I can't speak directly to efficiency as it's hardly my specialty, but I would suggest a change to this chunk:

    while($line = <F>) 
        {
            for $hit ($line =~ /matchdata/gi) 
            {
           
                push @files, $name;

                }
         }
         close F;
[download]

Logically it seems like you want to check to see if the file you're currently examining has data that needs to be updated (you're saving the filename in @files for later processing) -- but you're pushing the filename onto your list once for each match in the file! Once you find a match in a particular file you should (based on my understanding of your goal) (1) push the filename onto your to-be-processed stack, and then (2) move on to the next file. So:

    while($line = <F>) 
        {
            if ($line =~ /matchdata/i) 
            {
           
                push @files, $name;
                last;

                }
         }
         close F;
[download]

Note that I dropped the global flag for the match operator, too -- you just want to know if there's something there, anywhere, to be fixed later, and then move on.

Naturally, if I misunderstand your goals this could be way off base :)