From mboxrd@z Thu Jan  1 00:00:00 1970
From: Marek Vasut <marex@denx.de>
Date: Fri, 21 Sep 2012 15:53:26 +0200
Subject: [U-Boot] [PATCH 01/11] DM: add block device core
In-Reply-To: <1702411.xH8Xcq3Rms@bloomfield>
References: <1348169867-2917-1-git-send-email-morpheus.ibis@gmail.com>
	<201209211439.14753.marex@denx.de> <1702411.xH8Xcq3Rms@bloomfield>
Message-ID: <201209211553.26375.marex@denx.de>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Dear Pavel Herrmann,

> On Friday 21 of September 2012 14:39:14 Marek Vasut wrote:
> > Dear Pavel Herrmann,
> > 
> > > On Thursday 20 of September 2012 21:58:17 Marek Vasut wrote:
> > > > Dear Pavel Herrmann,
> > > > 
> > > > > This core will register all block devices (disk, cards, partitons)
> > > > > and provide unfied access to them, instead of current method with
> > > > > device + partition offset
> > > > > 
> > > > > Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
> > > > > ---
> > > > > 
> > > > >  Makefile                  |   1 +
> > > > >  drivers/blockdev/Makefile |  42 ++++++++++++++++
> > > > >  include/dm/blockdev.h     | 121
> > > > > 
> > > > > ++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 164
> > > > > insertions(+)
> > > > > 
> > > > >  create mode 100644 drivers/blockdev/Makefile
> > > > >  create mode 100644 include/dm/blockdev.h
> > > > 
> > > > Why not use drivers/block/ ?
> > > 
> > > Because the drivers in drivers/block have a different purpose than
> > > blockdev.
> > 
> > Different, ok. Can you elaborate how is it different?
> 
> blockctrl is equivalent in purpose to drivers/block, just a new approach
> 
> blockctrl = AHCI, PIIX... whichever chip you have between SATA and PCI (or
> generally disk-bus and board-bus)

So this is for sata ? Or will it also by used for SD/USB flash discs?

> blockdev = disk, partition, SD card

Uh, let's say I understand (even if I don't see the correlation between 
partition and SD card)

> - something that does basic checks
> (range, possibility of operation) and submits operations to correct parent
> (blockctrl, MMC controller, whatnot).

Ascii art might help here greatly (how these pieces fall together). I think I do 
understand it though.

> Also this gets rid of all partition-related code in filesystems, because
> the access to a partition and to the whole disk is the same, no need to
> manually compute offsets every time (and you can support discontinuous
> partitions, if you chose to do so)
> 
> Pavel Herrmann

Best regards,
Marek Vasut