Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-18 Thread Vishwanathapura, Niranjana
On Thu, Dec 15, 2016 at 09:24:20PM -0700, Jason Gunthorpe wrote: >>+struct __hfi_vesw_info { >>+ u16 fabric_id; >>+ u16 vesw_id; >>+ >>+ u8 rsvd0[6]; >>+ u16 def_port_mask; >>+ >>+ u8 rsvd1[2]; >>+ u16 pkey; >>+ >>+ u8 rsvd2[4]; >>+ u32 u

Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Jason Gunthorpe
On Thu, Dec 15, 2016 at 06:59:47PM -0800, Vishwanathapura, Niranjana wrote: > We have made the hfi_vnic driver dependent on CONFIG_X86_64. Er, don't do that either? > >>+struct __hfi_vesw_info { > >>+ u16 fabric_id; > >>+ u16 vesw_id; > >>+ > >>+ u8 rsvd0[6]; > >>+ u16 def_port_mask;

Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Vishwanathapura, Niranjana
On Thu, Dec 15, 2016 at 10:01:09AM -0700, Jason Gunthorpe wrote: On Wed, Dec 14, 2016 at 11:59:35PM -0800, Vishwanathapura, Niranjana wrote: +/** + * union hfi_vnic_bypass_hdr - VNIC bypass header + * @slid: source lid + * @length: length of packet + * @becn: backward explicit congestion notific

Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Christoph Hellwig
On Thu, Dec 15, 2016 at 10:24:37AM -0700, Jason Gunthorpe wrote: > Do not use bitfields without providing a BE version of the bitfield. Do not use bitfields ever for protocol defintions, period.

Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Jason Gunthorpe
On Thu, Dec 15, 2016 at 05:21:05PM +, Hefty, Sean wrote: > > This goes on the network too? Also looks like it has endian problems. > > I don't think OPA supports BE systems, and I think it uses LE on the > wire for at least some portions of its protocol. This is a linux driver for a PCI devic

RE: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Hefty, Sean
> This goes on the network too? Also looks like it has endian problems. I don't think OPA supports BE systems, and I think it uses LE on the wire for at least some portions of its protocol.

Re: [RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Jason Gunthorpe
On Wed, Dec 14, 2016 at 11:59:35PM -0800, Vishwanathapura, Niranjana wrote: > +/** > + * union hfi_vnic_bypass_hdr - VNIC bypass header > + * @slid: source lid > + * @length: length of packet > + * @becn: backward explicit congestion notification > + * @dlid: destination lid > + * @sc: service clas

[RFC v2 03/10] IB/hfi-vnic: Virtual Network Interface Controller (VNIC) netdev

2016-12-15 Thread Vishwanathapura, Niranjana
HFI VNIC netdev function supports Ethernet functionality over Omni-Path fabric by encapsulating Ethernet packets inside Omni-Path packet header. It interfaces with the network stack to provide standard Ethernet network interfaces. It invokes HFI device's VNIC callback functions for HW access. Revi