From: Ken-ichirou MATSUZAWA Date: Fri, 29 Jan 2016 01:45:50 +0000 (+0900) Subject: netlink: not trim skb for mmaped socket when dump X-Git-Tag: v3.16.35~386 X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=c501d329b2d8052184cd63b2b7e86088f2f33187;p=thirdparty%2Fkernel%2Fstable.git netlink: not trim skb for mmaped socket when dump commit aa3a022094fac7f6e48050e139fa8a5a2e3265ce upstream. We should not trim skb for mmaped socket since its buf size is fixed and userspace will read as frame which data equals head. mmaped socket will not call recvmsg, means max_recvmsg_len is 0, skb_reserve was not called before commit: db65a3aaf29e. Fixes: db65a3aaf29e (netlink: Trim skb to alloc size to avoid MSG_TRUNC) Signed-off-by: Ken-ichirou MATSUZAWA Signed-off-by: David S. Miller Signed-off-by: Luis Henriques --- diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 7e1ac5b5de0a6..46cac8e31a16c 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -2740,7 +2740,8 @@ static int netlink_dump(struct sock *sk) * reasonable static buffer based on the expected largest dump of a * single netdev. The outcome is MSG_TRUNC error. */ - skb_reserve(skb, skb_tailroom(skb) - alloc_size); + if (!netlink_rx_is_mmaped(sk)) + skb_reserve(skb, skb_tailroom(skb) - alloc_size); netlink_skb_set_owner_r(skb, sk); len = cb->dump(skb, cb);