IB/addr: Fix setting source address in addr6_resolve()

IB/addr: Fix setting source address in addr6_resolve()

[Raju Rangoju]

Hi,

The commit eea40b8f624f25 has introduced a bug in rdma/core which broke iw_cxgb4/ipv6. 

So, please pull the below commit to 4.9.x, 4.11.x stable release as it fixes the issue.

commit 79e25959403e6a79552db28a87abed34de32a1df
Author: Roland Dreier <roland@purestorage.com>
Date:   Tue Jun 6 09:22:00 2017 -0700

    IB/addr: Fix setting source address in addr6_resolve()

    Commit eea40b8f624f ("infiniband: call ipv6 route lookup via the stub
    interface") introduced a regression in address resolution when connecting
    to IPv6 destination addresses.  The old code called ip6_route_output(),
    while the new code calls ipv6_stub->ipv6_dst_lookup().  The two are almost
    the same, except that ipv6_dst_lookup() also calls ip6_route_get_saddr()
    if the source address is in6addr_any.

    This means that the test of ipv6_addr_any(&fl6.saddr) now never succeeds,
    and so we never copy the source address out.  This ends up causing
    rdma_resolve_addr() to fail, because without a resolved source address,
    cma_acquire_dev() will fail to find an RDMA device to use.  For me, this
    causes connecting to an NVMe over Fabrics target via RoCE / IPv6 to fail.

    Fix this by copying out fl6.saddr if ipv6_addr_any() is true for the original
    source address passed into addr6_resolve().  We can drop our call to
    ipv6_dev_get_saddr() because ipv6_dst_lookup() already does that work.

    Fixes: eea40b8f624 ("infiniband: call ipv6 route lookup via the stub interface")
    Cc: <stable@vger.kernel.org> # 3.12+
    Signed-off-by: Roland Dreier <roland@purestorage.com>
    Acked-by: Paolo Abeni <pabeni@redhat.com>
    Signed-off-by: Doug Ledford mailto:dledford@redhat.com


Thanks,
Raju 

Re: IB/addr: Fix setting source address in addr6_resolve()

[Greg KH]

On Wed, Sep 13, 2017 at 10:55:25AM +0000, Raju  Rangoju wrote:
> Hi,
> 
> The commit eea40b8f624f25 has introduced a bug in rdma/core which broke iw_cxgb4/ipv6. 
> 
> So, please pull the below commit to 4.9.x, 4.11.x stable release as it fixes the issue.

4.11.x is end-of-life, but I have pulled this into the 4.9.x tree,
thanks.

greg k-h

Re: IB/addr: Fix setting source address in addr6_resolve()

[Greg KH]

On Wed, Sep 20, 2017 at 11:11:28AM +0200, Greg KH wrote:
> On Wed, Sep 13, 2017 at 10:55:25AM +0000, Raju  Rangoju wrote:
> > Hi,
> > 
> > The commit eea40b8f624f25 has introduced a bug in rdma/core which broke iw_cxgb4/ipv6. 
> > 
> > So, please pull the below commit to 4.9.x, 4.11.x stable release as it fixes the issue.
> 
> 4.11.x is end-of-life, but I have pulled this into the 4.9.x tree,
> thanks.

Oops, no, I dropped it as it caused the build to break, which is why I
didn't include it in the first place :(

I'm guessing you didn't test this?  Or if you did, can you please
provide a working backport?

thanks,

greg k-h

RE: IB/addr: Fix setting source address in addr6_resolve()

[Raju Rangoju]

Hi Greg,

Yes, I did test this patch, it needs an additional (minimal) change for the build to work. But, I didn't realize that I have to provide a backported one.

Anyway, I'll be sending a backported patch shortly.

Thanks,
Raju
-----Original Message-----
From: Greg KH [mailto:greg@kroah.com] 
Sent: 20 September 2017 14:42
To: Raju Rangoju <rajur@chelsio.com>
Cc: stable@vger.kernel.org; Doug Ledford <dledford@redhat.com>; pabeni@redhat.com; roland@purestorage.com; SWise OGC <swise@opengridcomputing.com>; Potnuri Bharat Teja <bharat@chelsio.com>
Subject: Re: IB/addr: Fix setting source address in addr6_resolve()

On Wed, Sep 20, 2017 at 11:11:28AM +0200, Greg KH wrote:
> On Wed, Sep 13, 2017 at 10:55:25AM +0000, Raju  Rangoju wrote:
> > Hi,
> > 
> > The commit eea40b8f624f25 has introduced a bug in rdma/core which broke iw_cxgb4/ipv6. 
> > 
> > So, please pull the below commit to 4.9.x, 4.11.x stable release as it fixes the issue.
> 
> 4.11.x is end-of-life, but I have pulled this into the 4.9.x tree, 
> thanks.

Oops, no, I dropped it as it caused the build to break, which is why I didn't include it in the first place :(

I'm guessing you didn't test this?  Or if you did, can you please provide a working backport?

thanks,

greg k-h