Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 5 Oct 2018 09:32:45 +0200
From:   Christoph Hellwig <hch@lst.de>
To:     Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
Cc:     linux-nvme@lists.infradead.org, linux-kernel@vger.kernel.org,
        Sagi Grimberg <sagi@grimberg.me>, Jens Axboe <axboe@fb.com>,
        Christoph Hellwig <hch@lst.de>,
        Potnuri Bharat Teja <bharat@chelsio.com>,
        Keith Busch <keith.busch@intel.com>,
        Hannes Reinecke <hare@suse.com>,
        "Martin K . Petersen" <martin.petersen@oracle.com>
Subject: Re: [PATCH] nvme: create 'paths' entries for hidden controllers
Message-ID: <20181005073245.GA24224@lst.de>
References: <20180928191720.3461-1-cascardo@canonical.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20180928191720.3461-1-cascardo@canonical.com>
User-Agent: Mutt/1.5.17 (2007-11-01)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Sep 28, 2018 at 04:17:20PM -0300, Thadeu Lima de Souza Cascardo wrote:
> When using initramfs-tools with only the necessary dependencies to mount
> the root filesystem, it will fail to include nvme drivers for a root on a
> multipath nvme. That happens because the slaves relationship is not
> present.
> 
> As discussed in [1], using slaves will break lsblk, because the slaves are
> hidden from userspace, that is, they have no real block device, just an
> entry under sysfs.
> 
> Introducing the paths subdir and using that on initramfs-tools makes it
> possible to now boot a system with nvme multipath as root.

Do we need documentation how these paths links are supposed to work?
Who is going to parse them?