WO2014120113A1 - Assigning processors to memory mapped configuration - Google Patents

Assigning processors to memory mapped configuration Download PDF

Info

Publication number
WO2014120113A1
WO2014120113A1 PCT/US2013/023552 US2013023552W WO2014120113A1 WO 2014120113 A1 WO2014120113 A1 WO 2014120113A1 US 2013023552 W US2013023552 W US 2013023552W WO 2014120113 A1 WO2014120113 A1 WO 2014120113A1
Authority
WO
WIPO (PCT)
Prior art keywords
processors
pci
service unit
memory mapped
mapped configuration
Prior art date
Application number
PCT/US2013/023552
Other languages
French (fr)
Inventor
Anys Bacha
Original Assignee
Hewlett-Packard Development Company
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hewlett-Packard Development Company filed Critical Hewlett-Packard Development Company
Priority to EP13873433.0A priority Critical patent/EP2951705A4/en
Priority to US14/763,501 priority patent/US9940291B2/en
Priority to PCT/US2013/023552 priority patent/WO2014120113A1/en
Priority to CN201380071515.4A priority patent/CN105264506B/en
Priority to TW102145163A priority patent/TWI512479B/en
Publication of WO2014120113A1 publication Critical patent/WO2014120113A1/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38Information transfer, e.g. on bus
    • G06F13/42Bus transfer protocol, e.g. handshake; Synchronisation
    • G06F13/4204Bus transfer protocol, e.g. handshake; Synchronisation on a parallel bus
    • G06F13/4221Bus transfer protocol, e.g. handshake; Synchronisation on a parallel bus being an input/output bus, e.g. ISA bus, EISA bus, PCI bus, SCSI bus
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26Power supply means, e.g. regulation thereof
    • G06F1/32Means for saving power
    • G06F1/3203Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3234Power saving characterised by the action undertaken
    • G06F1/3296Power saving characterised by the action undertaken by lowering the supply or operating voltage
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2213/00Indexing scheme relating to interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F2213/0024Peripheral component interconnect [PCI]

Definitions

  • I/O Input/Output
  • FIGS. 1 A and 1 B are example block diagrams of a device for assigning processors to a memory mapped configuration
  • FIG. 2 is another example block diagram of a device for assigning processors to a memory mapped configuration
  • FIG. 3 is an example block diagram of a computing device including instructions for assigning processors to a memory mapped configuration
  • FIG. 4 is an example flowchart of a method for assigning processors to a memory mapped configuration.
  • Embodiments may allow for supporting a single PCI segment during early boot and then using a processor to properly enable additional PCI segments prior to handing off system resources to an OS.
  • a master processor may quiesce one or more slave processors running on a device.
  • Each of the processors may have access to different buses of a Peripheral Component Interconnect (PCI) segment using a 32-bit based memory mapped configuration.
  • a service unit may switch to a 64-bit memory mapped configuration after the master processor indicates the one or more slaves processors are quiesced.
  • Each of the processors may be assigned to one of a plurality of instances of the PCI segment after the service unit switches to the 64-bit memory mapped configuration. Each of these instances may have access to all of the buses of the PCI segment.
  • embodiments may provide greater PCI bus numbers while still allowing for system firmware code that is dependent on a single PCI segment configuration to run unmodified. As a result, embodiments may be simple to implement because a complexity associated with re-writing 32-bit code to make it 64-bit compliant, is reduced or eliminated.
  • FIGS. 1A and 1 B are example block diagrams of a device 100 for assigning processors 120, 122-1 and 122-2 to a memory mapped configuration.
  • the device 100 may be, for example, a server, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a network device, a controller, a wireless device, or any other type of device that is to interface with a Peripheral Component Interconnect (PCI).
  • PCI Peripheral Component Interconnect
  • PCI may refer to part of the PCI Local Bus standard and/or a local computer bus for attaching hardware devices in a computer.
  • the PCI bus may support functions found on a processor bus, but in a standardized format that is independent of any particular processor.
  • Devices connected to the bus may appear to the processor to be connected directly to the processor bus, and may be assigned addresses in the processor's address space.
  • the devices may be fitted onto a motherboard itself (e.g. a planar device) or an expansion card that fits into a slot.
  • Example PCI cards may include network cards, sound cards, modems, extra ports such as USB or serial, TV tuner cards, disk controllers and the like.
  • PCI supports both 32-bit and 64-bit addresses for memory space.
  • BIOS or System firmware assigns regions of memory space in the PCI address domain to PCI peripheral devices.
  • a base address of a region is stored in the base address register of the device's PCI configuration space.
  • the size of each region is generally a power of two, and the assigned base address is aligned on a boundary equal to the size of the region.
  • PCI peripheral devices in memory space are memory-mapped into a host address domain, such as a processor's address space, so that data access to any device can be performed by the processor's native load or store instructions.
  • Each PCI peripheral device is identified by a bus number, a device number, and a function number. The PCI specification permits a single system or segment to host up to 256 buses, but 256 buses may not sufficient for many large systems, as explained above.
  • the device 100 includes a service unit 1 10 and a plurality of processors 120, 122-1 and 122-2.
  • the service unit 1 10 may include, for example, a hardware device including electronic circuitry for implementing the functionality described below, such as control logic and/or memory.
  • the service unit 1 10 may be implemented as a series of instructions encoded on a machine-readable storage medium and executable by a processor.
  • the processors 120, 122-1 and 122-2 may be, a central processing unit (CPU), a semiconductor-based microprocessor, a graphics processing unit (GPU), or other hardware devices suitable for retrieval and execution of instructions.
  • the service unit 1 10 may be a separate computer system within the device that allows some service tasks, such as system restart or PCI configuration.
  • the service unit 1 10 may be available as long as the service unit 1 10 is supplied with power, even when a main operating system (OS) of the device 100 is shut down, defective or not installed at all.
  • FIG. 1 A shows a state of the device 100 after the device 100 has started to boot from system firmware (not shown), such as a Basic Input/Output System (BIOS).
  • system firmware may consist of 32-bit instructions
  • the device 100 may be in a 32-bit mode.
  • each of the processors 120, 122-1 and 122-2 are to initially have access to different buses 132 of a Peripheral Component Interconnect (PCI) segment 130 using a 32-bit based memory mapped configuration (not shown).
  • PCI Peripheral Component Interconnect
  • bus may define a low level interface between a CPU and a peripheral device.
  • the master processor 120 may have access to busses 132 numbered 0-84 of the PCI segment 130
  • the first slave processor 122-1 may have access to busses 132 numbered 85-169 of the PCI segment 130
  • the second slave processor 122- 1 may have access to busses 132 numbered 170-255 of the PCI segment 130.
  • each of the processors 120, 122-1 and 122-2 may initially include memory mapped configurations that limit their access to only part of the PCI segment 130, where no two processors 120, 122-1 and 122-2 have access to the same buse 132.
  • the service unit 1 10 may then select one of the processors 120 to be a master processor.
  • the master processor 120 may next quiesce one or more slave processors 122-1 and 122-2 running on the device 100. While FIG. 2 shows two slave processors 122-1 and 122-2, embodiments may include more or less than two slave processors.
  • the term quiesce may refer pausing or altering the state of running processes on a computer or processor, such as those that might modify information related to memory accesses, such as a write request.
  • the quiesced processors 120, 122-1 and 122-2 may not transmit traffic over a fabric external to the processors 120, 122-1 and 122-2 and may not carry out operations via the memory mapped configuration.
  • the service unit 1 10 may switch the device 100 to a 64-bit memory mapped configuration via a sideband mechanism, such as a Platform Environment Control Interface (PECI) and/or a System Management Bus (SMBus).
  • PECI Platform Environment Control Interface
  • SMBs System Management Bus
  • the service unit 1 10 configures one or more resources of the device 100 to switch the PC segment 130 to the 64-bit memory mapped configuration. For example, the service unit 1 10 may reset a base number of a PCI domain of each of the processors 120 and 122-1 and 122-2 to zero, where the base number is stored in the memory mapped configuration of each of the processors 120 and 122-1 and 122-2.
  • the service unit 1 10 may configure each root-complex (not shown) to decode according to the 64-bit memory mapped configuration.
  • the term root complex may refer to a device that connects a processor and memory subsystem to a PCI switch fabric composed of one or more switch devices (not shown). Switching the device 100 to a 64-bit memory mapped configuration is explained below with respect to FIG. 2.
  • the service unit 1 10 may pre-enable (but not allow use of) the 64-bit memory mapped configuration, such as by programming addresses and setting registers (not shown) for the PCI. For example, the service unit 100 may set the registers to specify a range of the busses available to each of the processors 120, 122-1 and 122-2.
  • each of the processors 120, 122-1 and 122-2 is assigned one of a plurality of instances 130-0 to 130-2 of the PCI segment 130.
  • Each of the instances 130-0 to 130-2 is to have access to all of the buses 132.
  • the service unit 1 10 may be able create an additional, multi-bit instance field (not shown) within the memory-mapped configuration of each of the processors 120, 122-1 and 122-2. Further, as explained above, the service unit 1 10 may configure the root-complex 208 to decode according to the 64-bit memory mapped configuration.
  • a value "00" within the instance field of the master processor 120 may indicate that it is assigned to a first instance 130-0 of the PCI segment. While a value "01 " within the instance field of the first slave processor 122-1 may indicate that it is to a second instance 130-1 of the PCI segment. Similarly, a value "10" within the instance field of the second slave processor 122-2 may indicate that it is assigned to a third instance 130-0 of the PCI segment.
  • each of the processors 120, 122-1 and 122-2 may have access to a maximum number or greater of busses 132, as allowed per PCI segment 130 by the PCI specification, such as 256 busses.
  • the service unit 1 10 is to switch to the 64-bit memory mapped configuration not only after the device 100 has started to boot from a Basic Input/Output System (BIOS) but also prior to handing off system resources to an operating system (OS) of the device 100, such as a device driver.
  • BIOS Basic Input/Output System
  • OS operating system
  • embodiments may allow a plurality of processors 120, 122-2 and 122-2 on a single device 100 to have access to a greater number of PCI buses 132 without modifying system firmware, such as to another format like 64-bit mode.
  • FIG. 2 is another example block diagram of a device 200 for assigning processors to a memory mapped configuration.
  • the device 200 may be, for example, a server, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a network device, a controller, a wireless device, or any other type of device that is to interface with a Peripheral Component Interconnect (PCI).
  • PCI Peripheral Component Interconnect
  • the device 200 of FIG. 2 may include at least the functionality and/or hardware of the device 100 of FIG. 1 .
  • the device 200 of FIG. 2 may respectively include the functionality of the service unit 1 10, master processor 120 and slave processors 122-1 and 122-2 included in the device 100 of FIG. 1 . Further, the device 200 of FIG. 2 also includes a root complex 208, BIOS 210, an ACPI table/namespace resource 212 and PCI 214.
  • the BIOS 210 and ACPI table/namespace resource 212 may include, for example, a hardware device including electronic circuitry for implementing the functionality described below, such as control logic and/or memory.
  • the BIOS 210 and ACPI table/namespace resource 212 may be implemented as a series of instructions or microcode encoded on a machine- readable storage medium and executable by a processor.
  • the PCI 214 may provide a hardware and/or firmware interface for connecting one or more PCI peripheral devices (not shown) to the one or more the processors 120, 122-1 and 122-2 of the device 200.
  • the master processor 120 is shown to write to a shared memory 204 of the service unit 1 10 to indicate that the one or more slave processors 122-1 and 122-2 are quiesced.
  • the service unit 1 10 is to receive control of the device 200 from the master processor 120 after the one or more slave processor are quiesced 122-1 and 122-2
  • the service unit 1 10 is to return control of the device 200 to the processors 120, 122-1 and 122-2 after the service unit 1 10 switches to the 64- bit memory mapped configuration.
  • the service unit 1 10 may enable a flag at a scratch register 204-1 to 204-3 of one or more of the processors 120, 122-1 and 122-2 after the service unit 1 10 switches to the 64- bit memory mapped configuration.
  • the BIOS 210 may instruct the quiesced processors 120, 122-1 and 122-2 to poll their scratch registers 204-1 to 204-3 to determine if the flag is enabled. If the flag is enabled, the quiesced processors 120, 122-1 and 122-2 may transition from a quiesced state back to an active state.
  • the service unit 1 10 is to transfer control of the device to the processors 120, 122-1 and 122-2, if the flag is enabled.
  • each of the processors 120, 122-1 and 122-2 is assigned to have a different instance of a PCI segment, with each instance having access to all PCI buses.
  • a memory mapped configuration (MMCFG) 202-1 of the master processor 120 may have access to a first instance (PCI segment 0)
  • a MMCFG 202-2 of the first slave processor 122-1 may have access to a second instance (PCI segment 1 )
  • a MMCFG 202-3 of the second slave processor 122-2 may have access to a third instance (PCI segment 2), with each instance (PCI segments 0-2) having 256 buses (0- 255).
  • the master processor 120 may walk the PCI busses to assign resources to the PCI segments (PCI segments 0-2).
  • the master processor 120 may also construct Advanced Configuration and Power Interface (ACPI) static tables and namespace resources 212 to expose information related to the segments to an OS (not shown) of the device 200.
  • ACPI Advanced Configuration and Power Interface
  • the master processor 120 via the OS or firmware, may query all PCI buses at startup time (via PCI Configuration Space) to find out what devices are present and what system resources (memory space, I/O space, interrupt lines, etc.) each needs. Then the master processor 120 may allocate the resources and tell each device its allocation. While embodiments refer to 32 and 64 bit modes, embodiments are not limited thereto and may include other bit modes as well.
  • FIG. 3 is an example block diagram of a computing device 300 including instructions for assigning processors to a memory mapped configuration.
  • the computing device 300 includes a processor 310 and a machine-readable storage medium 320.
  • the machine-readable storage medium 320 further includes instructions 321 , 323, 325, 327 and 329 for assigning processors (not shown) to a memory mapped configuration.
  • the computing device 300 may be, for example, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a server, a network device, a controller, a wireless device, or any other type of device capable of executing the instructions 321 , 323, 325, 327 and 329.
  • the computing device 300 may include or be connected to additional components such as memories, controllers, processors etc.
  • the processor 310 may be, at least one central processing unit (CPU), at least one semiconductor-based microprocessor, at least one graphics processing unit (GPU), a microcontroller, special purpose logic hardware controlled by microcode or other hardware devices suitable for retrieval and execution of instructions stored in the machine-readable storage medium 320, or combinations thereof.
  • the processor 310 may fetch, decode, and execute instructions 321 , 323, 325, 327 and 329 to implement assigning the processors to the memory mapped configuration.
  • the processor 310 may include at least one integrated circuit (IC), other control logic, other electronic circuits, or combinations thereof that include a number of electronic components for performing the functionality of instructions 321 , 323, 325, 327 and 329.
  • IC integrated circuit
  • the machine-readable storage medium 320 may be any electronic, magnetic, optical, or other physical storage device that contains or stores executable instructions.
  • the machine-readable storage medium 320 may be, for example, Random Access Memory (RAM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a storage drive, a Compact Disc Read Only Memory (CD-ROM), and the like.
  • RAM Random Access Memory
  • EEPROM Electrically Erasable Programmable Read-Only Memory
  • CD-ROM Compact Disc Read Only Memory
  • the machine- readable storage medium 320 can be non-transitory.
  • machine-readable storage medium 320 may be encoded with a series of executable instructions for assigning the processors to the memory mapped configuration.
  • the instructions 321 , 323, 325, 327 and 329 when executed by a processor can cause the processor to perform processes, such as, the process of FIG. 4.
  • the initialize instructions 321 may be executed by the processor 310 to initialize a plurality of processors to access different buses of a Peripheral Component Interconnect (PCI) segment, each of the different buses assigned to only one of the plurality of processors.
  • the quiesce instructions 323 may be executed by the processor 310 to quiesce the plurality of processors.
  • the transfer instructions 325 may be executed by the processor 310 to transfer control of the device from at least one of the quiesced processors to a service unit (not shown).
  • the assign instructions 327 may be executed by the processor 310 to assign, via the service unit, each of the plurality of processors a memory mapped configuration that accesses all of the buses, with each of the processors assigned to different instances of the PCI segment, before an operating system (OS) of the device is loaded.
  • the return instructions 329 may be executed by the processor 310 to return control of the device from the service unit to at least one of the quiesced processors.
  • FIG. 4 is an example flowchart of a method 400 for assigning processors to a memory mapped configuration.
  • execution of the method 400 is described below with reference to the device 200, other suitable components for execution of the method 400 can be utilized, such as the device 100. Additionally, the components for executing the method 400 may be spread among multiple devices (e.g., a processing device in communication with input and output devices). In certain scenarios, multiple devices acting in coordination can be considered a single device to perform the method 400.
  • the method 400 may be implemented in the form of executable instructions stored on a machine- readable storage medium, such as storage medium 320, and/or in the form of electronic circuitry.
  • a plurality of processors 120, 122-1 and 122-2 receive 32-bit instructions from a BIOS 210 after a device starts a boot sequence. Then, at block 420, one of the plurality of processors 120 is selected to quiesce a remainder of the plurality of processors 122-1 and 122-2. Next, at block 430, the plurality of processors 120, 122-1 and 122-2 are assigned, such as by the service unit 1 10 to interface with a PCI 214 according to a 64-bit memory mapped configuration 202-1 to 202-3.
  • Each of the plurality of processors 120, 122-1 and 122-2 are to access separate buses 132 of a single PCI segment 130, before the assigning at block 430.
  • Each of the plurality of processors 120, 122-1 and 122-2 are to access all of the buses 132 of different instances 130-0 to 130-2 of the PCI segment 130 bus, after the assigning at block 430.
  • the plurality of quiesced processors 120, 122-1 and 122-2 are activated, before hardware interfacing via the PCI is initialized.
  • embodiments provide a method and/or device for postponing enablement of multiple PCI segments until hardware initialization is complete.
  • embodiments may provide greater PCI bus numbers while still allowing for system firmware code that is dependent on a Single PCI Segment configuration to run unmodified.
  • embodiments may be simple to implement because a complexity associated with re-writing 32-bit code to make it 64-bit compliant, is reduced or eliminated.
  • embodiments may allow for supporting a single PCI segment during early boot and then using a processor, such as a Service Processor, to properly enable additional PCI segments prior to handing off system resources to an OS.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Multi Processors (AREA)
  • Stored Programmes (AREA)

Abstract

Embodiments herein relate to assigning processors to a memory mapped configuration. The processors having access to different buses of a Peripheral Component Interconnect (PCI) segment are quiesced. The quiesced processors are assigned a memory mapped configuration.

Description

ASSIGNING PROCESSORS TO MEMORY MAPPED CONFIGURATION
BACKGROUND
[0001 ] Due to recent innovations in technology, users are seeking larger amounts of Input/Output (I/O) resources. However, due to I/O standards and/or bandwidth limitations, devices are unable to support a greater number of I/O devices to provide the I/O resources. System vendors are challenged to provide users with a greater amount of I/O resources.
BRIEF DESCRIPTION OF THE DRAWINGS
[0002] The following detailed description references the drawings, wherein:
[0003] FIGS. 1 A and 1 B are example block diagrams of a device for assigning processors to a memory mapped configuration;
[0004] FIG. 2 is another example block diagram of a device for assigning processors to a memory mapped configuration;
[0005] FIG. 3 is an example block diagram of a computing device including instructions for assigning processors to a memory mapped configuration; and
[0006] FIG. 4 is an example flowchart of a method for assigning processors to a memory mapped configuration.
DETAILED DESCRIPTION
[0007] Specific details are given in the following description to provide a thorough understanding of embodiments. However, it will be understood that embodiments may be practiced without these specific details. For example, systems may be shown in block diagrams in order not to obscure embodiments in unnecessary detail. In other instances, well-known processes, structures and techniques may be shown without unnecessary detail in order to avoid obscuring embodiments.
[0008] Due to the dynamics of the information era and its demanding nature, the bar for IT companies to produce equipment that can accommodate a large amount of I/O resources is constantly being raised. For example, having an ability to support a large amount of I/O resources is becoming highly valued for virtualization, as more guest operating systems (OS) try to claim ownership of such assets. Therefore, with a number of I/O cards available on device, such as high-end systems, increasing and more virtual functions being added to PCI Express (PCIe) devices, PCI bus numbers are becoming increasingly scarce within a single PCI segment.
[0009] By supporting multiple PCI segments, vendors may be able to provide more bus numbers. However, supporting multiple PCI segments generally requires use of memory mapped I/O ranges that are above 4GB, known as Memory Mapped Configuration (MMCFG). This poses a problem for early firmware code that needs to access resources in the MMCFG space while still in 32-bit mode. Also most vendors provide hardware initialization firmware packages that are configured for 32-bit support. Thus, supporting multiple PCI segments during early boot creates several challenges. Such challenges may include a need to modify vendor provided packages to operate in 64-bit mode, which in turn can lead to supporting more complex code that results in delaying product releases. [0010] Embodiments may allow for supporting a single PCI segment during early boot and then using a processor to properly enable additional PCI segments prior to handing off system resources to an OS. For example, a master processor may quiesce one or more slave processors running on a device. Each of the processors may have access to different buses of a Peripheral Component Interconnect (PCI) segment using a 32-bit based memory mapped configuration. A service unit may switch to a 64-bit memory mapped configuration after the master processor indicates the one or more slaves processors are quiesced. Each of the processors may be assigned to one of a plurality of instances of the PCI segment after the service unit switches to the 64-bit memory mapped configuration. Each of these instances may have access to all of the buses of the PCI segment.
[001 1 ] Thus, by postponing enablement of multiple instances of PCI segments until hardware initialization is complete, embodiments may provide greater PCI bus numbers while still allowing for system firmware code that is dependent on a single PCI segment configuration to run unmodified. As a result, embodiments may be simple to implement because a complexity associated with re-writing 32-bit code to make it 64-bit compliant, is reduced or eliminated.
[0012] This further may translate in to savings in development schedules and on time product releases. In addition, enabling multiple PCI segments according to embodiments allows for an easier exchange of firmware packages between hardware vendors. For example, vendors may be able construct their packages to work in 32-bit mode only, as opposed to handling multiple modes (e.g., 32-bit and 64-bit modes). Having packages support a single mode may reduce development and validation cycles needed to support multiple modes. Also, embodiments may use industry standard interfaces such as SMBus, thus allowing for ease of implementation.
[0013] Referring now to the drawings, FIGS. 1A and 1 B are example block diagrams of a device 100 for assigning processors 120, 122-1 and 122-2 to a memory mapped configuration. The device 100 may be, for example, a server, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a network device, a controller, a wireless device, or any other type of device that is to interface with a Peripheral Component Interconnect (PCI).
[0014] The term PCI may refer to part of the PCI Local Bus standard and/or a local computer bus for attaching hardware devices in a computer. The PCI bus may support functions found on a processor bus, but in a standardized format that is independent of any particular processor. Devices connected to the bus may appear to the processor to be connected directly to the processor bus, and may be assigned addresses in the processor's address space. The devices may be fitted onto a motherboard itself (e.g. a planar device) or an expansion card that fits into a slot. Example PCI cards may include network cards, sound cards, modems, extra ports such as USB or serial, TV tuner cards, disk controllers and the like.
[0015] PCI supports both 32-bit and 64-bit addresses for memory space. As the device 100 boots, BIOS or System firmware (not shown) assigns regions of memory space in the PCI address domain to PCI peripheral devices. A base address of a region is stored in the base address register of the device's PCI configuration space. The size of each region is generally a power of two, and the assigned base address is aligned on a boundary equal to the size of the region. PCI peripheral devices in memory space are memory-mapped into a host address domain, such as a processor's address space, so that data access to any device can be performed by the processor's native load or store instructions. Each PCI peripheral device is identified by a bus number, a device number, and a function number. The PCI specification permits a single system or segment to host up to 256 buses, but 256 buses may not sufficient for many large systems, as explained above.
[0016] In the embodiment of FIGS. 1A and 1 B, the device 100 includes a service unit 1 10 and a plurality of processors 120, 122-1 and 122-2. The service unit 1 10 may include, for example, a hardware device including electronic circuitry for implementing the functionality described below, such as control logic and/or memory. In addition or as an alternative, the service unit 1 10 may be implemented as a series of instructions encoded on a machine-readable storage medium and executable by a processor. The processors 120, 122-1 and 122-2 may be, a central processing unit (CPU), a semiconductor-based microprocessor, a graphics processing unit (GPU), or other hardware devices suitable for retrieval and execution of instructions.
[0017] For example, the service unit 1 10 may be a separate computer system within the device that allows some service tasks, such as system restart or PCI configuration. The service unit 1 10 may be available as long as the service unit 1 10 is supplied with power, even when a main operating system (OS) of the device 100 is shut down, defective or not installed at all. [0018] FIG. 1 A shows a state of the device 100 after the device 100 has started to boot from system firmware (not shown), such as a Basic Input/Output System (BIOS). As the system firmware may consist of 32-bit instructions, the device 100 may be in a 32-bit mode. Thus, each of the processors 120, 122-1 and 122-2 are to initially have access to different buses 132 of a Peripheral Component Interconnect (PCI) segment 130 using a 32-bit based memory mapped configuration (not shown).
[0019] As explained above, the PCI specification permits a single system or segment to host up to only 256 buses. The term bus may define a low level interface between a CPU and a peripheral device. For example, the master processor 120 may have access to busses 132 numbered 0-84 of the PCI segment 130, the first slave processor 122-1 may have access to busses 132 numbered 85-169 of the PCI segment 130 and the second slave processor 122- 1 may have access to busses 132 numbered 170-255 of the PCI segment 130. Thus, as shown in FIG. 1A, each of the processors 120, 122-1 and 122-2 may initially include memory mapped configurations that limit their access to only part of the PCI segment 130, where no two processors 120, 122-1 and 122-2 have access to the same buse 132.
[0020] The service unit 1 10 may then select one of the processors 120 to be a master processor. The master processor 120 may next quiesce one or more slave processors 122-1 and 122-2 running on the device 100. While FIG. 2 shows two slave processors 122-1 and 122-2, embodiments may include more or less than two slave processors. The term quiesce may refer pausing or altering the state of running processes on a computer or processor, such as those that might modify information related to memory accesses, such as a write request.
[0021 ] The quiesced processors 120, 122-1 and 122-2 may not transmit traffic over a fabric external to the processors 120, 122-1 and 122-2 and may not carry out operations via the memory mapped configuration. After the master processor 120 indicates the one or more slaves processors 122-1 and 122-2 are quiesced, the service unit 1 10 may switch the device 100 to a 64-bit memory mapped configuration via a sideband mechanism, such as a Platform Environment Control Interface (PECI) and/or a System Management Bus (SMBus).
[0022] The service unit 1 10 configures one or more resources of the device 100 to switch the PC segment 130 to the 64-bit memory mapped configuration. For example, the service unit 1 10 may reset a base number of a PCI domain of each of the processors 120 and 122-1 and 122-2 to zero, where the base number is stored in the memory mapped configuration of each of the processors 120 and 122-1 and 122-2.
[0023] Further, the service unit 1 10 may configure each root-complex (not shown) to decode according to the 64-bit memory mapped configuration. The term root complex may refer to a device that connects a processor and memory subsystem to a PCI switch fabric composed of one or more switch devices (not shown). Switching the device 100 to a 64-bit memory mapped configuration is explained below with respect to FIG. 2. In addition, the service unit 1 10 may pre-enable (but not allow use of) the 64-bit memory mapped configuration, such as by programming addresses and setting registers (not shown) for the PCI. For example, the service unit 100 may set the registers to specify a range of the busses available to each of the processors 120, 122-1 and 122-2.
[0024] As shown in FIG. 1 B, after the service unit 100 switches the device 100 to the 64-bit memory mapped configuration, each of the processors 120, 122-1 and 122-2 is assigned one of a plurality of instances 130-0 to 130-2 of the PCI segment 130. Each of the instances 130-0 to 130-2 is to have access to all of the buses 132. For example, by switching to the 64-bit memory mapped configuration, the service unit 1 10 may be able create an additional, multi-bit instance field (not shown) within the memory-mapped configuration of each of the processors 120, 122-1 and 122-2. Further, as explained above, the service unit 1 10 may configure the root-complex 208 to decode according to the 64-bit memory mapped configuration.
[0025] For example, a value "00" within the instance field of the master processor 120 may indicate that it is assigned to a first instance 130-0 of the PCI segment. While a value "01 " within the instance field of the first slave processor 122-1 may indicate that it is to a second instance 130-1 of the PCI segment. Similarly, a value "10" within the instance field of the second slave processor 122-2 may indicate that it is assigned to a third instance 130-0 of the PCI segment. By creating the instances 130-0 to 130-2 of the PCI segment 130 within the processors 120, 122-1 and 122-2, each of the processors 120, 122-1 and 122-2 may have access to a maximum number or greater of busses 132, as allowed per PCI segment 130 by the PCI specification, such as 256 busses.
[0026] The service unit 1 10 is to switch to the 64-bit memory mapped configuration not only after the device 100 has started to boot from a Basic Input/Output System (BIOS) but also prior to handing off system resources to an operating system (OS) of the device 100, such as a device driver. Thus, embodiments may allow a plurality of processors 120, 122-2 and 122-2 on a single device 100 to have access to a greater number of PCI buses 132 without modifying system firmware, such as to another format like 64-bit mode.
[0027] FIG. 2 is another example block diagram of a device 200 for assigning processors to a memory mapped configuration. The device 200 may be, for example, a server, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a network device, a controller, a wireless device, or any other type of device that is to interface with a Peripheral Component Interconnect (PCI). Further, the device 200 of FIG. 2 may include at least the functionality and/or hardware of the device 100 of FIG. 1 . For instance, the service unit 1 10, master processor 120 and slave processors 122-1 and 122-2 included in the device 200 of FIG. 2 may respectively include the functionality of the service unit 1 10, master processor 120 and slave processors 122-1 and 122-2 included in the device 100 of FIG. 1 . Further, the device 200 of FIG. 2 also includes a root complex 208, BIOS 210, an ACPI table/namespace resource 212 and PCI 214.
[0028] The BIOS 210 and ACPI table/namespace resource 212 may include, for example, a hardware device including electronic circuitry for implementing the functionality described below, such as control logic and/or memory. In addition or as an alternative, the BIOS 210 and ACPI table/namespace resource 212 may be implemented as a series of instructions or microcode encoded on a machine- readable storage medium and executable by a processor. The PCI 214 may provide a hardware and/or firmware interface for connecting one or more PCI peripheral devices (not shown) to the one or more the processors 120, 122-1 and 122-2 of the device 200.
[0029] In FIG. 2, the master processor 120 is shown to write to a shared memory 204 of the service unit 1 10 to indicate that the one or more slave processors 122-1 and 122-2 are quiesced. The service unit 1 10 is to receive control of the device 200 from the master processor 120 after the one or more slave processor are quiesced 122-1 and 122-2
[0030] The service unit 1 10 is to return control of the device 200 to the processors 120, 122-1 and 122-2 after the service unit 1 10 switches to the 64- bit memory mapped configuration. For example, the service unit 1 10 may enable a flag at a scratch register 204-1 to 204-3 of one or more of the processors 120, 122-1 and 122-2 after the service unit 1 10 switches to the 64- bit memory mapped configuration. In the meantime, the BIOS 210 may instruct the quiesced processors 120, 122-1 and 122-2 to poll their scratch registers 204-1 to 204-3 to determine if the flag is enabled. If the flag is enabled, the quiesced processors 120, 122-1 and 122-2 may transition from a quiesced state back to an active state. The service unit 1 10 is to transfer control of the device to the processors 120, 122-1 and 122-2, if the flag is enabled.
[0031 ] As explained above in FIG. 1 , after the service unit 1 10 switches to the 64-bit memory mapped configuration, each of the processors 120, 122-1 and 122-2 is assigned to have a different instance of a PCI segment, with each instance having access to all PCI buses. For example, a memory mapped configuration (MMCFG) 202-1 of the master processor 120 may have access to a first instance (PCI segment 0), a MMCFG 202-2 of the first slave processor 122-1 may have access to a second instance (PCI segment 1 ) and a MMCFG 202-3 of the second slave processor 122-2 may have access to a third instance (PCI segment 2), with each instance (PCI segments 0-2) having 256 buses (0- 255).
[0032] After the flag is enabled, the master processor 120 may walk the PCI busses to assign resources to the PCI segments (PCI segments 0-2). The master processor 120 may also construct Advanced Configuration and Power Interface (ACPI) static tables and namespace resources 212 to expose information related to the segments to an OS (not shown) of the device 200.
[0033] For example, the master processor 120, via the OS or firmware, may query all PCI buses at startup time (via PCI Configuration Space) to find out what devices are present and what system resources (memory space, I/O space, interrupt lines, etc.) each needs. Then the master processor 120 may allocate the resources and tell each device its allocation. While embodiments refer to 32 and 64 bit modes, embodiments are not limited thereto and may include other bit modes as well.
[0034] FIG. 3 is an example block diagram of a computing device 300 including instructions for assigning processors to a memory mapped configuration. In the embodiment of FIG. 3, the computing device 300 includes a processor 310 and a machine-readable storage medium 320. The machine-readable storage medium 320 further includes instructions 321 , 323, 325, 327 and 329 for assigning processors (not shown) to a memory mapped configuration. [0035] The computing device 300 may be, for example, a secure microprocessor, a notebook computer, a desktop computer, an all-in-one system, a server, a network device, a controller, a wireless device, or any other type of device capable of executing the instructions 321 , 323, 325, 327 and 329. In certain examples, the computing device 300 may include or be connected to additional components such as memories, controllers, processors etc.
[0036] The processor 310 may be, at least one central processing unit (CPU), at least one semiconductor-based microprocessor, at least one graphics processing unit (GPU), a microcontroller, special purpose logic hardware controlled by microcode or other hardware devices suitable for retrieval and execution of instructions stored in the machine-readable storage medium 320, or combinations thereof. The processor 310 may fetch, decode, and execute instructions 321 , 323, 325, 327 and 329 to implement assigning the processors to the memory mapped configuration. As an alternative or in addition to retrieving and executing instructions, the processor 310 may include at least one integrated circuit (IC), other control logic, other electronic circuits, or combinations thereof that include a number of electronic components for performing the functionality of instructions 321 , 323, 325, 327 and 329.
[0037] The machine-readable storage medium 320 may be any electronic, magnetic, optical, or other physical storage device that contains or stores executable instructions. Thus, the machine-readable storage medium 320 may be, for example, Random Access Memory (RAM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a storage drive, a Compact Disc Read Only Memory (CD-ROM), and the like. As such, the machine- readable storage medium 320 can be non-transitory. As described in detail below, machine-readable storage medium 320 may be encoded with a series of executable instructions for assigning the processors to the memory mapped configuration.
[0038] Moreover, the instructions 321 , 323, 325, 327 and 329 when executed by a processor (e.g., via one processing element or multiple processing elements of the processor) can cause the processor to perform processes, such as, the process of FIG. 4. For example, the initialize instructions 321 may be executed by the processor 310 to initialize a plurality of processors to access different buses of a Peripheral Component Interconnect (PCI) segment, each of the different buses assigned to only one of the plurality of processors. The quiesce instructions 323 may be executed by the processor 310 to quiesce the plurality of processors.
[0039] The transfer instructions 325 may be executed by the processor 310 to transfer control of the device from at least one of the quiesced processors to a service unit (not shown). The assign instructions 327 may be executed by the processor 310 to assign, via the service unit, each of the plurality of processors a memory mapped configuration that accesses all of the buses, with each of the processors assigned to different instances of the PCI segment, before an operating system (OS) of the device is loaded. The return instructions 329 may be executed by the processor 310 to return control of the device from the service unit to at least one of the quiesced processors. The PCI buses are to be walked and an Advanced Configuration and Power Interface (ACPI) setup (not shown), after control is returned the at least one of quiesced processor. [0040] FIG. 4 is an example flowchart of a method 400 for assigning processors to a memory mapped configuration. Although execution of the method 400 is described below with reference to the device 200, other suitable components for execution of the method 400 can be utilized, such as the device 100. Additionally, the components for executing the method 400 may be spread among multiple devices (e.g., a processing device in communication with input and output devices). In certain scenarios, multiple devices acting in coordination can be considered a single device to perform the method 400. The method 400 may be implemented in the form of executable instructions stored on a machine- readable storage medium, such as storage medium 320, and/or in the form of electronic circuitry.
[0041 ] At block 410, a plurality of processors 120, 122-1 and 122-2 receive 32-bit instructions from a BIOS 210 after a device starts a boot sequence. Then, at block 420, one of the plurality of processors 120 is selected to quiesce a remainder of the plurality of processors 122-1 and 122-2. Next, at block 430, the plurality of processors 120, 122-1 and 122-2 are assigned, such as by the service unit 1 10 to interface with a PCI 214 according to a 64-bit memory mapped configuration 202-1 to 202-3.
[0042] Each of the plurality of processors 120, 122-1 and 122-2 are to access separate buses 132 of a single PCI segment 130, before the assigning at block 430. Each of the plurality of processors 120, 122-1 and 122-2 are to access all of the buses 132 of different instances 130-0 to 130-2 of the PCI segment 130 bus, after the assigning at block 430. At block 440, the plurality of quiesced processors 120, 122-1 and 122-2 are activated, before hardware interfacing via the PCI is initialized.
[0043] According to the foregoing, embodiments provide a method and/or device for postponing enablement of multiple PCI segments until hardware initialization is complete. Thus, embodiments may provide greater PCI bus numbers while still allowing for system firmware code that is dependent on a Single PCI Segment configuration to run unmodified. As a result, embodiments may be simple to implement because a complexity associated with re-writing 32-bit code to make it 64-bit compliant, is reduced or eliminated. For example, embodiments may allow for supporting a single PCI segment during early boot and then using a processor, such as a Service Processor, to properly enable additional PCI segments prior to handing off system resources to an OS.

Claims

CLAIMS We claim:
1 . A device, comprising:
a master processor to quiesce one or more slave processors running on the device, each of the processors to have access to different buses of a Peripheral Component Interconnect (PCI) segment using a 32-bit based memory mapped configuration; and
a service unit to switch to a 64-bit memory mapped configuration after the master processor indicates the one or more slaves processors are quiesced, wherein
each of the processors is assigned to one of a plurality of instances of the PCI segment after the service unit switches to the 64-bit memory mapped configuration, each of the instances to have access to all of the buses of the PCI segment.
2. The device of claim 1 , wherein,
the service unit is to switch to the 64-bit memory mapped configuration via a sideband mechanism, and
the sideband mechanism is at least one of a Platform Environment Control Interface (PECI) and a System Management Bus (SMBus).
3. The device of claim 1 , wherein the service unit is to switch to the 64-bit memory mapped configuration after the device has started to boot from a Basic Input/Output System (BIOS) and prior to handing off system resources to an operating system (OS) of the device.
4. The device of claim 1 , wherein,
the master processor is to write to a shared memory of the service unit to indicate that the one or more slave processor are quiesced,
the service unit to receive control of the device from the master processor after the one or more slave processor are quiesced, and
the service unit to return control of the device to the processors after bus after the service unit switches to the 64-bit memory mapped configuration.
5. The device of claim 4, wherein,
the quiesced processors are to poll a scratch register of each of the quiesced processors to determine to determine if a flag is enabled, and
the service unit is to enable the flag after the service unit switches to the 64-bit memory mapped configuration.
6. The device of claim 5, wherein,
the quiseced processors are to transition from a quiesced state to an active state if the flag is enabled, and
the service unit is to transfer control of the device to the processors if the flag is enabled.
7. The device of claim 5, wherein after the flag is enabled, the master processor is to at least one of,
walk the PCI bus to assign resources to the PCI segments, and construct Advanced Configuration and Power Interface (ACPI) static tables and namespace resources to expose information related to the segments to the operating system (OS) of the device.
8. The device of claim 1 , wherein the service unit is to at least one of pre-enable the 64-bit memory mapped configuration, program addresses and set registers, after the master processor quiesces the one or more slave processors.
9. The device of claim 1 , wherein,
an access to the memory mapped configuration is in a 32-bit range before the service unit switches to the 64-bit memory mapped configuration, and
hardware of the device is initialized after the service unit switches to the 64-bit memory mapped configuration.
10. The device of claim 1 , wherein,
the quiesced processors do not transmit traffic over a fabric external to the processor, and
the quiesced processor do not carry out operations via the memory mapped configuration.
1 1 . The device of claim 1 , wherein,
the service unit configures one or more resources of the device to switch the PCI segment to the 64-bit memory mapped configuration, and
the configuring the one or more resources includes at least one of setting a base number of a PCI domain of each of the processors to zero, configuring each root-complex to decode the 64-bit memory mapped configuration.
12. A method, comprising:
receiving, at a plurality of processors, 32-bit instructions from a Basic Input/Output System (BIOS) after a device starts a boot sequence;
selecting one of the plurality of processors to quiesce a remainder of the plurality of processors;
assigning the plurality of processors to interface with a Peripheral Component Interconnect (PCI) according to a 64-bit memory mapped configuration; and
activating the plurality of quiesced processors, before hardware interfacing via the PCI is initialized.
13. The method of claim 12, wherein,
each of the plurality of processors are to access separate buses of a single PCI segment before the assigning, and
each of the plurality of processors are to access all of the buses of different instances of the PCI segment bus after the assigning.
14. A non-transitory computer-readable storage medium storing instructions that, if executed by a processor of a device, cause the processor to: initialize a plurality of processors to access different buses of a Peripheral Component Interconnect (PCI) segment, each of the different buses assigned to only one of the plurality of processors;
quiesce the plurality of processors;
transfer control of the device from at least one of the quiesced processors to a service unit;
assign, via the service unit, each of the plurality of processors a memory mapped configuration that accesses all of the buses, with each of the processors assigned to different instances of the PCI segment, before an operating system (OS) of the device is loaded; and
return control of the device from the service unit to at least one of the quiesced processors.
15. The non-transitory computer-readable storage medium of claim 14, the PCI buses are to be walked and an Advanced Configuration and Power Interface (ACPI) setup, after control is returned the at least one of quiesced processor.
PCT/US2013/023552 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration WO2014120113A1 (en)

Priority Applications (5)

Application Number Priority Date Filing Date Title
EP13873433.0A EP2951705A4 (en) 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration
US14/763,501 US9940291B2 (en) 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration
PCT/US2013/023552 WO2014120113A1 (en) 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration
CN201380071515.4A CN105264506B (en) 2013-01-29 2013-01-29 Processor is distributed to configuration memory map
TW102145163A TWI512479B (en) 2013-01-29 2013-12-09 Assigning processors to memory mapped configuration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/US2013/023552 WO2014120113A1 (en) 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration

Publications (1)

Publication Number Publication Date
WO2014120113A1 true WO2014120113A1 (en) 2014-08-07

Family

ID=51262685

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2013/023552 WO2014120113A1 (en) 2013-01-29 2013-01-29 Assigning processors to memory mapped configuration

Country Status (5)

Country Link
US (1) US9940291B2 (en)
EP (1) EP2951705A4 (en)
CN (1) CN105264506B (en)
TW (1) TWI512479B (en)
WO (1) WO2014120113A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114185607A (en) * 2022-02-14 2022-03-15 苏州浪潮智能科技有限公司 Starting control method, device and equipment for PCI (peripheral component interconnect) equipment in ARM (advanced RISC machines) server

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8402188B2 (en) 2008-11-10 2013-03-19 Micron Technology, Inc. Methods and systems for devices with a self-selecting bus decoder
TWI561986B (en) * 2015-12-21 2016-12-11 Mitac Computing Technology Corp Method for allocating memory resource
CN106951326B (en) * 2017-03-16 2023-01-06 腾讯科技(深圳)有限公司 File unlocking method and electronic equipment
US11038749B2 (en) * 2018-12-24 2021-06-15 Intel Corporation Memory resource allocation in an end-point device
TWI764082B (en) 2019-02-14 2022-05-11 美商萬國商業機器公司 Method, computer system and computer program product for interrupt signaling for directed interrupt virtualization
WO2020164935A1 (en) 2019-02-14 2020-08-20 International Business Machines Corporation Directed interrupt virtualization with running indicator
CA3130164A1 (en) 2019-02-14 2020-08-20 International Business Machines Corporation Directed interrupt for multilevel virtualization
EP3924819A1 (en) 2019-02-14 2021-12-22 International Business Machines Corporation Directed interrupt for multilevel virtualization with interrupt table
TWI727607B (en) 2019-02-14 2021-05-11 美商萬國商業機器公司 Method, computer system and computer program product for directed interrupt virtualization with interrupt table
WO2020164820A1 (en) 2019-02-14 2020-08-20 International Business Machines Corporation Directed interrupt virtualization
TWI759677B (en) 2019-02-14 2022-04-01 美商萬國商業機器公司 Method, computer system and computer program product for directed interrupt virtualization with fallback
TWI710911B (en) * 2019-05-02 2020-11-21 緯創資通股份有限公司 Electronic system, host device and control method
US11386020B1 (en) * 2020-03-03 2022-07-12 Xilinx, Inc. Programmable device having a data processing engine (DPE) array
US20220335003A1 (en) * 2021-04-19 2022-10-20 Advanced Micro Devices, Inc. Master-Slave Communication with Subdomains
CN114328323A (en) * 2021-12-01 2022-04-12 北京三快在线科技有限公司 Data transfer unit and data transmission method based on data transfer unit

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050114639A1 (en) * 2001-05-11 2005-05-26 Zimmer Vincent J. Hardened extensible firmware framework to support system management mode operations using 64-bit extended memory mode processors
US20070097949A1 (en) * 2005-10-27 2007-05-03 Boyd William T Method using a master node to control I/O fabric configuration in a multi-host environment
US7356636B2 (en) * 2005-04-22 2008-04-08 Sun Microsystems, Inc. Virtualized PCI switch
US20110276779A1 (en) * 2010-05-05 2011-11-10 International Business Machines Corporation Memory mapped input/output bus address range translation
US8250348B2 (en) * 2005-05-19 2012-08-21 International Business Machines Corporation Methods and apparatus for dynamically switching processor mode

Family Cites Families (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5481738A (en) * 1992-02-20 1996-01-02 International Business Machines Corporation Apparatus and method for communicating a quiesce and unquiesce state between elements of a data processing complex
US6092170A (en) * 1996-11-29 2000-07-18 Mitsubishi Denki Kabushiki Kaisha Data transfer apparatus between devices
US6070207A (en) * 1998-01-26 2000-05-30 Intel Corporation Hot plug connected I/O bus for computer system
US6738882B1 (en) * 1999-11-30 2004-05-18 Hewlett-Packard Development Company, L.P. Concurrent multi-processor memory testing beyond 32-bit addresses
WO2001061935A1 (en) * 2000-02-17 2001-08-23 Conexant Systems, Inc, Cable modem having a programmable media access controller
US6990685B1 (en) 2000-06-15 2006-01-24 Dell Products L.P. System and method for tracking bootable devices
US6704831B1 (en) * 2000-11-16 2004-03-09 Sun Microsystems, Inc. Method and apparatus for converting address information between PCI bus protocol and a message-passing queue-oriented bus protocol
US6848046B2 (en) * 2001-05-11 2005-01-25 Intel Corporation SMM loader and execution mechanism for component software for multiple architectures
US7185135B1 (en) 2002-07-12 2007-02-27 Cypress Semiconductor Corporation USB to PCI bridge
US7953074B2 (en) 2003-01-21 2011-05-31 Emulex Design And Manufacturing Corporation Apparatus and method for port polarity initialization in a shared I/O device
US7155379B2 (en) * 2003-02-25 2006-12-26 Microsoft Corporation Simulation of a PCI device's memory-mapped I/O registers
JP4783907B2 (en) 2005-01-07 2011-09-28 国立大学法人京都大学 Optical sensor and manufacturing method thereof
US7447822B2 (en) * 2005-12-12 2008-11-04 Inventec Corporation Hot-plug control system and method
GB0623276D0 (en) 2006-11-22 2007-01-03 Transitive Ltd Memory consistency protection in a multiprocessor computing system
US7650259B2 (en) * 2007-10-01 2010-01-19 International Business Machines Corporation Method for tuning chipset parameters to achieve optimal performance under varying workload types
US7827333B1 (en) * 2008-02-04 2010-11-02 Nvidia Corporation System and method for determining a bus address on an add-in card
US8332593B2 (en) * 2008-12-24 2012-12-11 Nuon, Inc. Memory space management and mapping for memory area network
JP5401679B2 (en) 2009-02-19 2014-01-29 株式会社日立製作所 Computer system, management method and management server
US8615645B2 (en) * 2010-06-23 2013-12-24 International Business Machines Corporation Controlling the selectively setting of operational parameters for an adapter
US8593629B2 (en) 2011-02-17 2013-11-26 Hewlett-Packard Development Company, L.P. Apparatus for performing SERS
WO2012161683A1 (en) 2011-05-20 2012-11-29 Hewlett-Packard Development Company L.P. Surface enhanced raman spectroscopy sensor, system and method of sensing
US8949474B1 (en) * 2011-11-21 2015-02-03 Marvell International Ltd. Method for inter-chip and intra-chip addressing using port identifiers and address mapping
US8645594B2 (en) * 2012-06-29 2014-02-04 Intel Corporation Driver-assisted base address register mapping
US9626324B2 (en) * 2014-07-08 2017-04-18 Dell Products L.P. Input/output acceleration in virtualized information handling systems

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050114639A1 (en) * 2001-05-11 2005-05-26 Zimmer Vincent J. Hardened extensible firmware framework to support system management mode operations using 64-bit extended memory mode processors
US7356636B2 (en) * 2005-04-22 2008-04-08 Sun Microsystems, Inc. Virtualized PCI switch
US8250348B2 (en) * 2005-05-19 2012-08-21 International Business Machines Corporation Methods and apparatus for dynamically switching processor mode
US20070097949A1 (en) * 2005-10-27 2007-05-03 Boyd William T Method using a master node to control I/O fabric configuration in a multi-host environment
US20110276779A1 (en) * 2010-05-05 2011-11-10 International Business Machines Corporation Memory mapped input/output bus address range translation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2951705A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114185607A (en) * 2022-02-14 2022-03-15 苏州浪潮智能科技有限公司 Starting control method, device and equipment for PCI (peripheral component interconnect) equipment in ARM (advanced RISC machines) server
CN114185607B (en) * 2022-02-14 2022-06-14 苏州浪潮智能科技有限公司 Starting control method, device and equipment for PCI (peripheral component interconnect) equipment in ARM (advanced RISC machines) server

Also Published As

Publication number Publication date
US20150356049A1 (en) 2015-12-10
EP2951705A1 (en) 2015-12-09
CN105264506B (en) 2018-01-26
US9940291B2 (en) 2018-04-10
TW201430572A (en) 2014-08-01
EP2951705A4 (en) 2016-11-02
TWI512479B (en) 2015-12-11
CN105264506A (en) 2016-01-20

Similar Documents

Publication Publication Date Title
US9940291B2 (en) Assigning processors to memory mapped configuration
US9229730B2 (en) Multi-chip initialization using a parallel firmware boot process
US8589902B2 (en) Policy description technique in UEFI firmware
KR101574403B1 (en) Combined virtual graphics device
US8595723B2 (en) Method and apparatus for configuring a hypervisor during a downtime state
US8166288B2 (en) Managing requests of operating systems executing in virtual machines
US10860332B2 (en) Multicore framework for use in pre-boot environment of a system-on-chip
US20090265708A1 (en) Information Processing Apparatus and Method of Controlling Information Processing Apparatus
US8037210B2 (en) Computer and method for directly accessing computer hardware by virtual system
EP1256055B1 (en) Computer system including a memory access controller for using non-system memory storage resources during system boot time
JP5308522B2 (en) Memory management for hypervisor loading
EP2409234A1 (en) Inter operating system memory hotswap to support memory growth in a non-virtualized system
RU2611018C2 (en) Common method of constructing a virtual pci-device and a virtual mmio-device
US9417886B2 (en) System and method for dynamically changing system behavior by modifying boot configuration data and registry entries
US10620963B2 (en) Providing fallback drivers for IO devices in a computing system
WO2000019292A2 (en) Upgrade card for a computer system
WO2005038652A1 (en) Optimization of smi handling and initialization
US10437762B2 (en) Partitioned interconnect slot for inter-processor operation
US10853284B1 (en) Supporting PCI-e message-signaled interrupts in computer system with shared peripheral interrupts
US9910677B2 (en) Operating environment switching between a primary and a secondary operating system
US11714756B2 (en) Information handling systems and methods to improve the security and performance of a shared cache memory contained within a multi-core host processor
US20060242351A1 (en) Method and apparatus for loading instructions into high memory
Banik et al. Understanding the BIOS and Minimalistic Design
US11829772B2 (en) Heterogeneous compute domains with an embedded operating system in an information handling system
CN117075973A (en) Novel Linux boot starting and guiding method and system based on RISC-V server CPU

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 201380071515.4

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13873433

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 14763501

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

REEP Request for entry into the european phase

Ref document number: 2013873433

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2013873433

Country of ref document: EP