[openib-general] ip_ib_mc_map?

[Michael S. Tsirkin]

> Quoting Steve Wise <swise at opengridcomputing.com>:
> Subject: Re: ip_ib_mc_map?
> 
> On Thu, 2007-02-01 at 11:09 +0200, Michael S. Tsirkin wrote:
> > > Quoting Steve WIse <swise at opengridcomputing.com>:
> > > Subject: Re: ip_ib_mc_map?
> > > 
> > > On Wed, 2007-01-31 at 17:53 -0800, Sean Hefty wrote:
> > > > Steve Wise wrote:
> > > > > Perhaps there's no backport for this to rhel4u4?
> > > > 
> > > > I would have thought so, but I really don't know.  The function is called from 
> > > > net/ipv4/arp.c, and not directly by ipoib.  So, I don't know how the backport 
> > > > patches typically handle this.
> > > > 
> > > > - Sean
> > > 
> > > Here's what I see:
> > > 
> > > ip_ib_mc_map() is called directly from cma_join_ib_multicast(), which is
> > > added to the ofed_1_2 cma.c via patch file:
> > > kernel_patches/fixes/sean_multicast_1.patch
> > > 
> > > So when I compiled ofed_1_2 on rhel4u4, the cma wouldn't load because
> > > there is no ip_ib_mc_map() in rhel4u4.  
> > > 
> > > So you need a backport patch for this to work on rhel4u4.  Probably many
> > > of the older kernels.
> > 
> > I think this breakage is U4 specific. Someone at RH went to the trouble to
> > rip all of IB related stuff out of the U4 kernel.
> > 
> > I think just calling ip_tr_mc_map on U4 instead will be enough.
> > 
> 
> I changed cma.c to call ip_tr_mc_map() and I can at least load rdma_cm
> now.  I didn't test any mcast, but the rdma-cm is working over iwarp...

So this could be a macro in kernel_addons, unless someone from
Voltaire is willing to step up with a more elaborate implementation.

-- 
MST