Qualcomm Incorporated

Qualcomm Innovation Center, Inc.

Sergei Larin

Anshuman Das Gupta

Lucian Codrescu

Shankar Kalpathi Easwaran

Hemant Kulkarni

Matthew C Curtis

Tobias Edler Von Koch

Ron Lieberman

A system for translating compressed instructions to instructions in an executable format is described. A translation unit is configured to decompress compressed instructions into a native instruction format using X and Y indices accessed from a memory, a translation memory, and a program specified mix mask. A level 1 cache is configured to store the native instruction format for each compressed instruction. The memory may be configured as a paged instruction cache to store pages of compressed instructions intermixed with pages of uncompressed instructions. Methods of determining a mix mask for efficiently translating compressed instructions is also described. A genetic method uses pairs of mix masks as genes from a seed population of mix masks that are bred and may be mutated to produce pairs of offspring mix masks to update the seed population. A mix mask for efficiently translating compressed instructions is determined from the updated seed population.

Methods and apparatus for storage and translation of entropy encoded software embedded within a memory hierarchy

A method for optimizing source code comprises optimizing the source code of files from a computer program at link-time, and receiving, at a linker, a customized linker script defining output sections for files of an executable version of the files of the computer program. The method comprises adding, to intermediate representation files having global or local symbols, metadata comprising default section assignment information for the symbols and recording, for symbols in machine code files, an origin path and an output section. The method further comprises parsing, by the compiler, the intermediate representation files, recording the symbols and related symbol information comprising default section assignment and dependency information of the intermediate representation files, assigning output sections to the symbols based on the default section assignments and instructions from the customized linker script, and linking optimized code of the files of the computer program based on the assigned output sections.

Link time program optimization in presence of a linker script

A method of compressing a sequence of program instructions begins by examining a program instruction stream to identify a sequence of two or more instructions that meet a parameter. The identified sequence of two or more instructions is replaced by a selected type of layout instruction which is then compressed. A method of decompressing accesses an X-index and a Y-index together as a compressed value. The compressed value is decompressed to a selected type of layout instruction which is decoded and replaced with a sequence of two or more instructions. An apparatus for decompressing includes a storage subsystem configured for storing compressed instructions, wherein a compressed instruction comprises an X-index and a Y-index. A decompressor is configured for translating an X-index and Y-index accessed from the storage subsystem to a selected type of layout instruction which is decoded and replaced with a sequence of two or more instructions.

Methods and apparatus for storage and translation of an entropy encoded instruction sequence to executable form

An example system for compiling a source file includes an optimizer that identifies a segment of code in a first source file as a potential optimization opportunity. The first source file includes high-level source code. The example system also includes a compiler that identifies a compilation record storing data indicating that the segment of code is an actual optimization opportunity and generates a representation of the high-level source code in accordance with the actual optimization opportunity. The data is based on a previous compilation of a second source file.

Whole-program optimization using data from previous compilation runs

A method for link time optimization comprises parsing, by a compiler, an intermediate representation file to determine what symbols are present in the intermediate representation file. The method comprises providing the symbols to a linker and creating, by the linker, a symbol use tree of all the symbols that are present in the intermediate representation file and other symbols in binary code received by the linker. The method further comprises discarding, by the linker, any received objects for which no use can be identified and all dependencies of the objects. The method includes providing, from the linker to the compiler, a preserve list of symbols, the preserve list comprising a list of symbols proven used by the objects and the intermediate representation files. The method comprises compiling the intermediate representation files and the objects based on the preserve list of symbols, and deleting, by the linker, any remaining unused objects.

System and method for link time optimization

Systems and methods for static code scheduling are disclosed. A method can include receiving an intermediate representation of source code, building a directed acyclic graph (DAG) for the intermediate representation, and creating chains of dependent instructions from the DAG for cluster formation. The chains are merged into clusters and each node in the DAG is marked with an identifier of a cluster it is part of to generate a marked instruction DAG. Instruction DAG scheduling is then performed using information about the clusters to generate an ordered intermediate representation of the source code.

Growing community of inventors

Austin, TX, United States of America

Sergei Larin