PatentDe  


Dokumentenidentifikation EP1390842 17.01.2008
EP-Veröffentlichungsnummer 0001390842
Titel MULTITHREADPROZESSOR MIT REGISTERZUWEISUNG BASIERT AUF EINEM ANZAHL VON AKTIVEN THREADS
Anmelder Intel Corporation, Santa Clara, Calif., US
Erfinder ROSENBLUTH, Mark, Uxbridge, MA 01569, US;
WOLRICH, Gilbert, Framingham, MA 01701, US;
BERNSTEIN, Debra, Sudbury, MA 01776, US
Vertreter derzeit kein Vertreter bestellt
DE-Aktenzeichen 60223917
Vertragsstaaten AT, BE, BG, CH, CY, CZ, DE, DK, EE, ES, FI, FR, GB, GR, IE, IT, LI, LU, MC, NL, PT, SE, SK, TR
Sprache des Dokument EN
EP-Anmeldetag 27.08.2002
EP-Aktenzeichen 027687276
WO-Anmeldetag 27.08.2002
PCT-Aktenzeichen PCT/US02/27273
WO-Veröffentlichungsnummer 2003019358
WO-Veröffentlichungsdatum 06.03.2003
EP-Offenlegungsdatum 25.02.2004
EP date of grant 05.12.2007
Veröffentlichungstag im Patentblatt 17.01.2008
IPC-Hauptklasse G06F 9/38(2006.01)A, F, I, 20051017, B, H, EP
IPC-Nebenklasse G06F 9/46(2006.01)A, L, I, 20051017, B, H, EP   

Beschreibung[en]
CROSS REFERENCE TO RELATED APPLICATIONS

This application claims priority from U.S. Provisional Patent Application Ser. No. 60/315,144 (Attorney Docket No. 10559-579P01), filed August 27, 2001.

BACKGROUND

Typically, hardware implementations of multithreaded microprocessors provide for use by each thread a fixed number of resources, such as registers, program counters, and so forth.

WO 01/41530 discloses a multithreaded processor for executing multiple threads in which a state machine that outputs multiple configuration bits, where each bit represents an associated threads execution activity. Functional units are allocated as each thread's respective bit indicates that that thread as active or inactive.

Depending on the amount of parallelism in an application program executing on the microprocessor, some of the threads may not be used. Consequently, the resources of the unused threads and, more specifically, the power and silicon area consumed by those resources, are wasted.

STATEMENT OF INVENTION

An aspect of the invention provides a method of allocating resources in a multithreaded processor as set forth in claim 1. Another aspect of the invention provides a processor as set forth in claim 8.

DESCRIPTION OF DRAWINGS

  • FIG. 1 shows a block diagram of a communication system employing a processor having multithreaded microengines to support multiple threads of execution.
  • FIG. 2 shows a block diagram of the microengine (of FIG. 1).
  • FIG. 3 shows a microengine Control and Status Register (CSR) used to select a number of "in use" threads.
  • FIG. 4 shows a schematic diagram of a dual-bank implementation of a General Purpose Registers (GPR) file (of the microengine of FIG. 2) that uses a selected number of "in use" threads to allocate registers to threads.
  • FIG. 5 shows a table of thread GPR allocations for eight "in use" threads and four "in use" threads.
  • FIGS. 6A and 6B show the partition of registers in the GPR file in accordance with the thread GPR allocations for eight "in use" threads and four "in use" threads, respectively.

DETAILED DESCRIPTION

Referring to FIG. 1, a communication system 10 includes a processor 12 coupled to one or more I/O devices, for example, network devices 14 and 16, as well as a memory system 18. The processor 12 is multi-threaded processor and, as such, is especially useful for tasks that can be broken into parallel subtasks or functions. In one embodiment, as shown in the figure, the processor 12 includes multiple microengines 20, each with multiple hardware controlled program threads 22 that can be simultaneously active and independently work on a task. In the example shown, there are "n" microengines 20, and each of the microengines 20 is capable of processing multiple program threads 22, as will be described more fully below. In the described embodiment, the maximum number "N" of context threads supported is eight, but other maximum amount could be provided. Preferably, each of the microengines 20 is connected to and can communicate with adjacent microengines.

The processor 12 also includes a processor 24 that assists in loading microcode control for other resources of the processor 12 and performs other general-purpose computer type functions such as handling protocols and exceptions. In network processing applications, the processor 24 can also provide support for higher layer network processing tasks that cannot be handled by the microengines 20. In one embodiment, the processor 24 is a StrongARM (ARM is a trademark of ARM Limited, United Kingdom) core based architecture. The processor (or core) 24 has an operating system through which the processor 24 can call functions to operate on the microengines 20. The processor 24 can use any supported operating system, preferably a real-time operating system. Other processor architectures may be used.

The microengines 20 each operate with shared resources including the memory system 18, a PCI bus interface 26, an I/O interface 28, a hash unit 30 and a scratchpad memory 32. The PCI bus interface 26 provides an interface to a PCI bus (not shown). The I/O interface 28 is responsible for controlling and interfacing the processor 12 to the network devices 14, 16. The memory system 18 includes a Dynamic Random Access Memory (DRAM) 34, which is accessed using a DRAM controller 36 and a Static Random Access Memory (SRAM) 38, which is accessed using an SRAM controller 40. Although not shown, the processor 12 also would include a nonvolatile memory to support boot operations. The DRAM 34 and DRAM controller 36 are typically used for processing large volumes of data, e.g., processing of payloads from network packets. In a networking implementation, the SRAM 38 and SRAM controller 40 are used for low latency, fast access tasks, e.g., accessing look-up tables, memory for the processor 24, and so forth. The microengines 20 can execute memory reference instructions to either the DRAM controller 36 or the SRAM controller 40.

The devices 14 and 16 can be any network devices capable of transmitting and/or receiving network traffic data, such as framing/MAC devices, e.g., for connecting to 10/100BaseT Ethernet, Gigabit Ethernet, ATM or other types of networks, or devices for connecting to a switch fabric. For example, in one arrangement, the network device 14 could be an Ethernet MAC device (connected to an Ethernet network, not shown) that transmits packet data to the processor 12 and device 16 could be a switch fabric device that receives processed packet data from processor 12 for transmission onto a switch fabric. In such an implementation, that is, when handling traffic to be sent to a switch fabric, the processor 12 would be acting as an ingress network processor. Alternatively, the processor 12 could operate as an egress network processor, handling traffic that is received from a switch fabric (via device 16) and destined for another network device such as network device 14, or network coupled to such device. Although the processor 12 can operate in a standalone mode, supporting both traffic directions, it will be understood that, to achieve higher performance, it may be desirable to use two dedicated processors, one as an ingress processor and the other as an egress processor. The two dedicated processors would each be coupled to the devices 14 and 16. In addition, each network device 14, 16 can include a plurality of ports to be serviced by the processor 12. The I/O interface 28 therefore supports one or more types of interfaces, such as an interface for packet and cell transfer between a PHY device and a higher protocol layer (e.g., link layer), or an interface between a traffic manager and a switch fabric for Asynchronous Transfer Mode (ATM), Internet Protocol (IP), Ethernet, and similar data communications applications. The I/O interface 28 includes separate receive and transmit blocks, each being separately configurable for a particular interface supported by the processor 12.

Other devices, such as a host computer and/or PCI peripherals (not shown), which may be coupled to a PCI bus controlled by the PC interface 26 are also serviced by the processor 12.

In general, as a network processor, the processor 12 can interface to any type of communication device or interface that receives/sends large amounts of data. The processor 12 functioning as a network processor could receive units of packet data from a network device like network device 14 and process those units of packet data in a parallel manner, as will be described. The unit of packet data could include an entire network packet (e.g., Ethernet packet) or a portion of such a packet, e.g., a cell or packet segment.

Each of the functional units of the processor 12 is coupled to an internal bus structure 42. Memory busses 44a, 44b couple the memory controllers 36 and 40, respectively, to respective memory units DRAM 34 and SRAM 38 of the memory system 18. The I/O Interface 28 is coupled to the devices 14 and 16 via separate I/O bus lines 46a and 46b, respectively.

Referring to FIG. 2, an exemplary one of the microengines 20 is shown. The microengine (ME) 20 includes a control unit 50 that includes a control store 51, control logic (or microcontroller) 52 and a context arbiter/event logic 53. The control store 51 is used to store a microprogram. The microprogram is loadable by the processor 24.

The microcontroller 52 includes an instruction decoder and program counter units for each of supported threads. The The context arbiter/event logic 53 receives messages (e.g., SRAM event response) from each one of the share resources, e.g., SRAM 38, DRAM 34, or processor core 24, and so forth. These messages provides information on whether a requested function has completed.

The context arbiter/event logic 53 has arbitration for the eight threads. In one embodiment, the arbitration is a round robin mechanism. However, other arbitration techniques, such as priority queuing or weighted fair queuing, could be used.

The microengine 20 also includes an execution datapath 54 and a general purpose register (GPR) file unit 56 that is coupled to the control unit 50. The datapath 54 includes several datapath elements, e.g., and as shown, a first datapath element 58, a second datapath element 59 and a third datapath element 60. The datapath elements can include, for example, an ALU and a multiplier. The GPR file unit 56 provides operands to the various datapath elements. The registers of the GPR file unit 56 are read and written exclusively under program control. GPRs, when used as a source in an instruction, supply operands to the datapath 54. When use as a destination in an instruction, they are written with the result of the datapath 54. The instruction specifies the register number of the specific GPRs that are selected for a source or destination. Opcode bits in the instruction provided by the control unit 50 select which datapath element is to perform the operation defined by the instruction.

The microengine 20 further includes a write transfer register file 62 and a read transfer register file 64. The write transfer register file 62 stores data to be written to a resource external to the microengine (for example, the DRAM memory or SRAM memory). The read transfer register file 64 is used for storing return data from a resource external to the microengine 20. Subsequent to or concurrent with the data arrival, event signals 65 from the respective shared resource, e.g., memory controllers 36, 40, or core 24, can be provided to alert the thread that requested the data that the data is available or has been sent. Both of the transfer register files 62, 64 are connected to the datapath 54, the GPR file unit 56, as well as the control unit 50.

Also included in the microengine 20 is a local memory 66. The local memory 66, which is addressed by registers 68a, 68b, also supplies operands to the datapath 54. The local memory 66 receives results from the datapath 54 as a destination. The microengine 20 also includes local control and status registers (CSRs) 70 for storing local interthread and global event signaling information, as well as other information, and a CRC unit 72, coupled to the transfer registers, which operates in parallel with the execution datapath 54 and performs CRC computations for ATM cells. The local CSRs 70 and the CRC unit 72 are coupled to the transfer registers, the datapath 54 and the GPR file unit 56.

In addition to providing an output to the write transfer unit 62, the datapath 54 can also provide an output to the GPR file 56 over line 80. Thus, each of the datapath elements can return a result value from an executed.

The functionality of the microengine threads 22 is determined by microcode loaded (via the core processor 24) for a particular user's application into each microengine's control store 51. For example, in one exemplary thread task assignment, one thread is assigned to serve as a receive scheduler thread and another as a transmit scheduler thread, a plurality of threads are configured as receive processing threads and transmit processing threads, and other thread task assignments include a transmit arbiter and one or more core communication threads. Once launched, a thread performs its function independently.

Referring to FIG. 3, the CSRs 70 include a context enable register ("CTX_Enable") 90, which includes an "in use" contexts field 92 to indicate a pre-selected number of threads or contexts in use. The "in use" contexts field 92 stores a single bit, which when cleared (X=0) indicates all of the 8 available threads are in use, and which when set (X=1) indicates that only a predefined number, e.g., 4, more specifically, threads 0, 2, 4 and 6, are in use.

As shown in FIG. 4, the GPRs of the GPR file unit 56 may be physically and logically contained in two banks, an A bank 56a and a B bank 56b. The GPRs in both banks include a data portion 100 and an address portion 102. Coupled to each register address path 102 is a multiplexor 104, which receives as inputs a thread number 106 and register number 108 (from the instruction) from the control unit 50. The output of the multiplexor 104, that is, the form of the "address" provided to the address path 102 to select one of the registers 109, is controlled by an enable signal 110. The state of the enable signal 110 is determined by the setting of the "In_Use" Contexts bit in the field 92 of the CTX_Enable register 90.

Conventionally, each thread has a fixed percentage of the registers allocated to it, for example, one-eighth for the case of eight threads supported. If some threads are not used, the registers dedicated for use by those unused threads go unused as well.

In contrast, the use of the multiplexor 104 controlled by "in use" contexts configuration information in the CTX_Enable CSR 90 enables a re-partitioning of the number of bits of active thread number/instruction (register number) bits in the register address and therefore a re-allocation of registers to threads. More specifically, when the bit in field 92 is equal to a "0", the number of "in use" threads is 8, and the enable 110 controls the multiplexor 104 to select all of the bits of the active thread number 106 and all but the most significant bit from the register number 108 specified by the current instruction. Conversely, when the bit in field 92 is set to a "1", the number of "in use" threads is reduced by half, and the number of registers available for allocation is redistributed so that the number of registers allocated per thread is doubled.

FIG. 5 shows the thread allocation for a register file of 32 registers. For 8 threads, thread numbers 0 through 7, each thread is allocated a total of four registers. For 4 threads, thread numbers 0, 2, 4 and 6, each thread is allocated a total of eight registers.

FIGS. 6A and 6B show a register file (single bank, for example, register file 56a) having 32 registers available for thread allocation and re-allocation among a maximum of eight supported threads. In an 8-thread configuration 120, that is, the case of eight threads in use, shown in FIG. 6A, each of the threads is allocated four registers. The multiplexor 104 selects all three bits of the binary representation of the thread number and all bits except the most significant bit (that is, selects two bits (bits 0 and 1)) of the binary representation of the register number from the instruction because the enable 110 is low. For a 4-thread configuration 122, that is, when enable 110 is high and thus four threads, as illustrated in FIG. 6B, each of the four threads is allocated eight registers. The multiplexor 104 selects all but the least significant bit (in this case, selects two bits, bits 1 and 2) of the binary representation of the thread number and selects all three bits (bits 0-2) of the binary representation of the register number from the instruction. Thus, the address into the register file is a concatenation of bits of the currently active thread number with bits of the register number from the instruction, and the contributing number of bits from each is determined by the setting of the In_Use contexts bit 92 in the CTX_Enable register 90 (from FIG. 3).

Thus, the GPRs are logically subdivided in equal regions such that each context has relative access to one of the regions. The number of regions is configured in the In_Use contexts field 92, and can be either 4 or 8. Thus, a context-relative register number is actually associated with multiple different physical registers. The actual register to be accessed is determined by the context making the access request, that is, the context number concatenated with the register number, in the manner described above. Context-relative addressing is a powerful feature that enables eight or four different threads to share the same code image, yet maintain separate data. Thus, instructions specify the context-relative address (register number). For eight active contexts, the instruction always specifies registers in the range of 0-3. For four active contexts, the instruction always specifies registers in the range of 0-7.

Referring back to the table shown in FIG. 4, the absolute GPR register number is the register number that is actually used by the register address path (decode logic) to access the specific context-relative register. For example, with 8 active contexts, context-relative thread 0 for context (or thread) 2 is 8.

The above thread GPR allocation scheme can be extended to different numbers of threads (based on multiples of 2) and registers, for example, re-allocating a total of 128 registers from among a maximum number of 8 "in use" threads (16 registers each) to 4 "in use" threads (32 registers each), or re-allocating a total of 128 registers from among a maximum number of 16 "in use" threads (8 registers each) to 8 "in use" threads (16 registers each) .

Other embodiments are within the scope of the following claims.


Anspruch[de]
Verfahren für die Zuweisung von Ressourcen in einem Multi-Thread-Prozessor (12) umfassend: Bereitstellung von Ressourcen (56, 109) zur Verwendung durch Ausführungs-Threads (22), die von dem Multi-Thread-Prozessor (12) unterstützt werden, wobei die Ressourcen (56, 109) Register (109) in einer Universalregisterdatei (56) umfassen; und Anwendung von Konfigurationsinformationen auf eine Auswahl der Ressourcen (56, 109), um die Ressourcen unter den aktiven der Ausführungs-Threads (22) zuzuweisen, dadurch gekennzeichnet, dass die Konfigurationsinformationen ein Konfigurations-Bit (92) umfassen, das, wenn es gelöscht ist, alle der unterstützten Ausführungs-Threads (22) als aktive Threads angibt, und wenn es gesetzt ist, einen Teil der unterstützten Ausführungs-Threads als aktive Threads angibt. Verfahren nach Anspruch 1, wobei die Konfigurationsinformationen Folgendes umfassen: ein Konfigurations-Bit, das, wenn es gelöscht ist, alle der unterstützten Ausführungs-Threads als aktive Threads angibt, und wenn es gesetzt ist, die Hälfte der unterstützten Ausführungs-Threads als aktive Threads angibt. Verfahren nach Anspruch 1, wobei sich das Konfigurations-Bit (92) in einem Steuerungs- und Statusregister (70) befindet. Verfahren nach Anspruch 1, wobei die Universalregisterdatei (56) einen Adressdekodierungsteil (102) und einen Multiplexer (104) umfasst, der mit dem Adressdekodierungsteil verbunden ist, wobei der Multiplexer eine Thread-Nummer (106) und eine Registernummer (108) als Eingaben erhält und Bits der Tread-Nummer und der Registernummer basierend auf den Konfigurationsinformationen auswählt, um eine Adresse zu bilden, die einem der Register (109) entspricht. Verfahren nach Anspruch 4, wobei die Konfigurationsinformationen die Auswahl von allen außer dem Bit mit der niedrigsten Wertigkeit (Least Significant Bit, LSB) der Thread-Nummer (106) und allen Bits der Registernummer (108) angibt. Verfahren nach Anspruch 4, wobei die Konfigurationsinformationen die Auswahl von allen außer dem Bit mit der höchsten Wertigkeit (Most Significant Bit, LSB) der Registernummer (108) und allen Bits der Thread-Nummer (106) angibt. Verfahren nach Anspruch 4, wobei die gewählten Bits der Registernummer (108) eine Thread-bezogene Registernummer bilden. Prozessor umfassend: Ressourcen (56, 109) zur Verwendung durch Ausführungs-Threads (22), die vom Prozessor (12) unterstützt werden, wobei die Ressourcen (56, 109) Register (109) in einer Universalregisterdatei (56) umfassen; und einen Ressourcen-Selektor (110) für den Erhalt von Konfigurationsinformationen und die Zuweisung der Ressourcen (56, 109) unter den aktiven Threads der Ausführungs-Threads (22) basierend auf den Konfigurationsinformationen, dadurch gekennzeichnet, dass die Konfigurationsinformationen ein Konfigurations-Bit (92) umfassen, das, wenn es gelöscht ist, alle der unterstützten Ausführungs-Threads (22) als aktive Threads angibt, und wenn es gesetzt ist, einen Teil der unterstützten Ausführungs-Threads als aktive Threads angibt. Prozessor nach Anspruch 8, wobei die Konfigurationsinformationen Folgendes umfassen: ein Konfigurations-Bit, das, wenn es gelöscht wird, alle der unterstützten Ausführungs-Threads als aktive Threads angibt, und wenn es gesetzt ist, die Hälfte der unterstützten Ausführungs-Threads als aktive Threads angibt. Prozessor nach Anspruch 8, wobei sich das Konfigurations-Bit (92) in einem Steuerungs- und Statusregister (70) befindet. Prozessor nach Anspruch 8, wobei die Universalregisterdatei (56) einen Adressdekodierungsteil (102) umfasst, und der Ressourcen-Selektor (110) einen Multiplexer (104) umfasst, der mit dem Adressdekodierungsteil verbunden ist, wobei der Multiplexer eine Thread-Nummer (106) und eine Registernummer (108) als Eingaben erhält und Bits der Tread-Nummer und der Registernummer basierend auf den Konfigurationsinformationen auswählt, um eine Adresse zu bilden, die einem der Register (109) entspricht. Prozessor nach Anspruch 11, wobei die Konfigurationsinformationen die Auswahl von allen außer dem Bit mit der niedrigsten Wertigkeit (Least Significant Bit, LSB) der Thread-Nummer (106) und allen Bits der Registernummer (108) angibt. Prozessor nach Anspruch 11, wobei die Konfigurationsinformationen die Auswahl von allen außer dem Bit mit der höchsten Wertigkeit (Most Significant Bit, LSB) der Registernummer (108) und allen Bits der Thread-Nummer (106) angibt. Prozessor nach Anspruch 11, wobei die gewählten Bits der Registernummer (108) eine Thread-bezogene Registernummer bilden. Ein Computer-lesbares Medium umfassend Computer-Code-Mittel, die angepasst sind, um das Verfahren nach einem der Ansprüche 1 bis 7 durchzuführen, wenn das Programm auf einem Prozessor (12) ausgeführt wird. Ein System umfassend eine oder mehrere Netzwerkvorrichtungen (14, 16) und einen Prozessor (12) gemäß einem der Ansprüche 8 bis 14.
Anspruch[en]
A method of allocating resources in a multithreaded processor (12) comprising: providing resources (56, 109) for use by execution threads (22) supported by the multithreaded processor (12), wherein the resources (56, 109) comprise registers (109) in a general purpose register file (56); and applying configuration information to a selection of the resources (56, 109) to allocate the resources among active ones of the execution threads (22) characterized in that the configuration information comprises a configuration bit (92) which, when cleared, indicates all of the supported execution threads (22) as the active ones and, when set, indicates a portion of the supported execution threads as the active ones. The method of claim 1 wherein the configuration information comprises: a configuration bit which when cleared indicates all of the supported execution threads as the active ones and when set indicates half of the supported execution threads as the active ones. The method of claim 1, wherein the configuration bit (92) resides in a control and status register (70). The method of claim 1 wherein the general purpose register file (56) includes an address decode portion (102) and a multiplexor (104) coupled to the address decode portion, the multiplexor to receive a thread number (106) and a register number (108) as inputs and to select bits of the thread number and the register number based on the configuration information to form an address corresponding to one of the registers (109). The method of claim 4 wherein the configuration information indicates selection of all but the least signification bit of the thread number (106) and all bits of the register number (108). The method of claim 4 wherein the configuration information indicates selection of all but the most significant bit of the register number (108) and all bits of the thread number (106). The method of claim 4 wherein the selected bits of the register number (108) form a thread-relative register number. A processor comprising: resources (56, 109) for use by execution threads (22) supported by the processor (12), wherein the resources (56, 109) comprise registers (109) in a general purpose register file (56); and a resource selector (110) to receive configuration information and to allocate the resources (56, 109) among active ones of the execution threads (22) based on the configuration information, characterized in that the configuration information comprises a configuration bit (92) which, when cleared, indicates all of the supported execution threads (22) as the active ones and, when set, indicates a portion of the supported execution threads (22) as the active ones. The processor of claim 8 wherein the configuration information comprises: a configuration bit which when cleared indicates all of the supported execution threads as the active ones and when set indicates half of the supported execution threads as the active ones. The processor of claim 8, wherein the configuration bit (92) resides in a control and status register (70). The processor of claim 8 wherein the general purpose register file (56) includes an address decode portion (102) and the resource selector (102) includes a multiplexor (104) coupled to the address decode portion, the multiplexor to receive a thread number (106) and a register number (108) as inputs and to select bits of the thread number and the register number based on the configuration information to form an address corresponding to one of the registers (109). The processor of claim 11 wherein the configuration information indicates selection of all but the least signification bit of the thread number (106) and all bits of the register number (108). The processor of claim 11 wherein the configuration information indicates selection of all but the most significant bit of the register number (108) and all bits of the thread number (106). The processor of claim 11 wherein the selected bits of the register number (108) form a thread-relative register number. A computer readable medium comprising computer code means adapted to perform the method of any of claims 1 to 7 when said program is run on a processor (12). A system comprising one or more network devices (14, 16) and a processor (12) in accordance with any of claims 8 to 14.
Anspruch[fr]
Procédé d'affectation de ressources dans un processeur à unités d'exécution multiples (12), comprenant : la mise en place de ressources (56, 109) pour l'utilisation par des unités d'exécution (22) supportées par le processeur à unités d'exécution multiples (12), dans lequel les ressources (56, 109) comprennent des registres (109) dans un fichier de registres d'utilisation générale (56) ; et l'application d'informations de configuration à une sélection de ressources (56, 109) pour affecter les ressources parmi les unités actives des unités d'exécution (22), caractérisé en ce que

les informations de configuration comprennent un bit de configuration (22) qui, lorsqu'il est remis à zéro, indique toutes les unités d'exécution supportées (22) comme étant les unités actives et, lorsqu'il est mis à un, indique une portion des unités d'exécution supportées comme étant les unités actives.
Procédé selon la revendication 1, dans lequel les informations de configuration comprennent : un bit de configuration qui, lorsqu'il est remis à zéro, indique toutes les unités d'exécution supportées comme étant les unités actives et lorsqu'il est mis à un, indique la moitié des unités d'exécution supportées comme étant les unités actives. Procédé selon la revendication 1, dans lequel le bit de configuration (92) réside dans un registre de commande et d'état (70). Procédé selon la revendication 1, dans lequel le fichier de registre d'utilisation générale (56) comprend une portion de décodage adresse (102) et un multiplexeur (104) couplé à la portion de décodage adresse, le multiplexeur recevant un numéro d'unité d'exécution (106) et un numéro de registre (108) en tant qu'entrées et devant sélectionner les bits du numéro d'unité d'exécution et le numéro de registre en fonction des informations de configuration pour former une adresse correspondant à l'un des registres (109). Procédé selon la revendication 4, dans lequel les informations de configuration indiquent la sélection de tous les bits sauf le bit de poids faible du numéro d'unité d'exécution (106) et tous les bits du numéro de registre (108). Procédé selon la revendication 4, dans lequel les informations de configuration indiquent la sélection de tous les bits sauf le bit de poids fort du numéro de registre (108) et tous les bits du numéro d'unité d'exécution (106). Procédé selon la revendication 4, dans lequel les bits sélectionnés du numéro de registre (108) forment un numéro de registre relatif à l'unité d'exécution. Processeur comprenant : des ressources (56, 109) pour l'utilisation par les unités d'exécution (22) supportées par le processeur (12), dans lequel les ressources (56, 109) comprennent des registres (109) dans un fichier de registre d'utilisation générale (56) ; et un sélecteur de ressource (110) pour recevoir des informations de configuration et pour affecter les ressources (56, 109) parmi les unités actives des unités d'exécution (22) en fonction des informations de configuration, caractérisé en ce que,

les informations de configuration comprennent un bit de configuration (92) qui, lorsqu'il est remis à zéro, indique toutes les unités d'exécution supportées (22) comme étant les unités actives et, lorsqu'il est mis à un, indique une portion des unités d'exécution supportées (22) comme étant les unités actives.
Processeur selon la revendication 8, dans lequel les informations de configuration comprennent : un bit de configuration qui, lorsqu'il a été remis à zéro, indique toutes les unités d'exécution supportées comme étant les unités actives et lorsqu'il a été mis à un, indique la moitié des unités d'exécution supportées comme étant les unités actives. Processeur selon la revendication 8, dans lequel le bit de configuration (92) réside dans un registre de commande et d'état (70). Processeur selon la revendication 8, dans lequel le fichier de registre d'utilisation générale (56) comprend une portion de décodage adresse (102) et le sélecteur de ressource (110) comprend un multiplexeur (104) couplé à la portion de décodage adresse, le multiplexeur recevant un numéro d'unité d'exécution (106) et un numéro de registre (108) en tant qu'entrées et devant sélectionner les bits du numéro d'unité d'exécution et le numéro de registre en fonction des informations de configuration pour former une adresse correspondant à l'un des registres (109). Processeur selon la revendication 11, dans lequel les informations de configuration indiquent le sélection de tous les bits sauf le bit de poids faible du numéro d'unité d'exécution (106) et tous les bits du numéro de registre (108). Processeur selon la revendication 11, dans lequel les informations de configuration indiquent la sélection de tous les bits sauf le bit de poids fort du numéro de registre (108) et tous les bits du numéro de l'unité d'exécution (106). Processeur selon la revendication 11, dans lequel les bits sélectionnés du numéro de registre (108) forment un numéro de registre relatif à l'unité d'exécution. Un support lisible par un ordinateur, comprenant des moyens de code d'un programme d'ordinateur adaptés pour mettre en oeuvre le procédé selon l'une quelconque des revendications 1 à 7 quand ledit programme est exécuté par un processeur (12). Un système comprenant un ou plusieurs périphériques de réseau (14, 16) et un processeur (12) selon l'une quelconque des revendications 8 à 14.






IPC
A Täglicher Lebensbedarf
B Arbeitsverfahren; Transportieren
C Chemie; Hüttenwesen
D Textilien; Papier
E Bauwesen; Erdbohren; Bergbau
F Maschinenbau; Beleuchtung; Heizung; Waffen; Sprengen
G Physik
H Elektrotechnik

Anmelder
Datum

Patentrecherche

Patent Zeichnungen (PDF)

Copyright © 2008 Patent-De Alle Rechte vorbehalten. eMail: info@patent-de.com