Re: [Qemu-devel] [PATCH 0/3] RDMA error handling

["Dr. David Alan Gilbert" <dgilbert@redhat.com>]

* Michael R. Hines (mrhines@digitalocean.com) wrote:
> Reviewed-by: Michael R. Hines <michael@hinespot.com>
> 
> (By the way, I no longer work for IBM and no longer have direct access to RDMA hardware. If someone is willing to let me login to something that does in the future, I don't mind debugging things. I just don't have any hardware of my own anymore to debug, and the last time I tried to use software RDMA it was an unpleasurable experience.)

Thanks; I did hear a rumour that SoftRoCE was going to go upstream, but
it doesn't seem to have happened yet.

Dave

> 
> /*
>  * Michael R. Hines
>  * Senior Engineer, DigitalOcean.
>  */
> 
> On 09/23/2016 02:14 PM, Dr. David Alan Gilbert (git) wrote:
> > From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> > 
> > lp: https://bugs.launchpad.net/qemu/+bug/1545052
> > 
> > The RDMA code tends to hang if the destination dies
> > in the wrong place;  this series doesn't completely fix
> > that, but in cases where the destination knows there's
> > been an error, it makes sure it tells the source and
> > that cleans up quickly.
> > If the destination just dies, then the source still hangs
> > and I still need to look at better ways to fix that.
> > 
> > Dave
> > 
> > Dr. David Alan Gilbert (3):
> >    migration/rdma: Pass qemu_file errors across link
> >    migration: Make failed migration load set file error
> >    migration/rdma: Don't flag an error when we've been told about one
> > 
> >   migration/rdma.c   |  9 ++++++++-
> >   migration/savevm.c | 19 ++++++++++++-------
> >   2 files changed, 20 insertions(+), 8 deletions(-)
> > 
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK