The
vn_fullpath()
function makes a
"best effort"
attempt to generate a string pathname for
the passed vnode; the resulting path, if any, will be relative to
the root directory of the process associated with the passed thread pointer.
The
vn_fullpath()
function
is implemented by inspecting the VFS name cache, and attempting to
reconstruct a path from the process root to the object.
This process is necessarily unreliable for several reasons: intermediate
entries in the path may not be found in the cache; files may have more
than one name (hard links), not all file systems use the name cache
(specifically, most synthetic file systems do not); a single name may
be used for more than one file (in the context of file systems covering
other file systems); a file may have no name (if deleted but still
open or referenced).
However, the resulting string may still be more useable to a user than
a vnode pointer value, or a device number and inode number.
Code consuming the results of this function should anticipate (and
properly handle) failure.
Its arguments are:
|
| vp
|
| |
The vnode to search for.
No need to be locked by the caller.
|
| retbuf
|
| |
Pointer to a
char *
that
vn_fullpath()
may (on success) point at a newly
allocated buffer containing the resulting pathname.
|
| freebuf
|
| |
Pointer to a
char *
that
vn_fullpath()
may (on success) point at a buffer
to be freed, when the caller is done with
retbuf.
|
|
Typical consumers will declare two character pointers:
fullpath
and
freepath;
they will set
freepath
to
NULL,
and
fullpath
to a name to use
in the event that the call to
vn_fullpath()
fails.
After done with the value of
fullpath,
the caller will check if
freepath
is
non- NULL,
and if so, invoke
free(9)
with a pool type of
M_TEMP.
