55 results about "Software cache" patented technology

A method and apparatus for use in a computer system including a plurality of host computers including a root host computer and at least one child host computer. The root host computer exports at least a portion of the volume of storage to the at least one child host computer so they can share access to the volume of storage. In one embodiment, the volume of storage is stored on at least one non-volatile storage device. In another aspect, the volume of storage is made available to the root from a storage system. In a further aspect, at least a second portion of the volume of storage is exported from a child host computer to at least one grandchild host computer.

Mechanisms for optimized code generation targeting a high locality software cache are provided. Original computer code is parsed to identify memory references in the original computer code. Memory references are classified as either regular memory references or irregular memory references. Regular memory references are controlled by a high locality cache mechanism. Original computer code is transformed, by a compiler, to generate transformed computer code in which the regular memory references are grouped into one or more memory reference streams, each memory reference stream having a leading memory reference, a trailing memory reference, and one or more middle memory references. Transforming of the original computer code comprises inserting, into the original computer code, instructions to execute initialization, lookup, and cleanup operations associated with the leading memory reference and trailing memory reference in a different manner from initialization, lookup, and cleanup operations for the one or more middle memory references.

A computer-implemented method for classifying received packets using a hardware cache of evolving rules and a software cache having an original rule set. The method including receiving a packet, processing the received packet through a hardware-based packet classifier having at least one evolving rule to identify at least one cache miss packet, and processing the cache miss packet through a software based packet classifier including an original rule set. Processing the cache miss packet includes determining whether to expand at least one of the at least one evolving rules in the hardware-based packet classifier based on the cache miss packet. The determination includes determining whether an evolving rule has both the same action and lies entirely within one of the rule of the original rule set.

Dynamic translation of indirect branch instructions of a target application by a host processor is enhanced by including a cache to provide access to the addresses of the most frequently used translations of a host computer, minimizing the need to access the translation buffer. Entries in the cache have a host instruction address and tags that may include a logical address of the instruction of the target application, the physical address of that instruction, the code segment limit to the instruction, and the context value of the host processor associated with that instruction. The cache may be a software cache apportioned by software from the main processor memory or a hardware cache separate from main memory.

A method to efficiently pre-fetch and batch compiler-assisted software cache accesses is provided. The method reduces the overhead associated with software cache directory accesses. With the method, the local memory address of the cache line that stores the pre-fetched data is itself cached, such as in a register or well known location in local memory, so that a later data access does not need to perform address translation and software cache operations and can instead access the data directly from the software cache using the cached local memory address. This saves processor cycles that would otherwise be required to perform the address translation a second time when the data is to be used. Moreover, the system and method directly enable software cache accesses to be effectively decoupled from address translation in order to increase the overlap between computation and communication.

A compiler implemented software cache apparatus and method in which non-aliased explicitly fetched data are excluded are provided. With the mechanisms of the illustrative embodiments, a compiler uses a forward data flow analysis to prove that there is no alias between the cached data and explicitly fetched data. Explicitly fetched data that has no alias in the cached data are excluded from the software cache. Explicitly fetched data that has aliases in the cached data are allowed to be stored in the software cache. In this way, there is no runtime overhead to maintain the correctness of the two copies of data. Moreover, the number of lines of the software cache that must be protected from eviction is decreased. This leads to a decrease in the amount of computation cycles required by the cache miss handler when evicting cache lines during cache miss handling.

A highly predictable, low overhead and yet dynamic, memory allocation methodology for embedded systems with scratch-pad memory is presented. The dynamic memory allocation methodology for global and stack data (i) accounts for changing program requirements at runtime; (ii) has no software-caching tags; (iii) requires no run-time checks; (iv) has extremely low overheads; and (v) yields 100% predictable memory access times. The methodology provides that for data that is about to be accessed frequently is copied into the SRAM using compiler-inserted code at fixed and infrequent points in the program. Earlier data is evicted if necessary.

An apparatus and method for performing useful computations during a software cache reload operation is provided. With the illustrative embodiments, in order to perform software caching, a compiler takes original source code, and while compiling the source code, inserts explicit cache lookup instructions into appropriate portions of the source code where cacheable variables are referenced. In addition, the compiler inserts a cache miss handler routine that is used to branch execution of the code to a cache miss handler if the cache lookup instructions result in a cache miss. The cache miss handler, prior to performing a wait operation for waiting for the data to be retrieved from the backing store, branches execution to an independent subroutine identified by a compiler. The independent subroutine is executed while the data is being retrieved from the backing store such that useful work is performed.

A system and method to efficiently pre-fetch and batch compiler-assisted software cache accesses are provided. The system and method reduce the overhead associated with software cache directory accesses. With the system and method, the local memory address of the cache line that stores the pre-fetched data is itself cached, such as in a register or well known location in local memory, so that a later data access does not need to perform address translation and software cache operations and can instead access the data directly from the software cache using the cached local memory address. This saves processor cycles that would otherwise be required to perform the address translation a second time when the data is to be used. Moreover, the system and method directly enable software cache accesses to be effectively decoupled from address translation in order to increase the overlap between computation and communication.

A highly predictable, low overhead and yet dynamic, memory allocation methodology for embedded systems with scratch-pad memory is presented. The dynamic memory allocation methodology for global and stack data (i) accounts for changing program requirements at runtime; (ii) has no software-caching tags; (iii) requires no run-time checks; (iv) has extremely low overheads; and (v) yields 100% predictable memory access times. The methodology provides that for data that is about to be accessed frequently is copied into the SRAM using compiler-inserted code at fixed and infrequent points in the program. Earlier data is evicted if necessary.

The present invention provides for a method for computer program code optimization for a software managed cache in either a uni-processor or a multi-processor system. A single source file comprising a plurality of array references is received. The plurality of array references is analyzed to identify predictable accesses. The plurality of array references is analyzed to identify secondary predictable accesses. One or more of the plurality of array references is aggregated based on identified predictable accesses and identified secondary predictable accesses to generate aggregated references. The single source file is restructured based on the aggregated references to generate restructured code. Prefetch code is inserted in the restructured code based on the aggregated references. Software cache update code is inserted in the restructured code based on the aggregated references. Explicit cache lookup code is inserted for the remaining unpredictable accesses. Calls to a miss handler for misses in the explicit cache lookup code are inserted. A miss handler is included in the generated code for the program. In the miss handler, a line to evict is chosen based on recent usage and predictability. In the miss handler, appropriate DMA commands are issued for the evicted line and the missing line.

The present invention provides for a method for computer program code optimization for a software managed cache in either a uni-processor or a multi-processor system. A single source file comprising a plurality of array references is received. The plurality of array references is analyzed to identify predictable accesses. The plurality of array references is analyzed to identify secondary predictable accesses. One or more of the plurality of array references is aggregated based on identified predictable accesses and identified secondary predictable accesses to generate aggregated references. The single source file is restructured based on the aggregated references to generate restructured code. Prefetch code is inserted in the restructured code based on the aggregated references. Software cache update code is inserted in the restructured code based on the aggregated references. Explicit cache lookup code is inserted for the remaining unpredictable accesses. Calls to a miss handler for misses in the explicit cache lookup code are inserted. A miss handler is included in the generated code for the program. In the miss handler, a line to evict is chosen based on recent usage and predictability. In the miss handler, appropriate DMA commands are issued for the evicted line and the missing line.

Software-based cache coherence protocol. A processing unit may execute a memory request using a processor thread. In response to detecting a cache hit to shared or a cache miss associated with the memory request, a cache may provide both a trap signal and coherence information to the processor thread of the processing unit. After receiving the trap signal and the coherence information, the processor thread may perform a cache coherence operation for the memory request using at least the received coherence information. The processing unit may include a plurality of processor threads and a load balancer. The load balancer may receive coherence requests from one or more remote processing units and distribute the received coherence requests across the plurality of processor threads. The load balance may preferentially distribute the received coherence requests across the plurality of processor threads based on the operation state of the processor threads.

The invention provides a double-buffering type robot software log storage method. The method is characterized by comprising the following steps of: setting a log storage software function loading module, a log storage software function unloading module, a log storage software operation module, a log information software buffer, a log information software receiver and a log information software analyzer, wherein the log information software buffer comprises two log buffer areas with same storage spaces; when one log buffer area is full of data, switching an input pointer to the other log buffer area so as not to delay the data input; pointing an output pointer to the log buffer area full of data, outputting the log data information in the log buffer area full of data, and setting the log buffer area to be in an idle state. According to the method, the log data information can be stored, and the overall operation load of robot system software can be greatly reduced, so that benefit is brought to improve the software defect check efficiency.

Mechanisms for optimized code generation targeting a high locality software cache are provided. Original computer code is parsed to identify memory references in the original computer code. Memory references are classified as either regular memory references or irregular memory references. Regular memory references are controlled by a high locality cache mechanism. Original computer code is transformed, by a compiler, to generate transformed computer code in which the regular memory references are grouped into one or more memory reference streams, each memory reference stream having a leading memory reference, a trailing memory reference, and one or more middle memory references. Transforming of the original computer code comprises inserting, into the original computer code, instructions to execute initialization, lookup, and cleanup operations associated with the leading memory reference and trailing memory reference in a different manner from initialization, lookup, and cleanup operations for the one or more middle memory references.

A mechanism for efficient software cache accessing with handle reuse is provided. The mechanism groups references in source code into a reference stream with the reference stream having a size equal to or less than a size of a software cache line. The source code is transformed into optimized code by modifying the source code to include code for performing at most two cache lookup operations for the reference stream to obtain two cache line handles. Moreover, the transformation involves inserting code to resolve references in the reference stream based on the two cache line handles. The optimized code may be output for generation of executable code.

Reducing runtime coherency checking using global data flow analysis is provided. A determination is made as to whether a call is for at least one of a DMA get operation or a DMA put operation in response to the call being issued during execution of a compiled and optimized code. A determination is made as to whether a software cache write operation has been issued since a last flush operation in response to the call being the DMA get operation. A DMA get runtime coherency check is then performed in response to the software cache write operation being issued since the last flush operation.

The use of heap memory is optimized by extending a cache implementation with a CacheInterface base class. An instance of a ReferenceToCache is attached to the CacheInterface base class. The cache implementation is registered to a garbage collector application. The registration is stored as a reference list in a memory. In response to an unsatisfied cache allocation request, a garbage collection cycle is triggered to check heap occupancy. In response to exceeding a threshold value, the reference list is traversed for caches to be cleaned based upon a defined space constraint value. The caches are cleaned in accordance with the defined space constraint value.

In some embodiments, the invention involves a system and method relating to software caching with bounded-error delayed updates. Embodiments of the present invention describe a delayed-update software-controlled cache, which may be used to reduce memory access latencies and improve throughput for domain specific applications that are tolerant of errors caused by delayed updates of cached values. In at least one embodiment of the present invention, software caching may be implemented by using a compiler to automatically generate caching code in application programs that must access and / or update memory. Cache is accessed for a period of time, even if global data has been updated, to delay costly memory accesses. Other embodiments are described and claimed.

A cache memory system having a small-capacity and high-speed access cache memory provided between a processor and a main memory, including a software cache controller for performing software control for controlling data transfer to the cache memory in accordance with a preliminarily programmed software and a hardware cache controller for performing hardware control for controlling data transfer to the cache memory by using a predetermined hardware such that the processor causes the software cache controller to perform the software control but causes the hardware cache controller to perform the hardware control when it becomes impossible to perform the software control.

Process, cache memory, computer product and system for loading data associated with a requested address in a software cache. The process includes loading address tags associated with a set in a cache directory using a Single Instruction Multiple Data (SIMD) operation, determining a position of the requested address in the set using a SIMD comparison, and determining an actual data value associated with the position of the requested address in the set.

The invention provides a serial hub and a multi-serial high-speed communication method. The serial hub comprises a processing unit and a serial data switching unit, wherein the processing unit is connected with a host and is used for processing data and command signals transmitted from the host; the serial data switching unit is connected with the processing unit and a plurality of serial ports and is used for switching communication links or setting communication interface parameters according to control signals of the processing unit to ensure that a host port is connected with the corresponding serial port; the processing unit further comprises a universal asynchronous transceiver, a software buffer and a hardware buffer, wherein the universal asynchronous transceiver is connected with the host and is used for realizing the communication with the host, the software buffer is connected with the universal asynchronous transceiver and is used for storing overflow data, and the hardware buffer is connected with the software buffer and is used for overcoming time delay caused by the hardware buffer. The serial hub has the advantages of high communication speed, low error rate and wide application range.

A compiler implemented software cache apparatus and method in which non-aliased explicitly fetched data are excluded are provided. With the mechanisms of the illustrative embodiments, a compiler uses a forward data flow analysis to prove that there is no alias between the cached data and explicitly fetched data. Explicitly fetched data that has no alias in the cached data are excluded from the software cache. Explicitly fetched data that has aliases in the cached data are allowed to be stored in the software cache. In this way, there is no runtime overhead to maintain the correctness of the two copies of data. Moreover, the number of lines of the software cache that must be protected from eviction is decreased. This leads to a decrease in the amount of computation cycles required by the cache miss handler when evicting cache lines during cache miss handling.

Mechanisms for performing useful computations during a software cache reload operation are provided. With the illustrative embodiments, in order to perform software caching, a compiler takes original source code, and while compiling the source code, inserts explicit cache lookup instructions into appropriate portions of the source code where cacheable variables are referenced. In addition, the compiler inserts a cache miss handler routine that is used to branch execution of the code to a cache miss handler if the cache lookup instructions result in a cache miss. The cache miss handler, prior to performing a wait operation for waiting for the data to be retrieved from the backing store, branches execution to an independent subroutine identified by a compiler. The independent subroutine is executed while the data is being retrieved from the backing store such that useful work is performed.

Techniques are disclosed relating to controlling cache size and priority of data stored in the cache using machine learning techniques. A software cache may store data for a plurality of different user accounts using one or more hardware storage elements. In some embodiments, a machine learning module generates, based on access patterns to the software cache, a control value that specifies a size of the cache and generates time-to-live values for entries in the cache. In some embodiments, the system evicts data based on the time-to-live values. The disclosed techniques may reduce cache access times and / or improve cache hit rate.

A computer processor collects information for a dominant data access loop and reference code patterns based on data reference pattern analysis, and for pointer aliasing and data shape based on pointer escape analysis. The computer processor selects a candidate array for data splitting wherein the candidate array is referenced by a dominant data access loop. The computer processor determines a data splitting mode by which to split the data of the candidate array, based on the reference code patterns, the pointer aliasing, and the data shape information, and splits the data into two or more split arrays. The computer processor creates a software cache that includes a portion of the data of the two or more split arrays in a transposed format, and maintains the portion of the transposed data within the software cache and consults the software cache during an access of the split arrays.

A computer processor collects information for a dominant data access loop and reference code patterns based on data reference pattern analysis, and for pointer aliasing and data shape based on pointer escape analysis. The computer processor selects a candidate array for data splitting wherein the candidate array is referenced by a dominant data access loop. The computer processor determines a data splitting mode by which to split the data of the candidate array, based on the reference code patterns, the pointer aliasing, and the data shape information, and splits the data into two or more split arrays. The computer processor creates a software cache that includes a portion of the data of the two or more split arrays in a transposed format, and maintains the portion of the transposed data within the software cache and consults the software cache during an access of the split arrays.

The invention relates to a distributed storage access method initiated by a GPU (Graphics Processing Unit) in an AI (Artificial Intelligence) scene. The method comprises the following steps: S100: constructing a software cache module and an I / O (Input / Output) queue module in a GPU memory; s200, the GPU thread queries the target data through the software cache module, reads the target data if the target data is hit, and enters the next step if the target data is not hit; s300, when the I / O request is not hit, the repeated I / O requests initiated by multiple threads are automatically combined to generate a single effective I / O request, and the request is written into an I / O queue module; s400, the user mode distributed storage client monitors the written submission queue, reads the target data and then transmits the target data to the software cache module, and a completion queue of the I / O queue module is updated; and S500, the GPU thread polls the updated completion queue, and updates the reference count of the software cache module after reading the data. The main purpose of the invention is to realize high-throughput and low-delay data transmission without the participation of a CPU (Central Processing Unit).

The invention proposes a method and system for cleaning software cache. The right management module first applies to the user for the permission of cache cleaning, if user agrees to grant permission,the scanning module scans the storage space of the terminal device, scans the software cache files through the software cache directory, the cache file is compared with the original file by the comparison judgment module, If the corresponding original file does not exist, the cache file is marked as junk, Cleanup module cleans marked garbage files. The unmarked cache file is a normal file, and unusual software is marked by counting the usage of the normal file, and garbage files and unusual software are cleaned by a cleaning module, so that the invention can more flexibly and intelligently clean the cache file, and the accuracy and efficiency of the file cleaning are improved.