Content-Type: multipart/mixed; boundary="===============8367859576447490264==" MIME-Version: 1.0 From: Florian Westphal To: mptcp at lists.01.org Subject: [MPTCP] Re: [PATCH mptcp-next 3/3] mptcp: switch mptcp_poll to mptcp socket wait queue Date: Fri, 11 Oct 2019 12:06:40 +0200 Message-ID: <20191011100640.GM25052@breakpoint.cc> In-Reply-To: alpine.OSX.2.21.1910101702400.68386@crpeter1-mobl.amr.corp.intel.com X-Status: X-Keywords: X-UID: 2096 --===============8367859576447490264== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Mat Martineau wrote: > > - pr_debug("parent=3D%p", parent); > > + struct socket_wq *wq; > > = > > smp_mb__before_atomic(); > > set_bit(MPTCP_DATA_READY, &mptcp_sk(parent)->flags); > > smp_mb__after_atomic(); > > = > > - parent->sk_data_ready(parent); > > - } > > + rcu_read_lock(); > > + wq =3D rcu_dereference(parent->sk_wq); > > + if (skwq_has_sleeper(wq) && > > + mptcp_should_wake_parent(subflow)) { > > + wake_up_interruptible_all(&wq->wait); > > + } > > + rcu_read_unlock(); > = > Compared to sock_def_readable, this wakeup code doesn't call sk_wake_asyn= c. > Is SOCK_FASYNC irrelevant here (sk_wake_async doesn't do anything without > that sock flag)? Its not. Perhaps best to make this a simpler if (mptcp_should_wake_parent(subflow)) parent->sk_data_ready(parent); I will make this change in the next iteration. --===============8367859576447490264==--