International Business Machines Corporation

Fadi A Zaraket

Ali S El-Zein

Jason Raymond Baumgartner

Viresh Paruthi

Gabor Drasny

Hari Mony

Gabor Bobok

Wolfgang Roesner

Hussein Sharafeddin

Co-Optimization utilizing Symbolic Execution (COSE) works across components of an embedded design to optimize structures therein. COSE utilizes symbolic execution (SE) to analyze software components and defines a limited set of values that software feeds hardware as constraints. SE explores substantially all possible paths of execution of the code specifying a component. It accomplishes this by accumulating path conditions (PCs) and annotating them to the corresponding segments of the component. A PC is associated with a branch of code and consists of the conjunction of conditions over input and state variables necessary and sufficient for the branch to execute. These PCs define constraints that limit the set of values that software feeds hardware. These constraints are then propagated across the networks of the design and employ static analysis techniques such as constant propagation, redundancy removal, and don't care optimizations to reduce the hardware components.

Co-optimization of embedded systems utilizing symbolic execution

A method of verifying a software system includes receiving a description of a software system described utilizing a high-level modeling language, and responsive thereto, parsing the description and constructing an abstract syntax graph. The abstract syntax graph is transformed into a sequential logic representation of the software system. The sequential logic representation is formed by reference to a Hardware Description Language (HDL) library. Then, the sequential logic representation is transformed into a gate-level sequential logic representation. Following the transforming, the software system is verified based upon the gate-level sequential logic representation. Following verification, results of verification of the software system are output.

Sequential encoding for relational analysis (SERA) of a software model

A method, system and computer program product for modeling variables in subprograms of a HDL program. A subprogram is provided with an initial value of a variable of an element being modeled and the subprogram is stored in memory of a data processing system. In response to a subprogram call, a copy of the stored subprogram is provided to the requesting HDL program. During execution, the initial value of the variable in the provided copy of the subprogram may be modified by the HDL program, but the value retains unchanged in the stored subprogram.

Techniques for modeling variables in subprograms of hardware description language programs

A system for performing verification includes a means for: importing a design netlist containing component(s), computing output function(s) for the component(s), generating output equivalent state set(s) from the output function(s), identifying next-state function(s) for the component(s), means for producing image equivalent state set(s) for the next-state function(s), means for classifying output-and-image equivalent state set(s) for the image equivalent state set(s) and the output equivalent state set(s), getting a preimage from the next-state function(s) and the output-and-image equivalent state(s) to generate a preimage of the output-and-image equivalent state(s), partitioning over original state(s) of the component(s), and equivalent class input set(s) of the component(s). Moreover, the system includes a means for: selecting input representative(s) of the equivalent input set(s), forming an input map from the input representative(s), synthesizing the input map, and injecting the input map back into the netlist to generate a modified netlist.

Predicate-based compositional minimization in a verification environment

Unrolling the 'generate' statement of a hardware description language ('HDL') and displaying the unrolled HDL. For a conditional generate the condition is evaluated. If the statement is true the enclosed HDL code will be displayed. For an iterative generate, the enclosing HDL will be displayed as many times as specified by the bounds of the iteration scheme. This allows, for example, simulation value annotations for signals declared inside the generate statement, semantic navigation inside the generate statements, and allows the user to visualize what is included in the target design.

Unrolling hardware design generate statements in a source window debugger

A method includes pre-compilation operations on HDL source code files, creating a 'make it' file, on demand processing of the HDL source code in an HDL source browser, and resolving overloaded function and operator calls in an HDL source code browser debugger. Construct a list of all HDL files a list of HDL files to be processed. Send an HDL file in the list for compiling. If compilation is successful, branch to add the HDL file to an end of target file and that HDL file is removed from the list. The list is tested for remaining files and then a next file in the list is sent for compiling. After all files in the list have been processed, the HDL files which have been processed are checked for failures to compile and if any of said HDL files to be processed have failed to compile the method branches back to repeating the process until all runs are successful.

Method, system, and program product for pre-compile processing of hardware design language (HDL) source files

A method for performing verification includes importing a design netlist containing one or more components and computing one or more output functions for the one or more components. One or more output equivalent state sets are generated from the one or more output functions and one or more next-state functions for the one or more components are identified. One or more image equivalent state sets for the one or more next-state functions are produced and one or more output-and-image equivalent state sets are classified for the one or more image equivalent state sets and the one or more output equivalent state sets. One or more input representatives of the one or more equivalent input sets are selected and an input map is formed from the one or more input representatives. The input map is synthesized and injected back into the netlist to generate a modified netlist.

Growing community of inventors

Austin, TX, United States of America

Fadi A Zaraket