Correction apparatus, correction method, and computer product

A correction method executed by a computer, the correction method comprising:

• acquiring the execution time of an instruction in a given block among a block group that includes blocks obtained by dividing program code;
• detecting a first resource group designated by a tail instruction in a preceding block that is executed before the given block and a second resource group designated by a head instruction of the given block;
• identifying, when a resource common to the detected first resource group and the detected second resource group is an arithmetic unit and arithmetic units of a CPU (Central Processing Unit) that is subject to simulation by the program code is of a number that is less than the number of arithmetic units designated by the tail instruction and the head instruction, the arithmetic unit;
• calculating from the time when the identified arithmetic unit is used by the head instruction and the time when use of the identified arithmetic unit by the tail instruction ends, a delay period caused by the preceding block;
• correcting the acquired execution time of the instruction in the given block based on the delay period calculated by the calculating;
• and
• outputting the corrected execution time.

A non-transitory computer-readable medium storing therein a correction program that causes a computer to execute a process, the process comprising:

• acquiring the execution time of an instruction in a given block among a block group that includes blocks obtained by dividing program code;
• detecting a first resource group designated by a tail instruction in a preceding block that is executed before the given block and a second resource group designated by a head instruction of the given block;
• identifying, when a resource common to the detected first resource group and the detected second resource group is an arithmetic unit and arithmetic units of a CPU (Central Processing Unit) that is subject to simulation by the program code is of a number that is less than the number of arithmetic units designated by the tail instruction and the head instruction, the arithmetic unit;
• calculating from the time when the identified arithmetic unit is used by the head instruction and the time when use of the identified arithmetic unit by the tail instruction ends, a delay period caused by the preceding block;
• correcting the acquired execution time of the instruction in the given block based on the delay period calculated by the calculating;
• and
• outputting the corrected execution time.