Re: [PATCH RESEND] net/sctp: always initialise sctp_ht_iter::start_fail

2016-08-14 Thread David Miller
From: Vegard Nossum Date: Fri, 12 Aug 2016 09:50:51 +0200 > sctp_transport_seq_start() does not currently clear iter->start_fail on > success, but relies on it being zero when it is allocated (by > seq_open_net()). > > This can be a problem in the following sequence: > > open() // allocates

Re: [PATCH RESEND] net/sctp: always initialise sctp_ht_iter::start_fail

2016-08-12 Thread Marcelo Ricardo Leitner
On Fri, Aug 12, 2016 at 09:50:51AM +0200, Vegard Nossum wrote: > sctp_transport_seq_start() does not currently clear iter->start_fail on > success, but relies on it being zero when it is allocated (by > seq_open_net()). > > This can be a problem in the following sequence: > > open() // alloca

Re: [PATCH RESEND] net/sctp: always initialise sctp_ht_iter::start_fail

2016-08-12 Thread Neil Horman
On Fri, Aug 12, 2016 at 09:50:51AM +0200, Vegard Nossum wrote: > sctp_transport_seq_start() does not currently clear iter->start_fail on > success, but relies on it being zero when it is allocated (by > seq_open_net()). > > This can be a problem in the following sequence: > > open() // alloca

[PATCH RESEND] net/sctp: always initialise sctp_ht_iter::start_fail

2016-08-12 Thread Vegard Nossum
sctp_transport_seq_start() does not currently clear iter->start_fail on success, but relies on it being zero when it is allocated (by seq_open_net()). This can be a problem in the following sequence: open() // allocates iter (and implicitly sets iter->start_fail = 0) read() - iter->s