linux-kernel - Re: linux-next: build failure after merge of the net-next tree

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 11 Jul 2019 11:50:54 +1000
From:   Stephen Rothwell <sfr@...b.auug.org.au>
To:     Jason Gunthorpe <jgg@...lanox.com>
Cc:     Leon Romanovsky <leon@...nel.org>,
        Bernard Metzler <bmt@...ich.ibm.com>,
        Doug Ledford <dledford@...hat.com>,
        David Miller <davem@...emloft.net>,
        Networking <netdev@...r.kernel.org>,
        Linux Next Mailing List <linux-next@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: linux-next: build failure after merge of the net-next tree

Hi all,

On Wed, 10 Jul 2019 17:52:17 +0000 Jason Gunthorpe <jgg@...lanox.com> wrote:
>
> On Tue, Jul 09, 2019 at 09:43:46AM +0300, Leon Romanovsky wrote:
> > On Tue, Jul 09, 2019 at 01:56:36PM +1000, Stephen Rothwell wrote:  
> > > Hi all,
> > >
> > > After merging the net-next tree, today's linux-next build (x86_64
> > > allmodconfig) failed like this:
> > >
> > > drivers/infiniband/sw/siw/siw_cm.c: In function 'siw_create_listen':
> > > drivers/infiniband/sw/siw/siw_cm.c:1978:3: error: implicit declaration of function 'for_ifa'; did you mean 'fork_idle'? [-Werror=implicit-function-declaration]
> > >    for_ifa(in_dev)
> > >    ^~~~~~~
> > >    fork_idle
> > > drivers/infiniband/sw/siw/siw_cm.c:1978:18: error: expected ';' before '{' token
> > >    for_ifa(in_dev)
> > >                   ^
> > >                   ;
> > >    {
> > >    ~
> > >
> > > Caused by commit
> > >
> > >   6c52fdc244b5 ("rdma/siw: connection management")
> > >
> > > from the rdma tree.  I don't know why this didn't fail after I mereged
> > > that tree.  
> > 
> > I had the same question, because I have this fix for a couple of days already.
> > 
> > From 56c9e15ec670af580daa8c3ffde9503af3042d67 Mon Sep 17 00:00:00 2001
> > From: Leon Romanovsky <leonro@...lanox.com>
> > Date: Sun, 7 Jul 2019 10:43:42 +0300
> > Subject: [PATCH] Fixup to build SIW issue
> > 
> > Signed-off-by: Leon Romanovsky <leonro@...lanox.com>
> >  drivers/infiniband/sw/siw/siw_cm.c | 5 ++---
> >  1 file changed, 2 insertions(+), 3 deletions(-)
> > 
> > diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c
> > index 8e618cb7261f..c883bf514341 100644
> > +++ b/drivers/infiniband/sw/siw/siw_cm.c
> > @@ -1954,6 +1954,7 @@ static void siw_drop_listeners(struct iw_cm_id *id)
> >  int siw_create_listen(struct iw_cm_id *id, int backlog)
> >  {
> >  	struct net_device *dev = to_siw_dev(id->device)->netdev;
> > +	const struct in_ifaddr *ifa;
> >  	int rv = 0, listeners = 0;
> > 
> >  	siw_dbg(id->device, "id 0x%p: backlog %d\n", id, backlog);
> > @@ -1975,8 +1976,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog)
> >  			id, &s_laddr.sin_addr, ntohs(s_laddr.sin_port),
> >  			&s_raddr->sin_addr, ntohs(s_raddr->sin_port));
> > 
> > -		for_ifa(in_dev)
> > -		{
> > +		in_dev_for_each_ifa_rcu(ifa, in_dev) {
> >  			if (ipv4_is_zeronet(s_laddr.sin_addr.s_addr) ||  
> 
> Hum. There is no rcu lock held here and we can't use RCU anyhow as
> siw_listen_address will sleep.
> 
> I think this needs to use rtnl, as below. Bernard, please urgently
> confirm. Thanks
> 
> diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c
> index 8e618cb7261f62..ee98e96a5bfaba 100644
> --- a/drivers/infiniband/sw/siw/siw_cm.c
> +++ b/drivers/infiniband/sw/siw/siw_cm.c
> @@ -1965,6 +1965,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog)
>  	 */
>  	if (id->local_addr.ss_family == AF_INET) {
>  		struct in_device *in_dev = in_dev_get(dev);
> +		const struct in_ifaddr *ifa;
>  		struct sockaddr_in s_laddr, *s_raddr;
>  
>  		memcpy(&s_laddr, &id->local_addr, sizeof(s_laddr));
> @@ -1975,8 +1976,8 @@ int siw_create_listen(struct iw_cm_id *id, int backlog)
>  			id, &s_laddr.sin_addr, ntohs(s_laddr.sin_port),
>  			&s_raddr->sin_addr, ntohs(s_raddr->sin_port));
>  
> -		for_ifa(in_dev)
> -		{
> +		rtnl_lock();
> +		in_dev_for_each_ifa_rtnl(ifa, in_dev) {
>  			if (ipv4_is_zeronet(s_laddr.sin_addr.s_addr) ||
>  			    s_laddr.sin_addr.s_addr == ifa->ifa_address) {
>  				s_laddr.sin_addr.s_addr = ifa->ifa_address;
> @@ -1988,7 +1989,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog)
>  					listeners++;
>  			}
>  		}
> -		endfor_ifa(in_dev);
> +		rtnl_unlock();
>  		in_dev_put(in_dev);
>  	} else if (id->local_addr.ss_family == AF_INET6) {
>  		struct inet6_dev *in6_dev = in6_dev_get(dev);

So today this failed to build after I merged the rdma tree (previously
it didn;t until after the net-next tree was merged (I assume a
dependency changed).  It failed because in_dev_for_each_ifa_rcu (and
in_dev_for_each_ifa_rtnl) is only defined in a commit in the net-next
tree :-(

I have disabled the driver again.

-- 
Cheers,
Stephen Rothwell

Content of type "application/pgp-signature" skipped