Write a C/C program that performs the tasks described below.
The program should expect 3 cmd-line args:
filename
nbytes - number of bytes from the file to mmap into memory
nthreads - number of threads to create to solve the problem
For example:
. /p3 file1 100000000 4
The program will mmap 100,000,000 bytes of file1 into memory and use 4 threads to examine the bytes.
Note that the program will NOT read the file. Instead, it will simply mmap nbytes of the file into memory.
The main will not participate in the computation, but will create the specified number of threads, and wait for them to complete the computation, and the main will then print the answer. At most 8 threads will be specified to be created.
Use long int instead of simple int because there may be more than 4 GB of data. This implies that you may also want to use atol instead of atoi.
Use
"-Ofast"
as in p1 because performance matters. In addition to good single-thread performance, the program should also get reasonable speed-up with multiple threads.
Each file will contain a long string of letters like DNA, i.e. a,c,g,t The program should determine how many 20-character substrings contain more than 11 a's.
For example, in this short string, there are 2 20-character matches: tattataaagtagaaatataactgaaggttcagccgctgg attataaagtagaaatataa aaagtagaaatataactgaa
The program should print one line of output: TOTAL MATCHES n # where n is replaced by the correct number --------