| Message ID | 1457455018-17638-4-git-send-email-eli@mellanox.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On Tue, Mar 08, 2016 at 06:36:51PM +0200, Eli Cohen wrote: > The subnet prefix is a part of the port_info MAD returned and should be > available at the ib_port_attr struct. We define it here and provide a > default implementation in case the hardware driver does not provide one. > The subnet prefix is required when creating the address vector to access > the SA in networks where GRH must be used. > > Signed-off-by: Eli Cohen <eli@mellanox.com> > Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com> > Changes from V1: > Removed the hardware level implementation and implement as part of > ib_query_port. Looks reasonable Are you going to still send the mlx4/5 changes to avoid the other callback? Reviewed-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> > +++ b/include/rdma/ib_verbs.h > @@ -509,6 +509,7 @@ struct ib_port_attr { > u8 active_width; > u8 active_speed; > u8 phys_state; > + u64 subnet_prefix; > }; This turned out to be aligned then? Jason -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Mar 08, 2016 at 10:57:52AM -0700, Jason Gunthorpe wrote: > > Looks reasonable > > Are you going to still send the mlx4/5 changes to avoid the other > callback? I don't think it is necessary since the hardware driver will do the same thing to obtain the subnet prefix so I don't think it has any siginficant benefit. > > Reviewed-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> > Thanks > > +++ b/include/rdma/ib_verbs.h > > @@ -509,6 +509,7 @@ struct ib_port_attr { > > u8 active_width; > > u8 active_speed; > > u8 phys_state; > > + u64 subnet_prefix; > > }; > > This turned out to be aligned then? > Actually I missed this comment of yours. I will fix that as well. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 00da80e02154..7ba72799fc78 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -649,10 +649,23 @@ int ib_query_port(struct ib_device *device, u8 port_num, struct ib_port_attr *port_attr) { + union ib_gid gid; + int err; + if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) return -EINVAL; - return device->query_port(device, port_num, port_attr); + memset(port_attr, 0, sizeof(*port_attr)); + err = device->query_port(device, port_num, port_attr); + if (err || port_attr->subnet_prefix) + return err; + + err = ib_query_gid(device, port_num, 0, &gid, NULL); + if (err) + return err; + + port_attr->subnet_prefix = be64_to_cpu(gid.global.subnet_prefix); + return 0; } EXPORT_SYMBOL(ib_query_port); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index d83bc300c2b3..b3646e30e25d 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -509,6 +509,7 @@ struct ib_port_attr { u8 active_width; u8 active_speed; u8 phys_state; + u64 subnet_prefix; }; enum ib_device_modify_flags {