> which prints "f1". This reminded me of the fact that in Linux
> all filedescriptors have in fact names in the /proc directory,
> which kind of turns my original argument around: Since all fds
> have names, you can actually use the regular system calls to
> access them and are not required to use the f-calls instead. In
> his Example, Rainer demonstrates this by emulating the
> nonexistent fexecve() with a regular execve() on file descriptor 3.
>
Unfortunately, this works with some calls, but the semantics of
/proc/*/fd are a little screwy -- they don't dup() the file
descriptor, rather, they *open the same object again*. Meaning a
different file pointer, etc.
This is due to the way open is implemented in the VFS. I've been
meaning to hack up a different design and run it by the community, but
haven't gotten a chance yet. Perhaps in 2.5...
-hpa
-- <hpa@transmeta.com> at work, <hpa@zytor.com> in private!- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/