Commit 0b1208b1 authored Jul 02, 2010 by Aneesh Kumar K.V Committed by Eric Van Hensbergen Aug 02, 2010

fs/9p: destroy fid on failed remove



9P spec says:
"It is correct to consider remove to be a clunk with the
side effect of removing the file if permissions allow. "

So even if remove fails we need to destroy the fid.

Without this patch an rmdir on a directory with contents leave
the new cloned directory fid fid attached to fidlist. On umount
we dump the fids on the fidlist

~# rmdir /mnt2/test4/
rmdir: failed to remove `/mnt2/test4/': Directory not empty
~# umount /mnt2/
~# dmesg
[  228.474323] Found fid 3 not clunked

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Venkateswararao Jujjuri <jvrao@linux.vnet.ibm.com>
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>

parent a534c8d1

net/9p/client.c

+1 −2

Original line number	Diff line number	Diff line
		@@ -1240,9 +1240,8 @@ int p9_client_remove(struct p9_fid *fid)
		P9_DPRINTK(P9_DEBUG_9P, "<<< RREMOVE fid %d\n", fid->fid);

		p9_free_req(clnt, req);
		p9_fid_destroy(fid);

		error:
		p9_fid_destroy(fid);
		return err;
		}
		EXPORT_SYMBOL(p9_client_remove);

Admin message