From:   "kanchan" <joshi.k@samsung.com>
To:     "'Jens Axboe'" <axboe@kernel.dk>, <linux-kernel@vger.kernel.org>,
        <linux-block@vger.kernel.org>, <linux-nvme@lists.infradead.org>,
        <linux-fsdevel@vger.kernel.org>, <linux-ext4@vger.kernel.org>
Cc:     <prakash.v@samsung.com>
In-Reply-To: <6850a3ad-805f-4232-1bc7-f3249e1fba96@kernel.dk>
Subject: RE: [PATCH v4 4/7] block: introduce write-hint to stream-id
 conversion
Date:   Mon, 22 Apr 2019 19:06:49 +0530
Message-ID: <009c01d4f910$7db6ec40$7924c4c0$@samsung.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Thread-Index: AQK8/LoE/AuxJQqBdxKlIaVgudw56QGc9em7AVEQUzUCUdAPjaROKAng
Content-Language: en-us
Content-Type: text/plain; charset="utf-8"
CMS-TYPE: 105P
References: <1555523406-2380-1-git-send-email-joshi.k@samsung.com>
        <CGME20190417175358epcas1p41a0a4e349dfe0a70bdcc244161c71604@epcas1p4.samsung.com>
        <1555523406-2380-5-git-send-email-joshi.k@samsung.com>
        <6850a3ad-805f-4232-1bc7-f3249e1fba96@kernel.dk>
Sender: linux-ext4-owner@vger.kernel.org
Precedence: bulk

> Why not just use the bio write hint? We already disallow merging of dissi=
milar write hints, so req->bio->bi_write_hint is known to be identical with=
 the rest of the bio's in that chain.

Yes, that is better. Thanks for suggesting it.

-----Original Message-----
From: Jens Axboe =5Bmailto:axboe=40kernel.dk=5D=20
Sent: Wednesday, April 17, 2019 11:28 PM
To: Kanchan Joshi <joshi.k=40samsung.com>; linux-kernel=40vger.kernel.org; =
linux-block=40vger.kernel.org; linux-nvme=40lists.infradead.org; linux-fsde=
vel=40vger.kernel.org; linux-ext4=40vger.kernel.org
Cc: prakash.v=40samsung.com
Subject: Re: =5BPATCH v4 4/7=5D block: introduce write-hint to stream-id co=
nversion

On 4/17/19 11:50 AM, Kanchan Joshi wrote:
> This patch moves write-hint-to-stream-id conversion in block-layer.
> Earlier this was done by driver (nvme). Current conversion is of the=20
> form =22streamid =3D write-hint - 1=22, for both user and kernel streams.
> Conversion takes stream limit (maintained in request queue) into=20
> account. Write-hints beyond the exposed limit turn to 0.
> A new field 'streamid' has been added in request. While 'write-hint'
> field continues to exist. It keeps original value passed from upper=20
> layer, and used during merging checks.

Why not just use the bio write hint? We already disallow merging of dissimi=
lar write hints, so req->bio->bi_write_hint is known to be identical with t=
he rest of the bio's in that chain.

--
Jens Axboe