SUNRPC: Ensure we flush any closed sockets before xs_xprt_free()
authorTrond Myklebust <trond.myklebust@hammerspace.com>
Sun, 3 Apr 2022 19:58:11 +0000 (15:58 -0400)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 18 May 2022 08:26:57 +0000 (10:26 +0200)
commit54f6834b283d9b4d070b0639d9ef5e1d156fe7b0
treeadefe0cfb1136d5a90c3446efb87fb7ebb877bfd
parentcc57b2e4a7371b7232f365fcd3054fef73ef3ef1
SUNRPC: Ensure we flush any closed sockets before xs_xprt_free()

commit f00432063db1a0db484e85193eccc6845435b80e upstream.

We must ensure that all sockets are closed before we call xprt_free()
and release the reference to the net namespace. The problem is that
calling fput() will defer closing the socket until delayed_fput() gets
called.
Let's fix the situation by allowing rpciod and the transport teardown
code (which runs on the system wq) to call __fput_sync(), and directly
close the socket.

Reported-by: Felix Fu <foyjog@gmail.com>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Fixes: a73881c96d73 ("SUNRPC: Fix an Oops in udp_poll()")
Cc: stable@vger.kernel.org # 5.1.x: 3be232f11a3c: SUNRPC: Prevent immediate close+reconnect
Cc: stable@vger.kernel.org # 5.1.x: 89f42494f92f: SUNRPC: Don't call connect() more than once on a TCP socket
Cc: stable@vger.kernel.org # 5.1.x
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Cc: Meena Shanmugam <meenashanmugam@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/file_table.c
include/trace/events/sunrpc.h
net/sunrpc/xprt.c
net/sunrpc/xprtsock.c