Abstract: An aliasing technique allows transparently connecting multiple interconnects across a shared cross-connect interconnect, allowing devices on one interconnect to communicate with devices on another interconnect as if both interconnects were connected by a single interconnect bridge. Each interconnect appears to the cross-connect interconnect as a device on the cross-connect interconnect. Transactions between devices on different interconnects are aliased by a routing engine connected to the cross-connect interconnect for transmittal across the cross-connect interconnect and are invisible to other transactions on the cross-connect interconnect. Transactions between devices on the same interconnect are invisible to other interconnects. Cache coherent requests are supported by the use of additional attribute bits.

Abstract: A distributed power management technique allows controlling power states of devices separated from a power management controller, such as a processor, by an interconnect. The power management controller inserts power state information into an interconnect transaction. An interconnect connected device then extracts the power state information and modifies the power state of the device responsive to the power state information. The power state information can be extracted by a processor that then controls the power state of another device responsive to the power state information.

Abstract: Interconnect logic performs a transaction on an interconnect. The transaction may include multiple phases and the interconnect logic may include a counter state machine coupled to an interconnect state machine. The counter state machine may assert a signal to the interconnect state machine that may cause the interconnect state machine to prolong one or more phases of the transaction.

Abstract: Distributed peer-to-peer transactions are defined on an interconnect bus of a computer system according to an interconnect protocol. The transactions contain a completer device attribute data and a self-defining payload data. The transaction is identified as a peer-to-peer transaction by a command or an attribute data in the transaction. The transaction can be routed across a hierarchy of interconnect bus segments using the completer device address data. A handle can be used by an operating system of the computer system to indicate permission for the peer-to-peer transaction. Address information in a completer device address space can be provided within the peer-to-peer transaction or by a completer device driver for use by the completer device in processing the peer-to-peer transaction.

Abstract: An error correction code mechanism for the extensions to the peripheral component interconnect bus system (PCI-X) used in computer systems is fully backward compatible with the full PCI protocol. The error correction code check-bits can be inserted to provide error correction capability for the header address and attribute phases, as well as for burst and DWORD transaction data phases. The error correction code check-bits are inserted into unused attribute, clock phase, reserved, or reserved drive high portions of the AD and/or C/BE# lanes of the PCI-X phases.

Abstract: A target device in a computer bus system allocates resources by selecting a priority requester for allocation of scarce resources. In a non-bus arbiter configuration, the first initiator device to receive a retry response to a transaction request after the resources are exhausted is designated as a priority requester. In a bus arbiter configuration, the priority requester is chosen on a round-robin basis from initiator devices that received a retry response to the initiator's most recent transaction request. If only one resource is available when an initiator sends a transaction request, the initiator receives a retry response unless the initiator is the priority requester.

Abstract: An isochronous channel is configured on an interconnect bus between a first device and a second device. A first device requests an isochronous channel, required bandwidth, and a required service window size. If a service window of the required size at the required bandwidth is available, an isochronous bus controller sends the request to the second device. If the second device has a service window of the required size at the required, it accepts the isochronous channel request. The isochronous bus controller can be a collection of isochronous controllers, each controlling a subset of the interconnect bus. The isochronous bus controller then allocates bandwidth to the first device, notifying the first device to begin generating isochronous transactions, controlling access to the bus to ensure the first device does not exceed the bandwidth allocation. Further, the isochronous bus controller terminates the isochronous channel, if the first device stops sending isochronous transactions.

Abstract: As disclosed herein, an interface for a device adapted to couple to an interconnect may comprise decode and error check logic and a plurality of decode logic units. The decode and error check logic may receive error check bits and a target address from the interconnect and may determine whether the target address was received in error. At least one of the decode logic units also may receive the error check bits and correct the target address using the error check bits in parallel with the decode and error check logic determining whether the target address was received in error.

Abstract: Support for indicating and controlling transaction priority on a PCI-X bus. Embodiments of the invention provide indicia that can be set to communicate to PCI-X-to-PCI-X bridges and Completer that a transaction should be handled specially and scheduled ahead of any other transaction not having their corresponding indicia set. A special handling instruction allows the priority transaction to be scheduled first or early. The indicia are implemented by setting a bit(s) in an unused portion of a PCI-X attribute field, or multiplexed with a used portion, to schedule the associated transaction as the priority transaction over other transactions that do not have their corresponding bit set. The present invention can be used for interrupt messaging, audio streams, video streams, isochronous transactions, or for high performance, low bandwidth control structures used for communication in a multiprocessor architecture across PCI-X.

Abstract: Interconnect logic performs a transaction on an interconnect. The transaction may include multiple phases and the interconnect logic may include a counter state machine coupled to an interconnect state machine. The counter state machine may assert a signal to the interconnect state machine that may cause the interconnect state machine to prolong one or more phases of the transaction.

Abstract: A cyclic redundancy check (CRC) mechanism for the extensions (PCI-X) to the Peripheral Component Interconnect (PCI) bus system used in computer systems is fully backward compatible with the full PCI-X protocol. CRC check-bits are inserted to provide error detection capability for the header address and attribute phases, and for burst and DWORD transaction data phases. The CRC check-bits are inserted into unused attribute or clock (or target response) phases, or into reserved or reserved drive high portions (bits) of the AD, C/BE#, or into the parity lanes of the PCI-X phases.

Abstract: An aliasing technique allows transparently connecting multiple interconnects across a shared cross-connect interconnect, allowing devices on one interconnect to communicate with devices on another interconnect as if both interconnects were connected by a single interconnect bridge. Each interconnect appears to the cross-connect interconnect as a device on the cross-connect interconnect. Transactions between devices on different interconnects are aliased by a routing engine connected to the cross-connect interconnect for transmittal across the cross-connect interconnect and are invisible to other transactions on the cross-connect interconnect. Transactions between devices on the same interconnect are invisible to other interconnects. Cache coherent requests are supported by the use of additional attribute bits.

Abstract: A distributed power management technique allows controlling power states of devices separated from a power management controller, such as a processor, by an interconnect. The power management controller inserts power state information into an interconnect transaction. An interconnect connected device then extracts the power state information and modifies the power state of the device responsive to the power state information. The power state information can be extracted by a processor that then controls the power state of another device responsive to the power state information.

Abstract: A switching technique allows multiple interconnect bus devices to be connected to a single bus segment, even if the interconnect bus protocol only allows a one of the interconnect devices to be connected at any time. Each of the interconnect devices is connected to the interconnect bus segment with a switch, such that the interconnect device is electrically isolated from the interconnect bus segment when the switch is open. An interconnect sourcing agent connected to the interconnect bus segment controls the switches, closing the switch for one of the interconnect devices when a transaction is destined for that interconnect device, opening all of the other switches so that only one device is connected to the bus at any time.

Abstract: An interconnect switch provides full PCI compatibility while increasing performance via concurrency. The switch contains a primary bridge on a primary port. Secondary ports of the switch can be connected to secondary bridges and end devices. The switch can shadow registers associated with the secondary bridges. A transaction with a target address behind a secondary bridge is directly routed to the secondary port associated with the secondary bridge, using the shadowed registers. A transaction with a target address not behind a secondary bridge is routed to each of the other secondary ports. The transaction can be broadcast to all of the non-bridge secondary ports or can be routed successively to each of the non-bridge secondary ports until accepted. A tuning process can use positive acknowledgment of a transaction by an end device connected to a secondary port to directly route similar transactions to the same secondary port.

Abstract: An error correction code mechanism for the extensions to the peripheral component interconnect bus system (PCI-X) used in computer systems is fully backward compatible with the full PCI protocol. The error correction code check-bits can be inserted to provide error correction capability for the header address and attribute phases, as well as for burst and DWORD transaction data phases. The error correction code check-bits are inserted into unused attribute, clock phase, reserved, or reserved drive high portions of the AD and/or C/BE# lanes of the PCI-X phases.

Abstract: Distributed peer-to-peer transactions are defined on an interconnect bus of a computer system according to an interconnect protocol. The transactions contain a completer device attribute data and a self-defining payload data. The transaction is identified as a peer-to-peer transaction by a command or an attribute data in the transaction. The transaction can be routed across a hierarchy of interconnect bus segments using the completer device address data. A handle can be used by an operating system of the computer system to indicate permission for the peer-to-peer transaction. Address information in a completer device address space can be provided within the peer-to-peer transaction or by a completer device driver for use by the completer device in processing the peer-to-peer transaction.

Abstract: A target device in a computer bus system allocates resources by selecting a priority requester for allocation of scarce resources. In a non-bus arbiter configuration, the first initiator device to receive a retry response to a transaction request after the resources are exhausted is designated as a priority requester. In a bus arbiter configuration, the priority requester is chosen on a round-robin basis from initiator devices that received a retry response to the initiator's most recent transaction request. If only one resource is available when an initiator sends a transaction request, the initiator receives a retry response unless the initiator is the priority requester.

Abstract: An isochronous channel is configured on an interconnect bus between a first device and a second device. A first device requests an isochronous channel, required bandwidth, and a required service window size. If a service window of the required size at the required bandwidth is available, an isochronous bus controller sends the request to the second device. If the second device has a service window of the required size at the required, it accepts the isochronous channel request. The isochronous bus controller can be a collection of isochronous controllers, each controlling a subset of the interconnect bus. The isochronous bus controller then allocates bandwidth to the first device, notifying the first device to begin generating isochronous transactions, controlling access to the bus to ensure the first device does not exceed the bandwidth allocation. Further, the isochronous bus controller terminates the isochronous channel, if the first device stops sending isochronous transactions.

Abstract: Support for indicating and controlling transaction priority on a PCI-X bus. Embodiments of the invention provide indicia that can be set to communicate to PCI-X-to-PCI-X bridges and Completer that a transaction should be handled specially and scheduled ahead of any other transaction not having their corresponding indicia set. A special handling instruction allows the priority transaction to be scheduled first or early. The indicia are implemented by setting a bit(s) in an unused portion of a PCI-X attribute field, or multiplexed with a used portion, to schedule the associated transaction as the priority transaction over other transactions that do not have their corresponding bit set. The present invention can be used for interrupt messaging, audio streams, video streams, isochronous transactions, or for high performance, low bandwidth control structures used for communication in a multiprocessor architecture across PCI-X.