Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 29 Jun 2023 08:38:36 +0800
From:   Ming Lei <ming.lei@redhat.com>
To:     Damien Le Moal <dlemoal@kernel.org>
Cc:     Andreas Hindborg <nmi@metaspace.dk>,
        Hans Holmberg <Hans.Holmberg@wdc.com>,
        Aravind Ramesh <Aravind.Ramesh@wdc.com>,
        Jens Axboe <axboe@kernel.dk>,
        "open list:BLOCK LAYER" <linux-block@vger.kernel.org>,
        Christoph Hellwig <hch@infradead.org>,
        Matias Bjorling <Matias.Bjorling@wdc.com>,
        Andreas Hindborg <a.hindborg@samsung.com>,
        open list <linux-kernel@vger.kernel.org>, gost.dev@samsung.com,
        Minwoo Im <minwoo.im.dev@gmail.com>
Subject: Re: [PATCH v4 1/4] ublk: change ublk IO command defines to enum
Message-ID: <ZJzSjFbzzNxppH7p@ovpn-8-18.pek2.redhat.com>
References: <20230628190649.11233-1-nmi@metaspace.dk>
 <20230628190649.11233-2-nmi@metaspace.dk>
 <d23bf48c-5bc9-aab6-4ca2-ebbb24a0878e@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <d23bf48c-5bc9-aab6-4ca2-ebbb24a0878e@kernel.org>
Precedence: bulk

On Thu, Jun 29, 2023 at 07:47:47AM +0900, Damien Le Moal wrote:
> On 6/29/23 04:06, Andreas Hindborg wrote:
> > From: Andreas Hindborg <a.hindborg@samsung.com>
> > 
> > This change is in preparation for zoned storage support.
> > 
> > Signed-off-by: Andreas Hindborg <a.hindborg@samsung.com>
> > ---
> >  include/uapi/linux/ublk_cmd.h | 23 +++++++++++++++++------
> >  1 file changed, 17 insertions(+), 6 deletions(-)
> > 
> > diff --git a/include/uapi/linux/ublk_cmd.h b/include/uapi/linux/ublk_cmd.h
> > index 4b8558db90e1..471b3b983045 100644
> > --- a/include/uapi/linux/ublk_cmd.h
> > +++ b/include/uapi/linux/ublk_cmd.h
> > @@ -229,12 +229,23 @@ struct ublksrv_ctrl_dev_info {
> >  	__u64   reserved2;
> >  };
> >  
> > -#define		UBLK_IO_OP_READ		0
> > -#define		UBLK_IO_OP_WRITE		1
> > -#define		UBLK_IO_OP_FLUSH		2
> > -#define		UBLK_IO_OP_DISCARD	3
> > -#define		UBLK_IO_OP_WRITE_SAME	4
> > -#define		UBLK_IO_OP_WRITE_ZEROES	5
> > +enum ublk_op {
> > +	UBLK_IO_OP_READ = 0,
> > +	UBLK_IO_OP_WRITE = 1,
> > +	UBLK_IO_OP_FLUSH = 2,
> > +	UBLK_IO_OP_DISCARD = 3,
> > +	UBLK_IO_OP_WRITE_SAME = 4,
> > +	UBLK_IO_OP_WRITE_ZEROES = 5,
> > +	UBLK_IO_OP_ZONE_OPEN = 10,
> > +	UBLK_IO_OP_ZONE_CLOSE = 11,
> > +	UBLK_IO_OP_ZONE_FINISH = 12,
> > +	UBLK_IO_OP_ZONE_APPEND = 13,
> > +	UBLK_IO_OP_ZONE_RESET = 15,
> > +	__UBLK_IO_OP_DRV_IN_START = 32,
> > +	__UBLK_IO_OP_DRV_IN_END = 96,
> > +	__UBLK_IO_OP_DRV_OUT_START = __UBLK_IO_OP_DRV_IN_END,
> > +	__UBLK_IO_OP_DRV_OUT_END = 160,
> > +};
> 
> This patch does not do what the title says. You are also introducing the zone
> operations, and the very obscure __UBLK_IO_OP_DRV_XXX operations without an
> explanation. Also, why the "__" prefix for these ? I do not see the point...

It should be to reserve space for ublk passthrough OP.

> Given that this is a uapi, a comment to explain the less obvious commands would
> be nice.
> 
> So I think the change to an enum for the existing ops can be done either in
> patch 2 or as a separate patch and the introduction of the zone operations done
> in patch 3 or as a separate patch.

Also it might break userspace by changing to enum from macro for existed
definition, cause userspace may check something by '#ifdef UBLK_IO_OP_*',
so probably it is better to keep these OPs as enum, or at least keep
existed definition as macro.

Thanks,
Ming