Abstract: An apparatus and method for mediating a sequence of transactions across a fabric in a data processing system are implemented. A fabric bridge orders a preceding transaction and a subsequent transaction according to a predetermined protocol. Using the protocol a determination is made whether the subsequent transaction may be allowed to bypass the previous transaction, must be allowed to bypass the previous transaction, or must not be allowed to bypass the preceding transaction. Transactions include load/store (L/S) system memory and L/S to input/output (I/O) device, and direct memory access (DMA) to system memory and DMA peer-to-peer transactions.

Abstract: A method, system, and apparatus for preventing input/output (I/O) adapters used by an operating system (OS) image, in a logically partitioned data processing system, from fetching or corrupting data from a memory location allocated to another OS image within the data processing system is provided. A hypervisor prevents transmission of data between an input/output adapter in one of the logical partitions and memory locations assigned to other logical partitions during a direct memory access (DMA) operation by assigning each of the input/output adapters a range of I/O bus DMA addresses. The I/O adapters (IOAs) are connected to PCI host bridges via terminal bridges. A single terminal bridge may support multiple IOAs, in which case every terminal bridge has a plurality of sets of range registers, each associated with a respective one of the IOAs to which it is connected. An arbiter is provided which selects one of the input/output adapters to use the PCI bus.

Abstract: A methodology and implementing system are provided in which pipelined read transfers or PRTs are implemented. The PRTs include a request phase and a response phase. The PRT request phase involves a PRT request master delivering to a PRT request target, a source address, a destination address and the transfer size for the data being requested. In the PRT response phase, the PRT request target becomes a PRT response master, i.e. a PCI bus master, and initiates a completion of the transaction that was requested in the originating PRT request. Pipelined read transfers are ordered in accordance with a predetermined processing priority to improve performance and avoid deadlocks.

Abstract: A bus bridge including a buffer pool and steering logic where the buffer pool is organized as a plurality of buffers sets including at least first and second buffer sets and the steering logic is adapted to store transactions originating with a first peripheral device in the first buffer set and transactions originating with a second peripheral device in the second buffer set. Transactions may arrive via a secondary bus, such as a PCI bus, coupled to the bus bridge. The bridge further allows relaxed transaction ordering rules compared to conventional PCI transaction ordering rules by identifying transactions by grant signals and thus allows steering of transactions from the first and second devices to first and second buffer sets respectively. The bridge is suitably adapted for combining or merging two or more transactions within each buffer set. Each buffer set preferably includes one or more buffers for temporarily storing transactions arriving from the secondary bus and bound for a primary bus.

Abstract: A host bridge having a plurality of pre-defined registers used for injecting errors to a selected device so that other devices are not affected and normal systems operations can continue is disclosed. In accordance with the method and system of the present invention, device select lines from each device are brought into the host bridge individually for determining if an error is to be injected to a selected device. First, a register or a bit in a register in the host bridge is matched against an incoming bus operation for the type of bus operation, a load or a store, to inject the error upon. Next, a register having an initial or random value within the host bridge indicates which occurrence of the operation to inject the error. If the value of the register indicates that an error is to be injected, the load or store operation is delayed by forcing zero byte enables until the device identifier of the selected device may be checked against a device register within the host bridge.

Abstract: A bus bridge including a buffer pool comprised of a first and a second buffer sets. The first and second buffer sets are associated with first and second peripheral devices respectively. The bridge is configured to receive an interrupt and identify the interrupt source. A buffer set associated with the interrupt source is selected and transactions in the selected buffer set flushed prior to forwarding the interrupt to a processor. The bridge is preferably configured to identify the interrupt source by receiving a first interrupt signal from the first peripheral device and a second interrupt signal from the second peripheral device. Preferably, the bridge is configured to flush the transactions by pushing them into system memory via a primary bus such as a host bus of a processor.

Abstract: A method and apparatus is provided in which I/O data is tagged to identify an ordering of data transfer requests relative to other data transfer requests. Write and Read transaction requests are tagged for ordering relative to previous write requests. Current read and write transaction requests are selectively allowed to bypass earlier write transaction requests which have been temporarily delayed in transfer. In one embodiment, the bypass occurs in a bridge buffer positioned between I/O devices and a system memory. In another embodiment, the methodology is applied where a split read or write transaction includes reserved bits in the attribute fields which are utilized to indicate if the transaction is allowed to bypass previous write transactions.

Abstract: A method and system for supporting multiple Peripheral Component Interconnect (PCI) local buses through a single PCI host bridge having multiple PCI interfaces within a data-processing system are disclosed. In accordance with the method and system of the present invention, a processor and a system memory are connected to a system bus. One or more PCI local buses are connected to the system bus through a single PCI host bridge having bus and frequency control logic and bus clocks. The PCI local buses include sets of in-line electronic switches, dividing each PCI local bus into PCI local bus segments for supporting more PCI peripheral component slots then are called out by the PCI local bus standard. The sets of in-line electronic switches are open and closed in accordance with the bus and frequency control logic within the PCI host bridge thereby allowing the PCI peripheral component slots to operate at different bus frequencies, including bus frequencies higher than 66 MHz by using the bus clocks.

Abstract: A methodology and implementing system are provided in which pipelined read transfers or PRTs are implemented. The PRTs include a request phase and a response phase. The PRT request phase involves a PRT request master delivering to a PRT request target, a source address, a destination address and the transfer size for the data being requested. In the PRT response phase, the PRT request target becomes a PRT response master, i.e. a PCI bus master, and initiates a completion of the transaction that was requested in the originating PRT request.

Abstract: Device selects lines from each I/O device are brought into a PCI host bridge individually so that the device number of a failing device may be logged in an error register when an error is seen on the PCI bus. Until the error register is reset, subsequent load and store operations are delayed until the device number of the subject device may be checked against the error register. If the subject device is a previously failing device, the load/store operation to that device is prevented from completing, either by forcing bad parity or zeroing all byte enables. By forcing bad parity of zero byte enables, the I/O device will respond to the load or store request by activating its device select line, but will not accept store data. Operations to devices which are not logged in the error register are permitted to proceed normally, as are all load store operations when the error register is clear.

Abstract: A bus bridge coupled between primary and secondary busses including a buffer pool with first and second buffer sets and steering logic configured to direct transactions received from first and second peripheral devices to the first and second buffer sets respectively. The bridge is configured to push posted memory write transactions posted in the first buffer set onto the primary bus ahead of and in response to a read request transaction from the first peripheral device while leaving transactions in the second buffer set unaffected. In one embodiment, the steering logic is configured to receive first and second grant signals produced by arbitration logic of the bridge. The first and second grant signals indicate mastership of the secondary bus and the source of a subsequent transaction to be received via the secondary bus. The bridge and the secondary bus are suitably compliant with the PCI protocol. The primary bus may be the host bus of a processor unit or a peripheral bus such as a PCI bus.

Abstract: A peripheral interconnect for a computer system comprising a bridge, a peripheral bus, and a peripheral device, wherein at least one of these components is adapted to selectively operate in either a high performance mode or a low performance mode, the high performance mode using a first operating speed and a first protocol, and the low performance mode using a second operating speed which is lower than said first operating speed, and a second protocol which is different from the first protocol. The disclosed embodiment provides a high performance mode with a 100 MHz speed and a protocol that disallows pacing, and a low performance mode that uses a 66 MHz or 33 MHz speed and a standard PCI protocol that allows pacing. The high performance operating speed can be twice the low performance operating speed, by doubling the clock frequency and clocking data on only one clock edge, or by clocking data on both a rising edge and a falling edge of a clock signal while operating at the lower clock frequency.

Abstract: A method and system for supporting multiple Peripheral Component Interconnect (PCI) local buses through a single PCI host bridge having multiple PCI interfaces within a data-processing system are disclosed. In accordance with the method and system of the present invention, a processor and a system memory are connected to a system bus. First and second PCI local buses are connected to the system bus through a PCI host bridge. The first and second PCI local buses have sets of in-line electronic switches, dividing the PCI local buses into PCI local bus segments supporting a plurality of PCI peripheral component slots for connecting PCI devices. The sets of in-line electronic switches are open and closed in accordance with bus control logic within the PCI host bridge allowing up to fourteen or more PCI peripheral component slots for connecting up to fourteen PCI devices to have access through a single PCI host bridge to the system bus.

Abstract: A data processing system includes a processor, system memory and a number of peripheral devices, and one or more bridges which may connect between the processor, memory and peripheral devices and other hosts or peripheral devices such as in a network. A bridge, such as a PCI host bridge, connects between a primary bus (e.g system bus) and a secondary bus. The host bridge provides a dual host bridge function which creates two secondary bus interfaces. This allows increased loading capability under one dual host bridge compared to a lesser number of slots allowed under one normal host bridge. Also included is additional control logic for providing arbitration control and for steering transactions to the appropriate bus interface. Additionally, peer to peer support across the two secondary bus interfaces in provided.

Abstract: A method and apparatus are provided in which a control scheme is implemented to enable a PCI bus to operate more than two PCI slots into which PCI devices may be installed. The PCI slots are checked to determine if a PCI device is installed in the slots and the speed at which the installed PCI devices are capable of running. If any of the slots has a 33 MHz device installed in any of the slots, the system is enabled to run more than two slots, and all of the PCI devices will run at 33 MHz. When no 33 MHz cards or devices are installed in the PCI slots, and PCI devices are only installed in the first two slots, then the system is enabled to run only the first two slots at the speed of 66 MHz. In one alternative embodiment, a default configuration routine sets the PCI bus speed at one of the operating frequencies and modifies that default if it is determined during a system configuration cycle that another speed is more appropriate.

Abstract: A method and implementing computer system is provided in which PCI CONFIG.sub.-- ADDRESS and CONFIG.sub.-- DATA conventions are maintained in a large computer system and each PCI Host Bridge (PHB) CONFIG.sub.-- ADDRESS register and each PHB CONFIG.sub.-- DATA register have separate and system-unique addresses. In one example, the operating system provides a service to translate the device driver's configuration operation to a particular bus and device in the system, to a particular CONFIG.sub.-- ADDRESS or CONFIG.sub.-- DATA register of the PHB which has that device under it. By using this method, the hierarchical system can use architecture-independent routing of addresses down to the PHB that contains the appropriate CONFIG.sub.-- ADDRESS and CONFIG.sub.-- DATA registers.

Abstract: A method and system for supporting multiple Peripheral Component Interconnect (PCI) local buses through a single PCI host bridge having multiple PCI interfaces within a data-processing system are disclosed. In accordance with the method and system of the present invention, a processor and a system memory are connected to a system bus. First and second PCI local buses are connected to the system bus through a PCI host bridge. The first and second PCI local buses have sets of in-line electronic switches, dividing the PCI local buses into PCI local bus segments supporting a plurality of PCI peripheral component slots. The sets of in-line electronic switches are open and closed in accordance with bus control logic within the PCI host bridge allowing up to fourteen PCI peripheral component slots to have access through a single PCI host bridge to the system bus.

Abstract: A method of registering a newly added peripheral device with a computer system by responding with a status message from the device to a bus of the computer system, in response to an access attempt, and within a predetermined time period from the deasserting of the reset signal applied to device, so as to avoid stalling and thereby avoid the need to reboot the system in order to initialize the new peripheral device with the operating system. The device may be allowed to initially send a retry response, provided the response occurs during an initial latency period which is less than the predetermined time period. The invention also enables the peripheral device to respond to non-configuration cycles immediately following configuration completion. Internal logic of the peripheral device can be initialized after responding with the status message.

Abstract: A methodology and implementing system 101 are provided in which a PCI bus is enhanced to operate at a plurality of data transfer speeds, including for example, 133 MHz in order to accommodate subsystem boards operating at higher frequencies, while at the same time allowing normal 66 MHz PCI clocking for devices designed to operate at the lower 66 MHz standard PCI speed. Master strobe MSTB 303, 403 and target strobe TSTB signals 309, 411 are generated in a handshaking methodology to determine if a master data transaction requesting device and a target data transaction device are designed to operate at the higher data transfer frequency. Higher frequency capable devices or boards are run at the increased frequency when both the requesting master and the selected target devices request the higher transfer rate, and standard devices or boards are run at the lower standard PCI frequency, while both master and target devices are coupled to and run from the same multi-speed PCI bus 125.

Abstract: A method and system for translating peer-to-peer access across multiple Peripheral Component Interconnect (PCI) host bridges within a data-processing system are disclosed. In accordance with the method and system of the present invention, a processor and a system memory are connected to a system bus. A first and at least a second PCI local buses are also connected to the system bus via a first PCI host bridge and a second PCI host bridge, respectively. The two PCI local buses have bus transaction protocols that are different from those of the system bus. At least one PCI device is connected to each of the two PCI local buses, and shares data with the processor and the system memory. In addition, each PCI device shares data with the other PCI device as peer-to-peer devices across multiple PCI host bridges.