> -----Original Message----- > From: linux-rdma-owner@xxxxxxxxxxxxxxx [mailto:linux-rdma- > owner@xxxxxxxxxxxxxxx] On Behalf Of Qingmin Liu > Sent: Monday, May 14, 2018 6:11 PM > To: Doug Ledford <dledford@xxxxxxxxxx> > Cc: Jason Gunthorpe <jgg@xxxxxxxx>; Parav Pandit <parav@xxxxxxxxxxxx>; > Scott Branden <scott.branden@xxxxxxxxxxxx>; Ray Jui > <ray.jui@xxxxxxxxxxxx>; linux-rdma@xxxxxxxxxxxxxxx; Qingmin Liu > <qingmin.liu@xxxxxxxxxxxx> > Subject: [PATCH v1 1/2] infiniband: core: fix network namespace issue with > container -net=none > > With -net=none, docker container will use its own network namespace. > Container should find its own namespace but not init_net. > > Fixes: 1060f8653414 ("IB/{core/cm}: Fix generating a return AH for RoCE") > Signed-off-by: Qingmin Liu <qingmin.liu@xxxxxxxxxxxx> > --- > drivers/infiniband/core/addr.c | 2 +- > drivers/infiniband/core/cma.c | 2 +- > 2 files changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index > 88a7542d8c7b..a3abf9276a2b 100644 > --- a/drivers/infiniband/core/addr.c > +++ b/drivers/infiniband/core/addr.c > @@ -788,7 +788,7 @@ int rdma_addr_find_l2_eth_by_grh(const union ib_gid > *sgid, > > memset(&dev_addr, 0, sizeof(dev_addr)); > dev_addr.bound_dev_if = ndev->ifindex; > - dev_addr.net = &init_net; > + dev_addr.net = dev_net(ndev); > Thanks Qingmin for the patch. However this is not sufficient; when address is resolved net namespace can get deleted leads to the crash. I am waiting for recently posted gid reference count series to get merged to post next 3 series which addresses such issues. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
