摘要:
Selected regions of native instructions translated in a DBT environment from non-native instructions are compressed based on the independent compression of different fields of selected instructions using compression tables to reduce a length of selected fields. The regions of compressed instructions are stored and de-compressed into the native instructions during subsequent execution using de-compression tables. Specifically, for native instructions of a selected region, selected types of opcodes and/or operands may be compressed independently. The types may be selected by profiling the opcodes using benchmark programs and creating an opcode conversion table prior to compression, and scanning of the operands and creating an operand conversion table during compression of the opcodes.
摘要:
A method and apparatus for improving parallelism through optimal code replication is herein described. An optimal replication factor for code is determined based on costs associated with a plurality of replication factors. The code is replicated by the optimal replication factor, and then the code is potentially executed in parallel to obtain parallelized efficient execution.
摘要:
Methods and apparatuses enable on-demand instruction emulation via user-level exception handling. A non-supported instruction triggers an exception during runtime of a program. In response to the exception, a user-level or application-level exception handler is launched, instead of a kernel-level handler. Then the exception handler can execute at the application layer instead of the kernel level. The handler identifies the instruction and emulates the instruction, where emulation of the instruction is supported by the handler. Emulating the instructions enables the program to continue execution. Repeated instruction emulation is amortized via dynamic binary translation of hot code.
摘要:
Disclosed is a method for running a first code generated by a Software-based Redundant Multi-Threading (SRMT) compiler along with a second code generated by a normal compiler at runtime, the first code including a first function and a second function, the second code including a third function. The method comprises running the first function in a leading thread and a tailing thread (104); running the third function in a single thread (106), the leading thread calls the third function and running the second function in the leading thread and the tailing thread (108), the third function calls the second function. The present disclosure provides a mechanism for handling function calls wherein SRMT functions and binary functions can call each other irrespective of whether the callee function is a SRMT function or a binary function and thereby dynamically adjusts reliability and performance tradeoff based on run-time information and user selectable policies.
摘要:
Methods are disclosed to implement bit scan operations using properties of two's complement arithmetic and compute zero index instructions. A data value may be provided and the most-significant or least-significant bit may be determined using the methods set forth herein.
摘要:
A method and apparatus for dynamic binary translator to support precise exceptions with minimal optimization constraints. In one embodiment, the method includes the translation of a source binary application generated for a source instruction set architecture (ISA) into a sequential, intermediate representation (IR) of the source binary application. In one embodiment, the sequential IR is modified to incorporate exception recovery information for each of the exception instructions identified from the source binary application to enable a dynamic binary translator (DBT) to represent exception recovery values as regular values used by IR instructions. In one embodiment, the sequential IR may be optimized with a constraint on movement of an exception instruction downward past an irreversible instruction to form a non-sequential IR. In one embodiment, the non-sequential IR is optimized to form a translated binary application for a target ISA. Other embodiments are described and claimed.
摘要:
Methods and apparatus for optimizing a program undergoing dynamic binary translation using profile information are disclosed. A disclosed system optimizes foreign program instructions through an enhanced dynamic binary translation process. The foreign program instructions are translated into native program instructions. Loops within the native program instructions are instrumented with profiling instructions and optimized. The profiling information is collected during execution of the loop. After profiling information is collected, the loop may be further optimized by inserting prefetching instructions into the optimized loop. The prefetched loop is then linked back into the native program instructions and is executable.
摘要:
In one implementation, the invention involves a computer implemented method used in compiling a program. The method includes selecting conflict regions of the program. The method further includes performing invalidation profiling of load instructions with respect to certain ones of the conflict regions to determine invalidation rates of the load instructions. The method may further include a feedback step in which the invalidation rates are used by a scheduler of the compiler to determine whether to move the load instructions to target locations.
摘要:
A method and apparatus for efficient and consistent validation/conflict detection in a Software Transactional Memory (STM) system is herein described. A version check barrier is inserted after a load to compare versions of loaded values before and after the load. In addition, a global timestamp (GTS) is utilized to track a latest committed transaction. Each transaction is associated with a local timestamp (LTS) initialized to the GTS value at the start of a transaction. As a transaction commits it updates the GTS to a new value and sets versions of modified locations to the new value. Pending transactions compare versions determined in read barriers to their LTS. If the version is greater than their LTS indicating another transaction has committed after the pending transaction started and initialized the LTS, then the pending transaction validates its read set to maintain efficient and consistent transactional execution.
摘要:
Systems, methods, and apparatuses for data speculation execution (DSX) are described. In some embodiments, a hardware apparatus for performing DSX comprises a hardware decoder to decode an instruction, the instruction to include an opcode and an operand to store a portion of a fallback address, execution hardware to execute the decoded instruction to initiate a data speculative execution (DSX) region by activating DSX tracking hardware to track speculative memory accesses and detect ordering violations in the DSX region, and storing the fallback address.