NetBSD Problem Report #60637

From www@netbsd.org  Mon Aug 24 20:27:19 2026
Return-Path: <www@netbsd.org>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
	 key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256
	 client-signature RSA-PSS (2048 bits) client-digest SHA256)
	(Client CN "mail.netbsd.org", Issuer "YR2" (not verified))
	by mollari.NetBSD.org (Postfix) with ESMTPS id 449331A923B
	for <gnats-bugs@gnats.NetBSD.org>; Mon, 24 Aug 2026 20:27:19 +0000 (UTC)
Message-Id: <20260824202718.1D69B1A923C@mollari.NetBSD.org>
Date: Mon, 24 Aug 2026 20:27:18 +0000 (UTC)
From: joe@ns1.eloisystems.com
Reply-To: joe@ns1.eloisystems.com
To: gnats-bugs@NetBSD.org
Subject: destroying a vlan interface with a configured parent interface hangs
X-Send-Pr-Version: www-1.0
X-From4GNATS: "joe@ns1.eloisystems.com via gnats" <gnats-admin@NetBSD.org>

>Number:         60637
>Category:       kern
>Synopsis:       destroying a vlan interface with a configured parent interface hangs
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Aug 24 20:30:00 +0000 2026
>Last-Modified:  Tue Aug 25 20:15:01 +0000 2026
>Originator:     Emmanuel Nyarko
>Release:        NetBSD 11
>Organization:
NetBSD
>Environment:
amd64 x86
>Description:

The usual procedure to configuring a vlan interface is 

create the vlan interface, configure it with a vlan ID and a parent

when shutting it down, 

unconfigure it
and 
destroy it and this works without hanging.

but running destroy when not unconfigured hangs. 
>How-To-Repeat:

ifconfig vlan0 create
ifconfig vlan0 vlan 6 vlanif $interface
ifconfig vlan0 destroy
>Fix:
Not yet!

>Audit-Trail:
From: joe@ns1.eloisystems.com
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/60637: destroying a vlan interface with a configured parent
 interface hangs
Date: Tue, 25 Aug 2026 12:34:50 +0000

 On Mon, Aug 24, 2026 at 08:30:00PM +0000, gnats-admin@netbsd.org wrote:
 > Thank you very much for your problem report.
 > It has the internal identification `kern/60637'.
 > The individual assigned to look at your
 > report is: kern-bug-people. 
 > 
 > >Category:       kern
 > >Responsible:    kern-bug-people
 > >Synopsis:       destroying a vlan interface with a configured parent interface hangs
 > >Arrival-Date:   Mon Aug 24 20:30:00 +0000 2026

 Hi,

 The attched patch fixes this issue.

 Index: sys/net/if.c
 ===================================================================
 RCS file: /cvsroot/src/sys/net/if.c,v
 retrieving revision 1.538
 diff -u -r1.538 if.c
 --- sys/net/if.c	14 May 2026 08:05:48 -0000	1.538
 +++ sys/net/if.c	25 Aug 2026 12:27:04 -0000
 @@ -1356,12 +1356,10 @@
  	 * from pserialize read sections.  Note that we can't do
  	 * psref_target_destroy here.  See below.
  	 */
 -	IFNET_GLOBAL_LOCK();
  	ifindex2ifnet[ifp->if_index] = NULL;
  	TAILQ_REMOVE(&ifnet_list, ifp, if_list);
  	IFNET_WRITER_REMOVE(ifp);
  	pserialize_perform(ifnet_psz);
 -	IFNET_GLOBAL_UNLOCK();

  	if (ifp->if_slowtimo != NULL) {
  		struct if_slowtimo_data *isd = ifp->if_slowtimo_data;



 However, I might not be fully convinced yet if that's the best solution
 but i tracked down the hang and realized the vlan thread running if_detach forever waits
 on the "BIG LOCK" IFNET_GLOBAL_LOCK() because another thread is using it.


 My guess is maybe this lock doesn't need to be held here as it is causes a deadlock on vlan interfaces
 when destroying. or some other thread needs to release hold unto the BIG lock


From: Taylor R Campbell <riastradh@NetBSD.org>
To: Emmanuel Nyarko <joe@ns1.eloisystems.com>
Cc: gnats-bugs@NetBSD.org, netbsd-bugs@NetBSD.org
Subject: Re: kern/60637: destroying a vlan interface with a configured parent interface hangs
Date: Tue, 25 Aug 2026 19:04:01 +0000

 > Date: Mon, 24 Aug 2026 20:27:18 +0000 (UTC)
 > From: Emmanuel Nyarko <joe@ns1.eloisystems.com>
 > 
 > >Description:
 > 
 > The usual procedure to configuring a vlan interface is 
 > 
 > create the vlan interface, configure it with a vlan ID and a parent
 > 
 > when shutting it down, 
 > 
 > unconfigure it
 > and 
 > destroy it and this works without hanging.
 > 
 > but running destroy when not unconfigured hangs. 
 > >How-To-Repeat:
 > 
 > ifconfig vlan0 create
 > ifconfig vlan0 vlan 6 vlanif $interface
 > ifconfig vlan0 destroy

 I tried this in a VM with interface=vioif0 and it worked fine.

 Can you share the output of the following, before it hangs?

 1. ifconfig -a
 2. dmesg
 3. when it hangs, start crash(8) in another terminal (or start it
    before the hang) and run:
    - ps
    - show all tstiles/t

 Also, are you running with any local patches?

 > -	IFNET_GLOBAL_LOCK();
 >  	ifindex2ifnet[ifp->if_index] = NULL;
 >  	TAILQ_REMOVE(&ifnet_list, ifp, if_list);
 >  	IFNET_WRITER_REMOVE(ifp);
 >  	pserialize_perform(ifnet_psz);
 > -	IFNET_GLOBAL_UNLOCK();

 This is definitely not a fix!  Everything here -- ifindex2ifnet,
 ifnet_list, IFNET_WRITER_*, ifnet_psz -- must have access serialized
 by IFNET_GLOBAL_LOCK in order to coordinate the data structures with
 other threads.

From: joe@ns1.eloisystems.com
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/60637: destroying a vlan interface with a configured parent
 interface hangs
Date: Tue, 25 Aug 2026 20:10:16 +0000

 On Tue, Aug 25, 2026 at 07:05:02PM +0000, Taylor R Campbell via gnats wrote:
 > The following reply was made to PR kern/60637; it has been noted by GNATS.
 > 
 > From: Taylor R Campbell <riastradh@NetBSD.org>
 > To: Emmanuel Nyarko <joe@ns1.eloisystems.com>
 > Cc: gnats-bugs@NetBSD.org, netbsd-bugs@NetBSD.org
 > Subject: Re: kern/60637: destroying a vlan interface with a configured parent interface hangs
 > Date: Tue, 25 Aug 2026 19:04:01 +0000
 > 
 >  > Date: Mon, 24 Aug 2026 20:27:18 +0000 (UTC)
 >  > From: Emmanuel Nyarko <joe@ns1.eloisystems.com>
 >  > 
 >  > >Description:
 >  > 
 >  > The usual procedure to configuring a vlan interface is 
 >  > 
 >  > create the vlan interface, configure it with a vlan ID and a parent
 >  > 
 >  > when shutting it down, 
 >  > 
 >  > unconfigure it
 >  > and 
 >  > destroy it and this works without hanging.
 >  > 
 >  > but running destroy when not unconfigured hangs. 
 >  > >How-To-Repeat:
 >  > 
 >  > ifconfig vlan0 create
 >  > ifconfig vlan0 vlan 6 vlanif $interface
 >  > ifconfig vlan0 destroy
 >  
 >  I tried this in a VM with interface=vioif0 and it worked fine.

 so the funny case here is that it hangs 80 percent of time.
 i tried it five times, it hung four time and didn't one time.

 >  
 >  Can you share the output of the following, before it hangs?
 >  
 >  1. ifconfig -a
 >  2. dmesg
 >  3. when it hangs, start crash(8) in another terminal (or start it
 >     before the hang) and run:
 >     - ps
 >     - show all tstiles/t

 so I got it though DDB and this is what i got.

 * Lock 1 (initialized at netbsd:if_initialize+0x10e)
 lock address : ffffc674c3581f40
 type         : sleep/adaptive
 initialized  : netbsd:if_initialize+0x10e
 shared holds :                  0 exclusive:                  1
 shares wanted:                  0 exclusive:                  2
 relevant cpu :                  1 last held:                  1
 relevant lwp : 0xffffc674c4278800 last held: 0xffffc674c4278800
 last locked* : netbsd:if_detach+0x61
 unlocked     : netbsd:vlan_clone_destroy+0x5c
 owner field  : 0xffffc674c4278800 wait/spin:                1/0
 Turnstile: 
 => 0 waiting readers:
 => 2 waiting writers: 0xffffc675380a4c00 0xffffc674c2743000

 *** Locks wanted:

 workqueue_wait(ifnet_link_state_wq, &ifp->if_link_work);

 i got some printfs in and never run past this function.

 this is the function causing it i am now convinced.

 >  
 >  Also, are you running with any local patches?

 no patches. this is a GENERIC kernel using wm0 interface on a VM.

Home	PR Database Search