Al Viro 6b18662e23 9p connect fixes ... * if we fail in p9_conn_create(), we shouldn't leak references to struct file. Logics in ->close() doesn't help - ->trans is already gone by the time it's called. * sock_create_kern() can fail. * use of sock_map_fd() is all fscked up; I'd fixed most of that, but the rest will have to wait for a bit more work in net/socket.c (we still are violating the basic rule of working with descriptor table: "once the reference is installed there, don't rely on finding it there again"). Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>		2009-12-16 12:16:41 -05:00
..
client.c	9p: fix readdir corner cases	2009-11-02 08:43:45 -06:00
error.c	net/9p: insulate the client against an invalid error code sent by a 9p server	2009-08-17 16:39:54 -05:00
Kconfig	9p: disallow RDMA if RDMA CM isn't available	2009-01-15 16:39:42 -08:00
Makefile	9p: rdma: RDMA Transport Support for 9P	2008-10-22 18:47:39 -05:00
mod.c	9p: consolidate transport structure	2008-10-17 11:04:41 -05:00
protocol.c	9p: fix endian issues [attempt 3]	2009-02-06 22:07:41 -08:00
protocol.h	9p: rework client code to use new protocol support functions	2008-10-17 11:04:45 -05:00
trans_fd.c	9p connect fixes	2009-12-16 12:16:41 -05:00
trans_rdma.c	9p: minor comment fixes	2009-08-17 16:27:57 -05:00
trans_virtio.c	virtio: let header files include virtio_ids.h	2009-10-22 16:39:28 +10:30
util.c	9p: rework client code to use new protocol support functions	2008-10-17 11:04:45 -05:00