Linux – Grep and xargs: File name too long

Depends what you're doing exactly, but yes, ext file systems get slow when you've got a lot of files in one directory. Splitting the files into e.g. numbered subdirectories is one common way round this.

Does the ksh script explicitly define the path to awk? It could be via a variable ( AWK=/usr/bin/awk, invoked by $AWK). Try running which awk and alias awk, too.

Edit

Cygwin install process actually sets up /usr/bin as a hardlink to /bin. (Do ls -li /usr/ and ls -li / and you'll see the bin folders have the same inode number. Also, browse to the /usr/bin folder in windows explorer and you'll see it's empty.)

Your PATH is probably fine, but probably has /usr/bin before /bin, so it's finding awk via that route first.

Looking at your previous question, I think you might have a similar, shortcut related issue here. I tried renaming my awk.exe to dawk.exe in windows and then set up a shortcut called awk.exe pointing to it. Running awk then gives me the same error you can see. I had to force that issue, though; I think your Cygwin installation might be a bit broken, but it's hard to say how. Do you get similar errors from any other binaries?