Make sure to have 3 lines that has "blocksizex.x=your value" (where x.x corresponds to n value of 0.5, 1.0 and 1.5 billion) so that when the automatic grading script does a grep and sed, it gets the corresponding value and checks the performance of your sieve3/4 code. 

Example README file content:
blocksize0.5=1000
blocksize1.0=1000
blocksize1.5=1000

Linux command (for Preliminary testing):
grep "blocksize" README | sed 's/blocksize0.5=//' | sed 's/blocksize1.0=//' | sed  's/blocksize1.5=//' | awk -F"|" ' $1 ~ "^[0-9][0-9]*$" { print $1 }' | wc -l

Expected output:
3

Note: 
Run the above command in your directory that has the README file and verify if you get 3 as your output.

This is just to make sure that even if my best blocksize doesn't give the expected speed-up for your code, I would like to give credits for the best block size that you report. 

Thank you for your patience and cooperation in following the instructions !!