---
Bernard Metzler, PhD
Tech. Leader High Performance I/O, Principal Research Staff
IBM Zurich Research Laboratory
Saeumerstrasse 4
CH-8803 Rueschlikon, Switzerland
+41 44 724 8605
-----"Steve Wise" <swise@xxxxxxxxxxxxxxxxxxxxx> wrote: -----
>To: "Jason Gunthorpe" <jgg@xxxxxxxx>, "Bernard Metzler"
><bmt@xxxxxxxxxxxxxx>
>From: "Steve Wise" <swise@xxxxxxxxxxxxxxxxxxxxx>
>Date: 02/22/2019 04:41PM
>Cc: linux-rdma@xxxxxxxxxxxxxxx
>Subject: Re: [PATCH v5 02/13] SIW main include file
>
>On 2/21/2019 5:49 PM, Jason Gunthorpe wrote:
>> On Tue, Feb 19, 2019 at 11:08:52AM +0100, Bernard Metzler wrote:
>>> +struct siw_device {
>>> + struct ib_device base_dev;
>>> + struct net_device *netdev;
>>> + struct siw_dev_cap attrs;
>>> +
>>> + u32 vendor_part_id;
>>> + int numa_node;
>>> +
>>> + /* physical port state (only one port per device) */
>>> + enum ib_port_state state;
>>> +
>>> + spinlock_t lock;
>>> +
>>> + /* object management */
>>> + struct idr qp_idr;
>>> + struct idr cq_idr;
>>> + struct idr pd_idr;
>>> + struct idr mem_idr;
>> Please use xarray for all your idr needs, Matt is trying to get rid
>of
>> IDR. See the recent giant series from him, the conversion is not so
>> hard (though wait for the next -rc1 before doing it)
>
>
>Hey Bernard,
>
>What you could do to make forward progress until rdma/for-next is
>moved
>to 5.1-rc1, is to incorporate all the other comments, less the xarray
>change, and rebase this series onto the current rdma/for-next, which
>now
>has all the core dependent changes merged. I think that will allow
>the
>bot testing to progress as well.
>
yes, sounds like the way to go.
Thanks
Bernard.
