Abstract: A communication channel controller includes a queue, a memory map, and a scheduler. The queue to store a first memory transfer request received at the communication channel controller. The memory map stores information to identify a memory address range to be associated with a memory. The scheduler to compare a source address of the first memory transfer in the queue to the memory address range in the memory map to determine whether the source address of the first memory transfer request targets the memory, and in response allocate the first memory transfer request to a first communication channel of a plurality of communication channels in response to the first communication channel having all of its outstanding memory transactions to a common source address bank and source address page as a source address bank and a source address page of the first memory transfer request.

Abstract: A source processor can divide each packet of a data stream into multiple segments prior to communication of the packet, allowing a packet to be transmitted in smaller chunks. The source processor can process the segments for two or more packets for a given data stream concurrently, and provide appropriate context information in each segments header to facilitate in order transmission and reception of the packets represented by the individual segments. Similarly, a destination processor can receive the packet segments packets for an ordered data stream from a source processor, and can assign different contexts, based upon the context information in each segments header. When a last segment is received for a particular packet, the context for the particular packet is closed, and a descriptor for the packet is sent to a queue. The order in which the last segments of the packets are transmitted maintains order amongst the packets.

Abstract: A processor of a plurality of processors includes a processor core and a message manager. The message manager is in communication with the processor core. The message manager to receive a message from a second processor of the plurality of processors, to identify a classification rule for the message based on bits in a header of the message, and to create a queue identifier for the message using bits of a payload of the message, wherein the queue identifier is associated with a queue of the processor core.

Abstract: A method and circuit for a data processing system (200) provide a processor-based partitioned priority blocking mechanism by storing interrupt identifiers, partition identifiers, thread identifiers, and priority levels associated with accepted interrupt requests in special purpose registers (35-38) located at the processor core (26) to enable quick and efficient interrupt priority blocking on a partition basis.

Abstract: A method and circuit for a data processing system (20) provide a processor-based partitioned priority blocking mechanism by storing priority levels and associated partition information in special purpose registers (27-29) located at the processor core (26) to enable quick and efficient interrupt priority blocking on a partition basis.

Abstract: A method includes generating, by a first software process of the data processing system, a source partition descriptor for a DMA job which requires access to a first partition of a memory which is assigned to a second software process of the data processing system and not assigned to the first software process. The source partition descriptor comprises a partition identifier which identifies the first partition of the memory. The DMA unit receives the source partition descriptor and generates a destination partition descriptor for the DMA job. Generating the destination partition descriptor includes translating, by the DMA unit, the partition identifier to a buffer pool identifier which identifies a physical address within the first partition of the memory which is assigned to the second software process; and storing, by the DMA unit, the buffer pool identifier in the destination partition descriptor.

Abstract: A method and circuit for a data processing system provide a partitioned interrupt controller with an efficient deferral mechanism for processing partitioned interrupt requests by executing a control instruction to encode and store a delay command (e.g., DEFER or SUSPEND) in a data payload with a hardware-inserted partition attribute (LPID) for storage to a command register (25) at a physical address (PA) retrieved from a special purpose register (46) so that the partitioned interrupt controller (14) can determine if the delay command can be performed based on local access control information.

Abstract: A method and circuit for a data processing system provide virtualized instructions for accessing a partitioned device (e.g., 14, 61) by executing a control instruction (47, 48) to encode and store an access command (CMD) in a data payload with a hardware-inserted partition attribute (LPID) for storage to a command register (25) at a physical address (PA) retrieved from a special purpose register (46) so that the partitioned device (14, 61) can determine if the access command can be performed based on local access control information.

Abstract: A data processing system employs a scheduler to schedule pending memory access requests and a memory controller to service scheduled pending memory access requests. The memory access requests are asynchronously scheduled with respect to the clocking of the memory. The scheduler is operated using a clock signal with a frequency different from the frequency of the clock signal used to operate the memory controller. The clock signal used to clock the scheduler can have a lower frequency than the clock used by a memory controller. As a result, the scheduler is able to consider a greater number of pending memory access requests when selecting the next pending memory access request to be submitted to the memory for servicing and thus the resulting sequence of selected memory access requests is more likely to be optimized for memory access throughput.

Abstract: A DMA controller allocates space at a buffer to different DMA engines based on the length of time data segments have been stored at a buffer. This allocation ensures that DMA engines associated with a destination that is experiencing higher congestion will be assigned less buffer space than a destination that is experiencing lower congestion. Further, the DMA controller is able to adapt to changing congestion conditions at the transfer destinations.

Abstract: An integrated circuit device includes a processor core, and a controller. The processor core issues a command intended for a first thread of a plurality of threads. The controller initiates de-allocates hardware resources of the controller that are allocated to the first thread during a thread reset process for the first thread, returns a specified value to the processor core in response to the first command intended for the first thread during the thread reset process, drops responses intended for the first thread from other devices during the thread reset process, completes the thread reset process in response to a determination that all expected responses intended for the first thread have been either received or dropped, and continues to issue requests to other devices in response to commands from other threads of the plurality of threads and processing corresponding responses during the thread reset process.

Abstract: A data processing system employs an improved arbitration process in selecting pending memory access requests received from the one or more processor cores for servicing by the memory. The arbitration process uses memory timing and state information pertaining both to memory access requests already submitted to the memory for servicing and to the pending memory access requests which have not yet been selected for servicing by the memory. The memory timing and state information may be predicted memory timing and state information; that is, the component of the data processing system that implements the improved scheduling algorithm may not be able to determine the exact point in time at which a memory controller initiates a memory access for a corresponding memory access request and thus the component maintains information that estimates or otherwise predicts the particular state of the memory at any given time.

Abstract: In a data processing system having a plurality of resources and plurality of partitions, each partition including one or more resources of the plurality of resources, a method includes receiving an access request to a target resource of the plurality of resources; using a first set of transaction attributes of the access request to determine a partition identifier for the access request in which the partition identifier indicates a partition of the plurality of partitions which includes the target resource; using the partition identifier to determine access permissions for the partition indicated by the partition identifier; and based on the access permissions, determining whether or not the access request is permitted.

Abstract: A method includes generating, by a first software process of the data processing system, a source partition descriptor for a DMA job which requires access to a first partition of a memory which is assigned to a second software process of the data processing system and not assigned to the first software process. The source partition descriptor comprises a partition identifier which identifies the first partition of the memory. The DMA unit receives the source partition descriptor and generates a destination partition descriptor for the DMA job. Generating the destination partition descriptor includes translating, by the DMA unit, the partition identifier to a buffer pool identifier which identifies a physical address within the first partition of the memory which is assigned to the second software process; and storing, by the DMA unit, the buffer pool identifier in the destination partition descriptor.

Abstract: A method for controlling bandwidth in a direct memory access (DMA) unit of a computer processing system, the method comprising: assigning a DMA job to a selected DMA engine; starting a source timer; and issuing a request to read a next section of data for the DMA job. If a sufficient amount of the data was not obtained, allowing the DMA engine to wait until the source timer reaches a specified value before continuing to read additional data for the DMA job.

Abstract: A method for controlling bandwidth in a direct memory access (DMA) unit of a computer processing system, the method comprising: assigning a DMA job to a selected DMA engine; starting a source timer; and issuing a request to read a next section of data for the DMA job. If a sufficient amount of the data was not obtained, allowing the DMA engine to wait until the source timer reaches a specified value before continuing to read additional data for the DMA job.

Abstract: A method of multicast data transfer including accessing a source address to a source location of mapped memory which stores source data, accessing multiple destination addresses to corresponding destination locations of the mapped memory, and for each of at least one section of the source data, reading the section using the source address, storing the section into a local memory of a data transfer device, and writing the section from the local memory to each destination location in the mapped memory using the destination addresses. Separate source and destination attributes may be provided, so that the source and each destination may have different attributes for reading and storing data. The source and each destination may have any number of data buffers accessible by corresponding links provided in data structures supporting the data transfer. The source data may be divided into sections and handled section by section.

Abstract: A data processing system employs a scheduler to schedule pending memory access requests and a memory controller to service scheduled pending memory access requests. The memory access requests are asynchronously scheduled with respect to the clocking of the memory. The scheduler is operated using a clock signal with a frequency different from the frequency of the clock signal used to operate the memory controller. The clock signal used to clock the scheduler can have a lower frequency than the clock used by a memory controller. As a result, the scheduler is able to consider a greater number of pending memory access requests when selecting the next pending memory access request to be submitted to the memory for servicing and thus the resulting sequence of selected memory access requests is more likely to be optimized for memory access throughput.

Abstract: A data processing system employs an improved arbitration process in selecting pending memory access requests received from the one or more processor cores for servicing by the memory. The arbitration process uses memory timing and state information pertaining both to memory access requests already submitted to the memory for servicing and to the pending memory access requests which have not yet been selected for servicing by the memory. The memory timing and state information may be predicted memory timing and state information; that is, the component of the data processing system that implements the improved scheduling algorithm may not be able to determine the exact point in time at which a memory controller initiates a memory access for a corresponding memory access request and thus the component maintains information that estimates or otherwise predicts the particular state of the memory at any given time.

Abstract: In a data processing system having a plurality of resources and plurality of partitions, each partition including one or more resources of the plurality of resources, a method includes receiving an access request to a target resource of the plurality of resources; using a first set of transaction attributes of the access request to determine a partition identifier for the access request in which the partition identifier indicates a partition of the plurality of partitions which includes the target resource; using the partition identifier to determine access permissions for the partition indicated by the partition identifier; and based on the access permissions, determining whether or not the access request is permitted.