linux-mtd.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Artem Bityutskiy <dedekind1@gmail.com>
To: Tarosa <tarosa.y@gmail.com>
Cc: linux-mtd@lists.infradead.org
Subject: Re: Please teach UBI
Date: Tue, 14 Dec 2010 18:44:36 +0200	[thread overview]
Message-ID: <1292345076.2538.95.camel@localhost> (raw)
In-Reply-To: <4D0264BF.6060904@gmail.com>

On Sat, 2010-12-11 at 02:34 +0900, Tarosa wrote:
> Hello.
> 
> Because I do not usually use English, unknown sentences might be written.
> 
> It is a basic question.
> UBI is an array of the initial of "Unsorted block images". Please teach the
> meaning of this "Unsorted block images".

Well, you can consider physical erase blocks to be sorted. Then when you
put UBI on top, you have logical eraseblocks (LEBs). Each LEB contains
an "image" (data), and it is mapped to a PEB (block), and and UBI makes
these images to be unsorted in physical level.

So
1. block = physical eraseblock
2. block image - the contents of a physical eraseblocks.

Without UBI, "block images" are sorted.

With UBI, they become "unsorted", in a sense that UBI adds logical layer
with mapping, so on the physical level the "block images" become
unsorted.

Hence the name. It also means "where" in Latin. IBM guys invented it,
not me, so my understanding can be wrong.

> Was there a filesystem of "Sorted block images" in the past?

In physical eraseblocks.

> What after the filesystem is different as for not sorting block images?

UBI is not file-system at all. It is a layer between MTD and
file-system. UBIFS is a file-system which relies on UBI. It name comes
from UBI + FS = UBIFS.

> Where of UBI is unsorted block images?
> I do not understand. Please tell me.
> 
> There are another and a question.
> It is defined that CONFIG_MTD_UBI_WL_THRESHOLD is 4096. It is thought this is to
> mean WL is done as difference between a maximum  and a minimum  of the value of
> erace counter of PEBs are always in the range of 4096*2(=8192).

Not 8192, but 4096.

> Why is it 8192 ?
> In a value that is bigger than 8192, isn't it good? Or, isn't it good in a
> narrower range?

This is just a default. For modern flash you may need lower value, this
is configurable.

-- 
Best Regards,
Artem Bityutskiy (Артём Битюцкий)

      reply	other threads:[~2010-12-14 16:45 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-12-10 17:34 Please teach UBI Tarosa
2010-12-14 16:44 ` Artem Bityutskiy [this message]

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=1292345076.2538.95.camel@localhost \
    --to=dedekind1@gmail.com \
    --cc=linux-mtd@lists.infradead.org \
    --cc=tarosa.y@gmail.com \
    /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).