From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id DAE6B40BF5 for ; Mon, 24 Jun 2024 10:27:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1719224875; cv=none; b=Xb90X9qHTaiczZ3IN6i1M7K2q5Vt/7B+Qidxmc5J44BQmqYnoWX2q6hVyf3zbp15JFONEEWwCoDS3xrHaCbr8a+UdMfYUE66QzO4zDxdSmJP9bH7od07Gie6qtChc0UsXK0gfgTGqhHBhihsLKR1f1rn1/ByJvieClmPwS9Rk9g= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1719224875; c=relaxed/simple; bh=NWx14Ny1Soc9A6jsS6xyXOirDK9z1JLf9jWdnrkTqD0=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=ifm6d8tb3sSZwcJbIpWf/oJL/D3pULRJAHgezMnVikRc/eMRMzQeag+w/iHhP8x5TBAsZJEleCT9PJeQ8TOMzjJN5D7aaz6plcUi6eI+IBwesHdAKsqMueiCG2o+pZ+rm1T+PKVJDI8VM22r7MFkz8JWo1qwTgRU6iynLSdbT2U= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=Oz4VjPee; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="Oz4VjPee" Received: by smtp.kernel.org (Postfix) with ESMTPSA id E8321C2BBFC; Mon, 24 Jun 2024 10:27:54 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1719224875; bh=NWx14Ny1Soc9A6jsS6xyXOirDK9z1JLf9jWdnrkTqD0=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=Oz4VjPeeG1n6QYkJ7AMu02ou/qsR1O7Tr99fcy2fTRjGwvcbcd2ftEDJ1pGi/wQ/y vE2JypuROTpohbYTrS/Nt0nhpI4gpAbIXfIvMrf30WqHrkdkfWSsaLNjlDBJxUyrOO WVLkxF1rvOSMwvAMXx2FyQZiDRWYNxeUMWb5RPkzz7CWMkmvQ5R7KoeYJfQowX6tbi cWoHgL/2Y3RUEjU/S65BvjrC6NcK7zcb2bGU9C63rjBR46DbpAwU2pjFNDlVRBsd/6 Lj5/kHyqx2DvvSATvoE17aqDFJRDkGXlBHWCaE51fTQ9yCGqv0iek0B6rhjyiF1gvg 38HmzEALicQSA== Date: Mon, 24 Jun 2024 13:27:51 +0300 From: Leon Romanovsky To: Steffen Klassert Cc: Jianbo Liu , Eric Dumazet , netdev@vger.kernel.org Subject: Re: [PATCH v2 ipsec] xfrm: Fix unregister netdevice hang on hardware offload. Message-ID: <20240624102751.GE29266@unreal> References: Precedence: bulk X-Mailing-List: netdev@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: On Thu, Jun 20, 2024 at 08:47:24AM +0200, Steffen Klassert wrote: > When offloading xfrm states to hardware, the offloading > device is attached to the skbs secpath. If a skb is free > is deferred, an unregister netdevice hangs because the > netdevice is still refcounted. > > Fix this by removing the netdevice from the xfrm states > when the netdevice is unregistered. To find all xfrm states > that need to be cleared we add another list where skbs > linked to that are unlinked from the lists (deleted) > but not yet freed. > > Changes in v2: > > - Fix build with CONFIG_XFRM_OFFLOAD disabled. > - Fix two typos in the commit message. Changelog should be after "---" trailer marker. Thanks > > Fixes: d77e38e612a0 ("xfrm: Add an IPsec hardware offloading API") > Signed-off-by: Steffen Klassert > --- > include/net/xfrm.h | 36 +++++++------------------ > net/xfrm/xfrm_state.c | 61 +++++++++++++++++++++++++++++++++++++++++-- > 2 files changed, 69 insertions(+), 28 deletions(-) > > diff --git a/include/net/xfrm.h b/include/net/xfrm.h > index 77ebf5bcf0b9..7d4c2235252c 100644 > --- a/include/net/xfrm.h > +++ b/include/net/xfrm.h > @@ -178,7 +178,10 @@ struct xfrm_state { > struct hlist_node gclist; > struct hlist_node bydst; > }; > - struct hlist_node bysrc; > + union { > + struct hlist_node dev_gclist; > + struct hlist_node bysrc; > + }; > struct hlist_node byspi; > struct hlist_node byseq; > > @@ -1588,7 +1591,7 @@ void xfrm_state_update_stats(struct net *net); > static inline void xfrm_dev_state_update_stats(struct xfrm_state *x) > { > struct xfrm_dev_offload *xdo = &x->xso; > - struct net_device *dev = xdo->dev; > + struct net_device *dev = READ_ONCE(xdo->dev); > > if (dev && dev->xfrmdev_ops && > dev->xfrmdev_ops->xdo_dev_state_update_stats) > @@ -1946,13 +1949,16 @@ int xfrm_dev_policy_add(struct net *net, struct xfrm_policy *xp, > struct xfrm_user_offload *xuo, u8 dir, > struct netlink_ext_ack *extack); > bool xfrm_dev_offload_ok(struct sk_buff *skb, struct xfrm_state *x); > +void xfrm_dev_state_delete(struct xfrm_state *x); > +void xfrm_dev_state_free(struct xfrm_state *x); > > static inline void xfrm_dev_state_advance_esn(struct xfrm_state *x) > { > struct xfrm_dev_offload *xso = &x->xso; > + struct net_device *dev = READ_ONCE(xso->dev); > > - if (xso->dev && xso->dev->xfrmdev_ops->xdo_dev_state_advance_esn) > - xso->dev->xfrmdev_ops->xdo_dev_state_advance_esn(x); > + if (dev && dev->xfrmdev_ops->xdo_dev_state_advance_esn) > + dev->xfrmdev_ops->xdo_dev_state_advance_esn(x); > } > > static inline bool xfrm_dst_offload_ok(struct dst_entry *dst) > @@ -1973,28 +1979,6 @@ static inline bool xfrm_dst_offload_ok(struct dst_entry *dst) > return false; > } > > -static inline void xfrm_dev_state_delete(struct xfrm_state *x) > -{ > - struct xfrm_dev_offload *xso = &x->xso; > - > - if (xso->dev) > - xso->dev->xfrmdev_ops->xdo_dev_state_delete(x); > -} > - > -static inline void xfrm_dev_state_free(struct xfrm_state *x) > -{ > - struct xfrm_dev_offload *xso = &x->xso; > - struct net_device *dev = xso->dev; > - > - if (dev && dev->xfrmdev_ops) { > - if (dev->xfrmdev_ops->xdo_dev_state_free) > - dev->xfrmdev_ops->xdo_dev_state_free(x); > - xso->dev = NULL; > - xso->type = XFRM_DEV_OFFLOAD_UNSPECIFIED; > - netdev_put(dev, &xso->dev_tracker); > - } > -} > - > static inline void xfrm_dev_policy_delete(struct xfrm_policy *x) > { > struct xfrm_dev_offload *xdo = &x->xdo; > diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c > index 649bb739df0d..d531d2a1fae2 100644 > --- a/net/xfrm/xfrm_state.c > +++ b/net/xfrm/xfrm_state.c > @@ -49,6 +49,7 @@ static struct kmem_cache *xfrm_state_cache __ro_after_init; > > static DECLARE_WORK(xfrm_state_gc_work, xfrm_state_gc_task); > static HLIST_HEAD(xfrm_state_gc_list); > +static HLIST_HEAD(xfrm_state_dev_gc_list); > > static inline bool xfrm_state_hold_rcu(struct xfrm_state __rcu *x) > { > @@ -214,6 +215,7 @@ static DEFINE_SPINLOCK(xfrm_state_afinfo_lock); > static struct xfrm_state_afinfo __rcu *xfrm_state_afinfo[NPROTO]; > > static DEFINE_SPINLOCK(xfrm_state_gc_lock); > +static DEFINE_SPINLOCK(xfrm_state_dev_gc_lock); > > int __xfrm_state_delete(struct xfrm_state *x); > > @@ -683,6 +685,40 @@ struct xfrm_state *xfrm_state_alloc(struct net *net) > } > EXPORT_SYMBOL(xfrm_state_alloc); > > +#ifdef CONFIG_XFRM_OFFLOAD > +void xfrm_dev_state_delete(struct xfrm_state *x) > +{ > + struct xfrm_dev_offload *xso = &x->xso; > + struct net_device *dev = READ_ONCE(xso->dev); > + > + if (dev) { > + dev->xfrmdev_ops->xdo_dev_state_delete(x); > + spin_lock_bh(&xfrm_state_dev_gc_lock); > + hlist_add_head(&x->dev_gclist, &xfrm_state_dev_gc_list); > + spin_unlock_bh(&xfrm_state_dev_gc_lock); > + } > +} > + > +void xfrm_dev_state_free(struct xfrm_state *x) > +{ > + struct xfrm_dev_offload *xso = &x->xso; > + struct net_device *dev = READ_ONCE(xso->dev); > + > + if (dev && dev->xfrmdev_ops) { > + spin_lock_bh(&xfrm_state_dev_gc_lock); > + if (!hlist_unhashed(&x->dev_gclist)) > + hlist_del(&x->dev_gclist); > + spin_unlock_bh(&xfrm_state_dev_gc_lock); > + > + if (dev->xfrmdev_ops->xdo_dev_state_free) > + dev->xfrmdev_ops->xdo_dev_state_free(x); > + WRITE_ONCE(xso->dev, NULL); > + xso->type = XFRM_DEV_OFFLOAD_UNSPECIFIED; > + netdev_put(dev, &xso->dev_tracker); > + } > +} > +#endif > + > void __xfrm_state_destroy(struct xfrm_state *x, bool sync) > { > WARN_ON(x->km.state != XFRM_STATE_DEAD); > @@ -848,6 +884,9 @@ EXPORT_SYMBOL(xfrm_state_flush); > > int xfrm_dev_state_flush(struct net *net, struct net_device *dev, bool task_valid) > { > + struct xfrm_state *x; > + struct hlist_node *tmp; > + struct xfrm_dev_offload *xso; > int i, err = 0, cnt = 0; > > spin_lock_bh(&net->xfrm.xfrm_state_lock); > @@ -857,8 +896,6 @@ int xfrm_dev_state_flush(struct net *net, struct net_device *dev, bool task_vali > > err = -ESRCH; > for (i = 0; i <= net->xfrm.state_hmask; i++) { > - struct xfrm_state *x; > - struct xfrm_dev_offload *xso; > restart: > hlist_for_each_entry(x, net->xfrm.state_bydst+i, bydst) { > xso = &x->xso; > @@ -868,6 +905,8 @@ int xfrm_dev_state_flush(struct net *net, struct net_device *dev, bool task_vali > spin_unlock_bh(&net->xfrm.xfrm_state_lock); > > err = xfrm_state_delete(x); > + xfrm_dev_state_free(x); > + > xfrm_audit_state_delete(x, err ? 0 : 1, > task_valid); > xfrm_state_put(x); > @@ -884,6 +923,24 @@ int xfrm_dev_state_flush(struct net *net, struct net_device *dev, bool task_vali > > out: > spin_unlock_bh(&net->xfrm.xfrm_state_lock); > + > + spin_lock_bh(&xfrm_state_dev_gc_lock); > +restart_gc: > + hlist_for_each_entry_safe(x, tmp, &xfrm_state_dev_gc_list, dev_gclist) { > + xso = &x->xso; > + > + if (xso->dev == dev) { > + spin_unlock_bh(&xfrm_state_dev_gc_lock); > + xfrm_dev_state_free(x); > + spin_lock_bh(&xfrm_state_dev_gc_lock); > + goto restart_gc; > + } > + > + } > + spin_unlock_bh(&xfrm_state_dev_gc_lock); > + > + xfrm_flush_gc(); > + > return err; > } > EXPORT_SYMBOL(xfrm_dev_state_flush); > -- > 2.34.1 > >