公开(公告)号：US20060179226A1

公开(公告)日：2006-08-10

申请号：US11054450

申请日：2005-02-09

申请人： Guy Guthrie ,  Hugh Shen ,  William Starke ,  Derek Williams

发明人： Guy Guthrie ,  Hugh Shen ,  William Starke ,  Derek Williams

IPC分类号： G06F12/00

CPC分类号： G06F9/3834 ,  G06F9/30043 ,  G06F9/3824 ,  G06F12/0817

摘要： A system and method for re-ordering store operations from a processor core to a store queue. When a store queue receives a new processor-issued store operation from the processor core, a store queue controller allocates a new entry in the store queue. In response to allocating the new entry in the store queue, the store queue controller determines whether or not the new entry is dependent on at least one other valid entry in the store queue. In response to determining the new entry is dependent on at least one other valid entry in the store queue, the store queue controller inhibits requesting of the new entry to the RC dispatch logic until each valid entry on which the new entry is dependent has been successfully dispatched to an RC machine by the RC dispatch logic.

摘要翻译： 一种用于重新排序从处理器核到存储队列的存储操作的系统和方法。 当存储队列从处理器核心接收到新的处理器发出的存储操作时，存储队列控制器在存储队列中分配新的条目。 响应于在商店队列中分配新条目，商店队列控制器确定新条目是否依赖于商店队列中的至少一个其他有效条目。 响应于确定新条目取决于存储队列中的至少一个其他有效条目，存储队列控制器禁止向RC调度逻辑请求新条目，直到新条目依赖于其上的每个有效条目已经成功 通过RC调度逻辑调度到RC机器。

公开(公告)号：US20060085604A1

公开(公告)日：2006-04-20

申请号：US10965144

申请日：2004-10-14

申请人： Guy Guthrie ,  Sheldon Levenstein ,  William Starke ,  Derek Williams

发明人： Guy Guthrie ,  Sheldon Levenstein ,  William Starke ,  Derek Williams

IPC分类号： G06F12/00

CPC分类号： G06F12/0811 ,  G06F9/3004 ,  G06F9/30072 ,  G06F9/30087 ,  G06F9/3834 ,  G06F9/3851 ,  G06F12/0855

摘要： A processing unit for a multiprocessor data processing system includes a processor core including a store-through upper level cache, an instruction sequencing unit that fetches instructions for execution, a data register, and at least one instruction execution unit coupled to the instruction sequencing unit that concurrently executes multiple threads of instructions. The processor core, responsive to the at least one instruction execution unit executing a load-reserve instruction in a first thread that binds to a load target address in the store-through upper level cache during a reservation hazard window associated with a conflicting store-conditional operation of a second thread, causes a subsequent store-conditional operation of the first thread to a store target address matching the load target address to fail if the store-conditional operation of the second thread succeeds.

摘要翻译： 一种用于多处理器数据处理系统的处理单元，包括处理器核心，该处理器核心包括通过存储的上级高速缓存，指令执行指令排序单元，数据寄存器以及耦合到指令排序单元的至少一个指令执行单元， 同时执行多个指令线程。 所述处理器核心响应于所述至少一个指令执行单元在与冲突存储条件相关联的预留危险窗口期间执行在所述存储通过上级高速缓存中的绑定到加载目标地址的第一线程中的加载保留指令 如果第二线程的存储条件操作成功，则第二线程的操作使得第一线程的后续存储条件操作到与加载目标地址匹配的存储目标地址失败。

公开(公告)号：US20080091918A1

公开(公告)日：2008-04-17

申请号：US11952479

申请日：2007-12-07

申请人： Ravi Arimilli ,  Robert Cargnoni ,  Derek Williams ,  Kenneth Wright

发明人： Ravi Arimilli ,  Robert Cargnoni ,  Derek Williams ,  Kenneth Wright

IPC分类号： G06F15/76 ,  G06F9/02

CPC分类号： H04L69/16 ,  G06F15/17 ,  H04L12/42 ,  H04L67/1002 ,  H04L69/161 ,  H04W28/14

摘要： A processor communication register (PCR) contained within a multiprocessor cluster system provides enhanced processor communication. The PCR stores information that is useful in pipelined or parallel multi-processing. Each processor cluster has exclusive rights to store to a sector within the PCR and has continuous access to read its contents. Each processor cluster updates its exclusive sector within the PCR, instantly allowing all of the other processors within the cluster network to see the change within the PCR data, and bypassing the cache subsystem. Efficiency is enhanced within the processor cluster network by providing processor communications to be immediately networked and transferred into all processors without momentarily restricting access to the information or forcing all the processors to be continually contending for the same cache line, and thereby overwhelming the interconnect and memory system with an endless stream of load, store and invalidate commands.

摘要翻译： 包含在多处理器集群系统内的处理器通信寄存器（PCR）提供增强的处理器通信。 PCR存储在流水线或并行多处理中有用的信息。 每个处理器集群具有存储到PCR中的扇区的独占权限，并且具有连续访问以读取其内容。 每个处理器集群在PCR中更新其独占部分，立即允许集群网络内的所有其他处理器查看PCR数据中的更改，并绕过缓存子系统。 处理器集群网络中的效率得到提高，通过提供处理器通信来立即联网并传输到所有处理器中，而不会立即限制对信息的访问，或迫使所有处理器持续竞争相同的高速缓存行，从而压倒互连和内存 系统具有无限的加载流，存储和无效命令。

公开(公告)号：US20060265553A1

公开(公告)日：2006-11-23

申请号：US11130907

申请日：2005-05-17

申请人： Ravi Arimilli ,  Derek Williams

发明人： Ravi Arimilli ,  Derek Williams

IPC分类号： G06F13/28

CPC分类号： G06F12/0831

摘要： In response to receiving an initialization operation from an associated processor core that indicates a target memory block to be initialized, a cache memory determines a coherency state of the target memory block. In response to a determination that the target memory block has a data-invalid coherency state with respect to the cache memory, the cache memory issues on a interconnect a corresponding initialization request indicating the target memory block. In response to the initialization request, the target memory block is initialized within a memory of the data processing system to an initialization value. The target memory block may thus be initialized without the cache memory holding a valid copy of the target memory block.

摘要翻译： 响应于从指示要初始化的目标存储器块的相关联的处理器核心接收到初始化操作，高速缓存存储器确定目标存储器块的一致性状态。 响应于目标存储器块相对于高速缓冲存储器具有数据无效一致性状态的确定，高速缓冲存储器在互连上发出指示目标存储器块的对应的初始化请求。 响应于初始化请求，将目标存储器块在数据处理系统的存储器内初始化为初始化值。 因此，可以初始化目标存储器块，而不使高速缓冲存储器保持目标存储器块的有效副本。

公开(公告)号：US20080109816A1

公开(公告)日：2008-05-08

申请号：US11971959

申请日：2008-01-10

申请人： RAVI ARIMILLI ,  Robert Cargnoni ,  Derek Williams ,  Kenneth Wright

发明人： RAVI ARIMILLI ,  Robert Cargnoni ,  Derek Williams ,  Kenneth Wright

IPC分类号： G06F9/50

CPC分类号： G06F9/30101

摘要： A processor communication register (PCR) contained in each processor within a multiprocessor system provides enhanced processor communication. Each PCR stores identical processor communication information that is useful in pipelined or parallel multi-processing. Each processor has exclusive rights to store to a sector within each PCR and has continuous access to read the contents of its own PCR. Each processor updates its exclusive sector within all of the PCRs, instantly allowing all of the other processors to see the change within the PCR data, and bypassing the cache subsystem. Efficiency is enhanced within the multiprocessor system by providing processor communications to be immediately transferred into all processors without momentarily restricting access to the information or forcing all the processors to be continually contending for the same cache line, and thereby overwhelming the interconnect and memory system with an endless stream of load, store and invalidate commands.

摘要翻译： 包含在多处理器系统内的每个处理器中的处理器通信寄存器（PCR）提供增强的处理器通信。 每个PCR存储在流水线或并行多处理中有用的相同的处理器通信信息。 每个处理器具有存储在每个PCR内的扇区的专有权利，并且具有连续访问以读取其自己的PCR的内容。 每个处理器在所有PCR中更新其独占扇区，立即允许所有其他处理器查看PCR数据中的更改，并绕过缓存子系统。 通过提供处理器通信以立即转移到所有处理器中而不会立即限制对信息的访问或迫使所有处理器连续地竞争相同的高速缓存行，从而将互连和存储系统压倒在一起，从而在多处理器系统中提高效率 无限流的加载，存储和无效命令。

公开(公告)号：US20050251622A1

公开(公告)日：2005-11-10

申请号：US10825188

申请日：2004-04-15

申请人： Ravi Arimilli ,  Robert Cargnoni ,  Hugh Shen ,  Derek Williams

发明人： Ravi Arimilli ,  Robert Cargnoni ,  Hugh Shen ,  Derek Williams

IPC分类号： G06F9/312 ,  G06F9/38 ,  G06F12/00 ,  G06F12/08

CPC分类号： G06F12/0897 ,  G06F9/30043 ,  G06F9/3824 ,  G06F9/3834 ,  G06F12/0804

摘要： A method and processor system that substantially enhances the store gathering capabilities of a store queue entry to enable gathering of a maximum number of proximate-in-time store operations before the entry is selected for dispatch. A counter is provided for each entry to track a time since a last gather to the entry. When a new gather does not occur before the counter reaches a threshold saturation point, the entry is signaled ready for dispatch. By defining an optimum threshold saturation point before the counter expires, sufficient time is provided for the entry to gather a proximate-in-time store operation. The entry may be deemed eligible for selection when certain conditions occur, including the entry becoming full, issuance of a barrier operation, and saturation of the counter. The use of the counter increases the ability of a store queue entry to complete gathering of enough store operations to update an entire cache line before that entry is dispatched to an RC machine.

摘要翻译： 一种方法和处理器系统，其基本上增强了存储队列条目的存储收集能力，以便能够在该条目被选择用于发送之前收集最大数量的接近时间存储操作。 为每个条目提供一个计数器，以跟踪从上次收集到条目的时间。 当计数器达到阈值饱和点之前没有发生新的聚合时，该信号将被发出准备就绪。 通过在计数器到期之前定义最佳阈值饱和点，为入口提供足够的时间来收集即时存储操作。 当某些条件发生时，条目可能被视为有资格进行选择，包括条目变满，发出屏障操作和计数器的饱和。 计数器的使用增加了存储队列条目完成收集足够的存储操作以在将该条目分派到RC机器之前更新整个高速缓存行的能力。

公开(公告)号：US20070088926A1

公开(公告)日：2007-04-19

申请号：US11250022

申请日：2005-10-13

申请人： James Fields ,  Guy Guthrie ,  John Hollaway ,  Derek Williams

发明人： James Fields ,  Guy Guthrie ,  John Hollaway ,  Derek Williams

IPC分类号： G06F12/14

CPC分类号： G06F12/0831 ,  G06F12/0815 ,  G06F12/0824

摘要： A data processing system includes a plurality of requestors and a memory controller for a system memory. In response to receiving from the requestor a read-type request targeting a memory block in the system memory, the memory controller protects the memory block from modification, and in response to an indication that the memory controller is responsible for servicing the read-type request, the memory controller transmits the memory block to the requestor. Prior to receipt of the memory block by the requestor, the memory controller ends protection of the memory block from modification, and the requestor begins protection of the memory block from modification. In response to receipt of the memory block, the requestor ends its protection of the memory block from modification.

摘要翻译： 数据处理系统包括多个请求者和用于系统存储器的存储器控​​制器。 响应于从请求者接收到针对系统存储器中的存储器块的读取型请求，存储器控制器保护存储器块免受修改，并且响应于存储器控制器负责维护读取类型请求的指示 存储器控制器将该存储器块发送给请求者。 在请求者接收到存储器块之前，存储器控制器结束对存储器块的保护而不被修改，并且请求者开始保护存储器块免受修改。 响应于存储器块的接收，请求者结束其对存储器块的保护以免修改。

公开(公告)号：US20070033345A1

公开(公告)日：2007-02-08

申请号：US11195021

申请日：2005-08-02

申请人： Guy Guthrie ,  Derek Williams

发明人： Guy Guthrie ,  Derek Williams

IPC分类号： G06F12/00

CPC分类号： G06F12/0811 ,  G06F9/3004 ,  G06F9/30072 ,  G06F9/30087 ,  G06F9/3834 ,  G06F9/3851

摘要： A processing unit for a multiprocessor data processing system includes a processor core and a lower level cache including a reservation logic that records reservations of the processor core. The reservation logic passes or fails store-conditional operations received from the processor core based upon whether the processor core has reservations for target store addresses of the store-conditional operations. The processor core includes a store-through upper level cache, a reservation register, and sequencer logic that, by reference to the reservation register, fails a store-conditional operation without communication with said reservation logic.

摘要翻译： 一种用于多处理器数据处理系统的处理单元包括处理器核心和包含记录处理器核心预约的预约逻辑的下级高速缓存。 基于处理器核心是否具有对存储条件操作的目标存储地址的预留，预留逻辑通过或失败从处理器核心接收的存储条件操作。 处理器核心包括通过存储的上级缓存，预约寄存器和定序器逻辑，其通过参考预约寄存器而失败存储条件操作，而不与所述预留逻辑通信。

公开(公告)号：US20060090035A1

公开(公告)日：2006-04-27

申请号：US10970437

申请日：2004-10-21

申请人： Guy Guthrie ,  Hugh Shen ,  Derek Williams

发明人： Guy Guthrie ,  Hugh Shen ,  Derek Williams

IPC分类号： G06F12/00

CPC分类号： G06F9/30087 ,  G06F9/3004 ,  G06F9/3824 ,  G06F9/3834 ,  G06F9/3851

摘要： A method, system, and processor chip design for reducing the latency between completing a LARX operation and receiving the associated STCX operation to complete the update to the cache line. Each entry of the store queue of the issuing processor is provided an additional tracking bit (priority bit). The priority bit is set whenever a STCX operation is placed within the entry. During selection of an entry for dispatch by the arbitration logic, the arbitration logic scans the value of the priority bits of each eligible entry. An entry with the priority bit set is given priority in the selection process within architectural rules. That entry is then selected for dispatch as early as is possible within the established rules.

公开(公告)号：US20070168639A1

公开(公告)日：2007-07-19

申请号：US11333607

申请日：2006-01-17

申请人： John McCalpin ,  William Starke ,  Jeffrey Stuecheli ,  Derek Williams

发明人： John McCalpin ,  William Starke ,  Jeffrey Stuecheli ,  Derek Williams

IPC分类号： G06F12/00

CPC分类号： G06F12/0831 ,  G06F12/0833 ,  G06F12/10 ,  G06F2212/507

摘要： A data processing system includes at least first and second coherency domains coupled by an interconnect fabric. A memory coupled to the interconnect fabric includes an address translation table having a translation table entry utilized to translate virtual memory addresses to real memory addresses. The translation table entry also includes scope information for broadcast operations targeting addresses within a memory region associated with the translation table entry. Scope prediction logic within the first coherency domain predictively selects a scope of broadcast of an operation on an interconnect fabric of the data processing system by reference to the scope information within the address translation table entry.

摘要翻译： 数据处理系统至少包括由互连结构耦合的第一和第二相干域。 耦合到互连结构的存储器包括具有用于将虚拟存储器地址转换为实际存储器地址的转换表项的地址转换表。 转换表条目还包括针对与转换表条目相关联的存储器区域内的地址的广播操作的范围信息。 第一相干域内的范围预测逻辑通过参考地址转换表条目中的范围信息预测性地选择数据处理系统的互连结构上的操作的广播范围。