page 117. In the SRI-to-Xbar Buffer Count Register
(Function 3, Offset 70h), the virtual channels are subdivided into upstream (coherent HyperTransport)
and downstream (noncoherent HyperTransport) directions, since SRI must send packets into both
coherent HyperTransport and noncoherent HyperTransport links. Some buffers are allocated to a free
list pool to use the available resources more efficiently. The buffers in the free list pool are shared by
packets in multiple virtual channels. By default, no more than one buffer is allocated to a virtual
channel and the rest are allocated to the free list pool in the Free List Buffer Count Register (Function
3, Offset 7Ch). Any buffer increase to SRI can be added to the free list or allocated directly to a
virtual channel. The total count allocated through the Free List Buffer Count Register and the SRI-toXbar Buffer Count Register cannot exceed 10.
Table 25. Default Virtual Channel Command Buffer Allocation
Link
Request
Posted
Request
Response Probe
Number of
Command Buffers
Coherent HyperTransport™ Links
3
1
6
5
15
Non-coherent HyperTransport™ Links
6
5
4
0
15
SRI
2
3
3
0
8
MCT
0
0
8
3
11
Software can reallocate buffers by modifying buffer count registers (Function 0, Offsets 90h, B0h,
D0h, Function 3, Offsets 70h, 78h, 7Ch). The new buffer counts take effect after a warm reset. Since
hardware attempts to choose optimal settings, in general, these registers should not be changed.
114
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
The following information should be used for buffer reallocation:
1. Number of allocated command buffers for each link should not be greater than the number of
available command buffers for that link (see Table 23). Sum of allocated command buffers for three
HyperTransport links, SRI, and MCT should not be greater than 64.
2. If one HyperTransport link is not present, then other links can use all available buffers for each of
them (MCT and HyperTransport links can use one additional buffer, and SRI can use two additional
buffers).
3. If all links are present, some buffers from a noncoherent HyperTransport link can be reallocated to
coherent HyperTransport links. Table 26 shows command buffer allocation after two response buffers
from noncoherent HyperTransport link 0 are reallocated to coherent HyperTransport links 1 and 2.
Table 26. An Example of a Non Default Virtual Channel Command Buffer Allocation
Posted
Request
Response
Probe
Number of
Command Buffers
HyperTransport™ Link 0 6
5
2
0
13
HyperTransport™ Link 1 3
1
7
5
16
HyperTransport™ Link 2 3
1
7
5
16
SRI
2
3
3
0
8
MCT
0
0
8
3
11
Link
Request
4. If there are no noncoherent HyperTransport links, one ore more response buffers can be reallocated
from MCT to HyperTransport links.
5. In a multiprocessor system, number of coherent HyperTransport response buffers should be
increased first. The default buffer allocation is sufficient for a uniprocessor system.
SRI-to-XBAR Buffer Count Register
Function 3: Offset 70h
ReqD
10 9
reserved
8
URsp
7
6
5
4
Bits
Mnemonic
Function
R/W
Reset
31–30
DPReq
Downstream Posted Request Buffer Count
R/W
01b
29–28
DReq
Downstream Request Buffer Count
R/W
01b
27–26
reserved
R
0
R/W
01b
R
0
25–24
URspD
23–22
reserved
Chapter 3
Upstream Response Data Buffer Count
Memory System Configuration
3
2
reserved
reserved
DispRefReq
reserved
URspD
DReq
reserved
DPReq
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15
UPReq
SRI-to-XBAR Buffer Count Register
reserved
3.6.8.1
1
0
UReq
115
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
Bits
Mnemonic
Function
R/W
Reset
21:20
DispRefReq
Display Refresh Request Buffer Count
R/W
0
19–18
reserved
R
0
17–16
ReqD
Request Data Buffer Count
R/W
10b
15–10
reserved
R
0
9–8
URsp
Upstream Response Buffer Count
R/W
01b
7–6
reserved
R
0
5–4
UPReq
Upstream Posted Request Buffer Count
R/W
01b
3–2
reserved
R
0
1–0
UReq
Upstream Request Buffer Count
R/W
01b
September 2003
Field Descriptions
Upstream Request Buffer Count (UReq)—Bits 1–0. This field defines the number of upstream
request buffers available in the XBAR for use by the SRI.
Upstream Posted Request Buffer Count (UPReq)—Bits 5–4. This field defines the number of
upstream posted request buffers available in the XBAR for use by the SRI.
Upstream Response Buffer Count (URsp)—Bits 9–8. This field defines the number of upstream
response buffers available in the XBAR for use by the SRI.
Request Data Buffer Count (ReqD)—Bits 17–16. This field defines the number of request data
buffers available in the XBAR for use by the SRI.
Display Refresh Request Buffer Count (DispRefReq)—Bits 21-20. This field defines the number
of display refresh request buffers available in the XBAR for use by the SRI. See “Register
Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™ processors” on
page 19 for revision information about this field.
Upstream Response Data Buffer Count (URspD)—Bits 25–24. This field defines the number of
upstream response data buffers available in the XBAR for use by the SRI.
Downstream Request Buffer Count (DReq)—Bits 29–28. This field defines the number of
downstream request buffers available in the XBAR for use by the SRI.
Downstream Posted Request Buffer Count (DPReq)—Bits 31–30. This field defines the number of
downstream posted request buffers available in the XBAR for use by the SRI.
3.6.8.2
MCT-to-XBAR Buffer Count Register
MCT-to-XBAR Buffer Count Register
31
28 27
reserved
116
24 23
RspD
Function 3: Offset 78h
15 14
reserved
12 11
Prb
Memory System Configuration
8
Rsp
7
0
reserved
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bits
Mnemonic
Function
R/W
31–28
reserved
27–24
RspD
23–15
reserved
14–12
Reset
R
0
Response Data Buffer Count
R/W
8h
R
0
Prb
Probe Buffer Count
R/W
011b
11–8
Rsp
Response Buffer Count
R/W
8h
7–0
reserved
R
0
Field Descriptions
Response Buffer Count (Rsp)—Bits 11–8. This field defines the number of response buffers
available in the XBAR for use by the MCT.
Probe Buffer Count (Prb)—Bits 14–12. This field defines the number of probe buffers available in
the XBAR for use by the MCT.
Response Data Buffer Count (RspD)—Bits 27–24. This field defines the number of response data
buffers available in the XBAR for use by the MCT.
3.6.8.3
Free List Buffer Count Register
Free List Buffer Count Register
8
7
6
5
4
FReq
Function
R/W
Reset
R
0
SRI to XBAR Free Response Data Buffer Count
R/W
10b
R
0
SRI to XBAR Free Response Buffer Count
R/W
010b
R
0
FReq
SRI to XBAR Free Request Buffer Count
R/W
01b
FreeCmd
SRI Free Command Buffer Count
R/W
Bh
FRspD
reserved
14 13 12 11 10
reserved
Bits
Mnemonic
31–14
reserved
13–12
FRspD
11
reserved
10–8
FRsp
7–6
reserved
5–4
3–0
reserved
31
Function 3: Offset 7Ch
FRsp
3
0
FreeCmd
Field Descriptions
SRI Free Command Buffer Count (FreeCmd)—Bits 3–0. This field defines the number of free list
request or posted request buffers available in the SRI for use by the XBAR or CPU.
SRI to XBAR Free Request Buffer Count (FReq)—Bits 5–4. This field defines the number of free
list request buffers available in the XBAR for use by the SRI.
Chapter 3
Memory System Configuration
117
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
SRI to XBAR Free Response Buffer Count (FRsp)—Bits 10–8. This field defines the number of
free list response buffers available in the XBAR for use by the SRI.
SRI to XBAR Free Response Data Buffer Count (FRspD)—Bits 13–12. This field defines the
number of free list response data buffers available in the XBAR for use by the SRI.
3.6.9
XBAR-to-SRI Buffer Count Register
The Cross Bar Switch to System Request Interface (XBAR-to-SRI) Buffer Count register specifies
the number of command buffers for each virtual channel available in the SRI for use by the XBAR.
These buffers store commands for traffic routed to the SRI by the XBAR. The buffer counts take
effect on the next warm reset.
Since the XBAR must route packets in both the coherent HyperTransport technology fabric and down
the noncoherent HyperTransport chains, the usual virtual channels are further subdivided into
upstream (from noncoherent HyperTransport) and downstream (to noncoherent HyperTransport)
directions. In order to make more efficient use of the available resources, some of the buffers are
allocated onto a free list pool in which case these buffers can be used by packets issued in either
direction. See “Free List Buffer Count Register” on page 117. Any increase in the number of fixed
allocated buffers in the XBAR-to-SRI Buffer Count register must be compensated by a corresponding
reduction of the number of free list buffers specified in the Free List Buffer Count register.
When modifying buffer counts care must be taken to allocate a minimum number of buffers to each
virtual channel to avoid deadlock. Requests and Posted requests in both directions require at least one
buffer and Probes require at least two buffers to avoid deadlock. Since hardware attempts to choose
optimal settings, this register should not in general need to be changed.
Note: When the Probe Buffer Count is changed, care must be taken to avoid generation of system
management events from the time the new value is written into this register until it takes effect
(on the next warm reset).
DReq
23 22
reserved
20 19
DispRefReq
DPReq
31 30 29 28 27
Function 3: Offset 74h
16 15
reserved
12 11
Prb
7
reserved
6
4
UPReq
Bits
Mnemonic
Function
31–30
DPReq
Downstream Posted Request Buffer Count
R/W
01b
29–28
DReq
Downstream Request Buffer Count
R/W
01b
27–23
reserved
R
0
22:20
DispRefReq
Display Refresh Request Buffer Count
R/W
0
19–16
reserved
R
0
15–12
Prb
Probe Buffer Count
R/W
8h
118
Memory System Configuration
R/W
3
reserved
XBAR-to-SRI Buffer Count Register
2
0
UReq
Reset
Chapter 3
26094
Rev. 3.06
Bits
Mnemonic
11–7
reserved
6–4
UPReq
3
reserved
2–0
UReq
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Function
Upstream Posted Request Buffer Count
Upstream Request Buffer Count
R/W
Reset
R
0
R/W
001b
R
0
R/W
001b
Field Descriptions
Upstream Request Buffer Count (UReq)—Bits 2–0. This field defines the number of upstream
request buffers available in the SRI for use by the XBAR.
Upstream Posted Request Buffer Count (UPReq)—Bits 6–4. This field defines the number of
upstream posted request buffers available in the SRI for use by the XBAR.
Probe Buffer Count (Prb)—Bits 15–12. This field defines the number of probe buffers available in
the SRI for use by the XBAR.
Display Refresh Request Buffer Count (DispRefReq)—Bits 22-20. This field defines the number
of display refresh request buffers available in the SRI for use by the XBAR. See “Register
Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™ processors” on
page 19 for revision information about this field.
Downstream Request Buffer Count (DReq)—Bits 29–28. This field defines the number of
downstream request buffers available in the SRI for use by the XBAR.
Downstream Posted Request Buffer Count (DPReq)—Bits 31–30. This field defines the number of
downstream posted request buffers available in the SRI for use by the XBAR.
3.6.10
Display Refresh Flow Control Buffers
The Northbridge has a separate logical path from HyperTransport links to the system memory for
display refresh requests. This is necessary to support the bandwidth and latency requirements of
display refresh requests in systems where the display-refresh frame buffer is located in the system
memory.
A HyperTransport packet is defined as a display-refresh request packet when the read request has
isochronous bit, PassPW bit, and RspPassPW bit set, and coherent bit and SeqID cleared. All displayrefresh requests are marked as high priority requests and will have higher priority than other requests.
In order to accomplish a dedicated logical path to system memory a minimum number of buffers have
to be allocated in various queues.
DispRefReq (Function 3, Offset 70h) and DispRefReq (Function 3, Offset 74h) need to be set to
allocate the buffers for display refresh requests. At least one buffer should be allocated for display
refresh requests to avoid deadlock. Usually more than one buffer may be required to support the
necessary bandwidth. The buffer count written to the registers takes effect after a warm reset. See
“SRI-to-XBAR Buffer Count Register” on page 115 and “XBAR-to-SRI Buffer Count Register” on
page 118 for more information on setting these registers.
Chapter 3
Memory System Configuration
119
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
3.6.11
26094
Rev. 3.06
September 2003
Power Management Control Registers
These registers specify the processor response to STPCLK and HALT power management events.
Each STPCLK request received contains a 3-bit System Management Action Field (SMAF) which is
used as an index into the Power Management Control Low/High registers to select a Power
Management Mode (PMM) value to use for the duration of the STPCLK event. A SMAF value of 000
selects PMM0, a SMAF value of 001 selects PMM1, and so on. HALT is hardwired to use PMM7.
3.6.11.1
Power Management Mode (PMM) Value
The PMM value contains the following fields:
Bit
Name
Function
R/W
6–4
ClkSel
Clock Divisor Select
R/W
FID/VID Change Enable
R/W
3
reserved
2
FidVidEn
R
1
NBLowPwrEn
Northbridge Low Power Enable
R/W
0
CPULowPwrEn
CPU Low Power Enable
R/W
PMM Value Field Descriptions
Clock Divisor Select (ClkSel)—Bits 6–4. This field specifies the divisor to use when ramping down
the CPU clock or the Northbridge clock.
000b = Divide by 8
100b = Divide by 128
001b = Divide by 16
101b = Divide by 256
010b = Divide by 32
110b = Divide by 512
011b = Divide by 64
111b = reserved
FID/VID Change Enable (FidVidEn)—Bit 2. Enables a change in Frequency ID (FID) and/or
Voltage ID (VID). The CPU and the Northbridge clocks must also be ramped down (see
NBLowPwrEn and CPULowPwrEn).
0 = FID/VID change disabled
1 = FID/VID change enabled
Northbridge Low Power Enable (NBLowPwrEn)—Bit 1. Causes the Northbridge clock to ramp
down according to the clock divisor specified in ClkSel and puts the DRAM in self-refresh
mode. The CPU clock must also be ramped down (see CPULowPwrEn)
0 = Northbridge low power disabled
1 = Northbridge low power enabled
CPU Low Power Enable (CPULowPwrEn)—Bit 0. Causes the CPU clock to ramp down according
to the clock divisor specified in ClkSel.
0 = CPU low power disabled
1 = CPU low power enabled
120
Memory System Configuration
Chapter 3
Rev. 3.06
3.6.11.2
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Power Management Control Low Register
Power Management Control Low Register
Bits
Mnemonic
31
reserved
30–24
PMM3
23
reserved
22–16
PMM2
15
reserved
14–8
PMM1
7
reserved
6–0
PMM0
16 15 14
reserved
PMM3
reserved
24 23 22
reserved
31 30
Function 3: Offset 80h
PMM2
8
7
6
0
reserved
26094
PMM1
Function
Power Management Mode 3
Power Management Mode 2
Power Management Mode 1
Power Management Mode 0
PMM0
R/W
Reset
R
0
R/W
0
R
0
R/W
0
R
0
R/W
0
R
0
R/W
0
Field Descriptions
Power Management Mode 0 (PMM0)—Bits 6–0. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 1 (PMM1)—Bits 14–8. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 2 (PMM2)—Bits 22–16. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 3 (PMM3)—Bits 30–24. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Control High Register
Power Management Control High Register
Bits
Mnemonic
31
reserved
30–24
PMM7
23
reserved
22–16
PMM6
Chapter 3
16 15 14
PMM6
reserved
PMM7
reserved
24 23 22
reserved
31 30
Function 3: Offset 84h
8
PMM5
Function
Power Management Mode 7
Power Management Mode 6
Memory System Configuration
7
6
0
reserved
3.6.11.3
PMM4
R/W
Reset
R
0
R/W
0
R
0
R/W
0
121
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
Mnemonic
15
reserved
14–8
PMM5
7
reserved
6–0
PMM4
Function
Power Management Mode 5
Power Management Mode 4
26094
Rev. 3.06
R/W
Reset
R
0
R/W
0
R
0
R/W
0
September 2003
Field Descriptions
Power Management Mode 4 (PMM4)—Bits 6–0. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 5 (PMM5)—Bits 14–8. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 6 (PMM6)—Bits 22–16. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
Power Management Mode 7 (PMM7)—Bits 30–24. See “Power Management Mode (PMM) Value”
on page 120 for a description of these bits.
3.6.12
GART Aperture Control Register
This register contains the aperture size and enable for the graphics aperture relocation table (GART)
mechanism.
Bits
Mnemonic
31–7
reserved
6
5
4
3
1
0
GartSize
GartEn
reserved
6
DisGartCpu
7
DisGartIo
31
Function 3: Offset 90h
DisGartTblWlkPrb
GART Aperture Control Register
Function
R/W
Reset
R
0
DisGartTblWlkPrb
Disable GART Table Walk Probes
R/W
0
5
DisGartIo
Disable GART I/O Accesses
R/W
0
4
DisGartCpu
Disable GART CPU Accesses
R/W
0
3–1
GartSize
GART Size
R/W
0
0
GartEn
GART Enable
R/W
0
122
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Field Descriptions
GART Enable (GartEn)—Bit 0. Enables GART address translation for accesses falling within the
GART aperture. GartAperBaseAddr (Function 3, Offset 94h) and other related registers
should be initialized before GartEn is set.
GART Size (GartSize)—Bits 3–1. Defines the virtual address space to be allocated to the GART.
000b = 32 Mbytes
001b = 64 Mbytes
010b = 128 Mbytes
011b = 256 Mbytes
100b = 512 Mbytes
101b = 1 Gbyte
110b = 2 Gbyte
111b = reserved
Disable GART CPU Accesses (DisGartCpu)—Bit 4. Disables requests from CPUs from accessing
the GART.
Disable GART I/O Accesses (DisGartIo)—Bit 5. Disables requests from I/O devices from
accessing the GART.
Disable GART Table Walk Probes (DisGartTblWlkPrb)—Bit 6. Disables generation of probes for
GART table walks. This bit may be set to improve performance in cases where the GART
table entries are in address space which is marked uncacheable in processor MTRRs or page
tables. See “Register Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™
processors” on page 19 for revision information about this field.
3.6.13
GART Aperture Base Register
This register contains the base address of the aperture for the graphics aperture relocation table
(GART). The GART aperture base along with the GART aperture size define the GART aperture
address window. BIOS can place the GART aperture below the 4-gigabyte level in address space in
order to support legacy operating systems and legacy AGP cards (that do not support 64-bit address
space).
GART Aperture Base Register
31
Function 3: Offset 94h
15 14
reserved
0
GartAperBaseAddr[39:25]
Bits
Mnemonic
Function
31–15
reserved
14–0
GartAperBaseAddr[39:25] GART Aperture Base Address Bits 39–25
R/W
Reset
R
0
R/W
X
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Chapter 3
Memory System Configuration
123
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
GART Aperture Base Address Bits 39–25 (GartAperBaseAddr[39:25])—Bits 14–0. These bits
are used to compare to the incoming addresses to determine if they are in the aperture range.
They are active based on the aperture size. The remaining address bits are assumed to be 0.
[39:25] = 32 Mbytes
[39:26] = 64 Mbytes
[39:27] = 128 Mbytes
[39:28] = 256 Mbytes
[39:29] = 512 Mbytes
[39:30] = 1 Gbytes
[39:31] = 2 Gbytes
3.6.14
GART Table Base Register
This register contains the base address of the translation table for the graphics aperture relocation
table (GART). The GART table base points to the base address of a table of 32-bit GART page table
entries (PTEs) that contain the physical addresses to use for an incoming address that falls within the
GART aperture.
GART Table Base Register
Function 3: Offset 98h
31
4
GartTblBaseAddr[39:12]
Bits
3
0
reserved
Mnemonic
Function
R/W
Reset
31–4
GartTblBaseAddr[39:12]
GART Table Base Address Bits 39–12
R/W
X
3–0
reserved
R
0
“X” in the Reset column indicates that the field initializes to an undefined state after reset.
Field Descriptions
GART Table Base Address Bits 39–12 (GartTblBaseAddr[39:12])—Bits 31–4. These bits point to
the base of the table of GART PTEs to be used for GART address translation. Table 27 shows
the organization of each 32-bit PTE in the GART table.
Table 27. GART PTE Organization
Bits
Description
0
Valid
1
Coherent
3–2
reserved
11–4
PhysAddr[39:32]
124
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 27. GART PTE Organization
Bits
Description
31–12
PhysAddr[31:12]
3.6.15
GART Cache Control Register
This register controls the GART cache, which is used to cache the most recently translated GART
aperture accesses. The GART cache is enabled automatically whenever the GART aperture is
enabled.
31
2
reserved
Bits
Mnemonic
Function
R/W
Reset
31–2
reserved
1
GartPteErr
GART PTE Error
R
0
R/WC
0
0
InvGart
Invalidate GART
R/W
0
1
0
InvGart
Function 3: Offset 9Ch
GartPteErr
GART Cache Control Register
Field Descriptions
Invalidate GART (InvGart)—Bit 0. Setting this bit causes the GART cache to be invalidated. This
bit is cleared by hardware when the invalidation is complete.
GART PTE Error (GartPteErr)—Bit 1. This bit is set when an invalid PTE is encountered during a
table walk. Cleared by writing a 1.
3.6.16
Clock Power/Timing Low Register
This register controls the transition times between various voltage and frequency states. The value of
this register is maintained through a warm reset and is initialized to 0 on a cold reset.
Chapter 3
reserved
Memory System Configuration
8
7
6
4
3
reserved
LClkPLLLock (19–4)
11 10
reserved
16 15
ClkRampHyst
31
Function 3: Offset D4h
reserved
Clock Power/Timing Low Register
2
0
GPE
125
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Bits
26094
Rev. 3.06
Mnemonic
Function
R/W
31–16
LClkPLLLock
HyperTransport CLK PLL Lock Counter Bits 19–4
R/W
0
15–11
reserved
R
0
10-8
ClkRampHyst
7
reserved
6–4
reserved
3
reserved
2–0
GPE
Clock Ramp Hysteresis
SBZ
Good Phase Error
September 2003
Reset
R/W
0
R
0
R/W
0
R
0
R/W
0
Field Descriptions
Good Phase Error (GPE)—Bits 2–0. This field defines the BCLK PLL time until good phase error.
Counting occurs when at full frequency.
Revision B and earlier revision encodings are:
000b = 16 system clocks
001b = 400 system clocks
010b = 800 system clocks
011b = 1200 system clocks
100b = 1600 system clocks
101b = 2000 system clocks
110b = 2400 system clocks
111b = 4095 system clocks
Revision C encodings are:
000b = reserved
001b = 200 system clocks (1us)
010b = 400 system clocks (2us)
011b = 600 system clocks (3us)
100b = 800 system clocks (4us)
101b = 1600 system clocks (8us)
110b = 3000 system clocks (15us)
111b = 20000 system clocks (100us)
Clock Ramp Hysteresis (ClkRampHyst)—Bits 10-8. A non-zero value in this field enables a
hysteresis time which prevents the CPU clock grid from being ramped down after processing
a probe. It avoids unnecessary changes of the CPU clock grid when the probe arrival rate is
relatively low. See “Register Differences in Revisions of the AMD Athlon™ 64 and AMD
Opteron™ processors” on page 19 for revision information about this field.
000b = 0
001b = 125 ns
010b = 250 ns
011b = 375 ns
100b = 500 ns
126
Memory System Configuration
Chapter 3
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
101b = 750 ns
110b = 1000 ns
111b = 2000 ns
HyperTransport CLK PLL Lock Counter Bits 19–4 (LClkPLLLock)—Bits 31–16. This bit field
indicates how long it takes for the slowest HyperTransport technology clock PLL to ramp to
its new frequency and lock.
The 16 bits represent the most significant 16 bits of a 20-bit value. The number reflects the
number of system bus clocks (5 ns) to wait. It is only necessary to wait a short amount of time
in the event the frequency change is one that maintains the VCO frequency. In this case, the
PLL will be re-locked quickly.
3.6.17
Clock Power/Timing High Register
This register controls the transition times between various voltage and frequency states. The value of
this register is maintained through a warm reset and is initialized to 0 on a cold reset.
Clock Power/Timing High Register
28 27
RampVIDOff
reserved
31 30
Bits
Mnemonic
31
reserved
30–28
RampVIDOff
27–20
reserved
19–0
VSTime
Function 3: Offset D8h
20 19
reserved
0
VSTime
Function
R/W
Reset
R
0
Ramp VID Offset
R/W
0
R
0
R/W
0
Voltage Regulator Stabilization Time
Field Descriptions
Voltage Regulator Stabilization Time (VSTime)—Bits 19–0. This field indicates when the voltage
regulator is stable at the Ramp VID before ramping up the clocks. The count is the number of
5-ns system clock cycles.
Ramp VID Offset (RampVIDOff)—Bits 30–28. Defines the amount of extra voltage required while
the PLL is ramping for low power states. This “over-voltage” is applied only until the PLL has
phase locked to within specifications.
000b = 0 mV
001b = 25 mV
010b = 50 mV
011b = 75 mV
100b = 100 mV
Chapter 3
Memory System Configuration
127
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
101b = 125 mV
110b = 150 mV
111b = 175 mV
3.6.18
HyperTransport™ FIFO Read Pointer Optimization Register
This register allows the separation of read/write pointers in the HyperTransport technology receive/
transmit FIFOs to be changed from their default settings. The value of this register is maintained
through a warm reset and is initialized to 0 on a cold reset.
HyperTransport™ FIFO Read Pointer Optimization Register
Function 3: Offset DCh
31
8
Bits
Mnemonic
31–22
reserved
21–20
XmtRdPtrLdt2
19
reserved
18–16
RcvRdPtrLdt2
15–14
reserved
13–12
XmtRdPtrLdt1
11
reserved
10–8
RcvRdPtrLdt1
7–6
reserved
5–4
XmtRdPtrLdt0
3
reserved
2–0
RcvRdPtrLdt0
Function
4
R/W
Reset
R
0
R/W
0
R
0
Change Read Pointer For HyperTransport Link 2
Receiver
R/W
0
R
0
Change Read Pointer For HyperTransport Link 1
Transmitter
R/W
0
R
0
Change Read Pointer For HyperTransport Link 1
Receiver
R/W
0
R
0
R/W
0
R
0
R/W
0
Change Read Pointer For HyperTransport Link 2
Transmitter
Change Read Pointer For HyperTransport Link 0
Transmitter
Change Read Pointer For HyperTransport Link 0
Receiver
3
2
0
RcvRdPtrLdt0
5
reserved
6
XmtRdPtrLdt0
7
reserved
RcvRdPtrLdt1
reserved
XmtRdPtrLdt1
reserved
16 15 14 13 12 11 10
RcvRdPtrLdt2
reserved
reserved
XmtRdPtrLdt2
22 21 20 19 18
Field Descriptions
Change Read Pointer For HyperTransport Link 0 Receiver (RcvRdPtrLdt0)—Bits 2–0. See
RcvRdPtrLdt2 for values.
Change Read Pointer For HyperTransport Link 0 Transmitter (XmtRdPtrLdt0)—Bits 5–4. See
XmtRdPtrLdt2 for values.
128
Memory System Configuration
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Change Read Pointer For HyperTransport Link 1 Receiver (RcvRdPtrLdt1)—Bits 10–8. See
RcvRdPtrLdt2 for values.
Change Read Pointer For HyperTransport Link 1 Transmitter (XmtRdPtrLdt1)—Bits 13–12.
See XmtRdPtrLdt2 for values.
Change Read Pointer For HyperTransport Link 2 Receiver (RcvRdPtrLdt2)—Bits 18–16.
Moves the read pointer for the HyperTransport receive FIFO closer to the write pointer
thereby reducing latency through the receiver.
000b = RdPtr assigned by hardware
001b = Move RdPtr closer to WrPtr by 1 HyperTransport clock period
010b = Move RdPtr closer to WrPtr by 2 HyperTransport clock periods
011b = Move RdPtr closer to WrPtr by 3 HyperTransport clock periods
100b = Move RdPtr closer to WrPtr by 4 HyperTransport clock periods
101b = Move RdPtr closer to WrPtr by 5 HyperTransport clock periods
110b = Move RdPtr closer to WrPtr by 6 HyperTransport clock periods
111b = Move RdPtr closer to WrPtr by 7 HyperTransport clock periods
AMD recommends setting this field to 5 for all coherent HyperTransport links and
noncoherent HyperTransport links to AMD chipsets. Optimal value for noncoherent
HyperTransport links to other chipsets needs to be determined by the developer and tested to
ensure stability.
Change Read Pointer For HyperTransport Link 2 Transmitter (XmtRdPtrLdt2)—Bits 21–20.
Moves the read pointer for the HyperTransport technology transmit FIFO closer to the write
pointer thereby reducing latency through the transmitter.
00b = RdPtr assigned by hardware
01b = Move RdPtr closer to WrPtr by 1 HyperTransport clock period
10b = Move RdPtr closer to WrPtr by 2 HyperTransport clock periods
11b = Move RdPtr closer to WrPtr by 3 HyperTransport clock periods
AMD recommends setting this field to 2 for all coherent HyperTransport links and
noncoherent HyperTransport links to AMD chipsets. Optimal value for noncoherent
HyperTransport links to other chipsets needs to be determined by the developer and tested to
ensure stability.
3.6.19
Thermtrip Status Register
The Thermtrip Status register provides status information regarding the THERMTRIP thermal sensor.
Chapter 3
Memory System Configuration
129
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Bits
Mnemonic
Function
31
SwThermtp
Software Thermtrip
30–14
reserved
13–8
DiodeOffset
7-6
reserved
5
ThermtpEn
4
reserved
3
ThermtpSense
2
reserved
1
Thermtp
0
reserved
Diode Offset
W
0
R
0
1
0
Thermtp
2
R
0
R
0
R
R
Thermtrip
3
Reset
R
Thermtrip Sense
4
R/W
R
Thermtrip Enabled
5
reserved
6
reserved
DiodeOffset
7
ThermtpSense
reserved
8
reserved
14 13
ThermtpEn
31 30
SwThermtp
September 2003
Function 3: Offset E4h
reserved
Thermtrip Status Register
Rev. 3.06
0
R
R
0
Field Descriptions
Thermtrip (Thermtp)—Bit 1. Set to 1 if a temperature sensor trip occurs and was enabled.
Thermtrip Sense (ThermtpSense)—Bit 3. Set to 1 if a temperature sensor trip occurs.
Thermtrip Enabled (ThermtpEn)—Bit 5. Indicates that the thermtrip temperature sensor is
enabled. When this bit is set to 1, a THERMTRIP High event will cause the hardware to shut
down the PLL, assert the THERMTRIP output pin and set the ThermtpHi bit. The
ThermtpSense bit is set for a THERMTRIP High event, irrespective of the state of
ThermtpEn.
Diode Offset (DiodeOffset[5:0])—Bits 13–8.Thermal diode offset is used to correct temperature
measurement made by an external temperature sensor. The offset is in 1 degree Celsius
increments and it should be subtracted from the temperature measurement. A correction to the
offset may be needed for some temperature sensors. Contact the temperature sensor vendor to
determine whether an offset correction is needed. The maximum allowable offset is provided
in the appropriate processor data sheet, and the maximum offset can vary for different
processors.
Software Thermtrip (SwThermtp)—Bit 31. Writing a 1 to this bit position induces a THERMTRIP
event. This bit is write-only and returns 0 when read. This is a diagnostic bit, and it should be
used for testing purposes only.
130
Memory System Configuration
Chapter 3
26094
Rev. 3.06
3.6.20
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Northbridge Capabilities Register
The Northbridge Capabilities register indicates whether or not this Northbridge is capable of certain
behavior.
Function
Bits
Mnemonic
31–9
reserved
8
MemCntCap
7
reserved
6–5
DramFreq
4
ChipKillEccCap
Chip-Kill ECC Capable
R
3
EccCap
ECC Capable
R
2
BigMPCap
Big MP Capable
R
1
MPCap
MP Capable
R
0
128BitCap
128-Bit DRAM Capable
R
Memory Controller Capable
3
2
1
MPCap
Reset
R
0
0
128BitCap
4
BigMPCap
5
R/W
R
R
Maximum DRAM Frequency
6
EccCap
7
ChipKillEccCap
reserved
8
DramFreq
9
reserved
31
Function 3: Offset E8h
MemCntCap
Northbridge Capabilities Register
0
R
Field Descriptions
128-Bit DRAM Capable (128BitCap)—Bit 0. This bit is set to 1 if the Northbridge is capable of
supporting a 128-bit DRAM interface.
MP Capable (MPCap)—Bit 1. This bit is set to 1 if the Northbridge is capable of supporting
multiprocessor systems.
Big MP Capable (BigMPCap)—Bit 2. This bit is set to 1 if the Northbridge is capable of supporting
multiprocessor systems greater than DP.
ECC Capable (EccCap)—Bit 3. This bit is set to 1 if the Northbridge is capable of supporting ECC.
Chip-Kill ECC Capable (ChipKillEccCap)—Bit 4. This bit is set to 1 if the Northbridge is capable
of supporting chip-kill ECC.
Maximum DRAM Frequency (DramFreq)—Bits 6–5. Indicates the maximum DRAM frequency
supported.
00b = No limit
01b = 166 MHz
10b = 133 MHz
11b = 100 MHz
Chapter 3
Memory System Configuration
131
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Memory Controller Capable (MemCntCap)—Bit 8. This bit is set to 1 if the Northbridge is
capable of supporting an on-chip memory controller.
132
Memory System Configuration
Chapter 3
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
4
DRAM Configuration
4.1
Programming Interface
This section describes how to program various DRAM controller registers. There are two principal
areas of conﬁguration:
•
Conﬁguration state information obtained via the DIMM Serial Presence Detect (SPD) ROMs.
•
All other conﬁguration state information.
There are certain restrictions for 3 DIMM population in AMD Athlon™ 64 desktop platforms that
must be enforced by the BIOS. Please refer to the Unbuffered DIMM Support table in the AMD
Athlon™ 64 Data Sheet, order# 24659, for allowable DIMM loading.
4.1.1
SPD ROM-Based Configuration
The SPD device is an EEPROM on the DIMM encoded by the DIMM manufacturer. The description
of the EEPROM is usually provided on a data sheet for the DIMM itself along with data describing
the memory devices used. The data describes conﬁguration and speed characteristics of the DIMM
and the SDRAM components mounted on the DIMM. The data sheet also contains the DIMM byte
values that are encoded in the SPD on the DIMM.
BIOS acquires the values encoded in the SPD ROM through the I/O hub, which obtains the
information through a secondary device connected to the I/O hub through the SMBus. This secondary
device communicates with the DIMM by means of the I2C bus.
The SPD ROM provides values for several DRAM timing parameters that are required by the DRAM
controller. These parameters are:
•
tCL: (CAS latency)
•
tRC: Active-to-Active/Auto Refresh command period
•
tRFC: Auto-Refresh-to-Active/Auto Refresh command period
•
tRCD: Active-to-Read-or-Write delay
•
tRRD: Active-Bank-A to-Active-Bank-B delay
•
tRAS: Active-to-Precharge delay
•
tRP: Precharge time
•
tREF: Refresh interval (function of DRAM density)
Chapter 4
DRAM Conﬁguration
133
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
4.1.1.1
26094
Rev. 3.06
September 2003
tCL (CAS Latency)
The number of memory clocks it takes a DRAM to return data after the read CAS_L is asserted
depends on the memory clock frequency. The value that BIOS programs into the memory controller is
a function of the target clock frequency. The target clock frequency is determined from the supported
CAS latencies at given clock frequencies of each DIMM. A suggested algorithm is as follows:
1. Determine all CAS latencies supported by each installed DIMM, deﬁned in SPD byte 18. One bit
corresponds to each supported CAS latency. SPD byte 18 speciﬁes CAS latencies with which
devices on the DIMM can reliably operate. BIOS should not assume that a device can operate
with either slower or faster CAS latency than those speciﬁed by the SPD. Typically, all DDR
DIMMs support CAS latencies of 2 and 2.5. Some DDR DIMMs may support CAS latencies of 3.
The DRAM controller is designed to support these CAS latencies. The SPD ROM allows the
manufacturer to indicate support for CAS latency 3.5. This value is not supported by the DRAM
controller and should be discarded.
2. Determine the maximum clock frequency at each supported CAS latency for each DIMM. The
minimum cycle time for the highest, the second highest, and the third highest supported CAS
latencies is deﬁned in SPD byte 9, 23, and 25, respectively. There is a possibility of incorrectly
programmed SPDs such that a cycle time from SPD byte 23 or 25, corresponding to a set bit in
SPD byte 18, is unimplemented. BIOS should discard this pair. The minimum cycle time has
1/10ns granularity.
3. Determine the best CAS latency and clock frequency combination. Find the highest clock
frequency supported by the slowest DIMM and determine the CAS latency at that operating
frequency. It is necessary to choose the highest CAS latency supported by all the DIMMs at the
target frequency.
The processor also provides silicon fuses that limit the frequency of the DRAM clock. BIOS can read
this information from the Northbridge Capabilities register (function 3, offset E8h[6:5]). BIOS needs
to assume that the highest supported frequency is given by the slowest DIMM or the Maximum
DRAM Frequency fuses, whichever is lower.
There is a possibility that there are two options: a 166 MHz tCL = 3 conﬁguration or a 133 MHz tCL
= 2 conﬁguration. The above algorithm would result in selecting the 166 MHz conﬁguration, but it is
known through performance analysis that this is not the best choice. Whenever tCL for the higher
frequency is 1 greater than tCL for the next lowest frequency, BIOS should select the lower
frequency. If the tCL difference is 0.5, then the higher frequency should be selected.
4.1.1.2
tRCD (RAS-to-CAS Delay)
This parameter is deﬁned in SPD byte 29 and it has 1/4ns granularity. BIOS should read and convert
this value into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 48h for DDR333, which maps to 18 ns or 3 clocks.
134
DRAM Conﬁguration
Chapter 4
26094
Rev. 3.06
4.1.1.3
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
tRAS (Active-to-Precharge Delay)
This parameter is deﬁned in SPD byte 30 and it has 1-ns granularity. BIOS should read and convert
this value into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 2Ah for DDR333, which maps to 42 ns or 7 clocks.
4.1.1.4
tRP (Precharge Command Period)
This parameter is deﬁned in SPD byte 27 and it has 1/4-ns granularity. BIOS should read and convert
this value into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 48h for DDR333, which maps to 18 ns or 3 clocks.
4.1.1.5
tRC (Active-to-Active/Auto-Refresh Command Period)
This parameter is deﬁned in SPD byte 41 and it has 1-ns granularity. BIOS should read and convert
this value into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 3Ch for DDR333, which maps to 60 ns or 10 clocks.
Some DIMMs may not include the tRC value in byte 41. In this case, BIOS will read either 00h or
FFh from the ROM and it should do the following:
•
At 100 MHz, tRC is assumed to be 46h, which maps to 70 ns or 7 clocks.
•
At 133 MHz, tRC is assumed to be 41h, which maps to 65 ns or 9 clocks.
•
At 166 MHZ, tRC is assumed to be 3Ch, which maps to 60 ns or 10 clocks.
•
At 200 MHZ, tRC is assumed to be 37h, which maps to 55 ns or 11 clocks.
4.1.1.6
tRRD (Active-to-Active of a Different Bank)
This parameter is deﬁned in SPD byte 28 and it has 1/4-ns granularity. BIOS should read this value
and convert into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 30h for DDR333, which maps to 12 ns or 2 clocks.
4.1.1.7
tRFC (Auto-Refresh-to-Active/Auto-Refresh Command Period)
This parameter is deﬁned in SPD byte 42 and it has 1-ns granularity. BIOS should read and convert
this value into a number of DRAM clocks using the target frequency obtained in “tCL (CAS
Latency)” on page 134. Typically, this value is 48h for DDR333 DRAMs between 64-Mbit and 512Mbit and 78h for DDR333 1-Gbit DRAMS.
Some DIMMs may not include the tRFC value in byte 42. In this case, BIOS will read either 00h or
FFh from the ROM and it should do the following:
•
At 100 MHz, tRFC is assumed to be 50h, which maps to 80 ns or 8 clocks.
•
At 133 MHz, tRFC is assumed to be 4Bh, which maps to 75 ns or 10 clocks.
Chapter 4
DRAM Conﬁguration
135
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
•
At 166 MHz, tRFC is assumed to be 48h, which maps to 72 ns or 12 clocks for 512-Mbit devices
or smaller. For 1-Gbit devices, it is assumed to be 78h, which maps to 120 ns or 20 clocks.
•
At 200 MHz, tRFC is assumed to be 46h, which maps to 70 ns or 14 clocks for 512-Mbit or
smaller devices. For 1-Gbit devices, it is assumed to be 78h, which maps to 120 ns or 24 clocks.
4.1.1.8
tREF (Refresh Rate)
The tREF is not an SPD ROM parameter but a ﬁeld that the DRAM controller requires and that can
be obtained from several bytes in the SPD ROM. Typically, the entire DRAM must have every row
refreshed at least once every 64 ms. The average time between two row refreshes is thus based on the
number of rows in the DRAM and the clock frequency.
The clock frequency was previously derived. The number of rows in the implementation can be read
from SPD byte 3. This provides the number of row address bits for each of the two possible chip
selects on a DIMM. (Note that each chip-select range of a two chip-select DIMM uses devices that
have the same number of rows). A design with 12 row address bits implies 4k rows per internal device
chip select. This maps to a refresh every 15.6 µs on average. There are also 8k (13 row address bits)
and 16k (14 row address bits) row devices. By knowing the number of rows and the frequency, the
DRAM controller conﬁguration register can be correctly programmed. All 4k row devices require an
average refresh interval of 15.6 µs and all 8k and 16k row devices require a 7.8 µs average refresh
interval.
4.1.1.9
Registered or Unbuffered DIMMs
SPD byte 21, bit 1 indicates whether the DIMM registers the address and commands or not. Refer to
the processor data sheet to determine the type of DIMMs supported. Systems with mixed DIMMs
(unbuffered and registered) are not supported.
4.1.1.10
DIMM Chip Select Density
The size of a DIMM module chip-select range can be obtained directly from the SPD ROM.
SPD byte 5 indicates whether the DIMM is composed of one or two chip selects. Byte 31 indicates
the size of the chip-select range or ranges on the DIMM (from 32 Mbyte to 2 Gbyte). See the SPD
ROM speciﬁcation for the encoding.
Note: If the DIMM uses two chip selects, then each chip-select range is the same size.
4.1.1.11
DIMM ECC Enable
SPD byte 11 indicates whether the DIMM supports ECC bits. A value of 02h indicates that ECC is
supported for all chip-select banks on the DIMM.
4.1.1.12
x4 DIMMs
The DRAM device width is largely inconsequential to the controller except when they are x4 devices.
The width of the devices on the DIMM is determined by the value of SPD byte 13. If the DIMM
136
DRAM Conﬁguration
Chapter 4
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
implements two chip-select banks, then each chip-select bank uses devices of the same width. See the
SPD ROM speciﬁcation for the encoding.
4.1.2
Non-SPD ROM-Based Configuration
Many other bit ﬁelds are also required by the DRAM controller conﬁguration registers, but these
values cannot be obtained from the SPD ROM. In many cases, these values will be hardcoded into the
BIOS, but in others they are functions of other bit values. This section describes how BIOS programs
each non-SPD related ﬁeld that is not hardcoded.
4.1.2.1
Twr (Write Recovery)
This is a true DRAM device timing parameter but it is not included in the SPD ROM. Therefore, it
will be hardcoded based on the DDR200, DDR266, DDR333, and DDR400 AC speciﬁcation, as
follows:
•
DDR200, DDR266: Twr is 2 clocks.
•
DDR333, DDR400: Twr is 3 clocks.
4.1.2.2
Twtr (Write to Read Delay)
This parameter has the following values:
•
DDR200, DDR266, DDR333: Twtr is 1 clock.
•
DDR400: Twtr is 2 clocks.
4.1.2.3
Trwt
This ﬁeld ensures read-to-write data-bus turnaround. This ﬁeld is a function of CAS latency and
clock frequency. It is also a function of the asynchronous round-trip loop delay for each of the
systems shown in Table 28.
Table 28.
Trwt Values
Number of Clocks
CAS Latency
System
200 MHz
166 MHz
133 MHz
100 MHz
CL = 2
128-bit interface
N/A
3
2
2
64-bit interface (registered DIMMS)
N/A
2
2
2
64-bit interface (unbuffered DIMMS)
3
3
3
3
128-bit interface
N/A
3
3
3
64-bit interface (registered DIMMS)
N/A
3
3
3
64-bit interface (unbuffered DIMMS)
4
4
4
4
CL = 2.5
Chapter 4
DRAM Conﬁguration
137
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 28.
26094
Rev. 3.06
September 2003
Trwt Values (Continued)
Number of Clocks
CAS Latency
System
200 MHz
166 MHz
133 MHz
100 MHz
CL = 3
128-bit interface
3
4
3
3
64-bit interface (registered DIMMS)
3
3
3
3
64-bit interface (unbuffered DIMMS)
4
4
4
4
4.1.2.4
Twcl (Write CAS Latency)
If the DIMM is registered, the write CAS latency is two clocks. If the DIMM is unbuffered, then write
CAS latency is one clock.
4.1.2.5
Maximum Asynchronous Latency
AsyncLat ﬁeld should be set to 9 ns for registered DIMM systems with DDR266 8-DIMM
conﬁguration, and to 8 ns for all other registered DIMM systems. AsyncLat ﬁeld should be set to 7 ns
for unbuffered systems with 3 DIMMs, and to 6 ns for all other unbuffered DIMM systems.
4.1.2.6
Read Preamble Time
The Read Preamble time is a function of the asynchronous round trip loop latency, the clock
frequency, DIMM type and process type.
Table 29.
RdPreamble Values
200 MHz
166 MHz
133 MHz
100 MHz
Registered DIMM slots
7 ns
7.5 ns
8 ns
9 ns
Unbuffered 1 or 2 DIMM slots1
5 ns
6 ns
7 ns
9 ns
Unbuffered 3 DIMM slots1
5.5 ns
6.5 ns
7.5 ns
9 ns
Notes:
1. The values for RdPreamble must be set based on the number of DIMM slots, independent of the
number of DIMMs actually populated. This allows the memory controller to compensate for worstcase round-trip delay from DIMMs in the farthest slot.
4.1.2.7
Memory Clock Enable
DRAM clocks must be enabled by BIOS. There are four clock enable bits which correspond on a oneto-one basis with the DIMMs (i.e., if DIMM0 is populated, memory clock 0 must be enabled).
DRAM clocks are grouped according to the requirements for unbuffered or registered DIMMs. For
example, when MC0_EN is set to enable DIMM 0 in an unbuffered DIMMs conﬁguration, the
memory controller will drive three clock pairs to DIMM 0. In a registered DIMM conﬁguration a
single clock pair is driven to each enabled DIMM.
138
DRAM Conﬁguration
Chapter 4
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
In DIMM conﬁgurations that have two DIMM sockets connected to the same command/address bus
(as in SODIMM conﬁgurations), but have different clock pairs routed to each of the DIMMs, BIOS
must enable both sets of clocks even if the DIMM 0 slot is unpopulated. For example, in an SODIMM
conﬁguration with two DIMM slots, BIOS should set both MC0_EN and MC1_EN even if only
DIMM 1 is populated. This is required in order to guarantee proper operation.
4.1.3
DRAM Initialization
Some DRAM conﬁguration ﬁelds must be programmed in a speciﬁc order. BIOS must set
DramConﬁgRegLo[8] last to start DRAM initialization after a cold reset not associated with suspend
to RAM mode. BIOS must set DramConﬁgRegLo[12] and DramConﬁgRegLo[13] last to exit selfrefresh mode after a cold reset associated with suspend to RAM mode. BIOS should not assert
LDTSTOP_L to change HyperTransport™ link width and frequency while DramConﬁgRegLo[8] or
DramConﬁgRegLo[12] are set.
For Revision C, BIOS must guarantee a delay between setting DramConﬁgRegHi[25] and setting
DramConﬁgRegLo[12] and DramConﬁgRegLo[13] when resuming from suspend to RAM mode.
The required delay is 110us for registered DIMMs and 10us for unbuffered DIMMs. Both DIMM
types require 10us for memory clock stabilization time. Additional 100us required for registered
DIMMs is for DIMM PLL stabilization time.
4.2
DRAM Configuration
This section shows in a quick reference format how each DRAM controller conﬁguration register
should be programmed.
BaseAddrReg[7:0][31:0] = See “DRAM CS Base Address Registers” on page 69.
BaseMaskReg[7:0][31:0] = See “DRAM CS Mask Registers” on page 72.
BankAddrReg[31:0] = See “DRAM Bank Address Mapping Register” on page 73.
DramTimingRegLo[28] = tWR = See “Twr (Write Recovery)” on page 137.
DramTimingRegLo[26:24] = tRP[2:0] = See “tRP (Precharge Command Period)” on page 135.
DramTimingRegLo[23:20] = tRAS[3:0] = See “tRAS (Active-to-Precharge Delay)” on page 135.
DramTimingRegLo[18:16] = tRRD[2:0] = See “tRRD (Active-to-Active of a Different Bank)” on
page 135.
DramTimingRegLo[14:12] = tRCD[2:0] = See “tRCD (RAS-to-CAS Delay)” on page 134.
DramTimingRegLo[11:8] = tRFC[3:0] = See “tRFC (Auto-Refresh-to-Active/Auto-Refresh
Command Period)” on page 135.
DramTimingRegLo[7:4] = tRC[3:0] = See “tRC (Active-to-Active/Auto-Refresh Command Period)”
on page 135.
Chapter 4
DRAM Conﬁguration
139
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
DramTimingRegLo[2:0] = tCL[2:0] = See “tCL (CAS Latency)” on page 134.
DramTimingRegHi[22:20] = tWCL[2:0] = See “Twcl (Write CAS Latency)” on page 138.
DramTimingRegHi[12:8] = tREF[4:0] = See “tREF (Refresh Rate)” on page 136.
DramTimingRegHi[6:4] = tRWT[2:0] = See “Trwt” on page 137.
DramTimingRegHi[0] = tWTR = See “Twtr (Write to Read Delay)” on page 137.
DramConﬁgRegLo[27:25] = BypMax[2:0] = 4h
DramConﬁgRegLo[24] = DisInRcvrs = 0h1
DramConﬁgRegLo[23:20] = x4DIMMs[3:0] = See “x4 DIMMs” on page 136.
DramConﬁgRegLo[19] = 32ByteEn = 0h2
DramConﬁgRegLo[18] = UnBuffDimm = See “Registered or Unbuffered DIMMs” on page 136.
DramConﬁgRegLo[17] = DimmEccEn = See “DIMM ECC Enable,”.
DramConﬁgRegLo[16] = 128/64 = 0h3
DramConﬁgRegLo[15:14] = RdWrQByp = 2h
DramConﬁgRegLo[13] = SelfRefStat = 0
DramConﬁgRegLo[12] = ExitSelfRef = 0h
DramConﬁgRegLo[8] = DramInit = 1h
DramConﬁgRegLo[2] = QFCEn = 0h
DramConﬁgRegLo[1] = DrvEn = 0h
DramConﬁgRegLo[0] = DllDis = 0h
DramConﬁgRegHi[29:26] = EnMemClk[3:0] = See “Memory Clock Enable” on page 138.
DramConﬁgRegHi[25] = MemClkRatioVal = 1h
DramConﬁgRegHi[22:20] = MemClk[2:0] = See “tCL (CAS Latency)” on page 134.
DramConﬁgRegHi[19] = DynIdleCtrEn = 1h
DramConﬁgRegHi[18:16] = IdleCycLimit[2:0] = 3h
DramConﬁgRegHi[11:8] = RdPreamble[3:0] = See “Read Preamble Time” on page 138.
DramConﬁgRegHo[3:0] = AsyncLat[3:0] = See “Maximum Asynchronous Latency” on page 138.
DramDelayLineLo[31:0] = 0000_0000h
ScrubControl[31:0] = DramScrubAddrHi[31:0] = DramScrubAddrLo[31:0] = 0000_0000h
140
DRAM Conﬁguration
Chapter 4
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Notes:
1. In systems that do not use the DRAM controller, it is preferable to disable DRAM input
receivers and output drivers and to leave input receivers unconnected. This bit should be set
in such systems.
2. This bit should be set for 64-bit interfaces (when DramConﬁgRegLo[16] is cleared) on a
platform with a graphics core (AGP or UMA) that issues 32-byte requests. This bit is
ignored for 128-bit interfaces (when DramConﬁgRegLo[16] is set).
3. This bit should be cleared for 64-bit interfaces and set for 128-bit interfaces. The value of
this bit is a function of the system design and cannot be determined via the SPD ROM.
Chapter 4
DRAM Conﬁguration
141
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
142
DRAM Conﬁguration
26094
Rev. 3.06
September 2003
Chapter 4
26094
Rev. 3.06
5
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Machine Check Architecture
The AMD Athlon™ 64 processor and AMD Opteron™ processor machine check mechanism allows
the processor to detect and report a variety of hardware (or machine) errors found when reading and
writing data, probing, cache-line ﬁlls and writebacks. These include parity errors associated with
caches and TLBs, ECC errors associated with caches and DRAM, as well as system bus errors
associated with reading and writing to the external bus.
Software can enable the processor to report machine check errors through the machine check
exception (See “#MC—Machine Check Exception” in AMD64 Architecture Programmer’s Manual,
Volume 2: System Programming). Most machine check exceptions do not allow reliable restarting of
the interrupted programs. However, error conditions are logged in a set of model-speciﬁc registers
(MSRs) that can be used by system software to determine the possible source of a hardware problem.
5.1
Determining Machine Check Support
The availability of machine check registers and support of the machine check exception is
implementation dependent. System software executes the CPUID instruction to determine whether a
processor implements these features. After CPUID is executed, the values of the machine check
architecture (MCA) bit and the machine check exception (MCE) bit loaded in the EDX register
indicate whether the processor implements the machine check registers and the machine check
exception, respectively. See “Processor Feature Identiﬁcation” in AMD64 Architecture Programmer’s
Manual, Volume 2: System Programming, and “CPUID” in AMD64 Architecture Programmer’s
Manual, Volume 3: General Purpose and System Instructions, for further information on the level of
machine check support.
Once system software determines that the machine check registers are available, it must determine the
extent of processor support for the machine check mechanism. This is accomplished by reading the
machine check capabilities register (MCG_CAP). See “Machine Check Global Capabilities Register”
in AMD64 Architecture Programmer’s Manual, Volume 2: System Programming, for more
information on the interpretation of the MCG_CAP contents.
5.2
Machine Check Errors
Machine check errors are either recoverable or irrecoverable. Recoverable errors are those that the
processor can correct and, thus, do not raise the machine check exception (#MC). However, the error
is still logged in the machine check MSRs and it is the responsibility of the system software to
periodically poll the machine check MSRs to determine if recoverable errors have occurred. If a
recoverable error has been logged in the machine check MSRs, a second recoverable error can
overwrite it.
Chapter 5
Machine Check Architecture
143
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Irrecoverable, or fatal, machine check errors cannot be corrected by the processor. If machine check is
enabled, they raise the machine check exception (#MC). If an irrecoverable error has been logged in
the machine check MSRs, a second recoverable or irrecoverable MCA error will not overwrite it but
will set a bit that indicates overﬂow.
In the case of both recoverable and irrecoverable MCA errors, the contents of the machine check
MSRs are maintained though a warm reset. This is helpful since in some cases it may not be possible
to invoke the machine check handler due to the error and this allows the BIOS or other system boot
software to recover and report the information associated with the error.
5.2.1
Sources of Machine Check Errors
The processor can detect errors from the following hardware blocks within the processor. Each block
forms an error reporting bank for the purpose of reporting machine check errors.
•
Data cache unit (DC)—Includes the cache structures that hold data and tags, the data TLBs, and
the data cache probing logic.
•
Instruction cache unit (IC)—Includes the instruction cache structures that hold instructions and
tags, the instruction TLBs, and the instruction cache probing logic.
•
Bus unit (BU)—Includes the system bus interface to the Northbridge and the level 2 cache.
•
Load/store unit (LS)—Includes logic used to manage loads and stores.
•
Northbridge unit (NB)—Includes the Northbridge and DRAM controller.
A scrubber is associated with the data cache in DC, the L2 cache tag array in bus unit, and the DRAM
in the Northbridge. A scrubber is a hardware widget that periodically wakes up during idle cache
cycles and inspects the next line of the array with which it is associated to look for errors. If it ﬁnds a
single bit ECC error, the scrubber corrects the error and prevents a regular access from encountering
the same error. This is important in the data cache, since all ECC errors encountered during regular
operation in the data cache are fatal. In the bus unit and the Northbridge, this scrubber function also
saves the additional time it would take to ﬁx single-bit ECC errors when they are encountered during
normal operation.
Even though 64-bits are read from the data cache during normal operation, a byte load will use one
byte, a word load will use two bytes, and a double word load will use 4 bytes. If a data cache ECC
error is detected in a byte that is not being used by a load, it is corrected like an error detected by the
data cache scrubber. This feature is referred to as "piggyback scrubbing".
Table 30 on page 145 shows the various sources of machine check errors that can be encountered in
the processor. ECC check on the lineﬁll data for IC data is done by the DC unit. However, the ECC
error is reported by IC which is the unit responsible for receiving the data. The data cache is protected
by ECC; however single bit ECC errors encountered by normal load and store accesses are not
corrected. When the data cache scrubber encounters a single-bit ECC error, it corrects it.
144
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
Table 30.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Sources of Machine Check Errors
Type of
Check
Unit
Error Source
DC
System line fill into the data cache
L2 cache line fill into the data cache
Yes—single bit error
ECC
Cache data—data array
ECC
Cache data—snoop tag array
IC
Parity
No—multiple bit error
Yes—if single bit ECC error is detected by
the scrubber
No—any other case different than single bit
ECC error detected by the scrubber
No
Cache data—tag array
No
L1 Data TLB—physical and virtual arrays
No
L2 Data TLB—physical and virtual arrays
No
System line fill into the instruction cache
Yes—single bit error
L2 cache line fill into the instruction cache
ECC
No—multiple bit error
Instruction cache—data array
Yes1
Instruction cache—tag array
Yes1
Instruction cache—snoop tag array
L1 Instruction TLB—physical and virtual
arrays
BU
Recoverable
No
Parity
Yes1
L2 Instruction TLB—physical and virtual
arrays
Yes1
System address out of range
No, but detection is precise
L2 cache data array
ECC
L2 cache—tag array (during scrubs)
Yes—single bit error
No—multiple bit error
L2 cache—tag array
Parity
No.
Single and multiple bit errors in an L2
cache tag can also be detected, but not
corrected.
System Address Out-of-Range
Read
Data
No
LS
System Address Out-of-Range
Read
Data
No. Loads are detected precisely and
stores are detected imprecisely.
NB
See “Machine Check Architecture Registers” on page 146 for information on Northbridge machine check errors.
Notes:
1. Instruction cache lines are invalidated and refetched.
Chapter 5
Machine Check Architecture
145
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
5.3
26094
Rev. 3.06
September 2003
Machine Check Architecture Registers
The processor architecture deﬁnes a set of model-speciﬁc registers (MSRs) to support the MCA
mechanism. They are used to conﬁgure the MCA functions and provide a way for the hardware to
report errors in a manner compatible with the machine check architecture. These registers include the
global MCA registers used to set up machine checks and additional banks of MSRs for recording
errors that are detected by the hardware blocks listed in “Sources of Machine Check Errors” on
page 144. They can be read and written using the RDMSR and WRMSR instructions. The following
is a complete list of MCA MSRs.
•
Global status and control registers
– Machine check capabilities MSR (MCG_CAP)
– Processor status MSR (MCG_STATUS)
–
•
Exception reporting control MSR (MCG_CTL)
Each error reporting bank (associated with a speciﬁc hardware block listed in “Sources of
Machine Check Errors” on page 144) contains the following registers:
– Error reporting control register (MCi_CTL)
– Error reporting status register (MCi_STATUS)
– Error reporting address register (MCi_ADDR)
– Machine check miscellaneous error information register (MCi_MISC)
The i in each register name corresponds to the number of a supported register bank. Each errorreporting register bank is associated with a speciﬁc processor unit (or group of processor units). The
number of error-reporting register banks is implementation-speciﬁc.
Software reads the MCG_CAP register to determine the number of supported register banks. The
AMD Athlon™ 64 and AMD Opteron™ processors support ﬁve banks. The ﬁrst error-reporting
register (MC0_CTL) always starts with MSR address 400h, followed by MC0_STATUS (401h),
MC0_ADDR (402h), and MC0_MISC (403h). Error-reporting-register MSR addresses are assigned
sequentially through the remaining supported register banks. Using this information, software can
access all error-reporting registers in an implementation-independent manner.
The global machine check registers as well as the generic form of each register in the error reporting
banks are now described. The speciﬁcs of each error reporting bank will be described in “Error
Reporting Banks” on page 152.
5.3.1
Global Machine Check Model-Specific Registers (MSRs)
The global MSRs supported by the machine check mechanism include the MCG_CAP, the
MCG_STATUS and the MCG_CTL registers.
146
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
5.3.1.1
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
MCG_CAP—Global Machine Check Capabilities Register
This read-only register indicates the capabilities of the processor machine check architecture.
Attempting to write to this register will result in a #GP exception. See AMD64 Architecture
Programmer’s Manual, Volume 2: System Programming, for more details.
MCG_CAP Register
MSR 0179h
63
32
reserved
9
8
7
0
MCG_CTL_P
31
reserved
Function
Count
Bit
Name
R/W
Reset
63–9
reserved
8
MCG_CTL_P
MCG_CTL Register Present
R
1
7–0
Count
Count
R
05h
0
Field Descriptions
Count (Count)—Bits 7–0. Number of error-reporting banks supported by the processor
implementation.
MCG_CTL Register Present (MCG_CTL_P)—Bit 8. Indicates if the MCG_CTL register is
present. When the bit is set to 1, the register is supported. When the bit is cleared to 0, the
register is unsupported.
5.3.1.2
MCG_STATUS—Global Machine Check Processor Status Register
This register contains basic information about the processor state after a machine check error is
detected. See AMD64 Architecture Programmer’s Manual, Volume 2: System Programming, for
more details.
MCG_STATUS Register
MSR 017Ah
63
32
reserved
Chapter 5
Machine Check Architecture
147
September 2003
31
3
reserved
Bit
Name
Function
R/W
63–3
reserved
2
MCIP
Machine Check in Progress
R/W
0
1
EIPV
Error IP Valid Flag
R/W
0
0
RIPV
Restart IP Valid Flag
R/W
0
2
1
EIPV
Rev. 3.06
0
RIPV
26094
MCIP
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Reset
0
Field Descriptions
Restart IP Valid Flag (RIPV)—Bit 0. When set, indicates if program execution can be restarted at
EIP pushed on stack.
Error IP Valid Flag (EIPV)—Bit 1. When set, indicates if the EIP pushed on stack is that of the
instruction which caused the detection of the machine check error.
Machine Check in Progress (MCIP)—Bit 2. When set, indicates that a machine check is in
progress.
5.3.1.3
MCG_CTL—Global Machine Check Exception Reporting Control Register
This register contains a global bit for each error-checking unit in the processor. Each bit enables the
reporting, through the MCA interface, of errors detected by that particular unit. See AMD64
Architecture Programmer’s Manual, Volume 2: System Programming, for more details.
MCG_CTL Register
MSR 017Bh
63
32
Bit
Name
Function
R/W
63–5
reserved
4
NBE
NB Register Bank Enable
R/W
0
3
LSE
LS Register Bank Enable
R/W
0
2
BUE
BU Register Bank Enable
R/W
0
1
ICE
IC Register Bank Enable
R/W
0
0
DCE
DC Register Bank Enable
R/W
0
148
2
1
0
ICE
3
DCE
reserved
4
BUE
5
LSE
31
NBE
reserved
Reset
0
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Field Descriptions
DC Register Bank Enable (DCE)—Bit 0. When set, indicates that the Data Cache register bank is
enabled.
IC Register Bank Enable (ICE)—Bit 1. When set, indicates that the Instruction Cache register bank
is enabled.
BU Register Bank Enable (BUE)—Bit 2. When set, indicates that the Bus Unit register bank is
enabled.
LS Register Bank Enable (LSE)—Bit 3. When set, indicates that the Load/Store register bank is
enabled.
NB Register Bank Enable (NBE)—Bit 4. When set, indicates that the Northbridge register bank is
enabled.
5.3.2
Error Reporting Bank Machine Check MSRs
The registers in each error reporting bank include MCi_CTL, MCi_CTL_MASK, MCi_STATUS and
MCi_ADDR. The MCi_MISC register is not supported in AMD Athlon™ 64 and AMD Opteron™
processors.
5.3.2.1
MCi_CTL—Machine Check Control Registers
The machine check control registers (MCi_CTL) contain an enable bit for each error source within an
error-reporting register bank. Setting an enable bit to 1 enables error-reporting for the speciﬁc feature
controlled by the bit, and clearing the bit to 0 disables error reporting for the feature. These registers
should generally be set to either all zeros or all ones.
…
Error-Reporting Register-Bank Enable Bits
…
Bit
Name
Function
R/W
Reset
63–0
EN63–EN0
Enables
R/W
0
2
1
0
EN0
EN63
63
EN1
MSRs 0400h, 0404h, 0408h, 040Ch, 0410h
EN2
MCi_CTL Registers
Field Descriptions
Enables (EN63–EN0)—Bits 63–0. If set, error reporting for the specific feature controlled by the bit
is enabled. Not all these bits may be implemented in each bank.
Chapter 5
Machine Check Architecture
149
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
5.3.2.2
26094
Rev. 3.06
September 2003
MCi_STATUS—Machine Check Status Registers
The machine check status register (MCi_STATUS) describes the error detected by its unit. It is
written by the processor and should be cleared to 0 by software; writing any other value to the register
causes a general protection (GP#) exception. When a machine check error occurs, the processor loads
an error code into bits [15:0] of the appropriate MCi_STATUS register. Errors are classiﬁed as either
system bus, cache, or TLB errors. All the status registers in each bank use the same general logging
format shown below.
MCi_STATUS Registers
MSRs 0401h, 0405h, 0409h, 040Dh, 0411h
32
PCC
ADDRV
MISCV
EN
UC
OVER
VAL
63 62 61 60 59 58 57 56
Other Information
31
16 15
Model-Specific Error Code
0
MCA Error Code
Bit
Name
Function
R/W
Reset
63
VAL
Valid
R/W
0
62
OVER
Status Register Overflow
R/W
0
61
UC
Uncorrected Error indication
R/W
0
60
EN
Error Condition Enabled
R/W
0
59
MISCV
Miscellaneous-Error Register Valid
R/W
0
58
ADDRV
Error-Address Register Valid
R/W
0
57
PCC
Processor-Context Corrupt
R/W
0
56–32
Other Information
R/W
0
31–16
Model-Specific Error Code
R/W
0
15–0
MCA Error Code
R/W
0
Field Descriptions
MCA Error Code—Bits 15–0. This field holds an error code when an error is detected.
Model-Specific Error Code—Bits 31–16. This field encodes model-specific information about the
error.
Other Information—Bits 56–32. This field holds model-specific error information.
Processor-Context Corrupt (PCC)—Bit 57. If set to 1, this bit indicates that the state of the
processor may be corrupted by the error condition. Reliable restarting might not be possible.
0 = Processor not corrupted
1 = Processor may be corrupted
150
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Error-Address Register Valid (ADDRV)—Bit 58. If set to 1, this bit indicates that the address saved
in the address register is the address where the error occurred.
0 = Address register not valid
1 = Address register valid
Miscellaneous-Error Register Valid (MISCV)—Bit 59. If set to 1, this bit indicates whether the
Miscellaneous Error register contains valid information for this error. This bit should always
be 0 since the Miscellaneous Error register is not implemented.
Error Condition Enabled (EN)—Bit 60. If set to 1, this bit indicates that MCA error reporting is
enabled for this error in the MCA Control register.
0 = Error checking not enabled
1 = Error checking enabled
Uncorrected Error indication (UC)—Bit 61. If set to 1, this bit indicates that the error was not
corrected by hardware.
0 = Error corrected
1 = Error not corrected
Status Register Overflow (OVER)—Bit 62. Set to 1 if the unit detects an error but the valid bit of
this register already set. Enabled errors are written over disabled errors, uncorrectable errors
are written over correctable errors. Uncorrectable errors are not overwritten.
0 = No error overflow
1 = Error overflow
Valid (VAL)—Bit 63. If set to 1, this bit indicates that a valid error has been detected by the unit. This
bit should be cleared to 0 by software after the register is read.
0 = No valid error detected
1 = Valid error detected
5.3.2.3
MCi_ADDR—Machine Check Address Registers
Each error-reporting register bank includes a machine check address register (MCi_ADDR) that the
processor uses to report the instruction memory address or data memory address responsible for the
machine check error. The contents of this register are valid only if the ADDRV bit in the
corresponding MCi_STATUS register is set to 1.
The address can be up to 48 bits wide. In reality, not all address bits may be valid and the address ﬁeld
of the MCi_ADDR varies in width. The address ﬁeld can hold either a virtual (linear) or physical
address, depending on the type or error. Some error types cause the processor to load valid values into
a subset of the address bits. The bit ranges depend on the values in the MCi_STATUS register, i.e., on
the type of error detected by the unit, which can be determined by examining the MCA error code
contained in the MCi_STATUS register.
Chapter 5
Machine Check Architecture
151
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
MCi_ADDR Registers
26094
Rev. 3.06
September 2003
MSRs 0402h, 0406h, 040Ah, 040Eh, 0412h
63
48 47
reserved
32
ADDR (47–32)
31
0
ADDR (31–0)
Bit
Name
63–48
reserved
47–0
ADDR
Function
R/W
Address
R/W
Reset
0
0
Field Descriptions
Address (ADDR)—Bits 47–0. Not all these bits may be valid. The valid bits depend on the type of
error registered in the corresponding MCi_STATUS register.
5.4
Error Reporting Banks
The AMD Athlon™ 64 and AMD Opteron™ processors have ﬁve error-reporting banks—DC, IC,
BU, LS, and NB. Each error reporting bank includes the following registers:
•
Machine check control register (MCi_CTL).
•
Machine check status register (MCi_STATUS).
•
Machine check address register (MCi_ADDR).
The general format of these registers was described in “Error Reporting Bank Machine Check MSRs”
on page 149. This section describes the speciﬁcs of each register as it relates to each error reporting
bank.
5.4.1
Data Cache (DC)
The data cache unit includes the level 1 data cache that holds data and tags, as well as two levels of
TLBs and data cache probing logic.
5.4.1.1
MC0_CTL—DC Machine Check Control Register
This register enables the reporting, via the MCA interface, of a variety of errors detected by the Data
Cache (DC) processor unit. For an error to be reported, both the global DC enable, MCG_CTL[DCE],
and the corresponding local enable shown below must be set. If the local enable is off, the error will
152
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
only be logged in MC0_STATUS, but not reported via the machine check exception. If the global DC
enable is off, no error will be logged or reported.
MC0_CTL Register
MSR 0400h
63
32
Bit
Name
Function
R/W
63–7
reserved
6
L2TP
L2 TLB Parity Errors
R/W
0
5
L1TP
L1 TLB Parity Errors
R/W
0
4
DSTP
Snoop Tag Array Parity Errors
R/W
0
3
DMTP
Main Tag Array Parity Errors
R/W
0
2
DECC
Data Array ECC Errors
R/W
0
1
ECCM
Multi-bit ECC Data Errors
R/W
0
0
ECCI
Single-bit ECC Data Errors
R/W
0
3
2
1
DECC
ECCM
0
ECCI
4
DMTP
reserved
5
L1TP
6
L2TP
31
DSTP
reserved
Reset
0
Field Descriptions
Single-bit ECC Data Errors (ECCI)—Bit 0. Report single-bit ECC data errors during data cache
line fills or TLB reloads from the internal L2 or the system.
Multi-bit ECC Data Errors (ECCM)—Bit 1. Report multi-bit ECC data errors during data cache
line fills or TLB reloads from the internal L2 or the system.
Data Array ECC Errors (DECC)—Bit 2. Report data cache data array ECC errors.
Main Tag Array Parity Errors (DMTP)—Bit 3. Report data cache main tag array parity errors.
Snoop Tag Array Parity Errors (DSTP)—Bit 4. Report data cache snoop tag array parity errors.
L1 TLB Parity Errors (L1TP)—Bit 5. Report data cache L1 TLB parity errors.
L2 TLB Parity Errors (L2TP)—Bit 6. Report data cache L2 TLB parity errors.
5.4.1.2
MC0_STATUS—DC Machine Check Status Register
The MC0_STATUS MSR describes the error that was detected by DC. The machine check
mechanism writes the status register bits when an error is detected and sets the register valid bit (bit
63) to 1 to indicate that the status information is valid. This register will be written even if error
Chapter 5
Machine Check Architecture
153
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
reporting for the detected error is not enabled. However, error reporting must be enabled for the error
to result in a machine check exception. Software is responsible for clearing this register.
MC0_STATUS Register
MSR 0401h
20 19
reserved
reserved
32
reserved
16 15
EXT_ERR_CODE
31
41 40 39
SCRUB
SYND
UECC
CECC
47 46 45 44
reserved
PCC
ADDRV
MISCV
EN
UC
OVER
VAL
63 62 61 60 59 58 57 56 55 54
0
ERR_CODE
Bit
Name
Function
R/W
Reset
63
VAL
Valid.
R/W
0
62
OVER
Second error detected
R/W
0
61
UC
Error not corrected
R/W
0
60
EN
Error reporting enabled
R/W
0
59
MISCV
Additional info in MCi_MISC
R/W
0
58
ADDRV
Error address in MCi_ADDR
R/W
0
57
PCC
Processor state corrupted by error
R/W
56–55
reserved
0
0
54–47
SYND
ECC Syndrome (7–0)
R/W
0
46
CECC
Correctable ECC error
R/W
0
45
UECC
Uncorrectable ECC error
44–41
reserved
40
SCRUB
39–20
reserved
Error detected on a scrub
R/W
0
R/W
0
R/W
0
0
19–16
EXT_ERR_CODE
Extended error code
R/W
0
15–0
ERR_CODE
Error subsection
R/W
0
Field Descriptions
Error Subsection (ERR_CODE)—Bits 15–0. Indicates in which subsection the error was detected
and what transaction initiated it. See Table 11 on page 99 for the format of the error code and
Tables 12 through 17 on pages 99–100 for descriptions of the subfields.
Extended Error Code (EXT_ERR_CODE)—Bits 19–16. Contains an extended error code.
DC/IC:
0000b = TLB parity error in physical array
154
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
0001b = TLB parity error in virtual array (multi-match error)
BU:
0000b = Bus or cache data array error
0010b = Cache tag array error
LS: Reserved
Error Detected on a Scrub (SCRUB)—Bit 40.
DC: If set, indicates that the error was detected on a scrub.
IC/BU/LS: Reserved.
Uncorrectable ECC Error (UECC)—Bit 45. If set, indicates that the error is an uncorrectable ECC
error.
Correctable ECC Error (CECC)—Bit 46. If set, indicates that the error is a correctable ECC error.
ECC Syndrome Bits 7–0 (SYND)—Bits 54–47.
DC: The lower 8 syndrome bits when an ECC error is detected.
IC/BU/LS: Reserved.
Processor State Corrupted By Error (PCC)—Bit 57. If set, indicates that the processor state may
have been corrupted by the error condition.
Error address in MCi_ADDR (ADDRV)—Bit 58. If set, indicates that the address saved in the
corresponding MCi_ADDR register is the address where the error occurred.
Additional Info in MCi_MISC (MISCV)—Bit 59. If set, indicates that the MCi_MISC register
contains additional info. This bit is always set to 0 on an AMD Athlon™ 64 or
AMD Opteron™ processor.
Error Reporting Enabled (EN)—Bit 60. If set, indicates that error reporting was enabled for this
error in the corresponding MCi_CTL register.
Error Not Corrected (UC)—61. If set, it indicates that the error was not corrected.
Second Error Detected (OVER)—Bit 62. Set if a second error is detected while the VAL bit of this
register is already set.
Valid (VAL)—Bit 63. Set if the information in this register is valid.
5.4.1.3
MC0_ADDR—DC Machine Check Address Register (MSR 0402h)
The contents of this register are valid only if the ADDRV bit in the MC0_STATUS register is set to 1.
As shown in “MCi_ADDR—Machine Check Address Registers” on page 151, the address can be up
to 48 bits wide. In reality, not all address bits may be valid and the address ﬁeld of the MC0_ADDR
Chapter 5
Machine Check Architecture
155
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
varies in width. The bit ranges depend on the values in the MC0_STATUS register (i.e., the type of
error detected) as shown in Table 31.
Table 31.
Valid MC0_ADDR Bits
Error Source
Access Type
Valid Address Bits
Snoop Tag Array
snoop
physical[11–6]
victim
Data Tag Array
load
physical[39–3]
store
Data Array
victim
physical[11–6]
snoop
load
physical[39–3]
store
L1 Data TLB
scrub
physical[11–3]
load, store
linear[47–12]
line-fill
physical[39–6]
L2 Data TLB
L2 Cache Data
System Data
5.4.2
Instruction Cache (IC)
The IC unit includes the level 1 instruction cache that holds instruction data and tags, as well as two
levels of TLBs and instruction cache probing logic.
5.4.2.1
MC1_CTL—IC Machine Check Control Register
This register enables the reporting, via the MCA interface, of a variety of errors detected by the
Instruction Cache (IC) processor unit. For an error to be reported, both the global IC enable,
MCG_CTL[ICE], and the corresponding local enable shown below must be set. If the local enable is
off, the error will only be logged in MC1_STATUS, but not reported via the machine check
exception. If the global IC enable is off, no error will be logged or reported.
MC1_CTL Register
MSR 0404h
63
32
156
Machine Check Architecture
3
IMTP
2
1
0
ECCI
4
ECCM
5
IDP
6
ISTP
7
L1TP
reserved
8
reserved
10 9
RDDE
31
L2TP
reserved
Chapter 5
26094
Rev. 3.06
Bit
Name
63–10
reserved
9
RDDE
8–7
reserved
6
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Function
R/W
Reset
Read Data Errors
R/W
L2TP
L2 TLB Parity Errors
R/W
0
5
L1TP
L1 TLB Parity Errors
R/W
0
4
ISTP
Snoop tag array parity errors
R/W
0
3
IMTP
Main tag array parity errors
R/W
0
2
IDP
Data array parity errors
R/W
0
1
ECCM
Multi-bit ECC data errors
R/W
0
0
ECCI
Single-bit ECC data errors
R/W
0
0
0
0
Field Descriptions
Single-bit ECC Data Errors (ECCI)—Bit 0. Report single-bit ECC data errors during instruction
cache line fills or TLB reloads from the internal L2 or the system.
Multi-bit ECC Data Errors (ECCM)—Bit 1. Report multi-bit ECC data errors during instruction
cache line fills or TLB reloads from the internal L2 or the system.
Data Array Parity Errors (IDP)—Bit 2. Report instruction cache data array parity errors.
Main Tag Array Parity Errors (IMTP)—Bit 3. Report instruction cache main tag array parity
errors.
Snoop Tag Array Parity Errors (ISTP)—Bit 4. Report instruction cache snoop tag array parity
errors.
L1 TLB Parity Errors (L1TP)—Bit 5. Report instruction cache L1 TLB parity errors.
L2 TLB Parity Errors (L2TP)—Bit 6. Report instruction cache L2 TLB parity errors.
Read Data Errors (RDDE)—Bit 9. Report system read data errors for an instruction cache fetch if
MC2_CTL[S_RDE_ALL] = 1.
5.4.2.2
MC1_STATUS—IC Machine Check Status Register (MSR 0405h)
The MC1_STATUS MSR describes the error that was detected by IC and is very similar to
MC0_STATUS. See “MC0_STATUS—DC Machine Check Status Register” on page 153 for a
description of the ﬁelds in this register.
5.4.2.3
MC1_ADDR—IC Machine Check Address Register (MSR 0406h)
The contents of this register are valid only if the ADDRV bit in the MC1_STATUS register is set to 1.
The address can be up to 48 bits wide. In reality, not all address bits may be valid and the address ﬁeld
of the MC1_ADDR varies in width. The bits ranges depend on the values in the MC1_STATUS
register (i.e., the type of error detected) and this is shown in Table 32.
Chapter 5
Machine Check Architecture
157
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 32.
26094
Rev. 3.06
September 2003
Valid MC1_ADDR Bits
Error Source
Access Type
Valid Address Bits
Snoop Tag Array
Snoop
Physical[39–6]
Victim
None
Code Read
Linear[47–6]
Victim
None
Instruction Data Array
Code Read
Linear[47–4]
L1 TLB
Code Read
Linear[47–12] for 4-Kbyte page
Instruction Tag Array
Linear[47–20] for 2-Mbyte page
L2 TLB
Code Read
Linear[47–12] for 4-Kbyte page
L2 Cache Data
Line-fill
Physical[39–6]
System Data
System Address Out of Range
5.4.3
None
Bus Unit (BU)
The bus unit consists of the system bus interface logic and the L2 cache.
5.4.3.1
MC2_CTL—BU Machine Check Control Register
This register enables the reporting, via the MCA interface, of a variety of errors detected in the Bus
processor unit (BU). For an error to be reported, both the global BU enable, MCG_CTL[BUE], and
the corresponding local enable shown below must be set. If the local enable is off, the error will only
be logged in MC2_STATUS, but not reported via the machine check exception. If the global BU
enable is off, no error will be logged or reported.
MC2_CTL Register
MSR 0408h
63
32
158
Machine Check Architecture
5
4
3
S_ECCM_HP
S_ECCM_TLB
S_ECC1_HP
S_ECC1_TLB
2
1
0
S_RDE_HP
6
S_RDE_TLB
7
S_RDE_ALL
8
L2T_PAR_ICDC
L2T_PAR_SNP
L2T_PAR_CPB
L2T_PAR_SCR
L2D_ECC1_TLB
L2D_ECC1_SNP
L2D_ECC1_CPB
L2D_ECCM_TLB
L2D_ECCM_SNP
L2T_ECC1_SCR
reserved
L2D_ECCM_CPB
20 19 18 17 16 15 14 13 12 11 10 9
L2T_ECCM_SCR
31
L2T_PAR_TLB
reserved
Chapter 5
26094
Rev. 3.06
Bit
Name
63–20
reserved
19
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Function
R/W
Reset
L2T_ECCM_SCR
L2 tag array multi-bit ECC Scrub
R/W
0
18
L2T_ECC1_SCR
L2 tag array 1-bit ECC Scrub
R/W
0
17
L2D_ECCM_CPB
L2 data array multi-bit ECC copyback
R/W
0
16
L2D_ECCM_SNP
L2 data array multi-bit ECC snoop
R/W
0
15
L2D_ECCM_TLB
L2 data array multi-bit ECC TLB reload
R/W
0
14
L2D_ECC1_CPB
L2 data array 1-bit ECC copyback
R/W
0
13
L2D_ECC1_SNP
L2 data array 1-bit ECC snoop
R/W
0
12
L2D_ECC1_TLB
L2 data array 1-bit ECC TLB reload
R/W
0
11
L2T_PAR_SCR
L2 tag array parity scrub
R/W
0
10
L2T_PAR_CPB
L2 tag array parity copyback
R/W
0
9
L2T_PAR_SNP
L2 tag array parity snoop
R/W
0
8
L2T_PAR_TLB
L2 tag array parity TLB reload
R/W
0
7
L2T_PAR_ICDC
L2 tag array parity IC or DC fetch
R/W
0
6
S_ECCM_HP
System data multi-bit ECC hardware prefetch
R/W
0
5
S_ECCM_TLB
System data multi-bit ECC TLB reload
R/W
0
4
S_ECC1_HP
System data 1-bit ECC hardware prefetch
R/W
0
3
S_ECC1_TLB
System data 1-bit ECC TLB reload
R/W
0
2
S_RDE_ALL
All system read data
R/W
0
1
S_RDE_TLB
System read data TLB reload
R/W
0
0
S_RDE_HP
System read data hardware prefetch
R/W
0
0
Field Descriptions
System Read Data Hardware Prefetch (S_RDE_HP)—Bit 0. Report system read data errors for a
hardware prefetch.
System Read Data TLB Reload (S_RDE_TLB)—Bit 1. Report system read data errors for a TLB
reload.
All System Read Data (S_RDE_ALL)—Bit 2. Report system read data errors for any operation
including a DC/IC fetch, TLB reload or hardware prefetch.
System Data 1-bit ECC TLB Reload (S_ECC1_TLB)—Bit 3. Report system data 1-bit ECC errors
for a TLB reload.
System Data 1-bit ECC Hardware Prefetch (S_ECC1_HP)—Bit 4. Report system data 1-bit ECC
errors for a hardware prefetch.
System Data Multi-bit ECC TLB Reload (S_ECCM_TLB)—Bit 5. Report system data multi-bit
ECC errors for a TLB reload.
System Data Multi-bit ECC Hardware Prefetch (S_ECCM_HP)—Bit 6. Report system data
multi-bit ECC errors for a hardware prefetch.
Chapter 5
Machine Check Architecture
159
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
L2 Tag Array Parity IC or DC Fetch (L2T_PAR_ICDC)—Bit 7. Report L2 tag array parity errors
for an IC or DC fetch.
L2 Tag Array Parity TLB Reload (L2T_PAR_TLB)—Bit 8. Report L2 tag array parity errors for a
TLB reload.
L2 Tag Array Parity Snoop (L2T_PAR_SNP)—Bit 9. Report L2 tag array parity errors for a snoop.
L2 Tag Array Parity Copyback (L2T_PAR_CPB)—Bit 10. Report L2 tag array parity errors for a
copyback.
L2 Tag Array Parity Scrub (L2T_PAR_SCR)—Bit 11. Report L2 tag array parity errors for a
scrub.
L2 Data Array 1-bit ECC TLB Reload (L2D_ECC1_TLB)—Bit 12. Report L2 data array 1-bit
ECC errors for a TLB reload.
L2 Data Array 1-bit ECC Snoop (L2D_ECC1_SNP)—Bit 13. Report L2 data array 1-bit ECC
errors for a snoop.
L2 Data Array 1-bit ECC Copyback (L2D_ECC1_CPB)—Bit 14. Report L2 data array 1-bit ECC
errors for a copyback.
L2 Data Array Multi-bit ECC TLB Reload (L2D_ECCM_TLB)—Bit 15. Report L2 data array
multi-bit ECC errors for a TLB reload.
L2 Data Array Multi-bit ECC Snoop (L2D_ECCM_SNP)—Bit 16. Report L2 data array multi-bit
ECC errors for a snoop.
L2 Data Array Multi-bit ECC Copyback (L2D_ECCM_CPB)—Bit 17. Report L2 data array
multi-bit ECC errors for a copyback.
L2 Tag Array 1-bit ECC Scrub (L2T_ECC1_SCR)—Bit 18. Report L2 tag array 1-bit ECC errors
for a scrub.
L2 Tag Array Multi-bit ECC Scrub (L2T_ECCM_SCR)—Bit 19. Report L2 tag array multi-bit
ECC errors for a scrub.
5.4.3.2
MC2_STATUS—BU Machine Check Status Register (MSR 0409h)
The MC2_STATUS MSR describes the error that was detected by BU and is very similar to
MC0_STATUS. See “MC0_STATUS—DC Machine Check Status Register” on page 153 for a
description of the ﬁelds in this register.
5.4.3.3
MC2_ADDR—BU Machine Check Address Register (MSR 040Ah)
The contents of this register are valid only if the ADDRV bit in the MC2_STATUS register is set to 1.
The address can be up to 48 bits wide. In reality, not all address bits may be valid and the address ﬁeld
of the MC2_ADDR varies in width. The bit ranges depend on the values in the MC2_STATUS
register (i.e., the type of error detected) and this is shown in Table 33.
160
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
Table 33.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Valid MC2_ADDR Bits
Error Source
Access Type
Valid Address Bits
L2 Cache—Data Array
Any
Physical[39–6]
L2 Cache—Tag Array
Any
MC2_ADDR[3–0] contains
encoded cache way
Physical[15–6] for 1-Mbyte L2
Physical[14–6] for 512-Kbyte L2
Physical[13–6] for 256-Kbyte L2
Physical[12–6] for 128-Kbyte L2
System Address Out of Range
Any
Physical[39–6]
5.4.4
Load Store Unit (LS)
5.4.4.1
MC3_CTL—LS Machine Check Control Register
This register enables the reporting, via the MCA interface, of two types of errors detected by the
Load/Store (LS) processor unit. For an error to be reported, both the global LS enable,
MCG_CTL[LSE], and the corresponding local enable shown below must be set. If the local enable is
off, the error will only be logged in MC3_STATUS, but not reported via the machine check
exception. If the global LS enable is off, no error will be logged or reported.
MC3_CTL Register
MSR 040Ch
63
32
reserved
reserved
Function
1
Bit
Name
R/W
63–2
reserved
1
S_RDE_S
Read Data Errors on Store
R/W
0
0
S_RDE_L
Read Data Errors on Load
R/W
0
0
SRDEL
2
SRDES
31
Reset
0
Field Descriptions
Read Data Errors on Load (S_RDE_L)—Bit 0. Report system read data errors on a load if
MC2_CTL[S_RDE_ALL] = 1.
Chapter 5
Machine Check Architecture
161
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Read Data Errors on Store (S_RDE_S)—Bit 1. Report system read data errors on a store if
MC2_CTL[S_RDE_ALL] = 1.
5.4.4.2
MC3_STATUS—LS Machine Check Status Register (MSR 040Dh)
The MC3_STATUS MSR describes the error that was detected by LS and is very similar to
MC0_STATUS. See “MC0_STATUS—DC Machine Check Status Register” on page 153 for a
description of the ﬁelds in this register.
5.4.4.3
MC3_ADDR—LS Machine Check Address Register (MSR 040Eh)
The contents of this register are valid only if the ADDRV bit in the MC3_STATUS register is set to 1.
The only type of error recorded by the LS machine check mechanism is a “system address out of
range” or read data error for which MC3_ADDR[39:0] store the physical address.
5.4.5
Northbridge (NB)
The Northbridge and DRAM memory controller are included in this block.
5.4.5.1
MC4_CTL—NB Machine Check Control Register (MSR 0410h)
This register enables the reporting, via the MCA interface, of the errors detected by the North Bridge
(NB) processor unit. “MCA NB Control Register” (Function 3, Offset 40h) maps to
MC4_CTL[31:0]. See “MCA NB Control Register” on page 92 for detailed information on this
register. For an error to be reported, both the global NB enable, MCG_CTL[NBE], and the
corresponding local enable must be set. If the local enable is off, the error will only be logged in
MC4_STATUS, but not reported via the machine check exception. If the global NB enable is off, no
error will be logged or reported.
5.4.5.2
MC4_STATUS—NB Machine Check Status Register (MSR 0411h)
The MC4_STATUS MSR describes the error that was detected by the Northbridge. “MCA NB Status
Low Register” (Function 3, Offset 48h) maps to MC4_STATUS MSR[31:0] and “MCA NB Status
High Register” (Function 3, Offset 4Ch) maps to MC4_STATUS MSR[63:32]. See “MCA NB Status
Low Register” on page 98 and “MCA NB Status High Register” on page 101 for more detailed
information.
5.4.5.3
MC4_ADDR—NB Machine Check Address Register (MSR 0412h)
The contents of this register are valid only if the ErrAddrVal bit in the MC4_STATUS register is set to
1. “MCA NB Address Low Register” (Function 3, Offset 50h) maps to MC4_ADDR[31:0] and
“MCA NB Address High Register” (Function 3, Offset 54h) maps to MC4_ADDR[63:32]. See
“MCA NB Address Low Register” on page 103 and “MCA NB Address High Register” on page 104
for a description of this register.
162
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
5.5
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Initializing the Machine Check Mechanism
Following is the general process system software should follow to initialize the machine check
mechanism:
1. Execute the CPUID and verify that the processor supports the machine check exception (MCE)
and MCA. MCE is supported when EDX bit 7 is set to 1, and MCA is supported when EDX bit 14
set to 1. Software should not proceed with initializing the machine check mechanism if MCE is
not supported.
2. If MCA is supported, system software should take the following steps:
a. Check to see if the MCG_CTL_P bit in the MCG_CAP register is set to 1. If it is, then the
MCG_CTL register is supported by the processor. When this register is supported, software
should set its enable bits to 1 for the machine check features it uses. Software can load
MCG_CTL with all 1s to enable all machine check features.
b. Read the COUNT ﬁeld from the MCG_CAP register to determine the number of errorreporting register banks supported by the processor. This is set to 5 in AMD Athlon™ 64 and
AMD Opteron™ processors since there are ﬁve blocks—DC, IC, BU, LS, and NB. For each
error-reporting register bank, software should set the enable bits to 1 in the MCi_CTL register
for the error types it wants the processor to report. Software can load each MCi_CTL register
bit with a 1 to enable all error-reporting mechanisms.
The error-reporting register banks are numbered from 0 to one less than the value found in the
MCG_CAP.COUNT ﬁeld. For example, when the COUNT ﬁeld indicates that 5 register banks
are supported, they are numbered 0 to 4.
c. For each error-reporting register bank, software should clear all status ﬁelds in the
MCi_STATUS register by writing all 0s to the register.
It is possible that valid error status is reported in the MCi_STATUS registers at the time
software clears them. The status can reﬂect fatal errors recorded before a processor reset or
errors recorded during the system power-up and boot process. Prior to clearing the
MCi_STATUS registers, software should examine their contents and log any errors found.
3. As a ﬁnal step in the initialization process, system software should enable the machine check
exception by setting CR4.MCE (bit 6) to 1.
5.6
Using Machine Check Features
System software can detect and handle machine check errors using two methods:
•
Software can periodically examine the machine check status registers for reported errors, and log
any errors found.
•
Software can enable the machine check exception (#MC). When an uncorrectable error occurs,
the processor immediately transfers control to the machine check exception handler. In this case,
system software provides a machine check exception handler that, at a minimum, logs detected
Chapter 5
Machine Check Architecture
163
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
errors. The exception handler can be designed for a speciﬁc processor implementation or can be
generalized to work on multiple implementations.
5.6.1
Handling Machine Check Exceptions
The processor uses the interrupt control-transfer mechanism to invoke an exception handler after a
machine check exception occurs. This requires system software to initialize the interrupt-descriptor
table (IDT) with either an interrupt gate or a trap gate that references the interrupt handler.
At a minimum, the machine check exception handler must be capable of logging errors for later
examination. Because most machine check errors are not recoverable, the ability to log errors can be
sufﬁcient for implementation of the handler. More thorough exception-handler implementations can
analyze errors to determine if each error is recoverable. If a recoverable error is identiﬁed, the
exception handler can attempt to correct the error and restart the interrupted program.
Machine check exception handlers that attempt to correct recoverable errors must be thorough in their
analysis and the corrective actions they take. The following guidelines should be used when writing
such a handler:
•
All status registers in the error-reporting register banks must be examined to identify the cause or
causes of the machine check exception. Read the COUNT ﬁeld from MCG_CAP to determine the
number of status registers supported by the processor. The status registers are numbered from 0 to
one less than the value found in the MCG_CAP.COUNT ﬁeld. For example, if the COUNT ﬁeld
indicates ﬁve status registers are supported, they are numbered MC0_STATUS to MC4_STATUS.
•
Check the valid bit in each status register (MCi_STATUS.VAL). The MCi_STATUS register does
not need to be examined when its valid bit is clear.
•
Check the valid MCi_STATUS registers to see if error recovery is possible. Error recovery is not
possible when:
– The processor-context corrupt bit (MCi_STATUS.PCC) is set to 1.
– The error-overﬂow status bit (MCi_STATUS. OVER) is set to 1. This bit indicates that more
than one machine check error has occurred, but only one error is reported by the status register.
If error recovery is not possible, the handler should log the error information and return to the
operating system.
•
Check the MCi_STATUS.UC bit to see if the processor corrected the error. If UC=1, the processor
did not correct the error, and the exception handler must correct the error prior to restarting the
interrupted program. If the handler cannot correct the error, it should log the error information and
return to the operating system.
•
When identifying the error condition, portable exception handlers should examine only the MCA
error-code ﬁeld of the MCi_STATUS register. See the error codes in tables 12 through 17 for
information on interpreting this ﬁeld.
164
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
If the MCG_STATUS.RIPV bit is set to 1, the interrupted program can be restarted reliably at the
instruction-pointer address pushed onto the exception-handler stack. If RIPV = 0, the interrupted
program cannot be restarted reliably, although it may be possible to restart it for debugging purposes.
•
When logging errors, particularly those that are not recoverable, check the MCG_STATUS.EIPV
bit to see if the instruction-pointer address pushed onto the exception-handler stack is related to
the machine check error. If EIPV = 0, the address is not guaranteed to be related to the error.
•
Prior to exiting the machine check handler, be sure to clear MCG_STATUS.MCIP to 0. MCIP
indicates that a machine check exception occurred. If this bit is set when another machine check
exception occurs, the processor enters the shutdown state.
•
When an exception handler is able to, at a minimum, successfully log an error condition, clear the
MCi_STATUS registers prior to exiting the machine check handler. Software is responsible for
clearing at least the MCi_STATUS.VAL bit.
•
Additional machine check exception-handler portability can be added by having the handler use
the CPUID instruction to identify the processor and its capabilities. Implementation-speciﬁc
software can be added to the machine check exception handler based on the processor information
reported by CPUID.
5.6.1.1
Reporting Correctable Machine Check Errors
Machine check exceptions do not occur if the error is correctable by the processor. If system software
wishes to log and report correctable machine check errors, a system-service routine must be provided
to check the contents of the machine check status registers for correctable errors. The service routine
can be invoked by system software on a periodic basis, or it can be manually invoked by the user as
needed.
Assuming that the processor supports the machine check registers, a service routine that reports
correctable errors should perform the following:
1. Examine each status register (MCi_STATUS) in the error-reporting register banks. For each
MCi_STATUS register with a set valid bit (VAL = 1), the service routine should:
a. Save the contents of the MCi_STATUS register.
b. Save the contents of the corresponding MCi_ADDR register if MCi_STATUS.ADDRV = 1.
c. Save the contents of the corresponding MCi_MISC register if MCi_STATUS.MISCV = 1.
d. Check to see if MCG_STATUS.MCIP = 1, indicating that the machine check exception
handler is in progress. If this is the case, then the machine check exception handler has called
the service routine to log the errors. In this situation, the error-logging service routine should
determine whether or not the interrupted program is restartable, and report the determination
back to the exception handler. The program is not restartable if either of the following is true:
– MCi_STATUS.PCC = 1, indicating the processor context is corrupted.
– MCG_STATUS.RIPV = 0, indicating the interrupted program cannot be restarted reliably
at the instruction-pointer address pushed onto the exception-handler stack.
Chapter 5
Machine Check Architecture
165
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
2. Once the information found in the error-reporting register banks is saved, the MCi_STATUS
register should be cleared to 0. This allows the processor to properly report any subsequent errors
in the MCi_STATUS registers.
3. In multiprocessor conﬁgurations, the service routine can save the processor node identiﬁer. This
can help locate a failing multiprocessor-system component, which can then be isolated from the
rest of the system.
166
Machine Check Architecture
Chapter 5
26094
Rev. 3.06
6
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
System Management Mode (SMM)
System management mode (SMM) is an operating mode entered when system management interrupt
SMI is asserted. The SMI is handled by a dedicated interrupt handler pointed to by processor model
speciﬁc registers. SMM is used for system control activities such as power management. These
activities are transparent to conventional operating systems (such as MS-DOS and Windows®
operating system). SMM is used by the BIOS and specialized low level device drivers. The code and
data for SMM are stored in the SMM memory area, which may be isolated from the main memory
accesses using special processor functions.
This chapter describes the SMM state save area, entry into and exit from SMM, exceptions and
interrupts in SMM, and memory allocation and addressing in SMM.
6.1
SMM Overview
The processor enters SMM after the system logic asserts the SMI interrupt and the processor
recognizes SMI active. The processor may be programmed to send a special bus cycle to the system,
indicating that it is entering SMM mode. The processor saves its state into the SMM memory state
save area and jumps to the SMM service routine. The processor returns from SMM by executing the
RSM instruction in the SMM service routine. The processor restores its state from the SMM state
save area and resumes execution of the instruction following the point where it entered SMM. The
processor may be programmed to send a special bus cycle to the system, indicating that it is exiting
SMM mode.
6.2
Operating Mode and Default Register Values
The software environment after entering SMM mode has the following characteristics:
•
Addressing and operation in Real mode.
•
4-Gbyte segment limits.
•
Default 16-bit operand, address, and stack sizes (instruction preﬁxes can override these defaults).
•
Control transfers that do not override the default operand size truncate the EIP to 16 bits.
•
Far jumps or calls cannot transfer control to a segment with a base address requiring more than 20
bits, like in Real mode segment-base addressing, unless a change is made into protected mode
•
A20M# is disabled. A20M# assertion or deassertion have no effect on SMM mode.
•
Interrupt vectors use the Real mode interrupt vector table.
•
The IF ﬂag in EFLAGS is cleared (INTR is not recognized).
Chapter 6
System Management Mode (SMM)
167
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
•
The TF ﬂag in EFLAGS is cleared.
•
The NMI and INIT interrupts are masked.
•
Debug register DR7 is cleared (debug traps are disabled).
26094
Rev. 3.06
September 2003
Figure 2 shows the default map of the SMM memory area. It consists of a 64-Kbyte area, between
0003_0000h and 0003_FFFFh. The top 32 Kbytes (0003_8000–0003_FFFFh) must be populated
with RAM. The default code-segment (CS) base address for the area (called the SMM_BASE
address) is 0003_0000h. The top 512 bytes (0003_FE00–0003_FFFFh) are the SMM state save area.
The default entry point for the SMM service routine is 0003_8000h.
Fill Down
SMM
state save
Area
0003_FFFFh
0003_FE00h
32-Kbyte
Minimum
SMM
Service Routine
Service Routine Entry Point
0003_8000h
SMM_BASE Address (CS)
0003_0000h
Figure 2.
168
Default SMM Memory Map
System Management Mode (SMM)
Chapter 6
26094
Rev. 3.06
6.3
Table 34.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM State Save Definition
SMM Save State (Offset FE00–FFFFh)
Offset (Hex) from SMM_BASE1 Contents
Size
Allowable Access
Selector
Word
Read-Only
FE02h
Attributes
Word
FE04h
Limit
Doubleword
FE08h
Base
Quadword
Selector
Word
FE12h
Attributes
Word
FE14h
Limit
Doubleword
FE18h
Base
Quadword
Selector
Word
FE22h
Attributes
Word
FE24h
Limit
Doubleword
FE28h
Base
Quadword
Selector
Word
FE32h
Attributes
Word
FE34h
Limit
Doubleword
FE38h
Base
Quadword
Selector
Word
FE42h
Attributes
Word
FE44h
Limit
Doubleword
FE48h
Base
Quadword
Selector
Word
FE52h
Attributes
Word
FE54h
Limit
Doubleword
FE58h
Base
Quadword
reserved
2 Bytes
FE62h
reserved
Word
FE64h
Limit
Word
FE66h–FE67h
reserved
2 Bytes
FE68h
Base
Quadword
FE00h
FE10h
FE20h
FE30h
FE40h
FE50h
FE60h–FE61h
ES
CS
SS
DS
FS
GS
GDTR
Read-Only
Read-Only
Read-Only
Read-Only
Read-Only
Read-Only
Notes:
1. The offset for the SMM-revision identifier is compatible with previous implementations.
Chapter 6
System Management Mode (SMM)
169
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 34.
26094
Rev. 3.06
September 2003
SMM Save State (Offset FE00–FFFFh) (Continued)
Offset (Hex) from SMM_BASE1 Contents
Size
Allowable Access
Selector
Word
Read-Only
FE72h
Attributes
Word
FE74h
Limit
Doubleword
FE78h
Base
Quadword
reserved
2 Bytes
FE82h
reserved
Word
FE84h
Limit
Word
FEB6h–FEB7h
reserved
2 Bytes
FE88h
Base
Quadword
Selector
Word
FE92h
Attributes
Word
FE94h
Limit
Doubleword
FE98h
Base
Quadword
FE70h
FE80h–FEB1h
FE90h
LDTR
IDTR
TR
Read-Only
Read-Only
FEA0h–FEBFh
reserved
32 Bytes
—
FEC0h–FEC3h
SMM I/O Trap
Doubleword
Read-Only
FEC4h–FEC7h
reserved
4 Bytes
—
FEC8h
I/O Instruction Restart
Byte
Read/Write
FEC9h
Auto-Halt Restart
Byte
FECAh—FECFh
reserved
6 Bytes
—
FED0h
EFER
Quadword
Read-Only
FED8h—FEFBh
reserved
36 Bytes
—
Doubleword
Read-Only
FEFCh
SMM-Revision
FF00h
SMM_BASE
Doubleword
Read/Write
FF04h—FF47h
reserved
68 Bytes
—
FF48h
CR4
Quadword
Read-Only
FF50h
CR3
Quadword
FF58h
CR0
Quadword
FF60h
DR7
Quadword
FF68h
DR6
Quadword
FF70h
RFLAGS
Quadword
Identifier1
Read-Only
Read/Write
Notes:
1. The offset for the SMM-revision identifier is compatible with previous implementations.
170
System Management Mode (SMM)
Chapter 6
26094
Rev. 3.06
Table 34.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM Save State (Offset FE00–FFFFh) (Continued)
Offset (Hex) from SMM_BASE1 Contents
Size
Allowable Access
FF78h
RIP
Quadword
Read/Write
FF80h
R15
Quadword
FF88h
R14
Quadword
FF90h
R13
Quadword
FF98h
R12
Quadword
FFA0h
R11
Quadword
FFA8h
R10
Quadword
FFB0h
R9
Quadword
FFB8h
R8
Quadword
FFC0h
RDI
Quadword
FFC8h
RSI
Quadword
FFD0h
RBP
Quadword
FFD8h
RSP
Quadword
FFE0h
RBX
Quadword
FFE8h
RDX
Quadword
FFF0h
RCX
Quadword
FFF8h
RAX
Quadword
Read/Write
Notes:
1. The offset for the SMM-revision identifier is compatible with previous implementations.
After recognizing the SMI assertion, the processor saves almost the entire integer processor state to
memory. Exceptions are: the ﬂoating point state, the model speciﬁc registers, and CR2. Any processor
state not saved in the save state must be saved and restored by the SMM handler. With the
AMD Athlon™ 64 processor and AMD Opteron™ processor extension of register size to 64-bits, the
SMM save state has changed considerably compared to the legacy 32-bit version. One exception is
the SMM Revision Identiﬁer, which is located in the same position as before, for proper SMM
identiﬁcation.
Note: The save state will always be 64-bit, regardless of the operating mode (32-bit or 64-bit).
6.4
SMM Initial State
After storing the save state, the processor starts executing the handler at address SMM_BASE +
08000h. It starts with the entry state listed in Table 35.
Chapter 6
System Management Mode (SMM)
171
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 35.
26094
Rev. 3.06
September 2003
SMM Entry State
Initial Contents
Register
Selector
Base
Limit
CS
SMM_BASE[19:4]
SMM_BASE[31:0]
4 Gbytes
DS
0000h
0000_0000h
4 Gbytes
ES
0000h
0000_0000h
4 Gbytes
FS
0000h
0000_0000h
4 Gbytes
GS
0000h
0000_0000h
4 Gbytes
SS
0000h
0000_0000h
4 Gbytes
General-Purpose Registers
Unmodified
EFLAGS
0000_0002h
EIP
0000_8000h
CR0
Bits 0, 2, 3, and 31 cleared (PE, EM, TS, and PG); remainder is unmodified
CR4
0000_0000h
GDTR
Unmodified
LDTR
Unmodified
IDTR
Unmodified
TR
Unmodified
DR7
0000_0400h
DR6
Undefined
6.5
SMM-Revision Identifier
The SMM-Revision Identiﬁer speciﬁes the version of SMM and the processor extensions.
SMM-Revision Identifier
Offset FEFCh
reserved
IOTRAP
18 17 16 15
BRL
31
0
REV
Bit
Name
Function
31–18
reserved
SBZ
R
17
BRL
SMM Base Relocation Supported
R
16
IOTRAP
SMM I/O Trap Supported
R
1
15–0
REV
SMM Revision Level
R
0064h
172
System Management Mode (SMM)
R/W
Reset
1
Chapter 6
26094
Rev. 3.06
6.6
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM Base Address
SMM_BASE register holds the base of the system management memory region, and its default value
is 30000h. The value of this register is stored in the save state on entry into SMM and it is restored on
returning from SMM. The ﬁrst SMI handler instruction is fetched at SMM_BASE + 8000h. The 16
bit code segment selector is formed on entering SMM from SMM_BASE[19–4]. SMM_BASE[31–
20] and SMM_BASE[3–0] have to be 0; if not, the CS base will not be equal to the (CS selector <<
4), and attempts to load the CS_BASE into other descriptors will not work properly, since the selector
cannot properly represent the base. If there is a far branch in the SMM handler, it will only be able to
address the lower 1M of memory and will not be able to restore the SMM base to a value above 1M,
unless it ﬁrst switches to protected mode.
The SMM base address can be changed in two ways:
•
The SMM base address, at offset FF00h in the SMM state save area, can be changed by the SMM
service routine. The RSM instruction will update the SMM_BASE with the new value.
•
The SMM_BASE is mapped to MSR C001_0111h and WRMSR instruction can be used to
update it.
SMM_BASE Address
Offset FF00h
MSR C001_0111h
63
32
reserved
31
0
SMM_BASE
Bit
Name
Function
63–32
reserved
RAZ
31–0
SMM_BASE
System management mode base
6.7
R/W
R/W
Auto Halt Restart
During entry into SMM, the auto halt restart byte at offset FEC9h in the SMM state save area
indicates whether SMM was entered from the Halt state. Before returning from SMM, the halt restart
byte bit can be written by the SMM service routine to specify whether the return from SMM should
take the processor back to the Halt state or to the instruction-execution state speciﬁed by the SMM
state save area (the instruction after the HLT).
Chapter 6
System Management Mode (SMM)
173
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
If the return from SMM takes the processor back to the Halt state, the HLT instruction is not refetched
and reexecuted, but the Halt special bus cycle is sent to the system on the bus and the processor enters
the HLT state.
Auto Halt Restart
0
reserved
HLT
7
Offset FEC9h
Bit
Name
Function
R/W
7–1
reserved
SBZ
R/W
0
HLT
HLT Restart Byte
R/W
Field Description
HLT Restart Byte (HLT)—Bit 0. On entry:
0 = Entered from normal x86 instruction boundary
1 = Entered from HLT State
After entry and before returning:
0 = Return to SMM State
1 = Return to HLT State
6.8
SMM I/O Trap and I/O Restart
If the assertion of SMI is recognized on the boundary of an I/O instruction, the I/O trap doubleword at
offset FEC0h in the SMM state save area contains information about the executed I/O instruction. If
the Valid bit is equal to 0, not all information is valid.
This is used in the I/O restart implementation. When an I/O access is done to a device that may be
unavailable, the system can assert SMI and trap the I/O instruction. The system can then determine
which device is not responding by using the I/O port information in the I/O Trap doubleword. It can
then ﬁgure out why the device is not responding, ﬁx the device, and re-execute the I/O instruction. It
can reconstruct and then re-execute the I/O instruction in the SMM handler by using the SMM I/O
trap information. More likely, it can use the SMM I/O restart mechanism to cause the processor to
restart the I/O instruction immediately after the RSM.
174
System Management Mode (SMM)
Chapter 6
Rev. 3.06
6.8.1
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM I/O Trap
Bit
Name
Function
31–16
PORT
Trapped I/O Port
6
5
4
3
2
1
reserved
REP
STR
V
BRP
7
SZ8
PORT
12 11 10 9
NMIM
16 15
TF
31
SZ16
Offset FEC0h
SZ32
SMM I/O Trap
R/W
R
15–12
BRP
I/O Breakpoint Matches
R
11
TF
EFLAGS TF Value
R
10
NMIM
NMI Mask
R
9–7
reserved
RAZ
R
6
SZ32
Port Access was 32-bit
R
5
SZ16
Port Access was 16-bit
R
4
SZ8
Port Access was 8-bit
R
3
REP
Repeated Port Access
R
2
STR
String Based Port Access
R
1
V
I/O Trap Word Valid Bit
R
0
R/W
Port Access Type
0 = Write (OUT instruction)
1 = Read (IN instruction)
R
6.8.2
0
R/W
26094
SMM I/O Restart Byte
The processor initializes the I/O trap restart slot to 00h on entry into SMM. If SMM is entered as a
result of a trapped I/O instruction, the processor indicates the validity of the I/O instruction by setting
or clearing bit 1 of the I/O trap doubleword at offset FEC0h in the SMM state save area. The SMM
service routine should test bit 1 of the I/O trap doubleword to determine if a valid I/O instruction was
being executed when entering SMM and before writing the I/O trap restart slot. If the I/O instruction
is valid, the SMM service routine can safely rewrite the I/O trap restart slot with the value FFh,
causing the processor to re-execute the trapped I/O instruction when the RSM instruction is executed.
If the I/O instruction is invalid, writing the I/O trap restart slot has undeﬁned results.
If a second SMI is asserted and a valid I/O instruction is trapped by the ﬁrst SMM handler, the CPU
services the second SMI prior to re-executing the trapped I/O instruction. The second entry into SMM
will not have bit 1 of the I/O trap doubleword set, and the second SMM service routine must not
rewrite the I/O trap restart slot.
During a simultaneous SMI I/O instruction trap and debug breakpoint trap, the processor ﬁrst
responds to the SMI and postpones recognizing the debug exception until after the RSM instruction
returns from SMM. If the debug registers DR3–DR0 are used while in SMM, they must be saved and
restored by the SMM handler. The processor automatically saves and restores DR7 and DR6. If the I/
Chapter 6
System Management Mode (SMM)
175
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
O trap restart slot in the SMM state save area contains the value FFh when the RSM instruction is
executed, the debug trap does not occur until after the I/O instruction is re-executed.
SMM I/O Restart Byte
7
Offset FEC8h
0
RST
Bit
Name
Function
R/W
7–0
RST
I/O Restart Byte
00h = Do not restart
FFh = Restart I/O instruction
R/W
6.9
Exceptions and Interrupts in SMM
When SMM is entered, the processor masks INTR, NMI, SMI, INIT, and A20M interrupts. The
processor clears the IF ﬂag to disable INTR interrupts. To enable INTR interrupts within SMM, the
SMM handler must set the IF ﬂag to 1. A20M is disabled so that address bit 20 is always generated
externally when in SMM.
Generating an INTR interrupt is a method for unmasking NMI interrupts in SMM. The processor
recognizes the assertion of NMI within SMM immediately after the completion of an IRET
instruction. The NMI can be enabled by using an INTR interrupt. Once NMI is recognized within
SMM, NMI recognition remains enabled until SMM is exited, at which point NMI masking is
restored to the state it was in before entering SMM.
Because the IF ﬂag is cleared when entering SMM, the HLT instruction should not be executed in
SMM without ﬁrst setting the IF bit to 1. Setting this bit to 1 enables the processor to exit the Halt
state by means of an INTR interrupt.
The processor still responds to the DBREQ and STPCLK interrupts as well as to all exceptions that
might be caused by the SMM handler.
6.10
Protected SMM and ASeg/TSeg
System management memory, SMRAM, is deﬁned by two ranges. The ASeg range is located at a
ﬁxed range from A0000h–BFFFFh. The TSeg range is located at a variable base. The size of the TSeg
range is controlled by a variable mask. SMRAM provides a safe location for system management
code and data that is not readily accessible by applications. The SMM interrupt handler can be
located in one of these two ranges if protection is needed, or it can be located outside these ranges in
the rest of memory.
176
System Management Mode (SMM)
Chapter 6
26094
Rev. 3.06
6.10.1
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM_MASK Register
This register holds the mask of the TSeg range as well as conﬁguration information for both the ﬁxed
ASeg range and the variable TSeg range.
SMM_MASK Register
MSR C001_0113h
63
40 39
32
Bit
Name
Function
63–40
reserved
RAZ
39–17
MASK
SMRAM TSeg Range Mask
16–15
reserved
RAZ
14–12
TMTypeDram
SMRAM TSeg Range Memory Type
11
reserved
RAZ
4
3
2
1
AClose
TValid
0
AValid
5
TClose
6
AMTypteIcWc
7
reserved
8
AMTypeDram
reserved
MASK (14–0)
12 11 10
TMTypeDram
17 16 15 14
reserved
31
MASK (22–15)
TMTypeIoWc
reserved
R/W
R/W
R/W
10–8
AMTypeDram
SMRAM ASeg Range Memory Type
7–6
reserved
RAZ
R/W
5
TMTypeIoWc
Non-SMRAM TSeg Range Memory Type
R/W
4
AMTypteIcWc
Non-SMRAM ASeg Range Memory Type
R/W
3
TClose
Send TSeg Range Data Accesses to Non-SMRAM
R/W
2
AClose
Send ASeg Range Data Accesses to Non-SMRAM
R/W
1
TValid
Enable TSeg SMRAM Range
R/W
0
AValid
Enable ASeg SMRAM Range
R/W
SMRAM TSeg Range Mask (MASK)—Bits 39–17. Mask of the SMRAM TSeg Range.
SMRAM TSeg Range Memory Type (TMTypeDram)—Bits 14–12. Memory Type for the
SMRAM TSeg Range.
SMRAM ASeg Range Memory Type (AMTypeDram)—Bits 10–8. Memory Type for the SMRAM
ASeg Range.
Non-SMRAM TSeg Range Memory Type (TMTypeIoWc)—Bit 5. Memory Type for the nonSMRAM TSeg Range.
Non-SMRAM ASeg Range Memory Type (AMTypteIcWc)—Bit 4. Memory Type for the nonSMRAM ASeg Range.
Chapter 6
System Management Mode (SMM)
177
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Send TSeg Range Data Accesses to Non-SMRAM (TClose)—Bit 3. Send Data Accesses to the
TSeg Range to Non-SMRAM.
Send ASeg Range Data Accesses to Non-SMRAM (AClose)—Bit 2. Send Data Accesses to the
ASeg Range to Non-SMRAM.
Enable TSeg SMRAM Range (TValid)—Bit 1. Enables the TSeg SMRAM Range.
Enable ASeg SMRAM Range (AValid)—Bit 0. Enables the ASeg SMRAM Range.
6.10.2
SMM_ADDR Register
This register holds the base of the variable TSeg range.
SMM_ADDR Register
MSR C001_0112h
63
40 39
reserved
31
32
ADDR (22–15)
17 16
reserved
ADDR (14–0)
Bit
Name
Function
63–40
reserved
RAZ
39–17
ADDR
Base of the SMRAM TSeg Range
16–0
reserved
RAZ
6.10.3
0
R/W
R/W
SMM ASeg
The SMM ASeg address range is A0000–BFFFFh. The ASeg is enabled by the AValid bit in the
SMM_MASK MSR.
When the ASeg is enabled, the MTRR memory maps are not used for the addresses in A0000–
BFFFFh memory range. Instead, settings in the SMM_MASK register and whether the processor is in
SMM or not determine how those addresses are handled. If not in SMM, the addresses are mapped to
I/O space and do not access DRAM. The memory type used is either UC or WC, based on the
AMTypeIoWc bit in the SMM_MASK register. When inside SMM, the accesses to the ASeg memory
range are directed to DRAM.
The memory type is determined using the normal memory type encoding and is speciﬁed in
AMTypeDram. Table 36 lists these types.
178
System Management Mode (SMM)
Chapter 6
26094
Rev. 3.06
Table 36.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SMM ASeg-Enabled Memory Types
Out of SMM
Address Range
In SMM
AMTypeIoWc=0
AMTypeIoWc=1
0–9FFFh
Normal MTRR
Normal MTRR
Normal MTRR
A0000–BFFFFh
DRAM, AMTypeDram
I/O, UC
I/O, WC
C0000h–Max
Normal MTRR
Normal MTRR
Normal MTRR
The SMM save state and code can be cached. After leaving SMM, the same addresses will bypass the
caches and DRAM and access the I/O memory subsystem. This protects the SMM memory from
corruption by programs outside of SMM.
6.10.4
SMM TSeg
The TSeg is similar to the ASeg except it provides more programability and therefore allows more
space to be allocated in a different area of memory for the SMM handler. It is enabled by the TValid
bit in the SMM_MASK register. It uses the SMM_ADDR register to specify its base. It can be used
with ASeg or by itself. When it is used with ASeg, the two spaces should not overlap. If they do
overlap, the memory types should be consistent in both the ASeg map and the TSeg map for the
overlapping addresses. Otherwise, undeﬁned behavior will occur.
The SMM TSeg begins at the address speciﬁed in SMM_ADDR concatenated with 0s in the lower 17
bits, forcing the TSeg to begin on a 128-Kbyte boundary. The ending of TSeg is speciﬁed by the
SMM_Mask[39:17]. The SMM_Mask and SMM_ADDR function as the variable-range MTRRs.
Each address generated by the processor is ANDed with the SMM_MASK and compared to the
SMM_ADDR ANDed with the mask. If the values are equal, the address is within the TSeg range.
For example, suppose you wish to create a TSeg starting at the 1-Mbyte address boundary and
extending 256 Kbytes. The SMM_ADDR register would be set to 0010_0000h and the SMM_MASK
to FFFC_0002h. This will make the TSeg range from 0010_0000–0013_FFFFh.
The TSeg memory type table is similar to the ASeg memory type table. When not in SMM and TSeg
is enabled, the addresses within the TSeg range are directed to I/O space with either a UC memory
type or a WC memory type, based on the TMTypeIoWc bit in the SMM_MASK register. When
within SMM, the accesses are directed to DRAM with a memory type speciﬁed by TMTypeDram.
Table 37.
SMM TSeg-Enabled Memory Types
Out of SMM
Address Range
In SMM
TMTypeIoWc=0
TMTypeIoWc=1
<TSeg Range
Normal MTRR
Normal MTRR
Normal MTRR
TSeg Range
DRAM, TMTypeDram
I/O, UC
I/O, WC
> TSeg Range
Normal MTRR
Normal MTRR
Normal MTRR
Chapter 6
System Management Mode (SMM)
179
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
6.10.5
26094
Rev. 3.06
September 2003
Closing SMM
Sometimes within SMM code with ASeg or TSeg enabled, there is a requirement to access the I/O
space at the same address as the current SMM segment. That is typically only accessible outside of
SMM. To accomplish this function, the Aclose and Tclose bits from SMM_MASK register are used.
When the Aclose bit is set, data cache accesses to the ASeg that would normally go to DRAM are
redirected to I/O, with the memory type speciﬁed by AMTypeIoWc.
The same function applies to the TSeg. Instruction cache accesses and Page Directory/Table accesses
still access the SMM code in DRAM. When the SMM handler is done accessing the I/O space, it must
clear the appropriate close bit. Failure to do so and then issuing an RSM will probably cause the
processor to enter shutdown, as the save state will be read from I/O space.
6.10.6
Locking SMM
The SMM registers can be locked by setting the SMMLOCK (HWCR, bit 0). Once set, the
SMM_BASE, the SMM_ADDR, all but the two close bits of SMM_MASK and the
RSMSPCYCDIS, SMISPCYCDIS, and SMMLOCK bits of HWCR are locked and cannot be
changed. The only way to unlock the SMM registers is to assert reset. This provides security to the
SMM mechanism. The BIOS can lock the SMM environment after setting it up so that it can not be
tampered with.
6.11
SMM Special Cycles
Special cycles can be initiated on entry and exit from SMM to acknowledge to the system that these
transitions are occurring. This allows secure spaces external to the processor to be secured within the
SMM environment as well. It also helps with synchronizing the SMI interrupt across multiple
processor systems. These special cycles can be disabled by setting the SMISPCYCDIS to 1 to disable
the entry special cycle and setting the RSMSPCYCDIS bit to 1 to disable the exit special cycle.
180
System Management Mode (SMM)
Chapter 6
26094
Rev. 3.06
7
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Advanced Programmable Interrupt
Controller (APIC)
All AMD Athlon™ 64 and AMD Opteron™ processor-based systems must support 64-bit operating
systems. As a result, all platforms must support APIC and declare it to the operating system in the
appropriate tables.
The local APIC contains logic to receive interrupts from a variety of sources and to send interrupts to
other local APICs, as well as registers to control its behavior and report status.
Interrupts can be received from:
•
HyperTransport™ I/O devices, including the I/O hub (I/O APICs)
•
Other local APICs (inter-processor interrupts)
•
APIC timer
•
Performance counters
•
Legacy local interrupts from the I/O hub (INTR and NMI)
•
APIC internal errors
The APIC timer, performance counters, local interrupts, and internal errors are all considered local
interrupt sources, and their routing is controlled by local vector table entries. These entries assign a
message type and vector to each interrupt, allow them to be masked, and track the status of the
interrupt.
I/O and inter-processor interrupts have their message type and vector assigned at the source and are
unaltered by the local APIC. They carry a destination field and a mode bit that together determine
which local APICs will accept them. The destination mode (DM) bit specifies if the interrupt request
packet should be handled in Physical or Logical destination mode.
•
In physical destination mode, if the destination field is FFh, the interrupt is a broadcast and is
accepted by all local APICs. Otherwise, the interrupt is only accepted by the local APIC whose
APIC ID matches the destination field of the interrupt. Physical mode allows up to 255 APICs to
be addressed individually.
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
181
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
•
26094
Rev. 3.06
September 2003
In logical destination mode, a local APIC accepts interrupts selected by the logical destination
register (LOG_DEST) and the destination field of the interrupt using either cluster or flat format,
as configured by the destination format register.
– If flat destinations are in use, bits 7–0 of the LOG_DEST field are checked against bits 7–0 of
the arriving interrupt’s Destination Field. If any bit position is set in both fields, this APIC is a
valid destination. Flat format allows up to 8 APICs to be addressed individually.
– If cluster destinations are in use, bits 7–4 of the LOG_DEST field are checked against bits 7–
4 of the arriving interrupt’s destination field, identifying the cluster. If all of bits 7–4 match,
then bits 3–0 of the LOG_DEST and the interrupt destination are checked for any bit positions
that are set in both fields, choosing processors within the cluster. If both conditions are met, this
APIC is a valid destination. Cluster format allows 15 clusters of 4 APICs each to be addressed.
– In both flat and cluster formats, if the destination field is FFh, the interrupt is a broadcast and
is accepted by all local APICs.
7.1
Interrupt Delivery
SMI, NMI, INIT, Startup, and External interrupts are classified as non-vectored interrupts.
When an APIC accepts a non-vectored interrupt, it is handled directly by the processor instead of
being queued in the APIC. When an APIC accepts a fixed or lowest priority interrupt, it sets the bit in
the interrupt request register (IRR) corresponding to the vector in the interrupt. (For local interrupt
sources, this comes from the vector field in that interrupt’s local vector table entry.) If a subsequent
interrupt with the same vector arrives when the corresponding IRR bit is already set, the two
interrupts are collapsed into one. Vectors 0–15 are reserved.
7.2
Vectored Interrupt Handling
The task priority register (TPR_PRI) and processor priority register (PROC_PRI) each contain an 8bit priority, divided into a main priority (bits 7–4) and a priority sub-class (3–0). The task priority is
assigned by software to set a threshold priority at which the processor will be interrupted.
To calculate processor priority, an 8-bit ISR vector is set based on the highest bit set in the in-service
register (ISR). Like the TPR_PRI and PROC_PRI, this vector is divided into a main priority (7–4) and
priority sub-class (3–0). The main priority of the TPR_PRI and ISR are compared and the highest is
the PROC_PRI, as follows:
If (TPR_PRI[7:4] >= ISRVect[7:4])PROC_PRI = TPR_PRI
Else
PROC_PRI = {ISRVect[7:4],0h}
The resulting processor priority is used to determine if any accepted interrupts (indicated by IRR bits)
are high enough priority to be serviced by the processor. When the processor is ready to service an
interrupt, the highest bit in the IRR is cleared, and the corresponding bit is set in the ISR. The
182
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
corresponding trigger mode register (TMR) bit is set if the interrupt is level-triggered and cleared if
edge-triggered.
When the processor has completed service for an interrupt, it performs a write to the end of interrupt
(EOI) register, clearing the highest ISR bit and causing the next-highest interrupt to be serviced. If the
corresponding TMR bit is set, EOI messages will be sent to all APICs to complete service of the
interrupt at the source.
7.3
Spurious Interrupts
In the event that the task priority is set to or above the level of the interrupt to be serviced, the local
APIC will deliver a spurious interrupt vector to the processor, as specified by the spurious interrupt
vector register. The ISR will be unchanged and no EOI will occur.
7.3.1
Spurious Interrupts Caused by Timer Tick Interrupt
A typical interrupt is asserted until it is serviced. Interrupt is deasserted when software clears the
interrupt status bit within the interrupt service routine. Timer tick interrupt is an exception, since it is
deasserted regardless of whether it is serviced or not. A BIOS programs the 8254 Programmable
Timer to generate an 18.2 Hz square wave. This square wave represents the timer tick interrupt
(PITIRQ) and in some interrupt configurations it is connected to the 8259 Programmable Interrupt
Controller (PIC) IRQ0 input.
The processor is not always able to service interrupts immediately (i.e. when interrupts are masked by
clearing EFLAGS.IM or when the processor is in debug mode). The method of interrupt delivery to
the processor (wire or messages) determines system behavior for the timer tick interrupt connected to
the PIC after the processor has not been able to service it for some time. The PIC output INTR is
identical to PITIRQ when interrupts are not serviced.
If interrupts are delivered to the processor using a wire, and the processor is not able to service the
timer tick interrupt for some time, timer tick interrupts asserted during that time will be lost. The
following cases are possible when the processor is ready to service interrupts:
•
INTR is deasserted and it is not serviced by the processor. This will happen almost 50 percent of
the time.
•
INTR is asserted and it is serviced by the processor. This will happen almost 50 percent of the
time.
•
INTR is asserted, and it is detected by the processor. The processor sends the interrupt
acknowledge cycle, but when the PIC receives it, INTR is deasserted, and the PIC sends a
spurious interrupt vector. This will happen a very small percentage of the time.
The probability of spurious interrupts, when interrupts are delivered to the processor using a wire, is
very low. An example of a configuration that uses wire to deliver interrupts to the processor has
PITIRQ connected to PIC IRQ0, and PIC INTR connected to LINT0 of an AMD Athlon processor.
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
183
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
If interrupts are delivered to the processor using messages, and the processor is not able to service the
timer tick interrupt for an extended period of time, the INTR caused by the first timer tick interrupt
asserted during that time is delivered to the local APIC in ExtInt mode and latched, and the
subsequent timer tick interrupts are lost. The following cases are possible when the processor is ready
to service interrupts:
•
An ExtInt interrupt is pending, and INTR is asserted. This results in timer tick interrupt servicing.
This occurs 50 percent of the time.
•
An ExtInt interrupt is pending, and INTR is deasserted. The processor sends the interrupt
acknowledge cycle, but when the PIC receives it, INTR is deasserted, and the PIC sends a
spurious interrupt vector. This occurs 50 percent of the time.
There is a 50 percent probability of spurious interrupts when interrupts are delivered to the processor
using messages. An example of an AMD Athlon™ 64 or AMD Opteron™ processor configuration
that uses messages to deliver interrupts to the processor has PITIRQ connected to PIC IRQ0, PIC
INTR connected to IOAPIC IRQ0, and interrupts are delivered to the processor by HyperTransport
messages. HyperTransport messages are always used to deliver PIC interrupts to the processor in a
system with AMD Athlon™ 64 or AMD Opteron™ processors. An example of an AMD Athlon
processor configuration that uses messages to deliver interrupts to the processor has PITIRQ
connected to PIC IRQ0, PIC INTR connected to IOAPIC IRQ0, and interrupts are delivered to the
processor by APIC 3-wire messages.
7.4
Lowest-Priority Arbitration
Fixed and non-vectored interrupts are accepted by their destination APICs without arbitration.
Delivery of lowest-priority interrupts requires all APICs to arbitrate to determine which one will
accept the interrupt. If focus processor checking is enabled (bit 9 of the spurious interrupt vector
register cleared), then the focus processor for an interrupt will always accept the interrupt. A
processor is the focus of an interrupt if it is already servicing that interrupt (corresponding ISR bit is
set) or if it already has a pending request for that interrupt (corresponding IRR bit is set). If there is no
focus processor for an interrupt, or focus processor checking is disabled, then each APIC will
calculate an arbitration priority value, and the one with the lowest result will accept the interrupt.
To calculate arbitration priority, an 8-bit IRR Vector is set based on the highest bit set in the IRR. Like
the ISR, TPR_PRI, and PROC_PRI, this vector is divided into a main priority (7–4) and priority subclass (3–0). The main priority of the TPR_PRI, ISR, and IRR are compared and the highest is the new
ARB_PRI, as follows:
If
(TPR_PRI[7:4] >= IRRVect[7:4] and
TPR_PRI[7:4] > ISRVect[7:4])
ARB_PRI = TPR_PRI
Else if (IRRVect[7:4] > ISRVect[7:4]) ARB_PRI = {IRRVect[7:4],0h}
Else
ARB_PRI = {ISRVect[7:4],0h}
184
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
7.5
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Inter-Processor Interrupts
In order to redirect an interrupt to another processor, originate an interrupt to another processor, or
allow a processor to interrupt itself, the interrupt command register (ICR) provides a mechanism for
generating interrupts. A write to the low doubleword of the ICR causes an interrupt to be generated,
with the properties specified by the ICR low and ICR high fields.
7.6
APIC Timer Operation
The local APIC contains a 32-bit timer, controlled by the Timer LVT entry, initial count, and divide
configuration registers. The processor bus clock is divided by the value in the timer divide
configuration register to obtain a time base for the timer. When the timer initial count register is
written, the value is copied into the timer current count register. The current count register is
decremented at the rate of the divided clock. When the count reaches 0, a timer interrupt is generated
with the vector specified in the Timer LVT entry. If the Timer LVT entry specifies periodic operation,
the current count register is reloaded with the initial count value, and it continues to decrement at the
rate of the divided clock. If the mask bit in the timer LVT entry is set, timer interrupts are not
generated.
7.7
State at Reset
At power-up or reset, all registers take on the values listed in their descriptions. SMI, NMI, INIT,
Startup, and LINT interrupts may be accepted.
When the APIC is software-disabled, pending interrupts in the ISR and IRR are held, but further
fixed, lowest-priority, and ExtInt interrupts will not be accepted. All LVT entry mask bits are set and
cannot be cleared.
When a processor accepts an INIT interrupt, the APIC is reset as at power-up, with the exception that
APIC_ID is unaffected.
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
185
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
7.8
26094
Rev. 3.06
September 2003
Register Summary
All APIC registers are accessed with memory reads and writes of (APIC_BASE+Offset).
APIC_BASE is MSR 001Bh and defaults to 00_FEE0_0000h.
The APIC registers and their offsets are listed in Table 38.
Table 38. APIC Register Summary
Offset (Hex)
Mnemonic
Name
20
APIC_ID
APIC ID Register
30
APIC_VER
APIC Version Register
80
TPR_PRI
Task Priority Register
90
ARB_PRI
Arbitration Priority Register
A0
PROC_PRI
Processor Priority Register
B0
EOI
End Of Interrupt Register
D0
LOG_DEST
Logical Destination Register
E0
DEST_FORMAT
Destination Format Register
F0
SPUR_INTR_VEC
Spurious Interrupt Vector Register
100-170
ISR
In-Service Registers
180-1F0
TMR
Trigger Mode Registers
200-270
IRR
Interrupt Request Registers
280
ERR_STAT
Error Status Register
300
ICRLO
Interrupt Command Register Low (bits 31–0)
310
ICRHI
Interrupt Command Register High (bits 63–32)
320
TIMER_LVT
Timer Local Vector Table Entry
340
PERF_CNT_LVT
Performance Counter Local Vector Table Entry
350
LINT0_LVT
Local Interrupt 0 Local Vector Table Entry
360
LINT1_LVT
Local Interrupt 1 Local Vector Table Entry
370
ERROR_LVT
Error Local Vector Table Entry
380
INIT_CNT
Timer Initial Count Register
390
CURR_CNT
Timer Current Count Register
3E0
TIMER_DVD_CFG
Timer Divide Configuration Register
186
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
7.8.1
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
APIC ID Register
APIC_ID Register
31
Offset 20h
24 23
0
APICID
Bit
reserved
Name
Function
R/W
Reset
31–24
APICID
APIC Identification
R/W
Node ID
23–0
reserved
R/O
00_0000h
Field Descriptions
APIC Identification (APICID)—Bits 31–24. Software must ensure that all APICs are assigned
unique APIC IDs. When both ApicExtId and ApicExtBrdCst in the HyperTransport™
Transaction Control Register are set, all 8 bits of APIC ID are used. When either ApicExtID
or ApicExtBrdCst is clear, only bits 3–0 of APIC ID are used, and bits 7–4 are reserved. Node
ID is initially 00h if this is the boot strap processor or 07h for all other nodes.
7.8.2
APIC Version Register
APIC_VER Register
31
Offset 30h
24 23
reserved
Bit
Name
31–24
reserved
23–16
MaxLVTEntry
15–8
reserved
7–0
Version
16 15
MaxLVTEntry
Function
Maximum LVT Entry
Version
8
7
reserved
0
Version
R/W
Reset
R/O
00h
R/O
04h
R/O
00h
R/O
10h
Field Descriptions
Max LVT Entry—Bits 23–16. This field indicates the number of entries in the Local Vector Table
minus one.
Version—Bits 7–0. This field indicates the version number of this APIC implementation.
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
187
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
7.8.3
26094
Rev. 3.06
September 2003
Task Priority Register
TPR_PRI Register
Offset 80h
31
8
7
reserved
Bit
Name
31–8
reserved
7–0
Priority
0
Priority
Function
R/W
Reset
R/O
0000_00h
Priority
R/W
00h
Field Descriptions
Priority—Bits 7–0. This field is assigned by software to set a threshold priority at which the
processor will be interrupted.
7.8.4
Arbitration Priority Register
ARB_PRI Register
Offset 90h
31
8
7
reserved
Bit
Name
31–8
reserved
7–0
Priority
0
Priority
Function
Priority
R/W
Reset
R/O
0000_00h
R/O
00h
Field Descriptions
Priority—Bits 7–0. This field indicates the processor’s current priority, for a task being serviced, an
interrupt being serviced, or an interrupt that is pending, and is used to arbitrate between
processors to determine which will accept a lowest-priority interrupt request.
7.8.5
Processor Priority Register
PROC_PRI Register
Offset A0h
31
8
7
reserved
188
Advanced Programmable Interrupt Controller (APIC)
0
Priority
Chapter 7
26094
Bit
Rev. 3.06
Name
31–8
reserved
7–0
Priority
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Function
Priority
R/W
Reset
R/O
0000_00h
R/O
00h
Field Descriptions
Priority—Bits 7–0. This field indicates the processor’s current priority servicing a task or interrupt,
and is used to determine if any pending interrupts should be serviced.
7.8.6
End of Interrupt Register
This register is written by the software interrupt handler to indicate that servicing of the current
interrupt is complete.
EOI Register
Offset B0h
31
0
reserved
Bit
Name
31–0
reserved
7.8.7
Function
R/W
Reset
W/O
XXXX_XXXXh
Logical Destination Register
LOG_DEST Register
31
Offset D0h
24 23
0
Destination
Bit
reserved
Name
Function
R/W
Reset
31–24
Destination
Destination
R/W
00h
23–0
reserved
R/O
00_0000h
Field Descriptions
Destination (Destination)—Bits 31–24. This field contains this APIC’s destination identification,
and is used to determine which interrupts should be accepted.
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
189
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
7.8.8
26094
Rev. 3.06
September 2003
Destination Format Register
DEST_FORMAT Register
31
Offset E0h
28 27
0
Format
Bit
reserved
Name
Function
R/W
Reset
31–28
Format
Format
R/W
Fh
27–0
reserved
R/O
FFF_FFFFh
Field Descriptions
Format (Format)—Bits 31–28. 0h and Fh are the only allowed values. This field controls which
format to use when accepting interrupts with a logical destination mode.
0h = Cluster destinations are used.
Fh = Flat destinations are used.
7.8.9
Spurious Interrupt Vector Register
reserved
Bit
Name
10 9
8
APICSWEn
31
Offset F0h
FocusDisable
SPUR_INTR_VEC Register
7
0
Vector
Function
R/W
Reset
R/O
0000_00h
Focus Disable
R/W
0
31–10
reserved
9
FocusDisable
8
APICSWEn
APIC Software Enable
R/W
0
7–0
Vector
Vector
R/W
FFh
Field Descriptions
Focus Disable (FocusDisable)—Bit 9. This bit disables focus processor checking during lowestpriority arbitrated interrupts.
APIC Software Enable (APICSWEn)—Bit 8. When APICSWEnable is cleared, SMI, NMI, INIT,
Startup, and LINT interrupts may be accepted, pending interrupts in the ISR and IRR are held,
but further Fixed, Lowest-Priority, and ExtInt interrupts will not be accepted. All LVT entry
Mask bits are set and cannot be cleared.
190
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Vector—Bits 7–0. When ApicExtSpur in the HyperTransportTM Transaction Control Register is set,
bits 3–0 of Vector are writable. When ApicExtSpur is clear, bits 3–0 are read-only 1111b. This
field contains the vector that is sent to the processor in the event of a spurious interrupt.
7.8.10
In-Service Registers
ISR Register
Offset 100h
31
16 15
InServiceBits
0
reserved
Bit
Name
Function
R/W
Reset
31–16
InServiceBits
In-Service Bits
R/O
0000h
15–0
reserved
R/O
0000h
Field Descriptions
In-Service Bits (InServiceBits)—Bits 31–16. These bits are set when the corresponding interrupt is
being serviced by the CPU.
ISR Registers
Offsets 170h, 160h, 150h, 140h, 130h, 120h, 110h
31
0
InServiceBits
Bit
Name
Function
R/W
Reset
31–0
InServiceBits
In-Service Bits
R/O
0000_0000h
Field Descriptions
In-Service Bits (InServiceBits)—Bits 31–0. These bits are set when the corresponding interrupt is
being serviced by the CPU. Interrupts are mapped as follows:
Chapter 7
ISR
Interrupt
number
Offset 110h
Offset 120h
Offset 130h
Offset 140h
Offset 150h
Offset 160h
Offset 170h
63–32
95–64
127–96
159–128
191–160
223–192
255–224
Advanced Programmable Interrupt Controller (APIC)
191
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
7.8.11
26094
September 2003
Trigger Mode Registers
TMR Register
Offset 180h
31
16 15
TriggerModeBits
Bit
Rev. 3.06
0
reserved
Name
Function
R/W
31–16
TriggerModeBits
Trigger Mode Bits
15–0
reserved
Reset
R/O
0000h
R/O
0000h
Field Descriptions
Trigger Mode Bits (TriggerModeBits)—Bits 31–16. The Trigger Mode bit for each corresponding
interrupt is updated when an interrupt enters servicing. It is 0 for edge-triggered interrupts and 1 for
level-triggered interrupts.
TMR Registers
Offsets 1F0h, 1E0h, 1D0h, 1C0h, 1B0h, 1A0h, 190h
31
0
Trigger Mode Bits
Bit
Name
Function
R/W
Reset
31–0
TriggerModeBits
Trigger Mode Bits
R/O
0000_0000h
Field Descriptions
Trigger Mode Bits (TriggerModeBits)—Bits 31–0. The Trigger Mode bit for each corresponding
interrupt is updated when an interrupt enters servicing. It is 0 for edge-triggered interrupts and 1 for
level-triggered interrupts. Interrupts are mapped as follows:
192
TMR
Interrupt
number
Offset 190h
Offset 1A0h
Offset 1B0h
Offset 1C0h
Offset 1D0h
Offset 1E0h
Offset 1F0h
63–32
95–64
127–96
159–128
191–160
223–192
255–224
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
7.8.12
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Interrupt Request Registers
IRR Register
Offset 200h
31
16 15
RequestBits
0
reserved
Bit
Name
Function
R/W
Reset
31–16
RequestBits
Interrupt Requests Bits
R/O
0000h
15–0
reserved
R/O
0000h
Field Descriptions
Interrupt Request Bits (RequestBits)—Bits 31–16. Request bits are set when the corresponding
interrupt is accepted by the APIC.
IRR Registers
Offsets 270h, 260h, 250h, 240h, 230h, 220h, 210h
31
0
RequestBits
Bit
Name
Function
R/W
Reset
31–0
RequestBits
Interrupt Requests Bits
R/O
0000_0000h
Field Descriptions
Interrupt Request Bits (RequestBits)—Bits 31–0. Request bits are set when the corresponding
interrupt is accepted by the APIC. Interrupts are mapped as follows:
Chapter 7
IRR
Interrupt
number
Offset 210h
Offset 220h
Offset 230h
Offset 240h
Offset 250h
Offset 260h
Offset 270h
63–32
95–64
127–96
159–128
191–160
223–192
255–224
Advanced Programmable Interrupt Controller (APIC)
193
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
7.8.13
26094
Rev. 3.06
September 2003
Error Status Register
This register must be written to trigger an update before it can be read. Each write causes the internal
error state to be loaded into this register, clearing the internal error state. A second write before
another error occurs causes this register to be cleared.
Bit
Name
4
3
Function
R/W
Reset
R/O
0000_00h
Illegal Register Address
W/R
0
31–8
reserved
7
IllegalRegAddr
6
RcvdIllegalVector
Received Illegal Vector
W/R
0
5
SentIllegalVector
Sent Illegal Vector
W/R
0
4
reserved
R/O
0
3
RcvAcceptError
Receive Accept Error
W/R
0
2
SendAcceptError
Send Accept Error
W/R
0
1–0
reserved
R/O
00b
2
1
0
reserved
5
SendAcceptError
6
RcvAcceptError
reserved
7
reserved
8
SentIllegalVector
31
RcvdIllegalVector
Offset 280h
IllegalRegAddr
ERR_STAT Register
Field Descriptions
Illegal Register Address (IllegalRegAddr)—Bit 7. This bit indicates that an access to a nonexistent
register location within this APIC was attempted.
Received Illegal Vector (RcvdIllegalVector)—Bit 6. This bit indicates that this APIC received a
message with an illegal Vector (00h to 0Fh for fixed and lowest priority interrupts).
Sent Illegal Vector (SentIllegalVector)—Bit 5. This bit indicates that this APIC attempted to send a
message with an illegal Vector (00h to 0Fh for fixed and lowest priority interrupts).
Receive Accept Error (RcvAcceptError)—Bit 3. This bit indicates that a message received by this
APIC was not accepted by this or any other APIC.
Send Accept Error (SendAcceptError)—Bit 2. This bit indicates that a message sent by this APIC
was not accepted by any APIC.
194
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
7.8.14
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Interrupt Command Register Low
ICRLO Register
Offset 300h
Bit
Name
31–20
reserved
19–18
DestShrthnd
DM
DlvryStat
reserved
TM
Level
reserved
reserved
20 19 18 17 16 15 14 13 12 11 10
DestShrthnd
31
8
7
Msg Type
0
Vector
Function
R/W
Reset
R/O
000h
Destination Shorthand
R/W
00b
17–16
reserved
R/O
00b
15
TM
Trigger Mode
R/W
0
14
Level
Level
R/W
0
13
reserved
R/O
0
12
DlvryStat
Delivery Status
R/O
0
11
DM
Destination Mode
R/W
0
10–8
MsgType
Message Type
R/W
000b
7–0
Vector
Vector
R/W
00h
Field Descriptions
Destination Shorthand (DestShrthnd)—Bits 19–18. This field provides a quick way to specify a
destination for a message. If All Including Self or All Excluding Self are used, then DM is
ignored and physical is automatically used.
00b =Destination Field
01b =Self
10b =All Including Self
11b =All Excluding Self (Note that this sends a message with a destination encoding of all 1’s,
so if lowest priority is used, the message could end up being reflected back to this APIC.)
Trigger Mode (TM)—Bit 15. This bit can be 0 for Edge or 1 for Level.
Level (Level)—Bit 14. This bit can be 0 for deasserted or 1 or asserted.
Delivery Status (DlvryStat)—Bit 12. This bit is set to indicate that the interrupt has not yet been
accepted by the destination CPU(s).
Destination Mode (DM)—Bit 11. This bit can be 0 for Physical or 1 for Logical.Message Type
(MsgType)—Bits 10–8. The encoding for this field is as follows:
000b = Fixed
001b = Lowest Priority
010b = SMI
011b = Remote Read
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
195
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
100b = NMI
101b = INIT
110b = Startup
111b = External Interrupt
Note: Not all combinations of ICR fields are valid. Only those combinations listed in Table 39 are
valid.
Table 39. Valid Combinations of ICR Fields
Message Type
(MsgType)
Trigger Mode
(TM)
Level
(Lvl)
Destination Shorthand
(DestShrthnd)
Fixed
Edge
Level
Edge
Level
X
X
Assert
X
Assert
X
X
X
Dest or All Excluding Self
Dest or All Excluding Self
Dest or All Excluding Self
Lowest Priority, SMI, NMI, INIT
Startup
Note: X indicates a “don’t care”.
Vector (Vector)—Bits 7–0. This field contains the vector that will be sent for this interrupt source.
7.8.15
Interrupt Command Register High
ICRHI Register
63
Offset 310h
56 55
32
DestinationField
reserved
Bit
Name
Function
R/W
Reset
63–56
DestinationField
Destination Field
R/W
00h
55–32
reserved
R/O
00_0000h
Field Descriptions
Destination Field (DestinationField)—Bits 63–56. This field contains the destination encoding used
when the destination shorthand is 00b.
196
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
7.8.16
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Timer Local Vector Table Entry
TIMER_LVT Register
Offset 320h
Bit
Name
31–18
reserved
17
Mode
16
Mask
15–13
reserved
12
DlvryStat
11–8
reserved
7–0
Vector
13 12 11
reserved
DlvryStat
reserved
Mask
18 17 16 15
Mode
31
8
7
0
reserved
Vector
Function
R/W
Reset
R/O
0000h
Mode
R/W
0
Mask
R/W
1
R/O
000b
Delivery Status
R/O
0
R/O
0h
Vector
R/W
00h
Field Descriptions
Mode (Mode)—Bit 17. This bit is 0 for One-shot and 1 for Periodic.
Mask (Mask)—Bit 16. If this bit is set, this LVT Entry will not generate interrupts.
Delivery Status (DlvryStat)—Bit 12. This bit is set to indicate that the interrupt has not yet been
accepted by the CPU.
Vector (Vector)—Bits 7–0. This field contains the vector that will be sent for this interrupt source.
7.8.17
Performance Counter Local Vector Table Entry
PERF_CNT_LVT Register
Offset 340h
Bit
Name
31–17
reserved
16
Mask
15–13
reserved
12
DlvryStat
reserved
reserved
reserved
13 12 11 10
DlvryStat
17 16 15
Mask
31
8
7
Msg Type
0
Vector
Function
R/W
Reset
R/O
0000h
Mask
R/W
1
R/O
000b
Delivery Status
R/O
0
11
reserved
R/O
0
10–8
MsgType
Message Type
R/W
000b
7–0
Vector
Vector
R/W
00h
Chapter 7
Advanced Programmable Interrupt Controller (APIC)
197
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Mask (Mask)—Bit 16. If this bit is set, this LVT Entry will not generate interrupts.
Delivery Status (DlvryStat)—Bit 12. This bit is set to indicate that the interrupt has not yet been
accepted by the CPU.
Message Type (MsgType)—Bits 10–8. Only Message Types 000b (Fixed) and 100b (NMI) are legal.
Vector (Vector)—Bits 7–0. This field contains the vector that will be sent for this interrupt source.
7.8.18
Local Interrupt 0 (Legacy INTR) Local Vector Table Entry Register
LINT0_LVT Register
Offset 350h
Bit
Name
Function
31–17
reserved
16
Mask
15
TM
14
RmtIRR
13
PinPol
12
DlvryStat
Delivery Status
11
reserved
reserved
DlvryStat
PinPol
TM
reserved
RmtIRR
17 16 15 14 13 12 11 10
Mask
31
8
7
0
MsgType
Vector
R/W
Reset
R/O
0000h
Mask
R/W
1
Trigger Mode
R/W
0
Remote IRR
R/O
0
Pin Polarity
R/W
0
R/O
0
R/O
0
10–8
MsgType
Message Type
R/W
000b
7–0
Vector
Vector
R/W
00h
Field Descriptions
Mask (Mask)—Bit 16. If this bit is set, this LVT Entry will not generate interrupts.
Trigger Mode (TM)—Bit 15. This bit is set for level-triggered interrupts and clear for edge-triggered
interrupts.
Remote IRR (RmtIRR)—Bit 14. If trigger mode is level, remote IRR is set when the interrupt has
begun service. Remote IRR is cleared when the EOI has occurred.
Pin Polarity (PinPol)—Bit 13. This bit is not used because LINT interrupts are delivered by
HyperTransport™ messages instead of individual pins.
Delivery Status (DlvryStat)—Bit 12. This bit is set to indicate that the interrupt has not yet been
accepted by the CPU.
198
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Message Type (MsgType)—Bits 10–8. Only Message Types 000b (Fixed), 100b (NMI), and 111b
(External Interrupt) are legal.
Vector (Vector)—Bits 7–0. This field contains the vector that will be sent for this interrupt source.
7.8.19
Local Interrupt 1 (Legacy NMI) Local Vector Table Entry
LINT1_LVT Register
Offset 360h
Bit
Name
31–17
reserved
16
reserved
DlvryStat
PinPol
TM
reserved
RmtIRR
17 16 15 14 13 12 11 10
Mask
31
8
7
MsgType
0
Vector
Function
R/W
Reset
R/O
0000h
Mask
Mask
R/W
1
15
TM
Trigger Mode
R/W
0
14
RmtIRR
Remote IRR
R/O
0
13
PinPol
Pin Polarity
R/W
0
12
DlvryStat
Delivery Status
R/O
0
11
reserved
R/O
0
10–8
MsgType
Message Type
R/W
000b
7–0
Vector
Vector
R/W
00h
Field Descriptions
These fields are the same as those for Local Interrupt 0.
7.8.20
Error Local Vector Table Entry
31
reserved
Bit
Name
31–17
reserved
16
Mask
15–13
reserved
12
DlvryStat
11—8
reserved
7–0
Vector
Chapter 7
17 16 15
13 12 11
DlvryStat
Offset 370h
Mask
ERROR_LVT Register
reserved
8
7
reserved
0
Vector
Function
R/W
Reset
R/O
0000h
Mask
R/W
1
R/O
000b
Delivery Status
R/O
0
R/O
0h
Vector
R/W
00h
Advanced Programmable Interrupt Controller (APIC)
199
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Mask (Mask)—Bit 16. If this bit is set, this LVT Entry will not generate interrupts.
Delivery Status (DlvryStat)—Bit 12. This bit is set to indicate that the interrupt has not yet been
accepted by the CPU.
Vector (Vector)—Bits 7–0. This field contains the vector that will be sent for this interrupt source.
7.8.21
Timer Initial Count Register
INIT_CNT Register
Offset 380h
31
0
Count
Bit
Name
Function
R/W
Reset
31–0
Count
Initial Count Value
R/W
0000_0000h
Field Descriptions
Count (Count)—Bits 31–0. This field contains the value copied into the current count register when
the timer is loaded or reloaded.
7.8.22
Timer Current Count Register
CURR_CNT Register
Offset 390h
31
0
Count
Bit
Name
Function
R/W
Reset
31–0
Count
Current Count Value
R/O
0000_0000h
Field Descriptions
Count (Count)—Bits 31–0. This field contains the current value of the counter.
200
Advanced Programmable Interrupt Controller (APIC)
Chapter 7
Rev. 3.06
Timer Divide Configuration Register
Offset 3E0h
31
4
reserved
Bit
Name
31–4
reserved
3
Div[3]
2
reserved
1–0
Div[1–0]
3
2
Function
R/W
Reset
R/O
0000_00h
Div[3]
R/W
0
R/O
0
Div[1–0]
R/W
00b
1
0
Div[1-0]
TIMER_DVD_CFG Register
reserved
7.8.23
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Div[3]
26094
Field Descriptions
Div[3] and Div[1–0]—Bits 3 and 1–0. The Div bits are encoded as follows:
Div[3]
0
0
0
0
1
1
1
1
Chapter 7
Div[1–0]
00
01
10
11
00
01
10
11
Resulting Timer Divide
2
4
8
16
32
64
128
1
Advanced Programmable Interrupt Controller (APIC)
201
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
202
26094
Advanced Programmable Interrupt Controller (APIC)
Rev. 3.06
September 2003
Chapter 7
26094
Rev. 3.06
8
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
HyperTransport™ Technology
Configuration and Enumeration
An AMD Opteron™ processor or AMD Athlon™ 64 processor is a node connected to other nodes
with coherent HyperTransport™ links. Function 0 HyperTransport technology conﬁguration register
values for each node must be initialized. A node is identiﬁed in the Node ID register (Function 0,
Offset 60h) with a number from 0 to 7. Routing tables are used by the node to decide whether to
accept a packet and/or forward it through any or all of its HyperTransport links.
Three examples of coherent HyperTransport technology initialization sequence are shown in this
chapter: 1-node initialization, 2-node initialization, and a dynamic coherent initialization that can
enumerate any number of nodes.
8.1
Initial Configuration Steps
For coherent HyperTransport technology initialization, the following steps are required before
enumeration:
1. Coherent HyperTransport initialization is only performed by Node 0 (BSP). All the other nodes
(APs) should bypass the HyperTransport initialization process.
2. Clearing Function 0, Offset 6Ch[0] enables the routing table for Node 0, and allows access to the
memory controller on Node 0.
3. Node 0 is by deﬁnition connected to the HyperTransport I/O Hub, which means that Node 0 owns
the compatibility chain. The link number that connects to HyperTransport I/O Hub should be
written to Function 0, Offset 64h[SbLink].
4. Count coherent HyperTransport links on Node 0.
To perform the step 4 above, it is necessary to detect whether each HyperTransport link on a node is
connected, and if the connected link type is coherent or noncoherent. The AMD Opteron™ processor
supports up to three links, and the AMD Athlon™ 64 processor supports one link. For each
HyperTransport link, the following steps are performed to detect the link connection status and type:
1. Check whether the Link Connect Pending bit is set in Function 0, Offset 98h[4].
2. If the Link Connect Pending bit is clear, check whether the Link Connected bit is set in Function
0, Offset 98h[0].
If the Link Connected bit is set, the testing port is connected to other node.
3. Check whether the Initialization Complete bit is set in Function 0, Offset 98h[1].
4. If the bit is set, check the Non Coherent bit in Function 0, Offset 98h[2].
The bit value 0 indicates a coherent HyperTransport link, while value 1 indicates a noncoherent
link.
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
203
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
5. Record the number of coherent and noncoherent ports and their respective port numbers.
8.2
One-Node Coherent HyperTransport™
Technology Initialization
The number of nodes that exist in the system is determined on page 203. The following conﬁguration
steps should be executed in single (one-node) processor systems:
1. Set Function 0, Offset 68h[LimitCldtCfg] bit to enable limiting the extent of coherent
HyperTransport conﬁguration space based on the number of nodes in the system.
2. Disable read/write/ﬁll probes in Function 0, Offset 68h[10, 3:0], since single processor systems
do not need probes.
8.3
Two-Node Coherent HyperTransport™
Technology Initialization
For two-node systems, Node ID and Routing Tables need to be written for both nodes, as follows.
1. Initialize Node 0 Routing Table (RT) rows 0 and 1. Since the coherent HyperTransport link
number on Node 0 is already known, RT can be generated using the backbone-ﬁrst algorithm. See
“Algorithm” on page 207 for the backbone-ﬁrst routing algorithm.
2. Initialize Node 0 RT row 7. Since Node Id of Node 1 is 7 by default, this ID will be used
temporarily to conﬁgure Node 1. RT row 7 is the same as RT row 1.
3. Verify if Node 0-to-Node 7 link is established by reading Vendor ID/Device ID of Node 7.
4. Detect the connected coherent HyperTransport link number on Node 7 by reading Default Link in
Function 0, Offset 6Ch[3:2].
5. Initialize Node 7 RT row 0 and 1. Since the link number on Node 7 is known, RT can be generated
using the backbone-ﬁrst algorithm.
6. Renumber Node 7 to Node 1 by programming Function 0, Offset 60h[2:0].
7. Reset row 7 on Node 0 to the default value.
8. Write CPU Count and Node Count in Function 0, Offset 60h[19:16, 6:4].
9. Set Function 0, Offset 68h[LimitCldtCfg] bit for Node 0 and Node 1.
204
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
8.4
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Generic HyperTransport™ Technology
Configuration
This section presents a complete dynamic algorithm for initialization of the coherent HyperTransport
technology fabric. This algorithm describes conﬁguration register initialization for proper routing of
HyperTransport technology trafﬁc through the coherent fabric.
First, system conﬁguration assumptions are made. Second, the fabric initialization algorithm is
presented. Last, a sub-algorithm for writing rows of routing tables is described. See Chapter 3,
“Memory System Conﬁguration” for details about conﬁguration registers.
The described algorithm assumes one of the following system conﬁgurations:
1-node -> 1x1
2-node -> 2x1
4-node -> 2x2
6-node -> 2x3
8-node -> 2x4
For example, the 8-node system diagram is shown in Figure 3, where the node numbers are assigned
by the algorithm. Other conﬁgurations may be preferable to the one shown in Figure 3, but they are
not addressed.
Node
7
Node
6
Node
5
Node
4
Node
3
Node
2
Node
1
Node
0
Note: This is an 8-node system configuration. On the right side, from bottom to top, the nodes are numbered 0, 2, 4, and 6,
while on the left, also from bottom to top, they are 1, 3, 5, and 7. The horizontal links (between nodes 0 and 1, 2 and
3, 4 and 5, and 6 and 7) are called “across”. The link from 0 to 2 is “up” (similarly for other links from i to i+2) and from
2 to 0 is “down” (similarly for other links from i+2 to i). The slot formation for 2, 4, or 6 nodes is obtained by deleting
2-node horizontal rows, starting at the top of the figure.
Figure 3.
An Eight-Node Configuration
The algorithm determines node numbers and initializes backbone-ﬁrst routing tables, which provide
performance that is reasonably close to optimal in numerous settings. The idea of the backbone-ﬁrst
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
205
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
algorithm is explained below and refers only to the request and broadcast tables. The response tables
produce paths in the reverse order from paths produced by the other two tables. If the nodes and links
are laid out in one of the slot or planar conﬁgurations, then this algorithm will produce the expected
node numbers.
In the following description, a read or write to a register on Node n is always accomplished by an
access to the register offset in conﬁguration space, function 0, device 24+n, where n is the target Node
ID. For example, to write a value into the Node ID register of Node n, function 0, device 24+n, offset
60h is used to form the address. Until a node’s routing tables have been written, its routing table
causes all messages to all devices from 24 to 31 to be accepted locally.
Conﬁguration is viewed as a ladder extending upward, as shown in the Figure 3, where the backbone
(upward) direction is perpendicular to the “rungs”. A data structure in table format for each node has
information about adjacent links: “up”, “across”, and “down”. The optimal routing table is backboneﬁrst, since request paths follow along the length of the ladder, only going along a rung at the very end
of the path (if necessary). This table is kept in the BSP address space, although it contains information
for all nodes.
Section “Writing Routing Table Rows” on page 213 describes how to use link information—“up”,
“across”, “down”—to create the backbone-ﬁrst routing tables used in various steps of the algorithm.
The BSP will execute code from the ﬁrmware ROM to initialize the coherent HyperTransport
technology fabric.The assumption is that each node has the following registers that can be accessed
by the BSP.
1. NodeId[2:0], which has default value of 0 on the BSP node and 7 on the other nodes.
2. Routing tables that indicate to which links to route a packet and whether to route the packet to the
internal node based on the destination NodeID of the packet. These have default value “route to
this node”, which results in accepting all commands and not forwarding them.
3. A RoutingTableDisable (RTD) bit forces all commands to be accepted and all responses to be
returned back to the link that delivered the request, regardless of the routing tables. If a processor
on that node is sending messages through the crossbar, then this assumption could be incorrect
(see step 6 below).
4. DefaultLink ﬁeld (Function 0, Offset 6Ch) that records the link from which the last request was
received.
5. A LinkType register indicating how many connected links the node has and whether the links are
coherent or noncoherent.
6. A RequestDisable bit that prevents the node from making a request. This bit is cleared on the BSP
coming out of microcode reset, and should be set on all other nodes.
The code below is only executed on the BSP. The following assumptions are made for each node.
1. The NodeID is ignored when determining where to route a command (including conﬁguration
reads and writes). Determining routing and acceptance is only based on the node’s routing table,
the destination node in the command, and the destination unit in the command.
206
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
2. The result of a read or write to conﬁguration space is independent of the destination node in the
command. For example, if a node accepts a command that reads its Node ID, then it will respond
with the value of its Node ID register even if the destination was a device not corresponding to
that Node ID.
Node LDTi Type registers (i=0,1,2) are read to determine its connected coherent HyperTransport links
not marked as UP.
The following steps are grouped into three blocks, starting with Nodes 0, 1, 2 and 3, then Nodes 4 and
5, and ﬁnally Nodes 6 and 7.
8.4.1
Algorithm
Pseudo-code for fabric initialization is described in this section. Steps taken after the initialization of
Node IDs and routing tables are not described.
Subroutine ﬁll_routing_table(NodeN, rowr), described in “Writing Routing Table Rows” on
page 213, writes row r of the routing table for Node N. On a few occasions, a row of a routing table is
written directly without using this subroutine.
Northbridge registers on nodes other than Node0 should not be accessed before microcode on AP
processors has set the RequestDisable bit. Thus, the ﬁrst access to an AP node is a read of the
RequestDisable bit (Function 0, Offset 6Ch). This bit should be polled until it is set. A possible ﬁrst
access to each node is marked with “(ﬁrst_access)”. The value returned for the Function 0, Offset
6Ch register read can be 0FFFFFFFFh, indicating a time-out. In that case, the link is treated as not
present.
The algorithm comments are in parentheses.
BLOCK A: Complete the first four nodes, stopping short if there are fewer than four
nodes
A1. If Node0 has no coherent HyperTransport links other than possibly one marked as UP, then it is a
single-node system. There is no need to change the routing tables or set the NodeId, except for the
case where a link L is marked as UP. In that case, rows 0 and 1 of Node0’s routing table should be
written after marking the UP link as “across” and marking one of the other two links as “up” and the
other as “down.” Exit.
EXIT IF ONLY ONE NODE OR UNIPROCESSOR.
A2. If Node0 has only one coherent HyperTransport link, which is not UP, then it is a two-node
system. The following steps should be executed.
A2.1. Mark the coherent link of Node0 as “across” and mark the other two links of Node0 as
“down” and “up”, arbitrarily.
A2.2. Call fill_routing_table(Node0,rowi) for i=0,1.
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
207
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
A2.3. Write a NodeId of 1 to Device 25 (first_access).
A2.4. Read the default link of Node1, and mark it as “across”. Mark the remaining two links of
Node1 as “down” and “up”, arbitrarily.
A2.5. Call fill_routing_table(Node1,rowi) for i=0,1.
A2.6. Clear the RTD bit for Node1.
A2.7. Exit.
EXIT IF ONLY 2 NODES.
(It is known at this point that Node0 has exactly two coherent links, since one link is an noncoherent
link leading to the HyperTransport I/O Hub.)
A3. For each coherent link L of Node0, determine the number of coherent links in the node on the
other side of L, as follows.
A3.1. Write row 7 of Node0’s routing table to a value that causes requests and responses destined for Device 31 (Node7) to be routed through link L.
A3.2. Read the number of coherent links on Device 31 (first_access), and save the association
of link L with this number.
(It is known at this point that Node0 has exactly two coherent links, since one link is a noncoherent
link to the HyperTransport I/O Hub. It is assumed that there are minimum four nodes, where each of
the two links connects Node0 to a node with minimum two coherent links. The node across from
Node0 has exactly two coherent links, since it is in the “corner” of the conﬁguration. It is also known
whether the system is limited to four nodes: there are more than four nodes if and only if one of the
nodes connected to Node0 has three coherent links (the one labeled as Node2 at the end of the
algorithm).)
A4. Mark the coherent links of Node0 and write rows 0 to 2 of its routing table.
A4.1. Mark as “across” the coherent link from Node0 that points to a node with only two coherent links (as saved in Step A3). If there are two such nodes, i.e., if it is a four-node system,
then the “across” link is set to be the smaller of the two coherent link numbers of Node0.
(Note that the numbering will be correct for both the slot and planar four-node configurations.)
A4.2. Mark as “up” the other coherent link of Node0, and as “down” the noncoherent link of
Node0.
A4.3. Call fill_routing_table(Node0,rowi) for i=0,1,2. (A temporary value for row 3 will be
written in Step A7.1, and row 2 is the last one written here.)
208
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
A5. Complete the node across from Node0 by writing its NodeID to 1, marking its links, writing its
routing table, and clearing its Routing Table Disable (RTD) bit.
A5.1. Write a NodeID value of 1 to Device 25.
A5.2. Read the default link at Node1 (which will be the one pointing to Node0). Then mark that
link as “across”.
A5.3. Mark the remaining connected coherent link of Node1 as “up” and its last unmarked link
as “down”.
A5.4. Call fill_routing_table(Node1,rowi) for i=0,1,2,3.
A5.5. Clear Node1’s RTD.
A6. Write a NodeID of 2 to the node “up” from Node0, and then mark “down” link of Node2.
A6.1. Write a NodeID value of 2 to Device 26.
A6.2. Mark the link at Node2 down to Node0 as “down”, by reading the default link as in Step
A5.2 above.
A7. Write a NodeID of 3 to the node “up” from Node1, and mark its link to Node1 as “down”, as
follows.
A7.1. Write row 3 of Node0’s routing table so that messages destined for Node3 are sent
“across”.
A7.2. Mark the link at Node3 down to Node1 as “down” by reading the default link of Device
27 (first_access).
A7.3. Write a NodeID of 3 to Device 27.
A8. Call fill_routing_table(Node0,row3).
(At this point Node0 initialization is complete: it has the correct NodeID and routing table, and its
coherent links are appropriately marked “across” and “up”. Node1 initialization has also been
completed (Step A5). Node2 and Node3 have Node IDs and still have their RTD bits set.)
A9. If it is a four-node system (because Node2 has two coherent links, as discussed in the comment of
Step A3), then the following steps should be executed.
A9.1. Mark as “across” the remaining coherent link of Node2 (the one not marked as “down”),
and then as “up” the final link of Node2.
A9.2. Call fill_routing_table(Node2,rowi) for i=0 to 3 and then clear Node2’s RTD.
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
209
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
A9.3. Read the default link of Node3 and mark it as “across”. (Since Node0 has a “backbonefirst” routing table, the path of that read from Node0 to Node3 is by Node2.)1
A9.4. Set the remaining link of Node3 (the one not set just above or in Step A7.2) to “up”.
A9.5. Call fill_routing_table(Node3,rowi) for i=0 to 3 and then clear Node3’s RTD.
A9.6. Exit.
EXIT IF ONLY 4 NODES.
(If this point is reached, then there are more than four nodes, of which the ﬁrst four rows of Node0
and Node1 routing tables (Steps A4.3/A8 and A5, respectively) are complete. Node2 is partly
complete (Step A6): it has its NodeID of 2 and its link to Node0 has been marked “down”. The
NodeID ﬁeld has been written for Node3 and its “down” link has been marked (Step A7). Also, RTD
is still set for Node2 and Node3. The other two Node2 links need to be marked and routing table
initialized, and Node3 handling needs to be ﬁnished, before moving to other nodes (starting with Step
B1 below).)
A10. Call fill_routing_table(Node1,rowi) for i=4,5 and call fill_routing_table(Node0,row4).
A11. Finish marking the links of Node2, write its routing table, and clear its RTD.
A11.1. Mark the lower number of the remaining two links of Node2 as “across” and the other as
“up”. This is a guess.
A11.2. Call fill_routing_table(Node2,rowi) for i=0,3 and clear Node2’s RTD.
A11.3. Check the guess and modify it if necessary.
A11.3.1. Read the NodeID of Device 27. (Node2 will route this through the link
marked as “across”, which was a guess.)
A11.3.2. If the NodeID just read is 3, then the guess was correct; call
ﬁll_routing_table(Node2,rowi) for i=1,2,4,5. Otherwise, switch the marking of the
“across” and “up” links of Node2, and then call ﬁll_routing_table(Node2,rowi) for i=1
to 5. (Note that row 0 does not need to be modiﬁed either way. Also, if the read times
out, then the result will be 7. In that case, it can be assumed that the read was done
through the link pointing up.)
A12. Finish marking the links of Node3, write its routing table, and clear its RTD.
A12.1. Read the default link register of Node3. (Now that Node0’s and Node2’s routing tables
1. Steps A9.3 and A9.4 could possibly be replaced by a single step that marks the remaining coherent link of
Node3 as “across”. The following case: Node2 and Node3 both have 3 coherent links, and one of Node2’s
links is not connected, results in the four-node case. Similar cases are ignored in other places.
210
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
are correctly defined, that link connects across to Node2.) Mark that link as “across”, and
mark the remaining unmarked link of Node3 as “up”.
A12.2. Call fill_routing_table(Node3,rowi) for i=0 to 5 and clear Node3’s RTD.
(It is known at this point that there are more than four nodes. Nodes 0 through 3 have now received
their Node IDs and rows 0 through 5 of their routing tables, and their routing tables are enabled (i.e.,
their RTD bits are cleared). Row 5 of Node0’s routing table has not been written yet, since it can
change in Step B3.1.)
Block B: Complete Node4 and Node5
B1. Write the NodeID of Node4 and mark its “down” link.
B1.1. Write a NodeID of 4 to Device 28 (first_access).
B1.2. Read the default link of Device 28 (Node4) and mark it as “down”.
B2. If Node4 has only two coherent links, then complete both Node4 and Node5 and then exit.
B2.1. Mark the unmarked coherent link of Node4 as “across”, and then mark its remaining link
as “up”.
B2.2. Call fill_routing_table(Node4,rowi) for i=0 to 5 and clear Node4’s RTD.
B2.3. Call fill_routing_table(Node0,row5).
B2.4. Write a NodeID of 5 to Device 29 (first_access).
B2.5. Read the default link of Node5 (Device 29), and mark it as “across”.
B2.6. Mark the remaining coherent link of Node5 as “down” 2, and then mark the final link of
Node5 as “up”.
B2.7. Call fill_routing_table(Node5,rowi) for i=0 to 5 and clear Node5’s RTD.
B2.8. Exit.
EXIT IF ONLY 6 NODES.
(At this point, Nodes 0 through 3 are initialized, Node4 has its NodeID, the “down” link of Node4 is
marked, and Node4 is known to have three coherent links. Also, the RTD bit is set for Node4 and all
greater nodes. Row 5 of Node0’s routing table is not written.)
2. If Node5 has three active coherent links, which can happen if Node4 has three links marked as active coherent
links but the one “up” is not functioning, then a temporary routing table row change as in Step B3 can be
made.
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
211
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
B3. Write a NodeID of 5 to the node across from Node4 and mark that node’s “down” link. (The
“across” link of Node4 has not been marked yet, and messages are not routed through Node4 to
Node5.)
B3.1. Write row 5 of Node0’s routing table to point “across” for requests (so that requests from
Node0 to Device 29 will follow the path 0->1->3->N where N is the node “up” from
Node3.)3
B3.2. Read the default link of Device 29 (first_access) and mark that link as “down” for Node5.
B3.3. Write a NodeID of 5 to Device 29.
B4. Call fill_routing_table(Node0,rowi) for i=5,6,7 and call fill_routing_table(NodeN,rowi) for
N=1,2,3 and i=6,7.
B5. Finish marking the links of Node4, write its routing table, and clear its RTD. This step and the
next are completely analogous to Steps A11 and A12.
B5.1. Mark the lower number of the two unmarked links of Node4 as “across” and the other as
“up”. This is just a guess.
B5.2. Call fill_routing_table(Node4,rowi) for i=0,5 and clear Node4’s RTD.
B5.3. Check if the guess in Step B5.1 was correct, and if not, then correct it.
B5.3.1. Read the NodeID of Device 29. (Node4 will route this write through the link
marked as “across”, which was a guess.)
B5.3.2. If the NodeID read is 5, then the guess was correct; call
ﬁll_routing_table(Node4,rowi) for i=1 to 4 and for i=6,7. Otherwise, switch the marking of the “across” and “up” links of Node4, and then call
ﬁll_routing_table(Node4,rowi) for i=1 to 7. (Row 0 need not be re-written. And the
case of timing should be OK, as in Step 11.3.2.)
B6. Finish marking the links of Node5, write its routing table, and clear its RTD.
B6.1. Read the default link register of Node5. Mark that link as “across”, and mark the remaining unmarked link of Node5 as “up”.
B6.2. Call fill_routing_table(Node5,rowi) for i=0 to 7 and clear Node5’s RTD.
3. Step B3 approach will cause responses to take the path 5->3->2->0, which is not the reverse of the request
path from 0->1->3->5. That is acceptable, although the tracing will not function equally well. This can be
ﬁxed by modifying Node2’s routing table instead of Node 0’s, so that the request path is 0->2->3->5.
212
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
(It is known at this point that Node4 has 3 coherent links, so there are presumably 8 nodes.4 Rows 0 to
7 have been written for the routing tables of Nodes 0 to 5, and the RTD bits are clear for those nodes.)
Block C: Complete Node6 and Node7
C1. Write to Device 30 a NodeID of 6 (first_access).
C2. Read the default link of Device 30 (Node6) and mark it as “down”.
C3. There should be exactly one unmarked, active coherent link in Node6. Mark it as “across”, and
then mark the remaining link as “up”.5
C4. Call fill_routing_table(Node6,rowi) for i=0 to 7 and clear Node6’s RTD.
C5. Read the default link of Device 31 (first_access) and mark it as “across”. (NodeID does not have
to be written, since the reset microcode has already written a NodeID of 7.)
C6. There should only be one unmarked active coherent link on Node7. Mark this link as “down”, and
then mark the remaining link as “up”.
C7. Call fill_routing_table(Node7,rowi) for i=0 to 7 and clear Node7’s RTD.
Completion
At this point, all nodes have their Node IDs and routing tables. Also, it is known which NodeID was
last written; this is the maximum NodeID, M. Adjust the broadcast tables of nodes M and M-1 as it is
described in “Writing Routing Table Rows” on page 213. Then write every node’s NodeID register, in
particular its NodeCount.
During the execution of this algorithm, conﬁguration space accesses are sent to corresponding
coherent nodes, not to the compatibility bus. Bit LimitCldtCfg (Function 0, Offset 68h) is set near the
end to enable range based on node count. For example, if there are two nodes (Devices 24 and 25),
then accesses to Devices 26 through 31 will go to the compatibility bus.
Finally, the RequestDisable bit is cleared at each AP processor.
Before deciding the maximum NodeID and adjusting the broadcast tables, functionality of all links
must be veriﬁed.
8.4.1.1
Writing Routing Table Rows
Routine ﬁll_routing_table(NodeN,rowr) which writes a given row r of Node N’s routing table is
described in this section. The following assumption are made: the routing tables for all nodes on the
4. A check that there are 8 nodes can be done by checking that the node “up” from Node4 has two connected
coherent links. If the check fails, jump from Step C5 to 6-node case.
5. If the case that there is no unmarked active coherent link in Node6 is considered, the algorithm can be exited
with a 6-node system.
Chapter 8
HyperTransport™ Technology Conﬁguration and Enumeration
213
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
path up to Node N (not including Node N) have been written, and the Node N links have been
correctly marked as “up”, “down”, or “across”.
Statement “put a 1 in the P position”, means that if link L is associated with P at node N (where P is
“up”, “down”, or “across”), then bit position L+1 is assigned the value 1.
REQUEST TABLE. Row r of the “backbone-ﬁrst” request table is computed as follows.
If r = N, then the value is 0001b. Else if (r XOR 1) = N, then put a 1 in the “across” position. (This is
testing that r and N differ in the unit digits only, so they are across from each other.) Else if (r < N),
then put a 1 in the “down” position. Else put a 1 in the “up” position.
RESPONSE TABLE. For the response table, the following rule is used (“short-ﬁrst” or “rung-ﬁrst”
rather than “backbone-ﬁrst”).
If r = N, then the value is 0001b. Else if ((r XOR N) AND 0001b) != 0, i.e. r and N are in different
“rungs”, then put a 1 in the “across” position. Else if (r < N), then put a 1 in the “down” position. Else
put a 1 in the “up” position.
BROADCAST TABLE. The ﬁnal value for each row of Node N’s broadcast routing table must
comply with the following requirements:
1. If N is 0 or 1 then the “down” bit is 0.
2. Let M be the maximum node number. If N is M or M-1 then the “up” bit is 0. Note: Since M is not
always known at the time routing tables are written, this step should be performed at the end of
the process.
3. The least signiﬁcant bit “route to this node” is 1.
The rules for constructing row r of the broadcast table for Node N that should be applied before
ﬁnalizing the result of 1., 2., and 3. above. For the broadcast table, unlike the request and response
tables, r represents the source of the message, not the destination.
If r = N, then the value is 1111b. Else if ((r XOR N) AND 0001b) != 0, i.e., r and N are in different
“rungs”, then return 0001b. Else if r < N, put a 1 in the “up” and “across” positions. Else put a 1 in the
“down” and “across” positions.
214
HyperTransport™ Technology Conﬁguration and Enumeration
Chapter 8
26094
Rev. 3.06
9
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Power and Thermal Management
AMD Athlon™ 64 and AMD Opteron™ processors support ACPI-compliant power management for
all classes of systems from notebook PCs to multiprocessor servers. Table 40 lists various levels of
power management. Table 41 on page 216 lists ACPI-state support by system class. This chapter
covers processor-level power management and processor-speciﬁc aspects of system-level power
management.
Table 40.
Power Management Categories
Power Management Category
Comments
System Level
Coarse level power management applied to the entire system, typically after a predefined period of inactivity, or in response to a user
action such as pressing the system power button.
• System Sleep States (ACPI-defined S-states)
Processor Level
• Processor Power States (ACPI-defined C-states)
• Processor Performance States (ACPI-defined P-states)
• Software Transparent Power Management (Hardware enforced
throttling).
Device Level
• Device Power States (ACPI-defined D-states)
• Device Performance States
• Software Transparent Power Management.
Bus Level
• Bus Power States
• Software Transparent Power Management
Sub-system Level
Chapter 9
• Typically not used in PC systems, and beyond the scope of this
document.
Power and Thermal Management
215
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 41.
26094
Rev. 3.06
September 2003
ACPI-State Support by System Class
Mobile Systems
Uniprocessor
Desktop PCs
Multiprocessor
Systems
G0/S0/C0: Working
Yes
Yes
Yes
G0/S0/C0: Processor performance state (Pstate) transitions under OS control.
Yes
Yes1
Not with current
operating systems
G0/S0/C0: Thermal clock throttling (I/O hub
hardware enforced)
Yes
Yes
Revision B:No1
Revision C:Yes
G0/S0/C0: Thermal clock throttling (operating system enforced)
Yes
Yes
No
G0/S0/C1: Halt
Yes
Yes
Yes
G0/S0/C2: Stop Grant Caches snoopable
Yes1,2
No
No
G0/S0/C3: Stop Grant Caches not snoopable
Yes1,2
No
No
G1/S1: Stand By (Powered On Suspend)
Yes
Yes
Yes
G1/S3: Stand By (Suspend to RAM)
Yes
Yes
Revision B:No1
Revision C:Yes
G1/S4: Hibernate (Suspend to Disk)
Yes
Yes
Yes
G2/S5: Shut Down, Turn Off (Soft Off)
Yes
Yes
Yes
G3 Mechanical Off
Yes
Yes
Yes
ACPI State Support
Notes:
1. Not supported by some revisions of the silicon. Refer to Revision Guide for AMD Athlon™ 64 and
AMD Opteron™ Processors, order# 25759.
2. See “Advanced Programmable Interrupt Controller (APIC)” on page 181.
9.1
Stop Grant
The processor and chipset use HyperTransport™ technology STPCLK and Stop Grant system
management messages to sequence into all power management states except for Halt. These
messages carry a 3-bit System Management Action Field (SMAF) to differentiate the various reasons
for placing the processor into the Stop Grant state. Table 42 maps Stop Grant SMAF codes to ACPI
states and actions that the BIOS is required to conﬁgure the processor to take for each ACPI state.
216
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
Table 42.
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Required SMAF Code to Stop Grant Mapping
Reason for
STPCLK
Message
SMAF Field
of STPCLK
and Stop
Grant
Messages
Mechanism that Forces the I/O Hub to
Send the STPCLK Message
Processors Programmed
Response after Entering the
Stop Grant State
(See Table 52 Power
Management Control
Registers.)
C2 Stop Grant
Caches
Snoopable
000b
Sent in response to a read of the ACPIdefined P_LVL2 register.
Ramp the CPU clock grid down
when no probes need to be serviced. (CPU Low Power Enable)
C3 Stop Grant
Caches not
Snoopable.
Used only by
mobile systems.
001b
Sent either in response to a read of the
ACPI-defined P_LVL3 register or in
response to a write to the Link Frequency
Change and Resize LDTSTOP_L Command register in the I/O hub.
Ramp the CPU clock grid frequency down. (CPU Low Power
Enable)
Sent either in response to a VID/FID
Change special cycle from the processor
or in response to a write to the Link Frequency Change and Resize LDTSTOP_L
Command register.
After LDTSTOP_L assertion,
place memory into self-refresh,
ramp the processor clock grids
down, then drive new FID values to PLL. (CPU Low Power
Enable, Northbridge Low Power
Enable, FID/VID Change
Enable)
VID/FID Change 010b
Or
Link Width/
Frequency
changes
After LDTSTOP_L is asserted,
place memory into self-refresh
and ramp down the processor
host bridge and memory controller clock grid. (Northbridge
Low Power Enable)
S1 Sleep state
011b
Sent in response to writing the S1 value
to the SLP_TYP[2:0] field and setting the
SLP_EN bit of the ACPI-defined PM1
control register in the I/O hub.
Same response as C3.
S3 Sleep state
100b
Sent in response to writing the S3 value
to the SLP_TYP[2:0] field of the PM1
control register.
Same response as C3. Additionally, after LDTSTOP_L has
been asserted, the I/O hub will
power off the main power
planes.
Throttling
101b
Occurs based upon hardware initiated
thermal throttling or ACPI-controlled
throttling.
Ramp down the CPU grid by
the programmed amount. (CPU
Low Power Enable)
S4/S5
110b
Sent in response to writing the S4 or S5
value to the SLP_TYP[2:0] field of the
PM1 control register.
Same response as S3 for processor. Additionally, all power
will be removed from processor
during S4 and S5.
Chapter 9
Power and Thermal Management
217
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 42.
Reason for
STPCLK
Message
Rev. 3.06
September 2003
Required SMAF Code to Stop Grant Mapping (Continued)
SMAF Field
of STPCLK
and Stop
Grant
Messages
Reserved for use 111b
by processor
9.2
26094
Mechanism that Forces the I/O Hub to
Send the STPCLK Message
No I/O hub STPCLK message uses this
SMAF code. The power management
register field corresponding to this SMAF
code is used by the processor in
response to executing the Halt instruction.
Processors Programmed
Response after Entering the
Stop Grant State
(See Table 52 Power
Management Control
Registers.)
Same response as C2, except a
Halt special cycle is broadcast.
C-States
As Table 40 on page 215 indicates, the processor supports ACPI-deﬁned processor power states,
which are referred to as C-states. Table 41 on page 216 indicates which C-states are supported by
system class. The operating system will place the processor into C-states when the processor is idle
an operating-system-determined percentage of the time.
9.2.1
C1 Halt State
C1 is the Halt state. C1 is entered after the HLT instruction has been executed. The BIOS conﬁgures
the processor to enter a low-power state during C1, in which the CPU core clock grid is ramped down
from its operating frequency. See Table 52 on page 242.
9.2.2
C2 and C3
C2 is the Stop Grant state in which the processor caches can be snooped. When the processor is in the
Stop Grant state and there is no probe activity to the processor’s caches, the CPU core clock grid is
ramped down from its operating frequency. For more information, see Table 52 on page 242.
C3 is the Stop Grant state in which the processor’s caches cannot be snooped. The chipset asserts
LDTSTOP_L during the C3 state. The chipset may require BIOS to enable LDTSTOP_L assertion
for the C3 state. After LDTSTOP_L assertion, the processor’s system memory is placed into selfrefresh mode and the processor’s clock grids, including the host bridge/memory controller clock grid,
are ramped down from their operational frequency. For more information, see Table 52 on page 242.
9.3
Throttling
The BIOS must declare a DUTY_WIDTH value of zero in the Fixed ACPI Description table if the
system does not support the _PSV object as part of a processor ACPI thermal zone. Throttling is not
218
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
supported by some revisions of the silicon. Refer to AMD Revision Guide for AMD Athlon™ 64 and
AMD Opteron™ Processors, order# 25759.
Chipsets must not be conﬁgured to assert LDTSTOP_L during throttling.
9.4
Processor ACPI Thermal Zone
This section will be part of a future revision of this document.
9.5
Processor Performance States
Processor performance states (P-states) are valid operating combinations of processor core voltage
and frequency. The hardware supporting P-state transitions in AMD processors is referred to as
AMD PowerNow!™ technology for mobile systems and AMD Cool’n’Quiet™ technology for
desktop systems. In this document all descriptions of AMD PowerNow!™ technology, software and
driver apply to AMD Cool’n’Quiet™ technology, software, and driver.
•
For operating systems without native support for P-state transitions (legacy operating systems)
using Revision C and subsequent processors, AMD PowerNow! software is used to perform Pstate transitions. In this case, a BIOS generated Performance State Block (PSB) is used by AMD
PowerNow! software to determine what P-states are supported by the processor in the system.
•
For operating systems that have native support for P-state transitions using Revision C and
subsequent processors, a processor-speciﬁc driver is used by the operating system to make P-state
transitions. In this case, BIOS-provided ACPI-deﬁned P-state objects inform the operating system
that P-state transitions are supported by the system, which P-states are supported, and when given
P-states are available for use by the operating system.
The processor data sheet speciﬁes the valid P-states for each processor. The BIOS vendor must only
use P-states deﬁned in that document for a given processor when building the ACPI P-state objects
and PSB for use by higher-level software.
Processors that support P-state transitions provide MSR C001_0041h (FIDVID_CTL) and MSR
C001_0042h (FIDVID_STATUS) for initiating P-state transitions and verifying that they are
complete. If these two MSRs are accessed in a processor that does not support P-state transitions, then
a general protection fault occurs.
Chipsets provide support for P-state transitions. Refer to the chipset documentation for information
on chipset conﬁguration and P-state-related considerations.
9.5.1
BIOS Requirements for P-State Transitions
P-state transitions can be used only if they are supported by the processor and by the system.
BIOS is required to:
Chapter 9
Power and Thermal Management
219
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
•
Determine that the processor supports P-state transitions.
•
Have a valid set of P-states for the processor, based on the processor data sheet.
•
Take chipset or system limitations into account before providing the ACPI-deﬁned P-state objects
described in 9.6 on page 230, or the PSB described in 9.7 on page 239 of this document.
•
If P-states are supported and a Revision C or higher revision processor is present, BIOS is
required to provide both:
– ACPI-deﬁned P-state objects for operating systems that support native P-state control.
– a PSB in support of AMD PowerNow! software with legacy operating systems.
The BIOS must not provide the ACPI P-state objects or the PSB if:
•
The processor does not support P-state transitions.
•
The processor is Revision B.
•
The chipset or system does not support P-state transitions.
•
The BIOS does not have a set of valid P-states derived from the processor data sheet.
9.5.1.1
Step 1: Determine Processor Support for P-State Transitions
The BIOS determines if a processor supports P-state transitions by executing the CPUID extended
function 8000_0007h.
If the value returned in EDX[2:1] is:
•
11b, then the part is capable of performing P-state transitions.
•
00b, then the part is not capable of performing P-state transitions.
Refer to the following documents regarding CPUID and determining which processor is in the
system:
•
AMD Processor Recognition Application Note, order# 20734
•
CPUID Guide for the AMD Athlon™ 64 and AMD Opteron™ processors, order# 25481
9.5.1.2
Step 2: Match the Processor to a Valid Set of P-States
If CPUID extended function 8000_0007h returned EDX[2:1] = 11b, the BIOS must correlate the
processor in the system to a valid set of P-states from the processor data sheet. Each processor is
uniquely identiﬁed by the following:
•
CPUID extended function 8000_0001h:
– Family number and extended family number (when family number ﬁeld = 1111b).
– Model number and extended model number (when family number ﬁeld = 1111b).
– Stepping (Revision).
220
Power and Thermal Management
Chapter 9
26094
•
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
FIDVID_STATUS MSR (C001_0042h):
– MaxVID: The nominal voltage for the maximum P-state is the voltage selected by MaxVID
minus the ramp voltage offset (RVO).
– StartVID: This corresponds to the nominal voltage at which the processor starts operation after
a cold reset.
– MaxFID: This corresponds to the frequency that the processor operates at when in the maximum
performance state.
– StartFID: This corresponds to the frequency that the processor starts operation at after a cold
reset.
These seven parameters are used by the BIOS to match the processor in the system to a valid set of Pstates (voltage and frequency combinations) listed in the processor data sheet.
9.5.1.3
Step 3: Determine System Support for P-State Transitions
The BIOS software must also determine whether the system supports P-state transitions. The BIOS
must take into account chipset/motherboard components and potential system side effects associated
with P-state transitions that could preclude the use of P-state transitions.
9.5.2
BIOS-Initiated P-State Transitions
Processors that do not support P-state transitions and processors intended for desktop systems power
up to the maximum P-state. Processors intended for mobile systems power up in the minimum Pstate. BIOS can transition mobile processors from the minimum P-state to a higher P-state during the
POST routine. If BIOS performs a P-state transition it must follow the P-state transition algorithm
deﬁned in 9.5.5 on page 222. The BIOS never initiates P-state transitions after passing control to the
operating system.
9.5.3
BIOS Support for Operating System/CPU Driver-Initiated P-State
Transitions
Operating systems that have native control for processor P-states exclusively use the presence of
BIOS provided ACPI 2.0 deﬁned processor P-state objects to determine if processor P-states are
supported in a system. If the ACPI objects do not exist, then the processor driver is not loaded.
In operating systems that do not have native support for processor P-state transitions, an AMD
deﬁned PSB provided by the BIOS informs AMD PowerNow! technology software that P-state
transitions are supported.
Once it has been determined that a processor and system support P-state transitions, the BIOS must
provide both:
•
ACPI-2.0-deﬁned processor P-state objects for operating systems with native P-state support. See
Section 9.6 on page 230.
Chapter 9
Power and Thermal Management
221
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
•
26094
Rev. 3.06
September 2003
AMD-deﬁned PSB for use with AMD PowerNow! software and operating systems that do not
have native P-state support. See Section 9.7 on page 239.
9.5.4
Processor Driver Requirements
The processor driver that supports native operating system policy and control of P-state objects is
required to use the ACPI 2.0 P-state objects as deﬁned in this document. The processor driver is not
permitted to use the legacy PSB tables described in section 9.7 on page 239.
9.5.5
P-State Transition Sequence
This section describes the P-state transition algorithm for the AMD Athlon™ 64 and
AMD Opteron™ processors.
9.5.5.1
FID to VCO Frequency Relationship
Table 43 and Table 44 provide the core frequency-to-VCO frequency relationship for
AMD Athlon™ 64 and AMD Opteron™ processors. These processors do not support frequency
transitions in VCO frequency steps greater than 200 MHz. The processor driver, and the
AMD PowerNow! driver use these tables when making P-state transitions.
Only one P-state is allowed in Table 43. The one P-state consists of a frequency from the “Minimum
Core Frequency” column in Table 43 and is the minimum P-State supported by the processor. For
processors that support P-state transitions, the processor data sheet has a table of valid P-states based
on ordering part number that dictates which frequency in Table 43 is used for the minimum P-state
supported by the processor.
Table 43 additionally deﬁnes Portal Core Frequencies. The “Portal Core Frequencies in Table 44”
column of Table 43 deﬁnes:
•
The core frequencies in Table 44 to which direct transitions can be made from the minimum core
frequency in Table 43.
•
The core frequencies in Table 44 that support transitions to the minimum core frequency in
Table 43.
The processor supports direct transitions between the minimum core frequency in Table 43 and any of
the core frequencies from Table 44 listed in the Portal Core Frequencies column associated with the
minimum core frequency.
If the minimum P-state from the low table (Table 43) has a VCO frequency > 1600 MHz, then the
VCO/core frequency of all P- states in the high table (Table 44) must be >= the VCO frequency of the
minimum P-state minus 200 MHz.
222
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Example 1:
Given a processor with a minimum core frequency of 800 MHz and a maximum core frequency of
2000 MHz, to transition from 800 MHz to 2000 MHz, the processor driver will (in the order
speciﬁed):
1. Transition the processor from the 800 MHz core frequency to 1800 MHz core frequency.
1800 MHz is a portal frequency to and from 800 MHz as deﬁned in Table 43, because the VCO
frequency change from an 800 MHz core frequency to an 1800 MHz core frequency is <= 200
MHz.
2. Transition the processor core frequency from 1800 MHz to 2000 MHz (VCO frequency change is
<= 200 MHz according to Table 44).
Note: To simplify this example, the voltage transitions and isochronous relief times are not
described in this example.
Example 2:
Given a processor with a minimum core frequency of 1400 MHz and a maximum core frequency of
3400 MHz, to transition from 1400 MHz to 3400 MHz, the processor driver will (in the order
speciﬁed):
1. Transition the processor from the core frequency of 1400 MHz to a core frequency of 3000 MHz.
3000 MHz is a portal frequency to and from 1400 MHz as deﬁned in Table 43, because the VCO
frequency change from 1400 MHz core frequency to 3000 MHz core frequency is <= 200 MHz.
2. Transition the processor core frequency from 3000 MHz to 3200 MHz (VCO frequency change is
<= 200 MHz per Table 44).
3. Transition the processor core frequency from 3200 MHz to 3400 MHz (VCO frequency change is
<= 200 MHz per Table 44).
Note: Note: to simplify this example, the voltage transitions and isochronous relief times are not
described in this example.
Table 43.
Low FID Frequency Table (< 1600 MHz)
Portal Core
Frequencies in
Table 44
FID[5:0]
Minimum Core
Frequency MHz
VCO Frequency
MHz
000000b
800
1600
1600, 1800
000010b
1000
2000
1800, 2000, 2200
000100b
1200
2400
2200, 2400, 2600
000110b
1400
2800
2600, 2800, 3000
Chapter 9
Power and Thermal Management
223
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 44.
Rev. 3.06
September 2003
High FID Frequency Table (>= 1600 MHz)
FID[5:0]
Core Frequency
MHz
VCO Frequency
MHz
001000b
1600
1600
001010b
1800
1800
001100b
2000
2000
001110b
2200
2200
010000b
2400
2400
010010b
2600
2600
010100b
2800
2800
010110b
3000
3000
011000b
3200
3200
011010b
3400
3400
011100b
3600
3600
011110b
3800
3800
100000b
4000
4000
100010b
4200
4200
100100b
4400
4400
100110b
4600
4600
101000b
4800
4800
101010b
5000
5000
9.5.5.2
26094
P-state Transition Algorithm
The P-state transition algorithm has three phases. During phase 1 the processor voltage is transitioned
to the level required to support frequency transitions. During phase 2 the processor frequency is
transitioned to frequency associated with the OS-requested P-state. During phase 3 the processor
voltage is transitioned to the voltage associated with the OS-requested P-state.
Figure 4 shows the high-level P-state transition ﬂow. Figure 5 is a high-level timing diagram depicting
voltage and frequency stepping associated with each phase of a P-state transition. Figure 6 shows the
core voltage transition ﬂow for phase 1 of a P-state transition. Figure 7 shows the core frequency
transition ﬂow for phase 2 of a P-state transition. Figure 8 shows the core voltage transition ﬂow for
phase 3 of a P-state transition.
The algorithm shown in Figure 4 through Figure 8 describes the 3-phases of a P-state transition in the
context of Windows XP and the associated processor driver, but the algorithm also applies to the
AMD PowerNow! driver and any other software that performs P-state transitions (such as a debug
tool).
224
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
OS: determines a P-state transition is required and calls the
processor driver to make the transition to a specified P-state.
Processor driver: informs the OS that the requested P-state
transition is complete, and begins the transition.
Phase 1:
Processor driver: perform a series of VID-only transitions each increasing the
voltage by the maximum voltage step (MVS) to change the voltage:
• To the voltage for the requested P-state plus the voltage indicated by the ramp
voltage offset (RVO), if the requested P-state is greater in frequency than the
current P-state.
• To the voltage indicated by the CurrVID plus the voltage indicated by the RVO,
if the requested P-state is lower in frequency than the current P-state.
Software counts off voltage stabilization time (VST) after each voltage step.
Phase 2:
Processor driver: changes the processor core frequency to
the frequency for the OS-requested P-state.
Phase 3:
Processor driver: performs a VID-only change to the VID for
the OS requested P-state.
Figure 4.
Chapter 9
High-Level P-state Transition Flow
Power and Thermal Management
225
00011
00010
00010
00000
00001
1.525V
VDD
1.500V
1.475V
MVS = 25 mV
1.45V
VST = 100µs
VST = 100µs
1.550V
RVO
removed.
VST = 100µs
1.500 V
VST = 100µs
FID[5:0]
000000
Core Frequency
(MHz)
800
VCO Frequency
1600
(MHz)
Current
P-State
800 [email protected]
001010
001100
1800
2000
1800
2000
StpGntTOCnt
Phase 1:
Increase the voltage.
Add RVO
StpGntTOCnt
Phase 2:
Change frequency.
Processor driver
counts IRT.
VST = 100µs
Processor driver
counts IRT.
Power and Thermal Management
Example P-State Transition Timing Diagram
1.550V
Current
P-State
2000 [email protected]
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Figure 5.
226
00100
VID[4:0]
Phase 3:
Transition VID
to remove RVO
26094
Rev. 3.06
Chapter 9
September 2003
Not drawn to scale.
This figure depicts a P-state transition from a P-state of 800 MHz at 1.45 V to a P-state of 2000 MHz at 1.50 V.
For MVS see section 9.6.2.1.2.
For StpGntTOCnt see section 9.5.5.2.2.
For IRT see section 9.6.2.1.5.
For RVO and VID see section 9.6.2.1.4.
For VST see section 9.6.2.1.1.
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Is the CUR_VID <= the VID for the
P-state requested by the OS?
No
Yes
Increase the core voltage by the amount
indicated by RVO (see section 9.6.2.1.4).
Yes
Is RVO > 0?
Initiate a VID-only transition to
decrease the VID code by the
amount indicated by MVS (see
section 9.6.2.1.2). This transition
is made by following the steps in
section 9.5.5.2.1.
Set a software count =RVO
(RVO_STEPS)
Count off the VST
(see Section 9.6.2.1.1)
Is RVO_STEPS = 0?
Yes
No
Initiate a VID-only transition
to decrease the VID code
by 1. This transition is made
by following the steps in
section 9.5.5.2.1
Count off the VST
(see section 9.6.2.1.1).
Decrement RVO_STEPS.
Figure 6.
Chapter 9
Phase 1: Core Voltage Transition Flow
Power and Thermal Management
227
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Referring to the FID Frequency tables, is
the VCO frequency change between
CurrFID and the FID associated with the
OS-requested P-state <= 200 MHz?
26094
Rev. 3.06
September 2003
No
Is the OS requested
FID > CurrFID?
Yes
No
Yes
Is CurrFID > the FID values in Table 43, “Low
FID Frequency Table”?
Initiate a FID-only transition to the FID
associated with the OS-requested P-state.
No
Yes
No
Read the FIDVID_STATUS MSR. Is
the FidVidPending bit = 0?
Initiate a FID only
transition to the
highest portal
frequency that
corresponds to
CurrFID.
Yes
Processor driver counts off the
isochronous relief time (IRT)
specified by the _PSS object.
Initiate a FID only
transition to the next
higher FID in the High
FID Frequency table.
Initiate a FID only
transition to the next
lower FID in the High
FID Frequency table.
Core frequency is now at the
frequency requested by the OS.
No
Read the FIDVID_STATUS MSR.
Is the FidVidPending bit = 0?
Yes
Processor driver counts off
the isochronous relief time.
Figure 7.
228
Phase 2: Core Frequency Transition Flow
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Initiate a VID-only transition directly to the VID code
for the P-state requested by the OS. This transition
is made by following the steps in section 9.5.5.2.1.
Count off the VST (See section 9.6.2.1.1).
Figure 8.
Phase 3: Core Voltage Transition Flow
9.5.5.2.1 Changing the VID
Note: Software must hold the FID constant when changing the VID.
To change the processor voltage:
1. Write the following values to FIDVID_CTL (MSR C001_0041h):
– NewVID ﬁeld (bits 12–8) with the VID code associated with the target voltage
– NewFID ﬁeld (bits 5–0) with the CurrFID value indicated in the FIDVID_STATUS MSR
– StpGntTOCnt ﬁeld (bits 51–32) with 1h, which corresponds to 5 nsec. For Revision B
processors, this setting results in voltage transitions causing the minimum possible memory
access latency. This ﬁeld has no effect on VID changes for Revision C AMD Athlon™ 64 and
AMD Opteron™ processors, since the processor drives the VID[4:0] in response to the MSR
write without the processor ﬁrst being placed into the Stop Grant state.
– InitFidVid bit (bit 16) set to 1. Setting this bit initiates the VID change.
– Clear all other bits to 0.
2. Loop on reading the FidVidPending bit (bit 31) of FIDVID_STATUS (MSR C001_0042h) until
the bit returns 0. The FidVidPending bit stays set to 1 until the new VID code has been driven to
the voltage regulator.
9.5.5.2.2 Changing the FID
The AMD Athlon™ 64 and AMD Opteron™ processors support direct FID transitions only between
FIDs that represent a VCO frequency change less then or equal to 200 MHz. To change between FIDs
that represent a VCO frequency change greater than 200 MHz, software must make multiple
transitions each less than or equal to a 200 MHz change in VCO frequency. For information about
allowable FID transitions, see Tables 43 and 44 on page 224.
Chapter 9
Power and Thermal Management
229
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Note: Software must hold the VID constant when changing the FID.
To change the core frequency:
1. Write the following values to FIDVID_CTL (MSR C001_0041h):
– NewVID ﬁeld (bits 12–8) with the CurrVID value reported in the FIDVID_STATUS MSR.
– NewFID ﬁeld (bits 5–0) with the FID code associated with the target frequency.
– StpGntTOCnt ﬁeld (bits 51–32) with a value corresponding to the processor PLL lock time.
The PLL lock time is speciﬁed by the processor data sheet and refers to all components of PLL
lock including frequency lock, phase lock, and settling time. PLL lock time is communicated
to the processor driver through the PLL_LOCK_TIME ﬁeld of the _PSS object. PLL lock time
is in microseconds. To translate the PLL lock time to an StpGntTOCnt value, multiply
PLL_LOCK_TIME by 1000 to get nanoseconds, then divide by 5 which is the clock period of
the counter in ns. Therefore, StpGntTOCnt value = PLL_LOCK_TIME ∗1000/5. For example,
a PLL lock time of 2 µs results in an StpGntTOCnt value of 400 decimal and 190h.
– InitFidVid bit (bit 16) set to 1. Setting this bit initiates the P-state transition.
– Clear all other bits to 0.
2. Loop on reading the FidVidPending bit (bit 31) of FIDVID_STATUS (MSR C001_0042h) until
the bit returns 0. The FidVidPending bit stays set to 1 until the new FID code is in effect.
Figure 7 on page 228 illustrates the transition ﬂow for core frequencies, including the requirement to
wait the isochronous relief time between FID steps.
9.6
ACPI 2.0 Processor P-State Objects
For systems that support P-state transitions, the BIOS must provide the following ACPI 2.0 P-state
objects to support operating systems that provide native support for processor P-state transitions.
These subsections discuss the speciﬁc implementation of these objects.
9.6.1
_PCT (Performance Control)
The _PCT object is generically described in section 8.3.3.1 of the ACPI 2.0 speciﬁcation. The _PCT,
_PSS, and _PPC objects (deﬁned in the next sections) are all placed after the Processor object in the
\_PR name space for operating systems that have native support for processor P-states. This section
provides BIOS speciﬁcs regarding the use of _PCT.
The _PCT object speciﬁes the control register and status register used to initiate and verify P-state
transitions. The processor’s P-state transition protocol is abstracted from the operating system by the
processor driver, so this object does not declare the MSRs used for FID_Change protocol, but rather
declares these two registers to be functional ﬁxed hardware. Functional ﬁxed hardware means that the
mechanism is speciﬁc to the processor, and the processor driver knows how to initiate and verify
P-state transitions because the processor driver is written by the processor vendor or written based
upon documentation supplied to the operating system vendor by the processor vendor.
230
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
The following is a sample _PCT object. Sections referenced are in the ACPI 2.0 speciﬁcation and
ACPI 2.0 Errata Document, revision 1.4.
Name (_PCT, Package (2)
// Performance Control Object
{
//
ResourceTemplate () {Register(FFixedHW, 0, 0, 0)}
// Control
Buffer () {
0x82,
// B0-Generic Register Descriptor (Sec. 6.4.3.7)
0xC,0,
// B1:2-length (from _ASI through _ADR fields)
0x7F,
// B3-Address space ID, _ASI, SystemIO
0,
// B4-Register Bit Width, _RBW
0,
// B5-Register Bit Offset, _RBO
0,
// B6-Reserved. Must be 0.
0,0,0,0,0,0,0,0, // B7:14-register address, _ADR (64bits)
0x79,0},
// B15:16-End Tag (Section 6.4.2.8)
//
ResourceTemplate () {Register(FFixedHW, 0, 0, 0)}
// Status
Buffer () {
0x82,
// B0-Generic Register Descriptor (Section
6.4.3.7)
0xC,0,
0x7F,
0,
0,
0,
0,0,0,0,0,0,0,0,
0x79,0},
}) // End of _PCT object
9.6.2
//
//
//
//
//
//
//
B1:2-length (2 bytes)
B3-Address space ID, _ASI, SystemIO
B4-Register Bit Width, _RBW
B5-Register Bit Offset, _RBO
B6-Reserved. Must be 0.
B7:14-register address, _ADR (64bits)
B15:16-End Tag (Section 6.4.2.8)
_PSS (Performance-Supported States)
The _PSS object is generically described in Section 8.3.3.2 of the ACPI 2.0 speciﬁcation. This object
follows the _PCT object in the \_PR name space in support of operating systems with native support
for processor P-states. This section provides the BIOS speciﬁcs regarding use of _PSS.
The _PSS object speciﬁes the performance states that are supported by the system.
Based upon the maximum voltage and frequency supported by the processor and system
requirements, the BIOS will build the _PSS object to specify the performance states supported by the
system.
The BIOS creates the _PSS object based on the table of valid P-states in the processor data sheet. The
BIOS is required to maintain a set of Performance State Tables (PSTs) based on the processor data
sheet for all processors capable of P-state transitions. Only the PST entries that correspond to the
processor found in the speciﬁc system by the BIOS during the POST routine are used by the BIOS to
create the _PSS object for that system. The _PSS object could have as many performance states as the
processor data sheet indicates are supported. For example, the _PSS object could provide
performance state deﬁnitions P0–P2:
Chapter 9
Power and Thermal Management
231
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
P0: 2400 MHz at 1.20 V
P1: 1600 MHz at 1.10 V
P2: 800 MHz at 1.00 V
The generic _PSS object description has the following format:
Name (_PSS, Package()
{
// Field Name Field Type
Package () // Performance State 0 Definition – P0
{
CoreFreq,
// DWordConst
Power,
// DWordConst
TransitionLatency,
// DWordConst
BusMasterLatency, // DWordConst
Control,
// DWordConst
Status
// DWordConst
},
Package ()
// Performance State n Definition – Pn
{
CoreFreq,
// DWordConst
Power,
// DWordConst
TransitionLatency,
// DWordConst
BusMasterLatency, // DWordConst
Control,
// DWordConst
Status
// DWordConst
}
})
// End of _PSS object
Each performance state entry contains six data ﬁelds as follows:
•
CoreFreq is the core CPU operating frequency (in MHz).
•
Power is the typical power dissipation (in milliWatts). The BIOS must ﬁll out this ﬁeld for the
maximum performance state as speciﬁed in the processor data sheet. Estimates for power
consumption for lower P-states are acceptable.
•
TransitionLatency is the worst-case latency, in microseconds, that the CPU is unavailable during
a transition from any performance state to this performance state. During a P-state transition, the
CPU is unavailable for no more than 6 µs for each frequency step. While the total P-state
transition could take up to 2 ms, the operating system is not blocked during the transition.
Therefore the recommended value for this ﬁeld is 100 µs.
•
BusMasterLatency is the worst-case latency, in microseconds, that Bus Masters are prevented
from accessing memory during a transition from any performance state to this performance state.
This value is estimated at 7 µs.
232
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
•
Control is the value to be written to the _PSS Control Field in order to initiate a transition to the
performance state. The BIOS must ﬁll this out as described on page 233.
•
Status is the value that the processor driver can compare to a value read from the _PSS Status
Field to ensure that the transition to the performance state was successful. The BIOS must ﬁll this
out as described on page 236.
9.6.2.1
_PSS Control Field
The control ﬁeld of each performance state deﬁnition within the _PSS object contains the information
that the processor driver uses. This includes NewVID and NewFID values that must be written into
the FIDVID_CTL register (MSR C001_0041) to initiate transitions between P-states. _PSS Control
ﬁeld has the following format:
IRT
Res
31 30 29 28 27 26
RVO
Bits
20 19 18 17
PLL_LOCK_TIME
MVS
Field
VST
6
NewVID
5
0
NewFID
Description
31–30
IRT
Isochronous Relief Time
29–28
RVO
Ramp Voltage Offset
27
11 10
Reserved
Must be declared as zero.
26–20
PLL_LOCK_TIME
Phase-Locked Loop Time
19–18
MVS
Maximum Voltage Step
17–11
VST
Voltage Stabilization Time
10–6
NewVID
The VID value associated with the OS-requested P-state
5–0
NewFID
The FID value associated with the OS-requested P-state
The following sections describe the sub-ﬁelds of the _PSS control ﬁeld.
9.6.2.1.1 Voltage Stabilization Time
The Voltage Stabilization Time (VST) deﬁnes the number of microseconds (in 20 µs increments) that
are required for the voltage to increase by the amount speciﬁed by the MVS ﬁeld when the VID
outputs of the processor transition. The processor driver counts VST between VID steps that increase
the processor’s core voltage, not between steps that decrease it. Table 45 shows the actual
stabilization time for several VST values.
Table 45.
Sample VST Values
VST (Bits 17–11) Stabilization Time
7'h00
0 µs
7'h01
20 µs
...
Chapter 9
Power and Thermal Management
233
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 45.
26094
Rev. 3.06
September 2003
Sample VST Values
VST (Bits 17–11) Stabilization Time
7'h05
100 µs (BIOS default. This is the value used by all platforms.)
…
7'h64
2000 µs
…
7'h7e
2520 µs
7'h7f
2540 µs
9.6.2.1.2 Maximum Voltage Step
The Maximum Voltage Step (MVS) deﬁnes the maximum voltage increment the processor driver can
use when changing from a lower voltage to a higher voltage. For the processor driver, when
increasing core voltage, the next VID = CurrVID – 2MVS. Table 46 on page 234 shows the values and
increments for this ﬁeld.
For example, if the voltage is being increased, and
•
the current VID is 00100 (1.45 V),
•
the OS-requested target VID is 00000 (1.55 V),
•
and MVS is 0 (which selects 25mV voltage steps).
Then, the processor driver calculates the ﬁrst VID to which to transition as follows:
Next_VID = CurrVID - 2MVS = 00100 - 20 = 00100 - 1 = 00011 = 1.475 V
Next the processor driver transitions the VID to select 1.475 V.
After the VST is counted off, the processor driver calculates the next VID to which to transtion as
follows:
Next_VID = CurrVID - 2MVS = 00011 - 20 = 00010 = 1.500 V
This process iterates until the CurrVID is equal to the OS-requested target VID minus the VID
associated with the ramp voltage offset. After each VID transition the processor driver counts off the
voltage stabilization time before calculating the Next_VID.
Table 46.
MVS Values
MVS (Bits 19–18) Increment
234
00
25 mV (BIOS default. This is the required value
that must be used.)
01
50 mV (Reserved for future use)
10
100 mV (Reserved for future use)
11
200 mV (Reserved for future use)
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
9.6.2.1.3 PLL Lock Time
The 7-bit binary PLL_LOCK_TIME value deﬁnes the time required for the processor PLLs to lock in
microseconds. The PLL_LOCK_TIME value is speciﬁed in the processor data sheet.
9.6.2.1.4 Ramp Voltage Offset
The Ramp Voltage Offset (RVO) is the offset voltage applied to the nominal voltage when performing
P-state transitions. The RVO is speciﬁed in the processor data sheet. Table 47 lists the RVO values
and their corresponding offset voltages.
Table 47.
RVO Values
Offset
Voltage
Description
00
0 mV
Target P-state VID needs no adjustment for RVO.
01
25 mV
Decrement the target P-state VID by 1 to add the RVO.
10
50 mV
Decrement the target P-state VID by 2 to add the RVO. (BIOS default.
This is the required value.)
11
75 mV
Decrement the target P-state VID by 3 to add the RVO.
RVO (Bits 29–28)
For the processor driver, RVO is in VID increments. To increase the processor voltage the processor
driver must subtract the RVO ﬁeld from the VID. The MVS ﬁeld dictates the increments in which
RVO can be subtracted from VID.
For example, consider the case when the VID is 00100b (1.45 V), RVO is 10b (50 mV), MVS is 00b
(25 mV), and VST is 7’h05 (100 µs). Based upon these parameters, the voltage must be increased by
50 mV in two steps of 25 mV with a 100 µs delay after each VID change. The steps to apply the RVO
are:
1.
2.
3.
4.
The VID code is reduced to 00011b. This transitions the voltage from 1.45 V to 1.475 V.
The processor driver waits 100 µs as speciﬁed by VST.
The VID code is reduced to 00010b. This transitions the voltage from 1.475 V to 1.5 V.
The processor driver waits 100 µs as speciﬁed by VST.
Table 48 lists the VID codes and their corresponding voltages for the AMD Athlon™ 64 and
AMD Opteron™ processors.
Table 48.
VID Code Voltages
VID Code
(Bits 4–0)
Voltage
VID Code
(Bits 4–0)
Voltage
00000
1.550 V
10000
1.150 V
00001
1.525 V
10001
1.125 V
00010
1.500 V
10010
1.100 V
Chapter 9
Power and Thermal Management
235
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 48.
26094
Rev. 3.06
September 2003
VID Code Voltages (Continued)
VID Code
(Bits 4–0)
Voltage
VID Code
(Bits 4–0)
Voltage
00011
1.475 V
10011
1.075 V
00100
1.450 V
10100
1.050 V
00101
1.425 V
10101
1.025 V
00110
1.400 V
10110
1.000 V
00111
1.375 V
10111
0.975 V
01000
1.350 V
11000
0.950 V
01001
1.325 V
11001
0.925 V
01010
1.300 V
11010
0.900 V
01011
1.275 V
11011
0.875 V
01100
1.250 V
11100
0.850 V
01101
1.225 V
11101
0.825 V
01110
1.200 V
11110
0.800 V
01111
1.175 V
11111
Off
9.6.2.1.5 Isochronous Relief Time
The Isochronous Relief Time (IRT) is the amount of time the processor driver must count after each
FID change step in a given P-state transition. During each FID change step, system memory is
inaccessible to bus masters. While the processor driver is counting IRT between FID change steps,
bus masters have access to system memory. Table 49 lists the IRT values and their corresponding
times.
Table 49.
IRT Values
IRT (Bits 31–30)
Time
00
10 µs
01
20 µs
10
40 µs
11
80 µs (BIOS default)
9.6.2.2
_PSS Status Field
The status ﬁeld of each performance state deﬁnition within the _PSS object contains the information
required by the processor driver to verify that the transition has completed based upon a read of the
FIDVID_STATUS MSR. The _PSS Status ﬁeld is 32 bits, and the BIOS must map the
FIDVID_STATUS MSR CurrVID and CurrFID information to the _PSS status ﬁeld as shown in
Table 50 on page 237.
236
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
:
Table 50.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
_PSS Status Field
Bit
Name
Description
5–0
CurrFID
The FID value associated with the OS-requested P-state
10–6
CurrVID
The VID value associated with the OS-requested P-state
31–11
Reserved
These bits must be 0 and are ignored by the processor driver.
The CurrFID and CurrVID ﬁelds of the FIDVID_STATUS MSR are valid only when the
FidVidPending bit is 0.
9.6.3
_PPC (Performance Present Capabilities)
The _PPC object is used to limit the maximum P-state that the operating system can use at any given
time. This object follows the _PSS object in the \_PR name space for the operating systems that
support the ACPI 2.0 processor P-state objects.
_PPC is generically described in Section 8.3.3.3 of the ACPI 2.0 speciﬁcation.
This section provides BIOS speciﬁcs regarding use of _PPC.
If the OEM and BIOS vendor do not implement a _PPC that limits the maximum P-state under certain
conditions, then the BIOS must always return a value of 0 for the _PPC object. When a value of 0 is
returned for the _PPC object, all P-states speciﬁed by the _PSS object are available.
Operating systems with native support for processor P-state transitions and ACPI Thermal zones
make use of reduced P-states for passive cooling before making use of “throttling” with the stop grant
state. This is the case even when the _PPC object returns a value of 0.
9.6.3.1
Using _PPC to Limit the Maximum Processor P-State When battery
Powered
Processor performance exceeds the demands of most typical notebook PC applications. However,
there are basically idle scenarios where applications run instructions that incorrectly cause the
operating system to conclude that the processor is 100% utilized. This causes the operating system to
force the processor to its maximum performance state, wasting power with no beneﬁt to the user.
Also, static power consumption of the processor increases when the processor is operating in a P-state
above the minimum voltage level supported by the processor.
Therefore, it will increase the system’s battery-powered runtime to limit the maximum performance
state available when battery-powered.
9.6.3.2
Implementation Overview
An ACPI-compliant General-Purpose Event (GPE) input of the chipset (typically the I/O Hub) is
dedicated to causing an SCI whenever the notebook power source changes. This input will be referred
Chapter 9
Power and Thermal Management
237
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
to as AC available (ACAV). The system can control ACAV directly based upon the presence of an AC
adapter, or the embedded controller (EC) in the system can control ACAV.
Whenever the AC adapter is inserted or removed from the system, the chipset logic associated with
ACAV sets the status bit associated with ACAV, and if enabled causes an SCI to be asserted. Note: the
control method associated with ACAV status could be implemented in the EC. A chipset GPE
dedicated for AC adapter status is used for this description.
The BIOS provides the following ACPI ASL control methods and objects:
•
_PPC (Performance Present Capabilities)
•
_PSR (Power Source) is described in Section 11.3.1 of the ACPI 2.0 speciﬁcation. _PSR returns:
– 0x00000000 if the system is running on battery power.
– 0x00000001 if the system is running from an AC adapter.
•
\_SB.AC is a BIOS-deﬁned device object for the AC adapter.
•
_L12 is the control method associated with GPE 12.
The ﬂow associated with AC adapter insertion or removal is:
•
AC adapter is inserted or removed
•
The GPE Status bit associated with ACAV is asserted (for this example, GPE 12)
•
An SCI (system control interrupt is issued)
•
The OS/ACPI driver determines that GPE 12 was asserted and runs the associated control method
(_L12)
•
_L12
• Issues a Notify(\_PR.CPU0, 0x80) which forces the OS to re-evaluate the _PPC object.
– _PPC reads the state of the ACAV pin and returns:
0 if the system is AC-powered. All P-states are available.
n if the system is battery-powered, where n is the highest P-state is supported when
battery-powered. Only P-states n and lower are available. For example, if a processor
supports 3 P-states:
P0 = 2400 MHz at 1.2 V
P1 = 1600 MHz at 1.1 V
P2 = 800 MHz at 1.0 V
Then n = 2 will limit the maximum P-state to 800 MHz at 1.0 V when the system is
battery-powered.
– OS takes into account the available P-states and if necessary performs a P-state transition.
• Issues a Notify(\_SB.AC, 0x80) where \_SB.AC is the AC adapter device object.
238
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
–
•
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
OS runs _PSR to determine if the current power source is the AC adapter. PSR reads the
state of the ACAV pin and returns:
1 if the system is AC-powered.
0 if the system is battery-powered.
Normal operation continues.
9.6.4
PSTATE_CNT
PSTATE_CNT is an entry in the Fixed ACPI Description Table (FADT). A PSTATE_CNT entry of 0
informs the operating system that the BIOS in System Management Mode (SMM) does not perform
P-state transitions.
The BIOS must report a value of 0 in the PSTATE_CNT ﬁeld of the FADT. The only BIOS-controlled
P-state transition, if any, must be performed near the beginning of the POST routine before control is
passed to the operating system. All subsequent transitions are made by system software not the BIOS.
System software is either the AMD PowerNow! technology software (for Microsoft® operating
systems prior to the Windows® XP operating system) or the operating system and associated
processor driver (for the Windows XP operating system and subsequent Microsoft operating
systems).
Note: SMM is not used to perform P-state transitions.
9.6.5
CST_CNT
CST_CNT is an entry in the FADT. If non-zero, this ﬁeld contains the value the operating system
writes to the SMI_CMD register to indicate operating system support for the _CST object and C
States Changed notiﬁcation.
The BIOS must report a value of 0 in the CST_CNT ﬁeld of the FADT. AMD platforms only support
ACPI 1.0b processor power state functionality. Processor power states are referred to as “C” states by
the ACPI speciﬁcations.
9.7
BIOS Support for AMD PowerNow!™ Software
with Legacy Operating Systems
If BIOS determines it has a processor and system that support P-state transitions, it provides a PSB
that is comprised of a header and the processor speciﬁc Performance State Table (PST) that matches
the processor in the system.
If the BIOS cannot determine that the system and processor support P-state transitions as deﬁned in
9.5.1 on page 219, then the BIOS must not provide a PSB. Changing the PSB Signature ﬁeld to all 0s
is an effective way of removing the PSB.
Chapter 9
Power and Thermal Management
239
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 51 deﬁnes the PSB and PST structures.
Table 51. Performance State Block Structure
Name
Length
(Bytes)
Field Purpose
PSB Header
Signature
10
Start of the PSB - Always set this field to “AMDK7PNOW!”
TableVersion
1
The version of the table as defined by AMD. Set to 14h for version 1.4. If
AMD PowerNow! software does not recognize the TableVersion, it will not
attempt to make P-state transitions.
Flags
1
All bits are reserved and written to zero.
VST
2
Voltage Stabilization Time. The amount of time required for the processor’s
core voltage regulator to increase the processor’s core voltage the increment
specified by MVS. VST is specified in 20 µs increments. The default is 05h
(100 µs). All other values are reserved.
Reserved1
1
Bits 1–0: Ramp Voltage Offset (RVO) has the same definition as for the _PSS
ACPI object.
Bits 3–2: Isochronous Relief Time (IRT) has the same definition as for the
_PSS ACPI object.
Bits 5–4: Maximum Voltage Step (MVS) has the same definition as for the
_PSS ACPI object.
Bits 7–6: Number of available P-states when the system is battery powered
has the following definition:
00b = all P-states are available
01b = only the minimum P-state is available
10b = 2 lowest P-states are available
11b = 3 lowest P-states are available.
NumPST
1
The total number of PSTs in the PSB. This value must be set to 01h.
CPUID
4
CPUID (EAX of CPUID extended function 8000_0001h) of the processor that
this PST belongs to. This provides processor family, extended family, model,
extended model, and stepping (revision) fields.
PLL Lock Time
1
Processor PLL Lock time in microseconds.The PLL Lock time is specified in
the processor data sheet addendum. For example, PLL_LOCK_TIME of 2 µs
is represented as 00000010b.
MaxFID
1
MaxFID[5:0] as reported by the MaxFID field of the FIDVID_STATUS MSR.
MaxVID
1
MaxVID[4:0] as reported by the MaxVID field of the FIDVID_STATUS MSR.
NumPStates
1
The number of FID, VID combinations in the PST. This field must be >= 1h.
FID
1
FID[5:0] code corresponding to the frequency of the lowest performance state
that the processor supports. This is defined in the processor data sheet.
VID
1
VID[4:0] code corresponding to the voltage of the lowest performance state
that the processor supports. This is defined in the processor data sheet.
PST Header
240
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
Table 51.
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Performance State Block Structure (Continued)
Name
Length
(Bytes)
-----------
---------
Field Purpose
FID, VID pairs are concatenated here so that the total number of pairs is equal
to the NumPstates field. Each subsequent FID, VID pair is appended in
ascending order (i.e., the last FID,VID pair corresponds to the maximum performances state supported by the processor).
For processors with a family number of less than Fh, refer to BIOS Requirements for
AMD PowerNow! Technology Application Note, order# 25264.
9.8
System Configuration for Power Management
9.8.1
Chipset Configuration for Power Management
Chipset conﬁguration is covered in chipset related documentation. The BIOS is required to:
•
Ensure that the SMAF code to STPCLK and Stop Grant mapping for all components based on
HyperTransport technology is conﬁgured as described in Table 42 on page 217.
•
For single processor systems, the HyperTransport links are tri-stated in response to LDTSTOP_L
assertion.
•
Chipset power management features are appropriately enabled based upon system class.
•
LDTSTOP_L assertion during FID_Change:
– For Revision B processors is at least 4µs but less than 10 µs for UP systems.
– For Revision C and subsequent processors is at least 2µs but less than 6µs. Note: for UMA
chipsets 2µs is ideal and may be required in some cases.
9.8.2
Processor Configuration for Power Management
BIOS conﬁgures the processor to enable power management during the POST routine. Table 52 on
page 242 lists processor memory system conﬁguration register settings that the BIOS must initialize.
Chapter 9
Power and Thermal Management
241
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 52.
26094
Rev. 3.06
September 2003
Configuration Register Settings for Power Management
Functionality
Register
Do not enable
HyperTransport™ links to
be tri-stated in response to
LDTSTOP_L assertion
LDT0 Link Control
(function 0: offset 84h[13])
(LdtStopTriEn)
Mobile
Setting
UP Desktop
Setting
Multiprocessor
Setting
0b
0b
0b
6307_6361h
2307_0000h
Revision B:
2117_0000h
LDT1 Link Control
(function 0: offset A4h[13])
(LdtStopTriEn)
LDT2 Link Control
(function 0: offset C4h[13])
(LdtStopTriEn)
Map STPCLK actions to
SMAF codes:
SMAF 0 = C2
Power Management Control Low (function 3: offset
80h)
Revision C:
2307_0000h
SMAF 1 = C3
SMAF 2 = VID/FID Change
SMAF 3 = S1
Map STPCLK actions to
SMAF codes:
SMAF 4 = S3
Power Management Control High (function 3: offset 84h)
6113_4113h
2113_2113h
SMAF 5 = Throttling
Revision B:
0013_0013h
Revision C1:
0013_2113h
SMAF 6 = S4/S5
SMAF 7 is not used for
STPCLK, the corresponding field is used for HALT.
BCLK and LCLK PLL frequency lock
Clock Power/Timing Low
(function 3: offset D4h)
Sets “Ramp VID” (0 mV)
Clock Power/Timing High
and Voltage Regulator Sta- (function 3:offset D8h)
bilization time (0 s). The
processor driver performs
this function, not hardware.
Revision B:
000D_0001h
Revision C with unbuffered memory:
000D_0701h
Revision C with registered DIMMs:
04E2_0707h
0000_0000h
0000_0000h
0000_0000h
Notes:
1. Optionally, a value of 0113_2113h can be used. This will significantly reduce processor power
during halt and may reduce system performance.
242
Power and Thermal Management
Chapter 9
26094
Rev. 3.06
9.8.2.1
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Additional BIOS Support Requirements for S3
During the S3 state, power is removed from the processor’s CPU core, host bridge, and memory
controller. It is required that the BIOS restore the state of the processor’s memory controller upon
resume from S3 prior to having the memory controller bring system memory out of self-refresh mode.
The system must provide non volatile storage for the memory controller conﬁguration registers
during the S3 state. For UP and DP systems, this storage will be provided in the chipset. There are
two opportunities for the BIOS to store the memory controller conﬁguration to this nonvolatile
memory.
•
During POST as the memory controller is being initially conﬁgured—this is preferred, since it
does not require SMM support.
•
In SMM just prior to entry to the S3 state—BIOS uses an SMI trap on writes to the ACPI PM1
control register to get control of the system just prior to entry to S3. After storing the memory
controller conﬁguration registers, I/O instruction restart is used to place the system into the S3
state. This is not preferred because it relies on SMM.
Processor Function 2 conﬁguration space registers at offsets 40–98h must be stored in nonvolatile
RAM for support of resume from S3.
9.8.2.2
ACPI Tables/Objects
FADT entries for C-state and throttling support and throttling is listed in Table 53 on page 243. Some
legacy operating systems will not use C3 if support for C2 is not declared.
Table 53.
FADT Table Entries
Field
Mobile
UP Desktop
Multiprocessor
PSTATE_CNT
0
0
0
CST_CNT
0
0
0
P_LVL2_LAT
> 100 if C2 is not supported
1 if C2 is supported
> 100
> 100
P_LVL3_LAT
> 1000 if C3 is not supported
10 if C3 is supported
> 1000
> 1000
0 if _PSV is not used
3 if _PSV is used
0 if _PSV is not used
3 if _PSV is used
0
DUTY_WIDTH
Chapter 9
Power and Thermal Management
243
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
244
Power and Thermal Management
26094
Rev. 3.06
September 2003
Chapter 9
26094
Rev. 3.06
10
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Performance Monitoring
AMD Athlon™ 64 and AMD Opteron™ processors support the performance-monitoring features
introduced in earlier processor implementations. These features allow the selection of events to be
monitored, and include a set of corresponding counter registers that track the frequency of monitored
events. Software tools can use these features to identify performance bottlenecks, such as sections of
code that have high cache-miss rates or frequently mis-predicted branches. This information can then
be used as a guide for improving or eliminating performance problems through software
optimizations or hardware-design improvements.
For a general description of how to use these performance monitoring features, refer to the “Debug
and Performance Resources” section in Volume 2: System Programming of the AMD64 Architecture
Programmers Manual, order# 24593.
10.1
Performance Counters
The processor provides four 48-bit performance counters. Each counter can monitor a different event.
The accuracy of the counters is not ensured. Some events are reserved. When a reserved event is
selected, the results are undefined.
Performance counters are used to count specific processor events, such as data-cache misses, or the
duration of events, such as the number of clocks it takes to return data from memory after a cache
miss. During event counting, the processor increments the counter when it detects an occurrence of
the event. During duration measurement, the processor counts the number of processor clocks it takes
to complete an event. Each performance counter can be used to count one event, or measure the
duration of one event at a time.
The processor supports four performance counters, PerfCtrn. The implementation of each counter
supports 48 bits of resolution. This section shows the format of the PerfCtrn registers.
Only code executing at privilege level 0 can directly write the performance counters using the
WRMSR instruction.The PerfCtrn registers are model-specific registers that can be read using a
special read performance-monitoring counter instruction, RDPMC. The RDPMC instruction loads
the contents of the PerfCtrn register specified by the ECX register, into the EDX register and the EAX
register. The high 32 bits are loaded into EDX, and the low 32 bits are loaded into EAX. RDPMC can
be executed only at CPL=0, unless system software enables use of the instruction at all privilege
levels. RDPMC can be enabled for use at all privilege levels by setting CR4.PCE (the performancemonitor counter-enable bit) to 1. When CR4.PCE = 0 and CPL > 0, attempts to execute RDPMC
result in a general-protection exception (#GP).
The performance counters can also be read and written by system software running at CPL=0 using
the RDMSR and WRMSR instructions, respectively. Writing the performance counters can be useful
if software wants to count a specific number of events, and then trigger an interrupt when that count is
Chapter 10
Performance Monitoring
245
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
reached. An interrupt can be triggered when a performance counter overflows. Software should use
the WRMSR instruction to load the count as a two’s-complement negative number into the
performance counter. This causes the counter to overflow after counting the appropriate number of
times.
The performance counters are not assured of producing identical measurements each time they are
used to measure a particular instruction sequence, and they should not be used to take measurements
of very small instruction sequences. The RDPMC instruction is not serializing, and it can be executed
out-of-order with respect to other instructions around it. Even when bound by serializing instructions,
the system environment at the time the instruction is executed can cause events to be counted before
the counter value is loaded into EDX:EAX.
PerfCtr0–PerfCtr3 Registers
C001_0004h, C001_0005h, C001_0006h, C001_0007h
63
48 47
reserved
32
CTR (47–32)
31
0
CTR (31–0)
Bit
Name
Function
63–48
Reserved
RAZ
R
47–0
CTR
Counter value
R
10.2
R/W
Performance Event-Select Registers
Performance Event-Select registers (PerfEvtSeli) are used to specify the events counted by the
performance counters, and to control other aspects of their operation. Each performance counter
supported by the implementation has a corresponding event-select register that controls its operation.
This section shows the format of the PerfEvtSeln registers.
The performance event-select registers can be read and written only by system software running at
CPL=0 using the RDMSR and WRMSR instructions, respectively. Any attempt to read or write these
registers at CPL>0 causes a general-protection exception to occur.
To accurately start counting with the write that enables the counter, disable the counter when
changing the event and then enable the counter with a second MSR write.
The edge count mode will increment the counter when a transition happens on the monitored event. If
the event selected is changed without disabling the counter, an extra edge will be falsely detected
when the first event is a static 0 and the second event is a static one. To avoid this false edge detection,
disable the counter when changing the event and then enable the counter with a second MSR write.
246
Performance Monitoring
Chapter 10
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
The EVENT_MASK field specifies the event or event duration in a processor unit to be counted by
the corresponding PerfCtri register. Table 55 on page 248 shows the function unit encoding of
EVENT_MASK[7:5]. The UNIT_MASK field is used to further specify or qualify a monitored event.
Table 55 on page 248 shows the events and unit masks supported by the processor. The events that
can be counted are implementation dependent. For more up-to-date information on the events in the
latest processor revisions, refer to the Revision Guide for the AMD Athlon™ 64 and AMD Opteron™
processors, order# 25759.
PerfEvtSel0–PerfEvtSel3 Registers
C001_0000h, C001_0001h,
C001_0002h, C001_0003h
63
32
reserved
USR
OS
E
PC
INT
reserved
CNT_MASK
EN
24 23 22 21 20 19 18 17 16 15
INV
31
8
7
UNIT_MASK
0
EVENT_MASK
)
Bit
Name
Function
63–32
reserved
RAZ
R/W
31–24
CNT_MASK
Counter Mask
R/W
23
INV
Invert Counter Mask
R/W
22
EN
Enable Counter
R/W
21
reserved
SBZ
20
INT
Enable APIC Interrupt
R/W
19
PC
Pin Control
R/W
18
E
Edge Detect
R/W
17
OS
Operating-System Mode
R/W
16
USR
User Mode
R/W
15–8
UNIT_MASK
Event Qualification
R/W
7–0
EVENT_MASK
Unit and Event Selection
R/W
Field Descriptions
Event Selection (EVENT_MASK)—Bits 7–0. EVENT_MASK[7:5] selects a processor functional
unit. EVENT_MASK[4:0] selects an event within the processor unit.
Event Qualification (UNIT_MASK)—Bits 15–8. Each UNIT_MASK bit further specifies or
qualifies the event specified by EVENT_MASK. All events selected by UNIT_MASK are
simultaneously monitored.
User Mode (USR)—Bit 16. Count events in user mode (at CPL > 0).
Operating-System Mode (OS)—Bit 17. Count events in operating-system mode (at CPL = 0).
Chapter 10
Performance Monitoring
247
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Edge Detect (E)—Bit 18. 1=Edge detection. 0=Level detection.
Pin Control (PC)—Bit 19. With a value of 1, toggles the performance monitor pin PMi when
PerfCtri register overflows. With a value of 0, toggles the performance monitor pin PMi each
time PerfCtri register is incremented.
Enable APIC Interrupt (INT)—Bit 20. Enables APIC interrupt when PerfCtri register overflows.
Enable Counter (EN)—Bit 22. Enables performance monitor counter.
Invert Counter Mask (INV)—Bit 23. See CNT_MASK.
Counter Mask (CNT_MASK)—Bits 31–24. a) When CNT_MASK is 0, the corresponding PerfCtri
register is incremented by the number of events occurring in a clock cycle. Maximum number
of events in one cycle is 3. b) When CNT_MASK values are from 1 to 3 and INV is 0, the
corresponding PerfCtri register is incremented by 1, if the number of events occurring in a
clock cycle is greater than or equal to CNT_MASK. When CNT_MASK values are from 1 to
3 and INV is 1, the corresponding PerfCtri register is incremented by 1, if the number of
events occurring in a clock cycle is less than CNT_MASK. c) CNT_MASK values from 4 to
255 are reserved.
Table 54. Processor Functional Unit Encoding
EVENT_MASK[7:5] Description
0
Selects event from FP unit
1
Selects event from LS unit
2
Selects event from DC unit
3
Selects event from BU unit
4
Selects event from IC unit
5
Reserved
6
Selects event from FR unit
7
Selects event from NB unit
Table 55. Performance Monitor Events
EVENT_MASK
[7:0]
Encoded
EVENT_MASK
[7:5]
00h
FP
248
UNIT_
MASK
bit
Description
Dispatched FPU ops (Revision B and later revisions)
0
Add pipe ops excluding junk ops
1
Multiply pipe ops excluding junk ops
2
Store pipe ops excluding junk ops
3
Add pipe junk ops
4
Multiply pipe junk ops
Performance Monitoring
Chapter 10
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
Encoded
EVENT_MASK
[7:5]
UNIT_
MASK
bit
Description
5
Store pipe junk ops
7–6
Reserved
01h
FP
Cycles with no FPU ops retired (Revision B and later
revisions)
02h
FP
Dispatched FPU ops that use the fast flag interface (Revision
B and later revisions)
03h to 1Fh
20h
Reserved
LS
Segment register load
0
ES
1
CS
2
SS
3
DS
4
FS
5
GS
6
HS
7
Reserved
21h
LS
Microarchitectural resync caused by self-modifying code
22h
LS
Microarchitectural resync caused by snoop
23h
LS
LS buffer 2 full
24h
LS
Locked operation (Revision B and earlier revisions)
0
Number of lock instructions executed (Revision C and later
revisions)
1
Number of cycles spent in the lock request/grant stage
(Revision C and later revisions)
2
Number of cycles a lock takes to complete once it is nonspeculative and is the oldest load/store operation (nonspeculative cycles in Ls2 entry 0) (Revision C and later
revisions)
7-3
Reserved
25h
LS
Microarchitectural late cancel of an operation
26h
LS
Retired CFLUSH instructions
27h
LS
Retired CPUID instructions
28h to 3Fh
Reserved
40h
DC
Access (includes microcode scratchpad accesses)
41h
DC
Miss
42h
DC
Refill from L2
Chapter 10
Performance Monitoring
249
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
43h
44h
Encoded
EVENT_MASK
[7:5]
UNIT_
MASK
bit
Description
0
Invalid
1
Shared
2
Exclusive
3
Owner
4
Modified
7–5
Reserved
DC
Refill from system
0
Invalid
1
Shared
2
Exclusive
3
Owner
4
Modified
7–5
Reserved
DC
Copyback
0
Invalid
1
Shared
2
Exclusive
3
Owner
4
Modified
7–5
Reserved
45h
DC
L1 DTLB miss and L2 DTLB hit
46h
DC
L1 DTLB miss and L2 DTLB miss
47h
DC
Misaligned data reference
48h
DC
Microarchitectural late cancel of an access
49h
DC
Microarchitectural early cancel of an access
4Ah
DC
One bit ECC error recorded found by scrubber
4Bh
4Ch
250
0
Scrubber error
1
Piggyback scrubber errors
7–2
Reserved
DC
DC
Dispatched prefetch instructions
0
Load
1
Store
2
NTA
3–7
Reserved
DCACHE accesses by locks (Revision C and later revisions)
Performance Monitoring
Chapter 10
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
Encoded
EVENT_MASK
[7:5]
UNIT_
MASK
bit
Description
0
Number of dcache accesses by lock instructions (Revision C
and later revisions)
1
Number of dcache misses by lock instructions (Revision C
and later revisions)
7-2
Reserved
4Dh to 5Fh
7Dh
7Eh
7Fh
Reserved
BU
Internal L2 request
0
IC fill
1
DC fill
2
TLB reload
3
Tag snoop request
4
Cancelled request
7–5
Reserved
BU
Fill request that missed in L2
0
IC fill
1
DC fill
2
TLB reload
7–3
Reserved
BU
Fill into L2
0
Dirty L2 victim
1
Victim from L1
7–2
Reserved
80h
IC
Fetch
81h
IC
Miss
82h
IC
Refill from L2
83h
IC
Refill from system
84h
IC
L1ITLB miss and L2ITLB hit
85h
IC
L1ITLB miss and L2ITLB miss
86h
IC
Microarchitectural resync caused by snoop
87h
IC
Instruction fetch stall
88h
IC
Return stack hit
89h
IC
Return stack overflow
8Ah to BFh
Reserved
C0h
FR
Retired x86 instructions including exceptions and interrupts
C1h
FR
Retired uops
Chapter 10
Performance Monitoring
251
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
Encoded
EVENT_MASK
[7:5]
C2h
FR
Retired branches including exceptions and interrupts
C3h
FR
Retired branches mispredicted
C4h
FR
Retired taken branches
C5h
FR
Retired taken branches mispredicted
C6h
FR
Retired far control transfers (always mispredicted)
C7h
FR
Retired resyncs (non control transfer branches)
C8h
FR
Retired near returns
C9h
FR
Retired near returns mispredicted
CAh
FR
Retired taken branches mispredicted only due to address
miscompare
CBh
FR
Retired FPU instructions (Revision B and later revisions)
CCh
UNIT_
MASK
bit
Description
0
x87 instructions
1
Combined MMX™ and 3DNow!™ instructions
2
Combined packed SSE and SSE2 instructions
3
Combined scalar SSE and SSE2 instructions
4–7
Reserved
Retired fastpath double op instructions (Revision B and later
revisions)
FR
0
With low op in position 0
1
With low op in position 1
2
With low op in position 2
3–7
Reserved
CDh
FR
Interrupts masked cycles (IF=0)
CEh
FR
Interrupts masked while pending cycles (INTR while IF=0)
CFh
FR
Taken hardware interrupts
D0h
FR
Nothing to dispatch (decoder empty)
D1h
FR
Dispatch stalls (D2h–DAh combined)
D2h
FR
Dispatch stall from branch abort to retire
D3h
FR
Dispatch stall for serialization
D4h
FR
Dispatch stall for segment load
D5h
FR
Dispatch stall when reorder buffer is full
D6h
FR
Dispatch stall when reservation stations are full
D7h
FR
Dispatch stall when FPU is full
D8h
FR
Dispatch stall when LS is full
D9h
FR
Dispatch stall when waiting for all to be quiet
252
Performance Monitoring
Chapter 10
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
Encoded
EVENT_MASK
[7:5]
UNIT_
MASK
bit
DAh
FR
Dispatch stall when far control transfer or resync branch is
pending
DBh
FR
FPU exceptions (Revision B and later revisions)
Description
0
x87 reclass microfaults
1
SSE retype microfaults
2
SSE reclass microfaults
3
SSE and x87 microtraps
4–7
Reserved
DCh
FR
Number of breakpoints for DR0
DDh
FR
Number of breakpoints for DR1
DEh
FR
Number of breakpoints for DR2
DFh
FR
Number of breakpoints for DR3
E0h
NB
Memory controller page access event
0
Page hit
1
Page miss
2
Page conflict
7–3
Reserved
E1h
NB
Memory controller page table overflow
E2h
NB
Memory controller DRAM command slots missed (in
MemClks)
E3h
NB
Memory controller turnaround
E4h
EBh
Chapter 10
0
DIMM turnaround
1
Read to write turnaround
2
Write to read turnaround
7–3
Reserved
NB
Memory controller bypass counter saturation
0
Memory controller high priority bypass
1
Memory controller low priority bypass
2
DRAM controller interface bypass
3
DRAM controller queue bypass
7–4
Reserved
NB
Sized commands
0
NonPostWrSzByte
1
NonPostWrSzDword
2
PostWrSzByte
Performance Monitoring
253
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Table 55. Performance Monitor Events (Continued)
EVENT_MASK
[7:0]
ECh
F6h
F7h
F8h
254
Encoded
EVENT_MASK
[7:5]
UNIT_
MASK
bit
Description
3
PostWrSzDword
4
RdSzByte
5
RdSzDword
6
RdModWr
7
Reserved
NB
Probe result
0
Probe miss
1
Probe hit
2
Probe hit dirty without memory cancel
3
Probe hit dirty with memory cancel
7–4
Reserved
NB
HyperTransport™ bus 0 bandwidth
0
Command sent
1
Data sent
2
Buffer release sent
3
Nop sent
7–4
Reserved
NB
HyperTransport™ bus 1 bandwidth
0
Command sent
1
Data sent
2
Buffer release sent
3
Nop sent
7–4
Reserved
NB
HyperTransport™ bus 2 bandwidth
0
Command sent
1
Data sent
2
Buffer release sent
3
Nop sent
7–4
Reserved
Performance Monitoring
Chapter 10
26094
Rev. 3.06
11
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
BIOS Checklist
The checklist in this chapter reviews information that is described elsewhere and is required by BIOS
developers to properly incorporate AMD Athlon™ 64 and AMD Opteron™ processors into systems.
11.1
CPUID
Use the CPUID instruction to properly identify the processor.
•
Determine the processor type, stepping, and available features by using functions 0000_0001h
and 8000_0001h of the CPUID instruction.
•
Boot-up display - The processor name should be displayed according to the processor name string
deﬁned in CPUID Guide for the AMD Athlon™ 64 and AMD Opteron™ processors, order#
25481.
For more information on the CPUID instruction, see AMD Processor Recognition Application Note,
order# 20734 and CPUID Guide for the AMD Athlon™ 64 and AMD Opteron™ processors, order#
25481.
11.2
CPU Speed Detection
The BIOS can use the time-stamp counter (TSC) to clock a timed operation and compare the result to
the Real-Time Clock (RTC) to determine the operating frequency. See the example of frequencydetermination assembler code available on the AMD web site.
11.3
HyperTransport™ Link Frequency Selection
The BIOS should initialize HyperTransport™ link frequency (Link ﬁeld, Function 0, Offsets 88h,
A8h, C8h) with the minimum of the following frequencies:
•
the maximum frequency deﬁned by the link frequency capability bits (LnkFreqCap ﬁeld,
Function 0, Offsets 88h, A8h, C8h),
•
the maximum frequency values speciﬁed in the processor and chipset data sheets,
•
the maximum frequency supported by the platform.
Chapter 11
BIOS Checklist
255
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
11.4
26094
Rev. 3.06
September 2003
Multiprocessing Capability Detection
The multiprocessing capability of the AMD Opteron™ processor is determined by the MPCap and
BigMPCap bits in the Northbridge Capability Register (Function 3, Offset E8h).
Multiprocessing
Capability
MPCap
BigMPCap
UP Capable
0
0
DP Capable
1
0
MP Capable
1
1
During POST, the BIOS checks the multiprocessing capability of AMD Opteron™ processors, and
conﬁgures the system accordingly.
Multiprocessing capability detection is not required in a UP system.
All processors must be DP capable or MP capable in a DP system. If any processor is only UP
capable, the BIOS must conﬁgure the BSP as a UP processor, and must not initialize the AP.
All processors must be MP capable in an MP system. If any processor is not MP capable, the BIOS
must conﬁgure the BSP as a UP processor, and must not initialize APs.
If all processors do not have adequate multiprocessing capability for a DP or an MP system, the BIOS
must display the following message:
*********** Warning: non-MP Processor ***********
The processor(s) installed in your system are not multiprocessing capable. Now
your system will halt.
If all processors have adequate multiprocessing capability for a DP or an MP system, but have
different model numbers or operate at different frequencies, the BIOS must conﬁgure the BSP as a
UP processor, and must not initialize APs. The BIOS must display the following message:
****** Warning: non-matching MP Processors ******
AMD multiprocessing-capable processors have been detected in your system.
However, all processors must be the same model and frequency to run in
multiprocessing mode. Now your system will halt.
After the error message is displayed, the BIOS must halt the system without further initialization.
11.5
Model-Specific Registers (MSRs)
Access only the MSRs that are implemented in the processor.
•
256
Follow the processor MSR programming sequence described in this document. This includes
setting the HWCR, SYSCFG, MTRRs, IORRs, clock control MSR, Top of Memory, and other
registers.
BIOS Checklist
Chapter 11
26094
Rev. 3.06
11.6
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Machine Check Architecture (MCA)
The processor supports a machine check architecture that allows reporting and handling of system
errors through a consistent interface.
•
The global MCA Status register must be cleared (all machine check features disabled).
•
MC0_CTL must be set to all 1s because some do not set these bits.
•
In addition, the local MCA status and address registers for the following ﬁve reporting blocks
must also be cleared (all error and parity bits disabled):
– BU—Bus Unit
– IC—Instruction Cache Unit
– DC—Data Cache Unit
– LS—Load Store Unit
– NB—Northbridge Unit
After reset, the BIOS needs to determine if the cause of the reset was a power-on or soft reset.
•
When power-on reset occurs, the BIOS must clear the MCA registers.
•
Otherwise, the BIOS should check for any MCA information that may be related to the soft reset.
For more information on MCA, see Chapter 5, “Machine Check Architecture.”
11.7
Memory Map
11.7.1
I/O and Memory Type and Range Registers (IORRs, MTRRs)
The memory-type and range registers control the access and cacheability of memory regions in the
processor.
•
Follow the processor MTRR and IORR programming sequence that is described in this document.
•
To initialize MTTRs in BIOS, set the default memory cache type for all addresses to uncacheable,
then use a variable range MTRR to set all physical memory as cacheable (writeback), and ﬁnally
use the ﬁxed-range MTRRs to handle the special cases in the 640-Kbyte–1-Mbyte region.
MTRRs should be modiﬁed while cache is disabled. Modifying MTRRs while cache is enabled
will result in undeﬁned behavior.
•
The GART driver (miniport) should program the AGP aperture.
•
The BIOS does not map the cacheability of the video frame buffer and AGP aperture space; it is
done by the operating system, video, and AGP drivers.
Chapter 11
BIOS Checklist
257
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
11.7.2
26094
Rev. 3.06
September 2003
Memory Map Registers (MMRs)
The processor contains several memory map registers to support the appropriate software view of
memory.
•
Program the correct values in the MSRs, such as Top of Memory (TOP_MEM).
11.8
•
Cache Testing and Programming
BIOS must correctly maintain the tag parity and data ECC enable bits.
11.9
Memory System Configuration Registers
Node conﬁguration space contains registers that deﬁne RAM, PCI memory, and x86 I/O address
paths.
•
Address mapping registers are contained in function 1 of each node.
•
Cold boot and reset sends code fetch and other addresses to the compatibility link of Node0. At
cold boot and reset, these mapping registers are disabled.
•
Once enabled, these registers distribute addresses according to register mappings, which include
node and link routing. See Chapter 8, “HyperTransport™ Technology Conﬁguration and
Enumeration.”
•
Thereafter, carefully set registers to reﬂect current needs for code fetch and memory access, i.e.,
do not map BIOS E0000h and F0000h space to DRAM until BIOS exits execution from the ROM
chip. Otherwise, the system will hang. Likewise, memory-mapped I/O space should not be
mapped above 1 Mbyte until BIOS exits execution from the ROM chip, or else the system will
hang.
•
Memory-mapped I/O (MMIO) and PCI I/O mappings must include address space needed to move
option ROM code to RAM, as well as the required run-time resources. This could require setting
and resetting the MMIO mappings during enumeration and device initialization. Conﬁguration
space access of a device with or without an option ROM is deﬁned by the HyperTransport
technology conﬁguration mapping.
•
Multihost bus multiprocessor systems require that MMIO mappings direct relevant addresses to
the buses of devices through the node/link path to the bus.
•
Application processors and the bootstrap processor in multiprocessor systems must be mapped
separately with paths to target address space.
258
BIOS Checklist
Chapter 11
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
11.10 XSDT Table
The Extended System Description Table (XSDT) table deﬁned in ACPI 2.0 must be implemented in
the BIOS to support 64-bit operating systems for AMD Athlon™ 64 and AMD Opteron™ processor
based systems.
11.11 Detect Target Operating Mode Callback
The operating system notiﬁes the BIOS what the expected operating mode is with the Detect Target
Operating Mode callback (INT 15, function EC00h). Based on the target operating mode, the BIOS
can enable or disable mode speciﬁc performance and functional optimizations that are not visible to
system software.
This callback does not change the operating mode; it only declares the target mode to the BIOS. It
should be executed only once by the BSP before the ﬁrst transition into long mode.
The default operating mode assumed by the BIOS is Legacy Mode Target Only. If this is not the target
operating mode, system software must execute this callback to change it before transitioning to long
mode for the ﬁrst time. If the target operating mode is Legacy Mode Target Only, the callback does
not need to be executed.
The Detect Target Operating Mode callback inputs are stored in the AX and BL registers. AX has a
value of EC00h, selecting the Detect Target Operating Mode function. One of the following values in
the BL register selects the operating mode:
• 01h — Legacy Mode Target Only. All enabled processors will operate in legacy mode only.
• 02h — Long Mode Target Only. All enabled processor will switch into long mode once.
• 03h — Mixed Mode Target. Processors may switch between legacy mode and long mode, or
the preferred mode for system software is unknown. This value instructs the BIOS to use
settings that are valid in all modes.
• All other values are reserved.
The Detect Target Operating Mode callback outputs are stored in the AH register and CF (carry ﬂag
in the EFLAGS register), and the values of other registers are not modiﬁed. The following output
values are possible:
•
AH = 00h and CF = 0, if the callback is implemented and the value in BL is supported.
•
AH = 00h and CF = 1, if the callback is implemented and the value in BL is reserved. This
indicates an error; the target operating mode is set to Legacy Mode Target Only.
•
AH = 86h and CF = 1, if the callback is not supported.
Chapter 11
BIOS Checklist
259
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
11.12 SMM Issues
The processor includes support for system management mode (SMM). The functionality of the
AMD Athlon™ 64 processor or AMD Opteron™ processor SMM is a superset of the Pentium®
processor functionality.
•
The AMD Athlon™ 64 and AMD Opteron™ processors implement the SMM remapping and
control registers as model-speciﬁc registers on the CPU.
•
Implement the SMM state-save area in a manner compatible with the description of SMM found
in Chapter 6, “System Management Mode (SMM).”
Program the model-speciﬁc registers to set the SMM memory base, local address, destination address,
memory type, size and control, etc. See Chapter 6, “System Management Mode (SMM).”
260
BIOS Checklist
Chapter 11
26094
Rev. 3.06
12
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Processor Configuration Registers
This chapter includes descriptions of two types of model-speciﬁc registers, as follows:
•
General model-speciﬁc registers (see page 261)
•
AMD Athlon™ 64 and AMD Opteron™ model-speciﬁc registers (see page 284)
12.1
General Model-Specific Registers
Table 56 is a listing of the general model-speciﬁc registers supported by the processor, presented in
ascending hexadecimal address order. Register descriptions follow the table, organized according to
the following functions:
•
System software (see page 263)
•
Memory typing (see page 265)
•
APIC (see page 282)
•
Machine check architecture (see page 146)
•
Software debug (see page 283)
•
Performance monitoring (see page 284)
Table 56.
General MSRs
Address
Register Name
Description
0010h
TSC
page 284
001Bh
APIC_BASE
page 282
002Ah
EBL_CR_POWERON
page 282
00FEh
MTRRcap
page 265
0174h
SYSENTER_CS
page 263
0175h
SYSENTER_ESP
page 264
0176h
SYSENTER_EIP
page 264
0179h
MCG_CAP
page 147
017Ah
MCG_STATUS
page 147
017Bh
MCG_CTL
page 148
Chapter 12
Processor Conﬁguration Registers
261
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Table 56.
26094
Rev. 3.06
General MSRs (Continued)
Address
Register Name
Description
01D9h
DebugCtl
page 283
01DBh
LastBranchFromIP
page 283
01DCh
LastBranchToIP
page 283
01DDh
LastExceptionFromIP
page 283
01DEh
LastExceptionToIP
page 283
0200–020Eh Even
MTRRphysBase[7:0]
page 266
0201–020Fh Odd
MTRRphysMask[7:0]
page 266
0250h
MTRRfix64K_00000
page 267
0258h
MTRRfix16K_80000
page 268
0259h
MTRRfix16K_A0000
page 269
0268h
MTRRfix4K_C0000
page 271
0269h
MTRRfix4K_C8000
page 272
026Ah
MTRRfix4K_D0000
page 273
026Bh
MTRRfix4K_D8000
page 274
026Ch
MTRRfix4K_E0000
page 275
026Dh
MTRRfix4K_E8000
page 276
026Eh
MTRRfix4K_F0000
page 278
026Fh
MTRRfix4K_F8000
page 279
0277h
PAT
page 280
02FFh
MTRRdefType
page 281
0400h
MC0_CTL
page 152
0401h
MC0_STATUS
page 153
0402h
MC0_ADDR
page 155
0403h,
MC0_MISC
not supported
0404h
MC1_CTL
page 156
0405h
MC1_STATUS
page 157
0406h
MC1_ADDR
page 157
0407h
MC1_MISC
not supported
262
September 2003
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
Table 56.
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
General MSRs (Continued)
Address
Register Name
Description
0408h
MC2_CTL
page 156
0409h
MC2_STATUS
page 160
040Ah
MC2_ADDR
page 160
040Bh
MC2_MISC
not supported
040Ch
MC3_CTL
page 161
040Dh
MC3_STATUS
page 162
040Eh
MC3_ADDR
page 162
040Fh
MC3_MISC
not supported
0410h
MC4_CTL
page 162
0411h
MC4_STATUS
page 162
0412h
MC4_ADDR
page 162
0413h
MC4_MISC
not supported
12.1.1
System Software Registers
12.1.1.1
SYSENTER_CS Register
This register contains the code segment selector used by the SYSENTER and SYSEXIT instructions.
See the SYSENTER and SYSEXIT section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
SYSENTER_CS Register
MSR 0174h
63
32
reserved
31
16 15
reserved
0
SYSENTER_CS
Bit
Mnemonic
Function
R/W
Reset
63–32
reserved
RAZ
R
0
31–16
15–0
reserved
SBZ
R/W
0
SYSENTER_CS
SYSENTER/SYSEXIT code segment selector
R/W
0
Chapter 12
Processor Conﬁguration Registers
263
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
SYSENTER/SYSEXIT Code Segment Selector (SYSENTER_CS)—Bits 15–0.
12.1.1.2
SYSENTER_ESP Register
This register contains the stack pointer used by the SYSENTER and SYSEXIT instructions. See the
SYSENTER and SYSEXIT section in Volume 2 of the AMD64 Architecture Programmer’s Manual
for more information.
SYSENTER_ESP Register
MSR 0175h
63
32
reserved
31
0
SYSENTER_ESP
Bit
Mnemonic
Function
63–32
reserved
RAZ
31–0
SYSENTER_ESP
SYSENTER/SYSEXIT stack pointer
R/W
Reset
R
0
R/W
0
Field Descriptions
SYSENTER/SYSEXIT Stack Pointer (SYSENTER_ESP)—Bits 31–0.
12.1.1.3
SYSENTER_EIP Register
This register contains the instruction pointer used by the SYSENTER and SYSEXIT instructions. See
the SYSENTER and SYSEXIT section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
SYSENTER_EIP Register
MSR 0176h
63
32
reserved
31
0
SYSENTER_EIP
264
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bit
Mnemonic
Function
63–32
reserved
RAZ
31–0
SYSENTER_EIP
SYSENTER/SYSEXIT instruction pointer
R/W
Reset
R
0
R/W
0
Field Descriptions
SYSENTER/SYSEXIT Instruction Pointer (SYSENTER_EIP)—Bits 31–0.
12.1.2
Memory Typing Registers
12.1.2.1
MTRRcap Register
This is a read-only register that returns information about the processors MTRR capabilities. See the
“Using MTRRs” section in Volume 2 of the AMD64 Architecture Programmer’s Manual for more
information.
The MTRRcap register is a read-only status register. Attempting to modify this register will result in a
#GP(0).
MTRRcap Register
MSR 00FEh
63
32
Bit
8
MtrrCapFix
reserved
11 10 9
reserved
31
MtrrCapWc
reserved
7
0
MtrrCapVCnt
Mnemonic
Function
R/W
63–11
reserved
RAZ
R
10
MtrrCapWc
Write-combining memory type
R
9
reserved
RAZ
R
8
MtrrCapFix
Fixed range register
R
7–0
MtrrCapVCnt
Variable range registers count
R
Field Descriptions
Variable Range Registers Count (MtrrCapVCnt)—Bits 7–0. Indicates number of variable range
registers.
Fixed Range Registers (MtrrCapFix)—Bit 8. Indicates fixed range register capability.
Chapter 12
Processor Conﬁguration Registers
265
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Write-Combining Memory Type (MtrrCapWc)—Bit 10. Indicates write-combining memory type
capability.
12.1.2.2
MTRRphysBasei Registers
These registers deﬁne the base address and memory type for each of the variable MTRRs. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Attempting to modify reserved bits in MTRRphysBasei will result in a #GP(0).
MTRRphysBase0–7 Registers
MSRs 0200h, 0202h, 0204h, 0206h,
0208h, 020Ah, 020Ch, 020Eh
63
40 39
reserved
31
PhyBase 27–20
12 11
PhyBase 19–0
Bit
Mnemonic
Function
63–40
reserved
MBZ
39–12
PhyBase
Base address
11–8
reserved
MBZ
7–0
Type
Memory type
32
8
7
0
reserved
Type
R/W
Reset
0
R/W
U
0
R/W
U
Field Descriptions
Memory Type (Type)—Bits 7–0. Specifies memory type for this memory range.
Base Address (PhysBase)—Bits 39–12. Specifies base address for this memory range
12.1.2.3
MTRRphysMaski Registers
These registers deﬁne the address mask and valid bit for each of the variable MTRRs. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Attempting to modify reserved bits in MTRRphysMaski will result in a #GP(0).
266
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
MTRRphysMask0–7 Registers
MSRs 0201h, 0203h, 0205, 0207h,
0209h, 020Bh, 020Dh, 020Fh
63
40 39
32
reserved
PhyMask 27–20
12 11 10
PhyMask 19–0
Bit
Mnemonic
Function
63–40
reserved
MBZ
39–12
PhysMask
0
Valid
31
reserved
R/W
Reset
Address mask
R/W
U
R/W
11
Valid
MTRR is valid
10–0
reserved
MBZ
0
0
0
Field Descriptions
MTRR is Valid (Valid)—Bit 11. Indicated MTRR is valid.
Address Mask (PhysMask)—Bits 39–12. Specifies the address mask for this memory range.
12.1.2.4
MTRRfix64K_00000 Register
This register controls the memory types for the ﬁrst 512 Kbyte of physical memory. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix64K_00000 Register
63
56 55
MtrrFix64k7xxxxMemType
31
48 47
MtrrFix64k6xxxxMemType
24 23
MtrrFix64k3xxxxMemType
MSR 0250h
MtrrFix64k5xxxxMemType
16 15
MtrrFix64k2xxxxMemType
40 39
8
MtrrFix64k1xxxxMemType
32
MtrrFix64k4xxxxMemType
7
0
MtrrFix64k0xxxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix64k7xxxxMemType
Memory type address 70000–7FFFFh
R/W
U
55–48
MtrrFix64k6xxxxMemType
Memory type address 60000–6FFFFh
R/W
U
47–40
MtrrFix64k5xxxxMemType
Memory type address 50000–5FFFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Chapter 12
Processor Conﬁguration Registers
267
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Bit
Mnemonic
Function
R/W
Reset
39–32
MtrrFix64k4xxxxMemType
Memory type address 40000–4FFFFh
R/W
U
31–24
MtrrFix64k3xxxxMemType
Memory type address 30000–3FFFFh
R/W
U
23–16
MtrrFix64k2xxxxMemType
Memory type address 20000–2FFFFh
R/W
U
15–8
MtrrFix64k1xxxxMemType
Memory type address 10000–1FFFFh
R/W
U
7–0
MtrrFix64k0xxxxMemType
Memory type address 00000–0FFFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address 00000–0FFFFh (MtrrFix64k0xxxxMemType)—Bits 7–0. Memory type
for physical address 00000–0FFFFh.
Memory Type Address 10000–1FFFFh (MtrrFix64k1xxxxMemType)—Bits 15–8. Memory type
for physical address 10000–1FFFFh.
Memory Type Address 20000–2FFFFh (MtrrFix64k2xxxxMemType)—Bits 23–16. Memory type
for physical address 20000–2FFFFh.
Memory Type Address 30000–3FFFFh (MtrrFix64k3xxxxMemType)—Bits 31–24. Memory type
for physical address 30000–3FFFFh.
Memory Type Address 40000–4FFFFh (MtrrFix64k4xxxxMemType)—Bits 39–32. Memory type
for physical address 40000–4FFFFh.
Memory Type Address 50000–5FFFFh (MtrrFix64k5xxxxMemType)—Bits 47–40. Memory type
for physical address 50000–5FFFFh.
Memory Type Address 60000–6FFFFh (MtrrFix64k6xxxxMemType)—Bits 55–48. Memory type
for physical address 60000–6FFFFh.
Memory Type Address 70000–7FFFFh (MtrrFix64k7xxxxMemType)—Bits 63–56. Memory type
for physical address 70000–7FFFFh.
12.1.2.5
MTRRfix16K_80000 Register
This register controls the memory types for physical memory addresses 80000–9FFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix16K_80000 Register
63
56 55
MtrrFix16k9CxxxMemType
268
MSR 0258h
48 47
MtrrFix16k98xxxMemType
40 39
MtrrFix16k94xxxMemType
Processor Conﬁguration Registers
32
MtrrFix16k90xxxMemType
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
31
24 23
MtrrFix16k8CxxxMemType
16 15
MtrrFix16k88xxxMemType
8
MtrrFix16k84xxxMemType
7
0
MtrrFix16k80xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix16k9CxxxMemType Memory type address 9C000–9FFFFh
R/W
U
55–48
MtrrFix16k98xxxMemType
Memory type address 98000–9BFFFh
R/W
U
47–40
MtrrFix16k94xxxMemType
Memory type address 94000–97FFFh
R/W
U
39–32
MtrrFix16k90xxxMemType
Memory type address 90000–93FFFh
R/W
U
31–24
MtrrFix16k8CxxxMemType Memory type address 8C000–8FFFFh
R/W
U
23–16
MtrrFix16k88xxxMemType
Memory type address 88000–8BFFFh
R/W
U
15–8
MtrrFix16k84xxxMemType
Memory type address 84000–87FFFh
R/W
U
7–0
MtrrFix16k80xxxMemType
Memory type address 80000–83FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address 80000–83FFFh (MtrrFix16k80xxxMemType)—Bits 7–0. Memory type
for physical address 80000–83FFFh.
Memory Type Address 84000–87FFFh (MtrrFix16k84xxxMemType)—Bits 15–8. Memory type
for physical address 84000–87FFFh.
Memory Type Address 88000–8BFFFh (MtrrFix16k88xxxMemType)—Bits 23–16. Memory type
for physical address 88000–8BFFFh.
Memory Type Address 8C000–8FFFFh (MtrrFix16k8CxxxMemType)—Bits 31–24. Memory
type for physical address 8C000–8FFFFh.
Memory Type Address 90000–93FFFh (MtrrFix16k90xxxMemType)—Bits 39–32. Memory type
for physical address 90000–93FFFh.
Memory Type Address 94000–97FFFh (MtrrFix16k94xxxMemType)—Bits 47–40. Memory type
for physical address 94000–97FFFh.
Memory Type Address 98000–9BFFFh (MtrrFix16k98xxxMemType)—Bits 55–48. Memory type
for physical address 98000–9BFFFh.
Memory Type Address 9C000–9FFFFh (MtrrFix16k9CxxxMemType)—Bits 63–56. Memory
type for physical address 9C000–9FFFFh.
12.1.2.6
MTRRfix16K_A0000 Register
This register controls the memory types for physical memory addresses A0000–BFFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
Chapter 12
Processor Conﬁguration Registers
269
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
MTRRfix16K_A0000 Register
63
56 55
MtrrFix16kBCxxxMemType
31
MtrrFix16kACxxxMemType
MSR 0259h
48 47
MtrrFix16kB8xxxMemType
24 23
September 2003
MtrrFix16kB4xxxMemType
16 15
MtrrFix16kA8xxxMemType
40 39
8
MtrrFix16kA4xxxMemType
32
MtrrFix16kB0xxxMemType
7
0
MtrrFix16kA0xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix16kBCxxxMemType Memory type address BC000–BFFFFh
R/W
U
55–48
MtrrFix16kB8xxxMemType
Memory type address B8000–BBFFFh
R/W
U
47–40
MtrrFix16kB4xxxMemType
Memory type address B4000–B7FFFh
R/W
U
39–32
MtrrFix16kB0xxxMemType
Memory type address B0000–B3FFFh
R/W
U
31–24
MtrrFix16kACxxxMemType Memory type address AC000–AFFFFh
R/W
U
23–16
MtrrFix16kA8xxxMemType
Memory type address A8000–ABFFFh
R/W
U
15–8
MtrrFix16kA4xxxMemType
Memory type address A4000–A7FFFh
R/W
U
7–0
MtrrFix16kA0xxxMemType
Memory type address A0000–A3FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address A0000–A3FFFh (MtrrFix16kA0xxxMemType)—Bits 7–0. Memory type
for physical address A0000–A3FFFh.
Memory Type Address A4000–A7FFFh (MtrrFix16kA4xxxMemType)—Bits 15–8. Memory type
for physical address A4000–A7FFFh.
Memory Type Address A8000–ABFFFh (MtrrFix16kA8xxxMemType)—Bits 23–16. Memory
type for physical address A8000–ABFFFh.
Memory Type Address AC000–AFFFFh (MtrrFix16kACxxxMemType)—Bits 31–24. Memory
type for physical address AC000–AFFFFh.
Memory Type Address B0000–B3FFFh (MtrrFix16kB0xxxMemType)—Bits 39–32. Memory
type for physical address B0000–B3FFFh.
Memory Type Address B4000–B7FFFh (MtrrFix16kB4xxxMemType)—Bits 47–40. Memory
type for physical address B4000–B7FFFh.
Memory Type Address B8000–BBFFFh (MtrrFix16kB8xxxMemType)—Bits 55–48. Memory
type for physical address B8000–BBFFFh.
Memory Type Address BC000–BFFFFh (MtrrFix16kBCxxxMemType)—Bits 63–56. Memory
type for physical address BC000–BFFFFh.
270
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
12.1.2.7
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
MTRRfix4K_C0000 Register
This register controls the memory types for physical memory addresses C0000–C7FFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_C0000 Register
63
MSR 0268h
56 55
MtrrFix4kC7xxxMemType
31
48 47
MtrrFix4kC6xxxMemType
24 23
MtrrFix4kC3xxxMemType
MtrrFix4kC5xxxMemType
16 15
MtrrFix4kC2xxxMemType
40 39
8
MtrrFix4kC1xxxMemType
32
MtrrFix4kC4xxxMemType
7
0
MtrrFix4kC0xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kC7xxxMemType
Memory type address C7000–C7FFFh
R/W
U
55–48
MtrrFix4kC6xxxMemType
Memory type address C6000–C6FFFh
R/W
U
47–40
MtrrFix4kC5xxxMemType
Memory type address C5000–C5FFFh
R/W
U
39–32
MtrrFix4kC4xxxMemType
Memory type address C4000–C4FFFh
R/W
U
31–24
MtrrFix4kC3xxxMemType
Memory type address C3000–C3FFFh
R/W
U
23–16
MtrrFix4kC2xxxMemType
Memory type address C2000–C2FFFh
R/W
U
15–8
MtrrFix4kC1xxxMemType
Memory type address C1000–C1FFFh
R/W
U
7–0
MtrrFix4kC0xxxMemType
Memory type address C0000–C0FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address C0000–C0FFFh (MtrrFix4kC0xxxMemType)—Bits 7–0. Memory type
for physical address C0000–C0FFFh.
Memory Type Address C1000–C1FFFh (MtrrFix4kC1xxxMemType)—Bits 15–8. Memory type
for physical address C1000–C1FFFh.
Memory Type Address C2000–C2FFFh (MtrrFix4kC2xxxMemType)—Bits 23–16. Memory type
for physical address C2000–C2FFFh.
Memory Type Address C3000–C3FFFh (MtrrFix4kC3xxxMemType)—Bits 31–24. Memory type
for physical address C3000–C3FFFh.
Memory Type Address C40000–C4FFFh (MtrrFix4kC4xxxMemType)—Bits 39–32. Memory
type for physical address C4000–C4FFFh.
Chapter 12
Processor Conﬁguration Registers
271
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Memory Type Address C5000–C5FFFh (MtrrFix4kC5xxxMemType)—Bits 47–40. Memory type
for physical address C5000–C5FFFh.
Memory Type Address C6000–C6FFFh (MtrrFix4kC6xxxMemType)—Bits 55–48. Memory type
for physical address C6000–C6FFFh.
Memory Type Address C7000–C7FFFh (MtrrFix4kC7xxxMemType)—Bits 63–56. Memory type
for physical address C7000–C7FFFh.
12.1.2.8
MTRRfix4K_C8000 Register
This register controls the memory types for physical memory addresses C8000–CFFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_C8000 Register
63
MSR 0269h
56 55
MtrrFix4kCFxxxMemType
31
48 47
MtrrFix4kCExxxMemType
24 23
MtrrFix4kCBxxxMemType
MtrrFix4kCDxxxMemType
16 15
MtrrFix4kCAxxxMemType
40 39
8
MtrrFix4kC9xxxMemType
32
MtrrFix4kCCxxxMemType
7
0
MtrrFix4kC8xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kCFxxxMemType
Memory type for address CF000–CFFFFh
R/W
U
55–48
MtrrFix4kCExxxMemType
Memory type for address CE000–CEFFFh
R/W
U
47–40
MtrrFix4kCDxxxMemType
Memory type for address CD000–CDFFFh
R/W
U
39–32
MtrrFix4kCCxxxMemType
Memory type for address CC000–CCFFFh
R/W
U
31–24
MtrrFix4kCBxxxMemType
Memory type for address CB000–CBFFFh
R/W
U
23–16
MtrrFix4kCAxxxMemType
Memory type for address CA000–CAFFFh
R/W
U
15–8
MtrrFix4kC9xxxMemType
Memory type for address C9000–C9FFFh
R/W
U
7–0
MtrrFix4kC8xxxMemType
Memory type for address C8000–C8FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address C8000–C8FFFh (MtrrFix4kC8xxxMemType)—Bits 7–0. Memory type
for physical address C8000–C8FFFh.
Memory Type Address C9000–C9FFFh (MtrrFix4kC9xxxMemType)—Bits 15–8. Memory type
for physical address C9000–C9FFFh.
272
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Memory Type Address CA000–CAFFFh (MtrrFix4kCAxxxMemType)—Bits 23–16. Memory
type for physical address CA000–CAFFFh.
Memory Type Address CB000–CBFFFh (MtrrFix4kCBxxxMemType)—Bits 31–24. Memory
type for physical address CB000–CBFFFh.
Memory Type Address CC0000–CCFFFh (MtrrFix4kCCxxxMemType)—Bits 39–32. Memory
type for physical address CC000–CCFFFh.
Memory Type Address CD000–CDFFFh (MtrrFix4kCDxxxMemType)—Bits 47–40. Memory
type for physical address CD000–CDFFFh.
Memory Type Address CE000–CEFFFh (MtrrFix4kCExxxMemType)—Bits 55–48. Memory
type for physical address CE000–CEFFFh.
Memory Type Address CF000–CFFFFh (MtrrFix4kCFxxxMemType)—Bits 63–56. Memory
type for physical address CF000–CFFFFh.
12.1.2.9
MTRRfix4K_D0000 Register
This register controls the memory types for physical memory addresses D0000–D7FFF. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_D0000 Register
63
MSR 026Ah
56 55
MtrrFix4kD7xxxMemType
31
48 47
MtrrFix4kD6xxxMemType
24 23
MtrrFix4kD3xxxMemType
MtrrFix4kD5xxxMemType
16 15
MtrrFix4kD2xxxMemType
40 39
8
MtrrFix4kD1xxxMemType
32
MtrrFix4kD4xxxMemType
7
0
MtrrFix4kD0xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kD7xxxMemType
Memory type address D7000–D7FFFh
R/W
U
55–48
MtrrFix4kD6xxxMemType
Memory type address D6000–D6FFFh
R/W
U
47–40
MtrrFix4kD5xxxMemType
Memory type address D5000–D5FFFh
R/W
U
39–32
MtrrFix4kD4xxxMemType
Memory type address D4000–D4FFFh
R/W
U
31–24
MtrrFix4kD3xxxMemType
Memory type address D3000–D3FFFh
R/W
U
23–16
MtrrFix4kD2xxxMemType
Memory type address D2000–D2FFFh
R/W
U
15–8
MtrrFix4kD1xxxMemType
Memory type address D1000–D1FFFh
R/W
U
7–0
MtrrFix4kD0xxxMemType
Memory type address D0000–D0FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Chapter 12
Processor Conﬁguration Registers
273
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Memory Type Address D0000–D0FFFh (MtrrFix4kD0xxxMemType)—Bits 7–0. Memory type
for physical address D0000–D0FFFh.
Memory Type Address D1000–D1FFFh (MtrrFix4kD1xxxMemType)—Bits 15–8. Memory type
for physical address D1000–D1FFFh.
Memory Type Address D2000–D2FFFh (MtrrFix4kD2xxxMemType)—Bits 23–16. Memory type
for physical address D2000–D2FFFh.
Memory Type Address D3000–D3FFFh (MtrrFix4kD3xxxMemType)—Bits 31–24. Memory type
for physical address D3000–D3FFFh.
Memory Type Address D40000–D4FFFh (MtrrFix4kD4xxxMemType)—Bits 39–32. Memory
type for physical address D4000–D4FFFh.
Memory Type Address D5000–D5FFFh (MtrrFix4kD5xxxMemType)—Bits 47–40. Memory type
for physical address D5000–D5FFFh.
Memory Type Address D6000–D6FFFh (MtrrFix4kD6xxxMemType)—Bits 55–48. Memory type
for physical address D6000–D6FFFh.
Memory Type Address D7000–D7FFFh (MtrrFix4kD7xxxMemType)—Bits 63–56. Memory type
for physical address D7000–D7FFFh.
12.1.2.10 MTRRfix4K_D8000 Register
This register controls the memory types for physical memory addresses D8000–DFFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_D8000 Register
63
MSR 026Bh
56 55
MtrrFix4kDFxxxMemType
31
48 47
MtrrFix4kDExxxMemType
24 23
MtrrFix4kDBxxxMemType
MtrrFix4kDDxxxMemType
16 15
MtrrFix4kDAxxxMemType
40 39
8
MtrrFix4kD9xxxMemType
32
MtrrFix4kDCxxxMemType
7
0
MtrrFix4kD8xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kDFxxxMemType
Memory type address DF000–DFFFFh
R/W
U
55–48
MtrrFix4kDExxxMemType
Memory type address DE000–DEFFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
274
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bit
Mnemonic
Function
R/W
Reset
47–40
MtrrFix4kDDxxxMemType
Memory type address DD000–DDFFFh
R/W
U
39–32
MtrrFix4kDCxxxMemType
Memory type address DC000–DCFFFh
R/W
U
31–24
MtrrFix4kDBxxxMemType
Memory type address DB000–DBFFFh
R/W
U
23–16
MtrrFix4kDAxxxMemType
Memory type address DA000–DAFFFh
R/W
U
15–8
MtrrFix4kD9xxxMemType
Memory type address D9000–D9FFFh
R/W
U
7–0
MtrrFix4kD8xxxMemType
Memory type address D8000–D8FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address D8000–D8FFFh (MtrrFix4kD8xxxMemType)—Bits 7–0. Memory type
for physical address D8000–D8FFFh.
Memory Type Address D9000–D9FFFh (MtrrFix4kD9xxxMemType)—Bits 15–8. Memory type
for physical address D9000–D9FFFh.
Memory Type Address DA000–DAFFFh (MtrrFix4kDAxxxMemType)—Bits 23–16. Memory
type for physical address DA000–DAFFFh.
Memory Type Address DB000–DBFFFh (MtrrFix4kDBxxxMemType)—Bits 31–24. Memory
type for physical address DB000–DBFFFh.
Memory Type Address DC0000–DCFFFh (MtrrFix4kDCxxxMemType)—Bits 39–32. Memory
type for physical address DC000–DCFFFh.
Memory Type Address DD000–DDFFFh (MtrrFix4kDDxxxMemType)—Bits 47–40. Memory
type for physical address DD000–DDFFFh.
Memory Type Address DE000–DEFFFh (MtrrFix4kDExxxMemType)—Bits 55–48. Memory
type for physical address DE000–DEFFFh.
Memory Type Address DF000–DFFFFh (MtrrFix4kDFxxxMemType)—Bits 63–56. Memory
type for physical address DF000–DFFFFh.
12.1.2.11 MTRRfix4K_E0000 Register
This register controls the memory types for physical memory addresses E0000–E7FFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_E0000 Register
63
56 55
MtrrFix4kE7xxxMemType
Chapter 12
MSR 026Ch
48 47
MtrrFix4kE6xxxMemType
40 39
MtrrFix4kE5xxxMemType
Processor Conﬁguration Registers
32
MtrrFix4kE4xxxMemType
275
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
31
8
24 23
MtrrFix4kE3xxxMemType
16 15
MtrrFix4kE2xxxMemType
MtrrFix4kE1xxxMemType
Rev. 3.06
September 2003
7
0
MtrrFix4kE0xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kE7xxxMemType
Memory type for physical addr E7000–E7FFFh
R/W
U
55–48
MtrrFix4kE6xxxMemType
Memory type for physical addr E6000–E6FFFh
R/W
U
47–40
MtrrFix4kE5xxxMemType
Memory type for physical addr E5000–E5FFFh
R/W
U
39–32
MtrrFix4kE4xxxMemType
Memory type for physical addr E4000–E4FFFh
R/W
U
31–24
MtrrFix4kE3xxxMemType
Memory type for physical addr E3000–E3FFFh
R/W
U
23–16
MtrrFix4kE2xxxMemType
Memory type for physical addr E2000–E2FFFh
R/W
U
15–8
MtrrFix4kE1xxxMemType
Memory type for physical addr E1000–E1FFFh
R/W
U
7–0
MtrrFix4kE0xxxMemType
Memory type for physical addr E0000–E0FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address E0000–E0FFFh (MtrrFix4kE0xxxMemType)—Bits 7–0. Memory type
for physical address E0000–E0FFFh.
Memory Type Address E1000–E1FFFh (MtrrFix4kE1xxxMemType)—Bits 15–8. Memory type
for physical address E1000–E1FFFh.
Memory Type Address E2000–E2FFFh (MtrrFix4kE2xxxMemType)—Bits 23–16. Memory type
for physical address E2000–E2FFFh.
Memory Type Address E3000–E3FFFh (MtrrFix4kE3xxxMemType)—Bits 31–24. Memory type
for physical address E3000–E3FFFh.
Memory Type Address E40000–E4FFFh (MtrrFix4kE4xxxMemType)—Bits 39–32. Memory
type for physical address E4000–E4FFFh.
Memory Type Address E5000–E5FFFh (MtrrFix4kE5xxxMemType)—Bits 47–40. Memory type
for physical address E5000–E5FFFh.
Memory Type Address E6000–E6FFFh (MtrrFix4kE6xxxMemType)—Bits 55–48. Memory type
for physical address E6000–E6FFFh.
Memory Type Address E7000–E7FFFh (MtrrFix4kE7xxxMemType)—Bits 63–56. Memory type
for physical address E7000–E7FFFh.
12.1.2.12 MTRRfix4K_E8000 Register
This register controls the memory types for physical memory addresses E8000–EFFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
276
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
MTRRfix4K_E8000 Register
63
MSR 026Dh
56 55
MtrrFix4kEFxxxMemType
31
48 47
MtrrFix4kEExxxMemType
24 23
MtrrFix4kEBxxxMemType
40 39
MtrrFix4kEDxxxMemType
16 15
MtrrFix4kEAxxxMemType
MtrrFix4kE9xxxMemType
8
32
MtrrFix4kECxxxMemType
7
0
MtrrFix4kE8xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kEFxxxMemType
Memory type for physical addr EF000–EFFFFh
R/W
U
55–48
MtrrFix4kEExxxMemType
Memory type for physical addr EE000–EEFFFh
R/W
U
47–40
MtrrFix4kEDxxxMemType
Memory type for physical addr ED000–EDFFFh
R/W
U
39–32
MtrrFix4kECxxxMemType
Memory type for physical addr EC000–ECFFFh
R/W
U
31–24
MtrrFix4kEBxxxMemType
Memory type for physical addr EB000–EBFFFh
R/W
U
23–16
MtrrFix4kEAxxxMemType
Memory type for physical addr EA000–EAFFFh
R/W
U
15–8
MtrrFix4kE9xxxMemType
Memory type for physical addr E9000–E9FFFh
R/W
U
7–0
MtrrFix4kE8xxxMemType
Memory type for physical addr E8000–E8FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address E8000–E8FFFh (MtrrFix4kE8xxxMemType)—Bits 7–0. Memory type
for physical address E8000–E8FFFh.
Memory Type Address E9000–E9FFFh (MtrrFix4kE9xxxMemType)—Bits 15–8. Memory type
for physical address E9000–E9FFFh.
Memory Type Address EA000–EAFFFh (MtrrFix4kEAxxxMemType)—Bits 23–16. Memory
type for physical address EA000–EAFFFh.
Memory Type Address EB000–EBFFFh (MtrrFix4kEBxxxMemType)—Bits 31–24. Memory
type for physical address EB000–EBFFFh.
Memory Type Address EC0000–ECFFFh (MtrrFix4kECxxxMemType)—Bits 39–32. Memory
type for physical address EC000–ECFFFh.
Memory Type Address ED000–EDFFFh (MtrrFix4kEDxxxMemType)—Bits 47–40. Memory
type for physical address ED000–EDFFFh.
Memory Type Address EE000–EEFFFh (MtrrFix4kEExxxMemType)—Bits 55–48. Memory
type for physical address EE000–EEFFFh.
Memory Type Address EF000–EFFFFh (MtrrFix4kEFxxxMemType)—Bits 63–56. Memory
type for physical address EF000–EFFFFh.
Chapter 12
Processor Conﬁguration Registers
277
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
12.1.2.13 MTRRfix4K_F0000 Register
This register controls the memory types for physical memory addresses F0000–F7FFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_F0000 Register
63
MSR 026Eh
56 55
MtrrFix4kF7xxxMemType
31
48 47
MtrrFix4kF6xxxMemType
24 23
MtrrFix4kF3xxxMemType
40 39
MtrrFix4kF5xxxMemType
16 15
MtrrFix4kF2xxxMemType
MtrrFix4kF1xxxMemType
32
MtrrFix4kF4xxxMemType
8
7
0
MtrrFix4kF0xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kF7xxxMemType
Memory type address F7000–F7FFFh
R/W
U
55–48
MtrrFix4kF6xxxMemType
Memory type address F6000–F6FFFh
R/W
U
47–40
MtrrFix4kF5xxxMemType
Memory type address F5000–F5FFFh
R/W
U
39–32
MtrrFix4kF4xxxMemType
Memory type address F4000–F4FFFh
R/W
U
31–24
MtrrFix4kF3xxxMemType
Memory type address F3000–F3FFFh
R/W
U
23–16
MtrrFix4kF2xxxMemType
Memory type address F2000–F2FFFh
R/W
U
15–8
MtrrFix4kF1xxxMemType
Memory type address F1000–F1FFFh
R/W
U
7–0
MtrrFix4kF0xxxMemType
Memory type address F0000–F0FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address F0000–F0FFFh (MtrrFix4kF0xxxMemType)—Bits 7–0. Memory type
for physical address F0000–F0FFFh.
Memory Type Address F1000–F1FFFh (MtrrFix4kF1xxxMemType)—Bits 15–8. Memory type
for physical address F1000–F1FFFh.
Memory Type Address F2000–F2FFFh (MtrrFix4kF2xxxMemType)—Bits 23–16. Memory type
for physical address F2000–F2FFFh.
Memory Type Address F3000–F3FFFh (MtrrFix4kF3xxxMemType)—Bits 31–24. Memory type
for physical address F3000–F3FFFh.
Memory Type Address F40000–F4FFFh (MtrrFix4kF4xxxMemType)—Bits 39–32. Memory
type for physical address F4000–F4FFFh.
278
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Memory Type Address F5000–F5FFFh (MtrrFix4kF5xxxMemType)—Bits 47–40. Memory type
for physical address F5000–F5FFFh.
Memory Type Address F6000–F6FFFh (MtrrFix4kF6xxxMemType)—Bits 55–48. Memory type
for physical address F6000–F6FFFh.
Memory Type Address F7000–F7FFFh (MtrrFix4kF7xxxMemType)—Bits 63–56. Memory type
for physical address F7000–F7FFFh.
12.1.2.14 MTRRfix4K_F8000 Register
This register controls the memory types for physical memory addresses F8000–FFFFFh. See the
“Memory Type Range Registers” section in Volume 2 of the AMD64 Architecture Programmer’s
Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0).
MTRRfix4K_F8000 Register
63
MSR 026Fh
56 55
MtrrFix4kFFxxxMemType
31
48 47
MtrrFix4kFExxxMemType
24 23
MtrrFix4kFBxxxMemType
40 39
MtrrFix4kFDxxxMemType
16 15
MtrrFix4kFAxxxMemType
MtrrFix4kF9xxxMemType
8
32
MtrrFix4kFCxxxMemType
7
0
MtrrFix4kF8xxxMemType
Bit
Mnemonic
Function
R/W
Reset
63–56
MtrrFix4kFFxxxMemType
Memory type address FF000–FFFFFh
R/W
U
55–48
MtrrFix4kFExxxMemType
Memory type address FE000–FEFFFh
R/W
U
47–40
MtrrFix4kFDxxxMemType
Memory type address FD000–FDFFFh
R/W
U
39–32
MtrrFix4kFCxxxMemType
Memory type address FC000–FCFFFh
R/W
U
31–24
MtrrFix4kFBxxxMemType
Memory type address FB000–FBFFFh
R/W
U
23–16
MtrrFix4kFAxxxMemType
Memory type address FA000–FAFFFh
R/W
U
15–8
MtrrFix4kF9xxxMemType
Memory type address F9000–F9FFFh
R/W
U
7–0
MtrrFix4kF8xxxMemType
Memory type address F8000–F8FFFh
R/W
U
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Memory Type Address F8000–F8FFFh (MtrrFix4kF8xxxMemType)—Bits 7–0. Memory type
for physical address F8000–F8FFFh.
Memory Type Address F9000–F9FFFh (MtrrFix4kF9xxxMemType)—Bits 15–8. Memory type
for physical address F9000–F9FFFh.
Chapter 12
Processor Conﬁguration Registers
279
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Memory Type Address FA000–FAFFFh (MtrrFix4kFAxxxMemType)—Bits 23–16. Memory
type for physical address FA000–FAFFFh.
Memory Type Address FB000–FBFFFh (MtrrFix4kFBxxxMemType)—Bits 31–24. Memory
type for physical address FB000–FBFFFh.
Memory Type Address FC0000–FCFFFh (MtrrFix4kFCxxxMemType)—Bits 39–32. Memory
type for physical address FC000–FCFFFh.
Memory Type Address FD000–FDFFFh (MtrrFix4kFDxxxMemType)—Bits 47–40. Memory
type for physical address FD000–FDFFFh.
Memory Type Address FE000–FEFFFh (MtrrFix4kFExxxMemType)—Bits 55–48. Memory
type for physical address FE000–FEFFFh.
Memory Type Address FF000–FFFFFh (MtrrFix4kFFxxxMemType)—Bits 63–56. Memory type
for physical address FF000–FFFFFh.
12.1.2.15 PAT Register
This register contains the eight page attribute ﬁelds used for specifying memory types for pages. See
the “Page-Attribute Table Mechanism” section in Volume 2 of the AMD64 Architecture
Programmer’s Manual for more information.
Setting any range to an undeﬁned memory type will result in a #GP(0). Setting any reserved bits will
result in a #GP(0).
PAT Register
63
MSR 0277h
59 58
reserved
31
56 55
PA7
27 26
reserved
51 50
reserved
24 23
PA3
48 47
PA6
19 18
reserved
Bit
Mnemonic
Function
63–59
reserved
MBZ
43 42
reserved
16 15
PA2
PA5
11 10
reserved
58–56
PA7
Memory type for Page Attribute index 7
55–51
reserved
MBZ
50–48
PA6
Memory type for Page Attribute index 6
47–43
reserved
MBZ
42–40
PA5
Memory type for Page Attribute index 5
39–35
reserved
MBZ
34–32
PA4
Memory type for Page Attribute index 4
31–27
reserved
MBZ
280
40 39
Processor Conﬁguration Registers
reserved
8
PA1
35 34
7
PA4
3
reserved
R/W
32
2
0
PA0
Reset
0
R/W
0
0
R/W
7
0
R/W
4
0
R/W
6
0
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bit
Mnemonic
Function
R/W
Reset
26–24
PA3
Memory type for Page Attribute index 3
R/W
0
23–19
reserved
MBZ
18–16
PA2
Memory type for Page Attribute index 2
15–11
reserved
MBZ
10–8
PA1
Memory type for Page Attribute index 1
7–3
reserved
MBZ
2–0
PA0
Memory type for Page Attribute index 0
0
R/W
7
0
R/W
4
0
R/W
6
12.1.2.16 MTRRdefType Register
This register enables the MTRRs and deﬁnes the default memory type for memory not within one of
the MTRR ranges. See the “Memory Type Range Registers” section in Volume 2 of the AMD64
Architecture Programmer’s Manual for more information.
Setting MtrrDefMemType to an undeﬁned memory type will result in a #GP(0). Setting any reserved
bits will result in a #GP(0).
MTRRdefType Register
MSR 02FFh
63
32
reserved
8
7
reserved
reserved
MtrrDefTypeFixEn
12 11 10 9
MtrrDefTypeEn
31
0
MtrrDefMemType
Bit
Mnemonic
Function
R/W
Reset
63–12
reserved
MBZ
11
MtrrDefTypeEn
Enable MTRRs
R/W
0
10
MtrrDefTypeFixEn
Enable Fixed Range MTRRs
R/W
0
9–8
reserved
7–0
MtrrDefMemType
Default Memory Type
R/W
0
0
0
* Memory types must be set to values consistent with system hardware.
Field Descriptions
Default Memory Type (MtrrDefMemType)—Bit 7–0.
Enable Fixed Range MTRRs (MtrrDefTypeFixEn)—Bit 10.
Chapter 12
Processor Conﬁguration Registers
281
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Enable MTRRs (MtrrDefTypeEn)—Bit 11.
12.1.3
APIC Registers
12.1.3.1
APIC_BASE Register
This register enables APIC and deﬁnes the APIC base address. It also identiﬁes the Boot Strap
Processor (BSP).
Setting any reserved bits will result in a #GP(0).
APIC_BASE Register
MSR 001Bh
63
40 39
reserved
ApicBase 27–20
Bit
Mnemonic
Function
63–40
reserved
MBZ
8
BSP
ApicBase 19–0
reserved
12 11 10 9
ApicEn
31
32
R/W
7
0
reserved
Reset
0
39–12
ApicBase
APIC Base Address
R/W
0x0FEE00
11
ApicEn
Enable APIC
R/W
0
10–9
reserved
MBZ
8
BSP
Boot Strap Processor
R/W
1 if uniprocessor
or bsp of
multiprocessor
system
7–0
reserved
MBZ
0
0
* APIC configuration must be consistent with system hardware.
Field Descriptions
Boot Strap Processor (Bsp)—Bit 8.
Enable APIC (ApicEn)—Bit 11.
APIC Base Address (ApicBase)—Bits 39–12.
12.1.3.2
EBL_CR_POWERON Register
This read-only register contains the APIC cluster ID. See the “APIC Initialization” section in
Volume 2 of the AMD64 Architecture Programmer’s Manual for more information.
282
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Attempting to write to this register will result in a #GP(0).
EBL_CR_POWERON Register
MSR 002Ah
63
32
reserved
18 17 16 15
ApicClusterID
31
reserved
Bit
Mnemonic
Function
63–18
reserved
MBZ
17–16
ApicClusterID
APIC Cluster ID
15–0
reserved
MBZ
0
reserved
R/W
R
Field Descriptions
APIC Cluster ID (ApicClusterID)—Bits 17–16.
12.1.4
Software Debug Registers
The AMD Athlon™ 64 and AMD Opteron™ processors incorporate extensive debug features. These
include the following control, status, and control-transfer recording MSRs.
•
DebugCtl Register (MSR 01D9h)—Provides additional debug controls over control-transfer
recording and single stepping, as well as external-breakpoint reporting and trace messages.
•
LastBranchFromIP Register (MSR 01DBh)—Loaded with the segment offset of the branch
instruction.
•
LastBranchToIP Register (MSR 01DCh)—Holds the target rIP of the last branch that occurred
before an exception or interrupt.
•
LastExceptionFromIP Register (MSR 01DDh)—Holds the source rIP of the last branch that
occurred before the exception or interrupt.
•
LastExceptionToIP Register (MSR 01DEh)—Holds the target rIP of the last branch that occurred
before the exception or interrupt.
For more detailed information on the use of these MSRs, see the “Debug and Performance
Resources” section in Volume 2 of the AMD64 Architecture Programmer’s Manual.
Chapter 12
Processor Conﬁguration Registers
283
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
12.1.5
Performance Monitoring Registers
12.1.5.1
TSC Register
26094
Rev. 3.06
September 2003
The time-stamp counter (TSC) register maintains a running count of the number of internal processor
clock cycles executed after a reset. It is incremented by 1 on each internal processor clock. When the
TSC overﬂows its 64-bit range, it wraps around to 0. See the “Time-Stamp Counter” section in
Volume 2 of the AMD64 Architecture Programmer’s Manual for more information.
TSC Register
MSR 0010h
63
32
PCLKS 63–32
31
0
PCLKS 31–0
Bit
Mnemonic
Function
R/W
Reset
63–0
PCLKS
Running Clock Cycle Count
R/W
0
Field Descriptions
Processor Clock Cycles (PCLKS)—Bits 63–0. Running count of number of internal processor clock
cycles.
12.2
AMD Athlon™ 64 processor and AMD Opteron™
Processor Model-Specific Registers
Table 57 is a listing of the model-speciﬁc registers supported by the AMD Athlon™ 64 processor and
AMD Opteron™ processor, presented in ascending hexadecimal address order. Register descriptions
follow the table, organized according to the following functions:
•
Features (see page 286)
•
Identiﬁcation (see page 292)
•
Memory typing (see page 292)
•
I/O range registers (see page 293)
•
System call extension registers (see page 295)
•
Segmentation (see page 297)
284
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
•
System management (see page 167)
•
Power management (see page 299)
Table 57.
AMD Athlon™ 64 Processor and AMD Opteron™ Processor MSRs
Address
Register Name
Description
C000_0080h
EFER
page 286
C000_0081h
STAR
page 295
C000_0082h
LSTAR
page 296
C000_0083h
CSTAR
page 296
C000_0084h
SF_MASK
page 297
C000_0100h
FS.Base
page 297
C000_0101h
GS.Base
page 298
C000_0102h
KernelGSbase
page 298
C001_0000h–C001_0003h
PerfEvtSeli
page 246
C001_0004h–C001_0007h
PerfCtri
page 245
C001_0010h
SYSCFG
page 286
C001_0015h
HWCR
page 288
C001_0016h, C001_0018h
IORRBase[1:0]
page 293
C001_0017h, C001_0019h
IORRMask[1:0]
page 294
C001_001Ah
TOP_MEM
page 292
C001_001Dh
TOP_MEM2
page 293
C001_001Eh
MANID
page 292
C001_001Fh
NB_CFG
page 290
C001_0041h
FIDVID_CTL
page 299
C001_0042h
FIDVID_STATUS
page 300
C001_0050–C001_0053h
IOTRAP_ADDRi
page 302
C001_0054h
IOTRAP_CTL
page 303
C001_0111h
SMM_BASE
page 173
C001_0112h
SMM_ADDR
page 178
C001_0113h
SMM_MASK
page 177
Chapter 12
Processor Conﬁguration Registers
285
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
12.2.1
Feature Registers
12.2.1.1
EFER Register
26094
Rev. 3.06
September 2003
This register controls which extended features are enabled. See the “Extended Feature Enable
Register (EFER)” section in Volume 2 of the AMD64 Architecture Programmer’s Manual for more
information.
The LMA bit is a read-only status bit. When writing EFER, the processor will signal a #GP(0) if an
attempt is made to change LMA from its previous value. BIOS should leave this register alone unless
it needs to use one of the features enabled by this register.
EFER Register
MSR C000_0080h
63
32
LME
7
1
reserved
Bit
Mnemonic
Function
R/W
Reset
63–12
reserved
MBZ
R/W
0
11
NXE
No-Execute Page Enable
R/W
0
10
LMA
Long Mode Active
R
0
9
reserved
MBZ
8
LME
Long Mode Enable
7–1
reserved
RAZ
0
SYSCALL
System Call Extension Enable
R
0
R/W
0
R
0
R/W
0
0
SYSCALL
8
LMA
reserved
12 11 10 9
NXE
31
reserved
reserved
Field Descriptions
System Call Extension Enable (SCE)—Bit 0. Enables the system call extension
Long Mode Enable (LME)—Bit 8. Enables the long mode feature
Long Mode Active (LMA)—Bit 10. Indicates the long mode feature is active
No-Execute Page Enable (NXE)—Bit 11. Enables the no-execute page feature
12.2.1.2
SYSCFG Register
This register controls the system conﬁguration.
286
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
The MtrrFixDramModEn bit should be set to 1 during BIOS initalization of the ﬁxed MTRRs, then
cleared to 0 for operation.
SYSCFG Register
MSR C001_0010h
63
32
SetDirtyEnS
reserved
SetDirtyEnO
12 11 10 9
ClVicBlkEn
ChxToDirtyDis
SysUcLockEn
MtrrFixDramEn
MtrrVarDramEn
reserved
MtrrFixDramModEn
22 21 20 19 18 17 16 15
MtrrTom2En
31
8
7
SetDirtyEnE
reserved
5
SysVicLi
mit
4
0
SysAckLimit
Bit
Mnemonic
Function
R/W
Reset
R
0
BIOS
63–22
reserved
RAZ
21
MtrrTom2En
Top of Memory Address Register 2 Enable (reserved)
R/W
0
0
20
MtrrVarDramEn
Top of Memory Address Register and I/O Range
Register Enable
R/W
0
1
19
MtrrFixDramModEn
RdDram and WrDram Bits Modification Enable
R/W
0
0
18
MtrrFixDramEn
Fixed RdDram and WrDram Attributes Enable
R/W
0
1
17
SysUcLockEn
System Interface Lock Command Enable
R/W
1
1
16
ChxToDirtyDis
Change to Dirty Command Disable
R/W
0
0
15–12
reserved
RAZ
R
0
11
ClVicBlkEn
Revision B and earlier revisions: ClVicBlk System
Interface Command Enable
Revision C: Reserved
R/W
0
0
10
SetDirtyEnO
CleanToDirty Command for O->M State Transition
Enable
R/W
1
1
9
SetDirtyEnS
SharedToDirty Command for S->M State Transition
Enable
R/W
1
1
8
SetDirtyEnE
SharedToDirty Command for E->M State Transition
Enable
R/W
0
0
7–5
SysVicLimit
Outstanding Victim Bus Command Limit
R/W
000b
000b
4–0
SysAckLimit
Outstanding Bus Command Limit
R/W
00001b 00001b
Field Descriptions
Outstanding Bus Command Limit (SysAckLimit)—Bit 4–0. Limits maximum number of
outstanding bus commands.
Outstanding Victim Bus Command Limit (SysVicLimit)—Bit 7–5. Limits maximum number of
outstanding victim bus commands.
Chapter 12
Processor Conﬁguration Registers
287
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
SharedToDirty Command for E->M State Transition Enable (SetDirtyEnE)—Bit 8. Enables
SharedToDirty commands for E->M state transitions.
SharedToDirty Command for S->M State Transition Enable (SetDirtyEnS)—Bit 9. Enables
SharedToDirty commands for S->M state transitions.
CleanToDirty Command for O->M State Transition Enable (SetDirtyEnO)—Bit 10. Enables
CleanToDirty commands for O->M state transitions.
ClVicBlk System Interface Command Enable (ClVicBlkEn)—Bit 11. Enables ClVicBlk system
interface command.
Change to Dirty Command Disable (ChxToDirtyDis)—Bit 16. Disables change to dirty
commands, evicts line from DC instead.
System Interface Lock Command Enable (SysUcLockEn)—Bit 17. Enables lock commands on
system interface.
Fixed RdDram and WrDram Attributes Enable (MtrrFixDramEn)—Bit 18. Enables fixed
MTRR RdDram and WrDram attributes.
RdDram and WrDram Bits Modification Enable (MtrrFixDramModEn)—Bit 19. Enables
modification of RdDram and WrDram bits in fixed MTRRS.
Top of Memory Address Register and I/O Range Register Enable (MtrrVarDramEn)—Bit 20.
Enables use of top of memory address register and the I/O range registers.
Top of Memory Address Register 2 Enable (MtrrTom2En)—Bit 21. Enables use of top of
memory address register 2 (reserved).
12.2.1.3
HWCR Register
This register controls the hardware conﬁguration.
Operating systems that maintain page tables in uncacheable memory (UC memory type) must set the
TLBCACHEDIS bit to insure proper operation.
BIOS and SMM handlers may use the WRAP32DIS bit to access physical memory above 4 Gbytes
without switching into 64-bit mode. By setting WRAP32DIS to a 1 in conjunction with setting the
expanded FS or GS base registers, BIOS can size all of physical memory from legacy mode. To do so,
BIOS would write a >32 bit base to the FS or GS base register using WRMSR. Then it would address
2 Gbytes from one of those bases using normal memory reference instructions with a FS or GS
override preﬁx. However, the INVLPG, FST, and SSE store instructions generate 32-bit addresses in
legacy mode, regardless of the state of WRAP32DIS.
The MCi_STATUS_WREN bit can be used to debug Machine Check handlers. When
MCi_STATUS_WREN bit is set, privileged software can write non-zero values to the Machine Check
MSRs (MSRs 017Ah, 0401h, 0405h, 0409h, 040Dh, 0411h, 0402h, 0406h, 040Ah, 040Eh, 0412h)
without generating exceptions, and then simulate a machine check using the "int 18" instruction.
288
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Setting a reserved bit in these MSRs does not generate an exception when this mode is enabled.
However, setting a reserved bit may result in undeﬁned behavior.
HWCR Register
MSR C001_0015h
63
32
Bit
Mnemonic
Function
63–32
reserved
31–30
29–24
3
2
1
SLOWFENCE
R/W
Reset
RAZ
R
0
reserved
SBZ
R/W
0
START_FID
Startup FID Status
R
0
23–20
reserved
RAZ
R
0
19
reserved
SBZ
R/W
0
18
MCi_STATUS_WREN
MCi Status Write Enable
R/W
0
17
WRAP32DIS
32-bit Address Wrap Disable
R/W
0
16
reserved
15
SSEDIS
SSE Instructions Disable
R/W
0
14
RSMSPCYCDIS
Special Bus Cycle On RSM Disable
R/W
0
13
SMISPCYCDIS
Special Bus Cycle On SMI Disable
R/W
0
0
SMMLOCK
4
reserved
5
TLBCACHEDIS
reserved
6
INVD_WBINVD
7
FFDIS
8
reserved
9
DISLOCK
HLTXSPCYCEN
SMISPCYCDIS
SSEDIS
RSMSPCYCDIS
reserved
reserved
WRAP32DIS
START_FID
20 19 18 17 16 15 14 13 12 11
MCi_STATUS_WREN
24 23
reserved
reserved
31 30 29
IGNNE_EM
reserved
0
12
HLTXSPCYCEN
Enable Special Bus Cycle On Exit From HLT
R/W
0
11–9
reserved
SBZ
R/W
0
8
IGNNE_EM
IGNNE Port Emulation Enable
R/W
0
7
DISLOCK
Disable x86 LOCK prefix functionality
R/W
0
6
FFDIS
TLB Flush Filter Disable
R/W
0
5
reserved
SBZ
R/W
0
4
INVD_WBINVD
INVD to WBINVD Conversion
R/W
0
3
TLBCACHEDIS
Cacheable Memory Disable
R/W
0
2
reserved
SBZ
R/W
0
1
SLOWFENCE
Slow SFENCE Enable
R/W
0
0
SMMLOCK
SMM Code Lock
R/W
0
Field Descriptions
SMM Code Lock (SMMLOCK)—Bit 0. Locks SMM code in TSeg range by making it read-only.
Chapter 12
Processor Conﬁguration Registers
289
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Slow SFENCE Enable (SLOWFENCE)—Bit 1. Enable slow sfence.
Cachable Memory Disable (TLBCACHEDIS)—Bit 3. Disable performance improvement that
assumes that the PML4, PDP, PDE and PTE entries are in cacheable memory. If page tables
are uncachable, TLBCACHEDIS must be set to ensure correct functionality.
INVD to WBINVD Conversion (INVD_WBINVD)—Bit 4. Convert INVD to WBINVD.
TLB Flush Filter Disable (FFDIS)—Bit 6. Disable TLB flush filter.
Disable LOCK (DISLOCK)—Bit 7. Disable x86 LOCK prefix functionality.
IGNNE Port Emulation Enable (IGNNE_EM)—Bit 8. Enable emulation of IGNNE port.
Special Bus Cycle On HLT Exit Enable (HLTXSPCYCEN)—Bit 12. Enables special bus cycle
generation on exit from HLT. See “Register Differences in Revisions of the AMD Athlon™
64 and AMD Opteron™ processors” on page 19 for revision information about this field.
Special Bus Cycle On SMI Disable (SMISPCYCDIS)—Bit 13. Disables special bus cycle on SMI.
Special Bus Cycle On RSM Disable (RSMSPCYCDIS)—Bit 14. Disables special bus cycle on
RSM.
SSE Instructions Disable (SSEDIS)—Bit 15. Disables SSE instructions.
32-bit Address Wrap Disable (WRAP32DIS)—Bit 17. Disable 32-bit address wrapping.
MCi Status Write Enable (MCi_STATUS_WREN)—Bit 18. When set, writes by software to
Machine Check Error Status MSRs, Machine Check Error Address MSRs, and Machine
Check MISC MSRs (if implemented) do not cause general protection faults. Such writes
update all implemented bits in these registers. When clear, writing a non-zero pattern to these
registers causes a general protection fault. See the description above for more details.
Startup FID Status (START_FID)—Bits 29–24. Status of the startup FID.
12.2.1.4
NB_CFG Register
Software must perform a read-modify-write to this register to change its value.
NB_CFG Register
MSR C001_001Fh
37 36 35
reserved
290
Processor Conﬁguration Registers
DisDatMsk
63
32
reserved
Chapter 12
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
10 9
DisCohLdtCfg
31 30
reserved
Bit
Mnemonic
63–37
reserved
36
DisDatMsk
35–32
reserved
31
DisCohLdtCfg
30–10
reserved
9
EnRefUseFreeBuf
8–0
reserved
8
0
EnRefUseFreeBuf
26094
Function
Disable Data Mask
Disable Coherent HyperTransport Configuration
Accesses
Enable Display Refresh to Use Free List Buffers
reserved
R/W
Reset
R/W
0
R/W
0
R/W
0
R/W
0
R/W
0
R/W
0
R/W
0
Field Descriptions
Enable Display Refresh to Use Free List Buffers (EnRefUseFreeBuf)—Bit 9. Enables display
refresh request to use free list buffers. See “Register Differences in Revisions of the AMD
Athlon™ 64 and AMD Opteron™ processors” on page 19 for revision information about this
field.
Disable Coherent HyperTransport Configuration Accesses (DisCohLdtCfg)—Bit 31. Disables
automatic routing of PCI configuration accesses to the processor configuration registers.
When set, PCI configuration space accesses which fall within the hard-coded range reserved
for AMD Athlon™ 64 and AMD Opteron™ processor registers (see “Memory System
Configuration Registers” on page 27) are instead routed via the configuration address maps.
This can be used to effectively hide the configuration registers from software if they are routed
to the I/O hub, where they will then get a master abort. It can also be used to provide a means
for an external chip to route processor configuration accesses according to some scheme other
than the hard-coded version described in “Memory System Configuration Registers” on
page 27. When used, this bit needs to be set on all processors in a system. PCI configuration
accesses should not be generated if this bit is not set on all processors.
Disable Data Mask (DisDatMsk)—Bits 36. Disables DRAM data masking function. For all sized
write requests a DRAM read is performed before writing the data. See “Register Differences
in Revisions of the AMD Athlon™ 64 and AMD Opteron™ processors” on page 19 for
revision information about this field.
Chapter 12
Processor Conﬁguration Registers
291
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
12.2.2
Identification Registers
12.2.2.1
MANID Register
26094
Rev. 3.06
September 2003
This status register holds the mask manufacturing identiﬁcation number.
MANID Register
MSR C001_001Eh
63
32
reserved
10 9
reserved
Function
8
ReticleSite
31
7
4
MajorRev
Bit
Mnemonic
63–10
reserved
9–8
ReticleSite
Reticle site
R
7–4
MajorRev
Major mask set revision number
R
3–0
MinorRev
Minor mask set revision number
R
3
0
MinorRev
R/W
Field Descriptions
Minor Mask Set Revision Number (MinorRev)—Bits 3–0.
Major Mask Set Revision Number (MajorRev)—Bits 7–4.
Reticle Site (ReticleSite)—Bits 9–8.
12.2.3
Memory Typing Registers
12.2.3.1
TOP_MEM Register
This register holds the address of the top of memory. When enabled, this address indicates the ﬁrst
byte of I/O above DRAM.
TOP_MEM Register
MSR C001_001Ah
63
40 39
reserved
292
Processor Conﬁguration Registers
32
TOM 16–9
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
31
23 22
0
TOM 8–0
reserved
Bit
Mnemonic
Function
63–40
reserved
RAZ
39–23
TOM
Top of Memory
22–0
reserved
RAZ
R/W
Reset
R/W
Field Descriptions
Top of Memory (TOM)—Bits 39–23. Address of top of memory (8-Mbyte granularity).
12.2.3.2
TOP_MEM2 Register
This register holds the second top of memory address. When enabled by setting the MtrrTom2En bit
in the SYSCFG register, this address indicates the ﬁrst byte of I/O above a second allocation of
DRAM that starts at 4 Gbytes. Hence, the address in TOP_MEM2 should be set above 4 Gbytes.
TOP_MEM2 Register
MSR C001_001Dh
63
40 39
reserved
31
32
TOM2 16–9
23 22
0
reserved
TOM2 8–0
Bit
Mnemonic
Function
63–40
reserved
RAZ
39–23
TOM2
Second Top of Memory
22–0
reserved
RAZ
R/W
Reset
R/W
Field Descriptions
Second Top of Memory (TOM2)—Bits 39–23. Address of second top of memory (8-Mbyte
granularity).
12.2.4
I/O Range Registers
12.2.4.1
IORRBasei Registers
These registers hold the bases of the variable I/O ranges.
Chapter 12
Processor Conﬁguration Registers
293
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
IORRBase0–1 Registers
26094
Rev. 3.06
September 2003
MSRs C001_0016h, C001_0018h
63
40 39
32
12 11
Base 19–0
5
reserved
Bit
Mnemonic
Function
R/W
63–40
reserved
RAZ
39–12
Base
Base address
11–5
reserved
RAZ
4
RdDram
Read from DRAM
R/W
3
WrDram
Write to DRAM
R/W
2–0
reserved
RAZ
4
3
2
WrDram
31
Base 27–20
RdDram
reserved
0
reserved
Reset
R/W
Field Descriptions
Write to DRAM (WrDram)—Bit 3. If set, stores write to DRAM, otherwise I/O for this range.
Read from DRAM (RdDram)—Bit 4. If set, fetches read from DRAM, otherwise from I/O for this
range.
Base Address (Base)—Bits 39–12. Base address for this range.
12.2.4.2
IORRMaski Registers
These registers hold the masks of the variable I/O ranges.
IORRMask0–1 Registers
MSRs C001_0017h, C001_0019h
63
40 39
32
reserved
31
12 11 10
Mask 19–0
Bit
Mnemonic
Function
63–40
reserved
RAZ
39–12
Mask
Address mask
294
Mask 27–20
V
Processor Conﬁguration Registers
0
reserved
R/W
Reset
R/W
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Bit
Mnemonic
Function
R/W
11
V
Enables variable I/O range registers
R/W
10–0
reserved
RAZ
Reset
Field Descriptions
Variable I/O Range (V)—Bit 11. Enables variable I/O range register.
Address Mask (Mask)—Bits 39–12. Address mask.
12.2.5
System Call Extension Registers
The system call extension is enabled by setting bit 0 in the EFER register. This feature adds the
SYSCALL and SYSRET instructions which can be used in ﬂat addressed operating systems as low
latency system calls and returns. See the “SYSCALL and SYSRET” section in Volume 2 of the
AMD64 Architecture Programmer’s Manual for more information.
12.2.5.1
STAR Register
This register holds the target address used by the SYSCALL instruction and the code and stack
segment selector bases used by the SYSCALL and SYSRET instructions.
STAR Register
MSR C000_0081h
63
48 47
SysRetSel
32
SysCallSel
31
0
Target
Bit
Mnemonic
Function
R/W
63–48
SysRetSel
SYSRET CS and SS
R/W
47–32
SysCallSel
SYSCALL CS and SS
R/W
31–0
Target
SYSCALL target address
R/W
Reset
Field Descriptions
SYSCALL Target Address (Target)—Bits 31–0.
SYSCALL CS and SS (SysCallSel)—Bits 47–32.
SYSRET CS and SS (SysRetSel)—Bits 63–48.
Chapter 12
Processor Conﬁguration Registers
295
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
12.2.5.2
26094
Rev. 3.06
September 2003
LSTAR Register
The address stored in this register must be in canonical form (if not canonical, a #GP fault will occur).
LSTAR Register
MSR C000_0082h
63
32
LSTAR 63–32
31
0
LSTAR 31–0
Bit
Mnemonic
Function
R/W
63–0
LSTAR
Long Mode Target Address
R/W
Reset
Field Descriptions
Long Mode Target Address (LSTAR)—Bits 63–0. Target address for 64-bit mode calling programs.
12.2.5.3
CSTAR Register
The address stored in this register must be in canonical form (if not canonical, a #GP fault ﬁll occur).
CSTAR Register
MSR C000_0083h
63
32
CSTAR 63–32
31
0
CSTAR 31–0
Bit
Mnemonic
Function
R/W
63–0
CSTAR
Compatibility mode target address
R/W
Reset
Field Descriptions
Compatibility Mode Target Address (CSTAR)—Bits 63–0. Target address for compatibility mode
calling programs.
296
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
12.2.5.4
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
SF_MASK Register
This register holds the EFLAGS mask used by the SYSCALL instruction. Each one in this mask will
clear the corresponding EFLAGS bit when executing the SYSCALL instruction.
SF_MASK Register
MSR C000_0084h
63
32
reserved
31
0
MASK
Bit
Mnemonic
Function
R/W
63–32
reserved
RAZ
31–0
MASK
SYSCALL Flag Mask
Reset
R/W
Field Descriptions
SYSCALL Flag Mask (MASK)—Bits 31–0.
12.2.6
Segmentation Registers
12.2.6.1
FS.Base Register
This register provides access to the expanded 64 bit FS segment base. The address stored in this
register must be in canonical form (if not canonical, a #GP fault ﬁll occur).
FS.Base Register
MSR C000_0100h
63
32
FS_BASE (63–32)
31
0
FS_BASE (31–0(
Bit
Mnemonic
Function
R/W
63–0
FS_BASE
Expanded FS segment base
R/W
Chapter 12
Processor Conﬁguration Registers
Reset
297
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Expanded FS Segment Base (FS_BASE)—Bits 63–0.
12.2.6.2
GS.Base Register
This register provides access to the expanded 64 bit GS segment base. The address stored in this
register must be in canonical form (if not canonical, a #GP fault ﬁll occur).
GS.Base Register
MSR C000_0101h
63
32
GS_BASE 63–32
31
0
GS_BASE 31–0
Bit
Mnemonic
Function
R/W
63–0
GS_BASE
Expanded GS segment base
R/W
Reset
Field Descriptions
Expanded GS Segment Base (GS_BASE)—Bits 63–0.
12.2.6.3
KernelGSbase Register
This register holds the kernel data structure pointer which can be swapped with the GS_BASE
register using the new 64-bit mode instruction SwapGS. See the “SWAPGS Instruction” section in
Volume 2 of the AMD64 Architecture Programmer’s Manual for more information. The address
stored in this register must be in canonical form (if not canonical, a #GP fault will occur).
KernelGSbase Register
MSR C000_0102h
63
32
KernelGSBase 63–32
31
0
KernelGSBase 31–0
Bit
Mnemonic
Function
R/W
63–0
KernelGSBase
Kernel data structure pointer
R/W
298
Processor Conﬁguration Registers
Reset
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
Field Descriptions
Kernel Data Structure Pointer (KernelGSBase)—Bits 63–0.
12.2.7
Power Management Registers
Mobile processors contain AMD PowerNow!™ technology hardware that allows the processor
operating voltage and frequency to by dynamically controlled for power management purposes. Use
CPUID function 8000_0007h (Get Advanced Power Management Feature Flags) to determine the
thermal and power management capabilities of the processor. Refer to “Processor Performance
States” on page 219 regarding the use of the FIDVID_STATUS and FIDVID_CTL MSRs.
12.2.7.1
FIDVID_CTL Register
This register holds conﬁguration information relating to power management control. Accessing this
register is allowed if the device supports either FID control or VID control as indicated by CPUID. If
neither FID control nor VID control is indicated, accessing the FIDVID_CTL register will cause a
GP# fault.
This register is used to specify new Frequency ID (FID) and/or Voltage ID (VID) codes to which to
change on the next FID/VID change. This register is also used to control how long to wait following a
FID/VID change before the PLL and voltage regulator are stable at the new FID/VID. This register is
persistent through a warm reset and is initialized to 0 on a cold reset.
FIDVID_CTL Register
52 51
32
reserved
StpGntTOCnt
17 16 15
InitFidVid
31
reserved
reserved
Bit
Mnemonic
Function
63–52
reserved
MBZ
51–32
StpGntTOCnt
Stop Grant Time-Out Count
31–17
reserved
MBZ
16
InitFidVid
Initiate FID/VID Change
15–13
reserved
MBZ
12–8
NewVID
New VID
7–6
reserved
MBZ
5–0
NewFID
New FID
Chapter 12
13 12
8
NewVID
Processor Conﬁguration Registers
7
6
5
0
reserved
63
MSR C001_0041h
R/W
NewFID
Reset
0
R/W
0
0
W
0
0
R/W
0
R
0
R/W
0
299
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
Field Descriptions
Stop Grant Time-Out Count (StpGntTOCnt)—Bits 51-32. This field carries a count of system
clock cycles (5 ns) that must elapse from the time a new FID is applied until the time that the
PLL is stable at the new FID.
Initiate FID/VID Change (InitFidVid)—Bits 16. Writing this bit to a 1 initiates a FID/VID change.
In a multiprocessor system, only the bootstrap processor sees this bit written as a 1. Writing
this bit to a 1 initiates the FID change special bus cycle; in Revision B and earlier revisions,
the VID change special bus cycle was also initiated. This is a write-only bit and always reads
as 0. The status of the resulting FID/VID change can be determined from the
FIDVID_STATUS register.
New VID (NewVID)—Bits 12–8. This field is the new VID to transition to. If an attempt is made to
write a NewVID value that corresponds to a voltage greater than the voltage that MaxVID
corresponds to in the FIDVID_STATUS register then the MaxVID value is written instead.
See Table 48 on page 235 for VID values and their corresponding voltages.
New FID (NewFID)—Bits 5–0. This field is the new FID to transition to. If an attempt is made to
write a NewFID value greater than MaxFID in the FIDVID_STATUS register then the
MaxFID value is written instead. See Table 58 for FID code translations.
Table 58. FID Code Translations
Reference
Clock
Multiplier
6-Bit FID Code
Reference
Clock
Multiplier
6-Bit FID Code
4x
00_0000b
15x
01_0110b
5x
00_0010b
16x
01_1000b
6x
00_0100b
17x
01_1010b
7x
00_0110b
18x
01_1100b
8x
00_1000b
19X
01_1110b
9x
00_1010b
20X
10_0000b
10x
00_1100b
21X
10_0010b
11x
00_1110b
22x
10_0100b
12x
01_0000b
23x
10_0110b
13x
01_0010b
24x
10_1000b
14x
01_0100b
25x
10_1010b
Note: Encodings not listed are reserved. Using reserved encodings will cause
unpredictable behavior.
12.2.7.2
FIDVID_STATUS Register
This register holds status of the current FID, the current VID, pending changes, and maximum values.
Accessing this register is allowed if the device supports either FID control or VID control as indicated
300
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
via CPUID. If neither FID control nor VID control is indicated, accessing the FIDVID_STATUS
register will cause a GP# fault.
See Table 58 on page 300 for a complete list of FID values and their corresponding reference clock
multiplier values. See Table 48 on page 235 for VID values and their corresponding voltages.
FIDVID_STATUS Register
48 47
reserved
24 23 22 21
MaxRampVID
reserved
reserved
FidVidPending
31 30 29 28
MaxVID
45 44
reserved
40 39
StartVID
16 15 14 13
MaxFID
Bit
Mnemonic
Function
63–53
reserved
RAZ
37 36
32
reserved
8
StartFID
7
6
CurrVID
5
0
reserved
53 52
reserved
63
MSR C001_0042h
CurrFID
R/W
Reset
R
0
52–48
MaxVID
Max VID
R
0
47–45
reserved
RAZ
R
0
44–40
StartVID
Startup VID
R
0
39–37
reserved
RAZ
R
0
36–32
CurrVID
Current VID
R
0
31
FidVidPending
FID/VID Change Pending
R
0
30–29
reserved
RAZ
R
0
28–24
MaxRampVID
Max Ramp VID
R
0
23–22
reserved
RAZ
R
0
21–16
MaxFID
Max FID
R
0
15–14
reserved
RAZ
R
0
13–8
StartFID
Startup FID
R
0
7–6
reserved
RAZ
R
0
5–0
CurrFID
Current FID
R
0
Field Descriptions
Max VID (MaxVID)—Bits 52–48. This field indicates the VID code associated with the maximum
voltage software can select.
Startup VID (StartVID)—Bits 44–40. This field indicates the startup VID.
Current VID (CurrVID)—Bits 36–32. This field indicates the current VID.
Chapter 12
Processor Conﬁguration Registers
301
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
FID/VID Change Pending (FidVidPending)—Bit 31. This bit indicates that a FID/VID change is
pending. It is set to 1 by hardware when the InitFidVid bit is set in the FIDVID_CTL register.
It is cleared by hardware when the FID/VID change has completed.
Max Ramp VID (MaxRampVID)—Bit 28–24. This field indicates the max Ramp VID.
Max FID (MaxFID)—Bit 21–16. This field indicates the max FID.
Startup FID (StartFID)—Bit 13–8. This field indicates the startup FID.
Current FID (CurrFID)—Bit 5–0. This field indicates the current FID.
12.2.8
IO and Configuration Space Trapping to SMI
IO and conﬁguration space trapping to SMI is a mechanism for executing SMI handler if a speciﬁc
IO access to one of the speciﬁed addresses is detected. Access address and access type checking is
done before IO instruction execution. If the access address and access type match one of the speciﬁed
IO address and access types, the SMI handler is executed, the IO instruction is not executed, and a
breakpoint set on the IO instruction is not taken. The IO instruction can be executed inside the SMI
handler. This mechanism has higher priority than the SMI interrupt. If the trap is not taken, check for
SMI interrupt is done after the IO instruction has been executed.
Conﬁguration accesses are special IO accesses. An IO access is deﬁned as a conﬁguration access
when IO instruction address bits 31-0 are CFCh, CFDh, CFEh, or CFFh. The access address for a
conﬁguration space access is the current value of doubleword IO register CF8h. The access address
for an IO access that is not a conﬁguration access is equivalent to the IO instruction address, bits 31–
0.
The access address is compared with SmiAddr, and the instruction access type is compared with the
enabled access types deﬁned by ConﬁgSMI, SmiOnRdEn, and SmiOnWrEn. Access address bits 230 can be masked with SmiMask. Fields SmiAddr, SmiMask, ConﬁgSMI, SmiOnRdEn, and
SmiOnWrEn are deﬁned in IOTRAP_ADDRi registers (i = 0,1,2,3).
An SMI handler executed as a result of IO and conﬁguration space trapping will by default only be
detected by the processor executing the IO instruction. The user has the ability to enable an SMI
special bus cycle and RSM special bus cycle generation.
IO and conﬁguration space trapping to SMI applies only to single IO instructions; it does not apply to
string and REP IO instructions.
12.2.8.1
IOTRAP_ADDRi Registers
See “Register Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™ processors”
on page 19 for revision information about this register.
302
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
IOTRAP_ADDRi Registers
ConfigSmi
SmiOnWrEn
SmiOnRdEn
63 62 61 60
MSRs C001_0050h, C001_0051h,
C001_0052h, C001_0053h
56 55
32
reserved
SmiMask
31
0
SmiAddr
Bit
Mnemonic
Function
R/W
Reset
63
SmiOnRdEn
Enable SMI on IO Read
R/W
0
62
SmiOnWrEn
Enable SMI on IO Write
R/W
0
61
ConfigSmi
Configuration Space SMI
R/W
0
60-56
reserved
RAZ
55–32
SmiMask
SMI Mask
R/W
0
31–0
SmiAddr
SMI Address
R/W
0
R
Field Descriptions
Enable SMI on IO Read (SmiOnRdEn)—Bits 63. Enables SMI generation on a read access.
Enable SMI on IO Write (SmiOnWrEn)—Bit 62. Enables SMI generation on a write access.
Configuration Space SMI (ConfigSmi)—Bit 61. 1 = configuration access (see “IO and
Configuration Space Trapping to SMI” on page 302 for more information on configuration
access detection); 0 = IO access different than configuration access.
SMI Mask (SmiMask)—Bits 55–32. SMI IO trap mask. 0 = mask address bit; 1 = do not mask
address bit.
SMI Address (SmiAddr)—Bits 31–0. SMI IO trap address.
12.2.8.2
IOTRAP_CTL Register
See “Register Differences in Revisions of the AMD Athlon™ 64 and AMD Opteron™ processors”
on page 19 for revision information about this register.
IOTRAP_CTL Register
MSR C001_0054h
63
32
reserved
Chapter 12
Processor Conﬁguration Registers
303
7
6
5
SmiEn_3
SmiSts_3
SmiEn_2
IoTrapCtlRsmSpcEn
IoTrapEn
IoTrapCtlSmiSpcEn
reserved
reserved
R/W
4
3
2
Bit
Mnemonic
Function
63–16
reserved
RAZ
15
IoTrapEn
IO Trap Enable
14
IoTrapCtlSmiSpcEn
IO Trap Control SMI Special Cycle Enable
R/W
0
13
IoTrapCtlRsmSpcEn
IO Trap Control RSM Special Cycle Enable
R/W
0
12-8
reserved
RAZ
7
SmiEn_3
SMI Enable 3
6
SmiSts_3
SMI Status 3
R/W
0
5
SmiEn_2
SMI Enable 2
R/W
0
4
SmiSts_2
SMI Status 2
R/W
0
3
SmiEn_1
SMI Enable 1
R/W
0
2
SmiSts_1
SMI Status 1
R/W
0
1
SmiEn_0
SMI Enable 0
R/W
0
0
SmiSts_0
SMI Status 0
R/W
0
1
0
SmiEn_0
8
16 15 14 13 12
SmiSts_0
31
September 2003
SmiSts_1
Rev. 3.06
SmiEn_1
26094
SmiSts_2
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Reset
R
R/W
0
R
R/W
0
Field Descriptions
IO Trap Enable (IoTrapEn)—Bit 15. Enable IO and configuration space trapping.
IO Trap Control SMI Special Cycle Enable (IoTrapCtlSmiSpcEn)—Bit 14. Enable SMI special
bus cycle generation when SMI handler is entered as a result of IO and configuration space
trapping.
IO Trap Control RSM Special Cycle Enable (IoTrapCtlRsmSpcEn)—Bit 13. Enable RSM
special bus cycle generation when SMI handler is entered as a result of IO and configuration
space trapping.
SMI Enable 3 (SmiEn_3)—Bit 7. Enable SMI generation if IO access matches access specified in
MSR C001_0053h.
SMI Status 3 (SmiSts_3)—Bit 6. Set if IO access matched access specified in MSR C001_0053h.
SMI Enable 2 (SmiEn_2)—Bit 5. Enable SMI generation if IO access matches access specified in
MSR C001_0052h.
SMI Status 2 (SmiSts_2)—Bit 4. Set if IO access matched access specified in MSR C001_0052h.
SMI Enable 1 (SmiEn_1)—Bit 3. Enable SMI generation if IO access matches access specified in
MSR C001_0051h.
304
Processor Conﬁguration Registers
Chapter 12
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
SMI Status 1 (SmiSts_1)—Bit 2. Set if IO access matched access specified in MSR C001_0051h.
SMI Enable 0 (SmiEn_0)—Bit 1. Enable SMI generation if IO access matches access specified in
MSR C001_0050h.
SMI Status 0 (SmiSts_0)—Bit 0. Set if IO access matched access specified in MSR C001_0050h.
Chapter 12
Processor Conﬁguration Registers
305
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
306
Processor Conﬁguration Registers
26094
Rev. 3.06
September 2003
Chapter 12
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Glossary
128-bit media instructions. Instructions that use the 128-bit XMM registers. These are a combination
of SSE and SSE2 instruction sets.
64-bit media instructions. Instructions that use the 64-bit MMX™ registers. These are primarily a
combination of MMX and 3DNow!™ instruction sets, with some additional instructions from the
SSE and SSE2 instruction sets.
16-bit mode. Legacy mode or compatibility mode in which a 16-bit address size is active. See legacy
mode and compatibility mode.
32-bit mode. Legacy mode or compatibility mode in which a 32-bit address size is active. See legacy
mode and compatibility mode.
64-bit mode. A submode of long mode. In 64-bit mode, the default address size is 64 bits and new features, such as register extensions, are supported for system and application software.
absolute. Said of a displacement that references the base of a code segment rather than an instruction
pointer. Contrast with relative.
AH–DH. The high 8-bit AH, BH, CH, and DH registers. Compare AL–DL.
AL–DL. The low 8-bit AL, BL, CL, and DL registers. Compare AH–DH.
AL–r15B. The low 8-bit AL, BL, CL, DL, SIL, DIL, BPL, SPL, and R8B–R15B registers, available in
64-bit mode.
biased exponent. The sum of a floating-point value’s exponent and a constant bias for a particular
floating-point data type. The bias makes the range of the biased exponent always positive, which
allows reciprocation without overﬂow.
byte. Eight bits.
clear. To write a bit-value of 0. Compare set.
compatibility mode. A submode of long mode. In compatibility mode, the default address size is 32
bits and legacy 16-bit and 32-bit applications run without modiﬁcation.
b. A bit, as in 1 Mb for one megabit, or lsb for least-signiﬁcant bit.
B. A byte, as in 1 MB for one megabyte, or LSB for least-signiﬁcant byte.
canonical address. AMD Athlon™ 64 and AMD Opteron™ processors implement 48-bit virtual
addresses. A virtual address having all 1s or all 0s in bit 63 through the most significant implemented bit (bit 47 for AMD Athlon™ 64 processor and AMD Opteron™ processor) is said to be in
canonical form. Accessing a virtual address that is not in canonical form results in a fault.
commit. To irreversibly write, in program order, an instruction’s result to software-visible storage,
such as a register (including ﬂags), the data cache, an internal write buffer, or memory.
CPL. Current privilege level.
Chapter
Glossary
307
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
CRn. Control register number n.
CS. Code segment register.
direct. Referencing a memory location whose address is included in the instruction’s syntax as an
immediate operand. The address may be an absolute or relative address. Compare indirect.
dirty data. Data held in the processor’s caches or internal buffers that is more recent than the copy
held in main memory.
displacement. A signed value that is added to the base of a segment (absolute addressing) or an
instruction pointer (relative addressing). Same as offset.
doubleword. Two words, or four bytes, or 32 bits.
double quadword. Eight words, or 16 bytes, or 128 bits. Also called octword.
DP. Two processors or dual processors.
eAX–eSP. The 16-bit AX, BX, CX, DX, DI, SI, BP, SP registers or the 32-bit EAX, EBX, ECX, EDX,
EDI, ESI, EBP, ESP registers. Compare rAX–rSP.
EFER. Extended features enable register.
effective address size. The address size for the current instruction after accounting for the default
address size and any address-size override preﬁx.
effective operand size. The operand size for the current instruction after accounting for the default
operand size and any operand-size override preﬁx.
eFLAGS. 16-bit or 32-bit ﬂags register. Compare rFLAGS.
EFLAGS. 32-bit (extended) ﬂags register.
eIP. 16-bit or 32-bit instruction-pointer register. Compare rIP.
EIP. 32-bit (extended) instruction-pointer register.
element. See vector.
exception. An abnormal condition that occurs as the result of executing an instruction. The processor’s response to an exception depends on the type of the exception. For all exceptions except 128bit media SIMD ﬂoating-point exceptions and x87 ﬂoating-point exceptions, control is transferred
to the handler (or service routine) for that exception, as defined by the exception’s vector. For
floating-point exceptions defined by the IEEE 754 standard, there are both masked and unmasked
responses. When unmasked, the exception handler is called, and when masked a default response
is provided instead of calling the handler.
FLAGS. 16-bit ﬂags register.
flush. When referring to caches, (1) to writeback, if modified, and invalidate, as in “flush the cache
line”; when referring to the processor pipeline, (2) to invalidate, as in “ﬂush the pipeline”; or (3) to
change a value, as in “ﬂush to zero”.
GDT. Global descriptor table.
308
Glossary
Chapter
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
GDTR. Global descriptor table register.
GPRs. General-purpose registers. For the 16-bit data size, these are AX, BX, CX, DX, DI, SI, BP, SP.
For the 32-bit data size, these are EAX, EBX, ECX, EDX, EDI, ESI, EBP, ESP. For the 64-bit data
size, these include RAX, RBX, RCX, RDX, RDI, RSI, RBP, RSP, and R8–R15.
IDT. Interrupt descriptor table.
IDTR. Interrupt descriptor table register.
IGN. Ignore. Field is ignored.
indirect. Referencing a memory location whose address is in a register or other memory location. The
address may be an absolute or relative address. Compare direct.
IP. 16-bit instruction-pointer register.
IRB. The virtual-8086 mode interrupt-redirection bitmap.
IST. The long-mode interrupt-stack table.
IVT. The real-address mode interrupt-vector table.
LDT. Local descriptor table.
LDTR. Local descriptor table register.
legacy x86. The legacy x86 architecture. See “Related Documents” on page 18 for descriptions of the
legacy x86 architecture.
legacy mode. An operating mode of the AMD64 architecture in which existing 16-bit and 32-bit
applications and operating systems run without modification. A processor implementation of the
AMD64 architecture can run in either long mode or legacy mode. Legacy mode has three submodes, real mode, protected mode, and virtual-8086 mode.
long mode. An operating mode unique to the AMD64 architecture. A processor implementation of the
AMD64 architecture can run in either long mode or legacy mode. Long mode has two submodes,
64-bit mode and compatibility mode.
lsb. least-signiﬁcant bit.
LSB. least-signiﬁcant byte.
main memory. Physical memory, such as RAM and ROM (but not cache memory) that is installed in
a particular computer system.
mask. (1) A control bit that prevents the occurrence of a floating-point exception from invoking an
exception handling routine. (2) A ﬁeld of bits used for a control purpose.
MBZ. Must be 0. If software attempts to set an MBZ bit to 1, a general-protection exception (#GP)
occurs.
memory. Unless otherwise speciﬁed, main memory.
ModRM. A byte following an instruction opcode that specifies address calculation based on mode
(mod), register (r), and memory (m) variables.
Chapter
Glossary
309
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
moffset. A direct memory offset. I.e., a displacement that is added to the base of a code segment (for
absolute addressing) or to an instruction pointer (for addressing relative to the instruction pointer,
as in RIP-relative addressing).
MP. More than two processors.
msb. Most-signiﬁcant bit.
MSB. Most-signiﬁcant byte.
MSR. Model-speciﬁc register.
multimedia instructions. A combination of 128-bit media instructions and 64-bit media instructions.
octword. Same as double quadword.
offset. Same as displacement.
overflow. The condition in which a floating-point number is larger in magnitude than the largest,
ﬁnite, positive or negative number that can be represented in the data-type format being used.
packed. See vector.
PAE. Physical-address extensions.
physical memory. Actual memory, consisting of main memory and cache.
probe. A check for an address in a processor’s caches or internal buffers. External probes originate
outside the processor, and internal probes originate within the processor.
processor. This term refers to AMD AthlonTM 64 processor architecture and AMD OpteronTM processor architecture. This document covers both classes of devices. For details about differences
between them, see the AMD AthlonTM 64 Processor Data Sheet, order# 24659 and the
AMD OpteronTM Processor Data Sheet, order# 23932.
protected mode. A submode of legacy mode.
quadword. Four words, or eight bytes, or 64 bits.
r8–r15. The 8-bit R8B–R15B registers, or the 16-bit R8W–R15W registers, or the 32-bit R8D–R15D
registers, or the 64-bit R8–R15 registers.
rAX–rSP. The 16-bit AX, BX, CX, DX, DI, SI, BP, SP registers, or the 32-bit EAX, EBX, ECX, EDX,
EDI, ESI, EBP, ESP registers, or the 64-bit RAX, RBX, RCX, RDX, RDI, RSI, RBP, RSP registers. The “r” variable should be replaced by nothing for 16-bit size, “E” for 32-bit size, or “R” for
64-bit size.
RAX. 64-bit version of EAX register.
RAZ. Read as zero (0), regardless of what is written.
RBP. 64-bit version of EBP register.
RBX. 64-bit version of EBX register.
RCX. 64-bit version of ECX register.
310
Glossary
Chapter
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
RDI. 64-bit version of EDI register.
RDX. 64-bit version of EDX register.
real-address mode. See real mode.
real mode. A short name for real-address mode, a submode of legacy mode.
relative. Referencing with a displacement (also called offset) from an instruction pointer rather than
the base of a code segment. Contrast with absolute.
REX. An instruction preﬁx that speciﬁes a 64-bit operand size and provides access to additional registers.
rFLAGS. 16-bit, 32-bit, or 64-bit ﬂags register. Compare RFLAGS.
RFLAGS. 64-bit ﬂags register. Compare rFLAGS.
rIP. 16-bit, 32-bit, or 64-bit instruction-pointer register. Compare RIP.
RIP-Relative Addressing. 4-bit instruction-pointer register. Addressing relative to the 64-bit RIP
instruction pointer. Compare moffset.
RSI. 64-bit version of ESI register.
RSP. 64-bit version of ESP register.
SBZ. Should be zero. If software attempts to set an SBZ bit to 1, it results in undeﬁned behavior.
set. To write a bit-value of 1. Compare clear.
SIB. A byte following an instruction opcode that specifies address calculation based on scale (S),
index (I), and base (B).
SIMD. Single instruction, multiple data. See vector.
SP. Stack pointer register.
SS. Stack segment register.
SSE. Streaming SIMD extensions instruction set. See 128-bit media instructions and 64-bit media
instructions.
SSE2. Extensions to the SSE instruction set. See 128-bit media instructions and 64-bit media instructions.
sticky bit. A bit that is set or cleared by hardware and that remains in that state until explicitly changed
by software.
TOP. x87 top-of-stack pointer.
TPR. Task-priority register (CR8), a new register introduced in the AMD64 architecture to speed
interrupt management.
TR. Task register.
TSS. Task state segment.
Chapter
Glossary
311
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
26094
Rev. 3.06
September 2003
underflow. The condition in which a floating-point number is smaller in magnitude than the smallest
non-zero, positive or negative number that can be represented in the data-type format being used.
UP. One processor or uniprocessor.
vector. (1) A set of integer or ﬂoating-point values, called elements, that are packed into a single operand. Most of the 128-bit and 64-bit media instructions use vectors as operands. Vectors are also
called packed or SIMD (single-instruction multiple-data) operands. (2) An index into an interrupt
descriptor table (IDT), used to for accessing exception handlers. Compare exception.
virtual-8086 mode. A submode of legacy mode.
word. Two bytes, or 16 bits.
x86. See legacy x86.
312
Glossary
Chapter
26094
Rev. 3.06
September 2003
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
Index of Register Names
B
CS Base 4 Register (Function 2: Offset 50h) 69
CS Base 5 Register (Function 2: Offset 54h) 69
CS Base 6 Register (Function 2: Offset 58h) 69
CS Base 7 Register (Function 2: Offset 5Ch) 69
CS Mask 0 Register (Function 2: Offset 60h) 72
CS Mask 1 Register (Function 2: Offset 64h) 72
CS Mask 2 Register (Function 2: Offset 68h) 72
CS Mask 3 Register (Function 2: Offset 6Ch) 72
CS Mask 4 Register (Function 2: Offset 70h) 72
CS Mask 5 Register (Function 2: Offset 74h) 72
CS Mask 6 Register (Function 2: Offset 78h) 72
CS Mask 7 Register (Function 2: Offset 7Ch) 72
Delay Line Register (Function 2: Offset 98h) 88
Limit 0 Register (Function 1: Offset 44h) 58
Limit 1 Register (Function 1: Offset 4Ch) 58
Limit 2 Register (Function 1: Offset 54h) 58
Limit 3 Register (Function 1: Offset 5Ch) 58
Limit 4 Register (Function 1: Offset 64h) 58
Limit 5 Register (Function 1: Offset 6Ch) 58
Limit 6 Register (Function 1: Offset 74h) 58
Limit 7 Register (Function 1: Offset 7Ch) 58
Scrub Address High Register (Function 3: Offset 60h 113
Scrub Address Low Register (Function 3: Offset 5Ch) 112
Timing High Register (Function 2: Offset 8Ch) 80
Timing Low Register (Function 2: Offset 88h) 78
BU Machine Check
Address Register (MSR 040Ah) 160
Control Register (MSR 0408h) 158
Status Register (MSR 0409h) 160
C
Capabilities Pointer Register (Function 0: Offset 34h) 31
Class Code/Revision ID Register (Function 0: Offset 08h) 30
Class Code/Revision ID Register (Function 1: Offset 08h) 55
Class Code/Revision ID Register (Function 2: Offset 08h) 68
Class Code/Revision ID Register (Function 3: Offset 08h) 91
Clear Interrupts Register (Function 3: Offset B4h) 125
Clock Power/Timing High Register (Function 3: Offset D8h) 127
Clock Power/Timing Low Register (Function 3: Offset D4h) 125
Config
Base and Limit 0 Register (Function 1: Offset E0h) 64
Base and Limit 1 Register (Function 1: Offset E4h) 64
Base and Limit 2 Register (Function 1: Offset E8h) 64
Base and Limit 3 Register (Function 1: Offset ECh) 64
Configuration Address Register 0CF8h 25
D
DC Machine Check
Address Register (MSR 0402h) 155
Control Register (MSR 0400h) 152
Status (MSR 0401h) 153
Device/Vendor ID register (Function 0: Offset 00h) 29
Device/Vendor ID Register (Function 1: Offset 00h) 55
Device/Vendor ID Register (Function 2: Offset 00h) 67
DeviceVendor ID Register (Function 3: Offset 00h) 90
Display Refresh Flow Control Buffers 119
DRAM
Bank Address Mapping Register (Function 2: Offset 80h) 73
Base 0 Register (Function 1: Offset 40h) 57
Base 1 Register (Function 1: Offset 48h) 57
Base 2 Register (Function 1: Offset 50h) 57
Base 3 Register (Function 1: Offset 58h) 57
Base 4 Register (Function 1: Offset 60h) 57
Base 5 Register (Function 1: Offset 68h) 57
Base 6 Register (Function 1: Offset 70h) 57
Base 7 Register (Function 1: Offset 78h) 57
Config High Register (Function 2: Offset 94h) 85
Config Low Register (Function 2: Offset 90h) 82
CS Base 0 Register (Function 2: Offset 40h) 69
CS Base 1 Register (Function 2: Offset 44h) 69
CS Base 2 Register (Function 2: Offset 48h) 69
CS Base 3 Register (Function 2: Offset 4Ch) 69
G
GART
Aperture Base Register (Function 3: Offset 94h) 123
Aperture Control Register (Function 3: Offset 90h) 122
Cache Control Register (Function 3: Offset 9Ch) 125
Table Base Register (Function 3: Offset 98h) 124
Global Machine Check Capabilities Register (MSR 0179h) 147
Global Machine Check Exception Reporting Control Register
(MSR 017Bh) 148
Global Machine Check Processor Status Register (MSR
017Ah) 147
H
Header
Type Register (Function 0: Offset 0Ch) 31
Type Register (Function 1: Offset 0Ch) 56
Type Register (Function 2: Offset 0Ch) 69
Type Register (Function 3: Offset 0Ch) 91
HyperTransport
Initialization Control Register (Function 0: Offset 6Ch) 40
Read Pointer Optimization Register (Function 3: Offset
DCh) 128
Transaction Control Register (Function 0: Offset 68h) 36
Index of Register Names
313
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
I
IC Machine Check
Address Register (MSR 0406h) 157
Control Register (MSR 0404h) 156
Status Register (MSR 0405h) 157
L
LDT0
Buffer Count Register (Function 0: Offset 90h) 49
Bus Number Register (Function 0: Offset 94h) 51
Capabilities Register (Function 0: Offset 80h) 42
Feature Capability Register (Function 0: Offset 8Ch) 48
Frequency/Revision Register (Function 0: Offset 88h) 47
Link Control Register (Function 0: Offset 84h) 43
Type Register (Function 0: Offset 98h) 51
LDT1
Buffer Count Register (Function 0: Offset B0h) 49
Bus Number Register (Function 0: Offset B4h) 51
Capabilities Register (Function 0: Offset A0h) 42
Feature Capability Register (Function 0: Offset ACh) 48
Frequency/Revision Register (Function 0: Offset A8h) 47
Link Control Register (Function 0: Offset A4h) 43
Type Register (Function 0: Offset B8h) 51
LDT2
Buffer Count Register (Function 0: Offset D0h) 49
Bus Number Register (Function 0: Offset D4h) 51
Capabilities Register (Function 0: Offset C0h) 42
Feature Capability Register (Function 0: Offset CCh) 48
Frequency/Revision Register (Function 0: Offset C8h) 47
Link Control Register (Function 0: Offset C4h) 43
Type Register (Function 0: Offset D8h) 51
LS Machine Check
Address Register (MSR 040Eh) 162
Control Register (MSR 040Ch) 161
Status Register (MSR 040Dh) 162
M
MC0_CTL Register (MSR 0400h) 152
MC0_STATUS Register (MSR 0401h) 153
MC1_CTL Register (MSR 0404h) 156
MC2_CTL Register (MSR 0408h) 158
MC3_CTL Register (MSR 040Ch) 161
MCA
NB Address High Register (Function 3: Offset 54h) 104
NB Address Low Register (Function 3: Offset 50h) 103
NB Configuration Register (Function 3: Offset 44h) 95
NB Control Register (Function 3: Offset 40h) 92
NB Status High Register (Function 3: Offset 4Ch) 101
NB Status Low Register (Function 3: Offset 48h) 98
MCG_CAP Register (MSR 0179h) 147
MCG_CTL Register (MSR 017Bh) 148
MCG_STATUS Register (MSR 017Ah) 147
MCi_ADDR Register (MSRs 0402h, 0406h, 040Ah, 040Eh,
314
26094
Rev. 3.06
September 2003
0412h) 151
MCi_CTL Registers (MSRs 0400h, 0404h, 0408h, 040Ch,
0410h) 149
MCi_STATUS Registers (MSRs 0401h, 0405h, 0409h, 040Dh,
0411h) 150
Memory-Mapped I/O
Base 0 Register (Function 1: Offset 80h) 60
Base 1 Register (Function 1: Offset 88h) 60
Base 2 Register (Function 1: Offset 90h) 60
Base 3 Register (Function 1: Offset 98h) 60
Base 4 Register (Function 1: Offset A0h) 60
Base 5 Register (Function 1: Offset A8h) 60
Base 6 Register (Function 1: Offset B0h) 60
Base 7 Register (Function 1: Offset B8h) 60
Limit 0 Register (Function 1: Offset 84h) 61
Limit 1 Register (Function 1: Offset 8Ch) 61
Limit 2 Register (Function 1: Offset 94h) 61
Limit 3 Register (Function 1: Offset 9Ch) 61
Limit 4 Register (Function 1: Offset A4h) 61
Limit 5 Register (Function 1: Offset ACh) 61
Limit 6 Register (Function 1: Offset B4h) 61
Limit 7 Register (Function 1: Offset BCh) 61
N
NB Machine Check
Address Register (MSR 0412h) 162
Control Register (MSR 0410h) 162
Status Register (MSR 0411h) 162
NB_CFG Register (MSR C001_001Fh) 290
Node ID Register (Function 0: Offset 60h) 34
Northbridge
Capabilities Register (Function 3: Offset E8h) 131
P
PCI I/O
Base 0 Register (Function 1: Offset C0h) 62
Base 1 Register (Function 1: Offset C8h) 62
Base 2 Register (Function 1: Offset D0h) 62
Base 3 Register (Function 1: Offset D8h) 62
Limit 0 Register (Function 1: Offset C4h) 63
Limit 1 Register (Function 1: Offset CCh) 63
Limit 2 Register (Function 1: Offset D4h) 63
Limit 3 Register (Function 1: Offset DCh) 63
Performance
Monitor 0 Register (Function 3: Offset A0h) 125
Monitor 1 Register (Function 3: Offset A4h) 125
Monitor 2 Register (Function 3: Offset A8h) 125
Monitor 3 Register (Function 3: Offset ACh) 125
Power Management
Control High Register (Function 3: Offset 84h) 121
Control Low Register (Function 3: Offset 80h) 121
Index of Register Names
26094
Rev. 3.06
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
September 2003
R
Registers
General MSRs 261
Model Specific Registers 285
Power Management 242
Routing Table
Node 0 Register (Function 0: Offset 40h)
Node 1 Register (Function 0: Offset 44h)
Node 2 Register (Function 0: Offset 48h)
Node 3 Register (Function 0: Offset 4Ch)
Node 4 Register (Function 0: Offset 50h)
Node 5 Register (Function 0: Offset 54h)
Node 6 Register (Function 0: Offset 58h)
Node 7 Register (Function 0: Offset 5Ch)
32–33
33
33
33
33
33
33
33
S
Scrub Control Register (Function 3: Offset 58h) 108, 110
SMM
Save State (Offset FE00–FFFFh) 169
SRI-to-XBAR Buffer Count Register (Function 3: Offset
70h) 113
Status/Command Register (Function 0: Offset 04h) 30
T
Thermtrip Status Register (Function 3: Offset E4h) 129
U
Unit ID Register (Function 0: Offset 64h) 35
X
XBAR-to-SRI Buffer Count Register (Function 3: Offset
74h) 118
Index of Register Names
315
BIOS and Kernel Developer’s Guide for the AMD Athlon™ 64 and
AMD Opteron™ Processors
316
Index of Register Names
26094
Rev. 3.06
September 2003

Open as PDF

Similar pages: BIOS and Kernel Developer’s Guide (BKDG); AMD AMD-761; INTEL 82443GX; ETC RGE7500PL; INTEL X38; INTEL 82G35; INTEL 326769-002; AMD Opteron™ A1100 Series Processor ACPI Porting Guide; ETC L64364NL; ZMD ZMD31050_09; ETC 25759; NSC QR0001; AMD ADA3200DAK4BW; AMD ADAFX51CEP5AK; A1PROS AI7152S; ETC AMDOPTERON; ETC AMD-8111; AGERE W3030; IDT 79RC32364; AD AD9389KSTZ-80; MigratingExchangeToND8.pdf