Re: [PATCH net-next] sctp: reset ret in again path in sctp_for_each_transport

2018-01-23 Thread David Miller
From: Xin Long Date: Tue, 23 Jan 2018 18:22:25 +0800 > Commit 97a6ec4ac021 ("rhashtable: Change rhashtable_walk_start to > return void") only initialized ret for the first time, when going > to again path, the next tsp could be NULL. Without resetting ret, > cb_done would

Re: [PATCH net-next] sctp: reset ret in again path in sctp_for_each_transport

2018-01-23 Thread Neil Horman
On Tue, Jan 23, 2018 at 06:22:25PM +0800, Xin Long wrote: > Commit 97a6ec4ac021 ("rhashtable: Change rhashtable_walk_start to > return void") only initialized ret for the first time, when going > to again path, the next tsp could be NULL. Without resetting ret, > cb_done would be called with tsp

Re: [PATCH net-next] sctp: reset ret in again path in sctp_for_each_transport

2018-01-23 Thread Marcelo Ricardo Leitner
On Tue, Jan 23, 2018 at 06:22:25PM +0800, Xin Long wrote: > Commit 97a6ec4ac021 ("rhashtable: Change rhashtable_walk_start to > return void") only initialized ret for the first time, when going > to again path, the next tsp could be NULL. Without resetting ret, > cb_done would be called with tsp

[PATCH net-next] sctp: reset ret in again path in sctp_for_each_transport

2018-01-23 Thread Xin Long
Commit 97a6ec4ac021 ("rhashtable: Change rhashtable_walk_start to return void") only initialized ret for the first time, when going to again path, the next tsp could be NULL. Without resetting ret, cb_done would be called with tsp as NULL. A kernel crash was caused by this when running sctpdiag