WO2010044409A1 - マルチルートpciエクスプレススイッチ、その起動方法、及び、マルチルートpciマネージャプログラム - Google Patents
マルチルートpciエクスプレススイッチ、その起動方法、及び、マルチルートpciマネージャプログラム Download PDFInfo
- Publication number
- WO2010044409A1 WO2010044409A1 PCT/JP2009/067758 JP2009067758W WO2010044409A1 WO 2010044409 A1 WO2010044409 A1 WO 2010044409A1 JP 2009067758 W JP2009067758 W JP 2009067758W WO 2010044409 A1 WO2010044409 A1 WO 2010044409A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- pci
- root
- switch
- virtual
- pci express
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 25
- 230000004913 activation Effects 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 5
- 238000012546 transfer Methods 0.000 claims description 5
- 230000004044 response Effects 0.000 claims description 3
- 230000003213 activating effect Effects 0.000 claims 1
- 230000002093 peripheral effect Effects 0.000 abstract description 2
- 238000011144 upstream manufacturing Methods 0.000 description 17
- 238000010586 diagram Methods 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
- G06F13/38—Information transfer, e.g. on bus
- G06F13/40—Bus structure
- G06F13/4004—Coupling between buses
- G06F13/4022—Coupling between buses using switching circuits, e.g. switching matrix, connection or expansion network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/4401—Bootstrapping
- G06F9/4411—Configuring for operating with peripheral devices; Loading of device drivers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
- G06F13/38—Information transfer, e.g. on bus
- G06F13/40—Bus structure
- G06F13/4063—Device-to-bus coupling
- G06F13/4068—Electrical coupling
- G06F13/4081—Live connection to bus, e.g. hot-plugging
Definitions
- the present invention relates to a computing device used in a computer and a network device, and a PCI express press switch that extends a PCI express interface of a networking device.
- the present invention relates to an MRA (Multi-Root Aware) PCI Express switch for connecting a plurality of computers, a method for starting the switch, and a multi-root PCI manager program.
- MRA Multi-Root Aware
- PCI Peripheral component interconnect
- CPU central processing unit
- hard disk hard disk
- graphic controller graphics controller
- networking devices such as a forwarding engine and a network controller.
- PCI Express which serializes a PCI parallel bus and performs communication by a packet system, has been standardized.
- Non-patent document 1 describes a method of connecting a plurality of CPUs on a system constituted by the PCI express switch.
- FIG. 1 shows a system configuration diagram that can be realized in Non-Patent Document 1
- FIGS. 2 and 3 show the configuration of an MRA (Multi-Root Aware) PCI Express switch proposed in Non-Patent Document 1.
- MRA Multi-Root Aware
- the multi-root PCI manager software MR-PCIM 52 for managing the configuration state of the switch is required on any of the CPUs connected to the switch shown in FIG.
- FIG. 1 shows a system configuration using two MRA PCI Express switches.
- 2 and 3 show the internal configuration of the MRA PCI Express switches 511 and 512.
- FIG. 1 shows a system configuration using two MRA PCI Express switches.
- the MRA PCI express switch 511 is a PCI-PCI bridge 5111, 5112, 5122, 5113, 5114, 5115, a switch between bridges different from the conventional PCI express switch in order to accommodate a plurality of route complexes and MR endpoints.
- the virtual PCI express switches 5116, 5117, 5118, and 5119 that perform processing, the MRA controller logic 5120 that is set and controlled from the MR-PCIM 52, and a setting register 5121 that stores the setting information are included.
- Another MRA PCI express switch 512 for expanding the number of connection ports is similarly a PCI-PCI bridge 5123, 5124, 5133 different from the conventional PCI express switch, and a virtual PCI express switch for performing a switching process between the bridges. 5126, 5127, and 5129, MRA controller logic 5120 that is set and controlled by the MR-PCIM 52, and a setting register 5131 that stores the setting information.
- MRA PCI Express Switch in order to configure a tree similar to the conventional PCI Express for each of the multiple root complexes, which virtual PCI Express switch is used for each root complex, multiple PCI-PCI bridges are used. It is necessary to assign a VH (Virtual ⁇ Hierarchy) number for use in the network. Since switch processing is performed for each virtual PCI express switch, there is no interference with other virtual PCI express switches.
- VH Virtual ⁇ Hierarchy
- VH numbers are allocated, and it is considered that multiple PCI-PCI bridges are connected to one link. With different VH numbers, credit control, configuration information separation, and the like are performed, and a plurality of PCI-PCI bridge usage numbers on one port are indicated.
- These settings are assigned by the MR-PCIM to the setting register which is the control information of the MRA controller logic.
- 4 and 5 show configuration examples of the setting registers 5121 and 5131 of the MRA PCI express switches 511 and 512, respectively.
- the MR-PCIM is connected to Port: 1 to which the CPU 101 and the root complex 102 are connected.
- the route of the number VH0 is used in the MR PCI-PCI bridge for setting from the MR-PCIM.
- the setting information of the setting register includes, for the virtual PCI express switch number, information for configuring the virtual PCI express switch, switch port use number, use VH number, UP / Down of upstream PCI-PCI bridge or downstream PCI-PCI bridge : DN information is stored for the number of ports.
- the virtual PCI Express switch 1 5116 is switch number 1 and is port 1.
- the MR complex upstream PCI-PCI bridge 5111, the virtual PCI express switch 1 5116, the MR compatible downstream PCI-PCI bridge 5113, the MR endpoint 505, the downstream PCI-PCI bridge 5114, The end point 106 and the MR compliant downstream PCI-PCI bridge 5122 can be accessed.
- Access to the end of the MR compatible upstream / downstream PCI-PCI bridge 5122 can be made by setting the setting register 5131 of the MRA PCI express switch 512.
- the setting register 5131 Port 2 / VH 0 / DN and Port 3 / VH 0 / UP are set for the virtual PCI Express switch 2 5127. Therefore, the MR compatible upstream / downstream PCI-PCI bridge 5122 is connected to the MR compatible upstream / downstream PCI-PCI bridge 5124, the virtual PCI Express switch 2 5127, the MR compatible downstream PCI-PCI bridge 5133, and the MR end point 506. Accessible from the root complex 102.
- the virtual PCI express switch 2 5117, the virtual PCI express switch 3 5118 in the setting register 5121, and the virtual PCI express switch 1 5126 in the setting register 5131 are also set as shown in FIGS. With these settings, the PCI tree shown in FIG. 6 is built for the I / O device with respect to the CPU 101, and the PCI tree shown in FIG.
- the MR-PCIM does not configure the PCI tree unless the MR-PCIM takes into account the switch connection state and the logical connection state and sets the setting register. Cannot start until complete. Therefore, depending on the scale of the system, it takes a very long time to start the root complex system, and the operability of the system is significantly reduced.
- FIG. 8 shows the MR-PCIM 52 sequence. After the MR-PCIM is activated, it first grasps how many types of switches are connected and how many links the switch has (step S1). This can be grasped only by referring to the register of the switch.
- step S2 it is necessary to grasp the physical connection status and the topology in which the switches are connected (step S2). Since this operation is performed while considering whether the topology state is a loop or the like, it takes time to search depending on the structure of the kibo and the topology. Next, it is confirmed whether the I / O is connected to the tip of the topology, and if it is connected, the type of the I / O is grasped (step S3). This operation also takes time depending on the number of I / Os. Thereafter, after the setting requested by the user, that is, the setting of the table information shown in FIGS. 4 and 5 (step S4) is completed, the use of the switch is finally permitted to the root complex (step S5), and the CPU 101 and the CPU 1011 are activated. become able to.
- the start sequence of the root complex of the CPU 101 and the CPU 1011 is as shown in FIG. 9 by the operation of the MR-PCIM shown in FIG. Depending on the scale and configuration of the system, it takes a very long time, while MR-PCIM grasps the connection status (step S11), while MR-PCIM grasps the type of I / O (step S12).
- the root complex will wait a long time. Subsequent MR-PCIM tree setting is completed (step S13), and when the use permission is issued from MR-PCIM (step S14), the root complex is finally reset and the BIOS can be started. (Step S15).
- the configuration register includes a bus number lower limit value 18001, a bus number upper limit value 18002, a 32-bit I / O address lower limit value 18003, a 32-bit I / O address upper limit value 18004, a 32-bit memory lower limit value 18005, and a 32-bit memory upper limit value 18006.
- the information is transferred to the OS, and then the OS is started (step S17).
- the OS is started (step S17).
- the I / O driver and configuration existing on the set PCI tree are executed.
- the activation of the OS is completed and the software can be used (step S18).
- An object of the present invention is to provide a multi-root PCI express switch that can be activated without waiting for a root complex, regardless of the scale of the system, in an MRA PCI express switch that accommodates a plurality of root complex tems. It is to improve.
- the multi-root PCI express switch is a multi-root PCI express switch that can be connected to a plurality of root complexes, and is based on the topology of the switch connection destination and the physical connection status.
- a setting register that stores information necessary for the setting of the virtual switch bridge
- a virtual switch bridge control unit that stores information necessary for constructing a virtual PCI tree regardless of the state of the setting register.
- a multi-root PCI express switch activation method is a multi-root PCI express switch activation method connected to a plurality of root complexes, and includes a switch that configures the system when the system is activated.
- a multi-root PCI manager program is a multi-root PCI manager program that causes a computer to manage a multi-root PCI express switch connected to a plurality of root complexes, and the multi-root PCI manager program is activated when the system is started.
- Setting information necessary for constructing a virtual PCI tree in the multi-root PCI express switch based on information on the number of switches constituting the express switch and the number of links of the switch; and on the basis of the set information, Granting the root complex permission to use the multi-root PCI express switch; confirming the topology and physical connection status of the connection destination of the multi-root PCI express switch; Has a step of setting the information required for Lee built the multi-route PCI Express switch, the.
- the root complex in a multi-route compatible PCI Express switch, at the time of system startup, the root complex can be started without waiting until the topology of the connection destination and the physical connection status can be grasped, and the system startup time is shortened. Can be realized.
- the reason is that a virtual PCI tree is constructed before the topology and physical connection status of the connection destination can be grasped, and the root complex can be activated based on the virtual PCI tree.
- the root complex in a multi-route compatible PCI Express switch, can be started at a fixed timing regardless of the scale of the system, and reset and power management control for the root complex are simplified.
- Can be The reason is that a virtual PCI tree is constructed before the topology and physical connection status of the connection destination can be grasped, and the root complex can be activated based on the virtual PCI tree.
- MRA PCI Express switch This is the internal configuration of the MRA PCI Express switch. This is the internal configuration of the MRA PCI Express switch. It is a structural example of the setting register of MRA PCI Express switch. It is a structural example of the setting register of MRA PCI Express switch. It is a PCI tree that can be seen from the CPU in the MRA PCI Express switch. It is a PCI tree that can be seen from the CPU in the MRA PCI Express switch. It is a figure explaining the subject in the operation sequence of MR-PCIM. It is a figure explaining the subject in the operation
- FIG. 1 is an internal configuration diagram of an MRA extended PCI Express switch according to an embodiment of the present invention.
- FIG. 1 is an internal configuration diagram of an MRA extended PCI Express switch according to an embodiment of the present invention.
- FIG. It is a block diagram of the multistage virtual switch bridge controller of the MRA extended PCI Express switch according to an embodiment of the present invention. It is a figure which shows the structural example of the PCI switch bridge virtual configuration register information by one Embodiment of this invention.
- FIG. 5 is a diagram illustrating an MR-PCIM operation sequence of the MRA extended PCI Express switch according to an embodiment of the present invention.
- a multi-root PCI express switch In a multi-root PCI express switch, a multi-root PCI-PCI bridge that can be connected to the plurality of root complexes, and a plurality of PCI-PCI bridges provided corresponding to the plurality of root complexes and including the multi-root PCI-PCI bridge It is preferable to further include a virtual PCI express switch for performing connection between the bridges.
- the virtual switch bridge control unit preferably includes a PCI switch bridge virtual configuration register set based on information on the number of switches constituting the system and the number of links of the switches.
- the multi-root PCI express switch further includes an actual configuration register in which address information necessary for data transfer is stored, and the virtual switch bridge control unit performs the PCI switch bridge virtual configuration based on the setting of the setting register. It is preferable to further include a virtual register synchronization control unit that performs synchronization control between the register and the actual configuration register.
- the virtual switch bridge control unit issues a hot plug event for matching the virtual PCI tree with the actual PCI tree according to the setting state of the multi-root PCI express. It is preferable to further include a hot plug event issuing unit.
- the multi-root PCI express switch activation method is as described in the second aspect.
- the multi-root PCI express switch holds information on the virtual PCI tree and replaces the virtual PCI tree with a real-configuration PCI tree. It is preferable that the multi-root PCI express switch has a starting method.
- a multi-root PCI manager program of a host computer creates a virtual PCI tree based on information on the number of switches constituting the system and the number of links of the switch when the system is activated.
- a process for setting information necessary for construction in the multi-root PCI express switch is performed, the root complex starts a BIOS system, constructs the virtual PCI tree, and the multi-root PCI manager program
- After confirming the topology of the connection destination and confirming the connection status, information necessary for constructing an actual configuration PCI tree is set in the multi-root PCI express switch, and the root complex is configured by the actual configuration PCI.
- Tree Based on the information set in order to built, it is preferable the virtual PCI tree is a multi-root PCI Express switch activation method of replacing the PCI tree of the actual configuration.
- the multi-root PCI manager program grasps the number of switches constituting the system and the number of links of the switch by referring to a register in the multi-root PCI express switch. Then, information necessary for constructing a virtual PCI tree is set, a VH (Virtual Hierarchy) number is fixedly assigned to the root complex to which the host computer is connected, and the host computer is connected. In response to the use permission, the root complex that has received the use permission activates the BIOS system, constructs a virtual PCI tree, and the multi-root PCI manager.
- VH Virtual Hierarchy
- the topology of the connection destination is confirmed to grasp the physical connection status, whether or not the I / O is connected, and if connected, the I / O It is preferable that the method is a multi-root PCI express switch activation method that performs a process of grasping the type of O.
- the multi-root PCI manager program allows the multi-root PCI express switch to respond to the setting in order to grasp the physical connection status and the type of I / O.
- this is a multi-root PCI express switch activation method that issues a hot plug event and causes the root complex to match the virtual PCI tree with the actual PCI tree.
- the multi-root PCI manager program is preferably as described in the third aspect.
- the computer-readable storage medium that records the multi-root PCI manager program is preferably a computer-readable storage medium that records the multi-root PCI manager program according to the eleventh aspect.
- FIG. 11 shows a configuration diagram of an MRA extended PCI Express system using the present invention.
- the CPU 101, the CPU 108, the route complex 102, the memory 103, the route complex 109, and the memory 110 which are also used in the system composed of the conventional MRA PCI express switch, are used as the MRA extended PCI express switch 9511.
- the MR end point 505, the end point 106, and the end point 107 are connected and connected to the MRA expansion PCI Express switch 9511.
- the MRA expansion PCI express switch 9512 is connected to a CPU 1011, a route complex 1021, a memory 1031, and an MR end point 506.
- the root complex and end point that can be connected are a conventional PCI Express compatible device and an MR end point defined in Non-Patent Document 1, and can communicate between devices as in FIG.
- the MRA extended PCI Express switch uses the upstream / downstream / MR compatible downstream PCI-PCI bridge, MRA controller logic, and configuration registers that are also used in MRA PCI Express.
- An upstream / downstream PCI-PCI bridge constituting the switch and a multi-stage virtual switch bridge control unit that displays the PCI-PCI bridge as virtual I / O are provided.
- FIG. 12 shows the configuration of the MRA extended PCI express switch 9511
- FIG. 13 shows the configuration of the MRA extended PCI express switch 9512.
- the MRA extended PCI express switches 9511 and 9512 are configured such that, in addition to the conventional MRA PCI express switches 511 and 512, a multistage virtual switch bridge control unit 81 is provided.
- VH allocation is performed in the same way as the MRA PCI Express switch.
- the setting from the MR-PCIM also sets the setting register via the VH0 port and the MRA controller logic, so that the virtual PCI Express switch to be used and its port, VH number, upstream or downstream are determined.
- the PCI Express tree is constructed for the root complex, and access to the end point becomes possible. Accordingly, the process of transferring data between the root complex and the endpoints in the MRA extended PCI express switches 9511 and 9512 is performed in the same manner as the conventional MRA PCI express switch.
- the feature of the MRA extended PCI Express switch is that the multi-stage virtual switch bridge controller shows multi-stage virtual switches and bridges for the root complex.
- the MRA extended PCI express switch 9511 shown in FIG. 12 the MR complex upstream PCI-PCI bridge 5111 and the upstream PCI-PCI bridge 5112 are connected to the root complex 102 and the root complex 109 in a multistage virtual switch. Configure the bridge.
- the MRA extended PCI express switch 9512 shown in FIG. 13 a virtual switch and a bridge are configured ahead of the upstream PCI-PCI bridge 5123 with respect to the root complex 1021.
- FIG. 14 shows details of the internal configuration of the multistage virtual switch bridge control unit 81.
- the multistage virtual switch bridge control unit 81 includes a configuration transmission / reception control unit 821, a multistage virtual switch virtual bridge control unit 822, PCI switch bridge virtual configuration register information 824, virtual register synchronization control 825, and a hot plug event issue unit 823.
- the configuration transmission / reception control unit 821 includes a configuration transmission / reception control unit 821, a multistage virtual switch virtual bridge control unit 822, PCI switch bridge virtual configuration register information 824, virtual register synchronization control 825, and a hot plug event issue unit 823.
- the multistage virtual switch bridge control unit 81 obtains setting information indicating how many switches the system is composed of and the number of links of the switch from the MR-PCIM at the time of system startup.
- the switch and bridge stage number setting 812 is set for the multistage virtual switch virtual bridge controller 822.
- the multi-stage virtual switch virtual bridge controller 822 makes it appear that there are virtual upstream / downstream PCI-PCI bridges and PCI bridges as seen from the root complex.
- a plurality of bridge virtual configuration register information 824 are configured.
- the root complex has a start permission, the BIOS starts, and the PCI tree search program runs. Thereafter, configuration access occurs to the MRA extended PCI Express switch, and the bridge configuration packet 813 is received via the upstream PCI-PCI bridge.
- the PCI switch bridge virtual configuration register information 824 is accessed.
- FIG. 15 shows a detailed configuration of the PCI switch bridge virtual configuration register information 824.
- This register has the same configuration as the configuration register 1502 held by the actual PCI-PCI bridge shown in FIG. 10, and this register allows a virtual PCI-PCI bridge to the root complex to be configured. As a PCI bridge, it can be shown to the root complex.
- the multistage virtual switch bridge control unit 81 refers to the setting register of the MRA extended PCI express switch to grasp the configuration state of the actual switch, and when the actual PCI-PCI bridge becomes available.
- the virtual register synchronization control 825 performs synchronization control by copying and mirroring register information to the virtual PCI-PCI bridge corresponding to the actual PCI-PCI bridge and the switch internal configuration register 814. .
- the hot plug event issuing unit 823 has a virtual PCI -Issue hot-remove (disconnect) of PCI bridge and hot-add (grant) of actual PCI-PCI bridge to match from virtual state to actual use state .
- FIG. 16 shows an operation sequence of the MR-PCIM 952 that manages the MRA extended PCI Express switches 9511 and 9512.
- the MR-PCIM 952 first knows what type of switches are connected and how many links are connected to the switches (step S21). This can be grasped only by referring to the register of the switch.
- the VH number of the port to which the root complex is connected is fixedly assigned as an initial value (step S22). Since the link to which the root complex is connected is not shared, there is no problem even if an arbitrary fixed number is set for the VH number.
- the MR-PCIM permits the root complex to use the switch (step S23).
- step S24 grasp the physical connection status of the topology in which the switch is connected (step S24), confirm the I / O connected to the tip of the topology and grasp the type (Step S25), the setting desired by the user, that is, the table information shown in FIGS. 4 and 5 is set (step S26).
- FIG. 17 shows an operation sequence of the root complex for the operation of the MR-PCIM 952 described above.
- the CPU 101 and the root complex of the CPU 1011 receive an instruction from the MR-PCIM for permission to use the switch immediately after the system is activated (step S31), and activate the BIOS system (step S32).
- the activation permission is issued to the root complex after the MR-PCIM is activated immediately after a certain time, and the root complex is almost reset and starts the BIOS. Will be able to.
- the upstream PCI-PCI bridge beyond the port to which the root complex is connected has not been set by the setting register yet, so the virtual PCI Express switch is not used and multistage virtual switch bridge control is performed.
- the multi-stage virtual switch bridge provided by the unit is recognized and activated. Since the multistage virtual switch bridge receives information on the number of virtual switch stages and the number of links immediately after MR-PCIM activation, a virtual switch bridge has already been configured.
- FIG. 18 shows a PCI tree composed of a virtual switch and a bridge immediately after startup, as viewed from the root complex 102.
- the end link is a virtual bridge.
- the BIOS recognizes the PCI tree configured by this virtual switch bridge, configures the virtual switch bridge, and then starts the OS.
- MR-PCIM 952 grasps the time-consuming physical connection status and grasps the I / O type while the root complex is set up according to the instruction from MR-PCIM.
- the system startup time is shortened by performing the processing related to the time of the BIOS and OS startup, which takes time as the root complex processing.
- the setting register of the switch is read as the tree setting, the VH number is set to the VS number, and the actual PCI- The PCI bridge can be used.
- the virtual register synchronization control 825 and the hot plug event issuing unit 823 shown in FIG. 14 perform hot plug processing for information synchronization and configuration change when an actual PCI-PCI bridge is used. Match to the same state as system operation. That is, as shown in steps S36 to S39 in FIG.
- the MR-PCIM grasps the physical connection status (step S36), grasps the type of I / O (step S37), and the tree requested by the user Information necessary for the setting is set in the setting register (95121, 95131) (step S38). Further, the virtual PCI tree is reconfigured by the hot plug event so as to match the actual PCI tree (step S39).
- FIG. 19 shows a part to be deleted on the PCI tree.
- the net line portion is a portion that has been found to be unnecessary as a PCI tree by the setting of VH / VS.
- This part is configured so that the hot-plug event issuing unit 823 performs hot-remove processing, and the actual configuration is the same PCI tree state.
- the virtual PCI-PCI bridge and virtual bridge are changed to the actual PCI-PCI bridge and real I / O with Hot-Remove (detach) and Hot-Add (add) while maintaining the state of the PCI tree.
- the configuration is the same as that shown in FIG.
- the location that took time to grasp the physical connection status and I / O type which was a problem with the conventional MRA PCI Express switch, indicates that the OS is running while the root complex is running the BIOS. Since the activation time is used effectively, the activation time of the system is shortened, and the user can use the software in the same PCI tree as the conventional system.
- the multi-root PCI express switch can be connected to a plurality of root complexes (102, 109, 1021) as shown in FIGS.
- a setting register (95121, 95131) for storing information necessary for setting the PCI tree based on the topology of the connection destination of the switch and the physical connection status, and the status of the setting register
- the virtual switch bridge controller 81 stores information necessary for constructing a virtual PCI tree.
- the multi-root PCI express switch includes the virtual switch bridge control unit 81 that stores information necessary for constructing a virtual PCI tree.
- the root complex can be activated without waiting until the physical connection status can be grasped.
- the time until the root complex is activated can be made constant regardless of the scale of the system.
- the multi-root PCI express switch includes, for example, a multi-root PCI-PCI bridge (5113, 5133) that can be connected to a plurality of root complexes, as shown in FIGS.
- a plurality of bridges (5111, 5112, 5122, 5113, 5114, 5115, 5123, 5124) including a multi-root compatible PCI-PCI bridge (5113, 5133) provided corresponding to the plurality of root complexes (102, 109, 1021) , 5133), a virtual PCI express switch (5116, 5117, 5118, 5119, 5126, 5127, 5129) may be further provided.
- the multi-root PCI express switch activation method is, for example, a multi-route connected to a plurality of route complexes (102, 109, 1021 in FIG. 11) as shown in FIGS.
- a step of setting necessary information in the multi-root PCI express switch (S21, S22), a step of starting a BIOS system based on the set information and building a virtual PCI tree (S33); After confirming the topology and the connection status, the virtual PCI tree A step (S36 ⁇ S39) to replace the PCI tree of the actual configuration, the.
- the BIOS system can be started up in a short time. Also, since the virtually constructed PCI tree (FIG. 18) is replaced with the actual configuration PCI tree after confirming the topology and connection status of the connection destination (see FIG. 19), there is no inconsistency.
- the multi-root PCI manager program is connected to a plurality of root complexes (102, 109, 1021 in FIG. 11).
- 9511, 9512 is a multi-root PCI manager program for managing a computer (101 in FIG. 11), and the number of switches constituting the multi-root PCI express switch (9511, 9512) and the number of switch links (S21, S22) for setting information necessary for constructing a virtual PCI tree (for example, FIG. 18) in the multi-root PCI express switch (9511, 9512 in FIG.
- the multi-root PCI manager program can be installed in a computer via a storage medium such as a CD ROM, DVD, flexible disk, ROM, flash memory, RAM, or hard disk.
- the computer can execute a multi-root PCI manager program stored in the storage medium.
- the multi-root PCI manager it is necessary to construct a virtual PCI tree based on information on the number of switches and the number of links of switches that can be grasped only by referring to the switch registers at the time of system startup.
- Information is set in the multi-root PCI express switch, and based on this information, the root complex is permitted to use the multi-root PCI express switch. Therefore, the root complex can be activated early without depending on the scale of the system.
- the information set for constructing the virtual PCI tree is the information necessary for constructing the actual PCI tree after confirming the topology and physical connection status of the switch connection destination. Since the virtual PCI tree can be updated to the actual PCI tree by plug processing or the like, no inconsistency occurs.
- MR-PCIM in an MRA PCI Express switch that accommodates a plurality of root complexes, can immediately start the root complex without waiting for the time to grasp the physical configuration and logical configuration. System startup time can be shortened.
- the reason is that in addition to the configuration of the MRA PCI Express switch, it has a multi-stage virtual switch and a virtual switch bridge control unit that makes the virtual bridge appear in the root complex. This is because the MR-PCIM can permit the activation of the immediate root complex by using a typical switch and bridge.
- the MR-PCIM in the MRA PCI Express system, can grant the activation permission to the root complex immediately after activation at a fixed timing regardless of the scale of the system. Reset and power management control for the root complex can be simplified. The reason for this is that in addition to the configuration of the multi-root PCI express switch, it has a multi-stage virtual switch and a virtual switch bridge control unit that makes the virtual bridge appear in the root complex. This is because MR-PCIM can allow the immediate root complex to be activated by using virtual switches and bridges.
- the computer system and the network system connected by wire have been described in the present embodiment, but the present invention can be used for all system environments equipped with a PCI Express interface regardless of wired or wireless.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Computer Security & Cryptography (AREA)
- Stored Programmes (AREA)
- Bus Control (AREA)
Abstract
Description
本発明は、日本国特許出願:特願2008-266444号(2008年10月15日出願)の優先権主張に基づくものであり、同出願の全記載内容は引用をもって本書に組み込み記載されているものとする。
第1の側面に記載のとおりのマルチルートPCIエクスプレススイッチが好ましい。
マルチルートPCIエクスプレススイッチにおいて、前記複数のルートコンプレックスに対して接続可能なマルチルート対応PCI-PCIブリッジと、前記複数のルートコンプレックスに対応して設けられ、前記マルチルート対応PCI-PCIブリッジを含む複数のブリッジ間の接続を行う仮想PCIエクスプレススイッチと、をさらに備えることが好ましい。
マルチルートPCIエクスプレススイッチにおいて、前記仮想スイッチブリッジ制御部は、システムを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて設定されるPCIスイッチブリッジ仮想コンフィグレーションレジスタを備えることが好ましい。
マルチルートPCIエクスプレススイッチにおいて、データ転送時に必要なアドレス情報が格納された実コンフィグレーションレジスタをさらに備え、前記仮想スイッチブリッジ制御部が、前記設定レジスタの設定に基づいて、前記PCIスイッチブリッジ仮想コンフィグレーションレジスタと前記実コンフィグレーションレジスタとの同期制御を行う仮想レジスタ同期制御部をさらに備えることが好ましい。
マルチルートPCIエクスプレススイッチにおいて、前記仮想スイッチブリッジ制御部が、前記マルチルートPCIエクスプレスの設定状態に応じて、前記仮想的なPCIツリーを実構成のPCIツリーと整合させるためのホットプラグイベントを発行するホットプラグイベント発行部をさらに備えることが好ましい。
第2の側面に記載のとおりのマルチルートPCIエクスプレススイッチの起動方法であることが好ましい。
マルチルートPCIエクスプレススイッチの起動方法において、前記マルチルートPCIエクスプレススイッチが、前記仮想的なPCIツリーの情報を保持し、前記仮想的なPCIツリーを実構成のPCIツリーに置き換えさせる仮想スイッチブリッジ制御部を備えているマルチルートPCIエクスプレススイッチの起動方法であることが好ましい。
マルチルートPCIエクスプレススイッチの起動方法において、ホストコンピュータのマルチルートPCIマネージャプログラムが、前記システム起動時に、システムを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて仮想的なPCIツリーを構築するために必要な情報を前記マルチルートPCIエクスプレススイッチに設定する処理を行い、前記ルートコンプレックスが、BIOSシステムを起動し、前記仮想的なPCIツリーを構築し、前記マルチルートPCIマネージャプログラムが、前記接続先のトポロジを確認して接続状況を確認した後、実構成のPCIツリーを構築するために必要な情報を前記マルチルートPCIエクスプレススイッチに設定し、前記ルートコンプレックスが、前記実構成のPCIツリーを構築するために設定された情報に基づいて、前記仮想的なPCIツリーを実構成のPCIツリーに置き換えるマルチルートPCIエクスプレススイッチの起動方法であることが好ましい。
マルチルートPCIエクスプレススイッチの起動方法において、前記マルチルートPCIマネージャプログラムは、前記マルチルートPCIエクスプレススイッチ内のレジスタを参照することにより、前記システムを構成するスイッチの数、及び、スイッチのリンク数を把握して仮想的なPCIツリーを構築するために必要な情報を設定し、前記ホストコンピュータが接続されるルートコンプレックスに対してVH(Virtual Hierarchy)番号を固定で割り当てて、前記ホストコンピュータが接続されるルートコンプレックスに対してスイッチの利用許可を与え、前記利用許可に応答して、前記利用許可を受けたルートコンプレックスが前記BIOSシステムを起動し、仮想的なPCIツリーを構築し、前記マルチルートPCIマネージャプログラムは、前記ルートコンプレックスの処理と並行して前記接続先のトポロジを確認して物理的な接続状況を把握し、I/Oが接続されているか否か確認し、接続されている場合はI/Oの種類を把握する処理を行うマルチルートPCIエクスプレススイッチの起動方法であることが好ましい。
マルチルートPCIエクスプレススイッチの起動方法において、前記マルチルートPCIマネージャプログラムは、前記物理的な接続状況の把握、I/Oの種別の把握に、前記マルチルートPCIエクスプレススイッチが、その設定に応答してホットプラグイベントを発行して、前記ルートコンプレックスに前記仮想的なPCIツリーと実構成のPCIツリーとの整合を行わせるマルチルートPCIエクスプレススイッチの起動方法であることが好ましい。
マルチルートPCIマネージャプログラムとしては、第3の側面に記載のとおりとすることが好ましい。
マルチルートPCIマネージャプログラムを記録したコンピュータが読み取り可能な記憶媒体としては、上記形態11のマルチルートPCIマネージャプログラムを記録したコンピュータが読み取り可能な記憶媒体であることが好ましい。
102,109,1021 ルートコンプレックス
103,110,1031 メモリ
104 PCIエクスプレススイッチ
105,106,107 エンドポイント
505,506 MRエンドポイント
52,952 MR-PCIM
511,512 MRA PCIエクスプレススイッチ
5111 MR対応上流PCI-PCIブリッジ
5122,5124 MR対応上流・下流PCI-PCIブリッジ
5116,5117,5118,5119,5126,5127,5129 仮想PCIエクスプレススイッチ
5113,5133 MR対応下流PCI-PCIブリッジ
5114,5115 下流PCI-PCIブリッジ
5120 MRAコントローラ論理
5121,5131 設定レジスタ
5112,5123 上流PCI-PCIブリッジ
1502 コンフィグレーションレジスタ
18001 バス番号下限値
18002 バス番号上限値
18003 32ビットI/Oアドレス下限値
18004 32ビットI/Oアドレス上限値
18005 32ビットメモリ下限値
18006 32ビットメモリ上限値
18007 64ビットメモリ下限値
18008 64ビットメモリ上限値
18009 パケット転送ロジック
9511,9512 MRA拡張PCIエクスプレススイッチ
81 多段仮想スイッチブリッジ制御部
821 コンフィグレーション送受制御部
822 多段仮想スイッチ仮想ブリッジ制御部
823 ホットプラグイベント発行部
824 PCIスイッチブリッジ仮想コンフィグレーションレジスタ情報
825 仮想レジスタ同期制御
Claims (11)
- 複数のルートコンプレックスに対して接続可能なマルチルートPCIエクスプレススイッチにおいて、
スイッチの接続先のトポロジ、物理的な接続状況に基づいてPCIツリーの設定に必要な情報を格納する設定レジスタと、
前記設定レジスタの状態によらず、仮想的なPCIツリーを構築するために必要な情報を格納する仮想スイッチブリッジ制御部と、
を備えたマルチルートPCIエクスプレススイッチ。 - 前記複数のルートコンプレックスに対して接続可能なマルチルート対応PCI-PCIブリッジと、
前記複数のルートコンプレックスに対応して設けられ、前記マルチルート対応PCI-PCIブリッジを含む複数のブリッジ間の接続を行う仮想PCIエクスプレススイッチと、
をさらに備えた請求項1記載のマルチルートPCIエクスプレススイッチ。 - 前記仮想スイッチブリッジ制御部は、システムを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて設定されるPCIスイッチブリッジ仮想コンフィグレーションレジスタを備えた請求項1又は2記載のマルチルートPCIエクスプレススイッチ。
- データ転送時に必要なアドレス情報が格納された実コンフィグレーションレジスタをさらに備え、
前記仮想スイッチブリッジ制御部が、前記設定レジスタの設定に基づいて、前記PCIスイッチブリッジ仮想コンフィグレーションレジスタと前記実コンフィグレーションレジスタとの同期制御を行う仮想レジスタ同期制御部をさらに備えた請求項1乃至3いずれか1項記載のマルチルートPCIエクスプレススイッチ。 - 前記仮想スイッチブリッジ制御部が、前記マルチルートPCIエクスプレスの設定状態に応じて、前記仮想的なPCIツリーを実構成のPCIツリーと整合させるためのホットプラグイベントを発行するホットプラグイベント発行部をさらに備えた請求項1乃至4いずれか1項記載のマルチルートPCIエクスプレススイッチ。
- 複数のルートコンプレックスに接続されるマルチルートPCIエクスプレススイッチの起動方法であって、
システム起動時に、システムを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて仮想的なPCIツリーを構築するために必要な情報を前記マルチルートPCIエクスプレススイッチに設定するステップと、
前記設定された情報に基づいて、BIOSシステムを起動し、仮想的なPCIツリーを構築するステップと、
前記接続先のトポロジを確認して接続状況を確認した後、前記仮想的なPCIツリーを
実構成のPCIツリーに置き換えるステップと、
を有するマルチルートPCIエクスプレススイッチの起動方法。 - 前記マルチルートPCIエクスプレススイッチが、前記仮想的なPCIツリーの情報を保持し、前記仮想的なPCIツリーを実構成のPCIツリーに置き換えさせる仮想スイッチブリッジ制御部を備えている請求項6記載のマルチルートPCIエクスプレススイッチの起動方法。
- ホストコンピュータのマルチルートPCIマネージャプログラムが、前記システム起動時に、システムを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて仮想的なPCIツリーを構築するために必要な情報を前記マルチルートPCIエクスプレススイッチに設定する処理を行い、
前記ルートコンプレックスが、BIOSシステムを起動し、前記仮想的なPCIツリーを構築し、
前記マルチルートPCIマネージャプログラムが、前記接続先のトポロジを確認して接続状況を確認した後、実構成のPCIツリーを構築するために必要な情報を前記マルチルートPCIエクスプレススイッチに設定し、
前記ルートコンプレックスが、前記実構成のPCIツリーを構築するために設定された情報に基づいて、前記仮想的なPCIツリーを実構成のPCIツリーに置き換える
請求項6又は7記載のマルチルートPCIエクスプレススイッチの起動方法。 - 前記マルチルートPCIマネージャプログラムは、前記マルチルートPCIエクスプレススイッチ内のレジスタを参照することにより、前記システムを構成するスイッチの数、及び、スイッチのリンク数を把握して仮想的なPCIツリーを構築するために必要な情報を設定し、前記ホストコンピュータが接続されるルートコンプレックスに対してVH(Virtual Hierarchy)番号を固定で割り当てて、前記ホストコンピュータが接続されるルートコンプレックスに対してスイッチの利用許可を与え、
前記利用許可に応答して、前記利用許可を受けたルートコンプレックスが前記BIOSシステムを起動し、仮想的なPCIツリーを構築し、
前記マルチルートPCIマネージャプログラムは、前記ルートコンプレックスの処理と並行して前記接続先のトポロジを確認して物理的な接続状況を把握し、I/Oが接続されているか否か確認し、接続されている場合はI/Oの種類を把握する処理を行う請求項8記載のマルチルートPCIエクスプレススイッチの起動方法。 - 前記マルチルートPCIマネージャプログラムは、前記物理的な接続状況の把握、I/Oの種別の把握に、
前記マルチルートPCIエクスプレススイッチが、その設定に応答してホットプラグイベントを発行して、前記ルートコンプレックスに前記仮想的なPCIツリーと実構成のPCIツリーとの整合を行わせる請求項9記載のマルチルートPCIエクスプレススイッチの起動方法。 - 複数のルートコンプレックスに接続されるマルチルートPCIエクスプレススイッチをコンピュータに管理させるマルチルートPCIマネージャプログラムであって、
システム起動時に、前記マルチルートPCIエクスプレススイッチを構成するスイッチの数、及び、スイッチのリンク数の情報に基づいて仮想的なPCIツリーの構築に必要な情報を前記マルチルートPCIエクスプレススイッチに設定するステップと、
前記設定した情報に基づいて、前記ルートコンプレックスに前記マルチルートPCIエクスプレススイッチの使用許可を与えるステップと、
前記マルチルートPCIエクスプレススイッチの接続先のトポロジ、物理的な接続状況を確認して実構成のPCIツリー構築に必要な情報を前記マルチルートPCIエクスプレススイッチに設定するステップと、を有するマルチルートPCIマネージャプログラム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010533906A JP5440507B2 (ja) | 2008-10-15 | 2009-10-14 | マルチルートpciエクスプレススイッチ、その起動方法、及び、マルチルートpciマネージャプログラム |
US13/060,980 US8719483B2 (en) | 2008-10-15 | 2009-10-14 | Multi-root PCI express switch, boot method thereof, and multi-root PCI manager program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2008-266444 | 2008-10-15 | ||
JP2008266444 | 2008-10-15 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010044409A1 true WO2010044409A1 (ja) | 2010-04-22 |
Family
ID=42106572
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2009/067758 WO2010044409A1 (ja) | 2008-10-15 | 2009-10-14 | マルチルートpciエクスプレススイッチ、その起動方法、及び、マルチルートpciマネージャプログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US8719483B2 (ja) |
JP (1) | JP5440507B2 (ja) |
WO (1) | WO2010044409A1 (ja) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100211717A1 (en) * | 2009-02-19 | 2010-08-19 | Hitachi, Ltd. | Computer system, method of managing pci switch, and management server |
US20100312943A1 (en) * | 2009-06-04 | 2010-12-09 | Hitachi, Ltd. | Computer system managing i/o path and port |
JP2011514590A (ja) * | 2008-02-26 | 2011-05-06 | ヒューレット−パッカード デベロップメント カンパニー エル.ピー. | ホスト列挙プロセスを実行するための方法及び装置 |
WO2012073304A1 (ja) * | 2010-11-29 | 2012-06-07 | 株式会社日立製作所 | 計算機システム、それに使用されるスイッチ及びパケット転送制御方法 |
JPWO2013175570A1 (ja) * | 2012-05-22 | 2016-01-12 | 富士通株式会社 | 情報処理装置、制御方法、及び制御プログラム |
CN109408454A (zh) * | 2018-11-01 | 2019-03-01 | 郑州云海信息技术有限公司 | 一种芯片管理的方法以及相关装置 |
US10437473B2 (en) * | 2016-05-27 | 2019-10-08 | Huawei Technologies Co., Ltd. | Storage system and method for scanning for devices |
Families Citing this family (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8554973B2 (en) * | 2010-04-23 | 2013-10-08 | Hitachi, Ltd. | Storage device and method for managing size of storage device |
US8949499B2 (en) * | 2010-06-24 | 2015-02-03 | International Business Machines Corporation | Using a PCI standard hot plug controller to modify the hierarchy of a distributed switch |
WO2012023149A2 (en) * | 2010-08-19 | 2012-02-23 | Ineda Systems Pvt. Ltd | Multi-root input output virtualization aware switch |
WO2012023152A2 (en) * | 2010-08-19 | 2012-02-23 | Ineda Systems Pvt. Ltd | Multi-processor electronic systems |
CN102819517A (zh) * | 2011-06-08 | 2012-12-12 | 鸿富锦精密工业(深圳)有限公司 | Pcie接口卡 |
US8843688B2 (en) * | 2012-09-11 | 2014-09-23 | International Business Machines Corporation | Concurrent repair of PCIE switch units in a tightly-coupled, multi-switch, multi-adapter, multi-host distributed system |
US9262363B2 (en) | 2012-11-30 | 2016-02-16 | Ciena Corporation | PCI and PCI express virtual hot plug systems and methods |
KR102007368B1 (ko) * | 2012-12-17 | 2019-08-05 | 한국전자통신연구원 | Pci 익스프레스 스위치 및 이를 이용한 컴퓨터 시스템 |
US9436630B2 (en) * | 2013-06-11 | 2016-09-06 | Western Digital Technologies, Inc. | Using dual phys to support multiple PCIe link widths |
US20150254082A1 (en) * | 2014-03-10 | 2015-09-10 | Plx Technology, Inc. | Remote booting over pci express using synthetic remote boot capability |
US9779050B2 (en) | 2015-04-27 | 2017-10-03 | Red Hat Israel, Ltd. | Allocating virtual resources to root PCI bus |
US10572426B2 (en) * | 2015-06-02 | 2020-02-25 | Nxp Usa, Inc. | System-level redundancy in PCI express equipment |
CN108139924B (zh) | 2016-05-31 | 2021-10-08 | 安华高科技股份有限公司 | 热插拔硬件及软件实施方案 |
US10133699B2 (en) | 2016-08-05 | 2018-11-20 | Red Hat Israel, Ltd. | Hot-plug of devices in virtualized computer systems |
KR102669219B1 (ko) | 2016-08-16 | 2024-05-28 | 삼성전자주식회사 | 종단 장치로부터 제공되는 부트 업 명령어에 기초하여 부팅 동작을 수행하도록 구성되는 전자 장치 |
US10708199B2 (en) | 2017-08-18 | 2020-07-07 | Missing Link Electronics, Inc. | Heterogeneous packet-based transport |
US11356388B2 (en) | 2017-08-18 | 2022-06-07 | Missing Link Electronics, Inc. | Real-time multi-protocol heterogeneous packet-based transport |
US10853299B2 (en) | 2017-09-15 | 2020-12-01 | Dell Products L.P. | Hot-plugged PCIe device configuration system |
CN108804230B (zh) * | 2018-06-11 | 2020-05-26 | 苏州浪潮智能科技有限公司 | 一种根据pci设备的类型分配资源的方法、装置及介质 |
CN111651515B (zh) * | 2020-08-10 | 2020-10-27 | 成都智明达电子股份有限公司 | 可对外供电的高速数据卸载设备及其使用方法 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001265598A (ja) * | 2000-03-15 | 2001-09-28 | Matsushita Electric Ind Co Ltd | 情報処理装置 |
JP2008046722A (ja) * | 2006-08-11 | 2008-02-28 | Nec Corp | I/o装置及び方法 |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4501916B2 (ja) | 2006-09-20 | 2010-07-14 | 日本電気株式会社 | I/o機器の共有システムと情報処理装置共有システム及びそれらに用いる方法 |
US8271604B2 (en) * | 2006-12-19 | 2012-09-18 | International Business Machines Corporation | Initializing shared memories for sharing endpoints across a plurality of root complexes |
US7657663B2 (en) * | 2006-12-19 | 2010-02-02 | International Business Machines Corporation | Migrating stateless virtual functions from one virtual plane to another |
US7813366B2 (en) * | 2006-12-19 | 2010-10-12 | International Business Machines Corporation | Migration of a virtual endpoint from one virtual plane to another |
US7979592B1 (en) * | 2007-02-09 | 2011-07-12 | Emulex Design And Manufacturing Corporation | Virtualization bridge device |
JP2011516964A (ja) * | 2008-04-01 | 2011-05-26 | ヒューレット−パッカード デベロップメント カンパニー エル.ピー. | Pciデバイス用pciメモリ空間の確保 |
US8359415B2 (en) * | 2008-05-05 | 2013-01-22 | International Business Machines Corporation | Multi-root I/O virtualization using separate management facilities of multiple logical partitions |
WO2009139489A1 (ja) * | 2008-05-16 | 2009-11-19 | 日本電気株式会社 | Pciエクスプレススイッチ、pciエクスプレスシステム、及びネットワーク制御方法 |
-
2009
- 2009-10-14 US US13/060,980 patent/US8719483B2/en active Active
- 2009-10-14 JP JP2010533906A patent/JP5440507B2/ja active Active
- 2009-10-14 WO PCT/JP2009/067758 patent/WO2010044409A1/ja active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001265598A (ja) * | 2000-03-15 | 2001-09-28 | Matsushita Electric Ind Co Ltd | 情報処理装置 |
JP2008046722A (ja) * | 2006-08-11 | 2008-02-28 | Nec Corp | I/o装置及び方法 |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011514590A (ja) * | 2008-02-26 | 2011-05-06 | ヒューレット−パッカード デベロップメント カンパニー エル.ピー. | ホスト列挙プロセスを実行するための方法及び装置 |
US8533381B2 (en) * | 2009-02-19 | 2013-09-10 | Hitachi, Ltd. | Computer system, method of managing PCI switch, and management server |
US20100211717A1 (en) * | 2009-02-19 | 2010-08-19 | Hitachi, Ltd. | Computer system, method of managing pci switch, and management server |
US20100312943A1 (en) * | 2009-06-04 | 2010-12-09 | Hitachi, Ltd. | Computer system managing i/o path and port |
US8407391B2 (en) * | 2009-06-04 | 2013-03-26 | Hitachi, Ltd. | Computer system managing I/O path and port |
JP5399570B2 (ja) * | 2010-11-29 | 2014-01-29 | 株式会社日立製作所 | 計算機システム、それに使用されるスイッチ及びパケット転送制御方法 |
WO2012073304A1 (ja) * | 2010-11-29 | 2012-06-07 | 株式会社日立製作所 | 計算機システム、それに使用されるスイッチ及びパケット転送制御方法 |
US9396150B2 (en) | 2010-11-29 | 2016-07-19 | Hitachi, Ltd. | Computer system and method utilizing a PCIe switch to control transfer of packets |
JPWO2013175570A1 (ja) * | 2012-05-22 | 2016-01-12 | 富士通株式会社 | 情報処理装置、制御方法、及び制御プログラム |
US9703590B2 (en) | 2012-05-22 | 2017-07-11 | Fujitsu Limited | Information processing apparatus including bridges that connect virtual machines and physical devices, and control method thereof |
US10437473B2 (en) * | 2016-05-27 | 2019-10-08 | Huawei Technologies Co., Ltd. | Storage system and method for scanning for devices |
CN109408454A (zh) * | 2018-11-01 | 2019-03-01 | 郑州云海信息技术有限公司 | 一种芯片管理的方法以及相关装置 |
CN109408454B (zh) * | 2018-11-01 | 2021-10-22 | 郑州云海信息技术有限公司 | 一种芯片管理的方法以及相关装置 |
Also Published As
Publication number | Publication date |
---|---|
US8719483B2 (en) | 2014-05-06 |
US20110185163A1 (en) | 2011-07-28 |
JPWO2010044409A1 (ja) | 2012-03-15 |
JP5440507B2 (ja) | 2014-03-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5440507B2 (ja) | マルチルートpciエクスプレススイッチ、その起動方法、及び、マルチルートpciマネージャプログラム | |
EP2927812B1 (en) | Method and apparatus for extending pcie bus domain | |
KR101324844B1 (ko) | 기능 변환 시스템 및 방법과 변환 수행 시스템 | |
JP5598148B2 (ja) | スイッチング装置、情報処理装置、及びスイッチング装置の制御方法 | |
US7822908B2 (en) | Discovery of a bridge device in a SAS communication system | |
JP5763873B2 (ja) | データ処理システムの複数のルート複合体間の通信のために共用メモリを初期設定するための方法、コンピュータ・プログラム、およびデータ処理システム | |
JP2019145153A (ja) | スイッチングデバイス、ペリフェラル・コンポーネント・インターコネクト・エクスプレスシステムおよびその初期化方法 | |
KR102007368B1 (ko) | Pci 익스프레스 스위치 및 이를 이용한 컴퓨터 시스템 | |
JP4799222B2 (ja) | 情報処理装置、情報処理システムおよび初期化方法 | |
US11995019B2 (en) | PCIe device with changeable function types and operating method thereof | |
US11983136B2 (en) | PCIe device and operating method thereof | |
US11928070B2 (en) | PCIe device | |
US8775712B2 (en) | Bus connecting device for connecting host with external device | |
JP2011113163A (ja) | Ioアクセス通信システムにおけるエンドポイント間通信制御装置および方法 | |
US20230318606A1 (en) | Interface device and method of operating the same | |
JP2001027920A (ja) | バスブリッジ回路及びそのパワーマネージメント方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09820587 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13060980 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2010533906 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 09820587 Country of ref document: EP Kind code of ref document: A1 |