Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
IronPort-SDR: ddakNfOhIG6FftB9Q2IiLbSOamzS2aty6ybLDwbAuO5z91AhapoE6KkX63ODIJalQMy6PLp3KP
 V9hzynsmciC3RsFrQjJ9HPhumdDPizscWdlGqduT2ulb1K8jX04FkTcyLKoxbAS0GO+D7E5t3a
 lQM0c6+rrMWLwZd8kJvx0r6byCUaXpknoC8S0Z3G3H2jTBXlS2AUqnVKusXZ70y23il0VyY99P
 mzp3SK1c/h+/kWN+aDkf/ElmaPVMhadRHgVKgL6J4eUVha7nvUycogT5DU5NAMndSfNuJSlxDI
 VD8=
From:   Damien Le Moal <Damien.LeMoal@wdc.com>
To:     Kanchan Joshi <joshi.k@samsung.com>
CC:     "axboe@kernel.dk" <axboe@kernel.dk>,
        "viro@zeniv.linux.org.uk" <viro@zeniv.linux.org.uk>,
        "bcrl@kvack.org" <bcrl@kvack.org>,
        "asml.silence@gmail.com" <asml.silence@gmail.com>,
        "hch@infradead.org" <hch@infradead.org>,
        "linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
        "mb@lightnvm.io" <mb@lightnvm.io>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "linux-aio@kvack.org" <linux-aio@kvack.org>,
        "io-uring@vger.kernel.org" <io-uring@vger.kernel.org>,
        "linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
        "selvakuma.s1@samsung.com" <selvakuma.s1@samsung.com>,
        "nj.shetty@samsung.com" <nj.shetty@samsung.com>,
        "javier.gonz@samsung.com" <javier.gonz@samsung.com>,
        Arnav Dawn <a.dawn@samsung.com>
Subject: Re: [PATCH v2 1/2] fs,block: Introduce RWF_ZONE_APPEND and handling
 in direct IO path
Thread-Topic: [PATCH v2 1/2] fs,block: Introduce RWF_ZONE_APPEND and handling
 in direct IO path
Thread-Index: AQHWSxStAvYl2c5t2kCK+9TjcP8ZXw==
Date:   Tue, 30 Jun 2020 07:56:46 +0000
Message-ID: <CY4PR04MB375162ABFFA5BB660869C57DE76F0@CY4PR04MB3751.namprd04.prod.outlook.com>
References: <1593105349-19270-1-git-send-email-joshi.k@samsung.com>
 <CGME20200625171834epcas5p226a24dfcb84cfa83fe29a2bd17795d85@epcas5p2.samsung.com>
 <1593105349-19270-2-git-send-email-joshi.k@samsung.com>
 <CY4PR04MB37511FB1D3B3491A2CED5470E7930@CY4PR04MB3751.namprd04.prod.outlook.com>
 <20200629183202.GA24003@test-zns>
 <CY4PR04MB3751213DD4B370F58A63368BE76F0@CY4PR04MB3751.namprd04.prod.outlook.com>
 <20200630074005.GA5701@test-zns>
 <CY4PR04MB37517AAE0B475F631C81B404E76F0@CY4PR04MB3751.namprd04.prod.outlook.com>
Accept-Language: en-US
Content-Language: en-US
wdcipoutbound: EOP-TRUE
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-Network-Message-Id: ef6c4885-49fe-4186-13bc-08d81ccb2350
X-MS-Exchange-CrossTenant-originalarrivaltime: 30 Jun 2020 07:56:46.7575
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: b61c8803-16f3-4c35-9b17-6f65f441df86
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: Oel1TyFBPffNjd/EW26jLc03XRmOBdxS7JoFdwWR9NHAj/qsr4NgDDUsfZmYD+aCwvys2pjVjOPJ/+38fqm/xw==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CY4PR0401MB3586
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 2020/06/30 16:53, Damien Le Moal wrote:=0A=
> On 2020/06/30 16:43, Kanchan Joshi wrote:=0A=
>> On Tue, Jun 30, 2020 at 12:37:07AM +0000, Damien Le Moal wrote:=0A=
>>> On 2020/06/30 3:35, Kanchan Joshi wrote:=0A=
>>>> On Fri, Jun 26, 2020 at 02:50:20AM +0000, Damien Le Moal wrote:=0A=
>>>>> On 2020/06/26 2:18, Kanchan Joshi wrote:=0A=
>>>>>> Introduce RWF_ZONE_APPEND flag to represent zone-append. User-space=
=0A=
>>>>>> sends this with write. Add IOCB_ZONE_APPEND which is set in=0A=
>>>>>> kiocb->ki_flags on receiving RWF_ZONE_APPEND.=0A=
>>>>>> Make direct IO submission path use IOCB_ZONE_APPEND to send bio with=
=0A=
>>>>>> append op. Direct IO completion returns zone-relative offset, in sec=
tor=0A=
>>>>>> unit, to upper layer using kiocb->ki_complete interface.=0A=
>>>>>> Report error if zone-append is requested on regular file or on sync=
=0A=
>>>>>> kiocb (i.e. one without ki_complete).=0A=
>>>>>>=0A=
>>>>>> Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>=0A=
>>>>>> Signed-off-by: SelvaKumar S <selvakuma.s1@samsung.com>=0A=
>>>>>> Signed-off-by: Arnav Dawn <a.dawn@samsung.com>=0A=
>>>>>> Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>=0A=
>>>>>> Signed-off-by: Javier Gonzalez <javier.gonz@samsung.com>=0A=
>>>>>> ---=0A=
>>>>>>  fs/block_dev.c          | 28 ++++++++++++++++++++++++----=0A=
>>>>>>  include/linux/fs.h      |  9 +++++++++=0A=
>>>>>>  include/uapi/linux/fs.h |  5 ++++-=0A=
>>>>>>  3 files changed, 37 insertions(+), 5 deletions(-)=0A=
>>>>>>=0A=
>>>>>> diff --git a/fs/block_dev.c b/fs/block_dev.c=0A=
>>>>>> index 47860e5..5180268 100644=0A=
>>>>>> --- a/fs/block_dev.c=0A=
>>>>>> +++ b/fs/block_dev.c=0A=
>>>>>> @@ -185,6 +185,10 @@ static unsigned int dio_bio_write_op(struct kio=
cb *iocb)=0A=
>>>>>>  	/* avoid the need for a I/O completion work item */=0A=
>>>>>>  	if (iocb->ki_flags & IOCB_DSYNC)=0A=
>>>>>>  		op |=3D REQ_FUA;=0A=
>>>>>> +=0A=
>>>>>> +	if (iocb->ki_flags & IOCB_ZONE_APPEND)=0A=
>>>>>> +		op |=3D REQ_OP_ZONE_APPEND;=0A=
>>>>>=0A=
>>>>> This is wrong. REQ_OP_WRITE is already set in the declaration of "op"=
. How can=0A=
>>>>> this work ?=0A=
>>>> REQ_OP_ZONE_APPEND will override the REQ_WRITE op, while previously se=
t op=0A=
>>>> flags (REQ_FUA etc.) will be retained. But yes, this can be made to lo=
ok=0A=
>>>> cleaner.=0A=
>>>> V3 will include the other changes you pointed out. Thanks for the revi=
ew.=0A=
>>>>=0A=
>>>=0A=
>>> REQ_OP_WRITE and REQ_OP_ZONE_APPEND are different bits, so there is no=
=0A=
>>> "override". A well formed BIO bi_opf is one op+flags. Specifying multip=
le OP=0A=
>>> codes does not make sense.=0A=
>>=0A=
>> one op+flags behavior is retained here. OP is not about bits (op flags a=
re).=0A=
>> Had it been, REQ_OP_WRITE (value 1) can not be differentiated from=0A=
>> REQ_OP_ZONE_APPEND (value 13).=0A=
>> We do not do "bio_op(bio) & REQ_OP_WRITE", rather we look at the=0A=
>> absolute value "bio_op(bio) =3D=3D REQ_OP_WRITE".=0A=
> =0A=
> Sure, the ops are not bits like the flags, but (excluding the flags) doin=
g:=0A=
> =0A=
> op |=3D REQ_OP_ZONE_APPEND;=0A=
> =0A=
> will give you op =3D=3D (REQ_OP_WRITE | REQ_OP_ZONE_APPEND). That's not w=
hat you want...=0A=
=0A=
And yes, REQ_OP_WRITE | REQ_OP_ZONE_APPEND =3D=3D REQ_OP_ZONE_APPEND... But=
 still=0A=
not a reason for not setting the op correctly :)=0A=
=0A=
=0A=
=0A=
-- =0A=
Damien Le Moal=0A=
Western Digital Research=0A=