Methods and apparatus for cache intervention

[0001] The present invention relates in general to cache memory and, in particular, to methods and apparatus for cache intervention.

[0002] In an effort to increase computational power, many computing systems are turning to multi-processor systems. A multi-processor system typically includes a plurality of microprocessors, a plurality of associated caches, and a main memory. In an effort to reduce bus traffic to the main memory, many multi-processor systems use a write-back (as opposed to a write-through) policy. A write-back policy is a cache procedure whereby a microprocessor may locally modify data in its cache without updating the main memory until the cache data needs to be replaced. In order to maintain cache coherency in such a system, a cache coherency protocol may be used.

[0003] One problem with a write-back policy is sourcing a read request from one cache when another cache is holding the requested memory block in a modified state (i.e., the data is dirty). If the requesting cache is allowed to read the data from main memory, the value of the data will be incorrect. In order to solve this problem, some protocols abort the read operation, require the cache with the dirty data to update the main memory, and then allow the requesting cache to retry the read operation. However, this process adds latency to the read operation and increases bus traffic to the main memory. In an effort to further reduce bus traffic to the main memory, other protocols allow a first cache that is holding locally modified data (i.e., dirty data) to directly supply a second cache that is requesting the same block, without updating main memory.

[0004] Features and advantages of the disclosed methods and apparatus will be apparent to those of ordinary skill in the art in view of the detailed description of certain embodiments which is made with reference to the drawings, a brief description of which is provided below.

[0005] FIG. 1 is a high level block diagram of a computer system illustrating an environment of use for the present invention.

[0006] FIG. 2 is a more detailed block diagram of the multi-processor illustrated in FIG. 1.

[0007] FIG. 3 is a flowchart of a process for cache intervention in a multi-processor system.

[0008] FIG. 4 is a state diagram of a MESI cache coherency protocol amended to include exclusive intervention and shared intervention.

[0009] In general, the methods and apparatus described herein provide for cache-to-cache block transfers from a first cache to a second cache (i.e., cache intervention) when the state of the transferred block is in a non-modified state (e.g., exclusive or shared). In a first embodiment, the first cache holds the memory block in an exclusive state prior to the block transfer, and the second cache does not hold the memory block. When a processor associated with the second cache attempts to read the block from a main memory, the first cache intervenes and supplies the block instead of main memory supplying the block. The memory block in the second cache is stored in a shared state. In addition, the state of the memory block in the first cache changes from exclusive to shared. In a second embodiment, a processor associated with a third cache attempts to read the block from the main memory while the first cache and the second both hold the memory block in the shared state. Either the first cache or the second cache is determined to be an arbitration winner, and the arbitration winner intervenes and supplies the block. In both embodiments, communications with main memory and power consumption are reduced.

[0010] A block diagram of a computer system 100 is illustrated in FIG. 1. The computer system 100 may be a personal computer (PC), a personal digital assistant (PDA), an Internet appliance, a cellular telephone, or any other computing device. For one embodiment, the computer system 100 includes a main processing unit 102 powered by a power supply 103. The main processing unit 102 may include a multi-processor unit 104 electrically coupled by a system interconnect 106 to a main memory device 108 and one or more interface circuits 110. For one embodiment, the system interconnect 106 is a address/data bus. Of course, a person of ordinary skill in the art will readily appreciate that interconnects other than busses may be used to connect the multi-processor unit 104 to the main memory device 108. For example, one or more dedicated lines and/or a crossbar may be used to connect the multi-processor unit 104 to the main memory device 108.

[0011] The multi-processor 104 may include any type of well known central processing unit (CPU), such as a CPU from the Intel Pentium family of microprocessors, the Intel Itanium family of microprocessors, and/or the Intel XScale family of processors. In addition, the multi-processor 104 may include any type of well known cache memory, such as static random access memory (SRAM). The main memory device 108 may include dynamic random access memory (DRAM) and/or non-volatile memory. For one embodiment, the main memory device 108 stores a software program which is executed by the multi-processor 104 in a well known manner.

[0012] The interface circuit(s) 110 may be implemented using any type of well known interface standard, such as an Ethernet interface and/or a Universal Serial Bus (USB) interface. One or more input devices 112 may be connected to the interface circuits 110 for entering data and commands into the main processing unit 102. For example, an input device 112 may be a keyboard, mouse, touch screen, track pad, track ball, isopoint, and/or a voice recognition system.

[0013] One or more displays, printers, speakers, and/or other output devices 114 may also be connected to the main processing unit 102 via one or more of the interface circuits 110. The display 114 may be cathode ray tube (CRTs), liquid crystal displays (LCDs), or any other type of display. The display 114 may generate visual indications of data generated during operation of the main processing unit 102. The visual displays may include prompts for human operator input, calculated values, detected data, etc.

[0014] The computer system 100 may also include one or more storage devices 116. For example, the computer system 100 may include one or more hard drives, a compact disk (CD) drive, a digital versatile disk drive (DVD), and/or other computer media input/output (I/O) devices.

[0015] The computer system 100 may also exchange data with other devices via a connection to a network 118. The network connection may be any type of network connection, such as an Ethernet connection, digital subscriber line (DSL), telephone line, coaxial cable, etc. The network 118 may be any type of network, such as the Internet, a telephone network, a cable network, and/or a wireless network.

[0016] A more detailed block diagram of the multi-processor unit 104 is illustrated in FIG. 2. Although certain signal names are used to describe this embodiment, a person of ordinary skill in the art will readily appreciate that the name of each of the signal lines described herein is irrelevant to the operation of the signal line. Similarly, although certain connection schemes and logic gates are used to describe this embodiment, a person of ordinary skill in the art will readily appreciate that many other connection schemes and/or logic gates may be used without departing from the scope or spirit of the present invention.

[0017] In the embodiment illustrated in FIG. 2, the multi-processor 104 includes a plurality of processing agents 200 and a memory controller 202 electrically coupled by a cache interconnect 204. The cache interconnect 204 may be any type of interconnect such as a bus, one or more dedicated lines, and/or a crossbar. Each of the components of the multi-processor 104 may be on the same chip or on separate chips. For one embodiment, the main memory 108 resides on a separate chip. Due to the memory controller 202, one processing agent 200 may communicate with another processing agent 200 via the cache interconnect 204 without the communication necessarily generating activity on the system interconnect 106. Typically, if activity on the system interconnect 106 is reduced, overall power consumption is reduced. This is especially true in an embodiment where the main memory 108 resides on a separate chip from the processing agents 200.

[0018] Each processing agent 200 may include a central processing unit (CPU) 206 and one or more cache(s) 208. As discussed above, each CPU 206 may be any type of well known processor such as an Intel Pentium processor. Similarly, each cache may be constructed using any type of well known memory, such as SRAM. In addition, each processing agent 200 may include more than one cache. For example, a processing agent may include a level 1 cache and a level 2 cache. Similarly, a processing agent may include an instruction cache and/or a data cache.

[0019] Each processing agent 200 may include at least one signal input and at least one signal output. For one embodiment, a hit out signal output is asserted when an agent 200 detects activity on the cache interconnect 204 associated with a memory location for which the agent 200 is currently holding a copy in its cache 208. For one embodiment, each agent snoops address lines on a cache interconnect bus and asserts hit out each time it sees an address associated with a memory block in its cache. For example, if a second agent initiates a read request, and a first agent holds a copy of the same memory block in its cache, the first agent may assert its hit out line.

[0020] For one embodiment, one or more of these hit out lines are connected to a hit in line on each processing agent 200. For one embodiment, all of the hit out lines are logically ORed together, by one or more OR gates 210, and the output of the OR gate(s) 210 is connected to each of the hit in lines as shown in FIG. 2. In this manner, an active processing agent 200 knows when the cache 208 of another processing agent 200 holds a memory block associated with an activity the active processing agent 200 is performing. However, the active processing agent 200 does not necessarily know which cache 208 holds the memory block. Each processing agent 200 may be structured to use this hit in line to initiate and/or cancel any activity the processing agent 200 is capable of performing. For example, an asserted hit in line may serve to cancel a read from main memory.

[0021] In addition, one or more of the hit out lines are may be connected to a back-off input on each processing agent 200. For one embodiment, a first processing agent 200 optionally includes a back-off input which is never asserted (e.g., the input is connected to logic zero). This processing agent 200 has the highest priority in an arbitration scheme described in detail below (i.e., no other agent ever tells this agent to back-off). A second processing agent 200 may include a back-off input which is connected only to the hit out of the first processing agent. This processing agent has the second highest priority (i.e., only the highest priority agent can tell this agent to back-off). If included in the system, a third processing agent 200 may include a back-off input which is connected to the output of a first OR gate 210. The inputs of the first OR gate 210 are in turn connected to the hit out signals of the first processing agent 200 and the second processing agent 200. This processing agent has the third highest priority (i.e., either of the highest priority agent and the second highest priority agent can tell this agent to back-off). If included in the system, a fourth processing agent 200 may include a back-off input which is connected to the output of a second OR gate 210. The inputs of the second OR gate 210 are in turn connected to the hit out signal of the third processing agent 200 and the output of the first OR gate 210. This processing agent 200 has the fourth highest priority (i.e., any of the first three agents can tell this agent to back-off). This pattern may continue for any number of processing agents 200 as shown in FIG. 2.

[0022] A flowchart of a process 300 for cache intervention is illustrated in FIGS. 3-4. Adjacent each operation in the illustrated process 300 is a block diagram illustrating example actions taken by each of a first cache 208, a second cache 208, a third cache 208, and a main memory 108 during the associated operation. For simplicity in description, only one short memory block is illustrated for each of the first cache 208, the second cache 208, the third cache 208, and the main memory 108. Although the process 300 is described with reference to the flowchart illustrated in FIGS. 3-4, a person of ordinary skill in the art will readily appreciate that many other methods of performing the acts associated with process 300 may be used. For example, the order of some of the operations may be changed without departing from the scope or spirit of the present invention. In addition, many of the operations described are optional, and many additional operations may occur between the operations illustrated.

[0023] For one embodiment, a write-back (as opposed to a write-through) policy is used. A write-back policy is a cache procedure whereby a cache agent 200 may locally modify data in its cache 208 without updating main memory 108 until the cache block needs to be replaced. In order to maintain cache coherency in such a system, a cache coherency protocol may be used.

[0024] In one embodiment, a MESI (i.e., modified, exclusive, shared, invalid) cache coherency protocol is followed. However, a person of ordinary skill in the art will readily appreciate that any cache coherency protocol which includes the equivalent of a non-modified state, an exclusive state, and/or a shared state may be used without departing from the scope or spirit of the present invention. For example, a MOESI, ESI, Berkeley, or Illinois cache coherency protocol may be used. In the well known MESI cache coherency protocol, an invalid block is a block that does not contain useful data (i.e., the block is effectively empty). An exclusive block is a block that is non-modified (i.e., the same as main memory) and only held by one cache 208 (e.g., the block was just read in from main memory for the first time). A modified block is a block that is dirty (i.e., different from main memory) and only held by one cache 208 (e.g., a new value was written to the cache copy, but not to main memory's copy). A shared block is a block that is held by more than one cache 208. If a MOESI type protocol is used, an owned state is added. An owned block is a block that is modified and shared (i.e., dirty and held by another cache). The owner of a block is responsible for eventually updating main memory 108 with the modified value (i.e., the owner is responsible for performing the write-back).

[0025] In one embodiment, the state of a cached memory block is recorded in a cache directory. In another embodiment, the state of a cached memory block is recorded in a tag associated with the cached memory block. In the MOESI cache coherency protocol there are five possible states. Accordingly, each state may be represented by a different digital combination (e.g., 000Modified, 001Owned, 010Exclusive, 011Shared, 100Invalid). Retagging a cached memory block is the act of changing the state of the cached memory block. For example, retagging a block from exclusive to shared may be accomplished by changing a tag associated with the block from 010 to 011. Of course, a person of ordinary skill in the art will readily appreciate that any method of storing and changing a cache block state may be used without departing from the scope and spirit of the present invention.

[0026] Generally, process 300 illustrates an example exclusive cache intervention and an example shared cache intervention. In the exclusive cache intervention example, the first cache holds a memory block in an exclusive state prior to a block transfer, and a second cache does not hold the memory block. When a processor associated with the second cache attempts to read the block from a main memory, the first cache intervenes and supplies the block instead of main memory supplying the block. For one embodiment, the memory block in the second cache is stored in a shared state. In addition, the state of the memory block in the first cache may change from exclusive to shared.

[0027] In the shared cache intervention example, a processor associated with a third cache attempts to read the block from the main memory while the first cache and the second both hold the memory block in the shared state. Either the first cache or the second cache is determined to be an arbitration winner, and the arbitration winner intervenes and supplies the block. Of course, any number of caches may be used with any type of arbitration scheme. In both examples, communications with main memory and power consumption are reduced.

[0028] The process 300 begins when a first processing agent 200 initiates a read request for a particular memory block (operation 302). In this example, the first cache 208 includes a position that is tagged invalid. Of course, a person of ordinary skill in the art will readily appreciate that a cache position need not be tagged invalid to be over-written, and many well known cache replacement protocols, such as least recently used (LRU), may be used to determine which cache position is to be over-written.

[0029] No other cache 208 currently holds the requested memory block (e.g., no hit is generated or a cache directory indicates that no other caches holds the requested block), so main memory 108 supplies the requested block (operation 304). This action requires the memory controller 202 to access the main memory 108 via the system interconnect 106. The cached block is may be tagged exclusive to indicate that no other cache 208 currently holds this block (operation 304).

[0030] If the second processing agent 200 initiates a read request for the same memory block, the first cache 208 detects a hit (e.g., by snooping the address bus shared by the first and second agents or using a cache directory) (operation 306). Because the first cache 208 is holding the block in the exclusive state (i.e., the block in the first cache is the same as the block in main memory), main memory 108 could be allowed to supply the block, as requested by the second processing agent 200. However, the first cache 208 may intervene and supply the block via the cache interconnect 204 in order to reduce traffic on the system interconnect 106 (operation 306). The memory blocks in both the first cache 208 and the second cache 208 may be tagged shared to indicate that another cache 208 also holds this memory block (operation 306). In other words, if either cache 208 writes to this block, the other cache 208 needs to be updated or invalidated. Significantly, in operation 306, a first processing agent 200 intervenes to supply a block held in an exclusive state to a second processing agent 200.

[0031] If the third processing agent 200 also initiates a read request for the same memory block, the first and second caches 208 both detect a hit (e.g., by snooping the address bus or via a cache directory) (operation 308). As a result, the second cache 208 may assert the back-off input of the first cache (operation 308). Because the first cache 208 and the second cache 208 are both holding the block in the shared state (i.e., the cache blocks are the same as the block in main memory), main memory 108 could be allowed to supply the block, as requested by the third processing agent 200. However, the second cache 208 may intervene and supplies the block via the cache interconnect 204 in order to reduce traffic on the system interconnect 106 (operation 308). The first cache 208 knows to let another cache 208 (i.e., the second cache) supply the block because the back-off input of the first cache is asserted. The memory block in the third cache 208 may be tagged shared to indicate that another cache 208 also holds this memory block (operation 308). Significantly, in operation 308, one processing agent 200 intervenes to supply a block held in a shared state to another processing agent 200, and the intervening agent 200 also asserts a signal to suppress yet another agent 200 from supplying the same block.

[0032] A state diagram 500 of a MESI cache coherency protocol amended to include exclusive intervention and shared intervention is illustrated in FIG. 4. In addition to the state transitions normally associated with the well known MESI cache coherency protocol, two transitions are modified and one transition is added.

[0033] First, a snoop push operation 502 is added to the exclusive-to-shared transition associated with a snoop hit on read. A snoop push operation is a cache operation in which a first cache supplies a memory block to a second cache instead of a main memory supplying the second cache. A cache following this amended protocol will intervene to supply an exclusive block to a requesting cache and change the state of the supplied block to shared.

[0034] Second, a shared-to-shared transition 504 associated with a snoop hit on read with no back-off is added, and this new transition includes a snoop push operation 506. A cache following this amended protocol will intervene to supply a shared block to a requesting cache without changing the state of the supplied block. This protocol could be followed, for example, by the cache that wins the arbitration in a shared block situation.

[0035] Third, the shared-to-shared transition 508 normally associated with a snoop hit on read is modified to additionally check if a back-off signal is asserted. There is no snoop push associated with this transition. Accordingly, a cache with a shared block that is told to back-off, will not place traffic on the cache interconnect 204. This modification to the standard MESI protocol allows another cache that does not receive a back-off signal to intervene in accordance with the new SHRNBO transition 504 without contention on the cache interconnect 204. Of course, a person of ordinary skill in the art will readily appreciate that other arbitration schemes may be similarly employed.

[0036] In summary, persons of ordinary skill in the art will readily appreciate that methods and apparatus for cache intervention has been provided. Systems implementing the teachings described herein may benefit from a reduction in memory latency, bus traffic, and power consumption.

[0037] The foregoing description has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the example embodiments disclosed. Many modifications and variations are possible in light of the above teachings. It is intended that the scope of the invention be limited not by this detailed description of example embodiments, but rather by the claims appended hereto.