> > As a general feature, however, is it reasonable to add a S_ITMP flag to
> > creat(2) to create temp files that do not have associated directory entries?
>
> <shudder> Good luck implementing it over NFS. OTOH... Something
> similar is possible, but I don't think that create() is the right
> interface for that. BTW, most of the times you really need not that, but
> anonymous mmap...
Where's the problem with nfs? In case of nfs, it will be kernel's duty
to try to generate unique name. It has to do it, anyway (look at
nfs_sillyrename), so I do not see why this is a problem.
> > i.e. the filesystem just allocates an inode and returns it, with the file
> > being deleted on close. We save on having to carefully find a unique filename
> > for the tempfile, when we're just going to unlink the thing anyway. Of course
> > this doesn't do anything for shell scripts.
Pavel
-- I'm really pavel@atrey.karlin.mff.cuni.cz. Pavel Look at http://atrey.karlin.mff.cuni.cz/~pavel/ ;-).- 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/