[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Yzcvwlq9QcssgDjJ@ziepe.ca>
Date: Fri, 30 Sep 2022 15:04:50 -0300
From: Jason Gunthorpe <jgg@...pe.ca>
To: Li Zhijian <lizhijian@...itsu.com>
Cc: Bob Pearson <rpearsonhpe@...il.com>,
Leon Romanovsky <leon@...nel.org>, linux-rdma@...r.kernel.org,
Zhu Yanjun <zyjzyj2000@...il.com>, yangx.jy@...itsu.com,
y-goto@...itsu.com, mbloch@...dia.com, liangwenpeng@...wei.com,
tom@...pey.com, tomasz.gromadzki@...el.com,
dan.j.williams@...el.com, linux-kernel@...r.kernel.org
Subject: Re: [for-next PATCH v5 03/11] RDMA: Extend RDMA kernel verbs ABI to
support flush
On Thu, Sep 29, 2022 at 02:21:24PM +0800, Li Zhijian wrote:
> we can see, IBV_ACCESS_REMOTE_ATOMIC and IBV_ACCESS_ON_DEMAND are
> tagged "if supported" . but currently kernel just returns EINVAL
> when user registers a MR with IB_ACCESS_ON_DEMAND to RXE.
>
> I wonder we should return -EOPNOTSUPP if the device doesn't support requested capabilities
Yes, unsupported combinations of access flags should trigger
EOPNOTSUPP
Jason
Powered by blists - more mailing lists