| phil-news-nospam@ipal.net 2007-06-24, 7:07 pm |
| On 23 Jun 2007 16:28:13 GMT Casper H.S. Dik <Casper.Dik@sun.com> wrote:
| phil-news-nospam@ipal.net writes:
|
|>If I create a file tree that just barely fits in PATH_MAX, then move it
|>inside of a longer directory name or deeper directory, I now have a name
|>that cannot be directly access if PATH_MAX is strictly complied with.
|
| It's nearly impossible for the OS to limit all created filenames to
| PATH_MAX; it would require the OS to know at all times the length
| of the longest child name and the length of the current pathname.
|
| Possible to achieve, but prohibitively expensive to compute in the
| general case.
I'm not expecting it to apply limits in the actual filesystem. What I am
interested in is if an implementation allows very very large strings to
be used as pathnames (> 4GB), what should PATH_MAX be defined as in headers?
--
|---------------------------------------/----------------------------------|
| Phil Howard KA9WGN (ka9wgn.ham.org) / Do not send to the address below |
| first name lower case at ipal.net / spamtrap-2007-06-23-2049@ipal.net |
|------------------------------------/-------------------------------------|
|