From: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
To: Artem Bityutskiy <dedekind1@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>,
Mike Frysinger <vapier@gentoo.org>,
Richard Weinberger <richard.weinberger@gmail.com>,
Michael Opdenacker <michael.opdenacker@free-electrons.com>,
linux-mtd@lists.infradead.org,
Piergiorgio Beruto <piergiorgio.beruto@gmail.com>,
Brian Norris <computersforpeace@gmail.com>,
David Woodhouse <dwmw2@infradead.org>, Willy Tarreau <w@1wt.eu>
Subject: Re: [PATCH v6 2/3] ubi-utils: Add ubiblkvol tool
Date: Tue, 18 Feb 2014 17:20:29 -0300 [thread overview]
Message-ID: <20140218202028.GA13799@localhost> (raw)
In-Reply-To: <1392710080.21319.52.camel@sauron.fi.intel.com>
On Tue, Feb 18, 2014 at 09:54:40AM +0200, Artem Bityutskiy wrote:
> On Sun, 2014-02-16 at 17:04 -0300, Ezequiel Garcia wrote:
> > With the addition of block device access to UBI volumes, we now
> > add a simple userspace tool to access the new ioctls.
>
> I noticed you call this driver "block device access to UBI volumes" or
> even "block device interface of UBI volumes". I am not sure we picked
> the best terminology.
Yes, given we've shaped this more like a new UBI feature, than a standalone
driver, I tried to move away from the "ubiblock driver" wording.
> Could we please discuss this a bit. Here are my
> thoughts.
>
Sure.
> Essentially, what you have implemented is a R/O block driver. This
> driver works on top of UBI volumes. It uses simple "1-to-1" mapping,
> which is basically its media format.
>
So far, so good.
> Someone, in theory, may implement a more sophisticated block driver
> which would be an R/W driver with good random I/O performance. This
> driver would not use 1-to-1 mapping. It would have internal block
> mapping tables, and garbage-collector.
>
Indeed. I've been playing with that idea, but (as we already discussed)
this simpler implementation is already interesting enough.
> There may be multiple drivers like this implemented.
>
Well, I thought the above as future improvements of the current proposal.
But I guess we can't know in advance, how are they going to be like.
> These multiple drivers would have the same user interface - the block
> API. But very different implementation, and different incompatible media
> format.
>
> What would be our terminology then, I wonder. What do yo think:
>
> * If your driver is ubiblock, how would those be called, any idea?
> * How would we distinguish between them when doing 'modprobe' ?o
I'm not sure about any of these.
Are you thinking in having multiple block emulation implementations living
together using this same design; i.e. all being part of the UBI driver?
I'm not sure that would be sane. I'm more inclined towards improvements of
the current code. Is this idea too narrow-minded?
> * How would we specify which one to use when using "ubiblkvol" ?
>
Supposing we have several options, I guess a parameter will do.
> BTW, I am not sure "ubiblkvol" is a good name, may be you have other
> ideas? Absent better ideas, may be calling it 'ubiblock' would be
> better, just like the name of the driver? This would at least use less
> of the "namespace".
>
Yes, could be. "ubiblock" looks a bit cleaner.
--
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
next prev parent reply other threads:[~2014-02-18 20:20 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-02-16 20:03 [PATCH v6 0/3] ubi: Add block interface Ezequiel Garcia
2014-02-16 20:03 ` [PATCH v6 1/3] ubi: Introduce block devices for UBI volumes Ezequiel Garcia
2014-02-17 10:45 ` Artem Bityutskiy
2014-02-17 11:27 ` Ezequiel Garcia
2014-02-17 11:34 ` Willy Tarreau
2014-02-25 15:30 ` Ezequiel Garcia
2014-02-17 15:10 ` Artem Bityutskiy
2014-02-17 15:49 ` Ezequiel Garcia
2014-02-17 15:22 ` Artem Bityutskiy
2014-02-18 20:32 ` Ezequiel Garcia
2014-02-16 20:04 ` [PATCH v6 2/3] ubi-utils: Add ubiblkvol tool Ezequiel Garcia
2014-02-18 7:54 ` Artem Bityutskiy
2014-02-18 20:20 ` Ezequiel Garcia [this message]
2014-02-16 20:04 ` [PATCH v6 3/3] UBI: Add block interface documentation Ezequiel Garcia
2014-02-17 10:19 ` [PATCH v6 0/3] ubi: Add block interface Artem Bityutskiy
2014-02-17 10:48 ` Ezequiel Garcia
2014-02-17 10:53 ` Artem Bityutskiy
2014-02-17 11:11 ` Ezequiel Garcia
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140218202028.GA13799@localhost \
--to=ezequiel.garcia@free-electrons.com \
--cc=computersforpeace@gmail.com \
--cc=dedekind1@gmail.com \
--cc=dwmw2@infradead.org \
--cc=linux-mtd@lists.infradead.org \
--cc=michael.opdenacker@free-electrons.com \
--cc=piergiorgio.beruto@gmail.com \
--cc=richard.weinberger@gmail.com \
--cc=thomas.petazzoni@free-electrons.com \
--cc=vapier@gentoo.org \
--cc=w@1wt.eu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).